Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2015 Sep 24;5:14372. doi: 10.1038/srep14372

Pathogenic Escherichia coli producing Extended-Spectrum β-Lactamases isolated from surface water and wastewater

Eelco Franz 1,a, Christiaan Veenman 1, Angela H A M van Hoek 1, Ana de Roda Husman 1, Hetty Blaak 1
PMCID: PMC4585870  PMID: 26399418

Abstract

To assess public health risks from environmental exposure to Extended-Spectrum β-Lactamases (ESBL)-producing bacteria, it is necessary to have insight in the proportion of relative harmless commensal variants and potentially pathogenic ones (which may directly cause disease). In the current study, 170 ESBL-producing E. coli from Dutch wastewater (n = 82) and surface water (n = 88) were characterized with respect to ESBL-genotype, phylogenetic group, resistance phenotype and virulence markers associated with enteroaggregative E. coli (EAEC), enteroinvasive E. coli (EIEC), enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), extraintesinal E. coli (ExPEC), and Shiga toxin-producing E. coli (STEC). Overall, 17.1% of all ESBL-producing E. coli were suspected pathogenic variants. Suspected ExPECs constituted 8.8% of all ESBL-producing variants and 8.3% were potential gastrointestinal pathogens (4.1% EAEC, 1.8% EPEC, 1.2% EIEC, 1.2% ETEC, no STEC). Suspected pathogens were significantly associated with ESBL-genotype CTX-M-15 (X2 = 14.7, P < 0.001) and phylogenetic group B2 (X2 = 23.5, P < 0.001). Finally, 84% of the pathogenic ESBL-producing E. coli isolates were resistant to three or more different classes of antibiotics. In conclusion, this study demonstrates that the aquatic environment is a potential reservoir of E. coli variants that combine ESBL-genes, a high level of multi-drug resistance and virulence factors, and therewith pose a health risk to humans upon exposure.

Escherichia coli is generally considered a commensal inhabitant of the gastrointestinal tract of humans and animals. Consequently, it is one of the most frequently used indicator bacterium for fecal contamination in environments. However, E. coli is also one of the most important causes of nosocomial-acquired and community-acquired infections in humans and can easily gain resistance to antibiotics consumed by humans and animals1. This includes resistance caused by extended-spectrum β-lactamases (ESBLs). Like other antibiotic resistances, genes for ESBLs are most often encoded on plasmids, which can readily be transferred between bacteria2. In the Netherlands, ESBL-producing E. coli strains are present among the commensal E. coli population in healthy individuals and food-producing animals3,4 and have been isolated from various surface waters5. Surface water is an important source for drinking water production and is used for recreation and irrigation of crops, providing exposure of humans to ESBL-E. coli6. Exposure to ESBL-producing pathogenic E. coli variants may directly result in hard-to-treat infection, also in healthy individuals.

E. coli is characterized by a rather clonal population structure in which isolates form four main phylogenetic groups (A, B1, B2, D)7. Pathogenic E. coli are scattered throughout the phylogenetic tree of the species but phylogenetic groups vary in their host association, presence of virulence factors and persistence in the non-host environment8,9,10. Strains isolated from non-host environments are generally more likely to belong to phylogenetic group B1 and A9,10. Strains of phylogenetic group B1 seem to be more host generalists whereas B2 strains being preferentially host-adapted and are more likely to harbor traits for extraintestinal infection8. Pathogenic E. coli can roughly be separated into two groups: those with primarily gastrointestinal disease and the ones causing extraintestinal infections. The intestinal pathogens can be further distinguished in diffusely adherent E. coli (DAEC), enteroaggregative E. coli (EAEC), enterohemorrhagic E. coli (EHEC; which in turn is a subgroup of Shiga toxin-producing E. coli or STEC), enteroinvasive E. coli (EIEC), enteropathogenic E. coli (EPEC), enterotoxigenic E. coli (ETEC), and extraintestinal E. coli (ExPEC) . ExPEC includes avian pathogenic E. coli (APEC), neonatal meningitis E. coli (NMEC) and uropathogenic E. coli (UPEC)11. The differences in the ability of E. coli strains to cause disease and the diverse syndromes caused by the various pathotypes can be attributed to specific virulence genes and the variability in which these occur among strains11.

Few studies have investigated the pathogenic potential of environmental ESBL-producing bacteria which may be directly linked to public health risk12. The presence of virulence factors, representative for different pathogenic groups of E. coli, among ESBL-E. coli isolated from wastewater and surface water was examined in the current study. In addition, the distribution of phylogenetic groups was determined in order to examine correlations between ESBL-production, virulence factors and genetic background. This sheds light on the transmission ecology and direct public health risks of exposure to waterborne ESBL-E. coli.

