Stephanie Schüller

TccP is an enterohaemorrhagic Escherichia coli O157:H7 type III effector protein that couples Tir to the actin‐cytoskeleton†

Subversion of host cell actin microfilaments is the hallmark of enterohaemorrhagic (EHEC) and enteropathogenic (EPEC) Escherichia coli infections. Both pathogens translocate the trans‐membrane receptor protein – translocated intimin receptor (Tir), which links the extracellular bacterium to the cell cytoskeleton. While both converge on neural Wiskott–Aldrich syndrome protein (N‐WASP), Tir‐mediated actin accretion by EPEC and EHEC differ in that TirEPEC requires both tyrosine phosphorylation and the host adaptor protein Nck, whereas TirEHEC is not phosphorylated and utilizes an unidentified linker. Here we report the identification of Tir‐cytoskeleton coupling protein (TccP), a novel EHEC effector that displays an Nck‐like coupling activity following translocation into host cells. A tccP mutant did not affect Tir translocation and focusing but failed to recruit α‐actinin, Arp3, N‐WASP and actin to the site of bacterial adhesion. When expressed in EPEC, bacterial‐derived TccP restored actin polymerization activity following infection of an Nck‐deficient cell line. TccP has a similar biological activity on infected human intestinal explants ex vivo. Purified TccP activates N‐WASP stimulating, in the presence of Arp2/3, actin polymerization in vitro. These results show that EHEC translocates both its own receptor (Tir) and an Nck‐like protein (TccP) to facilitate actin polymerization.

A major role for intestinal epithelial nucleotide oligomerization domain 1 (NOD1) in eliciting host bactericidal immune responses to Campylobacter jejuni

Campylobacter jejuni is the foremost cause of bacterial‐induced diarrhoeal disease worldwide. Although it is well established that C. jejuni infection of intestinal epithelia triggers host innate immune responses, the mechanism(s) involved remain poorly defined. Innate immunity can be initiated by families of structurally related pattern‐recognition receptors (PRRs) that recognize specific microbial signature motifs. Here, we demonstrated maximal induction of epithelial innate responses during infection with live C. jejuni cells. In contrast when intestinal epithelial cells (IECs) were exposed to paraformaldehyde‐fixed bacteria, host responses were minimal and a marked reduction in the number of intracellular bacteria was noted in parallel. These findings suggested a role for intracellular host–C. jejuni interactions in eliciting early innate immunity. We therefore investigated the potential involvement of a family of intracellular, cytoplasmic PRRs, the nucleotide‐binding oligomerization domain (NOD) proteins in C. jejuni recognition. We identified NOD1, but not NOD2, as a major PRR for C. jejuni in IEC. We also found that targeting intestinal epithelial NOD1 with small interfering RNA resulted in an increase in number of intracellular C. jejuni, thus highlighting a critical role for NOD1‐mediated antimicrobial defence mechanism(s) in combating this infection at the gastrointestinal mucosal surface.

Interaction of Shiga toxin from Escherichia coli with human intestinal epithelial cell lines and explants: Stx2 induces epithelial damage in organ culture

Shiga toxins (Stx) produced by Escherichia coli are associated with systemic complications such as haemolytic–uraemic syndrome. The mechanism of Stx translocation across the epithelial barrier is unknown as human intestinal epithelium lacks receptor Gb3. In this study, we have examined the interaction of purified Stx1 and 2 with Caco‐2 (Gb3+) and T84 (Gb3–) cell lines, and determined the effects of Stx on human intestine using in vitro organ culture (IVOC). Stx exposure caused inhibition of protein synthesis and apoptosis in Caco‐2 but not in T84 cells. However, both Stx1 and 2 were transported to the endoplasmic reticulum, and the Stx1 A‐subunit was cleaved in a furin‐dependent manner in both cell lines. Thus, a Gb3‐independent retrograde transport route exists in T84 cells for Stx that does not induce cell damage. IVOC demonstrated increased epithelial cell extrusion in response to exposure to Stx2, but not Stx1, in both small intestine and colon. Pretreatment of Stx2 with Stx2‐specific antibody abrogated this effect. Overlaying frozen sections with Stx showed lamina propria, but not epithelial, cell binding that paralleled Gb3 localization, and included endothelium and pericryptal myofibroblasts. This indicates that human intestinal epithelium may evince Stx2‐induced damage in the absence of Gb3 receptors, by an as yet unrecognized mechanism.

