Extended-spectrum β-lactamase prevalence and virulence factor characterization of enterotoxigenic Escherichia coli responsible for acute diarrhea in Nepal from 2001 to 2016

Multidrug-resistant (MDR) Gram-negative bacterial species are an increasingly dangerous public health threat, and are now endemic in many areas of South Asia. However, there are a lack of comprehensive data from many countries in this region determining historic and current MDR prevalence. Enterotoxigenic Escherichia coli (ETEC) is a leading cause of both acute infant diarrhea and traveler’s diarrhea in Nepal. The MDR prevalence and associated resistance mechanisms of ETEC isolates responsible for enteric infections in Nepal are largely unknown. A total of 265 ETEC isolates were obtained from acute diarrheal samples (263/265) or patient control samples (2/265) at traveler’s clinics or regional hospitals in Nepal from 2001 to 2016. Isolates were screened for antibiotic resistance, to include extended spectrum beta-lactamase (ESBL) production, via the Microscan Automated Microbiology System. ETEC virulence factors, specifically enterotoxins and colonization factors (CFs), were detected using multiplex PCR, and prevalence in the total isolate population was compared to ESBL-positive isolates. ESBL-positive isolates were assessed using multiplex PCR for genetic markers potentially responsible for observed resistance. A total of 118/265 (44.5%) ETEC isolates demonstrated resistance to ≥2 antibiotics. ESBL-positive phenotypes were detected in 40/265 isolates, with isolates from 2008, 2013, 2014, and 2016 demonstrating ESBL prevalence rates of 1.5, 34.5, 31.2, and 35.0% respectively. No difference was observed in overall enterotoxin characterization between the total ETEC and ESBL-positive populations. The CFs CS2 (13.6%), CS3 (25.3%), CS6 (30.2%), and CS21 (62.6%) were the most prevalent in the total ETEC population. The ESBL-positive ETEC isolates exhibited a higher association trend with the CFs CS2 (37.5%), CS3 (35%), CS6 (42.5%), and CS21 (67.5%). The primary ESBL gene identified was blaCTX-M-15 (80%), followed by blaSHV-12 (20%) and blaCTX-M-14 (2.5%). The beta-lactamase genes blaTEM-1 (40%) and blaCMY-2 (2.5%) were also identified. It was determined that 42.5% of the ESBL-positive isolates carried multiple resistance genes. Over 30% of ETEC isolates collected post-2013 and evaluated in this study demonstrated ESBL resistance. Persistent surveillance and characterization of enteric ETEC isolates are vital for tracking the community presence of MDR bacterial species in order to recommend effective treatment strategies and help mitigate the spread of resistant pathogens.


