Our Unit, over the years, has provided a unique and detailed analysis of how a pathogen subverts an epithelial barrier. Central to this process is a type III secretory system (TTSS) that delivers effectors dedicated to the various steps in Shigella pathogenesis (i.e. entry into epithelial cells via directed phagocytosis, escape into the cytosol and cell-to-cell spread following actin-dependent motility). We have developed in vitro and in vivo assays to model the epithelium, and unraveled a paradigm in which Shigella elicits inflammation and uses it to destabilize the epithelial barrier and to facilitate its invasion, and conversely dampens this inflammation to secure its survival, colonization and invasion capacities. Four discoveries have dominated this area: (i) the demonstration in the early nineties that Shigella causes apoptosis of macrophages, a pro-inflammatory process now called pyroptosis, (ii) the demonstration in early 2000’ of an intracellular system of microbial sensing (the Nod molecules) in epithelial cells, leading to their re-programming in order to produce pro-inflammatory mediators, (iii) the identification in the mid-2000’of a pool of Shigella effectors controlling the innate and adaptive responses, and (iv) the demonstration of the oxygen-dependant competence of Shigella to activate its TTSS while engaging cells. The whole process is controlled by a hierarchy of complex regulations that dictates the nature and timing of expression and delivery of effectors, including the assembly and activation of the TTSS itself. Capitalizing on these founding contributions, our current studies on Shigella pathogenesis address novel angles such as “life and cross-regulations at epithelial surfaces”, in other words how is the pathogen primed by the conditions met in the intestinal lumen and at the epithelial surface to optimally proceed to epithelial invasion? and “subversion of epithelial and immune functions” in other words how TTSS-injected effector molecules affect gut epithelial physiology (i.e. regeneration, differentiation, polarization, secretion), and manipulate innate and adaptive immune responses.
1 – Cross-talks at epithelial surface, reporters to track O2, mucus, and TTSS activation. Understanding how Shigella adapts to, and is primed for virulence by conditions prevailing at the gut epithelial surface. Thanks to Signature Tagged Mutagenesis, we could define key virulence factors that were otherwise undetectable in vitro (1). We recently showed that in the anaerobic environment of the gut lumen, Shigella was primed for invasion and expressed extended T3SS needles while reducing effectors secretion. This is mediated by FNR (fumarate and nitrate reduction), a regulator of anaerobic metabolism repressing the transcription of spa32 and spa33, in turn regulating TTSS secretion. We observed a zone of relative oxygenation adjacent to the epithelial surface allowing T3SS activation (2). We will continue to exploit a tagged-transposon library to identify mutations that affect the bacterial-epithelial cross-talks at their interface. In this context, we have recently identified a Shigella protein, ZapE, which is required to complete the final stage of bacterial division in anaerobic conditions that involves the reduction and disassembly of the FtsZ ring (3). We develop novel imaging tools to probe this strategic interface: We have designed a very fast folding and maturing variant of GFP that allows following its transcription under the control of a promoter responding to activation of the TTSS via the MxiE-IpgC complex (4). This original tool (5) will be essential to identify in vitro and in vivo, in real time, when and where the TTSS is active or inactive. We also develop novel tools to probe the environment at the bacterial-epithelial interface, such as a probe achieving one-step fluorescent staining of mucins (6), phosphorescent probes to “map” oxygen concentrations in vivo, and probes detecting/grading the anaerobic environments.
2 – Subversion of epithelial physiology. Shigella invasion induces major damage to receptor recycling and to the Golgi complex. We recently observed that Shigella induces tubulation of the Rab11-positive compartment, thereby affecting receptors recycling, and fragmentation of the Golgi complex, inducing inhibition of both secretion and retrograde transport. The mechanisms involve massive redistribution of cholesterol at the plasma membrane to the sites of Shigella entry. IpaB, a TTSS factor of Shigella that binds cholesterol, induces alterations of both Golgi and recycling compartments (7). This promising line will be continued with the aim to dissect the mechanisms of cholesterol recruitment from organelles to the plasmic membrane, thus getting into fundamental aspects of homeostasis of lipid membrane composition in cell compartments. We will carefully analyze cholesterol trafficking with superresolution microscopy combined to the study of vesicle trafficking in response to specific GTPases. We will also analyze how a pore forming molecule that binds cholesterol, like IpaB, can mobilize cell cholesterol resources at plasmic membrane. As Golgi disorganization was also observed in vivo, this new mechanism affecting the sorting and secretion of cell surface molecules, we will study how it contributes to epithelial barrier subversion, both in terms of rupture of the barrier capacity, and secretion of inflammatory mediators. In addition, we have shown that Shigella also subverts mucin secretion, particularly Muc-5AC, altering the physical properties of the gel and the degree of sulfation of the glycans that heavily decorate the protein moiety (8).