Coronin is involved in uptake of Mycobacterium bovis BCG in human macrophages but not in phagosome maintenance

By applying density gradient electrophoresis (DGE) to human macrophages infected with Mycobacterium bovis BCG, we were able to separate three different bacterial fractions representing arrested phagosomes, phagolysosomes and mycobacterial clumps. After further purification of the phagosomal population, we found that isolated phagosomes containing live BCG were arrested in maturation as they exhibited only low amounts of the lysosomal glycoprotein LAMP‐1 and processing of the lysosomal hydrolase cathepsin D was blocked. In addition, low amounts of MHC class I and class II molecules and the absence of HLA‐DM suggest sequestration of mycobacterial phagosomes from antigen‐processing pathways. We further investigated the involvement of the actin‐binding protein coronin in intracellular survival of mycobacteria and showed that human coronin, as well as F‐actin, were associated with early stages of mycobacterial phagocytosis but not with phagosome maintenance. Therefore, we conclude that the unique DGE migration pattern of arrested phagosomes is not as a result of retention of coronin, but that there are other proteins or lipids responsible for the block in maturation in human macrophages.

The Escherichia coli Common Pilus and the Bundle-Forming Pilus Act in Concert during the Formation of Localized Adherence by Enteropathogenic E. coli

Although the bundle-forming pilus (BFP) of enteropathogenic Escherichia coli (EPEC) mediates microcolony formation on epithelial cells, the adherence of BFP-deficient mutants is significantly abrogated, but the mutants are still adherent due to the presence of intimin and possibly other adhesins. In this study we investigated the contribution of the recently described E. coli common pilus (ECP) to the overall adherence properties of EPEC. We found that ECP and BFP structures can be simultaneously observed in the course (between zero time and 7 h during infection) of formation of localized adherence on cultured epithelial cells. These two pilus types colocalized at different levels of the microcolony topology, tethering the adhering bacteria. No evidence of BFP disappearance was found after prolonged infection. When expressed from a plasmid present in nonadherent E. coli HB101, ECP rendered this organism highly adherent at levels comparable to those of HB101 expressing the BFP. Purified ECP bound in a dose-dependent manner to epithelial cells, and the binding was blocked with anti-ECP antibodies, confirming that the pili possess adhesin properties. An ECP mutant showed only a modest reduction in adherence to cultured cells due to background expression levels of BFP and intimin. However, isogenic mutants not expressing EspA or BFP were significantly less adherent when the ecpA gene was also deleted. Furthermore, a DeltaespA DeltaecpA double mutant (unable to translocate Tir and to establish intimate adhesion) was at least 10-fold less adherent than the DeltaespA and DeltaecpA single mutants, even in the presence of BFP. A Delta bfp DeltaespA DeltaecpA triple mutant showed the least adherence compared to the wild type and all the isogenic mutant strains tested, suggesting that ECP plays a synergistic role in adherence. Our data indicate that ECP is an accessory factor that, in association with BFP and other adhesins, contributes to the multifactorial complex interaction of EPEC with host epithelial cells.

Characterization of Two Non-Locus of Enterocyte Effacement-Encoded Type III-Translocated Effectors, NleC and NleD, in Attaching and Effacing Pathogens

ABSTRACT Intestinal colonization by enteropathogenic and enterohemorrhagic Escherichia coli requires the locus of enterocyte effacement-encoded type III secretion system. We report that NleC and NleD are translocated into host cells via this system. Deletion mutants induced attaching and effacing lesions in vitro, while infection of calves or lambs showed that neither gene was required for colonization.

Tir phosphorylation and Nck/N-WASP recruitment by enteropathogenic and enterohaemorrhagic Escherichia coli during ex vivo colonization of human intestinal mucosa is different to cell culture models.

Tir, the translocated intimin receptor of enteropathogenic and enterohaemorrhagic Escherichia coli (EPEC and EHEC) and Citrobacter rodentium, is translocated into the host cell by a filamentous type III secretion system. Epithelial cell culture has demonstrated that Tir tyrosine phosphorylation is necessary for attaching effacing (A/E) lesion formation by EPEC and C. rodentium, but is not required by EHEC O157:H7. Recent in vivo work on C. rodentium has reported that Tir translocation, but not its phosphorylation, is necessary for colonization of the mouse colon. In this study we investigated the involvement of Tir and its tyrosine phosphorylation in EPEC and EHEC human intestinal colonization, N‐WASP accumulation and F‐actin recruitment using in vitro organ culture (IVOC). We showed that both EPEC and EHEC Tir are translocated into human intestinal epithelium during IVOC and that Tir is necessary for ex vivo intestinal colonization by both EPEC and EHEC. EPEC, but not EHEC, Tir is tyrosine phosphorylated but Tir phosphorylation‐deficient mutants still colonize intestinal explants. While EPEC Tir recruits the host adaptor protein Nck to initiate N‐WASP‐Arp2/3‐mediated actin polymerization, Tir derivatives deficient in tyrosine phosphorylation recruit N‐WASP independently of Nck indicating the presence of a tyrosine phosphorylation‐independent mechanism of A/E lesion formation and actin recruitment ex vivo by EPEC in man.

