| |
| J Clin Microbiol. 2008 January; 46(1): 118–129. Published online 2007 November 7. doi: 10.1128/JCM.01685-07. | PMCID: PMC2224289 |
Copyright © 2008, American Society for Microbiology Comparison of Molecular Typing Methods for Characterization of Staphylococcus epidermidis: Proposal for Clone Definition ![[down-pointing small open triangle]](corehtml/pmc/pmcents/x25BF.gif) † M. Miragaia, 1 J. A. Carriço, 2 J. C. Thomas, 3 I. Couto, 4,5 M. C. Enright, 3 and H. de Lencastre 1,6*Laboratory of Molecular Genetics, Instituto de Tecnologia Química e Biológica, Universidade Nova de Lisboa, Rua da Quinta Grande, nr. 6, 2780-156 Oeiras, Portugal,1 Knowledge Discovery in Bioinformatics Group, Instituto de Engenharia de Sistemas e Computadores: Investigação e Desenvolvimento, Rua Alves Redol nr. 9, Apartado 13069, 1000-029 Lisboa, Portugal,2 Department of Infectious Disease Epidemiology, Faculty of Medicine, Imperial College London, Old Medical School Building, St Mary's Hospital, Norfolk Place, London W2 1PG, United Kingdom,3 UEI Micobactérias, Instituto de Higiene e Medicina Tropical, Universidade Nova de Lisboa, Rua da Junqueira, nr. 96, 1349-008 Lisboa, Portugal,4 Centro de Recursos Microbiológicos, Faculdade de Ciências e Tecnologia, Quinta da Torre, 2829-516 Monte de Caparica, Portugal,5 Laboratory of Microbiology, The Rockefeller University, New York, New York 100216 Received August 23, 2007; Revised October 13, 2007; Accepted October 25, 2007. |
Abstract In the present study we give some direction on the selection of the most appropriate typing method(s) to be used for the characterization of Staphylococcus epidermidis, in view of the most recent findings on the evolution, population structure, and epidemiology of this species. In order to achieve this aim, quantitative assessment of the correlation of the results of three typing methods—pulsed-field gel electrophoresis (PFGE), multilocus sequence typing (MLST), and staphylococcal chromosomal cassette mec (SCCmec) typing, which target different regions of the chromosome that evolve at different rates—was performed. In order to evaluate the discriminatory ability and the strength and direction of the correlation of the different typing methods, Simpson's index of diversity (SID), the adjusted Rand coefficient (AR), and the Wallace coefficient (W) were calculated. PFGE was the most discriminatory method (SID = 99%), followed by MLST (SID = 90%) and SCCmec typing (SID = 75%). The values of AR and W (0.10 < AR < 0.30; 0.50 < W < 0.75) indicated that the partition of the same isolate collection by PFGE, MLST, and SCCmec typing provided results that had only a poor correlation with each other. However, the information provided by the combination of PFGE and SCCmec enabled the prediction of the results obtained by MLST at the level of the clonal complex with a high degree of precision (W > 0.90). We propose that clones of S. epidermidis be defined by the combination of the PFGE type followed by the SCCmec type, which provides reliable information on the short-term epidemiology and the ability to predict with consistency long-term clonal evolution. |
In recent years, the development and extensive use of molecular typing methods based on the direct analysis of genomic polymorphisms have greatly improved the understanding of the evolutionary history, population dynamics, and patterns of dissemination of bacteria ( 14, 37, 38). The basic principle of epidemiological typing is that isolates of an infectious agent that belong to the same chain of transmission are clonally related; that is, they descend from a common ancestor. During their evolutionary history, isolates within a given species diversify through point mutations, recombination, or the acquisition/deletion of mobile genetic elements, giving rise to extensive genomic and phenotypic diversity. Since the rates for mutation, recombination, and transfer of mobile elements as well as environmental pressures vary from species to species, the typing method selected and the threshold of marker similarity used to define a clone need to be adjusted to the species under investigation. Furthermore, space and time also need to be considered when the optimal epidemiological markers are selected. Ideally, typing methods should be reproducible and stable; should have a high discriminatory power and epidemiological concordance; should be versatile, easy to perform, easy to interpret; and should be cost-effective and time effective. Staphylococcus epidermidis is one of the most important nosocomial pathogens associated with catheter-related and other indwelling medical device-related infections. Moreover, approximately 70% of the S. epidermidis strains isolated in the hospital environment have acquired resistance to methicillin, and the majority of them are also resistant to almost all classes of antimicrobial agents ( 10, 20, 26). The most widely used method for characterizing S. epidermidis isolates is pulsed-field gel electrophoresis (PFGE), which has mainly been deployed to address short-term epidemiology issues. The molecular characterization of nosocomial S. epidermidis isolates by PFGE has revealed considerable genetic diversity within the population ( 1, 6, 11, 25, 31). Despite the diversity that has been observed, PFGE has been used to track the dissemination of particular S. epidermidis strains between different patients ( 27), wards ( 47), and hospitals ( 35), thus providing a valuable tool for the control of S. epidermidis infections in clinical settings. It was only recently that some key features of the evolutionary history, population structure, and long-term global epidemiology of S. epidermidis were revealed when an improved multilocus sequence typing (MLST) scheme was developed for S. epidermidis ( 41). Despite the distinctive genetic diversity of this species, the molecular characterization of a geographically and temporally diverse collection of S. epidermidis isolates by MLST revealed that the nosocomial S. epidermidis population was composed of nine clonal lineages that were found to be spread worldwide ( 30). Furthermore, in contrast to Staphylococcus aureus, in which the early stages of clonal diversification were found to occur mainly by point