3 – Subversion of the innate immune response. Regulation of danger signaling and ATP release by epithelial cells as a model system. We demonstrated earlier that Shigella induced peaks in intracellular calcium concentration, dependent upon a functional TTSS, and phospholipase-C-dependent opening of connexin hemichannels, allowing extracellular release of ATP. ATP, through signaling to the P2 family of purinergic receptors, paracrinally increased bacterial invasion and calcium signaling (9,10). Connexin-mediated release of ATP also constitutes a danger signal activating the inflammasome (11), and the differentiation of Th17 lymphocytes (12). We observed that following entry into cells, Shigella induced a sharp decrease of cellular ATP release through hemichannel closing. IpgD, a TTSS-injected phosphatidyl-inositol phosphatase, hydrolyses PI(4,5)P2 into PI5P, inducing closing of hemichannels (13). We will decipher the mechanisms of hemichannel closing by PI5P by further defining the opening/closing regulatory dynamics, including genome-wide siRNA silencing. We will also address the possibility of a direct action of PI5P on hemichannels reconstituted in Giant Unilamellar Vesicles. Dampening innate immune effectors. We continue to explore the functions of the TTSS-secreted effectors OspG (14) and OspF (15). We also wish to discover the functions of the members of the IpaH family of E3 ubiquitin ligases in Shigella (16, 17). Twelve ipaH genes encode molecules sharing a conserved catalytic E3-ubiquitin ligase domain and a variable Leucine Rich Repeats domain (LRR). Their function and eukaryotic substrates are yet mysterious with one exception, IKKgamma/NEMO (18). The diverse LRR domains may target various eukaryotic substrates. To address the function of these effectors, we need to determine in which context they are expressed and to find the substrate for their activity. We have established a combination of bacterial genetics, biochemistry and large-scale proteomics strategy to define the polyubiquitinome as it is remodeled in vitro in the course of Shigella infection, or by expression of individual IpaH molecules, to search for IpaH substrates.
4 – Subversion of the adaptive immune response. We recently demonstrated that Shigella invades activated, but not non-activated, human CD4+ T and B cells in vitro (19). Injection of effectors via the TTSS occurs in the absence of invasion. Chemoattractant-induced migration of Shigella-invaded/injected T cells is arrested and the TTSS effector IpgD is implicated (19). Studies using a mouse model of Shigella infection in the footpad and two-photon LASER scanning microscopy to image cell interactions in the draining lymph node have confirmed these results (20). For B lymphocytes, we showed that Shigella triggers death of particular B cell subsets in vitro, consistent with the reduction of B cell number observed in lymph nodes upon infection with Shigella. A TTSS effector, IpaD, triggers B cell death via TLR2 activation (21). Based on these results, we will further detail the impairment of T and B lymphocyte functions, and analyze other critical steps like the generation of efficient memory B cells and antibody-producing plasma cells. We will also decipher the molecular and cellular mechanisms evolved by Shigella to impair the formation of the immune synapse both in vitro and in vivo.
5 – Conclusion and perspectives. Our recent analysis of Shigella infection of human colonic tissue explants and of experimental infection of the Guinea pig’s colon – that is sensitive to Shigella infection – has shown strong tropism of invasive bacteria to the crypts. We are exploring how the invasive phenotype of Shigella may subvert, invade and cause inflammatory destruction of the colonic crypt. On these bases, we propose a novel vision of the pathogenesis of shigellosis that will combine studies addressing crypt invasion/inflammation, subversion of key physiological properties of the epithelium, destruction of epithelial capacities of restitution, and manipulation of innate and adaptive immune protection allowing to respectively achieve early colonization/invasion by dampening danger signals and pro-inflammatory pathways, and to establish an immunosuppressive environment resulting in poor adaptive immunity, with short-lasting memory and weak helper and cytotoxic T cell responses.