Previous Article | Next Article ![]()
Applied and Environmental Microbiology, January 2009, p. 154-163, Vol. 75, No. 1
0099-2240/09/$08.00+0 doi:10.1128/AEM.01649-08
Copyright © 2009, American Society for Microbiology. All Rights Reserved.
,
,
Hans-Henno Dörries,1,¶,
M. Neus Ayuso-Gabella,2
Miquel Salgot de Marçay,2
Valter Tandoi,3
Caterina Levantesi,3
Costantino Masciopinto,4
Emmanuel Van Houtte,5
Ulrich Szewzyk,1
Thomas Wintgens,6 and
Elisabeth Grohmann1*
Department of Environmental Microbiology, Technical University Berlin, Franklinstr. 29, D-10587 Berlin, Germany,1 Unitat d'Edafologia, Facultat de Farmàcia, Universitat de Barcelona, Av. Joan XXIII, s/n, E-08028 Barcelona, Spain,2 Water Research Institute, CNR, Via Reno 1, 00198 Rome, Italy,3 Water Research Institute, CNR, Bari, Italy,4 Intermunicipal Water Company of the Veurne Region, Doornpannestraat 1, B-8670 Koksijde, Belgium,5 Institut für Verfahrenstechnik, RWTH Aachen, Turmstr. 46, D-52056 Aachen, Germany6
Received 17 July 2008/ Accepted 5 November 2008
|
|
|---|
|
|
|---|
To monitor and guarantee the quality of reclaimed water, the possible track of pathogens and antibiotic resistance genes in the water reclamation and distribution system from the effluent of the WWTP to the end user needs to be investigated; thus, the number of pathogenic bacteria has to be measured at various stages of the process. However, the traditional detection methods for pathogenic bacteria, especially those for intracellular pathogens, show many drawbacks with regard to very special culture requirements and long cultivation time. For routine analysis of alternative water sources and for analysis in case of waterborne disease outbreak, rapid, sensitive, and specific techniques for identification and quantification of the pathogen under suspicion are required.
In a previous study of subsurface flow constructed wetlands, PCR-based methods for rapid detection of pathogenic bacteria, such as Campylobacter spp. and Yersinia enterocolitica, in treated wastewater were developed and applied (2, 3). In the present work, (i) the presence of six different antibiotic resistance genes conferring resistance to antibiotics with different mechanisms of action, (ii) the presence of three intracellular human pathogens, and (iii) fecal-contamination indicators were investigated in three artificial groundwater recharge systems. The work was conducted through the interdisciplinary project Reclaim Water from the Sixth European Union Research Framework Programme, contract number 018309. The resistance genes were selected due to their abundance and the persisting increase of resistant microorganisms in the environment (e.g., see references 7, 12, 15, 16, 22, 37, and 40): ampC (ampicillin resistance), mecA (methicillin resistance), blaSHV-5 (extended β-lactam resistance conferring broad resistance to penicillins and cephalosporins), ermB (erythromycin resistance), tetO (tetracycline resistance), and vanA (vancomycin resistance). Three pathogens were selected for the analysis: Yersinia enterocolitica, Helicobacter pylori, and Mycobacterium avium subsp. paratuberculosis. For all of these pathogens, surviving long-term in aquatic systems and causing waterborne diseases are well documented (6, 8, 23, 29, 30, 33, 43, 45). Additionally, the ability of Y. enterocolitica (39), H. pylori (1, 29, 32), and M. avium (5, 34, 46) to persist for a long time in a dormant or viable but nonculturable state in aquatic environments has been reported. Therefore, culture-independent, rapid, and specific tests are required for the detection of these potential water pollutants. The fate of Y. enterocolitica, H. pylori, and M. avium subsp. paratuberculosis was monitored quantitatively. For this purpose, quantitative molecular assays for the six antibiotic resistance genes and the three pathogens were developed. The technique consists of sample concentration, mechanical cell disruption, and total DNA extraction followed by real-time qPCR. The procedure was adapted to water samples with different concentrations of organic pollutants and was applied in a hygienic survey of three European artificial groundwater recharge systems differing in construction and operation.
|
|
|---|
(ii) Nardò.