mutations ( 8), analysis of S. epidermidis MLST data indicates that in S. epidermidis this occurs primarily by recombination ( 30). Another method that has recently helped to clarify the molecular epidemiology and evolution of methicillin-resistant S. epidermidis (MRSE) strains is the typing of the staphylococcal chromosomal cassette mec (SCC mec), which is the mobile element carrying the determinant of methicillin resistance ( mecA) ( 21). The molecular characterization of this element in diverse collections of S. epidermidis isolates identified five SCC mec types previously identified in S. aureus ( 18, 19, 32), as well as a high number of nontypeable and new SCC mec types ( 15, 16, 28, 30), indicating a high degree of genetic diversity within the SCC mec elements carried by S. epidermidis. Moreover, several findings, such as the high degree of variability observed in the specific region of integration of SCC mec ( orfX) ( 28) and the high degree of diversity found in the type of SCC mec associated with a particular PFGE type or sequence type (ST) ( 28, 30), support the frequent acquisition and/or loss of SCC mec by S. epidermidis. Recent discoveries regarding the population structure of S. epidermidis, namely, that the genetic diversity observed between S. epidermidis strains may derive from the rapid evolution of the chromosome through recombination and the frequent transfer of SCCmec, have a pronounced effect on the interpretation and understanding of the data generated by typing methods. Although a large volume of data has recently been collected through the characterization of S. epidermidis strains by PFGE, MLST, and SCCmec typing, no study in which the three methods were compared has been performed. Moreover, no guidelines exist on the optimal epidemiological markers or the threshold of marker similarity to be used for the definition of an S. epidermidis clone. In the present study we assess quantitatively the strength of the correlation between the results of PFGE, MLST, and SCC mec typing when characterizing S. epidermidis strains through the calculation of measures of concordance: the adjusted Rand coefficient (AR) ( 17, 36) and the Wallace coefficient ( W) ( 45). We anticipate that the finding of any congruence between the type assignments determined by these methods, which target different regions of the chromosome that evolve at different rates, will give an indication of the most appropriate typing method or combination of methods to be used for the definition of a clone in S. epidermidis. |
MATERIALS AND METHODS Bacterial isolates. A collection of 419 nosocomial S. epidermidis isolates collected between 1996 and 2001 from patients with disease and patient carriers attending hospitals in 17 different countries (Argentina, Bulgaria, Cape Verde, China, Colombia, Denmark, Greece, Hungary, Iceland, Italy, Japan, Mexico, Poland, Portugal, Spain, Taiwan, and Uruguay) ( 30) was characterized by PFGE. A total of 216 isolates, including 1 isolate of each PFGE subtype (Iceland and Denmark) or PFGE type (the remaining countries) and control strains ATCC 12228 ( 48) and RP62A ( 12), were selected for characterization by MLST; and at least one isolate of each ST was selected for SCC mec typing ( 30). S. aureus strains COL ( 33), N315 ( 24), HU25 ( 33), and WIS ( 19) were included as controls for SCC mec typing. DNA preparation. Agarose disks containing chromosomal DNA and genomic DNA for PCR were prepared as described previously ( 5, 28, 29). PFGE. SmaI PFGE patterns were determined by conventional protocols ( 5, 29). Visual assignment of the PFGE types and subtypes was performed according to established criteria ( 40). PFGE types were identified by letters, and subtypes were identified by letters followed by a numeric subscript. A database containing all the SmaI PFGE patterns was created by using Bionumerics software (version 4.5) from Applied Maths (Sint-Martens-Latem, Belgium), where band patterns over the multiple gels were normalized and compared. Clustering was performed by using the Dice similarity coefficient and the unweighted pair group method with arithmetic means (UPGMA), with 1.3% of tolerance and 0.8% optimization. PFGE types were automatically assigned (the PFGE type AA method) by using cutoff similarity values that varied from 60 to 100%. The types obtained by the PFGE type AA method are represented by numbers. MLST. MLST was carried out by using the new MLST scheme described by Thomas et al. ( 41). The internal fragments of seven housekeeping genes were amplified by PCR with primers whose sequences match highly conserved regions, and both strands of all amplicons were sequenced with an ABI Prism 3700 DNA sequencer by using BigDye (version 3) fluorescent terminators. Numbers for alleles and STs were assigned according to the S. epidermidis MLST database ( http://sepidermidis.mlst.net/). eBURST algorithm. The most likely patterns of evolutionary descent in our collection were assessed by using the eBURST algorithm ( http://eburst.mlst.net) ( 9) with previously defined settings ( 30). Clonal complexes are represented by the abbreviation CC, and singletons are represented by the abbreviation S. CC2 was subdivided into clusters I and II, as proposed previously ( 30), and cluster II was further separated into subclusters, represented by the cluster number followed by the subgroup founder number (II-5, II-6, II-85, and II-89). Analysis of SCCmec structure. The structures of the ccr and the mec complexes were determined by conventional PCRs, as described previously ( 19, 32). SCC mec types were defined by the combination of the type of ccr complex and the class of mec complex. SCC mec was considered nontypeable when the ccr or the mec complex, or both, were nontypeable, according to previously defined criteria ( 30). Genotypic diversity. Genotypic diversity was calculated by using Simpson's index of diversity (SID) ( 39). Confidence intervals were calculated as described by Grundmann et al. ( 13). Clustering concordance coefficients: AR and W. For comparison of the congruence between the type assignments determined by the different typing methods, AR ( 17, 36) and W ( 45) were calculated and applied as described previously ( 3). AR values close to 0 indicate a lack of congruence between the methods, while values close to 1 indicate a high level of congruence. W values close to 1 indicate that the results obtained by a given typing method enable one to make a good prediction of the clustering determined by the other method. |
