Multiparametric comparison of chromogenic-based culture methods used to assess the microbiological quality of drinking water and the mFC method combined with a molecular confirmation procedure

In the Province of Québec, the Programme d'accréditation des laboratoires d'analyse, is administered by the Centre d'expertise en analyse environnementale du Québec (CEAEQ), which certifies private, municipal and institutional laboratories. Since 2001, thermotolerant coliforms detection has been recommended by the CEAEQ for the assessment of the microbiological quality of drinking water. Indeed, in the presence of these microorganisms, drinking water is potentially contaminated with pathogens, and thus inadequate for human consumption.

However, the mFC method is not totally specific and non-fecal origin microorganisms can be detected with this method as false-positive results. Consequently, in 2010 the CEAEQ proposed amending their guidelines and since 2013 has required the measurement of the presence of Escherichia coli rather than thermotolerant coliforms, as recommended in the United States and many European countries (Standard Methods 1998, 2005; Government of Quebec 2013).

To comply with this new guideline, water testing companies must validate a new procedure able to detect the presence of E coli rather than thermotolerant coliforms in drinking water. Many methods are available to detect the presence of E. coli in water with high variability in cost. However, many studies show that methods based on the detection of the β-d-glucuronidase enzyme (the majority) fail to detect 5 to 8% of E. coli strains because they do not express this enzyme (Feng & Hartman 1982; Moberg et al. 1988; Hartman 1989; Martin et al. 1993). The mFC method based on lactose hydrolysis and high temperature incubation detects these E. coli strains. Accordingly, Maheux et al. (2011) suggest that it might be more advantageous to use the mFC method combined with an E. coli-specific real-time polymerase chain reaction (PCR) assay to confirm the identity of the isolated colonies (in only 1 hour) in the context of very low occurrence of water contamination and the higher cost of the β-d-glucuronidase-based methods.

In this study, we compared the sensitivity, ease of use, time to result and affordability of two methods (Colilert^® and USEPA Method 1604 on MI agar; USEPA 2002) both based on β-glucuronidase enzyme expression and approved by the USEPA to detect the presence of E. coli in water, as well as the new combination of methods (mFC + E. coli rtPCR) proposed by Maheux et al. (2011).

During the summers of 2011 and 2012, 968 1-L raw well water samples from individual households were collected in the Québec City region (Canada). For statistical validity, regions presenting a high concentration of E. coli-contaminated well water samples were targeted for sample collection in the present study. Each well water sample was divided in 100-mL subsamples for simultaneous testing using standard microbiological methods mFC agar, USEPA Method 1604 on MI agar and Colilert^® (see sections on membrane filtration and liquid culture methods). The identity of blue colonies recovered from mFC agar (maximum 10 per plate) was further confirmed by E. coli-specific rtPCR (Maheux et al. 2011; see sections on preparation of the DNA extract, rtPCR primers and rtPCR amplification).

The membrane filtration method was performed according to Maheux et al. (2009a). Two 100-mL volumes were filtered on Millipore filters with a standard platform manifold. The first filter was incubated on MI agar (BD, Franklin Lakes, NJ, USA) for 24 h ± 2 h at 35.0 ± 0.5 °C and the second on mFC agar (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) for 24 h ± 2 h at 44.5 ± 0.5 °C before determining colony counts and colour. Each preparation of MI agar and mFC agar plates was tested for performance using pure cultures of target and non-target microorganisms, as recommended by the USEPA microbiology methods manual (USEPA 1978). Tests to confirm the sterility of the filter membranes and buffer used to rinse the filtration apparatus were also performed (Standard Methods 2005).

For the detection of E. coli in water samples with the Colilert^® method (Colilert; IDEXX Laboratories Canada Corp., Toronto, Ontario, Canada), preparation, validation, storage and handling steps were all performed according to the manufacturer's instructions. Briefly, one snap pack containing the Colilert reagent was dissolved in 100 mL of sample water. The solution was then added to a Quanti-tray^®, sealed and incubated at 35.0 ± 0.5 °C for 24 h ± 2 h prior to the identification of E. coli and total coliform positive samples. E. coli presents a yellow coloration and fluorescence under UV light (λ = 365 nm), whereas total coliforms presents a yellow coloration. Positive and negative controls using pure cultures of target and non-target microorganisms were performed as recommended by the manufacturer.

rtPCR amplifications were performed using a bacterial suspension adjusted to a 0.5 McFarland standard (Fisher Scientific Company, Ottawa, Ontario, Canada). The cells were lysed using the BD Diagnostics-GeneOhm Rapid Lysis kit as recommended by the manufacturer (BD Diagnostics-GeneOhm, Québec, Quebec, Canada).