Nardò is a town of about 30,000 inhabitants located 50 km south of Lecce, Italy. The average rainfall is 500 to 550 mm per year. The artificial recharge method is based upon injection in a sinkhole, located in a karst area. Water used for recharge is made up of effluents from a municipal treatment plant (Galatone WWTP) and the surface draining water from its surrounding area. The water is collected and transported in an open channel (Asso channel, 140 to 200 liters/s) to the sinkhole. The injection rate is 12,000 to 17,000 m3/day. The reclaimed water is used for irrigation purposes.
(iii) Torreele.
Torreele, the intermunicipal water company of the Veurne region, supplies drinking water to the western part of the coastal region of Flanders, Belgium. The number of consumers grows from 60,000 to 250,000 in the summer season. The average rainfall is about 700 mm per year. The Torreele facility treats tertiary wastewater effluent by ultrafiltration and reverse osmosis. This water is then recharged into the dunes of the St-André water catchment via an infiltration pond and subsequently abstracted using 112 wells sited at least 35 m inland from the pond. Abstracted water is given conventional drinking water treatment, aeration, and rapid sand filtration, followed by UV disinfection prior to distribution. The combination of water reuse and groundwater recharge results in sustainable groundwater management of the St-André water catchment.
Sample processing.
Seven-liter water samples were taken from the WWTP effluents and 40-liter samples from all other sampling points. Six sampling points were probed at Torreele and five at Nardò and Sabadell. All samples were concentrated by filtration through one to three 0.45-µm cellulose acetate filters (14.2 cm in diameter) with a stainless steel in-line filter holder (Sartorius, Göttingen, Germany). The filter holder was washed twice with 70% ethanol before use. In order to prevent early clogging of the filters, highly polluted samples were prefiltered through sterile standard paper filters. Samples were taken in three or four sampling campaigns: in Torreele in January, July, and October 2007; in Nardò in February, May, and September 2007; and in Sabadell in January, March, June, and October 2007. In Torreele, samples were taken from six sampling points: (i) the effluent from the WWTP, (ii) water after ultrafiltration, (iii) the infiltration water before transport, (iv) the infiltration pond, (v) groundwater after infiltration, and (vi) drinking water. In Nardò, samples were taken from five different sampling points: (i) secondary effluent from the Galatone WWTP, (ii) the sinkhole (collecting Galatone WWTP secondary effluent and surface water flows), for delivery to the fractured aquifer (at a 27-m depth), (iii) a well at a 350-m distance from the sinkhole, (iv) a well at a 500-m distance from the sinkhole, and (v) reference groundwater. Groundwater samples were collected by a submerged pump 1 m below the water table. In Sabadell, samples were taken from five different sampling points: (i) secondary effluent from the Ripoll River WWTP, (ii) a Ripoll River reference point (sample of the river water before WWTP effluent discharge), (iii) infiltration water (sample of the river water after WWTP effluent discharge), (iv) recovered groundwater, and (v) disinfected groundwater ready to use (park irrigation and street cleaning). All filters were stored separately at 4°C and were analyzed within 1 week after sampling. The cell material was removed from the filter with a plastic scraper and resuspended in 5 to 20 ml phosphate-buffered saline buffer (8.0 g/liter NaCl, 0.2 g/liter KCl, 1.44 g/liter Na2HPO4, 0.24 g/liter KH2PO4, pH 7.4). The volume of the phosphate-buffered saline buffer was adjusted to the amount of cell material and then concentrated by centrifugation for 5 min at 10,000 x g, yielding a final cell suspension volume of 500 µl to 1,500 µl.
Determination of fecal-contamination indicators.