Materials and Methods

Isolates

ESBL-producing isolates were selected from an existing collection of isolates obtained from wastewater and surface water sampled between 2010 and 20125,13. Surface water samples (n = 20) were taken at ten different sites situated in four different regions in the Netherlands (including rivers, canals, lakes, North Sea), wastewater samples (n = 20) included influents and effluents from wastewater treatment plants (WWTP), an international airport WWTP, and from wastewater of health care institutions. Isolates were obtained by filtration of multiple volumes of water samples through 0.45 μm filters, followed by incubation of these filters on selective culture media for the isolation of ESB-producing E. coli: Tryptone Bile X-glucuronide medium supplemented with 1 μg/ml cefotaxime (TBX/CTX) or on ChromIDTM ESBL agar (Biomerieux). Incubation conditions were 18–24 hours at 36 ± 2 °C or 4 to 5 hours at 36 ± 2 °C followed by 18 to 19 hours at 44 ± 0.5 °C. Isolation procedures were based on standard isolation procedures for the selective isolation of E. coli from water and food using chromogenic media (NEN-EN-ISO 9308-1 and ISO 16649-2), adapted to enable selective growth of ESBL-producing variants. Variations in isolation procedures was related to isolates being obtained as part of different projects. Suspected ESBL-E. coli isolates (i.e. the ß-glucuronidase-positive colonies on TBX-CTX as well as on ChromIDTM ESBL agar) were further confirmed as E. coli by testing for indole-production using BBL Dry SlideTM (BD), and subsequently tested for ESBL-production by disk diffusion following CLSI guidelines14, using Sensi-DiscsTM (BD, Breda, the Netherlands). Zone diameters were determined for cefotaxime (30 μg) ± clavulanic acid (10 μg), ceftazidime (30 μg) ± clavulanic acid (10 μg). ESBL-producing isolates were defined as strains resistant to cefotaxime (zone diameter ≤22 mm) and/or ceftazidime (zone diameter ≤17 mm), and an increase in zone diameter of ≥5 mm with the disks containing clavulanic acid14.

From this pre-existing collection of confirmed ESBL-producing E. coli isolates, 93 wastewater and 93 surface water isolates, from 20 samples each, were selected more or less randomly, but taking into account established ABR profiles (when available), to minimize the chance of including duplicate isolates from the same sample. After characterization of phylogenetic groups, ESBL-genes, virulence genes and antibiotic resistance profiles was completed, some isolates were retrospectively identified as duplicates and omitted, leaving 88 wastewater (63 from WWTP, 15 from the international airport and 10 from health care institutions) and 82 surface water isolates for analyses.

Molecular characterization

Each isolate was characterized with respect to phylogenetic group, ESBL-genotype, and the presence of virulence factors. For this purpose, material from one single colony was suspended in Tris EDTA buffer (pH 8.0, Sigma-Aldrich, Zwijndrecht, the Netherlands), and DNA extract stored at −20 °C. PCRs were carried out using primers and conditions as described in Supplementary Table S1.

For phylogenetic group analysis, a multiplex PCR targeting chuA, yjaA genes and TspE4.C2 DNA fragment as described by Clermont et al.15 was used. Strains were sub-grouped according to Escobar-Páramo et al.16: subgroup A0, chuA−, yjaA−, TspE4.C2−; subgroup A1, chuA−, yjaA+, TspE4.C2−; group B1, chuA−, yjaA−, TspE4.C2+; subgroup B22, chuA+, yjaA+, TspE4.C2−; subgroup B23, chuA+, yjaA+, TspE4.C2+; subgroup D1, chuA+, yjaA−, TspE4.C2−; subgroup D2, chuA+, yjaA− , TspE4.C2+. For ESBL-genotyping, the presence of blaCTX-M-1-group, blaCTX-M-2-group, and blaCTX-M-9-group, blaOXA, blaSHV and blaTEM, was established by multiplex PCRs using primers described by Dalenne et al.17. Products of the expected size were treated with ExoSAP-IT (GE Healthcare, Hoevelaken, the Netherlands) and sequenced using the same primers used to generate PCR-products, and BigDye Terminator v3.1 Cycle Sequencing kit (Applied Biosystems, Bleiswijk, the Netherlands). Thus obtained partial sequences were compared with ESBL-gene sequences in the GenBank database and on the Lahey website (www.lahey.org/Studies). Partial sequence analysis allows identification of clusters of homologous alleles: CTX-M-1/61, CTX-M-3/22/66, CTX-M-9/51, CTX-M-14/21/83/90/113, CTX-M-15/28, CTX-M-27/98, CTX-M-32, CTX-M-55/57, SHV-12/129, TEM-52/92. For sake of annotation, the lowest allele number belonging to a cluster was assigned to identified genotypes. For analysis of virulence factors, PCRs were used targeting genes encoding the diarrheagenic virulence factors: Shigatoxin 1 and 2 (stx1, stx2) for STEC, intimin (eae) for EPEC, aggregative virulence regulator (aggR) for EAEC, invasion plasmid antigen (ipaH) for EIEC and heat labile enterotoxin (estA) and heat stabile enterotoxine (eltB) for ETEC18. The following markers were selected for identifying extraintestinal pathogenic E. coli (ExPEC): afimbrial adhesion (afa), F1C fimbriae (focG), cytolytic protein toxin (hlyD), iron acquisistion system (iutA), group 2 polysaccharide capsule (kpsMII), P fimbriae (papA) and S fimbriae (sfaS)5,13. ExPEC was defined as strains having three or more of these genes20.

Biofilm formation

Isolates were tested for their capacity to form biofilms using the procedure described by Wakimoto et al.21. Biofilm formation was defined as having an OD of at least four times that of negative control wells (i.e. moderate and strong biofilm formation according to Naves at al. 200822).