The sequences of the rtPCR primers evaluated in this study are shown in Table 1. Oligonucleotide primers were synthesized by Integrated DNA Technologies (Coralville, IA, USA).

Briefly, 1 μL of the standardized lysed bacterial suspension was transferred directly to a 24-μL mixture of rtPCR containing 50 mM KCl, 10 mM Tris–HCl (pH 9.0), 0.1% Triton X-100, 2.5 mM MgCl2, 0.4 μM of each primer, 0.2 μM probe, 200 μM of each deoxyribonucleoside triphosphate (GE Healthcare Bio-Sciences Inc., Baie d'Urfé, Quebec, Canada), 3.3 μg per μL of bovine serum albumin (BSA; Sigma-Aldrich Canada Ltd, Oakville, Ontario, Canada), 0.06 μ g/μL methoxsalen (Sigma-Aldrich Canada Ltd), 0.5U of Taq DNA polymerase (Promega, Madison, WI, USA), and TaqStart antibody (Clontech Laboratories, Mountain View, CA, USA). Decontamination of the rtPCR mixtures prior to rtPCR was achieved using the UV crosslinker Spectrolinker™ Model XL-1000 (Spectronics Corporation, Westbury, NY, USA; Maheux et al. 2009b). For each experiment, 1 μL of sterile water was added to the rtPCR mixture as a negative control. For tuf gene amplification, the rtPCR mixtures were subjected to thermal cycling (1 min at 95 °C and then 45 cycles of 2 s at 95 °C, 10 s at 58 °C, and 20 s at 72 °C) in a Rotor-Gene thermocycler (QIAGEN Inc., Mississauga, Ontario, Canada).

All individual results were recorded using Microsoft Excel 2010 software (Microsoft Corporation; Redmond, WA, USA) and the statistical analysis was performed using the SAS 9.3 program (SAS Institute Inc. 2011, Cary, NC).

All water samples were recorded as positive (1) or negative (0) for E. coli. McNemar's test was used to compare paired proportions with a 95% confidence interval. When the (two- sided) p-value was less than 0.05, it was concluded that there was a significant difference between both methods. For an overly conservative measure of agreement, Cohen's kappa coefficient was also used to measure the inter-rater agreement. Fleiss (1981) magnitude guidelines were used to characterize the κ values (>0.75 = excellent, 0.40 to 0.75 = fair to good, and <0.40 = poor).

The enzyme-based MI agar and Colilert^® test methods as well as the mFC + E. coli rtPCR assay were first tested with 100-mL subsamples obtained from 968 1-litre well water samples collected in the Québec City region during the summers of 2011 and 2012. The three methods yielded an E. coli positive signal for 111 (11.5%), 105 (10.8%), and 111 (11.5%) of the 968 well water samples tested, respectively (Table 2). In 2011, Maheux et al. showed that the mFC agar method detected 5.5% to 6.8% more E. coli colonies than the Readycult and Colilert enzymatic culture-based methods, respectively. Following these results, they suggested that the combination of mFC + E. coli rtPCR could be more sensitive than the enzymatic culture-based methods. The results obtained in the present study suggest that this better ubiquity (ability to detect all or most E. coli strains) of mFC agar is not transposed into better sensitivity, as the detection of E. coli-contaminated well water samples was statistically equivalent (Table 3).

