Comparison of Extended-Spectrum β-Lactamase-Producing Escherichia coli Isolates from Drinking Well Water and Pit Latrine Wastewater in a Rural Area of China

The present study was conducted to gain insights into the occurrence and characteristics of extended-spectrum beta-lactamase- (ESBL-) producing Escherichia coli (E. coli) from drinking well water in the rural area of Laiwu, China, and to explore the role of the nearby pit latrine as a contamination source. ESBL-producing E. coli from wells were compared with isolates from pit latrines in the vicinity. The results showed that ESBL-producing E. coli isolates, with the same antibiotic resistance profiles, ESBL genes, phylogenetic group, plasmid replicon types, and enterobacterial repetitive intergenic consensus-polymerase chain reaction (ERIC-PCR) fingerprints, were isolated from well water and the nearby pit latrine in the same courtyard. Therefore, ESBL-producing E. coli in the pit latrine may be a likely contributor to the presence of ESBL-producing E. coli in rural well water.


Introduction
The use of cephalosporins in clinical practices of humans and animals has contributed to the occurrence of extendedspectrum -lactamase-(ESBL-) producing bacteria across the world [1,2]. These bacteria are resistant to most betalactam antibiotics, such as first-, second-, third-, and fourthgeneration cephalosporins, which would increase medical costs and limit treatment options [3,4].
Many species of Gram-negative bacteria can produce ESBLs, but ESBLs are mainly detected in Enterobacteriaceae, particularly in Escherichia coli (E. coli) and Klebsiella spp. [5]. It is worrisome that not only are ESBL-producing E. coli associated with community-acquired infections, especially urinary tract infection [6,7], but also they have been isolated from food-producing animals and healthy human populations [8,9].
ESBL-producing E. coli in the intestinal tracts of humans and animals are easily excreted into the environment, particularly into surface water bodies. Numerous studies have shown that ESBL-producing E. coli can be detected in water bodies [10][11][12][13].
Human exposure to ESBL-producing E. coli in water bodies may easily occur, for example, when contaminated surface water is used for recreation, for irrigation of crops, or as a drinking water source [14,15]. Similarly, animal exposure to ESBL-producing E. coli in water bodies may also easily occur, for instance, when they drink contaminated surface water [1]. To limit the spread of ESBL-producing E. coli via water bodies, it is pivotal to understand the possible contamination sources.
To date, numerous researches about contamination sources of ESBL-producing E. coli have been focusing on hospital environments, animal farms, rivers and lakes, and wastewater treatment plants [10,12,[16][17][18][19]. But data about the prevalence and possible contamination sources of ESBLproducing E. coli in rural water well in undeveloped regions is very limited.
In undeveloped areas of China, plenty of domestic wells are used to supply drinking water. Impurities from the surface easily enter wells, so the rural wells are relatively easy to be contaminated by bacteria. More importantly, most bacteria, especially ESBL-producing E. coli, could contaminate well water coming from fecal material from humans and animals, for instance, from on-site sanitation systems, such as pit latrines and septic tanks [20]. But little information about ESBL-producing E. coli contamination in well waters is available in China. This study was therefore conducted to gain insights into the prevalence and characterization of ESBLproducing E. coli from drinking well water in the rural area of Laiwu, China, and to explore the role of the nearby pit latrine as a contamination source.

Well Selection and Homeowner Enrollment.
The sampling site was located in the rural area of Laiwu city of China, in which the majority of households have both a shallow drinking well (approximately 12-18 m in depth) and a pit latrine in their courtyards, and the distance between them is about 8-10 m. To minimize the influence of surrounding surface water and animal fecal material on the groundwater, water wells located near pit latrines and not within 100 m of rivers/lakes or animal farms were favored for selection in this study. In addition, the distance between two sampled wells is no less than 100 m.
Owners of water wells were sent a letter describing the study and were telephoned two days later to request their participation and confirm that their water well was near a pit latrine.

Sample Collection.
Between July and August 2014, a 500 mL well water sample was collected using a sterile bottle from each well (100 wells). During the same period, wastewater samples from 100 pit latrines located near the wells were obtained using the same method. All samples were immediately transported to our lab in an icebox and processed in 6 h.

Isolation and Identification of ESBL-Producing E. coli.
Each water sample was filtered using a membrane filter (0.45 m). The filter was then spread on MacConkey agar plate containing cefotaxime (4 g/mL) and incubated at 37 ∘ C overnight. A suspected E. coli colony was identified using API 20E (BioMérieux, Marcy-l'Etoile, France).

Detection of Beta-Lactamase Gene.
Based on the previously published reference [23], ESBL-producing E. coli isolates were subjected to multiplex polymerase chain reaction (PCR) to determine the presence/absence of genes encoding CTX-M, SHV, and TEM. PCR products were sequenced and compared with beta-lactamase gene sequences in the GenBank database and in the Lahey website (http://www .lahey.org/studies).

Phylogenetic
Groups. ESBL-producing E. coli isolates were assigned to eight groups, A, B1, B2, C, D, E, F, and clade I, according to the published work by Clermont and his colleagues [24].

ERIC-PCR.
ESBL-producing E. coli isolates were subjected to enterobacterial repetitive intergenic consensus-(ERIC-) PCR [27][28][29]. ERIC-PCR fingerprints were analyzed using the NTSYSpc software (version 2.02K, Applied Biostatistics, Inc., NY, USA). The dendrogram was constructed based on the average relatedness of the matrix using the algorithm of the unweighted pair-group method (UPGMA) in the SAHN program of the NTSYSpc software.