Analysis of antibiotic resistance

Isolates were screened for antibiotic susceptibility to a panel of 14 antibiotics of human and veterinary clinical relevance, using Sensi-disksTM according to the manufacturers’ instructions (Becton, Dickinson BV, Breda, the Netherlands). Included antibiotics were one or two representatives from eight classes: beta-lactams: ampicillin (penicillins), cefotaxime and ceftazidime (3rd generation cephalosporins); β-lactam/β-lactamase combination: amoxicillin/clavulanic acid; carbapenems: imipenem and meropenem; tetracyclines: tetracycline; (fluor)quinolones: ciprofloxacin and naladixic acid; aminoglycosides: streptomycin and gentamycin; folate pathway inhibitors: sulfisoxazole and trimethoprim; phenicols: chloramphenicol. Resistance was defined as having an inhibition zone smaller than or equal to the clinical break-points defined by CLSI14. Multi-drug resistance was defined as resistance to 3 or more different classes of antibiotics23.

Statistics

Differences in frequencies of genetic genotypes and pathogenicity markers among groups was evaluated using chi-squared tests (χ2) on contingency tables with a significance level of P = 0.05. Univariate analysis of variance was performed for inference on differences in average numbers of pathogenicity markers and multi-drug resistance between ESBL-genotypes and phylogenetic groups. Analyses were performed in IBM SPSS Statistics version 19.

Results

Distribution of phylogenetic groups among ESBL-producing E. coli from water

Among the ESBL-producing E. coli from both surface water and wastewater, ten different genotypes were represented. The most prevalent ESBL-genotypes in both types of water were CTX-M-15 (41%) and CTX-M-1 (26%), which together constituted 65% and 74% of all ESBL-producing E. coli in surface water and wastewater respectively. Other observed genotypes were CTX-M-14 (9.4%), SHV-12 (8.2%), TEM-52 (5.3%), CTX-M-3 (4.1%), CTX-M-9 (2.4%), CTX-M-27 (1.8%), CTX-M-32 (1.2%) and CTX-M-55 (0.6%). With the exception of the two least prevalent genotypes which were solely detected in wastewater, all were detected in both surface water and wastewater. With respect to phylogenetic groups, the majority of isolates were of group A (55%: A0 25%, A1, 30%), followed by group D (22%: D1 13%, D2 9%), B2 (13%: B22 0.6%, B23 12%), and B1 (10%). Among wastewater isolates, all phylogenetic (sub)groups were represented, while in the surface water isolates B1, B22 and D2 were not detected. In order to determine the distribution of ESBL-genotypes among phylogenetic groups, all isolates were pooled for analysis. With the exception of the two least prevalent ESBL-genotypes (CTX-M-32 and CTX-M-57), all ESBL-genes were detected in multiple E. coli backgrounds (Fig. 1). Phylogenetic groups generally associated with commensal E. coli (A and B1) as well as the pathogenic D group, had representatives among eight of the ten different ESBL-genotypes. By contrast, E. coli of phylogenetic (sub)groups B2 and D2 were each observed among four different ESBL-genotypes. Significant associations were identified between ESBL-genotype CTX-M-1 and phylogenetic groups A and D, CTX-M-15 and B2, and SHV-12 and D (Table 1).

Figure 1. Distribution of phylogenetic subgroups among ESBL-genotypes.

Figure 1

Open in a new tab

Different colors represent the percentages of isolates belonging to the different phylogenetic subgroups for each ESBL-genotype. At the x-axis, the total number of isolates per ESBL-genotype is indicated between brackets.

Table 1. Significant (χ2 P < 0.05) positive and negative associations between ESBL-genotype, virulence markers and phylogenetic groups.

Marker/group1 ESBL-genotype
 Phylogenetic group
CTX-M-1 SHV-12 CTX-M-15 CTX-M-14 CTX-M-27 A B1 B2 D
kpsMIII ns ns ns χ2 7.2; OR 3.9 (1.4–11) χ2 6.6; OR 13.4 (1.1–169) χ2 47.5; OR 0.06 (0.02–0.15) ns χ2 43.6; OR 27.1 (7.5–98.1) χ2 16.5 OR 4.6 (2.1–10.0)
iutA ns ns χ2 6.2; OR 2.2 (1.2–4.2) ns ns ns ns χ2 6.2; OR 3.1 (7.5–98.1) ns
hlyD ns ns ns ns ns χ2 5.1; OR 0.95 (0.90–1.00) χ2 7.3; OR 10.1 (1.3–76.7) ns ns
afa χ2 4.1; OR 0.2 (0.1–1.2) ns χ2 4.6; OR 3.0 (1.0–8.6) χ2 8.9; OR 5.2 (1.6–18) ns χ2 23.4; OR 0.78 (0.69–0.88) ns ns ns
A χ2 8.1; OR 2.9 (1.4–6.1) ns ns ns ns — — — —
B2 ns ns χ2 8.0; OR 3.7 (1.4–9.7) ns ns — — — —
D χ2 8.2; OR 0.19 (0.1–0.7) χ2 4.0; OR 3.0 (1.0–9.4) ns ns ns — — — —
Open in a new tab

χ2 is Chi-square value, OR is odds ratio with 95% confidence interval, ns is not significant.