Heterotrophic plate counts were determined by the APAT-CNR-IRSA, 7050, 2003 method (Nardò); APHA 9215 D (4) (Sabadell); and ISO 6222 (Torreele). Fecal coliforms were measured by APAT-CNR-IRSA, 7020, 2003 (Nardò); EPA accepted 40, CFR part 136 (Sabadell); and AFNOR BRD-07/1-07/93 (Torreele). Escherichia coli cells were quantified by application of most probable number (MPN)/100 ml (Environment Agency 2002: Methods for the Examination of Waters and Associated Materials, part 4 [http://www.environment-agency.gov.uk/]) (Nardò); by EPA accepted 40, CFR part 136 (Sabadell); and by the Colilert-18 technique (Torreele). Enterococci were determined by following the APAT-CNR-IRSA, 7040, 2003 method (Nardò); APHA 9230 C (4) (Sabadell); and ISO 7899/2 (Torreele). Clostridium spores were measured by determination of CFU/100 ml in sulfite-polymyxin-sulfadiazine agar (9a) (Nardò) and by a membrane filtration method derived from ISO 7937 (Sabadell and Torreele).
pH and electrical conductivity measurement.
pH was determined by the APHA 4500-H+ B electrometric method (4) (Nardò and Sabadell) and by following ISO 10523 (Torreele). Electrical conductivity was measured by APHA 2520 B (4) (Nardò), APHA 2510 B (4) (Sabadell), and ISO 7888 (Torreele).
Reference strains, plasmids, and culture techniques.
Yersinia enterocolitica (DSM 11502), Mycobacterium avium subsp. paratuberculosis (DSM 44133), and Helicobacter pylori (DSM 4867) were obtained from the DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen). Mycobacterium avium subsp. paratuberculosis and Helicobacter pylori were obtained as actively growing cultures from the DSMZ. A plasmid containing tetO was obtained from Marilyn Smith (University of Kansas Medical Center). Plasmid pIP501 (DSM 8629) containing ermB, Enterobacter cloacae (DSM 46348) containing ampC, Enterococcus faecium (DSM 17050) containing vanA, Klebsiella pneumoniae (DSM 16609) containing blaSHV-5 (for sulfhydryl variable, extended-spectrum β-lactamase), and Staphylococcus aureus subsp. aureus (DSM 13661) containing mecA were obtained from the DSMZ. Yersinia enterocolitica and the bacteria harboring the resistance genes were cultured overnight in tryptone soy broth at room temperature with shaking.
DNA extraction.
DNA was extracted from bacterial cultures and wastewater samples with a FastDNA spin kit for soil (MP Biomedicals Europe, Heidelberg, Germany) according to the manufacturer's instruction, with minor modifications. The spin filter centrifugation step with 500 µl salt-ethanol wash solution (SEWS-M buffer; MP Biomedicals Europe, Heidelberg, Germany) was performed at 14,000 x g for 2 min. DNA was eluted with 100 µl distilled water. The concentration and integrity of extracted DNA were verified by electrophoresis on 0.7% agarose gels. Depending on the DNA concentration, different serial dilutions of DNA in Tris-EDTA buffer containing 10 µg/ml herring sperm DNA were applied to the real-time qPCR assay.
Standards and internal amplification control for real-time qPCR assays.
Absolute quantifications were performed with the real-time qPCR assays for the detection of ampC, ermB, mecA, blaSHV-5, tetO, and vanA with serial diluted exogenous standards that consisted of purified PCR products and for the detection of Yersinia enterocolitica, Helicobacter pylori, and Mycobacterium avium subsp. paratuberculosis with serial diluted exogenous standards consisting of purified chromosomal DNA. The standards were generated by PCR with DNA from the corresponding reference strains and plasmids. The primers listed in Table 1 amplify PCR products that harbor the complete target sequences of the appropriate real-time qPCR assays. Each 50-µl reaction mixture contained 1.0 µM of each primer, 0.2 mM deoxynucleoside triphosphates, 2.5 U Taq DNA polymerase (GenScript, New Jersey), 1x reaction buffer [10 mM KCl, 10 mM (NH4)2SO4, 20 mM Tris-HCl, 2.0 mM MgSO4, 0.1% Triton X-100, pH 9.0, at 25°C], and template DNA. PCR was carried out with a Primus 96 plus thermocycler (MWG-Biotech, Ebersberg, Germany) with the following three segments: (i) initial denaturation at 95°C for 2 min; (ii) 35 cycles of 95°C for 30 s, 60°C for 45 s, and 72°C for 1 min; and (iii) final elongation at 72°C for 7 min. PCR products were purified with a QIAquick PCR purification kit (Qiagen, Hilden, Germany), analyzed on 2% agarose gels, and quantified with an AlphaImager instrument (Biozym, Hessisch Oldendorf, Germany). Serial dilutions of the standards were made with Tris-EDTA buffer containing 10 µg/ml herring sperm DNA. A VIC-labeled internal activity control included in a TaqMan exogenous internal positive control reagent VIC probe kit (Applied Biosystems, Foster City, CA) was applied to each real-time qPCR assay.
