Virulence factors, serogroups and antimicrobial resistance properties of Escherichia coli strains in fermented dairy products

Background From a clinical perspective, it is essential to know the microbial safety of fermented dairy products. Doogh and kashk are fermented dairies. These products are used by millions of people but their microbial qualities are unknown. Shiga toxin producing Escherichia coli (STEC) is one of the most commonly detected pathogens in the cases of food poisoning and food-borne illnesses. The present investigation was carried out in order to study the molecular characterization and antimicrobial resistance properties of STEC strains isolated from fermented dairy products. Methods Six hundred fermented dairy samples were collected and immediately transferred to the laboratory. All samples were cultured immediately and those that were E. coli-positive were analyzed for the presence of O157 , O26, O103, O111, O145, O45, O91, O113, O121 and O128 STEC serogroups, tetA, tetB, blaSHV, CITM, cmlA, cat1, aadA1, dfrA1, qnr, aac (3)-IV, sul1 and ereA antibiotic resistance genes and stx1, stx2, eaeA, ehly, cnf1, cnf2, iutA, cdtB, papA, traT, sfaS and fyuA virulence factors using PCR. Antimicrobial susceptibility testing was performed also using disk diffusion methodology with Mueller–Hinton agar. Results Fifty out of 600 (8.33%) dairy samples harbored E. coli. In addition, yoghurt was the most commonly contaminated dairy. O157 (26%) and O26 (12%) were the most commonly detected serogroups. A significant difference was found between the frequency of Attaching and Effacing E. coli and Enterohaemorrhagic E. coli (P <0.05). Stx1 (44%), eae (36%), papA (32%) stx2 (30%), and ehly (28%) were the most commonly detected virulence factors. The genes encode resistance against tetracycline (tetA and tetB) (76% and 70%, respectively), cephalothin (blaSHV) (38%), ampicillin (CITM) (36%) and gentamicin (aac (3)-IV) (32%) were the most commonly detected. High resistance levels to tetracycline (84%), penicillin (46%), ampicillin (38%) and streptomycin (36%) were observed. Conclusion Fermented dairy products can easily become contaminated by antibiotic resistant STEC strains. Our findings should raise awareness about antibiotic resistance in Iran. Clinicians should exercise caution when prescribing antibiotics, especially in veterinary treatments.

from these dairy products in their routine meal. Unfortunately, adequate heat and time were not performed in their traditionally producing.
A broad spectrum of Gram-negative bacterial species has been shown to produce Cytolethal Distending Toxin (CDT) [14], and three closely linked genes (cdtA, cdtB, and cdtC) are required for toxin expression [14].
Diseases caused by E. coli often require antimicrobial therapy; however, antibiotic-resistant strains of this bacterium cause longer and more severe illnesses than their antibiotic-susceptible counterparts. Several studies have shown that antibiotic resistance in E. coli has increased over time [2,16,[18][19][20][21]. In keeping with this, an epidemiological investigation in Iran revealed that STEC strains were the most commonly detected strains in patients with diarrhea and that there was a high incidence of resistance (85-100%) to commonly used antibiotics [22][23][24][25].
There were no data on the distribution of serogroups, virulence genes and the antimicrobial resistance properties of E. coli strains isolated from yoghurt, kashk and doogh. Therefore, the aim of the present study was to characterize E. coli strains isolated from yoghurt, kashk and doogh at the molecule level and investigate their susceptibility to 14 commonly used antibiotics.

Sampling and Escherichia coli identification
Overall 600 dairy products including 200 yoghurt, 200 doogh and 200 kashk samples were purchased from supermarkets and retailers in various parts of Iran at summer of 2012. All of these dairy products were made traditionally by native people and after collection were kept under refrigeration in plastic bags. Samples were transported under refrigeration (at 4-6°C) in thermal boxes containing ice packs. All samples were diluted in phosphate buffered saline (PBS, Merck, Germany). A 25 g portion of each sample was blended with 225 mL of nutrient broth (Merck, Germany) for 2 min, using a Stomacher lab blender and incubated at 37°C for 24 h. A 1 mL sample of the nutrient broth culture was mixed with 9 mL of MacConkey broth (Merck, Germany) and further incubated at 37°C for 24 h. One loop of each tube was streaked on MacConkey agar (Merck, Germany). Such colonies were confirmed as E. coli using standard biochemical tests (e.g., Indole, Methyl red, Voges-Proskauer and Citrate utilization tests). Colonies were confirmed as E. coli by PCR [26]. E. coli isolates were stored in Tryptic Soy Broth (TSB, Merck, Germany) containing 20% glycerol at −70°C for further characterization.

DNA extraction
Bacterial strains were grown overnight in Trypticase Soy Agar (TSA, Merck, Germany) at 37°C. A single colony was suspended in 100 μL of sterile distilled water. After boiling the suspension for 13 min, the suspension was frozen and centrifuged at 14,000 rpm for 15 min to pellet the cell debris [28]. The supernatant was used as a template for PCR amplification.
To detect serogroups and antibiotic resistance genes in the E. coli isolates, several PCR assays were used. The primer sequences are summarized in Table 1. A DNA thermo-cycler (Eppendorf Mastercycler, Eppendorf-Nethel-Hinz GmbH, Hamburg, Germany) was used in all PCR reactions. The amplified DNA products were electrophoresed on 2% agarose gels at 90 V for 3 h using 1× TBE (0.89 M Tris borate, 0.02 M EDTA, pH 8.3) as the running buffer, then stained with ethidium bromide (10 mg/ml). Gels were visualized using a UV gel documentation system (Uvitech, UK). DNAs of E. coli O157:K88ac:H19, CAPM 5933, O159: H20 and CAPM 6006 strains were used as positive controls and distilled water was used as a negative control.

