The Nck binding site and the NPY motif within Tir are conserved among typical EPEC, non-O157 EHEC and C. rodentium
(reviewed in Garmendia et al., 2005
; Frankel and Phillips, 2008
). The absence of the Nck binding site in Tir EHEC O157:H7 is compensated by expression of TccP/EspFU
(Campellone et al., 2004b
; Garmendia et al., 2004
), which similarly to Nck, activates N-WASP (Garmendia et al., 2006
; Campellone et al., 2008
; Cheng et al., 2008
; Sallee et al., 2008
). This suggests the existence of selective pressure for the preservation of these signalling pathways. However, despite intense investigations, the role of these signalling pathways in vivo
is not fully understood.
In this study we investigated the role of Tir tyrosine residues Y451 and Y471 during the entire infection cycle using the C. rodentium model. After confirming that C. rodentium expressing TirY451A, TirY471A and TirY451A/Y471A had the same in vitro phenotypes as their corresponding EPEC Y454 and Y474 Tir mutants (in terms of actin polymerization and Nck recruitment), we investigated their phenotype in vivo. We found that C. rodentium expressing the single TirY451A and TirY471A substitutions and the double TirY451A/Y471A mutation exhibited colonization dynamic, pathogen load, tissue targeting, pathology (i.e. colonic hyperplasia) and A/E lesion formation abilities that were indistinguishable from the parental wild-type strain. Importantly, we found that even though C. rodentium expressing TirY471A lost the ability to recruit Nck, N-WASP was still found recruited under attached bacteria in vivo. C. rodentium expressing TirY451A recruited both Nck and N-WASP to the site of bacterial attachment in vivo. In addition, A/E lesions induced by C. rodentium expressing any of the Tir mutants, but Tir1−33stop, showed similar accumulation of electron-dense material under adherent bacteria.
Consistent with our data, Deng et al. (2003
) have previously shown that C. rodentium
induced typical A/E lesions in vivo
at day 8 post oral challenge. However, as these studies were done using plasmid-encoded Tir, the level of colonization of the tir
mutant was greatly lower than that achieved by the wild-type strain. Using the human IVOC model, Schuller et al. (2007
) have shown that while Tir is essential for colonization of the human gut mucosa, mutation in Y474 of EPEC E2348/69 Tir abolished Nck recruitment but did not affect the ability of the bacteria to recruit N-WASP and to induce typical A/E lesions. Moreover, and consistent with our C. rodentium
data, EPEC E2348/69 expressing TirY454F/Y474F
still recruited N-WASP to the bacterial attachment sites on human IVOC. When considered together these data suggest that actin polymerization on mucosal surfaces is triggered by EPEC and C. rodentium
independently of Y454/1 and Y474/1 residues by a yet unidentified mechanism.
Recent reports have shown that Tir recruits IRTKS (Vingadassalom et al., 2009
) and/or IRSp53 (Weiss et al., 2009
) in an NPY-dependent manner. Therefore, we investigated if IRTKS or IRSp53 are recruited to the C. rodentium
attachment sites in vivo
. IRTKS was detected under attached bacteria and its recruitment was Y451-dependent. In contrast, no IRSp53 was detected under adherent bacteria in vivo
. Similarly we found that IRTKS, but not IRSp53, was recruited under adherent EHEC bacteria using human IVOC.
Adhesion of EHEC O157 to human (Garmendia et al., 2004
) and bovine (Girard et al., 2009b
) IVOC and bovine ileal loops (Vlisidou et al., 2006
) was reported to be Tir-dependent but independent of the Tir-IRTKS-TccP/EspFU
complex, which is consistent with our observation that EHEC adhered to human IVOC independently of IRTKS recruitment. In vivo
calf studies revealed no measurable differences in colonization levels between wild type and EHEC O157 ΔtccP/espFU
(Vlisidou et al., 2006
). However, in infant rabbits colonization efficiency of the tccP/espFU
EHEC O157 mutant was similar to the parent strain in the ileum but was reduced in the large bowel at 7 days post infection (Ritchie et al., 2008
). In gnotobiotic piglets, expression of TccP/EspFU
was associated with larger-size adherent bacterial aggregates (Ritchie et al., 2008
). Together, these results are consistent with the notion that the Tir NPY signalling pathway is not essential for the formation of A/E lesions and for the establishment of colonization, but promotes bacterial expansion from the initial infection sites.
In addition to the Tir Y451 and Y471 pathways, we investigated the role of the Tir N-terminal polyproline sequence as it was reported to play a role in the recruitment of kinases involved in Tir phosphorylation, Nck recruitment and actin polymerization. Contrary to its reported role in EPEC (Bommarius et al., 2007
), the polyproline sequence plays no role in recruitment of Nck and actin polymerization in C. rodentium in vitro
and in vivo
. These data are consistent with a previous report showing that deletion of the entire N-terminus of EPEC and EHEC Tir did not inhibit actin polymerization (Campellone et al., 2004a
Taken together, our results show that while Tir Y471 recruits Nck and Y451 recruits IRTKS, these signalling pathways are dispensable for colonization, colonic hyperplasia and A/E lesion formation. This conclusion raises a fundamental question: does the ability to activate these pathways benefit the bacterium? In order to address this experimentally, we performed mixed infection studies. By comparing the CIs of C. rodentium expressing mock or tir mutants, we have shown that in mixed infections of wild type with either of the single Y451 and Y471 Tir mutants the latter strains were similarly out-competed. Importantly, mixed infections of wild type and C. rodentium expressing the double TirY451A/Y471A mutant resulted in rapid decline of the mutant. These results suggest that TirY451A and TirY471A contribute independently to C. rodentium competitiveness in vivo. The fact that C. rodentium expressing TirY451A/471A is significantly more attenuated than each of the single mutants alone suggests a cooperative (accumulative) function of the Y451 and Y471 pathways. We conclude that Tir residues Y454/1 and Y474/1 contribute to in vivo competitiveness, probably during mixed infections.
In conclusion, our results show that despite defining EPEC, EHEC and C. rodentium infection, we lack basic knowledge of the mechanisms involved in A/E lesion formation. Moreover, we have shown that recruitment of N-WASP to the site of bacterial attachment in vivo occurs independently of Tir residues Y451 and Y471. While IRTKS is recruited to bacterial adhesion sites ex vivo and in vivo, it is not essential for A/E lesion formation or recruitment of N-WASP. Finally, our results show that although not involved in A/E lesion formation Tir residues Y451 and Y471 play an important role in pathogen host interaction.