|
View this table: [in a new window] |
TABLE 1. Primers used to produce real-time qPCR standards
|
|
View this table: [in a new window] |
TABLE 2. Primers and probes for real-time qPCR
|
|
|
|---|
|
View this table: [in a new window] |
TABLE 3. Assessment of water quality at different stages of the aquifer recharge process in Nardò, Sabadell, and Torreele
|
Accuracy of the real-time qPCR assays.
Calibration curves were generated using the threshold values of the standard amplification plots. The threshold values of the assays were linear over a measurement range of 6 orders of magnitude. Only real-time qPCR runs without inhibition of the internal activity control were included in the analyses. The efficiencies of the real-time qPCR assays for the pathogens were 98% for M. avium subsp. pseudotuberculosis, 90% for H. pylori, and 85% for Y. enterocolitica. The average values over all real-time qPCR assays were applied for determination of pathogen concentration at the three sites. The efficiencies of the real-time qPCR assays for the antibiotic resistance genes were 96% for tetO, 95% for blaSHV-5, 94% for ermB, 91% for mecA, 85% for vanA, and 78% for ampC. The average values over all real-time qPCR assays were applied for determination of antibiotic resistance gene concentration at the three sites. The detection limit of all real-time qPCR assays was 10 gene copies per reaction. The application of less than 10 gene copies of the resistance genes and the pathogenic bacteria (dilutions down to 1 gene copy per reaction were tested) gave results that were not accurately reproducible (data not shown). Therefore, only runs resulting in
10 gene copies per reaction were applied for the calculation of pathogen and resistance gene concentrations.
Quantitative monitoring of H. pylori, M. avium subsp. pseudotuberculosis, and Y. enterocolitica in three artificial aquifer recharge systems.
In the reclaimed water samples for potable use (Torreele) or irrigation purposes (Nardò and Sabadell), none of the three pathogens were detected. In February 2007 at Nardò sampling point 1, the effluent of the Galatone WWTP showed a signal positive for Y. enterocolitica. Nevertheless, the copy number was below the limit for accurate quantification (10 target genes/reaction) of the Y. enterocolitica 16S rRNA gene. At sampling point 2 (sinkhole), Y. enterocolitica was not detectable anymore. H. pylori and M. avium subsp. pseudotuberculosis were never detected in any of the samples, not even in the WWTP influents from Sabadell, Torreele, or Nardò (1-liter samples each).
Quantitative monitoring of six antibiotic resistance genes in three artificial aquifer recharge systems.
From January 2007 to October 2007, the reclamation sites Sabadell, Nardò, and Torreele were investigated in three (Torreele and Nardò) or four (Sabadell) sampling campaigns for the presence of the antibiotic resistance genes ampC, ermB, mecA, blaSHV-5, tetO, and vanA. The real-time qPCR results for all of the resistance genes are summarized in Tables 4, 5, and 6.
|
View this table: [in a new window] |
TABLE 4. Antibiotic resistance genes detected at Nardò
|
|
View this table: [in a new window] |
TABLE 5. Antibiotic resistance genes detected at Sabadell
|
|
View this table: [in a new window] |
TABLE 6. Antibiotic resistance genes detected at Torreele
|
|
|
|---|
Fecal-contamination analysis as well as monitoring of pH and electrical conductivity showed that Torreele fulfilled the criteria of the Flemish drinking water ordinance, which is in accordance with European Directive 98/83/EC with respect to all tested parameters. In only 1 of the 11 sampling campaigns did we observe 2 CFU enterococci/100 ml reclaimed water for potable use, but this was not confirmed and could be due to sampling errors (E. Van Houtte, personal communication). For the case studies where the aquifer was recharged for irrigation purposes, we obtained the following results. Spanish royal decree 1620/2007 for reclaimed water for park irrigation contains a single microbiological parameter, namely, that Escherichia coli has to be monitored and that levels are not allowed to exceed 200 CFU/100 ml water (41, 42). The E. coli threshold value was never reached in the Sabadell reclaimed water; in all sampling campaigns, it was actually clearly below this maximum permissible value. For the Italian case study site in Nardò, the Italian law of the main quality parameters for agricultural reuse of municipal wastewater (D. Lgs n. 185/03 [summarized in reference 25]) applies. It contains strict thresholds for two microbiological parameters, namely, a maximum of 10 CFU E. coli/100 ml water and no detectable CFU of Salmonella spp./100 ml water. The E. coli threshold was exceeded in four of the seven sampling campaigns in Nardò sampling point 4 (well at a 500-m distance from the sinkhole).