RESULTS Molecular characterization by PFGE. When the criteria of Tenover et al. ( 40), which are an arbitrary guide for the clustering of the S. aureus SmaI PFGE band patterns, were applied to the visual comparison of the SmaI PFGE band patterns, the 216 S. epidermidis isolates from this study were grouped into 65 different PFGE types and 193 subtypes. The PFGE type comprising the highest number of subtypes was PFGE type A (20 subtypes), followed by PFGE types C, FF, and Q (14, 13, and 12 subtypes, respectively). The remaining PFGE types identified in this study contained less than 10 subtypes. The most predominant PFGE type was PFGE type A (10%), followed by types C, FF, and Q (6% each) and types D and GG (5% each). The remaining 59 PFGE types were represented by less than 5% of the isolates (Fig. 1 and Table 1).  | FIG. 1.Dendrogram, constructed by the UPGMA method, representing the similarity of the S. epidermidis SmaI PFGE DNA restriction profiles identified by the PFGE type AA method. Representatives of all the 77 PFGE types identified by the PFGE type AA method among (more ...) |
 | TABLE 1. PFGE, MLST, and SCCmec results for strains belonging to CC2, minor CCs, and singletons identified in the strain collection |
Correspondence between visual and automatically assigned classification of PFGE DNA band patterns. The SmaI PFGE DNA band patterns of the 216 S. epidermidis strains were also analyzed by using Bionumerics software (see Materials and Methods). Clusters were defined through the construction of a dendrogram by using the Dice coefficient and the UPGMA clustering method and cutoff similarity values that varied between 60 and 100%. The congruence between the PFGE type assignments obtained by visual comparison (PFGE type visual assignment [the PFGE type VA method]) and the PFGE type AA method for different cutoff similarity values was assessed by calculation of AR and W (data not shown). The highest overall congruence between the PFGE type VA method and the comparison performed with Bionumerics software was obtained by using a cutoff similarity value of 79% (ARcutoff 79% = 0.3576). It was also at the cutoff value of approximately 79% that the highest value of W in both reading directions was obtained, meaning that this is the cutoff value at which both the PFGE type AA and the PFGE type VA method assignments provide the best prediction possible of the assignment previewed by the other method (WPFGE type VA→PFGE type AA = 0.3970; WPFGE type AA→PFGE type VA = 0.3662). By considering a cutoff similarity value of 79% in the UPGMA dendrogram, 77 PFGE types (PFGE types 0001 to 0077) were defined (Fig. 1). The most common PFGE types in the collection analyzed were PFGE types 0016 (16%) and 0010 (7%), followed by PFGE types 0001 (6%) and 0024 (5%). The remaining PFGE types were represented by less than 5% of the population. The majority of the strains assigned to the most-represented PFGE types by automatic classification (PFGE types 0001, 0010, and 0016) were also classified by the visual assignment as belonging to the most common PFGE types (PFGE types A, D, C, and FF) (see the data delimited by the gray square in Table SA1 in the supplemental material). For the subsequent analysis we maintained the visual assignment of SmaI PFGE DNA band patterns. Molecular characterization by MLST. Seventy-four different STs were identified among the 216 S. epidermidis isolates. The most represented ST was ST2, comprising 31% of all isolates (66 of 216 isolates), followed by ST59 (6%) and ST23 (5%). Each of the remaining STs comprised less than 5% of the isolates (Tables 1). The eBURST algorithm separated the 74 STs into one major CC (CC2), 8 minor CCs (CCs 1, 11, 21, 23, 33, 42, 49, and 66), and 13 singletons. We have previously proposed that CC2 be subdivided into two different clusters (clusters I and II), based on the finding that two groups of STs within this CC had contrasting recombination rates, distinct phylogenetic congruence, and different levels of complexity on descendence patterns ( 30). Turner and colleagues ( 42) recently suggested that in species for which a large CC was defined by eBURST analysis due to high recombination rates, like S. epidermidis, radial subgroups within the CC corresponded reasonably well with the true ancestry, whereas the interlinking of those subgroups was probably incorrect. In order to illustrate better the true ancestry, we propose that cluster II of the large CC2 be further subdivided into four subclusters that correspond to the most radial subgroups, each containing as its ancestor a subgroup founder: subgroups II-5, II-6, II-85, and II-89 (Fig. 2).  | FIG. 2.Proposal for subdivision of S. epidermidis CC2 identified by eBURST analysis. Each ST is represented by a black dot. White and gray dots correspond to group and subgroup founders, respectively. Single-locus variants are linked by lines, and the CC or (more ...) |
Molecular characterization by SCCmec typing. A total of 138 MRSE isolates, including at least 1 isolate of each ST, were selected for characterization by SCC mec typing. The molecular characterization of the 138 isolates identified 14 different SCC mec structures, including the 5 well-characterized SCC mec structures (SCC mec types I to V), 3 SCC mec structures with new associations between the mec complex class and the ccr complex type ( mec complex A associated with ccrC, mec complex C associated with ccrAB2, and mec complex B associated with ccrAB3), and 6 different SCC mec structures in which either the mec complex class or the ccr complex type, or both, were nontypeable (Table 1). Nearly half (41%) of the 138 S. epidermidis isolates harbored SCCmec type IV, whereas 27% carried SCCmec type III. On the other hand, a limited number of isolates carried SCCmec type V, I, or II (6%, 4%, and 4%, respectively). In addition, as many as 18% of the isolates analyzed carried nontypeable SCCmec structures (9%) or SCCmec structures with new associations between the ccr complex and the mec complex (9%). Genotypic diversity. The most discriminatory method used was PFGE (SID = 96.43% for the PFGE type and SID = 99.88% for the PFGE subtype), followed by MLST (SID = 89.64%), eBURST analysis (SID = 77.70%), and SCC mec typing (SID = 75.55%) (Table 2). When the information provided by the PFGE type or subtype was complemented with the information provided by SCC mec typing for 138 isolates, no significant increase in the discriminatory power was obtained (Table 2). | TABLE 2. Number of types and genotypic diversity of each typing method |