In terms of ease of use, Colilert was the simplest method to use. Indeed, unit-dosed packaging eliminates media preparation. Furthermore, there is no repeat testing due to clogged filters. Finally, its use does not require well-trained employees. The MI method, on the other hand, provides ease of use comparable to all membrane filtration methods. The medium must be prepared and quality control carried out on each batch. Employee training is also more important than the Colilert method. However, employees already using the membrane filtration equipment can easily use this method. The mFC + E. coli rtPCR assay presents the same advantages and disadvantages as the MI agar membrane filtration method. However, it requires highly trained employees to confirm the identity of a presumptive E. coli colony isolated on mFC by rtPCR. Furthermore, since the mFC agar method must be used in combination with the mEndo agar method to detect both E. coli and total coliforms in water, two membrane filtrations must be carried out for the same water samples, multiplying preparation time by two, quality control, sample treatment and the collection of results.

In terms of time to result, the Colilert and MI methods both required 24 hours prior to obtaining results without the confirmation step, whereas the mFC + E. coli assay needed an additional 1-hour confirmation step when blue colonies were observed on the mFC agar plates. Confirmation is obtained after touching the colony with a micropipette tip or a sterile toothpick, re-suspending it in saline buffer, transferring 1 μL of this suspension to a pre-prepared rtPCR mixture and transferring it to an rtPCR machine for a thermal cycling run of some 50 minutes. In this study, 136 samples from among the 968 well water samples tested resulted in blue colonies on mFC agar; 32 (23.5%) blue colonies isolated from mFC agar were not confirmed as E. coli by E. coli-specific rtPCR assay. Maheux et al. (2011) showed that these false-positive blue colonies are mostly Acinetobacter baumannii, Cronobacter sakazakii, E. albertii, E. fergusonii, and Klebsiella pneumoniae. This result explains why confirmation tests are required for the mFC agar method and not for the enzymatic-based methods. Nevertheless, it should be noted that for statistical validity, regions presenting the highest concentrations of E. coli-contaminated well water samples were targeted for the sample collection of the present study. Globally, the occurrence of E. coli in well water samples ranged mostly from 1% to 5% (Maheux et al. 2012). Compared to classical confirmation procedures that require 24 to 48 hours, a confirmation of E. coli identity by a 1-h rtPCR could represent a good alternative.

In terms of affordability, the MI agar and mFC (+ mEndo) agar methods are comparable. However, the cost of rtPCR confirmation when blue colonies are present makes MI agar more cost-effective since no confirmation is required using this method (Table 4). In all cases, the Colilert method is more expensive (six to 10 times more expensive than MI agar per water sample). However, when no trained employee and/or membrane filtration equipment is available, the Colilert method is more advantageous. For example, in Nunavik, potable water is routinely tested for total coliform and E. coli contamination using the Colilert method because it is more user-friendly and less equipment-intensive than standard membrane filtration-based methods (Edberg et al. 1989; Martin et al. 2007).

Consequently, using the one-hour E. coli-rtPCR assay to confirm the identity of E. coli colonies on mFC agar does not provide greater sensitivity than the Colilert and MI agar culture-based methods. Results showed that they are equivalent in term of sensitivity.

Globally, since environmental laboratories already possess the equipment for membrane filtration methods, the use of the MI agar method seems to be the best option for the assessment of drinking water quality by total coliform and E. coli detection. MI agar is more cost-effective than Colilert and mFC (+ mEndo) + E. coli rtPCR, easier to use and faster than methods requiring confirmation by molecular methods. However, when no trained employee and/or membrane filtration equipment is available, the Colilert method should be preferred.

At the present time, molecular technologies are expensive and not fully automated. However, the day will come when assessing drinking water quality by molecular methods will be more cost-effective and rapid. Furthermore, automation will not require well-trained employees. Then, molecular technologies will compete advantageously with culture-based technologies.

The results obtained in the present study are only applicable to drinking water samples. Results could be different with other types of water.

We thank Steve Charette (IBIS; Institut de Biologie Intégrative et des Systèmes, Université Laval, Québec City, Canada) for providing laboratory space. We also thank Hélène Crépeau (Département de mathématiques et statistique, Université Laval, Québec City (Québec), Canada) for the statistical analysis. This research project was partially supported by an Engage Grant (EG) 413660 from the Natural Sciences and Engineering Research Council of Canada (NSERC) and by the partners of the Research Chair on Drinking Water of Université Laval. The first author was supported by a postdoctoral fellowship from NSERC.