1The markers stx1, stx2, eae. aggR, papA, safS, focG, ipaH, eltB, astA, and phylogenetic group B1 showed no significant associations with any of the ESBL-genotypes or phylogenetic groups and were therefore omitted from this table.

Distribution of virulence markers among ESBL-producing E. coli isolates from water

Overall, 17.1% (29/170) of all ESBL-producing E. coli isolates from surface water and wastewater assessed in this study were potentially pathogenic, as defined by the presence of characteristic virulence markers (Table 2). Potential pathogens were significantly associated with ESBL-genotype genotype CTX-M-15 (21/29 among potential pathogens versus 48/141 among non-pathogens) (X2 14.7; P < 0.001; OR = 5.1). In addition, potential pathogens were significantly associated with phylogenetic group B2 (X2 = 23.5, P < 0.001, OR = 8.8).

Table 2. Prevalence of pathogenic variants among ESBL-producing E. coli.

Pathogenic category Genetic marker Surface water* (n = 82) Waste water* (n = 88) Total* (n = 170) Phylogenetic group/ESBL-genotype
EAEC aggR 5 (6.1%) 2 (2.2%) 7 (4.1%) A0/CTX-M-15 (1)
        A1/CTX-M-15 (2)
        A1/CTX-M-27 (1)
        D1/CTX-M-15 (1)
        D1/CTX-M-14 (1)
        D2/CTX-M-15 (1)
EIEC ipaH 0 (0%) 2 (2.3%) 2 (1.2%) B23/CTX-M-15 (1)
        D2/CTX-M-15 (1)
EPEC eae 0 (0%) 3 (3.3%) 3 (1.8%) B23/CTX-M-15 (1)
        B1/CTX-M-15 (2)
ETEC eltB or estA 2 (2.4%) 0 (0%) 2 (1.2%) A0/CTX-M-15 (2)
ExPEC ≥3 markers 5 (6.1%) 10 (11.4%) 15 (8.8%) B23/CTX-M-1 (3)
        B23/CTX-M-14 (1)
        D2/CTX-M-14 (2)
        A0/CTX-M-15 (1)
        B23/CTX-M-15 (6)
        D2/CTX-M-15 (2)
STEC stx1 or stx2 0 (0%) 0 (0%) 0 (0%) —
Total   12 (14.6%) 17 (19.3%) 29 (17.1%)  
Open in a new tab

*Indicated are the numbers and percentages of suspected pathogenic isolates, based on the presence of the indicated markers.

Suspected ExPECs (defined as isolates with three or more ExPEC markers) constituted 8.8% (15/170) of all ESBL-producing isolates: 6.1% of the surface water isolates and 11.4% of the wastewater isolates (Table 2). No significant associations were identified between ExPEC and ESBL-genotype. However, ExPEC was significantly associated with phylogenetic group B2 (X2 42.2, P < 0.001, OR = 23.8). Ninety-eight ESBL-producing isolates (58%) harbored at least one ExPEC marker with iutA (39%) and kpsMII (28%) the most frequently observed and detected in eight and seven of all ESBL-genotypes, respectively (Fig. 2). The markers afa (10%), papA (6.5%), and hlyD (2.4%) were only observed associated with CTX-M-1-, CTX-M-14- and/or CTX-M-15-genotypes; focG and sfaS were each observed only once (0.6% of isolates), in relation with CTX-M-1. Significant positive associations were observed between kpsMIII and ESBL-genotype CTX-M-14 and CTX-M-27, between iutA and CTX-M-15, and between afa and, CTX-M-15 and CTX-M-14; a negative association was observed between afa and CTX-M-1 (Table 1). In addition, significant associations were identified between kpsMIII and phylogenetic group A (negative), B2 (positive) and D (positive); between iutA and phylogenetic group B2 (positive); between hlyD and phylogenetic group A (negative) and B1 (positive); and between afa and phylogentic group A (negative) (Table 1).

Figure 2. Distribution of ExPEC markers among ESBL-producing E. coli isolates.

Figure 2

Open in a new tab

Indicated are the numbers of isolates carrying the indicated ExPEC markers, for each ESBL-genotype. At the x-axis, the total number of isolates per ESBL-genotype are indicated between brackets.

Suspected diarrheagenic variants constituted 8.3% of all ESBL-producing E. coli isolates from water (8.5% and 7.8% from surface water and wastewater, respectively) (Table 2). The most prevalent pathogenic variants were EAEC (4.1% of all isolates), EPEC (1.8%), EIEC (1.2%) and ETEC (1.2%). STEC were not detected among the 170 ESBL-producing E. coli analyzed. Only three aggR positive isolates showed strong biofilm formation (all three ODstrain/ODblanco = 12). These isolates belonged to phylogenetic group and ESBL genotype: A0/CTX-M-15, A1/CTX-M-15, A1/CTX-M-27. The other four aggR positive isolates did not distinguish themselves from the other E. coli with respect to biofilm formation (average ODstrain/ODblanco = 1.2). Potential diarrheagenic E. coli was significantly associated with ESBL-genotype CTX-M-15 (X2 = 8.3, P = 0.004, OR 2.8). No associations were found with phylogenetic group.