Statistical analyses
The data were analyzed using SPSS software (Version 17. SPSS Inc, United States) to find any significant correlation between incidences of virulence factors, serogroups and antibiotics resistance properties of E. coli strains isolated from dairy products. Statistical significance was regarded at a P value < 0.05.

Results
All of the dairy samples were tested using culture and PCR techniques. Distribution of E. coli strains in yoghurt, doogh and kashk is shown in Table 2. From 600 dairy samples, 50 (8.33%) were positive for E. coli. We found that yoghurt were the most commonly contaminated dairy (10%). We also found that O157 (26%) and O26 (12%) were the most commonly detected STEC O-serogroups among dairy samples. Distribution of putative virulence genes among E. coli strains isolated from dairy samples is shown in Table 3. Stx1 (44%) had the highest incidence, followed by eae (36%), papA (32%) and stx2 (30%). The Attaching and Effacing E. coli (AEEC) subtype was most commonly detected subtype ( Table 4). All of the EHEC strains of our study harbored all of the stx1, eae and ehly genes ( Table 4).
Dairy samples of the present study had the pH range of 4-6.1. The distribution of antimicrobial resistance genes within the STEC serogroups isolated from dairy samples is shown in Table 5. Genes that encode resistance to tetracycline, tetracycline, cephalothin, ampicillin and gentamicin antibiotics, i.e., tetA (76%), tetB (70%), blaSHV (38%), CITM (36%) and aac (3)-IV (32%) were the most common antibiotic resistance genes in STEC serogroups isolated from dairy samples. Antimicrobial resistance pattern in the STEC serogroup isolates from the dairy samples is shown in Table 6. STEC strains exhibited the highest level of resistance to tetracycline (84%), followed by penicillin (46%), cephalothin (42%), ampicillin (38%) and streptomycin (36%). We found that 50% of tested strains were resistant to more than one antibiotic.

Discussion
Our work has identified the high levels of contamination in yoghurt, doogh and kashk with E. coli strains. The distribution of E. coli strains in yoghurt, doogh and kashk were 10%, 7% and 8%, respectively. There were significant differences (P <0.05) in the incidence of E. coli strains between the yoghurt and doogh. Of the studies that have been conducted in this field [2,[43][44][45][46][47][48], several have shown a low distribution of E. coli strains in yoghurt [45][46][47][48]. However, Rahimi et al. (2011) [47] failed to detection E. coli strains in yogurt. Unfortunately, there were no published data on the presence of E. coli strains in kashk and doogh.
Another important finding relates to the distributions of several bacterial virulence factors in the dairy products of our study. We found statistically significant (P <0.05) differences between the incidence of stx1 and stx2 genes and between the EHEC and AEEC subtypes (P <0.05). The presence of multiple stx1, eaeA, and ehly genes was found in all of the EHEC strains (100%). Also, the presence of multiple stx1 and eaeA genes, stx2 and eaeA genes and finally stx1, stx2 and eaeA genes were found in 45.16%, 32.25% and 22.58% of AEEC strains, respectively. Similar  [55]. The same study found that out of the 77 E. coli isolates, 25 (32.46%) could be classified as Shigatoxigenic based on PCR results (11, 3 and 11 isolates were positive for stx1, stx2, and both stx1 and stx2, respectively) [56]. Virpari et al. (2013) [57] showed that out of 80 E. coli isolates, 25 isolates (31.25%) were positive for stx genes, of which 7 (8.75%) isolates were positive for stx1 gene only, while 12 (15.00%) isolates were positive for stx2 gene and 5 (6.25%) isolates were positive for both stx1 and stx2, 7 isolates (8.75%) were positive for eaeA gene and all isolates were negative for rfb O157 gene.
We found statistically significant (P <0.05) differences between the incidence of genes encode resistance to tetracycline and trimethoprim. There were no significant differences between the incidence of tetA and tetB and cat1 and cmlA genes. We found statistically significant (P <0.05) differences between the incidence of antibiotic resistance against tetracycline and enrofloxacin and also penicillin and trimethoprim.
Bacterial resistance against chloramphenicol and nitrofurantoin 30% and 22%, respectively. Chloramphenicol    (22) 12 (24) and nitrofurantoin are banned antibiotics and the high antibiotic resistances to these drugs detected in our study indicate that irregular and unauthorized use of them may have occurred in Iran. High bacterial resistances against chloramphenicol, nitrofurantoin, tetracycline, ampicillin and gentamicin have been reported previously [2,4,19].

Conclusions
In conclusion, we identified a large number of serogroups, virulence factors and antibiotic resistance genes and resistance to more than one antibiotic in the E. coli strains isolated from yoghurt, doogh and kashk. Our data indicate that O157 and non-O157 STEC strains are predominant in Iranian dairy samples. Our data revealed that the O157 serogroup, the stx1, stx2, eaeA and ehly putative virulence genes, the tetA, tetB, blaSHV and CITM antibiotic resistance genes, and resistance to tetracycline, penicillin, ampicillin and streptomycin were the most commonly detected characteristics of the E. coli strains isolated from yoghurt, doogh and kashk. Careful hygienic supervision on the processing and packaging of dairy products should be performed to reduce load of E. coli contamination. Adequate heat and time should also be performed in the traditionally production of dairy.

Competing interests
The authors declare that they have no competing interests.
Authors' contributions FSD carried out the molecular genetic studies, participated in the primers sequence alignment, design the study and writing and drafted the manuscript. FY and JM carried out the sampling and culture method. YV participated in the statistical analysis. All authors read and approved the final manuscript.