Shiga Toxin Interaction with Human Intestinal Epithelium

After ingestion via contaminated food or water, enterohaemorrhagic E. coli colonises the intestinal mucosa and produces Shiga toxins (Stx). No Stx-specific secretion system has been described so far, and it is assumed that Stx are released into the gut lumen after bacterial lysis. Human intestinal epithelium does not express the Stx receptor Gb3 or other Stx binding sites, and it remains unknown how Stx cross the intestinal epithelial barrier and gain access to the systemic circulation. This review summarises current knowledge about the influence of the intestinal environment on Stx production and release, Stx interaction with intestinal epithelial cells and intracellular uptake, and toxin translocation into underlying tissues. Furthermore, it highlights gaps in understanding that need to be addressed by future research.

Dissecting the role of the Tir:Nck and Tir:IRTKS/IRSp53 signalling pathways in vivo.

Attaching and effacing (A/E) lesions and actin polymerization, the hallmark of enteropathogenic Escherichia coli (EPEC), enterohemorrhagic E. coli (EHEC) and Citrobacter rodentium (CR) infections, are dependent on the effector Tir. Phosphorylation of TirEPEC/CR Y474/1 leads to recruitment of Nck and neural Wiskott–Aldrich syndrome protein (N‐WASP) and strong actin polymerization in cultured cells. TirEPEC/CR also contains an Asn‐Pro‐Tyr (NPY454/1) motif, which triggers weak actin polymerization. In EHEC the NPY458 actin polymerization pathway is amplified by TccP/EspFU, which is recruited to Tir via IRSp53 and/or insulin receptor tyrosine kinase substrate (IRTKS). Here we used C. rodentium to investigate the different Tir signalling pathways in vivo. Following infection with wild‐type C. rodentium IRTKS, but not IRSp53, was recruited to the bacterial attachment sites. Similar results were seen after infection of human ileal explants with EHEC. Mutating Y471 or Y451 in TirCR abolished recruitment of Nck and IRTKS respectively, but did not affect recruitment of N‐WASP or A/E lesion formation. This suggests that despite their crucial role in actin polymerization in cultured cells the Tir:Nck and Tir:IRTKS pathways are not essential for N‐WASP recruitment or A/E lesion formation in vivo. Importantly, wild‐type C. rodentium out‐competed the tir tyrosine mutants during mixed infections. These results uncouple the Tir:Nck and Tir:IRTKS pathways from A/E lesion formation in vivo but assign them an important in vivo role.

The ex vivo response of human intestinal mucosa to enteropathogenic Escherichia coli infection

In vitro organ culture (IVOC) represents a gold standard model to study enteropathogenic E. coli (EPEC) infection of human intestinal mucosa. However, the optimal examination of the bacterial–host cell interaction requires a directional epithelial exposure, without serosal or cut surface stimulation. A polarized IVOC system (pIVOC) was developed in order to overcome such limitations: apical EPEC infection produced negligible bacterial leakage via biopsy edges, resulted in enhanced colonization compared with standard IVOC, and showed evidence of bacterial detachment, as in natural rabbit EPEC infections. Examination of mucosal innate immune responses in pIVOC showed both interleukin (IL)‐8 mRNA and protein levels were significantly increased after apical EPEC infection. Increased IL‐8 levels mainly depended on flagellin expression as fliC‐negative EPEC did not elicit a significant IL‐8 response despite increased mucosal colonization compared with wild‐type EPEC. In addition, apical application of purified flagella significantly increased IL‐8 protein levels over non‐infected controls. Immunofluorescence staining of EPEC‐infected small intestinal biopsies revealed apical and basolateral distribution of Toll‐like receptor (TLR) 5 on epithelium, suggesting that EPEC can trigger mucosal IL‐8 responses by apical flagellin/TLR5 interaction ex vivo and does not require access to the basolateral membrane as postulated in cell culture models.