Enterohemorrhagic Escherichia coli (EHEC) in water from karst springs: detection with real-time PCR and isolation of strains

Within 2 months, two water sources in a karst area in Switzerland were sampled 9 times each, and analyzed by real-time PCR for 6 EHEC O-types, Shiga-like-toxin (stx1 and stx2) and intimin (eae) genes. With the exception of O111, 5 O-types were recorded regularly and at high frequencies (O26: 33.3 %; O157: 33.3 %; O104: 66.6 %; O103: 72.2 %; O145: 94.4 %). Genes for Shiga-like-toxins and intimin were almost omnipresent (stx1: 77.8 %; stx2: 83.3 %; eae: 77.8 %). Strain isolation was undertaken for O-groups 26, 103, 104, 145 and 157. Sample selection for strain isolation was based on Cq-values for the O-groups and stx1, stx2 and eae. From selected samples, frozen enrichment cultures were cultivated on EHLY-agar and 50 typical colonies screened for the O-type and genes encoding for stx1, stx2 and eae. With this approach, only one virulent EHEC-strain could be isolated (Escherichia coli O103, stx1 +; stx2 -; eae +). We carried out one extensive testing with 800 colonies of O-group O145, and no virulent strain was isolated. Our findings showed that PCR-results are not sufficient to formulate epidemiological conclusions and that the isolation of strains is necessary. However, as the detection procedure of EHEC in foods is cumbersome and expensive, the appropriateness of such an approach in official food control is a matter of debate.