No regulations for the presence of the human-pathogenic bacteria H. pylori, M. avium subsp. pseudotuberculosis, and Y. enterocolitica in drinking water or irrigation water exist. This is based on a general lack of data about particular pathogens, with the exception of Salmonella and Legionella spp., in drinking water, water for irrigation, and recreational water. However, our data demonstrate that there is no risk for contamination with H. pylori, Y. enterocolitica, and M. avium subsp. pseudotuberculosis when the reclaimed water from the three reclamation sites is used, as none of the pathogens were ever detected in the purified water. The presence of Y. enterocolitica, H. pylori, and M. avium subsp. pseudotuberculosis was also investigated in a WWTP in Salisbury (South Africa) and in samples from the world's largest sewage field in Mezquital Valley (Mexico). One-liter water samples were processed from each site. Influent samples from the Mexican site showed the presence of M. avium subsp. pseudotuberculosis at concentrations between 1.6 x 102 and 2.6 x 103 gene copies/100 ml (U. Böckelmann and S. Hergett, unpublished data). Y. enterocolitica was also present in some of the samples from both sites. However, the concentration was below the limit for accurate quantification (10 gene copies/reaction). H. pylori was not detected in any of the samples.
For the antibiotic resistance genes, the situation is the same. No regulations with respect to the presence of resistance genes in drinking water or water for irrigation or recreational purposes exist. Our study aimed to show the potential impact of reclaimed wastewater on the antibiotic resistance gene pool in the environment. To this end, we applied a potent rapid molecular technique, real-time qPCR, to measure the concentrations of antibiotic resistance genes in the reclaimed water. The method proved to be very efficient, reproducible, and sensitive. A minimum of 10 gene copies of the respective antibiotic resistance gene could be detected reproducibly per real-time qPCR reaction.
Only two resistance genes, tetO and ermB, were detected in all of the three reclamation sites. Chen and coworkers and Cauwerts and colleagues found that the levels of abundance of the erm resistance genes often appeared to be positively correlated with those of the tet resistance genes (11, 13). Martins da Costa and coworkers investigated antimicrobial resistance in Enterococcus spp. isolated in inflow, effluent, and sludge from municipal WWTPs in Portugal (26). They found that 34.6% of the isolates were resistant to tetracycline and 24.8% to erythromycin. Wastewater treatment in the Portuguese plants showed an enterococcal decrease up to 4 log. Nevertheless, more than 4.4 x 105 CFU/100 ml were present in the outflow of the plants (26). The reclamation sites investigated in this study were more efficient in the elimination of enterococcal contamination, with a maximum of 1.27 log CFU/100 ml for Sabadell and a maximum of 3.42 log CFU/100 ml for Nardò. We could also detect some correlation between the abundance of enterococci in the reclaimed water and the concentration of tetracycline and erythromycin resistance genes, respectively. In Nardò, in the February and May sampling campaign, tetO resistance genes were found in the reclaimed water at concentrations of 1.27 x 105 and 8.74 x 103, respectively. In these sampling campaigns, we noticed a twofold increase in the enterococcal contamination in comparison with that of the September sampling campaign, where neither tetO nor ermB was detected. The same scenario was observed in Sabadell: the highest tetO concentration in the irrigation water was measured in October 2007 (2.39 x 104 gene copies/100 ml). The October sampling campaign was the only one where enterococcal contamination was detected in the sprinkler water (1.27 log CFU/100 ml water). In Nardò, we additionally noticed a correlation between the concentration of total culturable bacteria at 22°C, the number of E. coli cells, and the abundance of tetO resistance genes. Sample 4 from the Nardò May sampling campaign (8.74 x 103 tetO copies/100 ml) showed the highest contamination by total bacteria and E. coli cells of all of the sampling campaigns (data not shown). In Sabadell, the highest tetO concentration was measured in October 2007. This is in agreement with the occurrence of contamination by coliforms and E. coli in the sample, in comparison with the absence of these indicators in the January and June samples, where no tetO and ermB genes were detected. There was no correlation between the presence of the oxacillin resistance gene, mecA, in the Sabadell samples from the January sampling campaign and elevated levels of total bacteria or enterococci (M. N. Ayuso-Gabella, personal communication).
