The main goal of this study was to determine the prevalence and distribution of pathogenicity island OI-122 in tEPEC and aEPEC. Four principle results emerged from our data: (1) the OI-122 modules were found in all tEPEC (except one isolate), but the levels of distribution varied among the OI-122 modules observed within isolates, (2) more aEPEC (57.7%) lake of OI-122 genes and the OI-122 modules was prevalent in tEPEC than in aEPEC isolates, (3) there was statistical evidence for the existence of
nleB,
efa/lifA,
sen and
pagC genes in tEPEC genotype, and (4) there was a high degree of genetic heterogeneity among OI-122 carrying EPEC strains. These results were consistent with those previously reported; suggesting that the acquisition of complete OI-122 island, OI-122 module, and OI-122 genes are evidence of horizontal gene transfer and the evolutionary dynamics among EPEC strains (
5,
15). While obtained of the OI-122 genes are associated with severe diseases and outbreaks (
6), our results could not indicate the association between OI-122 module and severity of disease because all of the strains analyzed in this study were isolated from children with acute diarrhea.
A complete OI-122 gene (carrying
efa1/lifA,
pagC,
sen, and
nleB) was more prevalent in tEPEC (18.75%) than in aEPEC (3.84%). Similar results were reported by Vieira et al. and Salvador et al. (
5,
15).
The
efa1/lifA gene (coding for EHEC factor for adherence protein) was absent in our aEPEC strains with one exception, this gene encodes a lymphostatin, which has been shown to have adhesive potentials and inhibitory effects in lymphocyte functions (
4-
6). An
efa1/lifA gene was found in 37.5% of tEPEC and in 3.8% of aEPEC strains in this study, the significant association with diarrhea was reported for the
efa1/lifA gene (
6). The
efa1/lifA frequency in aEPEC strains in Japan, Brazil, Norway, Australia and New Zealand were 32.5%, 30.4%, and 28.8% respectively, which it is higher than the results in the current study for aEPEC strains (
5-
7).
Module 2 was the prevalent module among tEPEC and aEPEC strains in this study, which it is consistent with previous reports (
5-
7). This module encodes
nleB (virulence factor) and
sen (a putative enterotoxin) genes (
16,
17). The
nleB detected in the majority (87.5%) of the tEPEC strains examined, suggesting that it represents a stable acquisition of the positive clonal lineages.
The
pagC gene (immunogen and bacterial survival factor within macrophages) from module 1 was present in strains carrying a complete OI-122 with all three modules and strains carrying the combination of this module and module 2. This module plays an important role (immunogen and bacterial survival factor within macrophages) in O157: H7 infections, (
18,
19) but its role in EPEC infection is unknown. The heterogeneous distribution of OI-122 modules was observed in EPEC strains in this work and that of other researchers, (
5-
7) probably due to the acquisition of OI-122 gene contents in a modular manner.
Most of the strains (76.19%) in this study belonged to the B1 phylogenetic group and only two isolates were the B2 phylogenetic group. This typing method classifies strains according to the presence or absence of the three genes
chuA,
yjaA, and
tspE4C2, and the only B2 group is positive for a
yjaA gene (
12). The significant negative association with diarrhea for the
yjaA gene reported in the study of Afset et al. (
6) and suggested that the presence of the
yjaA gene may be a marker for low diarrheagenic potential aEPEC strains. The
yjaA gene as the negative marker for diarrhea may be one possible reason for the association of B2 phylogenetic group with extraintestinal
E. coli infections (
20-
22).
The common serogroups were O128, O111, O55 and O44 in EPEC in this study, and 50% of tEPEC and 42.3% of aEPEC strains were untypeable with the traditional O antisera. The presence of the high number of untypeable strains in this study indicates that O serogrouping is not the valuable method in the diagnosis of EPEC infections.
The extensive heterogeneity among the tEPEC and aEPEC strains in this study was found by the MLVA analysis; differences in the MLVA type were identified between strains belonging to the same phylogenetic group or O serogroups. However, two strains shared a same (M23) MLVA type (
Table 2), these strains were isolated from a sister and brother, which are belong to the B1 phylogenetic group and O128 serogroup of aEPEC. Interestingly, while the OI-122 genes were absent in one of these strains, the other strain harbored the
nleB gene. MLVA typing identified seven clonal complexes (7CC) in EPEC strains, but clonal relationships were not found with OI-122 genes. The MLST results of aEPEC isolates from China also showed very heterogeneity (
23). Staples et al. (
24) analyzed EPEC isolates by MLST and MLVA. Their isolates were highly heterogeneous. They concluded that the human enteric EPEC population might be a complex of commensal or pathogenic strains. MLVA typing of aEPEC strains from poultry suggested that poultry strains were closer to bovine strains and were less similar to patient strains (
25).
In conclusion, the results of this study provide evidence for dynamic evolution regarding the OI-122 pathogenicity island in tEPEC and aEPEC strains. The OI-122 gene contents distribution in a modular manner in this study indicating that clinical, animal and cell line studies need for an understanding of the OI-122 genes roles in pathogenicity of EPEC strains. Since no specific results on the distribution of the OI-122 pathogenicity island in EPEC strains are available in Iran, the results of this study shed light on the roles of OI-122 genes in the pathogenicity and epidemiology of EPEC strains, particularly of the aEPEC strains.