Background
Multidrug-resistant (MDR) Gram-negative bacterial species are an established prominent international public health threat [1]. MDR pathogens have an especially high prevalence in South Asia where bacterial infections are common and antibiotic use is widely unregulated [2][3][4]. A lack of comprehensive surveillance programs may result in the unchecked spread of MDR Gram-negative bacterial species and their associated resistance mechanisms. Developing countries with an established national surveillance network, such as Nepal, do not have the resources to perform in-depth isolate characterization to comprehensively identify resistance mechanisms carried by MDR bacterial isolates [5][6][7]. It is imperative that both archived and current bacterial isolates from underdeveloped regions undergo extensive MDR characterization to inform national strategies designed to halt the continuing spread of these dangerous pathogens.
Diarrheagenic Escherichia coli is a ubiquitous presence in the developing world and is responsible for a range of enteric infections [8][9][10]. Enterotoxigenic E. coli (ETEC) is defined by production of 1-3 enterotoxins, and is one of the leading causes of both acute infant diarrhea and traveler's diarrhea worldwide, including Nepal [11][12][13][14]. ETEC attaches to and colonizes cells in the small intestine along the epithelial surface through the interaction of bacterial colonization factors (CFs) with the intestinal cell wall [15,16]. After the establishment of intestinal colonization, the bacterial cells produce heat-labile enterotoxins (LT) and/or heat-stable enterotoxins (STh or STp) that target vital cellular processes [8,17,18]. The production of these enterotoxins most commonly results in abdominal cramping and watery diarrhea, with more severe cases also presenting with fever and nausea. ETEC infections are typically self-limiting with resolution of symptoms 3-7 days after initial onset and a recommended treatment strategy of rest and oral rehydration therapy [8,19]. In spite of ongoing efforts to develop an ETEC vaccine, no viable vaccine candidates have been produced thus far [20,21]. ETEC virulence factors, such as enterotoxins and CFs, are potential targets for vaccines or therapeutic candidates, and determination of regional prevalence rates and specific virulence factors may serve to inform the development of globally effective vaccines [15,[18][19][20][21].
Severe ETEC cases presenting with cholera-like watery diarrhea that do not resolve through oral rehydration therapy may require treatment with antibiotics [8,10]. Initial treatment has historically been administration of first-line beta-lactams and quinolones. However, the evolution and global dissemination of multiple classes of antimicrobial resistance mechanisms has led to an increase of MDR ETEC infections that are insensitive to first-line antibiotics [11,22]. Beta-lactamases and Extended-spectrum β-lactamases (ESBLs) confer resistance to a number of antibiotic classes, such as penicillins, extended-spectrum cephalosporins, and monobactams [23,24]. Global dissemination of these mechanisms has occurred through clonal expansion and/or horizontal transfer of genetic resistance mechanisms between bacterial species [24]. Bacterial pathogens carrying ESBL resistance mechanisms are commonly associated with, and detected in, hospital-acquired infections. However, these antibiotic resistance mechanisms have spread to community-acquired isolates, and are increasingly identified in acute diarrhea and other enteric infections [12,[25][26][27][28][29][30]. ETEC infections are a large cause of acute diarrhea cases in Nepal, but prevalence rates of MDR ETEC isolates, both from a historic or current perspective, are unknown.
The current study determined antibiotic resistance and virulence factors of ETEC isolates previously identified from clinical acute diarrheal samples (263/265) or patient control samples (2/265) collected in Nepal between 2001 and 2016. The ETEC isolates were assayed to determine antimicrobial resistance profiles (including ESBL production) and virulence factor characterization, specifically enterotoxins and colonization factors. Antimicrobial resistance studies that encompass retrospective and current clinical isolates are vital to under standing the regional emergence of ESBL producing organisms, the existing treatment challenges, and the potential for ongoing dissemination. This study serves as a thorough investigation of antibiotic resistance in ETEC isolates from acute diarrheal samples in Nepal, and highlights the need for interventions that may address and stem the continual spread of MDR mechanisms in community settings.

Bacterial strains, media, and chemicals
A total of 265 ETEC isolates were previously isolated and identified from clinical acute diarrheal samples (263/265) or patient control samples (2/265) collected from hospitals and travel clinics in Nepal from 2001 to 2016. The isolates were archived for long-term storage at − 20°C in Luria Bertani (LB) broth + 20% glycerol at the Armed Forces Research Institute of Medical Sciences (AFRIMS) in Bangkok, Thailand. The archived ETEC isolates were grown from frozen glycerol stocks on MacConkey agar (MAC) plates overnight at 37°C. Isolated colonies were re-streaked on tryptic soy agar (TSA) plates, blood agar plates (BAP), or colonization factor agar (CFA) plates in preparation for further phenotypic analysis or DNA isolation.

Antimicrobial susceptibility testing
All ETEC isolates were screened for ESBL production using the Microscan, Walkaway 40 automated system with Negative Breakpoint Combo 34 panels as per manufacturer's instructions (Beckman Coulter, West Sacramento, CA, USA).
ESBL-positive isolates were identified by automatic susceptibility testing panel results by assessing bacterial susceptibility to cefotaxime and ceftazidime with or without clavulanic acid. ESBL-positive ETEC isolates were confirmed through disk diffusion assay against beta-lactam combination agents using both cefotaxime (30 μg) and ceftazidime (30 μg) with or without clavulanic acid (10 μg) following Clinical and Laboratory Standards Institute guidelines [31]. A ≥ 5 mm increase in zone diameter for either antimicrobial agent tested in combination with clavulanate and the zone diameter of the agent when tested along indicated ESBL production.

