TEXT

Top Abstract Text References

Shiga toxin-producing Escherichia coli (STEC) strains of different serotypes have been increasingly isolated from humans with disease and from healthy domestic animals (3, 6-8, 10, 11, 13, 17). Many of these isolates were typical STEC strains belonging to serotypes O26:H11, O103:H2, O111:H, O113:H21, H145:H, and O157:H7, which in humans can cause severe disease, such as hemorrhagic colitis (HC) and hemolytic-uremic syndrome (HUS) (11, 17). Numerous other serotypes of STEC either are associated with disease at low frequency or have not been implicated in disease (2, 10). Various virulence factors have been identified in STEC (12, 15, 19). Shiga toxins (Stx1 and Stx2) encoded by the stx₁ and stx₂ genes may contribute to the occurrence of diarrhea (15) and appear to be necessary for HC and HUS (9). Some STEC strains can tightly attach to epithelial cells of the intestine through an adhesin called intimin (2), which is encoded by the eae gene of attaching and effacing lesions (12). Almost all STEC strains produce an enterohemorrhagic E. coli-specific plasmid-encoded hemolysin, encoded by hlyA (19). However, only eae and stx₂ were significantly associated with isolates from serotypes found in humans or with isolates from serotypes implicated in severe human disease (2).

While various approaches have been used to develop isolation and detection procedures for STEC O157 (3, 8), effective isolation and detection of serotypes of STEC other than O157:H7 have apparently not been reported. Our group (6, 7) reported an effective, rapid, and simple method for isolation of STEC O26, O111, and O157 from feces and food samples, namely by culture on cefixime (0.05 mg/liter), tellurite (2.5 mg/liter), and sorbitol-MacConkey agar (CT-SMAC) (24) after hydrochloric acid treatment of samples, as well as enrichment cultures. This new method is based on the acid resistance of E. coli and the tellurite resistance of STEC O26, O111, and O157. One purpose of the present study was to determine if the culture on CT-SMAC after acid treatment is a useful method not only for isolating STEC O26, O111, and O157, but also for isolating other serotypes which can cause severe disease in humans. We also examined the relationship between tellurite resistance and virulence factors of STEC.

A total of 543 STEC strains, an enteroinvasive E. coli (EIEC) strain, two enterotoxigenic E. coli (ETEC) strains, and 18 enteropathogenic E. coli (EPEC) strains were collected from researchers in Japan and Germany and isolated in our laboratory (Table 1). The isolates were separated into three groups according to the grouping by Gyles et al. (10). Group 1 consisted of 321 human isolates and 92 bovine isolates, one food isolate, and 4 environmental isolates belonging to 8 serotypes which occur at moderate to high frequencies in severe human disease. In Japan, O121:H19 strains were isolated from sporadic cases of severe human disease and were classified into group 1. Group 2 consisted of 47 human isolates belonging to serotypes less frequently implicated in human disease. Group 3 consisted of 58 bovine isolates, two ovine isolates, 5 beef isolates, and 13 environmental isolates belonging to serotypes that have not been implicated in human disease.

The importance of gastric juice in controlling the outcome of food-borne infections is well recognized. To cause human illness, an invading organism must survive the acidic environment of the stomach before the organism reaches the intestine. Thus, acidity of gastric juice provides a first line of defense against food-borne pathogens (1). We reported that HCl treatment is effective for the separation of E. coli from other gram-negative bacteria, but the data covered only 17 strains (6). To confirm the effects of acid resistance in E. coli, bacterial suspensions of 10⁷ to 10⁹ CFU/ml (50 µl), cultured in Trypticase soy broth (BBL, Cockeysville, Md.) at 37°C for 18 h, were transferred to 0.125 N HCl-0.5% NaCl solutions (50 µl), mixed, and held for 30 s. Then 20-µl portions of each HCl-treated suspension were put into 1 ml of 0.067 M phosphate-buffered saline (pH 7.2), after which, a 100-fold dilution (10 µl) of those suspensions was placed on Trypticase soy agar (BBL) and then cultured at 37°C for 18 h. The viable count for all strains of E. coli decreased from 0 to 1 log₁₀ after exposure to 0.125 N HCl for 30 s, regardless of differences in serotypes and pathogenicity. Therefore, E. coli, including STEC, has resistance to HCl.

Although the mechanism of tellurite resistance in pathogens, including Corynebacterium diphtheriae, Staphylococcus aureus, Vibrio cholerae, and Shigella spp., is unknown, tellurite has been used to isolate these bacteria (14, 21). Thereafter, tellurite was used for selection of STEC O26, O111, and O157 from other members of the E. coli family (6, 7, 24). Data on growth of STEC, EPEC, EIEC, and ETEC on CT-SMAC are given in Table 1. A 100-fold dilution (10 µl) of bacterial suspensions of 10⁷ to 10⁹ CFU/ml was cultured on SMAC and CT-SMAC at 37°C for 18 h. O157 strains showed sorbitol-nonfermenting colonies, while other serotypes showed sorbitol-fermenting colonies, except for five strains: O78:H21, O103:H2, O121:H19, O145:H, and O121:H19 on CT-SMAC. Of 418 group 1 isolates, 402 grew on CT-SMAC, with no significant reduction in the size or number of colonies. However, 32 of 47 group 2 isolates, 69 of 78 group 3 isolates, and 18 of 20 other pathogenic E. coli isolates showed inhibited growth on CT-SMAC. The MICs of potassium tellurite (Daigo, Japan) for STEC are given in Table 2. A range of dilutions of tellurite were prepared in 20-ml volumes in a petri dish with Mueller-Hinton II agar (BBL). Ten microliters of cultures was placed on plates with each dilution of tellurite (final concentrations, 400 to 0.05 mg/liter), and the plates were incubated at 37°C for 18 h. The lowest dilution of tellurite that inhibited growth was recorded as the MIC. At a tellurite concentration of 3.1 mg/liter, which is near and over the concentration (2.5 mg/liter) inoculated in CT-SMAC, 402 group 1 isolates grew. To determine the MICs for 36 group 2 isolates, 63 group 3 isolates, and 16 other E. coli isolates, the tellurite concentration was under 3.1 mg/liter. The frequency of growth on CT-SMAC was significantly higher in group 1 isolates than in group 2 and 3 isolates (P < 0.01, respectively). Thus, selective isolation of group 1 isolates such as O26:H11, O111:H, and O157:H7, and also of O26:H, O103:H2, O121:H19, O145:H, and O157:H, was feasible by culture on CT-SMAC.