Our data on the human intracellular pathogens are consistent with the data of others, as in the case of Y. enterocolitica with Alexandrino and coworkers. They could not detect any Y. enterocolitica in wastewater samples or purified water from two different constructed wetlands in Germany by multiplex PCR (3). For H. pylori, the results vary with different countries and different wastewater purification systems: Shannon and coworkers could not detect H. pylori in any stage of municipal wastewater treatment in Canada in a real-time qPCR study (38), whereas a nested-PCR study of water systems for human use in Mexico City demonstrated the absence of H. pylori in water from dams used as water sources but found that 17% of treated wastewater samples contained the pathogen (27). For M. avium, no molecular studies of WWTPs are available.
It is well known that effluent released by WWTP can be a source for antibiotic resistance genes in the environment (16, 20). The best-investigated antibiotic resistance genes with respect to presence in the environment are the tetracycline resistance genes, including tetO. Ferreira da Silva and coworkers (16) showed that tetracycline-resistant enterococci were not eliminated during wastewater treatment consisting of primary and secondary activated sludge processes. The data are consistent with our Sabadell and Nardò results. At both case study sites, secondary WWTP effluent is processed by soil passage and dilution with surface waters. Both artificial aquifer recharge sites resulted in a release of enterococci and in a temporary release of tetO and ermB genes in the environment via the irrigation water. Volkmann et al. performed real-time qPCR analysis for the presence of the resistance genes ampC, vanA, and mecA in municipal wastewater in Germany. vanA was detected in 21% of the wastewater samples and ampC in 78%. mecA was not found in municipal wastewater but was found in two clinical wastewater samples (44). Ferreira da Silva and coworkers also looked for vancomycin-resistant enterococci in raw and treated wastewater, but no resistance among the enterococcal isolates was observed (16). The detection of ampC in wastewater is in agreement with our data from the Sabadell site, where ampC was found in two of the four sampling campaigns. vanA was never detected in our survey, not even in the raw wastewater (data not shown). The presence of mecA in all samples from the Sabadell January 2007 sampling campaign could be explained by the fact that the Sabadell WWTP receives clinical wastewater from the local hospital. For the resistance gene blaSHV-5, no comparable analyses have been conducted.
No clear trend in the extent of contamination by antibiotic resistance genes over the course of the year was observed for the three sites. Consistent for all three sites is that the lowest resistance gene contamination was found in the summer sampling campaign. No resistance genes were detected in Torreele in the July samples, in Sabadell in the irrigation water in June, or in Nardò in the irrigation water from September 2007.
The three investigated artificial aquifer recharge systems demonstrated different capacities of removal of fecal contaminators and antibiotic resistance genes. Our data demonstrated that tetO, ermB, and mecA can occur in groundwater derived from artificial recharge, suggesting that groundwater may be a potential source of antibiotic resistance in the food chain.
The skillful technical assistance of Katharina Knobel is highly acknowledged. We thank Catalina Maya Rendón and Bettina Genthe for providing wastewater samples and Bernd Böckelmann for valuable help with the graphic design.
Published ahead of print on 14 November 2008. ![]()
Supplemental material for this article may be found at http://aem.asm.org/. ![]()
Present address: Berliner Wasserbetriebe, Motardstrasse 35, 13629 Berlin, Germany. ![]()
U.B. and H.-H.D. contributed equally to the work. ![]()
¶ Present address: BIOTECON Diagnostics, D-14473 Potsdam, Germany. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»