Qualitative comparison of data obtained by the different typing methods. One major CC (subdivided into five different subclusters), 8 minor CCs, and 13 singletons were identified among the collection of isolates studied. The CC comprising the highest number of STs was CC2 cluster I, with 16 STs, whereas the remaining subclusters and minor CCs had between 2 and 9 STs. The ST containing the highest number of PFGE types, as defined by visual inspection, was ST2, with 12 different PFGE types (types A, CC, D, EE, EEE, FF, G, KK, N, PP, V, and WW). The remaining STs comprised between one and four PFGE types. The great majority of PFGE types identified in our study were associated with a particular ST or CC (60 of 65 PFGE types) (Table 1); however, we observed that some strains with possibly related SmaI PFGE restriction profiles could have different STs or could even belong to a different CC. This was the case for strains of PFGE type C that were assigned to ST36 and ST42 (both of which belong to CC42), ST22 (CC2 cluster I), and also ST60 (S60) and strains with PFGE type F that were assigned to ST1 (CC1), ST5 and ST54 (which belong to CC2 cluster II-5 and cluster I, respectively), and ST21 and ST52 (both of which belong to CC21). In addition, the lack of correspondence between the PFGE type and a particular ST or CC was also observed for strains with PFGE types X, FF, and P (Table 1). With respect to the SCCmec type distribution among the MRSE isolates in this study, we found that strains with the same CC, ST, PFGE type, and even PFGE subtype could harbor different SCCmec types. Strains of ST2 and ST22 had the highest number of different SCCmec types (five different types each), whereas strains belonging to the remaining 72 STs carried between one and three different SCCmec types. SCCmec type IV was carried by strains belonging to as many as seven different CCs (CCs 1, 2, 11, 21, 23, 49, and 66) and three singletons (S65, S72, and S82), including 37 STs and 31 PFGE types. Conversely, SCCmec types I, II, III, and V were present in a limited number of different genetic backgrounds: SCCmec type I was present in strains within CC42 (PFGE type C), SCCmec type II was detected in strains belonging to CC2 cluster I and cluster II-5 (PFGE types FF, H, and V) and CC11 (PFGE type X), SCCmec type III was found to be exclusively associated with strains belonging to CC2 (PFGE types A, C, D, FF, G, KK, N, PP, QQ, and V), and SCCmec type V was carried by strains within CC2 clusters I and II-5 (PFGE types Q and C) and S56 (PFGE type T). Quantitative measure of clustering concordance between the different typing methods. To measure the congruence between the type assignments of the different typing methods, AR and W were calculated for a subset of 138 isolates for which the results of all typing methods were available. The values of AR obtained indicated that among the methods used in this study to type the S. epidermidis strains, the highest overall congruence obtained was between MLST and PFGE (AR ST→PFGE type VA = 0.3119; AR ST→PFGE type AA = 0.5152) (Table 3). The overall congruence between SCC mec typing and MLST was comparably low (AR SCCmec→ST = 0.1879), and was even lower between SCC mec typing and PFGE (AR SCCmec→PFGE type VA = 0.0833; AR SCCmec→PFGE type AA = 0.1350). | TABLE 3. AR for the methods used to characterize the 138 S. epidermidis isolates |
The results obtained for W for the comparison of the different typing methods are shown in Table 4, and a figure representing the correspondence between the different typing methods was constructed on the basis of these values (Fig. 3). The values of concordance between the type assignment done by PFGE and MLST as defined by W are relatively low ( WPFGE type VA/AA→ST < 0.65); however, a high value of W was obtained for the clustering performed by PFGE and eBURST analysis ( WPFGE type VA→CC = 0.7914; W PFGE type AA→CC = 0.8529), indicating that two strains with the same PFGE type have a high probability of belonging to the same CC. When the information provided by PFGE type was complemented with the information from SCC mec typing, an even better correlation with the eBURST analysis clustering was obtained (W PFGE type VA + SCCmec→CC = 0.9212; W PFGE type AA + SCCmec→CC = 0.9379). | TABLE 4. W for the methods used to characterize 138 S. epidermidis isolates |
 | FIG. 3.Representation of the strength of concordance between the typing methods used to type a subset of 138 S. epidermidis isolates, as measured by the use of W. Circles include the type assignments performed by the typing methods used in this study: PFGE type (more ...) |
The values of AR and W between the PFGE type assignment and the clustering performed by MLST and SCC mec typing were generally higher for the automated assignment of PFGE types than for the partition of the collection done by using the visual comparison of the SmaI PFGE band patterns (Tables 3 and 4). |