Multidrug-resistance

Overall, 77% of the ESBL-producing E. coli isolates were resistant to three or more different classes of antibiotics (i.e. were multidrug-resistant (MDR)). The proportion MDR E. coli was slightly (but not statistically different) higher in suspected pathogenic variants compared to isolates not identified as pathogenic (84% vs. 75%, p > 0.1 Chi-square Test). Two of three suspected EAEC variants (both with strong biofilm forming capacity), all seven isolates carrying EIEC, EPEC, or ETEC markers, and 12 of 15 (80%) of suspected ExPEC variants were multidrug-resistant.

Suspected pathogenic variants carried on average 6.8 different resistance types (Table 3). No differences was observed between ExPEC and non-ExPEC (P = 0.814), nor between the different phylogenetic groups (P = 0.843). The various ESBL-genotypes differed in the number antibiotics the isolates were resistant to (P < 0.001), with highest level of multi-drug resistance among CTX-M-9, CTX-M-15, CTX-M-27 and SHV-12 (Fig. 3).

Table 3. Antibiotic resistance profiles of suspected pathogenic variants.

Pathogeniccategory Phylogenetic group/ESBL-genotype Sample IDa Origin ABR profileb No. resistant classes
EAEC A0/CTX-M-15 Sw5 Surface water AmCxCzTm 2
A1/CTX-M-15 Sw6 Surface water AmCxCzTeNaStSuTm 5
A1/CTX-M-27 Tp5 Airport WWTP AmCxCzTeNaGeStSuTmCh 6
EIEC B23/CTX-M-15 Tp3 mWWTP AmCxCzTeNaGeStSuTm 5
D2/CTX-M-15 Tp5 Airport WWTP AmCxCzTeCiNaGeSuTmCh 6
A1/CTX-M-27 Tp5 Airport WWTP AmCxCzTeNaGeStSuTmCh 6
EPEC B1/CTX-M-15 Tp5 Airport WWTP AmCxCzTeNaStSu 5
B1/CTX-M-15 Tp3 mWWTP AmCxCzNaTm 3
B23/CTX-M-15 Tp1.c mWWTP AmCxCzTeGeSuTm 4
ETEC A0/CTX-M-15 Sw4.a Surface water AmCxCzTeNaSuTm 4
A0/CTX-M-15 Sw4.b Surface water AmCxCzTeNaSuTm 4
ExPEC A0/CTX-M-15 Sw1 Surface water AmCxCzAxTe 3
B23/CTX-M-1 Sw1 Surface water AmCx 1
B23/CTX-M-1 Tp1.a mWWTP AmCxTeCiNaStSuTm 5
B23/CTX-M-1 Tp2 mWWTP AmCxAxTeNaGeStSuTm 6
B23/CTX-M-14 Sw2 Surface water AmCxTe 2
B23/CTX-M-14 Sw1 Surface water AmCxCzCiNaGe 3
B23/CTX-M-15 Sw3 Surface water AmCxCzTe 2
B23/CTX-M-15 Hc1 HCI AmCxCzAxTeCiNaGeStSuTm 6
B23/CTX-M-15 Hc2 HCI AmCxCzAxTeCiNaStSuTm 6
B23/CTX-M-15 Tp1.b mWWTP AmCxCzTeNaGeStSuTm 5
B23/CTX-M-15 Tp1.c mWWTP AmCxCiNaSuTm 3
B23/CTX-M-15 Hc2 HCI AmCxTeStSuTm 4
D2/CTX-M-14 Tp3 mWWTP AmCxNaGe 3
D2/CTX-M-14 Hc2 HCI AmCxCzTeNaGe 4
D2/CTX-M-15 Tp4 mWWTP AmCxCzTeCiNaTm 4
Open in a new tab

aNumbers are used to indicate different locations, the appending letters (a, b, or c) indicate different time-points; Sw = surface water, Tp = wastewater from WWTP, Hc = wastewater from health care institution.

bAm = ampicillin, Cx = cefotaxime, Cz = ceftazidime, Ax = amoxicillin + clavulanic acid, Te = tetracycline, Ci = ciprofloxacin, Na = nalidixic acid, Ge = gentamycin, St = streptomycin, Su = sulfisoxazol, Tm = trimethoprim, Ch = chloramphenicol, resistance was defined as disk diameters above clinical break-point (CLSI).

Figure 3. Boxplot of number of antimicrobial resistances (ABR) among isolates of with different ESBL-genotypes.

Figure 3

Open in a new tab

Solid horizontal line represents the median, the box displays the 25%–75% quartile range, the stems show the minimum and maximum values, the circle indicate an outlier.

Discussion

Besides health care settings and food, the environment is likely to have a role in the dissemination of ESBL-producing bacteria and may serve as an exposure route to humans. Previously recreational waters were identified as a potential exposure source of ESBL-producing E. coli5. Even though (outside the clinical setting) E. coli is generally considered a relatively harmless inhabitant of the human (and animal) gut, major public health risks may be associated with the spread of ESBL-producing commensal bacteria. Firstly, upon colonization ESBL-producing commensal bacteria may disseminate and transfer ESBL-encoding genes to intestinal pathogens through horizontal gene transfer24. Secondly, while relatively harmless for healthy individuals, these opportunistic bacteria may cause disease in more vulnerable individuals, such as hospitalized individuals, the elderly or newborns. Thirdly, exposure to ESBL-producing pathogenic E. coli variants may directly result in hard-to-treat infection, also in healthy individuals. The public health impact of exposure to ESBL-producing E. coli (and other AMR commensal bacteria) is determined by the sum of these individual risks.

