Development and validation of real-time PCR assays for a multi-centre study to assess the prevalence and epidemiology of shiga-toxin producing Escherichia coli in the Netherlands
Abstract number: o170
Schuurman T., Roovers A., van der Zwaluw W.K., van Zwet A.A., Sabbe L.J.M., Kooistra-Smid A.M.D., van Duynhoven Y.T.H.P.
Traditionally, laboratory surveillance for shiga-toxin producing Escherichia coli (STEC) has focused on the detection of E. coli O157 by stool culture on sorbitol-MacConkey agar (SMAC). Infections by E. coli O157 range from mild, self-limiting diarrhea to hemorrhagic colitis and hemolytic-uremic syndrome (HUS). Recently, non-O157 strains are increasingly associated with diarrhea and HUS in several European countries. Therefore, a nationwide screening program will take place in the Netherlands from November 2005 to November 2006. This screening will be real-time PCR-based (RT-PCR), with subsequent STEC isolation from the positive stools. Prior to the start of this program, RT- PCR assays were developed and validated.
Assays targeting the stx1 and stx2 genes were developed for both LightCycler (LC) and TaqMan (TM). The LC assay has been described by Bellin et al (J. Clin. Microbiol. 2001 370374). The TM stx1 assay was adapted from Jinneman et al. (Appl. Environ. Microbiol. 2003 63276333), whereas the stx2 assay was a new design. Stools were processed by a miniMAG stool protocol. The phocine herpes virus-1 was used as an internal control (IC). Both assays were validated with a panel of well characterized E. coli strains (n = 31) and non-E. coli strains (n = 38). Intra-, inter-assay variation and analytical sensitivity were assessed by dilution series (n = 8), spiked in 2 faecal matrices, analysed in 5-fold on the same day and once daily on 4 subsequent days.
Both assays proved specific for stx1 and stx2 genes and no cross-reaction was observed. The TM assay was capable of detecting approximately 10000 CFU/g of stool with a 100% hit rate for both semi-solid and liquid stools. Lower hit rates were observed at approximately 1000 CFU/g (22% and 67%). The LC assay proved to be 1 log less sensitive (100% hit rate) compared to the TM assay for semi-solid stools. Furthermore, the LC assay did not detect approximately 1000 CFU/g. Coefficients of variation (CV) were < 5% for both TM and LC assays.
Both TM and LC assays proved to be very reproducible, although their sensitivities were not equal. The lower sensitivity of the LC assay compared to TM is however still in line with other published fecal RT-PCR assays. Probably, it will not be detrimental to the screening, since it is still 1 to 2 log more sensitive compared to the reported sensitivity for SMAC screening. Both RT-PCR assays are currently in use in the Dutch STEC screening program.
|Session name:||XXIst ISTH Congress|
|Back to top|