DISCUSSION Typing methods have been extensively used to distinguish epidemiologically related from epidemiologically unrelated S. epidermidis isolates ( 7, 11, 31, 43, 49) and have recently enabled key features of the epidemiology of S. epidermidis to be revealed ( 28- 30). However, no guidelines for the selection of the optimal epidemiological markers to be used for the definition of an S. epidermidis clone have yet been defined. In an attempt to detect an accurate phylogenetic signal that would give confidence in the interpretation of the data and provide valuable information on the most appropriate method to be used for the definition of a clone, we performed a quantitative assessment of the congruence between three distinct typing methods when they were used to characterize the same S. epidermidis isolate collection. SCCmec acquisition and recombination as sources of genetic diversity. The molecular characterization of S. epidermidis isolates by PFGE, MLST, and SCC mec typing revealed a high degree of genetic diversity (0.75 < SID < 0.99), as was previously described by other authors using different typing methods ( 7, 11, 15, 16, 28, 30, 31, 41, 43, 46, 49). Nevertheless, the level of genetic diversity obtained in this particular study may be biased toward a higher level of diversity in relation to what would usually be found in the clinical setting, since the sample analyzed was initially selected to include a very diverse collection of isolates ( 30). To date, the origins of S. epidermidis genetic diversity have been explored only partially. However, according to the most recent findings on the epidemiology of S. epidermidis ( 28, 30), the high rates of recombination and the frequent acquisition of mobile genetic elements appear to be the most likely driving forces generating genomic polymorphisms in this species. PFGE demonstrated the most diversity (SID = 99.88), reflecting the sensitivity of the method to detecting changes throughout the genome. The extraordinarily fast pace at which S. epidermidis macrorestriction patterns evolve is well illustrated by the finding of chromosomal rearrangements in S. epidermidis strains isolated from single patients during infection episodes ( 11, 49). The frequent insertion/excision of IS 256 ( 49) and SCC mec ( 28, 30) may be contributing to the extensive polymorphism in the number and location of SmaI restriction sequences in the S. epidermidis chromosome ( 28, 49). The genotypic diversity of MLST observed (SID = 89.64%) may be a result of the shuffling of alleles among different S. epidermidis strains, promoted by recombination, which, in conjunction with the existence of a diverse pool of alleles for each locus, may have led to the emergence of a high number of STs ( 30). The origin of the high degree of genetic diversity of SCC mec carried by S. epidermidis observed in this and other studies ( 15, 16, 28, 30) is still unclear. The existence of regions of homology between different SCC elements ( 4, 22) suggests that the genetic diversity in SCC mec may mainly be generated by recombination events. It is possible that the high recombination rates and the high number of SCC mec acquisitions found in S. epidermidis ( 30), together with a high frequency of SCC mec carriage ( 10, 20, 26), may increase the potential for the generation of new SCC mec structures in this species. Despite the genetic diversity in the structure of SCCmec carried by S. epidermidis observed in this study, SCCmec type IV was carried by nearly half of the isolates, comprising as many as seven different S. epidermidis clonal lineages. The predominance and widespread distribution of SCCmec type IV in S. epidermidis suggests that this species may be a potential reservoir of SCCmec type IV for other staphylococcal species. Usefulness of automated assignment of PFGE types. In the present study, we attempted to give some indication of the most appropriate parameters for automated PFGE type assignment of S. epidermidis SmaI restriction band patterns using Bionumerics software. Our results showed that the parameters for which the type obtained by the PFGE type AA method best matched that obtained by the PFGE type VA method were a tolerance of 1.3%, an optimization of 0.8%, and a cutoff value of 79%, which agree with the settings defined for other gram-positive pathogens, such as Streptococcus pneumoniae ( 2) and Streptococcus pyogenes ( 3). Although it can be argued that the values of the concordance coefficients between the types obtained by PFGE type VA and AA obtained by using the parameters mentioned were low (AR = 0.3576; WPFGE typeVA→AA = 0.3970; WPFGE type AA→VA = 0.3662), we should call the attention to the fact that we were analyzing a collection of isolates biased toward a higher degree of genetic diversity. The existence of such diversity in the SmaI PFGE restriction band patterns makes automatic band matching between different strains particularly difficult to obtain. Moreover, the correctness of the automatic band matching performed may have been influenced by differences in the PFGE gel runs that may have generated systematic band shifts, which might have been interpreted as differences. We anticipate that higher values of concordance would be obtained if a more homogeneous collection of isolates was analyzed or if the PFGE gels were rerun under more standardized conditions. The conditions for the automation-based PFGE type definition proposed here are specific to this study collection; however, we predict that the same or similar settings should perform equally well for data generated by the same PFGE protocol, which will generate similarly resolved band patterns. The use of the PFGE type AA method, specifically in a species holding such a high degree of genetic diversity at the level of macrorestriction band patterns, such as S. epidermidis, is particularly welcome, since the PFGE type VA method can become particularly difficult and subject to error, especially when large numbers of PFGE macrorestriction patterns are compared. In this study, the usefulness of the automated system is well illustrated by the higher values of the concordance coefficients obtained between PFGE and the other methods when the PFGE type AA method was performed compared to the values obtained by the PFGE type VA method, which indicates that some errors may have been introduced during the visual inspection of the band patterns. Ability to predict CC from the combination of PFGE and SCCmec typing results. The more significant and interesting results from our study were obtained when AR and W were calculated for the quantitative measurement of the correlation between the different methods or combinations of methods. No good correlation between PFGE type, ST, and SCC mec type was detected (AR and W < 0.75). The lack of correlation between the three typing methods is apparent in the dendrogram constructed from the types