The current study showed that a considerable fraction of ESBL-producing E. coli from Dutch surface water and wastewater are potential diarrheagenic or extraintestinal pathogenic. Thus far, only few studies have examined the pathogenicity of non-clinical ESBL-producing E. coli strains. A recent study from South Korea reported 60% of ESBL-producing E. coli isolated from a river to be potentially pathogenic, which is markedly higher than our findings12. These difference in fractions of pathogenic variants between studies possibly reflects differences in water management or differences in antimicrobial selection pressure within humans since extraintestinal pathogenic E. coli primarily have a human reservoir11. Previously, we have shown concentrations of ESBL-producing E. coli ranging from 1.5 to 150 colony forming units (cfu)/liter in recreational waters5. Quantitative risk assessment based on these concentrations established an average probability of exposure to ESBL-producing E. coli of 0.20 per swimming event for children, and slightly lower for men (0.16) and women (0.13) (Schijven et al. unpublished). Combined with the results of the present study the probability of exposure to a pathogenic ESBL-producing E. coli becomes 0.03 (0.20 × 0.15) which, given frequent recreation in surface waters, can rise to a considerable cumulative probability of exposure. Recreational freshwater swimming has been identified as a significant risk factor for acquiring urinary tract infection caused by ESBL-producing enterobacteriaceae6. The results of the present study provides experimental evidence to this epidemiological association. Ingestion of surface water may lead to intestinal colonization of extraintestinal ESBL-E. coli and subsequent urinary tract infection.

The predominance of CTX-M-15 among ESBL-producing E. coli in Dutch water and its strong association with virulence factors is in agreement with CTX-M-15 being one of the most common ESBL-genotypes in humans and often being health-care related (i.e. community onset urinary tract infections or nosocomial infections), in the Netherlands and globally25,26,27. The strong association with virulence factors is also consistent with the observed association between CTX-M-15 and phylogenetic group B2 in the present as well as a previous study27, which is thought to be more virulent than other groups28,29. The global occurrence of CTX-M-15-producing ESBL-E. coli is partially associated with the spread of the pathogenic ST131 clone, causing urinary tract and bloodstream infections worldwide30. Two of the 15 B23/CTX-M-15 isolates included in the current study were characterized with respect to sequence type, and were also identified as ST131 (data not shown). The second prevalent ESBL-genotype among the Dutch water isolates was CTX-M-1. Although globally this genotype was not identified as a major human genotype26, in the Netherlands it has previously been recognized as highly prevalent, among humans (the second most prevalent genotype after CTX-M-15)25 as well as broilers31. The results from this study showed that the microbiological contents of surface- and wastewater provide a reflection of what is circulating among the human population.

This study did not detect any Shiga toxin-producing E. coli (STEC) which might be due to several reasons. ESBL-positive STEC seems to be a rare phenomenon35. In addition the isolation method used did not select for the isolation of the most frequent STEC serotype O157 since this type, in contrast to other E. coli, is β-glucuronidase-negative. It remains unknown why STEC still represent a minor subpopulation of strains that have acquired ESBL genes. A possible explanation might be the relatively low selection pressure in the bovine reservoir (due to lower use of antibiotics4), as well as in humans since clinical STEC infections are rarely treated with antibiotics. However, there may be a risk of STEC emerging in the aquatic environment, since Stx-phages are widely distributed and can be acquired by E. coli belonging to various phylogenetic and pathogenic groups, including EAEC (for example the German Stx-producing EAEC O104:H4 outbreak)32,33.

Biofilm formation may be an important contributory factor in persistent infections by allowing the bacteria to evade the local immune system and by preventing the transport of antibacterial factors, including antibiotics. Assays to quantify biofilm formation have been suggested as a possible method of screening for pathogenic EAEC strains34. In the current study strong biofilm formation was observed with three out of seven aggR positive isolates, which raises the question whether the remaining four isolates can truly be considered potential EAEC based on the detection of aggR only. However, extensive studies have been performed using PCR tests to find the right combination of genes identifying the “truly pathogenic” EAEC strains but no consensus has been reached on this matter35. The aggR gene is highly conserved among EAEC strains and has been found to be associated with diarrhea in several studies35. We conclude that all aggR positive strains identified in this study can be considered at least as potential EAEC. In addition to the diarrheagenic nature of EAEC these strains also have been associated with urinary tract infections (UTIs) made possible by a combination of EAEC (aggR) and ExpEC virulence factors (among which iutA)36. Three of the aggR positive isolates in the current study were simultaneously positive for the ExPEC marker kpsMIII and another 3 isolates positive for iutA, making these isolates potential hybrid EAEC-ExPEC strains.

In conclusion, this study has shown that environmental ESBL-producing E. coli not only are an indirect public health threat by being harmless commensal carriers of ESBL-genes, but also poses a direct threat upon exposure by actually carrying virulence factors representative for major pathogenic groups of E. coli. Surface waters and wastewater are a potential reservoir of E. coli combining ESBL-genes, high level of multi-drug resistance, and virulence factors. These findings demonstrate the role of the environment in the transmission ecology of pathogenic, and in particular extraintestinal pathogenic, ESBL-producing E. coli.

