Occurrence of non-O157 Shiga toxin-producing Escherichia coli in healthy cattle and goats and distribution of virulence genes among isolates

Shiga toxin-producing Escherichia coli (STEC) are receiving more attention mainly because they are zoonotic and food borne in nature. The objectives of the present study were to determine the occurrence of non-O157 Shiga toxin producing E. coli in cattle and goats and distribution of the virulence genes in the isolates. The overall isolation rates of non-O157 STEC was 15.27% (7/87) in cattle and 8.14% (22/144) in goats. Four serogroups namely O8, O26, O103 and O128 were detected. Of these serogroups, O8, O26 and O103 were common in both, cattle and goats, whereas O128 was carried by goats only. Using Duopath Verotoxin (DV) test, 52.63% of sero-positive E. coli isolates from cattle were positive for Stx1 shiga toxin whereas none of the isolates was positive for Stx2. Similarly, 42.9% (3/7) of sero-positive E. coli isolates from goats produced Stx1 and 14.3% (1/7) were positive for both, Stx1 and Stx2. The study on virulence genes showed that stx1 was commonly distributed among non-O157 STEC isolates of cattle (57.9%) and goats (57.1%).


INTRODUCTION
Shiga toxin-producing E. coli (STEC), also known as Verotoxin-producing E. coli (VTEC) are recognized as important food-borne pathogens (Armstrong et al., 1996).Shiga toxin-producing E. coli has been associated as a causative agent in several human diseases which may range from mild diarrhea to severe and life-threatening conditions, such as hemolytic-uremic syndrome (Paton and Paton, 2003).Serotype O157:H7 has been named as the most common STEC serotype linked to clinical diseases in humans, however, serotypes other than O157 (non-O157) are increasingly recognized as important cause of human morbidity and mortality (Bettelheim, 2007).Among non-O157 STEC strains, serogroups O8, O26, O91, O103, O111, O128 and O145 are frequently isolated from human beings (Bettelheim, 2007;Fairbrother and Nadeau, 2006;Oporto et al., 2008).
The pathogenic capacity of STEC resides in two potent cytotoxins, Stx1 and Stx2, encoded by stx 1 and stx 2 genes (Paton and Paton, 2003).Studies have shown that E. coli strains possessing stx 2 are potentially more virulent than strains carrying stx 1 (Todd et al., 1999;Tesh et al., 1993).Other virulence factors, such as intimin (eae) and enterohaemolysin (hlyA), may enhance the pathogenicity of organisms but are not required for strains to cause severe disease (Paton and Paton, 2003).
Cattle are considered to be the primary reservoir of both O157 and non-O157 STEC bacteria (Fairbrother and Nadeau, 2006).Cattle harbor STEC without suffering from any pathological symptoms (Blanco et al., 1997).
Wild animals have also been reported to harbor STEC (Fairbrother and Nadeau, 2006;Sanchez et al., 2009).
Among ASEAN countries, there are reports from Vietnam, Thailand and Malaysia on the presence of non-O157 STEC (Apun et al., 2006;Vu-Khac and Cornick, 2008;Suthienkul et al., 1990;Leelaporn et al., 2003).In Malaysia, both O157 and non-O157 STEC have been recovered from beef (Apun et al., 2006;Radu et al., 1998) and patients with diarrhea (Son et al., 1996).However, there is a lack of data on the presence of non-O157 STEC in farm animals.Therefore, the objective of the present study was to determine the occurrence of non-O157 STEC in cattle and goats and to detect the presence of virulence genes in the isolates.

Samples collection
Recto-anal mucosal swab (RAMS) samples were collected from beef cattle (n=144) in seven farms and goats (n= 87) in four farms around Selangor, Negri Sembilan and Johor states in Malaysia during July to December, 2008.All the animals sampled were apparently healthy except in one cattle farm, where the herd suffered mild diarrhea due to parasitic infection.

Reference strains
STEC O111 (ECL 6611, The E. coli Laboratory, Saint-Hyacinthe, Canada) was used as a positive control whereas non-STEC (ATCC 11775) was used as negative control strain.

Enrichment of samples
Each RAMS sample was immediately put into 10 ml of modified EC broth (Oxoid, UK) supplemented with novobiocin 20 mg/L (Sigma-Aldrich, St Louis, USA).All samples were kept in a box containing ice and transported to the Veterinary Public Health Laboratory within a period of 4-6 h.Upon arrival at the laboratory, all samples were immediately incubated aerobically at 37°C for 24 h.

Plating of enriched samples
A loopful of each broth culture was streaked onto sorbitol MacConkey (SMAC) agar (Oxoid, UK) followed by incubation at 37°C for 24 h.From each SMAC agar plate pink colonies resembling with that of E. coli were subcultured onto Chromocult coliform agar (Merck, Germany).From Chromocult coliform agar, presumptive E. coli isolates were identified based on β-Dglucuronidase (GUD) reaction (dark blue to violet colored colonies) and subjected to standard biochemical tests according to Jang et al. (2004).Confirmed E. coli isolates were kept in nutrient agar slant (Merck, Germany) at room temperature until used.

Detection of Shiga toxin production
Colonies were examined for Stx1 and Stx2 production using Duopath ® Verotoxins (DV) test (Merck, Germany) following the manufacturer's instruction.

