Escherichia coli O157:H7 does not require intimin to persist in pigs.

Pigs dually inoculated with Escherichia coli O157:H7 and an isogenic intimin deletion mutant were examined 1 h to 38 days postinoculation. Both strains were recovered from the alimentary tract without sustained significant differences between strains. The highest levels were recovered from the spiral colon early postinoculation and from the tonsil later in the study. There is evidence that intimin facilitates the persistence of Escherichia coli O157:H7 in adult ruminants (6). In cattle and sheep, E. coli O157:H7 strain 86-24 was recovered in higher numbers from feces 2 weeks and 1 month after experimental inoculation than was the isogenic intimin-deficient mutant. Strain 86-24 was present in seven out of eight sheep and calves, whereas only one sheep and one calf remained infected with the intimin-deficient mutant at 1 month postinfection. Intimin is necessary for the development of attachment-and-effacement lesions in E. coli O157:H7-infected newborn pigs (9, 11, 17, 22). It has been demonstrated that older pigs can be infected by E. coli O157:H7 and maintain the infection for at least 2 months with recovery of the inoculum E. coli O157:H7 from the feces (4). There have been no studies to determine what, if any, role intimin plays in the persistence of E. coli O157:H7 in such older pigs. Since pork can serve as a vehicle for food-borne pathogens (1, 12, 13, 24), it is important to define the factors that influence persistent colonization of E. coli O157:H7 in pigs. The objective of this study was to determine if intimin facilitates persistence of E. coli O157:H7 in older pigs. Ten-week-old conventional pigs were acclimated for 2 weeks to the facilities and antibiotic-free feed; preinoculation fecal samples were collected. The bacterial inoculum strains were E. coli O157:H7 strain 86-24, resistant to nalidixic acid (6, 15), and an intimin-deficient strain, E. coli O157:H7 86-24 Δeae10, that was resistant to streptomycin (6, 17). Bacterial strains were grown as previously described, and inoculum cultures were harvested, quantified, and frozen at −80°C until needed (4, 5, 20). The inoculum contained both the parent strain and the intimin mutant strain at a dose of 1010 CFU per strain. The inoculum was given via the feed to 49 pigs. Two pigs served as noninoculated controls. Four to six pigs were necropsied at scheduled intervals postinoculation (1, 6, 12, 24, and 48 h and 4, 12, 24, 31, and 38 days). Tissue samples from the soft palate adjacent to the tonsil, the tonsil, the stomach, the cecum, the spiral colon, and the rectal-anal junction and fecal samples were collected for bacteriologic analysis. The preinoculation fecal samples were processed and plated onto MacConkey agar, sorbitol MacConkey agar, and sorbitol MacConkey agar with streptomycin-novobiocin and nalidixic acid-novobiocin antibiotic combinations added to suppress E. coli strains other than the inoculum strains. Fecal and tissue samples were collected and processed for bacteriologic evaluation as previously described (3-7). Tonsil, cecum, and rectal-anal junction tissue samples were collected at necropsy and processed for histologic examination. Slides were stained with hematoxylin and eosin, and another set was stained for E. coli O157 antigen (8). Selected sections from the paraffin blocks were prepared for transmission electron microscopy by staining with 1% osmium tetroxide, cut into thin sections, and placed on nickel grids for viewing with a Philips 410 transmission electron microscope. The results were analyzed as correlated observations. Distribution of differences was used for statistical evaluation. The intimin-deficient strain count was subtracted from the parent strain count, resulting in a sample difference for each sample. If samples were only positive on enrichment, an arbitrary number of 49 was assigned for statistical analysis. The differences were averaged at each time point, and the standard deviation and probability interval were calculated. Each set of data was evaluated to determine if zero was within the probability interval. JMP (version 5.0.1a; SAS Institute, Inc.) was used for analysis, with a P value of <0.05 indicating a significant difference. Both inoculum strains of E. coli O157:H7 were recovered from all inoculated pigs, with little difference between strains, as determinated by bacterial counts or the number of tissues positive throughout the study (Table ​(Table1).1). The tonsil, soft palate, cecum, and spiral colon had recoverable levels of one or both inoculum strains throughout the study, with bacterial counts tending to be highest in the spiral colon (Fig. ​(Fig.1).1). The parent strain was isolated in a statistically significantly higher quantity than the intimin-deficient strain in six instances: from the rectal-anal junction at 6 h postinoculation, from the cecum at 24 and 48 h postinoculation, and from the spiral colon, rectal-anal junction, and feces at 48 h postinoculation. The parent strain was not isolated consistently more frequently or in greater quantity from tissue samples than was the intimin-deficient mutant strain (Table ​(Table1;1; Fig. ​Fig.11). FIG. 1. Bacterial isolation from tissues and feces. Shown are individual log10 bacterial counts of the E. coli O157:H7 parent strain (squares) and its intimin-deficient mutant (triangles) obtained by direct plating and enrichment of tissue and fecal samples from ... TABLE 1. Tissue and fecal recovery of inoculum strainsa Histologically, there were low numbers of bacteria intimately associated with the mucosa of cecal sections. Three of the cecal samples contained O157-antigen-positive organisms closely associated with enterocytes. The antigen-positive organisms could have been either of the inoculum strains. In some instances, the enterocyte brush border was interrupted and the cells were somewhat attenuated; however, evidence of effacement was not compelling. All three of these samples were from pigs necropsied at 6 h postinoculation. Sections from these samples were also examined by electron microscopy; no attaching-and-effacing lesions were observed. There have been many other reports of studies of E. coli O157:H7 infection and persistence in ruminants but few reports of studies of nonruminant species (2, 4-7, 14, 16, 18, 23). The inability to demonstrate reduced persistence by the intimin-deficient mutant strain in pigs contrasts with similar studies of these strains in cattle and sheep (6). In aggregate, the studies lead to the suggestion that intimin contributes more to persistence in ruminants than in swine. The complete mechanism of colonization by E. coli O157:H7 is unclear. Other Shiga-toxigenic E. coli strains lacking intimin presumably have other adhesions facilitating colonization and pathogenicity (10, 16, 19, 21). These or similar factors are likely active or present in strains possessing intimin as well. It is likely that unidentified host, as well as bacterial, factors contribute to the colonization relationship. These yet-to-be-described interactions may contribute to the predilection of ruminants to serve as a reservoir host for E. coli O157:H7. The results suggest that, in contrast to cattle, sheep, and neonatal pigs, intimin is not required for persistent colonization of E. coli O157:H7 in 12-week-old conventional pigs.