Functional studies of intimin in vivo and ex vivo: implications for host specificity and tissue tropism

Abstract

Intimin is an outer-membrane adhesin that is essential for colonization of the host gastrointestinal tract by attaching and effacing pathogens including enteropathogenic Escherichia coli (EPEC), enterohaemorrhagic E. coli (EHEC) and Citrobacter rodentium (CR). The N-terminus of intimin from the different strains is highly conserved while the C-terminus, which harnesses the active receptor-binding site, shows sequence and antigenic polymorphism. This diversity was used to define a number of distinct intimin types, the most common of which are alpha, beta and gamma. Intimin binds the type III secretion system effector protein Tir. However, a large body of evidence suggests that intimin also binds a host-cell-encoded receptor(s) (Hir), and interaction of different intimin types with Hir contributes to tissue and host specificity. The aims of this study were to compare the activity of the major intimin types (alpha, beta and gamma) in vivo and ex vivo, using the CR mouse model and in vitro organ culture (IVOC), and to determine their exchangeability. The results confirm that intimin gamma is not functional in the CR mouse model. In the pig, intimin beta can substitute for EPEC intimin alpha but when placed in an EHEC O157 : H7 background it does not produce an intimin alpha-like tropism, although some adhesion to the small and large intestine was observed. In contrast, in human IVOC, intimin beta in an EHEC background produces small intestinal colonization in a similar manner to intimin alpha.

Fig. 1.

Schematic representation (not to scale) showing the construction of recombinant intimin β and γ. The C-terminal regions including the receptor-binding regions of intimin α (eae280_α), intimin β (eae280_β) and intimin γ (eae280_γ) are represented by different shading. The percentage amino acid identity of each region of intimin β and intimin γ compared to that of intimin α is indicated. The position of the restriction site, SalI, used to construct the recombinant intimin is given in nucleotides and the position corresponding to SalI in intimin is represented as a solid line.

Fig. 2.

Colonization of mouse gastrointestinal tracts by different CR strains, as indicated by shedding of c.f.u. in stools. The levels of colonization are indicated by the viable bacterial counts (c.f.u., means±sem) from stool samples taken at different times for the 8 days post-challenge. (A) Strain DBS255(pICC327) (◊) was shed in stools at levels very similar to those of the wt strain (▪), whereas strain DBS255 (□) was not recovered after 2 days post-challenge. (B) Strain DBS255(pCVD438) (▵) was shed in stools at a slightly higher level than the wt strain (▪) on days 1 and 2 post-challenge. In contrast, strain DBS255(pICC55) (○) was shed at levels approximately 2 logs lower than those of the wt strain from days 2 to 8 post-challenge.

Fig. 3.

Caeca and colons of mice infected with different CR strains for 8 days. Mice infected with strains DBS255 and DBS255(pICC55) showed a full caecum and well-formed dark stools in the distal colon. In contrast, mice infected with wt, DBS255(pCVD438) and DBS255(pICC327) strains had shorter, thicker colons with a few diffuse watery stools and visible hyperplasia at the distal end. In addition, the caeca in these mice were often half-full or completely empty.

Fig. 4.

Virulence of CR strains in the mouse colon. (A) The total colon was weighed after the removal of all stools at day 8 post-challenge. Mice infected with DBS255(pCVD438) had colon weights that were not significantly different from those of mice infected with the wt CR. There was no significant difference between the colon weights of mice infected with DBS255 and DBS255(pICC55) and the colon weights of uninfected mice. In contrast, the colon weights of mice infected with strain DBS255(pICC327) were significantly greater than those of uninfected mice (P<0.001) but still slightly less than those of mice infected with the wt (P<0.05). (b) Mice infected with the wt strain and with DBS255(pCVD438) and DBS255(pICC327) all had similarly high pathogen burdens (around 10⁸–10⁹ c.f.u. per colon). In contrast, mice infected with DBS255(pICC55) had significantly lower bacterial loads (around 10⁴–10⁵ c.f.u. per colon; P<0.001), although the levels were still higher than those in mice infected with the DBS255 mutant, from which no challenge bacteria were recovered.

Fig. 5.

Immunofluorescence staining of CR-infected cryosectioned mouse colons. Staining was performed for CR (A, green) and Tir (B–E, green). Filamentous actin was visualized by phalloidin staining (blue), and bacteria and cell nuclei were counterstained with propidium iodide (red). Intimately adhering bacteria with translocated Tir underneath were observed on mouse colons infected with wt CR (B), DBS255(pICC438) (C) and DBS255(pICC327) (D). No adherent bacteria were detected on mouse colons infected with DBS255(pICC55) (E).

Fig. 6.

SEM of human IVOC. (A) Non-infected duodenal tissue showed smooth surface epithelium without any bacteria. In contrast, intimately attaching bacteria were present on duodenal mucosa infected with ICC170(pICC327) (B) and CVD206(pICC327) (C). Bars, 5 μm.

Fig. 7.

Representative micrographs of HPS-stained sections of porcine IVOC segments inoculated with the complemented mutant strains CVD206(pICC327) (A, ileal IVOC,), or ICC170(pICC327) (B, jejunal IVOC). Large foci of intimately adhering bacteria, along with mucosal irregularities, were observed for CVD206(pICC327) (arrowheads), whereas either smaller foci of intimately adhering bacteria, or individually adhering bacteria, not always associated with mucosal irregularities, were observed for ICC170(pICC327) (arrow). Bars, 500 μm.

Fig. 8.

Adherence of the complemented mutant strains CVD206(pICC327) and ICC170(pICC327) to porcine intestinal IVOC prepared from duodenum, jejunum, ileum, caecum and colon, as observed by SEM. Although CVD206(pICC327) induced typical A/E lesions in all intestinal sites investigated (**), complemented mutant strain ICC170(pICC327) showed only rare A/E lesions in the duodenum (arrowheads). Bacteria associated with the epithelium were observed in small to large foci (arrows) in all other intestinal sites investigated, without direct evidence of A/E lesions. Bars, 5 μm [CVD206(pICC326) and ICC170(pICC327) duodenum] or 10 μm [ICC170(pICC327) jejunum, ileum, caecum and colon].