Detection of virulence genes
Serogroups were further investigated for the presence of stx1, stx2, eaeA and ehlyA genes using multiplex Polymerase chain reaction (mPCR) assay as described by Fagan et al. (1999).Sero-positive E. coli isolates were cultured on SMAC agar and incubated aerobically for 24 h at 37°C.DNA from each isolate was extracted by InstaGene Matrix (Bio-Rad, Germany) according to the manufacturer's instruction.The PCR reaction was performed in a total volume of 50 µl containing 5 µl primer mix (0.2 µM of each primer), 25 µl multiplex PCR master mix (Qiagen, USA), 2 µl of DNA template and 18 µl of DNase/RNase free water.The assay was perform using thermalcycler (Bio-Rad, Germany) set with initial activation temperature as 95°C for 15 min followed by 35 cycles of 95°C for 20 s (denaturation), 58°C for 90 s (annealing), and 72°C for 90 s (extension).Amplified DNA fragments were then electrophoresed using 2% (w/v) agarose at 80 V for 90 min and stained with Gel-Red (Biotium, USA).The products on the gel were visualized using UV illumination (Bio-Rad, Germany).Primer sets used for mPCR are given in Table 1.

Shiga toxin production
Using DV test, 52.63% (10/19) of sero-positive E. coli  2).All three O157 isolates from cattle were found to be negative for Shiga toxin production.

DISCUSSION
In Malaysia, the occurrence of O157 STEC and non-O157 STEC in meat, meat products and hospitalized patients have been reported previously (Apan et al., 2006;Nazmul et al., 2008;Radu et al., 1998Radu et al., , 2001)).However, to our knowledge, this was the first study carried out in farm animals in Malaysia on the prevalence of STEC and non-O157 STEC serogroups O8, O26, O91, O103, O111, O128 and O148.
Although a wide range of animal species have been found to harbor STEC, ruminants are considered as the most common source for transmission of both O157 and non-O157 STEC to humans (Nataro and Kaper, 1998).
Worldwide, the prevalence rate of non-O157 STEC in cattle has been reported as 2.1 to 70.1% (Hussein and Bollinger, 2005) and 0.9 to 75% in goats (Orden et al., 2008;Wani et al., 2006;Zschock et al., 2000).Compared with the present study, lower isolation rates of non-O157 STEC in cattle were found in Spain, 4.1% (Blanco et al., 1997), France, 2.6% (Pradel et al., 2000), Brazil, 1.3% (Irino et al., 2005), Japan, 3.1-3.8%(Kijima-Tanaka et al., 2005;Miyao et al., 1998) and Hong Kong, 0.31% (Leung et al., 2001).None of the above mentioned serogroup was isolated in Canada (Wilson et al., 1992), India (Wani et al., 2006), Vietnam (Vu-Khac andCornick, 2008) and Siri Lanka (Tokhi et al., 1993).However, a Japanese study gave a similar isolation rate of 6.2% (Kobayashi et al., 2001) and a study in Argentina revealed even a higher rate of isolation, 13.5% (Meichtri et al., 2004).The difference of prevalence values for STEC among studies are due to differences in sampling methods, sample size and detection methods used (Oporto et al., 2008).The isolation media, SMAC agar was used to isolate E. coli from cattle and goats.The primary purpose of SMAC agar is to differentiate between sorbitol fermenting and sorbitol non-fermenting E. coli which is the core feature of most O157 strains (March and Ratnam, 1986).As most non-O157 STEC are very similar to other E. coli, it is suggested to use common culture media for E. coli such as SMAC or MacConkey agar (Bettelheim, 2007).
PCR analysis of 19 STEC isolates from cattle and seven from goats showed various distribution patterns of virulence genes among the isolates with the toxigenic profile stx 2 /eaeA/ehlyA significantly associated with E. coli O157 and non-O157 isolates.In this study, stx 1 and ehlyA were the predominant genes identified from cattle and goats, respectively.Oporto et al. (2008) reported that stx 2 was predominant in cattle, whereas in sheep the Using DV test, it was found that all non-O157 STEC produced only Stx1 toxin except for one isolate from goat that produced both Stx1 and Stx2.However, multiplex PCR detected six isolates producing stx 2 .Failure to produce Stx2 in DV test for stx 2 positive strains may possibly be due to the toxins produced by the bacteria which were below the detection limit of the test kit.Moreover, it was also reported that not all stx 1 /stx 2 positive strains produce Stx 1 /Stx 2 toxins (Bettelheim, 2007;Paton and Paton, 2003).Park et al. (2003) reported that DV test had 100% sensitivity resulting in no false positive when compared to primer EHEC assay (Meridian Bioscience, Ohio) as a gold standard.In the present study, detection of Stx1 was comparable using both assays, DV and mPCR, on the contrary Stx2 detection by DV test was poor compared to mPCR which was probably due to low level of toxin production thus was not detected by DV assay.
In conclusion, the present study demonstrated the prevalence of non-O157 in cattle and goats with distribution of virulence genes.These animals act as important reservoir for the pathogenic E. coli and can be a potential source for food contamination.The distribution of virulence factors among STEC strains from cattle and goats suggest that some of them represent a potential risk for human infections.Good management practices and control strategies at the production stage are crucial to avoid widespread distribution of the organisms.

Table 2 .
Distribution of virulence genes among STEC isolates.