Multiplex PCR for ESBL gene identification and DNA sequencing
ETEC isolates that tested positive for ESBL production were further characterized to determine potential genetic elements responsible for the observed antimicrobial resistance through multiplex PCR as described previously [33]. All PCR primers and conditions used for beta-lactamase or ESBL characterization were listed previously [33]. The genes identified from the multiplex PCR assays were sequenced to determine variant type. Genetic sequencing services were contracted through AITbiotech (Singapore) or First BASE (Selangor, Malaysia). The PCR products were prepared and sent for sequencing according to company instruction. The nucleotide sequences were analyzed with Sequencher software version 5.3 and BLAST software (http://www.Ncbi.nlm.nih.gov/BLAST).

Statistics
Statistical analyses were performed using GraphPad QuickCalcs. Experimental groups were analyzed using a chi-square with Fisher's exact test with a two-tailed p value of ≤0.05 considered statistically significant.

ETEC enterotoxin gene characterization and prevalence
Detection of the ETEC enterotoxin genes lt, sth, and/or stp was performed for all isolates. The enterotoxin gene lt was detected in a total of 132/265 (49.8%) isolates, sth was detected in a total of 152/265 (57.3%) isolates, and stp was detected in a total of 72/265 (27.5%) isolates. Multiple enterotoxin genes were detected in 88/265 (33.2%) isolates, and enterotoxin gene combinations prevalence are described in Table 2. The enterotoxin combination of sth only was significantly detected less frequently in ESBL-positive isolates compared to ESBLnegative isolates (15% vs. 30.9%; p = 0.0156). No other enterotoxin gene (by total percent detected or by combinations) demonstrated a statistically significant difference in prevalence rates between the two populations.
The beta-lactamase genes bla TEM-1 (16/40, 40%) and bla CMY-2 (1/40, 2.5%) were also shown to be carried in the ESBL-positive ETEC population. A total of 17/40 (42.5%) isolates were shown to carry multiple ESBL and/ or beta-lactamase genes (Table 4). Notably, the multiplex PCR screen did not identify any ESBL or beta-lactamase genes in one phenotypically ESBL-positive ETEC isolate. The multiplex PCR assay performed in this study screens for prevalent resistance genes, but is not a comprehensive

Discussion
MDR Gram-negative bacterial pathogens are a global public health threat [1,6]. ESBL-producing bacterial strains were first detected in, and were mostly limited to, hospital-associated infections, but began being detected in community-associated infections beginning in the early 2000's [25,27,28,34]. Initial reports of community-associated ESBL-producing E. coli strains, including reports originating from Nepal, are mainly comprised of pathogens isolated from urinary tract or blood stream infections [35][36][37][38][39]. However, community-associated infections isolated from enteric MDR pathogens are being increasingly reported in South and Southeast Asia [8,11,12,22,26,30]. ETEC is an enteric pathogen that results in both acute infant diarrhea and traveler's diarrhea [8,10,15]. ESBL-producing ETEC may result in treatment failures for infections that were previously easily treatable with first-line antibiotic administration. In the current study, only a single ESBL-positive ETEC isolate was identified from isolates archived in 2001-2009 from Nepal. However, over 30% of the ETEC isolates collected after 2013 phenotypically display ESBL production. This suggests that ESBL resistance mechanisms in Nepal have expanded relatively recently into community-associated bacterial pathogens, and represents a new threat to both the community as well as international travelers in the region. Notably, international travelers have been shown to be a potential vector for the spread of resistant bacterial pathogens through acquisition during travel, contributing to the global dissemination of MDR enteric pathogens [40]. The combination of established community prevalence and additional spread by infected travelers may contribute to local and global dissemination of ESBL-positive ETEC isolates.
ETEC virulence factors, such as enterotoxins and CFs, are of particular interest as potential targets for infection intervention as vaccine or therapeutic targets [15,18,21]. The expression of enterotoxins and CFs are essential for ETEC colonization of the small intestine and resulting pathogenesis. Blocking bacterial cell adhesion to the small intestine and/or neutralizing enterotoxins results in a limited ETEC infection and significantly reduces infection morbidity [15,18,21]. The currently reported prevalence of the enterotoxins LT, STh, and STp in this study were comparative to previous studies reviewed and analyzed by Isidean, et al. [18]. The most prevalent CFs in the current study were CS21, CS6, CS3, and CS2, while the most prevalent CFs in previously conducted studies reviewed by Isidean et al were CFA/1, CS6, and CS21 [18]. Notably, CFA/1 was detected at a lower rate in the current study than previously reported (6.8 vs 20%), indicating CFA/1 may not be a priority target for ETEC interventions in South Asia [18]. Overall, the current study suggests vaccines and therapeutics targeting the enterotoxins LT and STh, and the CF's CS21, CS6, CS2, and CS3 would likely be most efficacious against ETEC infections in South Asia. This study demonstrated that these virulence factors are also prevalent in the ESBL-positive ETEC population, a group that is important to target for vaccine efficacy as MDR ETEC continues to emerge and spread throughout the region. It is important that ongoing studies characterizing virulence factors in MDR ETEC populations are performed to determine relevant intervention strategies against currently circulating strains.
Many ESBL genes have been reported as actively circulating in South Asia in hospital-and community-associated infections, including bla CTX-M group 1 and group 9, bla PER , bla VEB , bla GES , and bla SHV [41]. Within the gene family bla CTX-M, bla CTX-M-15 and bla CTX-M-14 are the most common type of ESBL found worldwide [27,42]. Aligning with the global findings of gene prevalence, previous studies conducted in Nepal have identified bla CTX-M-15 as prevalent in hospital-associated E. coli infections, while bla TEM and bla SHV were also detected at a lower frequencies [43,44]. In the current study, bla CTX-M-15 was detected in the majority of ESBL-positive ETEC isolates. The genes bla SHV-12 , bla CTX-M-14 , bla TEM-1 , and bla CMY-2 were also identified at lower rates, and multiple resistance genes were detected in 40% of the characterized isolates. Importantly, ESBL resistance mechanisms are often located on mobile plasmids [42,45,46]. As such, enteric ETEC strains that are ESBL-positive may potentially serve as a community reservoir in Nepal for the dissemination of ESBL resistance Limitations of the current study include small sample sizes per year and the lack of clinical patient information. These limitations serve to highlight the need for additional long-term, comprehensive enteric pathogen surveillance to track ongoing MDR emergence and spread in community settings. ESBL-positive E. coli often have limited treatment options, such as carbapenem antibiotics. However, the efficacy of these treatment options are compromised by increased prevalence of carbapenemase resistance mechanisms, such as bla NDM-1 , bla KPC , and bla IMP . No carbapenemase-producing ETEC were identified in the current study, but continued surveillance is imperative for tracking if, or more likely when, community-associated MDR ETEC isolates acquire carbapenem-resistance mechanisms in Nepal.

Conclusions
MDR Gram-negative bacterial species pose a dangerous threat to global public health, including community-associated MDR enteric infections that are increasingly prevalent. The current study demonstrated that ESBL resistance mechanisms have spread to community-associated pathogens in Nepal, with over 30% of ETEC isolates collected after 2013 demonstrating ESBL production. Both retrospective and recent surveillance studies of ETEC isolates identified from acute diarrheal clinical samples are vital for tracking the community presence of emerging antibiotic resistance mechanisms. Consistent, long-term surveillance may lead to better informed treatment options, alternative therapeutic development, and community-based interventions to mitigate the spread of resistant bacterial strains beyond hospital environments.