obtained by PFGE type AA, where strains with the same SCC mec, ST, or CC are clustered apart (Fig. 1). The high levels of recombination that occur in S. epidermidis species ( 30) and the frequent acquisition of SCC mec by strains harboring the same ST and even the same PFGE type ( 28, 30) may be the explanation for this lack of an association. Although no good concordance between the three typing methods tested was established, an unexpectedly high correlation was found between the type obtained by PFGE type VA or AA and the CC, and the strength of this correlation was augmented when the data generated by PFGE type assignment were complemented with those generated by SCCmec typing. This result suggests that the same strong clonal signal is being detected by both the PFGE type-SCCmec combination and eBURST analysis of the MLST data. The explanation for such a correlation is not obvious; however, we can speculate that despite the frequent insertion/excision of SCC mec from the S. epidermidis chromosome, the rate of SCC mec acquisition by S. epidermidis must be lower than the rate of occurrence of events necessary for the emergence of new PFGE subtypes, allowing the detection of a strong phylogenetic signal when information from PFGE and SCC mec typing are combined. This indicates that, at least transiently, a specific SCC mec type is associated with a strain with a characteristic SmaI PFGE restriction profile. The ability of PFGE to successfully identify genetic lineages was already demonstrated for clonal species like S. aureus ( 34). However, it was surprising to verify that such an ability was also observed in a species like S. epidermidis, which holds such a high degree of genetic diversity in its SmaI DNA restriction band patterns. The fact that the strength of this correlation was amplified when the SCC mec typing results were added may be related to the existence of exclusive restriction-modification systems. The constraint on the exchange of genetic material between strains of different lineages provided by restriction-modification systems may have promoted the existence of specific sets of SCC mec types inside each CC, as suggested previously ( 23, 30, 44). A proposal for clone definition. Taking into consideration these new observations, we propose that clones within the S. epidermidis species be defined by the combination of the PFGE type followed by the SCC mec type, since among all the methods and combinations tested, the results of these methods were the most concordant with those of all other typing methods, providing reliable information on the short-term epidemiology and the ability to predict with consistency the long-term evolutionary history. Besides providing the more meaningful epidemiological information, both methods have high discriminatory powers, are easy to perform in the clinical laboratory, and are cost-effective. The major weakness in our proposal of clone definition probably resides in the difficulty of achieving the interlaboratory reproducibility of PFGE band patterns, which can hinder the interchange of data between different laboratories. For S. aureus, it was demonstrated that the laboratory-to-laboratory reproducibility of PFGE is possible if some minor technical problems are overcome ( 5). In order to increase the interlaboratory reproducibility and standardize the analysis of the S. epidermidis SmaI PFGE band patterns, we recommend that identical PFGE protocols be applied, that prototype strains of the most important S. epidermidis PFGE types be included in each PFGE run, and that automated analysis of the PFGE band patterns with the same or similar settings be performed. For the sake of future comparisons we propose that S. epidermidis clones be identified by the type obtained by the PFGE type AA method, followed by the SCCmec type, and the CC number (with reference to cluster/subclusters in case of CC2) (e.g., 28-V-CCII5). However, the fact that S. epidermidis has a highly dynamic chromosome, with the consequent frequent emergence of new PFGE types and the finding of new SCCmec structures, should be kept in mind. Therefore, the determination of STs for the new PFGE type-SCCmec type combinations should be performed. The work presented here illustrates well how the integration of data from a correctly curated database of PFGE band patterns with data obtained by other typing methods and the knowledge of congruence between them can guide the choice of the most appropriate and effective approach for inferring clonal relationships between S. epidermidis isolates. We believe that the application of this same methodology to other important clinically relevant bacteria will allow a broader view of how the results of different typing methods are related and how they can be used to effectively differentiate clonally related strains from unrelated strains. |
|
Acknowledgments Partial support for this study was provided by project POCTI/SAUESP/57841/2004 from the Fundacão para a Ciência e Tecnologia, Lisbon, Portugal, and project FCG/61052 from the Fundação Calouste Gulbenkian, Lisbon, Portugal, which were both awarded to H. de Lencastre, and a Wellcome Trust project grant awarded to M. C. Enright. Grant SFRH/BPD/26663/2005 from the Fundação para a Ciência e Tecnologia, Portugal, provided support for M. Miragaia. M. C. Enright is a Royal Society University Research Fellow. We thank K. Hiramatsu for providing strain N315 and T. Ito and W. Grubb for providing strain WIS, which were included in this study. |
|