-Lactamase Genes of ESBL-Producing E. coli.
Except that CTX-M-3 gene was restricted to pit latrine wastewater isolates, a similar distribution of -lactamases genes among both sources was found in this study. Twenty-two out of 23 well water isolates (95.6%) and 37 of 40 pit latrine wastewater isolates (92.5%) carried CTX-M genes. Fifteen out of 23 well water isolates (65.2%) and 30 of 40 pit latrine wastewater isolates (75.0%) carried TEM-1 genes, and most of these isolates combined with CTX-M genes. Among all ESBLproducing E. coli carrying CTX-M genes, the most prevalent ESBL gene was CTX-M-15 . No SHV genes were detected in this study (Table 3).

Phylogenetic Groups of ESBL-Producing E. coli.
The phylogenetic group distribution of ESBL-producing E. coli from well water and pit latrine wastewater was comparable. 39.1% and 26.1% of ESBL-producing E. coli from well water belonged to group A and group B1, respectively. Similarly, 32.5% and 22.5% of ESBL-producing E. coli from wastewater belonged to group A and group B1. 26.1% and 8.7% of Table 3: Distribution of -lactamase genes by source group among ESBL-producing E. coli isolates from well water and pit latrine wastewater. Note: TEM-1 is not an ESBL.  the well water isolates belonged to groups D and B2, respectively. 27.5% and 17.5% of the wastewater isolates belonged to groups D and B2, respectively (Table 4).
3.6. ERIC-PCR Analysis. ERIC analysis showed that genetic relatedness of all ESBL-producing E. coli isolates ranged between 57.0% and 100.0%. Overall, the isolates from well water and wastewater were extensively intermixed. Of note, seven isolates from five different wells showed 100.0% genetic similarity with isolates from the pit latrine in the same courtyard ( Figure 1).

Discussion
In this study, there were 7 matches (100.0% similarity) among ESBL-producing E. coli from two source groups in the same courtyard. These findings suggested that pit latrine wastewater might be an important source for ESBL-producing E. coli isolates from well water. High resistance rates of ESBL-producing E. coli forlactamases and other antibiotic classes were observed in Table 5: Distribution of plasmid replicons of ESBL-producing E. coli isolates from well water and pit latrine wastewater. 2 (8.7) 1 (2.5) this study, which is consistent with previous studies about ESBL-producing E. coli from water bodies of China or other countries [30][31][32]. Additionally, relatively high resistance rates to fluoroquinolone (72.5%), ciprofloxacin (62.5%), and gentamicin (62.5%) from wastewater samples were observed, which is related to the fact that those antimicrobials are widely used by humans in China [33,34]. Of note, compared with pit latrine wastewater, ESBL-producing E. coli from well water showed higher rates of resistance against tetracycline (78.3% versus 57.5%), which is frequently used in veterinary practice [33]. Therefore, the higher proportion of tetracycline resistant strains in well water isolates may be associated with the contamination of animal-borne bacteria. Fortunately, only 4.0% of wastewater isolates were resistant to amikacin, and 100.0% of the isolates were susceptible to imipenem and meropenem.
In both source groups, the predominant CTX-M-1 group gene was CTX-M-15 , and CTX-M-14 was the most prevalent CTX-M-9 group gene, which is in agreement with several other reports in China [33,35]. Additionally, CTX-M-15 and CTX-M-14 were used to represent the epidemic CTX-M-1 and CTX-M-9 groups in humans in Asia, respectively [36]. Of note, CTX-M-55 was relatively prevalent in isolates from both sources, which is frequently found in human and foodproducing animals in China [34,37]. CTX-M-64 was detected in isolates from both sources, which could have resulted from recombination of CTX-M-55 and CTX-M-14 [34,38,39]. In addition, TEM-1-type gene was found in both sources, but the lack of identification of TEM-1-type variants by direct sequencing was a limitation of the study.
Several studies have shown that ESBL-producing E. coli from patients, poultry, and environmental waters had the same ESBL genes, resistance profiles, and genomic backbone [33,40]. Similarly, ESBL-producing E. coli exhibiting identical antibiotic resistance profiles, ESBL genes, phylogenetic group, plasmid replicon types, and ERIC fingerprints were isolated from well water and the pit latrine in the same courtyard.  Sources  AMP  PRL  AMC  CF  CXM  CAZ  CRO  CPM  IPM  MEC  AMK  GM  NA  CIP  TE  C  FIB  HI1  P  L/M  Y  A/C  F  N  IncI1  FIA  A  D  B2  B1  CTXM-64  CTXM-55  CTXM-15  CTXM-14  CTXM-3  TEM- The result suggested that there is possible transmission of bacteria carrying resistance gene between pit latrine wastewater and well water.
There was a similar phylogroup distribution between both sources, and the prevalent groups were A and B1.
Phylogroup A is common in commensal strains from chickens [41]. Also, phylogroup B1 is prevalent among extraintestinal infectious isolates from humans, and group B2 is the most prevalent phylogroup among isolates from humans [39,42]. The existence of groups B1 and B2 isolates in well water might 6 BioMed Research International be an indication of virulent human isolates being spread into the water bodies.
With respect to mobile elements, such as plasmid replicons and -lactamases genes, the high overlap between ESBL-producing E. coli from both sources suggested that the transfer of mobile elements may have occurred from fecal wastewater source isolates to well water ones [43].

Conclusions
Together, these findings showed that ESBL-producing E. coli isolates with the same antibiotic resistance profiles, ESBL genes, phylogenetic group, plasmid replicon types, and ERIC-PCR fingerprints were isolated from well water and the nearby pit latrine in the same courtyard. Therefore, it could be concluded that pit latrine wastewater may be an important contributor to the presence of ESBL-producing E. coli in well water, and ESBL-producing E. coli may disseminate from pit wastewater to well water through long-term permeation or water runoff, especially during heavy rains.