Diarrheagenic Escherichia coli Markers and Phenotypes among Fecal E. coli Isolates Collected from Nicaraguan Infants

PCR and PhP typing of E. coli isolates.

E. coli isolates analyzed in the study were recovered from a study conducted from March 2005 to September 2006 in León, Nicaragua, that examined 526 stool samples from children with (n = 381) and without (n = 145) diarrhea. All 282 samples that were positive for at least one of the DEC types, namely, enterotoxigenic E. coli (ETEC), enteroaggregative E. coli (EAEC), enteropathogenic E. coli (EPEC), enterohemorrhagic E. coli (EHEC), and enteroinvasive E. coli (EIEC), according to a multiplex-PCR assay of the whole E. coli flora (12), were here further tested for the presence of DEC markers among eight randomly selected isolates per sample. Altogether, 2,164 E. coli colonies (1,563 from diarrheal samples and 601 from controls) were examined for eltB and estA for ETEC, eaeA for EPEC, vt1 and vt2 for EHEC, ial for EIEC, and CVD432 for EAEC. PCR assays were performed using a 25-μl reaction mixture with a PuRe Taq Ready-To-Go PCR bead (GE Healthcare, United Kingdom) as previously described (12). The PCR products (10 μl) were electrophoresed on a 1.5% agarose gel and visualized under UV light after ethidium bromide staining. A 100-bp DNA ladder (Invitrogen Life Technologies) was used as a molecular mass marker. The same eight isolates per sample were previously typed using biochemical fingerprinting with PhP-RE plates of the PhenePlate system (PhPlate; http://www.phplate.se ) (7) in order to define clonal groups and phenotypic diversity.

Student's t test was used to determine the statistical significance, where applicable.

DEC rates were calculated among the 282 DEC-positive samples as well as for all 4,009 isolates from all 526 previously studied samples (7) (Table 1). The EAEC (CVD432) marker was the most frequent and, surprisingly, seemed more common among isolates from controls than among isolates from diarrheal infants (Table 1). This observation is in agreement with other studies, where EAEC was also isolated from controls at frequencies equal to or higher than the rates seen with diarrheal children (1, 8, 11), indicating that such bacteria may be endemic in many parts of the world. In addition, the high phenotypic diversity shown among diarrheal as well as control isolates indicates that this group is very heterogeneous, as proposed in other studies (3, 5, 9), and it is probable that only certain strains carrying the CVD432 marker are diarrheagenic. However, such strains did not seem to be common in this study. ETEC (eltB and/or estA)-positive isolates were clearly correlated with diarrhea (P < 0.05), partly due to the presence of estA-positive isolates that were isolated only from patients with diarrhea (Table 1). The low phenotypic diversity among ETEC isolates carrying the estA marker was due to the fact that they were dominated by a single phenotype. The correlation between EPEC (eae)-positive isolates and diarrhea was less obvious (Table 1), and the diversity among isolates both from diarrhea patients and from controls was high. Also, other studies have shown that EPEC strains constitute a complex E. coli group whose members contain a diversity of virulence markers that make it difficult to determine which strains are truly pathogenic (2, 10). The few isolates positive for the EHEC (vt2) marker were found only in samples from diarrhea patients and showed low diversity, indicating that they could be members of virulent clonal groups.

Thus, our combined data from PCR detection of virulence markers and biochemical fingerprinting for detection of clonal groups have indicated that the diversity among Nicaraguan DEC-positive isolates was high for EAEC and EPEC, and those DEC types could not be significantly correlated with diarrhea. For ETEC and EHEC, the correlation to diarrhea was obvious, and, in addition, ETEC estA and EHEC isolates showed low levels of diversity due to the predominance of certain biochemical phenotypes. Thus, it seems that diarrhea was often caused not by the presence of any E. coli strain carrying a DEC marker but more probably by certain virulent DEC clones. This finding raises questions about the diagnostic value of the use of PCR to determine DEC virulence genes in stool samples without further analysis.