Genotyping of virulent Escherichia coli obtained from poultry and poultry farm workers using enterobacterial repetitive intergenic consensus-polymerase chain reaction

Aim: The aim of this study was to characterize virulent Escherichia coli isolated from different poultry species and poultry farm workers using enterobacterial repetitive intergenic consensus-polymerase chain reaction (ERIC-PCR) genotyping. Materials and Methods: Fecal swabs from different poultry species (n=150) and poultry farm workers (n=15) were analyzed for E. coli and screened for virulence genes (stx1, stx2, eaeA, and hlyA) by multiplex PCR. Virulent E. coli was serotyped based on their “O” antigen and then genotyped using ERIC-PCR. Results: A total of 134 E. coli isolates (122/150 from poultry and 12/15 from farm workers) were recovered. Virulence genes were detected in a total of 12 isolates. Serological typing of the 12 virulent E. coli revealed nine different serotypes (O2, O49, O60, O63, O83, O101, O120, UT, and Rough). ERIC-PCR genotyping allowed discrimination of 12 virulent E. coli isolates into 11 ERIC-PCR genotypes. The numerical index of discrimination was 0.999. Conclusion: Our findings provide information about the wide genetic diversity and discrimination of virulent E. coli in apparently healthy poultry and poultry farm workers of Andhra Pradesh (India) based on their genotype.

Bacterial genomes contain imperfect palindromic repeat sequences such as enterobacterial repetitive intergenic consensus (ERIC) sequence [9]. The distance between these repetitive sequences and frequency of repeated sequences vary with individual bacterial strains [10]. ERIC-PCR genotyping technique yields strain-specific unique multiband patterns obtained by the amplification of conserved ERIC sequences, thus discriminating individual bacterial strains [11]. ERIC-PCR technique can be effectively used to discriminate E. coli strains intra-serotypically into different genotypes [10].
Perusal of the available literature revealed the paucity of information regarding the intra-serotypic genetic diversity of virulent E. coli of poultry origin in India. Hence, the present research was undertaken to assess the genetic diversity of virulent E. coli isolated from poultry and poultry farm workers in Andhra Pradesh, India.

Ethical approval
Ethical approval is not necessary to pursue this study. However, samples were collected without harming birds. Informed consent was obtained from farm workers.

Sample collection and isolation of E. coli
Fecal swabs (n=150) from different poultry species, namely, chicken, ducks, quails, turkey, fancy birds (each 30), and poultry farm workers (n=15) were collected from three different poultry farms in Andhra Pradesh, India. Isolation and identification were done using Trypticase Soy Broth (Hi-Media), eosin methylene blue (EMB, HiMedia) agar, and MacConkey agar medium. The isolates were confirmed to be E. coli by standard biochemical tests [12] as well as by the PCR amplification of the E16S gene of E. coli using the primer pair (F, 5'-ATC AAC CGA GAT TCC CCC AGT-3' and R, 5'-TCA CTA TCG GTC AGT CAG GAG-3') [13]. Whole-cell DNA extraction was carried out by boiling and snap chilling method [14].

Detection of virulence genes by multiplex PCR
A multiplex (m)-PCR for the detection of stx 1 , stx 2 , eaeA, and hlyA genes in E. coli was carried out as described by Paton and Paton [1]. Primers used and their expected amplicon sizes are summarized in Table-1. The reaction was carried out in 10 µl volume containing 5.0 µl of PCR Master Mix (Takara Bio Inc., Japan), 0.22 µl of each forward and reverse primer (10.0 pmol/µl), 2.24 µl of nuclease-free water, and 1.0 µl of DNA template (75 ng/µl). Thermal cycling conditions were as follows -35 cycles, each consisting denaturation at 95°C for 60 s, annealing at 65°C for 120 s, and elongation at 72°C for 90 s (for first 10 cycles), decrementing annealing temperature to 60°C by cycle 15; denaturation at 95°C for 60 s, annealing at 60°C 120 s, and elongation at 72°C 90 s (from cycle 15-25), and incrementing elongation time from 90 s to 150 s from cycles 26 to 35.

Serotyping of virulent E. coli isolates
Serotyping of E. coli isolates positive for virulence genes was done at National Salmonella and Escherichia coli Center, Central Research Institute, Kasauli (Himachal Pradesh, India) on the basis of their "O" antigen.
Reaction was carried out in 25 µl volume using 12 μl of Master Mix (Takara Bio Inc.), 1.5 µl of DNA template (100 ng/μl), 1.0 μl of each primer (20 pmol/μl), and 9.5 μl of nuclease-free water. The DNA amplifications were performed in the Eppendorf thermal cycler (USA) with an initial denaturation at 95°C for 7 min, followed by 30 cycles of 30 s of denaturation at 94°C, 1 min annealing at 50°C, 8 min of elongation at 65°C, and 16 min of final elongation at 65°C. Standardization of the ERIC-PCR reactions was done using the DNA from reference strain E. coli (MTCC 1610).