REFERENCES 1. Bogado, I., A. Limansky, E. Sutich, P. Marchiaro, M. Marzi, J. Putero, and A. Viale. 2002. Molecular characterization of methicillin-resistant coagulase-negative staphylococci from a neonatal intensive care unit. Infect. Control Hosp. Epidemiol. 23:447-451. [PubMed]. 2. Carrico, J. A., F. R. Pinto, C. Simas, S. Nunes, N. G. Sousa, N. Frazao, H. de Lencastre, and J. S. Almeida. 2005. Assessment of band-based similarity coefficients for automatic type and subtype classification of microbial isolates analyzed by pulsed-field gel electrophoresis. J. Clin. Microbiol. 43:5483-5490. [PubMed]. 3. Carrico, J. A., C. Silva-Costa, J. Melo-Cristino, F. R. Pinto, H. de Lencastre, J. S. Almeida, and M. Ramirez. 2006. Illustration of a common framework for relating multiple typing methods by application to macrolide-resistant Streptococcus pyogenes. J. Clin. Microbiol. 44:2524-2532. [PubMed]. 4. Chongtrakool, P., T. Ito, X. X. Ma, Y. Kondo, S. Trakulsomboon, C. Tiensasitorn, M. Jamklang, T. Chavalit, J. H. Song, and K. Hiramatsu. 2006. Staphylococcal cassette chromosome mec (SCC mec) typing of methicillin-resistant Staphylococcus aureus strains isolated in 11 Asian countries: a proposal for a new nomenclature for SCC mec elements. Antimicrob. Agents Chemother. 50:1001-1012. [PubMed]. 5. Chung, M., H. de Lencastre, P. Matthews, A. Tomasz, I. Adamsson, M. Aires de Sousa, T. Camou, C. Cocuzza, A. Corso, I. Couto, A. Dominguez, M. Gniadkowski, R. Goering, A. Gomes, K. Kikuchi, A. Marchese, R. Mato, O. Melter, D. Oliveira, R. Palacio, R. Sa-Leao, I. Santos Sanches, J. H. Song, P. T. Tassios, and P. Villari. 2000. Molecular typing of methicillin-resistant Staphylococcus aureus by pulsed-field gel electrophoresis: comparison of results obtained in a multilaboratory effort using identical protocols and MRSA strains. Microb. Drug Resist. 6:189-198. [PubMed]. 6. de Mattos, E. M., L. A. Teixeira, V. M. Alves, C. A. Rezenda e Resende, M. V. da Silva Coimbra, M. C. da Silva-Carvalho, B. T. Ferreira-Carvalho, and A. M. Figueiredo. 2003. Isolation of methicillin-resistant coagulase-negative staphylococci from patients undergoing continuous ambulatory peritoneal dialysis (CAPD) and comparison of different molecular techniques for discriminating isolates of Staphylococcus epidermidis. Diagn. Microbiol. Infect. Dis. 45:13-22. [PubMed]. 7. Dominguez, M. A., J. Linares, A. Pulido, J. L. Perez, and H. de Lencastre. 1996. Molecular tracking of coagulase-negative staphylococcal isolates from catheter-related infections. Microb. Drug Resist. 2:423-429. [PubMed]. 8. Feil, E. J., J. E. Cooper, H. Grundmann, D. A. Robinson, M. C. Enright, T. Berendt, S. J. Peacock, J. M. Smith, M. Murphy, B. G. Spratt, C. E. Moore, and N. P. Day. 2003. How clonal is Staphylococcus aureus? J. Bacteriol. 185:3307-3316. [PubMed]. 9. Feil, E. J., B. C. Li, D. M. Aanensen, W. P. Hanage, and B. G. Spratt. 2004. eBURST: inferring patterns of evolutionary descent among clusters of related bacterial genotypes from multilocus sequence typing data. J. Bacteriol. 186:1518-1530. [PubMed]. 10. Fluit, A. C., F. J. Schmitz, and J. Verhoef. 2001. Multi-resistance to antimicrobial agents for the ten most frequently isolated bacterial pathogens. Int. J. Antimicrob. Agents 18:147-160. [PubMed]. 11. Galdbart, J. O., A. Morvan, N. Desplaces, and N. el Solh. 1999. Phenotypic and genomic variation among Staphylococcus epidermidis strains infecting joint prostheses. J. Clin. Microbiol. 37:1306-1312. [PubMed]. 12. Gill, S. R., D. E. Fouts, G. L. Archer, E. F. Mongodin, R. T. Deboy, J. Ravel, I. T. Paulsen, J. F. Kolonay, L. Brinkac, M. Beanan, R. J. Dodson, S. C. Daugherty, R. Madupu, S. V. Angiuoli, A. S. Durkin, D. H. Haft, J. Vamathevan, H. Khouri, T. Utterback, C. Lee, G. Dimitrov, L. Jiang, H. Qin, J. Weidman, K. Tran, K. Kang, I. R. Hance, K. E. Nelson, and C. M. Fraser. 2005. Insights on evolution of virulence and resistance from the complete genome analysis of an early methicillin-resistant Staphylococcus aureus strain and a biofilm-producing methicillin-resistant Staphylococcus epidermidis strain. J. Bacteriol. 187:2426-2438. [PubMed]. 13. Grundmann, H., S. Hori, and G. Tanner. 2001. Determining confidence intervals when measuring genetic diversity and the discriminatory abilities of typing methods for microorganisms. J. Clin. Microbiol. 39:4190-4192. [PubMed]. 14. Gutacker, M. M., B. Mathema, H. Soini, E. Shashkina, B. N. Kreiswirth, E. A. Graviss, and J. M. Musser. 2006. Single-nucleotide polymorphism-based population genetic analysis of Mycobacterium tuberculosis strains from 4 geographic sites. J. Infect. Dis. 193:121-128. [PubMed]. 15. Hanssen, A. M., G. Kjeldsen, and J. U. Sollid. 2004. Local variants of staphylococcal cassette chromosome mec in sporadic methicillin-resistant Staphylococcus aureus and methicillin-resistant coagulase-negative staphylococci: evidence of horizontal gene transfer? Antimicrob. Agents Chemother. 48:285-296. [PubMed]. 16. Hanssen, A. M., and J. U. Sollid. 2007. Multiple staphylococcal cassette chromosomes and allelic variants of cassette chromosome recombinases in Staphylococcus aureus and coagulase-negative staphylococci from Norway. Antimicrob. Agents Chemother. 51:1671-1677. [PubMed]. 17. Hubert, L., and P. Arabie. 1985. Comparing partitions. J. Classification 2:193-218. 18. Ito, T., Y. Katayama, K. Asada, N. Mori, K. Tsutsumimoto, C. Tiensasitorn, and K. Hiramatsu. 2001. Structural comparison of three types of staphylococcal cassette chromosome mec integrated in the chromosome in methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 45:1323-1336. [PubMed]. 19. Ito, T., X. X. Ma, F. Takeuchi, K. Okuma, H. Yuzawa, and K. Hiramatsu. 2004. Novel type V staphylococcal cassette chromosome mec driven by a novel cassette chromosome recombinase, ccrC. Antimicrob. Agents Chemother. 48:2637-2651. [PubMed]. 20. John, M. A., C. Pletch, and Z. Hussain. 2002. In vitro activity of quinupristin/dalfopristin, linezolid, telithromycin and comparator antimicrobial agents against 13 species of coagulase-negative staphylococci. J. Antimicrob. Chemother. 50:933-938. [PubMed]. 21. Katayama, Y., T. Ito, and K. Hiramatsu. 2000. A new class of genetic element, Staphylococcus cassette chromosome mec, encodes methicillin resistance in Staphylococcus aureus. Antimicrob. Agents Chemother. 44:1549-1555. [PubMed]. 22. Katayama, Y., F. Takeuchi, T. Ito, X. X. Ma, Y. Ui-Mizutani, I. Kobayashi, and K. Hiramatsu. 2003. Identification in methicillin-susceptible Staphylococcus hominis of an active primordial mobile genetic element for the staphylococcal cassette chromosome mec of methicillin-resistant Staphylococcus aureus. J. Bacteriol. 185:2711-2722. [PubMed]. 23. Kuhn, G., P. Francioli, and D. S. Blanc. 2006. Evidence for clonal evolution among highly polymorphic genes in methicillin-resistant Staphylococcus aureus. J. Bacteriol. 188:169-178. [PubMed]. 