Additional Information

How to cite this article: Franz, E. et al. Pathogenic Escherichia coli producing Extended-Spectrum β-Lactamases isolated from surface water and wastewater. Sci. Rep. 5, 14372; doi: 10.1038/srep14372 (2015).

Supplementary Material

Supplementary info Table S1
srep14372-s1.pdf (166KB, pdf)

Acknowledgments

The authors wish to thank Gretta Lynch and Patrick de Kruijf who isolated the included strains from water samples.

Footnotes

Author Contributions E.F. and H.B. conceived the experiments; C.V. and A.H. conducted the experiments; E.F., H.B. and A.R.H. analyzed the results. All authors reviewed the manuscript.

References

  1. Oteo J., Pérez-Vázquez M. & Campos J. Extended-spectrum β-lactamase producing Escherichia coli: Changing epidemiology and clinical impact. Current Opinion in Infectious Diseases 23, 320–326, 10.1097/QCO.0b013e3283398dc1 (2010). [DOI] [PubMed] [Google Scholar]
  2. Haenni M., Châtre P. & Madec J. Y. Emergence of Escherichia coli producing extended-spectrum AmpC β-lactamases (ESAC) in animals. Frontiers in Microbiolog1 5, 10.3389/fmicb.2014.00053 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Huijbers P. M. C. et al. Prevalence of extended-spectrum β-lactamase-producing Enterobacteriaceae in humans living in municipalities with high and low broiler density. Clinical Microbiology and Infection 19, E256–E259, 10.1111/1469-0691.12150 (2013). [DOI] [PubMed] [Google Scholar]
  4. Anonymous. NETHMAP/MARAN 2013 Consumption of antimicrobial agents and antimicrobial resistance among medically important bacteria in the Netherlands & Monitoring of Antimicrobial Resistance and Antibiotic Usage in Animals in the Netherlands in 2012. (Nijmegen, 2013). [Google Scholar]
  5. Blaak H. et al. Prevalence and characteristics of ESBL-producing E. coli in Dutch recreational waters influenced by wastewater treatment plants. Veterinary Microbiology 171, 448–459, 10.1016/j.vetmic.2014.03.007 (2014). [DOI] [PubMed] [Google Scholar]
  6. Søraas A., Sundsfjord A., Sandven I., Brunborg C. & Jenum P. A. Risk Factors for Community-Acquired Urinary Tract Infections Caused by ESBL-Producing Enterobacteriaceae -A Case-Control Study in a Low Prevalence Country. PLoS ONE 8, 10.1371/journal.pone.0069581 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chaudhuri R. R. & Henderson I. R. The evolution of the Escherichia coli phylogeny. Infection, Genetics and Evolution 12, 214–226 (2012). [DOI] [PubMed] [Google Scholar]
  8. White A. P. et al. Intergenic sequence comparison of Escherichia coli isolates reveals lifestyle adaptations but not host specificity. Applied and Environmental Microbiology 77, 7620–7632 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Méric G., Kemsley E. K., Falush D., Saggers E. J. & Lucchini S. Phylogenetic distribution of traits associated with plant colonization in Escherichia coli. Environmental Microbiology 15, 487–501 (2013). [DOI] [PubMed] [Google Scholar]
  10. Walk S. T., Alm E. W., Calhoun L. M., Mladonicky J. M. & Whittam T. S. Genetic diversity and population structure of Escherichia coli isolated from freshwater beaches. Environmental Microbiology 9, 2274–2288, 10.1111/j.1462-2920.2007.01341.x (2007). [DOI] [PubMed] [Google Scholar]
  11. Kaper J. B., Nataro J. P. & Mobley H. L. T. Pathogenic Escherichia coli. Nature Reviews Microbiology 2, 123–140, 10.1038/nrmicro818 (2004). [DOI] [PubMed] [Google Scholar]
  12. Jang J. et al. Pathogenic Escherichia coli strains producing extended-spectrum β-lactamases in the Yeongsan river basin of South Korea. Environmental Science and Technology 47, 1128–1136, 10.1021/es303577u (2013). [DOI] [PubMed] [Google Scholar]
  13. Blaak H. et al. Multidrug-resistant and extended spectrum beta-lactamase-producing Escherichia coli in dutch surface water and wastewater. PLoS ONE 10, 10.1371/journal.pone.0127752 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Clinical and Laboratory Standards Institute. Performance standards for antimicrobial susceptibility testing; twentieth informational supplement M100-S20. (CLSI, Wayne, PA, USA, 2010). [Google Scholar]
  15. Clermont O., Bonacorsi S. & Bingen E. Rapid and simple determination of the Escherichia coli phylogenetic group. Appl Environ Microbiol 66, 4555–4558 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Escobar-Páramo P. et al. Identification of forces shaping the commensal Escherichia coli genetic structure by comparing animal and human isolates. Environ Microbiol 8, 1975–1984, 10.1111/j.1462-2920.2006.01077.x (2006). [DOI] [PubMed] [Google Scholar]
  17. Dallenne C., Da Costa A., Decré D., Favier C. & Arlet G. Development of a set of multiplex PCR assays for the detection of genes encoding important β-lactamases in Enterobacteriaceae. J Antimicrob Chemother 65, 490–495, 10.1093/jac/dkp498 (2010). [DOI] [PubMed] [Google Scholar]