Analysis of DNA fingerprinting patterns
The PCR products were resolved by 2% agarose gel electrophoresis under 110V for 2 h [16]. Banding patterns were photographed using BIO-RAD Gel Documentation system (USA). The DNA fingerprints obtained were analyzed both by visual inspection and Image Lab Software (BIO-RAD). The position of bands was compared using 100 bp and 1 kb DNA ladder (Genei ™ , Bengaluru). Binary matrix was constructed based on the presence (1) or absence (0) of a particular band in the given strain. Dendrogram was constructed by "branch-and-bound method" using dollop program of PHYLIP software, version 3.6. Clusters were considered at a 70% similarity cutoff value. The similarity of banding patterns was determined using the Pearson's correlation coefficient. The numerical index of discrimination was calculated using Simpson's index of diversity, D = 1-1/N (N-1) ∑ n j (n j -1) , where D corresponds to the discriminatory power, N corresponds to the total number of strains, and n j corresponds to the number of strains belonging to the j th type [17].

Isolation and identification of E. coli
Of the 150 poultry fecal swab samples analyzed, a total of 122 (81.3%) E. coli isolates were recovered (26/30 from chicken, 24/30 ducks, 23/30 quails, 24/30 turkey, and 25/30 fancy birds). Of the 15 human fecal swab samples analyzed, a total of 12 (80%) E. coli were recovered. All the E. coli isolates were able to produce green metallic-sheen colonies on EMB agar, lactose-fermenting pink colonies on MacConkey agar, and amplified E16S gene giving 231 bp amplicon. The present results were in accordance with the findings

Virulence genes characterization
The multiplex-PCR assay revealed the presence of one or more virulence genes in a total of 12 (chicken, 3; ducks, 2; turkey, 2; fancy birds, 2; quails, 1; and humans, 2) isolates, with an overall frequency of 12% (10/122 E. coli) and 13% (2/15 E. coli) occurrence of virulence genes in poultry and poultry farm workers, respectively. All these 12 isolates were positive for both stx1 and eaeA genes, whereas hlyA gene was detected in eight isolates and stx2 gene in four isolates (Table-2). The previous studies on avian pathogenic E. coli by Parreira and Gyles [19] and Dutta et al. [7] revealed association of STEC with 53% and 23.8% of avian pathogenic E. coli isolates. However, the absence of STEC is also reported in another study where E. coli was recovered from apparently healthy birds [20]. In the present study, all the 12 E. coli isolates that were positive for at least one virulence gene carried both stx1 and eaeA genes, which is in agreement with the findings of other workers [7,19,20], who reported a higher percentage (>50%) of pathogenic E. coli isolates carrying stx1 and eaeA genes. However, in contrast to our study, Wani et al. [18] reported only 2.4% frequency of eaeA gene in E. coli from poultry.

Serological typing
Serotyping of 12 virulent E. coli isolates revealed nine different serotypes, of which O2, O60 and UT were predominant (two isolates each), followed by O49, O63, O83, O101, O120, and Rough (one isolate each). Detection of O2, O60, O120, and R serotypes among E. coli of avian origin in the present study was in accordance with a study from Jammu and Kashmir [18]. However, avian E. coli serotypes reported in other studies from India [6,7] were not detected in the present study, indicating the variable distribution of different serotypes of E. coli in different geographical regions of India.
Several investigations have used ERIC-PCR as a highly sensitive genomic fingerprinting tool for detecting and determining different strains of E. coli from diverse sources [11,15,21]. Soltani et al. recognized 65 different avian E. coli strains from 95 samples with 232 to 2690 bp bands on gel electrophoresis and confirmed that ERIC-PCR as a good technique for bacterial genotyping among cheap and simple molecular tools [22]. In determining genetic profiles of E. coli, ERIC-PCR technique was showed to have more discriminatory power than 16S rRNA gene analysis [23].
In the present study, ERIC-PCR profile dendrogram analysis discriminated virulent E. coli isolates into three major clusters (C 1 to C 3 ) for 70% similarity cutoff, namely, C 1 with four isolates (chicken, 2; fancy birds, 2), C 2 with two isolates (chicken, 1; quail, 1), and C 3 with five isolates (turkey, 2; duck, 2; and reference strain E. coli MTCC 1610) (Figure-1). The two human E. coli isolates were found to be unclustered (UC) with other isolates. Cluster analysis indicated wide intra-serotypic genetic diversity among the isolates affiliated to the same "O" serotype, as evidenced by their grouping under different clusters in the dendrogram. For example, isolates of O120 serotype, namely, Ch2 and Ch3 were grouped under clusters C 1 and C 2 , respectively, with different ERIC-PCR genotypes (E2 and E3) ( Table-2 and Figure-1). The numerical index of discrimination calculated for 12 isolates, using Simpson's index of diversity was 0.997, which corroborate with earlier studies [15,21], where ERIC-PCR was shown to have highly significant discriminatory power in characterizing the genetic diversity of E. coli strains.

Conclusion
The results presented in this paper and past studies emphasize the utility of the combination of serotyping with PCR-based genotyping tools in the epidemiological investigations of virulent E. coli strains. This is the first study in India to elucidate the intra-serotypic genetic diversity of virulent E. coli isolates from different poultry species as well as poultry farm workers, and ERIC-PCR was proved to be quick, sharp, and cost-effective fingerprint method for effective discrimination of E. coli strains based on their genotype.

Authors' Contributions
MSS and NMS prepared the study design and carried out the research under the supervision of TSR and MM. NMS collected the samples and executed the isolation. MSS and MM conducted the molecular part of the research. The manuscript was drafted and revised by MSS and NMS under the guidance of TSR and MM. All authors read and approved the final manuscript.