24. Kuroda, M., T. Ohta, I. Uchiyama, T. Baba, H. Yuzawa, I. Kobayashi, L. Cui, A. Oguchi, K. Aoki, Y. Nagai, J. Lian, T. Ito, M. Kanamori, H. Matsumaru, A. Maruyama, H. Murakami, A. Hosoyama, Y. Mizutani-Ui, N. K. Takahashi, T. Sawano, R. Inoue, C. Kaito, K. Sekimizu, H. Hirakawa, S. Kuhara, S. Goto, J. Yabuzaki, M. Kanehisa, A. Yamashita, K. Oshima, K. Furuya, C. Yoshino, T. Shiba, M. Hattori, N. Ogasawara, H. Hayashi, and K. Hiramatsu. 2001. Whole genome sequencing of meticillin-resistant Staphylococcus aureus. Lancet 357:1225-1240. [PubMed]. 25. Lina, B., F. Vandenesch, J. Etienne, B. Kreiswirth, and J. Fleurette. 1992. Comparison of coagulase-negative staphylococci by pulsed-field electrophoresis. FEMS Microbiol. Lett. 71:133-138. [PubMed]. 26. Melo-Cristino, J., et al. 1998. Antimicrobial resistance in staphylococci and enterococci in 10 Portuguese hospitals in 1996 and 1997. Microb. Drug Resist. 4:319-324. [PubMed]. 27. Milisavljevic, V., F. Wu, J. Cimmotti, J. Haas, P. Della-Latta, E. Larson, and L. Saiman. 2005. Genetic relatedness of Staphylococcus epidermidis from infected infants and staff in the neonatal intensive care unit. Am. J. Infect. Control 33:341-347. [PubMed]. 28. Miragaia, M., I. Couto, and H. de Lencastre. 2005. Genetic diversity among methicillin-resistant Staphylococcus epidermidis (MRSE). Microb. Drug Resist. 11:83-93. [PubMed]. 29. Miragaia, M., I. Couto, S. F. Pereira, K. G. Kristinsson, H. Westh, J. O. Jarlov, J. Carrico, J. Almeida, I. Santos-Sanches, and H. de Lencastre. 2002. Molecular characterization of methicillin-resistant Staphylococcus epidermidis clones: evidence of geographic dissemination. J. Clin. Microbiol. 40:430-438. [PubMed]. 30. Miragaia, M., J. C. Thomas, I. Couto, M. C. Enright, and H. de Lencastre. 2007. Inferring a population structure for Staphylococcus epidermidis from multilocus sequence typing (MLST) data. J. Bacteriol. 189:2540-2552. [PubMed]. 31. Nunes, A. P., L. M. Teixeira, C. C. Bastos, M. G. Silva, R. B. Ferreira, L. S. Fonseca, and K. R. Santos. 2005. Genomic characterization of oxacillin-resistant Staphylococcus epidermidis and S taphylococcus haemolyticus isolated from Brazilian medical centres. J. Hosp. Infect. 59:19-26. [PubMed]. 32. Okuma, K., K. Iwakawa, J. D. Turnidge, W. B. Grubb, J. M. Bell, F. G. O'Brien, G. W. Coombs, J. W. Pearman, F. C. Tenover, M. Kapi, C. Tiensasitorn, T. Ito, and K. Hiramatsu. 2002. Dissemination of new methicillin-resistant Staphylococcus aureus clones in the community. J. Clin. Microbiol. 40:4289-4294. [PubMed]. 33. Oliveira, D. C., A. Tomasz, and H. de Lencastre. 2001. The evolution of pandemic clones of methicillin-resistant Staphylococcus aureus: identification of two ancestral genetic backgrounds and the associated mec elements. Microb. Drug Resist. 7:349-361. [PubMed]. 34. Oliveira, D. C., A. Tomasz, and H. de Lencastre. 2002. Secrets of success of a human pathogen: molecular evolution of pandemic clones of meticillin-resistant Staphylococcus aureus. Lancet Infect. Dis. 2:180-189. [PubMed]. 35. Petinaki, E., I. Spiliopoulou, F. Kontos, M. Maniati, Z. Bersos, N. Stakias, H. Malamou-Lada, C. Koutsia-Carouzou, and A. N. Maniatis. 2004. Clonal dissemination of mupirocin-resistant staphylococci in Greek hospitals. J. Antimicrob. Chemother. 53:105-108. [PubMed]. 36. Rand, W. M. 1971. Objective criteria for the evaluation of clustering methods. J. Am. Stat. Assoc. 66:846-850. 37. Robinson, D. A., and M. C. Enright. 2004. Multilocus sequence typing and the evolution of methicillin-resistant Staphylococcus aureus. Clin. Microbiol. Infect. 10:92-97. [PubMed]. 38. Shopsin, B., M. Gomez, S. O. Montgomery, D. H. Smith, M. Waddington, D. E. Dodge, D. A. Bost, M. Riehman, S. Naidich, and B. N. Kreiswirth. 1999. Evaluation of protein A gene polymorphic region DNA sequencing for typing of Staphylococcus aureus strains. J. Clin. Microbiol. 37:3556-3563. [PubMed]. 39. Simpson, E. H. 1949. Measurement of species diversity. Nature 163:688. 40. Tenover, F. C., R. D. Arbeit, R. V. Goering, P. A. Mickelsen, B. E. Murray, D. H. Persing, and B. Swaminathan. 1995. Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing. J. Clin. Microbiol. 33:2233-2239. [PubMed]. 41. Thomas, J. C., M. R. Vargas, M. Miragaia, S. J. Peacock, G. L. Archer, and M. C. Enright. 2007. An improved multilocus sequence typing scheme for Staphylococcus epidermidis. J. Clin. Microbiol. 45:616-619. [PubMed]. 42. Turner, K. M., W. P. Hanage, C. Fraser, T. R. Connor, and B. G. Spratt. 2007. Assessing the reliability of eBURST using simulated populations with known ancestry. BMC Microbiol. 7:30. [PubMed]. 43. Van Eldere, J., W. E. Peetermans, M. Struelens, A. Deplano, and H. Bobbaers. 2000. Polyclonal staphylococcal endocarditis caused by genetic variability. Clin. Infect. Dis. 31:24-30. [PubMed]. 44. Waldron, D. E., and J. A. Lindsay. 2006. Sau1: a novel lineage-specific type I restriction-modification system that blocks horizontal gene transfer into Staphylococcus aureus and between S. aureus isolates of different lineages. J. Bacteriol. 188:5578-5585. [PubMed]. 45. Wallace, D. L. 1983. A method for comparing two hierarchial clusterings: comment. J. Am. Stat. Assoc. 78:569-576. 46. Wang, X. M., L. Noble, B. N. Kreiswirth, W. Eisner, W. McClements, K. U. Jansen, and A. S. Anderson. 2003. Evaluation of a multilocus sequence typing system for Staphylococcus epidermidis. J. Med. Microbiol. 52:989-998. [PubMed]. 47. Worthington, T., P. A. Lambert, and T. S. Elliott. 2000. Is hospital-acquired intravascular catheter-related sepsis associated with outbreak strains of coagulase-negative staphylococci? J. Hosp. Infect. 46:130-134. [PubMed]. 48. Zhang, Y. Q., S. X. Ren, H. L. Li, Y. X. Wang, G. Fu, J. Yang, Z. Q. Qin, Y. G. Miao, W. Y. Wang, R. S. Chen, Y. Shen, Z. Chen, Z. H. Yuan, G. P. Zhao, D. Qu, A. Danchin, and Y. M. Wen. 2003. Genome-based analysis of virulence genes in a non-biofilm-forming Staphylococcus epidermidis strain (ATCC 12228). Mol. Microbiol. 49:1577-1593. [PubMed]. 49. Ziebuhr, W., K. Dietrich, M. Trautmann, and M. Wilhelm. 2000. Chromosomal rearrangements affecting biofilm production and antibiotic resistance in a Staphylococcus epidermidis strain causing shunt-associated ventriculitis. Int. J. Med. Microbiol. 290:115-120. [PubMed]. |
| |
This article has been 
