Screening for verocytotoxin-producing Escherichia coli in Brussels

Verocytotoxin-producing Escherichia coli (VTEC) in the Brussels-Capital Region were screened in stool samples collected by 6 of the 9 hospital laboratories for microbiology. The participating laboratories were asked to submit ten stools per week. Samples were selected as follows: those from patients with HUS; from patients with a history of bloody diarrhea; from children Stool samples were cultured on SMAC and SMAC-CT. A colony sweep and, if positive, single colonies were tested for VTEC using a multiplex PCR targeting VT1, VT2, and their subtypes. Isolated VTEC were tested for eaeA, ehxA, and the OI 122 gene cluster. Latex agglutination was used to identify serogroups O157, O26, O103, O111, O121, and O145.
A total of 5151 stool samples were screened for VTEC between January and December 2008 with 76 (1.48%) positive PCR results. VTEC were isolated from 55 (72.3%) PCR positive stools. Twenty-seven (49.1%) isolates were positive for VT1, 16 (29.1%) for VT2 (including 4 VT2f), and 12 (21.8%) for both VT1 and VT2. The eaeA and ehxA genes were present in 36 (65.5%) and 38 (67.8%) out of 55 isolates, respectively. Only two isolates were positive for saa. Twenty-six (46.4%) isolates carried espP, while katP and etpD were only present in 11 (19.6%) isolates. All O157 VTEC (9/9) carried a complete OI 122 (COI 122), while all three OI 122 modules were present in only 17.4% (8/46) of the non-O157 VTEC, comprising 5 O111, 1 O103, and 2 isolates that did not agglutinate with these antisera. An incomplete OI 122 was detected in 16/46 (34.8%) isolates, and in 22/46 (47.8%) isolates OI 122 was not detected. Twenty-one (37.5%) isolates belonged to the six previously mentioned O serogroups, comprising 9 O157, 5 O26, 5 O111, and 2 O103.
The highest prevalence of VTEC was detected in stools from patients with a history of bloody diarrhea (7.7%) followed by stools containing macroscopic blood (2.1%). Prevalence among young children (1.82%) was significantly higher than in random diarrhea samples (1.05%) (p=0.022). VTEC was isolated in 1 out of 6 reported HUS cases. No VTEC outbreaks were detected as determined by PFGE. Forty-eight out of 76 (63.2%) positive samples were registered during the period June - September revealing a strong association between VTEC infection and summer months (p=0.005).
The majority of isolates (84%) belonged to non-O157 serogroups. The global prevalence makes it worth to screen for VTEC, at least in samples from patients with HUS, bloody diarrhea or a history of bloody diarrhea. The prevalence of COI 122 in these VTEC is low, however, the analysis of this pathogenicity island in addition to eaeA and ehxA genes allows us to better define the pathogenicity of VTEC isolates.
research supported by Brussels-Capital Region grant PRFB 2007-29 to G. Buvens