15-1 Introduction to microbial water analysis
Water, the universal solvent, is essential to life. In drier climes, people will even fight over it (look in a newspaper and read about water rights in Colorado and California). Critical to our modern civilization is the availability of a clean water supply for bathing, drinking and cooking. Unfortunately, many pathogens are transmitted through the water supply. Some of these disease-causing pests enter water from the feces of ill individuals and are then ingested and thereby transmitted to others. Diseases such as polio, typhoid, cholera, hepatitis, shigellosis, salmonellosis and others can spread in this manner. To assure a safe water supply, it is critical to monitor for the presence of these pathogens. However, it would be expensive and time consuming to check the water supply for all of them, instead, an indicator organism is used to assay for fecal contamination. Indicator organisms must have four properties to be useful for water analysis.
Coliforms come closest to fulfilling all these criteria and are the standard indicator organisms used to test for the biological pollution of water. Enterobacter and Klebsiella are able to survive and multiply in the environment and are therefore not the best indicators of fecal pollution. The sole habitat of E. coli and K. pneumoniae, termed fecal coliforms, is the intestines of warm blooded animals. Thus, fecal coliforms are good indicators of fecal pollution and can be differentiated from other coliforms by incubating on selective media at 44.5°C.
Using coliforms, the EPA has developed standards for clean, safe water. These standards vary, depending upon the waters intended use. Drinking water and the water in swimming pools must be of the highest purity. There can be no more than one positive sample (>1 coliform/100 ml) in 40 samples tested in a month and the concentration of fecal coliforms must be zero. But wait, we just said that some coliforms are present in the environment. How can these standards be met? In good quality well water most microorganisms are filtered out as the water percolates from the surface to the well. Unusually high numbers of coliforms in well water may indicate run-off from a polluted area. In the case of surface waters (rivers and lakes), filtration through a sand bed and chlorination remove most microbes. There may also be further steps that need to be taken to insure water safety depending on the treatment plant. Swimming pools, being open surface water, are often contaminated by organisms in the air or by swimming, bacteria-infested humans. Chlorine is added to keep numbers low. Note that the number of permissible coliforms is not zero. This would be difficult to achieve and would provide no additional gain in safety.
Natural bathing beaches and treated sewage are assayed for numbers of fecal coliforms. Total coliform counts are not used as a measure due to the near ubiquitous presence of Enterobacter and Klebsiella in the environment. If the count reaches > 400 fecal coliforms per 100 ml or a monthly geometric average of > 200 per 100 ml, it may indicate a problem in the sewage treatment process or that the beach should be closed. The latter often happens in heavily utilized beaches during the summer.
Testing for coliforms
Presently, several tests are in use to assay for coliforms in water, The oldest of these is the multiple tube fermentation test. In this test three steps are performed; the presumptive, confirmed, and completed tests. A moderately selective lactose broth medium (Lactose Lauryl Tryptose Broth), containing a Durham tube, is first used in the presumptive test to encourage the recovery and growth of potentially stressed coliforms in the sample. If harsher selective conditions are used, a deceptively low count may result. A tube containing both growth and gas is recorded as a positive result. It is possible for non-coliforms (Clostridium or Bacillus) to cause false positives in this medium and therefore all positive tubes are then inoculated into a more selective medium (Brilliant Green Lactose Broth or EC Broth) to begin the confirmed test.
The confirmed test medium effectively eliminates all organisms except true coliforms or fecal coliforms, depending upon the medium and incubation conditions. If a positive result is recorded in these tubes the completed test is begun by first streaking a loopful of the highest dilution tube which gave a positive result onto highly selective Eosin Methylene Blue (EMB) agar. After incubation, subsequent colonies are evaluated for typical coliform reactions.
The multiple tube fermentation test has the great disadvantage of taking 3-5 days to complete. If a municipality has a drinking water crisis, this is too long to wait. This has lead to development of faster, less complex tests. In the membrane filter technique 100 ml or greater of a test sample is passed through a filter with pores small enough to retain all bacteria in the sample. The filter is then placed on a selective medium that allows for the detection of coliforms. The advantages of this technique are the shorter time needed to complete the test (1 day vs. 3 to 4 days), its low cost, the higher accuracy in counting, since the colonies can be enumerated directly from the plate, and its simplicity. Disadvantages are that particulate samples (containing silt or other organic matter) quickly clog the filter, metals and phenols can stick to the filter inhibiting growth, and non-coliforms in the test sample may interfere with the formation of coliform colonies on the plate.
Recently, less complex tests for the detection of coliforms have become available. In the presence-absence test (P-A test), a large water sample (100 ml) is mixed with triple strength LLTB in a single culture bottle. Brom cresol purple is added as a pH indicator. If present, coliforms will ferment the lactose to acid and gas, turning the medium from purple to yellow. To detect coliforms and E. coli, the Colilert defined substrate test can be used. A 100 ml sample of water is mixed with a medium containing ortho-nitrophenyl-β,D-galactoside (ONPG) and 4-methyl umbelliferyl-β,D-glucoronide (MUG) as the only nutrients. If coliforms are present ONPG is metabolized, resulting in a yellow color. If E. coli is present, it will degrade MUG to a fluorescent product that can be detected by observation under long wave length UV-light. Both the P-A test and the Colilert defined substrate test are preliminary and any positive results will warrant further analysis of the offending sample.
The most probable number method of enumeration
In this experiment we will detect coliforms in a water sample using the multiple tube fermentation method. Enumeration of coliforms using this method involves inoculating multiple tubes with a 10-fold dilution series of the water sample and uses the most probable number (MPN) technique to estimate the population. To understand this technique, let us imagine preparing a 10-1 to 10-6 dilution series of a culture and inoculating 1 ml portions into tubes containing nutrient broth. After incubation, growth is observed in the tubes inoculated with the 10-1 to 10-5 dilutions, but not in the 10-6 tube. The number of organisms in the original sample is estimated to be less than 106, but greater than 105 bacteria per ml. By inoculating 3 or 5 tubes per dilution and using statistical analysis, a more accurate estimate of the bacterial concentration can be made. This is the basis of the MPN method. Realize that if a 10 ml solution contains 20 organisms, each 1 ml sample will probably not contain exactly 2 organisms. There will be some variation (0, 1, 2, 3 or 4 organisms/ml), but the total of ten 1 ml portions will add up to 20. This is why at lower dilutions, one tube inoculated with a certain dilution blank may show growth (receiving 1 or 2 organisms when inoculated) while other tubes inoculated with the same dilution do not (no organisms in the 1 ml). The number of organisms is assessed by counting the number of positives in the last three dilutions showing growth and then determining the MPN by following the directions on an appropriate table. An excellent explanation of the MPN method has been created by John Lindquist from UW-Madison.[Next]