Research Highlights
Detecting Biological Contaminants in Water, Using Rapid Polymerase Chain Reaction (PCR) TechnologiesIn the past, people in the United States have largely taken for granted the convenience of potable municipal water. However, the threat of intentional contamination of our water supplies is becoming a concern because of a rise in the number of terrorist acts around the world. As a result, there is much interest in technologies that can be used to detect a contamination event as well as dispel or confirm the credibility of a threat. One of these technologies uses PCR to determine the presence of specific bacteria in water in less than four hours. PCR involves enzyme-mediated reactions that replicate the deoxyribonucleic acid (DNA) of the target organisms. This replication process is facilitated through a series of temperature cycles and amplifies the amount of DNA until it reaches detectable levels. In June 2004, EPA tested three rapid PCR technologies:
EPA tested each technology’s ability to detect specific biological contaminants as well as its propensity to register false positive and false negative responses as a result of interfering compounds. Because rapid PCR technologies are expected to serve mainly as screening tools in water monitoring scenarios, this testing produces only qualitative results (i.e., results indicate only the presence or absence of a contaminant, not a concentration level). Evaluating each rapid PCR technology required two steps: a DNA extraction and purification step, followed by an amplification and detection step. Each of the three rapid PCR technologies was evaluated for:
Test DesignTable 1 identifies the rapid PCR technologies that were tested using various water types fortified Table 1. Technologies, Contaminants,
|
Technologies | Contaminants | Interfering Compounds |
---|---|---|
Applied Biosystems TaqMan® E. coli O157:H7 Detection System |
Escherichia coli | Humic Acid Fulvic Acid |
Idaho Technology, Inc. R.A.P.I.D.™ System |
Francisella tularensis Yersinia pestis Bacillus anthracis Brucella suis Escherichia coli |
|
Invitrogen Corporation PathAlert™ Detection Kits |
Francisella tularensis Yersinia pestis Bacillus anthracis |
Three types of water samples were tested in these evaluations: performance test (PT), drinking water (DW), and quality control (QC). PT samples were prepared with deionized (DI) water and fortified with the target contaminant, an interferent, or both the contaminant and interferent. Contaminant-only PT samples were tested in a series of concentrations that included the accepted lethal/infective dose and approximately 2, 5, 10, and 50 times the vendor-stated detection limit.
Both interference PT samples and DW samples were used to determine the technologies’ susceptibility to false positive and false negative results. DW samples were collected from four geographically diverse municipal sources that varied in source (ground water or surface water), treatment (filtered or unfiltered), and disinfection process (chlorination or chloramination). Because real-world application of rapid PCR screening relies on preconcentration of the water sample, approximately 100 L of DW was dechlorinated and then concentrated to 250 mL, using an ultrafiltration sample concentration method. The concentrated DW samples were analyzed after being spiked with each contaminant (one at a time) at a concentration approximately 10 times greater than each PCR’s detection limit, as well as with no contaminant spike.
All PT and DW samples were analyzed in quadruplicate. The results of each replicate sample set were reported as a ratio of the number of positive results to the total number of replicates (e.g., 0/4, 1/4). Method blank QC samples consisted of 10% of all samples. The number of positive and negative control samples supplied by the vendors varied.
Performance and Results
The accuracy of the technology was determined by dividing the number of positive responses by the overall number of analyses of spiked contaminant-only PT samples. The precision of the sample set replicates was determined by calculating the number of consistent responses for all the sample sets. Responses were considered consistent if all four replicates had the same results. Specificity was determined by dividing the number of negative responses by the total number of unspiked samples. A false positive rate was defined as the frequency of false positive results out of the total number of unspiked samples. A false negative rate was defined as the frequency of false negative results out of the total number of spiked PT (contaminant and interferent) samples and spiked DW samples. Table 2 summarizes the results of the evaluation parameters for each technology
Table 2. Summary of Results
Technology | Contaminant | Gene Target | Conc. Range (cfu/mL) | Accuracy | Precision | Specificity | Total False Positives | Total False Negatives |
---|---|---|---|---|---|---|---|---|
TaqMan ® E. coli O157:H7 Detection System |
Escherichia coli | N/A | 500-106 | 100% | 78% | 96% | 1 out of 24 | 0 out of 52 |
R.A.P.I.D.® System |
Francisella tularensis | 1 | 2×103-4×105 | 100% | 95% | 88% | 0 out of 24 | 0 out of 60 |
2 | 100% | 86% | 96% | 0 out of 24 | 4 out of 60 | |||
Yersinia pestis | 1 | 2×103-5×104 | 100% | 100% | 100% | 0 out of 24 | 0 out of 56 | |
2 | 100% | 100% | 100% | 0 out of 24 | 2 out of 56 | |||
Bacillus anthracis | 1 | 2×103-5×104 | 94% | 76% | 100% | 0 out of 24 | 2 out of 56 | |
2 | 94% | 86% | 100% | 0 out of 24 | 2 out of 56 | |||
3 | 100% | 95% | 100% | 0 out of 24 | 0 out of 56 | |||
Brucella suis | N/A | 2×103-5×104 | 88% | 90% | 100% | 0 out of 24 | 0 out of 56 | |
Escherichia coli | N/A | 2×103-5×104 | 100% | 100% | 100% | 0 out of 24 | 0 out of 52 | |
PathAlert™ Detection Kit |
Francisella tularensis | N/A | 2×104-5×105 | 100% | 95% | 96% | 0 out of 24 | 0 out of 60 |
Yersinia pestis | N/A | 2×102-5×103 | 100% | 95% | 96% | 0 out of 24 | 0 out of 56 | |
Bacillus anthracis | N/A | 2×104-5×105 | 100% | 95% | 100% | 0 out of 22 | 0 out of 56 |
N/A not applicable
In general, test concentrations ranged from 102 to 106 colony-forming units per milliliter (cfu/mL). The following is a summary of the testing:
- Minimum accuracy across all of the samples was 88%, with nine instances of 100% accuracy.
- Precision was 76% or above, with three instances of 100% repeatability.
- Specificity was 88% or above, with eight instances of 100% specificity.
- There was one false positive result and six false negative results.
- Four of six false negative results were due to matrix interferences.
- 100% accuracy, 100% precision, 100% specificity, no false positives, and no false negatives were achieved for two contaminants, Yersinia pestis and Escherichia coli, using the R.A.P.I.D.® System
Note that several bacteria were tested using different gene targets offered by Idaho Technology, Inc. (R.A.P.I.D.® System) and were therefore evaluated more than once by that technology. The infective/lethal doses for Yersinia pestis (0.28 cfu/mL), Bacillus anthracis (200 cfu/mL), Brucella suis (40 cfu/mL), and Escherichia coli (0.2 cfu/mL) were below the technologies’ detection limits and were not included in the above results
For more information about the rapid PCR technology’s ease of use, sample throughput, field portability, and cost, visit the NHSRC Web site at www.epa.gov/nhsrc, or view the full report for each technology.
Contact: Eric Koglin