  18. Antikainen J. et al. New 16-plex PCR method for rapid detection of diarrheagenic Escherichia coli directly from stool samples. Eur J Clin Microbiol Infect Dis 28, 899–908, 10.1007/s10096-009-0720-x (2009). [DOI] [PubMed] [Google Scholar]
  19. Jakobsen L. et al. Broiler chickens, broiler chicken meat, pigs and pork as sources of ExPEC related virulence genes and resistance in Escherichia coli isolates from community-dwelling humans and UTI patients. Int J Food Microbiol 142, 264–272, 10.1016/j.ijfoodmicro.2010.06.025 (2010). [DOI] [PubMed] [Google Scholar]
  20. Johnson J. R. & Stell A. L. Extended virulence genotypes of Escherichia coli strains from patients with urosepsis in relation to phylogeny and host compromise. Journal of Infectious Diseases 181, 261–272, 10.1086/315217 (2000). [DOI] [PubMed] [Google Scholar]
  21. Wakimoto N. et al. Quantitative biofilm assay using a microtiter plate to screen for enteroaggregative Escherichia coli. The American journal of tropical medicine and hygiene 71, 687–690 (2004). [PubMed] [Google Scholar]
  22. Naves P. et al. Measurement of biofilm formation by clinical isolates of Escherichia coli is method-dependent. J Appl Microbiol 105, 585–590, 10.1111/j.1365-2672.2008.03791.x (2008). [DOI] [PubMed] [Google Scholar]
  23. Magiorakos A. P. et al. Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance. Clin Microbiol Infect 18, 268–281, 10.1111/j.1469-0691.2011.03570.x (2012). [DOI] [PubMed] [Google Scholar]
  24. Hunter P. R., Wilkinson D. C., Catling L. A. & Barker G. C. Meta-analysis of experimental data concerning antimicrobial resistance gene transfer rates during conjugation. Applied and Environmental Microbiology 74, 6085–6090, 10.1128/AEM.01036-08 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Voets G. M. et al. Population Distribution of Beta-Lactamase Conferring Resistance to Third-Generation Cephalosporins in Human Clinical Enterobacteriaceae in The Netherlands. PLoS ONE 7, 10.1371/journal.pone.0052102 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Ewers C., Bethe A., Semmler T., Guenther S. & Wieler L. H. Extended-spectrum β-lactamase-producing and AmpC-producing Escherichia coli from livestock and companion animals, and their putative impact on public health: A global perspective. Clinical Microbiology and Infection 18, 646–655 (2012). [DOI] [PubMed] [Google Scholar]
  27. Cantón R. et al. Prevalence and spread of extended-spectrum β-lactamase-producing Enterobacteriaceae in Europe. Clinical Microbiology and Infection 14, 144–153, 10.1111/j.1469-0691.2007.01850.x (2008). [DOI] [PubMed] [Google Scholar]
  28. Le Gall T. et al. Extraintestinal virulence is a coincidental by-product of commensalism in b2 phylogenetic group Escherichia coli strains. Molecular Biology and Evolution 24, 2373–2384, 10.1093/molbev/msm172 (2007). [DOI] [PubMed] [Google Scholar]
  29. Picard B. et al. The link between phylogeny and virulence in Escherichia coli extraintestinal infection? Infection and Immunity 67, 546–553 (1999). [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Petty N. K. et al. Global dissemination of a multidrug resistant Escherichia coli clone. Proceedings of the National Academy of Sciences of the United States of America 111, 5694–5699, 10.1073/pnas.1322678111/-/DCSupplemental (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Huijbers P. M. C. et al. Methicillin-resistant Staphylococcus aureus and extended-spectrum and AmpC β-lactamase-producing Escherichia coli in broilers and in people living and/or working on organic broiler farms. Veterinary Microbiology 176, 120–125, 10.1016/j.vetmic.2014.12.010 (2015). [DOI] [PubMed] [Google Scholar]
  32. Muniesa M. & Jofre J. Abundance in sewage of bacteriophages that infect Escherichia coli O157:H7 and that carry the Shiga toxin 2 gene. Applied and Environmental Microbiology 64, 2443–2448 (1998). [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Tozzoli R. et al. Shiga toxin-converting phages and the emergence of new pathogenic Escherichia coli: a world in motion. Frontiers in cellular and infection microbiology 4, 80 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Bangar R. & Ballal M. Identification of enteroaggregative Escherichia coli in infants with acute diarrhea based on biofilm production in Manipal, south India. Indian Journal of Medical Sciences 62, 8–12 (2008). [PubMed] [Google Scholar]
  35. Jensen B. H., Olsen K. E. P., Struve C., Krogfelt K. A. & Petersen A. M. Epidemiology and clinical manifestations of enteroaggregative Escherichia coli. Clinical Microbiology Reviews 27, 614–630, 10.1128/CMR.00112-13 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Boll E. J. et al. Role of enteroaggregative Escherichia coli virulence factors in uropathogenesis. Infection and Immunity 81, 1164–1171, 10.1128/IAI.01376-12 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary info Table S1
srep14372-s1.pdf (166KB, pdf)

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES