Multiple antibiotic resistances among Shiga toxin producing Escherichia coli O157 in feces of dairy cattle farms in Eastern Cape of South Africa

Background Shiga toxin–producing Escherichia coli (STEC) O157:H7 is a well-recognized cause of bloody diarrhea and hemolytic-uremic syndrome (HUS). The ability of STEC strains to cause human disease is due to the production of Shiga toxins. The objectives of this study were to determinate the prevalence, serotypes, antibiotic susceptibility patterns and the genetic capability for Shiga toxin production in Escherichia coli (STEC) strains isolated from dairy cattle farms in two rural communities in the Eastern Cape Province of South Africa. Methods Fecal samples were collected between March and May 2014, from individual cattle (n = 400) in two commercial dairy farms having 800 and 120 cattle each. Three hundred presumptive isolates obtained were subjected to polymerase chain reactions (PCR) for identification of O157 serogroup and Shiga toxin producing genes (stx1, stx2) on genomic DNA extracted by boiling method. Susceptibility of the isolates to 17 antibiotics was carried out in vitro by the standardized agar disc-diffusion method. Results Based on direct PCR detection, 95 (31.7 %) isolates were identified as O157 serogroup. The genetic repertoire for Shiga toxin production was present in 84 (88.42 %) isolates distributed as stx1 (37), stx2 (38) and stx1/2 (9) respectively while 11 of the isolates did not harbor Shiga toxin producing genes. Multiple antibiotic resistances were observed among the isolates and genetic profiling of resistance genes identified blaampC 90 %, blaCMY 70 %, blaCTX-M 65 %, blaTEM 27 % and tetA 70 % and strA 80 % genes among the antimicrobial resistance determinants examined. Conclusion We conclude that dairy cattle farms in the Eastern Cape Province are potential reservoirs of antibiotic resistance determinants in the province.


Background
Escherichia coli is an important pathogen in cattle, medicine and public health [1], and Shiga toxin-producing strain (STEC) have emerged as important food-borne pathogens, especially serotype O157:H7. Human diseases caused by this serotype that produces STEC ranges from mild diarrhoea to haemorrhagic colitis and haemolytic uraemic syndrome (HUS) and typically it affects children, the elderly, and immunocompromised patients [2]. Healthy domestic ruminants such as cattle, sheep, and goats can harbor STEC and E. coli O157:H7 in their faeces and are thus natural reservoirs of these pathogens [3][4][5]. The pathogenicity of STEC resides in a number of virulence factors, including Shiga toxins (Stx1 and Stx2), intimin, enterohaemolysin, and the STEC autoagglutinating adhesin (Saa) [6]. Shiga toxin-producing (STEC) and enteropathogenic Escherichia coli (EPEC) represent two of the six different categories of diarrheagenic E. coli that can cause disease in humans [1]. STEC, which is defined by the production of two Shiga toxins, Stx1 and/or Stx2, is a zoonotic pathogen that is a major cause of diarrhea worldwide. Stx2 is more closely related to these diseases than stx1 [7]. There are three Stx1 subtypes (Stx1a, Stx1c, and Stx1d) and seven Stx2 subtypes (Stx2a, Stx2b, Stx2c, Stx2d, Stx2e, Stx2f, and Stx2g) according to the subtyping nomenclature proposal put forth at the 7th International Symposium on Shiga Toxin (Verocytotoxin)-Producing Escherichia coli Infection, held in Buenos Aires, in 2009.
The use of antimicrobial in animal feeds as growth promoters is common worldwide. Across the globe, a variety of antimicrobial agents are available for therapeutic use or as growth promotion in animals. Many studies have supported the claim that with the increased use of antimicrobial agents in animals and humans, an increased prevalence of resistant strains may be selected as a direct consequence of the antimicrobial use [8,9]. Humans, via the food chain, ingest a lot of bacteria originating from food-producing animals, which have been recognized as major reservoirs of E. coli habouring CTX-M β-lactamase an enzyme that confers resistance to β-lactam antibiotics [23]. Sasaki [10] have reported a high prevalence of CTX-M β-lactamase encoding gene among Enterobacteriaceae in stool specimens from healthy asymptomatic volunteers in a rural community in Thailand. Concerns exist about the potential spread of the β-lactam-CTX-M genes from food animal products to humans through the food chain. CTX-M βlactamase genes have been reported in E. coli from various food-producing animals worldwide raising a potential threat to public health [11][12][13][14][15][16]. Giving the frequent occurrence of O157 STEC as foodborne pathogen in North America and some parts of the western world, there is a clear need to gather data on the prevalence and distribution of STEC producing E.coli and the antibiotic resistance profiles of isolates of this organism recovered from faecal samples from commercial dairy cattle farms in the Eastern Cape of South Africa where pastoral farming is a major source of income for many families. This study therefore aimed at characterizing E.coli O157 isolates from fecal samples of two dairy cattle farms in the Eastern Cape Province of South Africa. Ability of isolates to produce Shiga toxin and their antibiotic susceptibility patterns as well as presence of some resistance determinants were screened by molecular approaches.

Ethical clearance
Ethical clearance was obtained from the University of Fort Hare ethics committee prior to sample collection and permission was sought from farmers from whose farms samples were collected.

Study population and sampling
Details on the study population and sampling procedures are as follows. Briefly, samples were collected from commercial dairy cattle farms in Nkonkobe District of Eastern Cape Province, South Africa. A total of 400 samples from two commercial dairy farms were collected for the study. Rectal fecal grab samples of approximately 10 g were collected from individual cattle using sterile gloves into appropriate capped containers. After collection, samples were shipped on ice to the University of Fort Hare Microbiology laboratory for immediate processing.

Preliminary sample processing
Approximately 1 g of each fecal sample was mixed in 9 ml of Trypticase soya broth (TSB) with 20 mg/L novobiocine and incubated for 6-8 h at 37°C. This was streaked out onto sorbitol MacConkey agar (SMAC) supplemented with 1 mg/L potassium tellurite and incubated for 18-24 h at 37°C. A pale colony each (sorbitol nonfermenters) was picked as presumptive E. coli O157 per sample. The pure colonies were each inoculated into separate TSB and incubated for 24 h at 37°C from which glycerol stock was made and then stored at −80°C for further analyses.

DNA extraction
Bacterial DNA was prepared as previously described by Bai [17]; briefly, bacterial culture from glycerol stock was resuscitated by an overnight growth in trypticase soya broth (TSB) (Oxoid Ltd, London, UK) at 37°C with slight agitation. From this culture, 2 ml was centrifuged for 5 min at 14,000 rpm and the pellet was washed with normal saline (0.85 % NaCl). After the addition of 150 μl of rapid lysis buffer (100 mM NaCl, 10 mM Tris-HCl pH8.3, 1 mM EDTA pH9.0; 1 % Triton X-100), the suspension was votexed, boiled for 15 min, centrifuged at 10,000 rpm, supernatants collected in a DNase free Eppendorf tube and stored at −20°C. These were then used as templates in all the polymerase chain reactions (PCRs) that were performed in this study.

Molecular serotyping and virulence typing
Molecular serotyping using the O-unit flippase gene (wzx) was performed: primers used for detection of O157 strains are shown in Table 1. E. coli O157:H7 ATCC 35150 served as the positive control. Specific primers were used to detect the presence of virulence genes encoding the Shiga-toxins (stx1 and stx2) as previously described by [17,18]. Amplification was performed using 25 μl of PCR mix containing 5 μl of bacterial DNA purified as described above, 12 μl of 2X Dream Taq Master Mix (Thermo Scientific), 10pmol of both forward and reverse primers and 6 μl of water of PCR grade. The conditions for PCR amplification performed in a thermal cycler (BioRad) were 94°C for 3 min, followed by 35 cycles of 93°C for 60 s, either 55°C for 60 s and 72°C for 60 s. The final cycle was followed by an extension step at 72°C for 7 min. The amplified products were visualized by standard gel electrophoresis using 5 μl of the PCR product on 2 % agarose gels in 0.5X TBE buffer (0.1 M Tris, 0.1 M boric acid and 0.002 M NaEDTA). Gels were stained using ethidiumbromide (1 mg/ml) and photographed under UV light in a transilluminator (ALLIANCE 4.7).

PCR profiling of resistance genes
Template DNA was prepared as previously stated above. The β-lactamase genes bla TEM , bla SHV, bla CMY-2 , bla-CTX-M , bla-CTX-M1 , bla-CTX-M9 and bla-ampC and two of the genes responsible for resistance to streptomycin (strA) and tetracycline (tetA) respectively were tested using specific primers in Table 2 Thong et al., 2010). PCR was performed in a 25-μl mixture of 12 μl of 2X Dream Taq Master Mix (Thermo Scientific, Pittsburgh, PA, USA), 10 pmol of each forward and reverse primer and 6 μl water of PCR grade. The PCR mixture was subjected to a 3-min denaturation step at 94°C, followed by 35 cycles of 45 s at 94°C, 45 s at 55°C/ 57°C/50°C depending on the primer set, and 60 s at 72°C , and a final elongation step of 7 min at 72°C. PCR

Results
Serotyping PCR-based molecular serotyping using primers designed for detection of O157 group identified 95 positive isolates as O157 serotype (gel not shown) out of the 320 presumptive isolates.

PCR-based detection of virulence genes
The 95 molecularly confirmed E.coli O157 isolates were analyzed by PCR for their Shiga toxin producing capabilities. Table 3 shows PCR results of the different virulence genes for 84 (88.45 %) isolates detected among the 95 confirmed O157 isolates. The remaining 11 (11.55 %) though belonging to O157 serogroup, did not harbor Shiga toxin genes and were therefore regarded as non-STEC strains. Distribution of virulence genes among the 84 STEC strains showed that 37 (44 %) isolates possessed the stx1 gene as shown in Fig. 1, 38 Table 4 below. All isolates showed reduced susceptibility to several antimicrobials agents in the panel.

PCR profiling of resistance genes Prevalence of AMR genes
Results of genetic profiling of the observed phenotypic resistances among the isolates showed predominance of bla ampC 90 %, bla CMY 70 %, bla CTX-M 65 %, and bla TEM 27 %, among the isolates that were resistance to ampicillin, amoxicillin/clavulanate, cephalothin, cefuroxime, ceftazidime. PCR amplification of the bla SHV , bla VEB , bla CTX-Mgroup1 and bla CTX-Mgroup9 did not yield any amplicon. The tet(A) and strA resistance genes were amplified from 70 % and 80 % respectively of the isolates that were phenotypically resistant to tetracycline, oxytetracycline and streptomycin.
Altogether, AMR genes were detected in all E. coli isolates. The most frequent resistance genes were tet(A), str(A), bla-ampC, bla-CMY-I (Table 5). With a very few exceptions, susceptibility test results were consistent with genotyping results. Gel photographs of some of the amplified resistance genes are shown in Figs. 2, 3 and 4. Overall, AMR genes were identified in 92 % of the O157 isolates recovered from studied samples. Resistance genes detected from isolates as shown in Table 5.

Discussions
The prevalence of STEC O157 serogroup in fecal samples collected from commercial dairy cattle was investigated. Escherichia coli O157:H7 (O157) is the Shiga toxin-producing E. coli (STEC) serotype most frequently isolated and most often associated with hemolytic uremic syndrome (HUS) in the United States [2]. According to the U.S. Centers for Disease Control and Prevention, an estimated 265,000 STEC infections occur each year in the United States. STEC O157 causes about 36 % of these infections and non-O157 STEC cause the rest [2]. STEC inhabits in the guts of ruminant animals, including cattle, goats, sheep, deer, and elk [2]. The major source for human illnesses is cattle and around 5-10 % of those who are diagnosed with STEC infection develop a potentially life-threatening complication known as hemolytic  uremic syndrome (HUS) with young children and the elderly more likely to develop severe illness and hemolytic uremic syndrome (HUS) than others. In this study, a total of 400 fecal samples were collected from two commercial dairy farms in the Eastern Cape Province of South Africa. These samples were analyzed for the presence of O157 E.coli strain. A total of 95 isolates confirmed by PCR targeting the O-unit flippase gene (wzx) were delineated to be O157 isolates. Results of the determination for the presence of Shiga toxin encoding gene (stx1 and stx2) among the 95 isolates showed that 35 (36.84 %) harbored the stx1 gene, 26 (27.4 %) were positive for stx2 while 9 (9.5 %) harbored both stx1 and stx2 genes. Twenty five (26.3 %) of the isolates were commensals as no Shiga toxin genes were detected in them.
According to Gyles [3], ruminants especially cattle and sheep are the major reservoir of STEC and individual animal could carry more than one serogroups of STEC. During processing, meat derived from infected animals may become contaminated by STEC contained fecal materials if they are mistakenly mixed with it. Barlow and Mellor [27] had reported the presence of STEC in fecal samples of cattle from Australia where a prevalence of 10 % was observed with E.coli O157 accounting for 1.7 % of all the isolates. It is possible for fresh farm produce to be contaminated with STEC where irrigation water or soil treated with farm effluent or manure is used in growing them. A prevalence as high as 33.5 % of STEC in bulk milk has been reported internationally [28]. It is also possible for STEC to survive for a long time in soil applied with manure from cattle and sheep. The possibility of water sources being contaminated by STEC is also very high as fecal materials of animal origin could be washed through storm drains into fresh waterbodies thus posing health challenges to the people who depend on such waterbodies for several uses. WHO [29] had reported waterborne transmission of STEC in both  There are increasing concerns about the use of antimicrobial products in food-producing animals and focus has been on human food safety because foods of animal origin are sometimes identified as the vehicles of food borne disease as well as resistant food borne pathogens carrying resistant genetic materials. We profiled the antimicrobial susceptibility of E.coli O157 isolates recovered from commercial dairy cattle that are constantly receiving (tylosin, advocin, ampicillin, tetracycline) antimicrobial agents. We observed a very high level of multiple antimicrobial resistances among the isolates and the most common resistance was to tetracycline. This is not surprising since tetracycline is often used as a first-line antimicrobial in disease prevention and growth promotion in food animals and its widespread use has likely contributed to high rates of resistance [30]. The frequency of tetracycline resistance among the E. coli isolates from the farms that we investigated was 96.8 %, which is within the range of values described in previous reports (68 to 93 %) [31,32]. Genetic profiling of the resistance determinants showed that the resistances were encoded by bla ampC , bla CMY , bla CTXM, bla TEM genes for the ESLBs (extended spectrum β-lactamases), while tetA and strA genes were responsible for tetracycline and streptomycin respectively. High prevalence of CTX-M βlactamase-encoding genes in Enterobacteriaceae has been reported in stool specimens from healthy asymptomatic volunteers in a rural community in Thailand [11]. The fact that bacteria which infect animals could also establish infections in humans poses concerns about the potential spread of the bla-CTX-M genes from food animal products to humans through the food chain. CTX-M β-lactamase has been increasingly reported in E. coli from various food-producing animals worldwide raising a potential threat to public health with the earliest account of CTX-M β-lactamase of food animal origin from Spain, where a CTX-M-14-producing E. coli was isolated from healthy chickens [33]. Since then, E.coli harboring CTX-M β-lactamase encoding gene has been reported from healthy cattle from Japan [34] and Hong Kong [35], and from sick or healthy cattle from France [36,37] and in the United States [38,39]. Similarly, E.coli strains isolated from pigs that have the genetic repertoire to produce CTX-M-β-lactamase have also been reported from Hong Kong [40], China [41], Spain [42], and France [36]. In this study, genetic resistance determinants were however not amplified from some of the isolates and this could be attributed to the fact that the genetic elements targeted in our PCR profiling were not responsible for the observed phenotypic resistances as there are numerous arrays of genes that encodes for resistances to the drugs. Besides, there are many resistance mechanisms like the efflux pump, intrinsic resistance, innate resistant or acquire resistance to one or few classes of antimicrobial agents. Our findings also showed that most of the isolates were susceptible to imipenem, amikacin, kanamycin, the quinolones (norfloxacin, ciprofloxacin and enrofloxxacin) and gentamicin. This finding is curious as regards the susceptibility to the quinolones because the farms that were sampled uses advocin (danofloxacin) which is a synthetic fluoroquinolone in the treatment of respiratory disease in chickens, cattle and pigs and ought to have selected for other quinolone Fig. 3 Agarose gel image of amplicons obtained from PCR with primers designed for bla-ampC resistance gene of E. coli isolates recovered from this study. Lane 1 is molecular size markers (100 bp), lane 2 is negative control (PCR mix without DNA) while lanes 3 to 18 are bla-ampC (198 bp) gene from O157 strains isolated in this study Fig. 2 Agarose gel images of amplicons obtained from PCR with primers designed for strA resistance gene of E. coli isolates recovered from this study. Lane 1 is molecular size markers (100 bp), lane 2 is negative control (PCR mix without DNA) while lanes 3 to 18 are strA (548 bp) gene from O157 strains isolated in this study resistances as they have similar structure and mode of action.
To understand the bases of high tetracycline resistance among our isolates, we screened for tet determinants. The predominance of the tet(A) efflux gene observed in our study is similar to that previously documented in coliforms (73 %) of human and animal origins by Marshall [43]. In this study, we observed also a very high resistance to streptomycin among the isolates which is in partial agreement with the reported 53 % resistance of E.coli O157 isolates from feedlots by Rao [44]. Also, Sawant [45] reported a prevalence of 93 % of tetracycline resistance in E.coli of dairy cattle with tet(B) accounting for the resistance determinant in 93 % of their isolates and this is similar to our finding though mediated by different variants of the gene.
The fact that bacteria from animals spread to the food products during slaughter and processing has been extensively documented [46][47][48][49]. The detection of E.coli resistant to antibiotic growth promoters (AGPs) in food products derived from animals where AGPs have been used therefore comes as no surprise. Resistant bacteria and active antibiotics, or active metabolites of antibiotics can also spread on farmland with manure. In this study, we isolated toxigenic O157 E.coli strains carrying S tx1 and stx2 genes that were also multi-resistant to many antibiotics some of which are medically important in human medicine. The possibility of these antibiotic resistant strains being shed into the environment and the eventual transmission of the resistance determinants to nonpathogenic environmental bacteria is high. Such transfer of resistance determinants could fuel the spread of antibiotic resistant bacteria (ARB) that could have grave implications on the health of humans and animals thus increasing the burden of disease in the community. There is therefore need for urgent policy formulations on the prudent use of antimicrobials in both human and veterinary medicine as failure in this regards could spell doom in the nearest future.

Conclusions
In conclusion, this study demonstrated that antimicrobial resistant (AMR) determinants were present in dairy cattle exposed to veterinary antimicrobials for both therapeutic and as growth promoting agents. The overall resistance rate was high and the isolates have the genetic repertoires to survive antimicrobial pressure. Though genetic capacity for Shiga toxins production have been reported among E.coli nonO157 serogroups, we did not profile for any other serogroups apart from O157 in the fecal samples analyzed. The observed phenotypic multiple resistances among the isolates were also genetically confirmed. There were high β-lactam resistances among the isolates and this poses health implications as these are the drugs of choice for the management of numerous Gram negative infections. The results of this study suggest that agricultural activities, specifically antimicrobial use may have a significant impact on AMR evolution in general. More studies with larger sample sizes and more precise AMR genes typing by DNA sequencing and molecular typing of bacterial strains are needed to further throw more light in this regard.