The eae, stx₂, and hlyA genes are found more frequently in STEC isolates from patients with severe disease than in other STEC populations (16, 18, 20), and the stx₁ gene may be associated with some STEC isolates of bovine origin (13). The association of four virulence-related genes with tellurite resistance was examined by multiplex PCR and growth on CT-SMAC of STEC isolates (Tables 1 and 3). Multiplex PCR amplification was used to detect the stx₁, stx₂, eae, and hlyA genes in all 543 STEC isolates and EIEC, ETEC, and EPEC isolates. The multiplex PCR amplification protocols used were those described by Fagan et al. (4). The primers used were those described by Gannon et al. (8) for stx₁, stx₂, and eae and by Fratamico et al. (5) for hlyA. The stx₁ and stx₂ genes were present in 79 and 57%, 65 and 67%, and 47 and 75% of the STEC isolates belonging to groups 1, 2, and 3, respectively. The stx₁ gene was more frequently detected in isolates of group 1 than in the isolates of groups 2 and 3 (0.05 > P > 0.01 and P < 0.01, respectively). In group 1, isolates belonging to serotypes O26:H11, O26:H, O103:H2, and O111:H were predominantly positive for stx₁ and isolates belonging to serotypes O121:H19, O157:H7, and O157:H were predominantly positive for stx₂. Two-thirds of O157:H7 and O157:H were positive for both stx₁ and stx₂. O145:H isolates had either the stx₁ gene or the stx₂ gene, but not both (Table 1). There was a slight difference (0.05 > P > 0.01) between growth on CT-SMAC and the presence of stx genes, because growth on CT-SMAC was observed in 83 and 76% of isolates with the stx₁ and stx₂ genes, respectively (Table 3). Thus, there was no close relationship between stx₁ and stx₂ and tellurite resistance in STEC.

The prevalence (99%) of eae in group 1 was significantly higher than the low prevalence (33 or 7.7%, respectively) of this gene in group 2 or 3 (P < 0.001, respectively). There was a significant difference (P < 0.001) between growth on CT-SMAC and the presence of eae. Growth on CT-SMAC was observed in 94% of the eae-positive isolates and in 16% of the eae-negative isolates (Table 3). The frequencies of occurrence of the hlyA gene in group 1, 2, and 3 isolates were 98, 65, and 82%, respectively (Table 1). The prevalence of hlyA in group 1 and 3 isolates was significantly higher than the prevalence of this gene in group 2 isolates (P < 0.01 and 0.05 > P > 0.01, respectively). There was a significant difference (P < 0.01) between growth on CT-SMAC and the presence of the hlyA gene. Growth on CT-SMAC was observed in 82% of the hlyA-positive isolates and 38% of the hlyA-negative isolates (Table 3). Although these data suggest an association of hlyA with tellurite resistance, there was no significant difference (P < 0.05) between growth on CT-SMAC and the presence of the hlyA gene in the eae-negative and eae-positive isolates (Table 3). These findings show significant differences between growth on CT-SMAC and the presence of the eae gene, but not the stx₁, stx₂, and hlyA genes. This close relationship between the presence of the eae gene in isolates and tellurite resistance of isolates showed that culture on CT-SMAC after HCl treatment facilitated isolation of eae-positive strains of STEC, such as group 1 isolates belonging to serotypes O26:H11, O26:H, O103:H2, O111:H, O121:H19, O145:H, O157:H7, and O157:H, associated with severe disease in humans. Because some strains (6%) of eae-positive STEC did not grow on CT-SMAC, selective media without tellurite should also be used for routine isolations of STEC.

The mechanism of tellurite resistance in STEC O157 is unknown, and whether it is chromosomally or plasmid mediated is also unclear (24). Although intrinsic tellurite resistance in members of the family Enterobacteriaceae is often plasmid mediated (23), it has been reported that chromosomally mediated resistance can be induced in tellurite-sensitive strains of E. coli, and possible mechanisms of resistance include reduced uptake via the phosphate transport pathway and reduction to metallic tellurium (22). In the present study, there was a significant difference between tellurite resistance and the presence of the eae gene encoded by chromosome, but not the presence of the hlyA gene encoded by plasmid. These findings suggest the possibility that tellurite resistance of STEC may be encoded by a gene close to or near lesions containing the eae gene on the chromosome.