Molecular characterisation of Shigella flexneri IcsA and the role of lipopolysaccharide O-antigen in actin-based motility.

Abstract

Shigella spp. cause bacillary dysentery through invasion of the colonic epithelium. Shigella flexneri IcsA (VirG) is a polarly localised, outer membrane (OM) protein that is essential for virulence. IcsA activates the host actin regulatory protein, neural Wiskott-Aldrich syndrome protein (N-WASP), which in-turn recruits the Arp2/3 complex that polymerises host actin. The resultant F-actin comet tails initiate bacterial actin-based motility (ABM) and intercellular spread. The N-terminal surface-exposed region of IcsA, referred to as the passenger domain (aa 53-758), is responsible for IcsA activity in ABM. A glycine-rich region (aa 140-307) within this passenger domain is involved in mediating N-WASP binding. This thesis sought to conduct a comprehensive study of IcsA structure-function. Linker-insertion mutagenesis was undertaken to randomly introduce in-frame insertions of 5 aa within the IcsA passenger domain. Forty-seven linker-insertion mutants (IcsA [subscript i]) mutants were isolated and expressed in S. flexneri ∆icsA. The resultant strains were characterised for IcsA protein production, cell surface-expression and localisation, as well as intercellular spreading, F-actin comet tail formation, and the recruitment of N-WASP. Linker-insertions between aa 595-716 of IcsA affected production and lie in a region homologous to the putative auto-chaperone domain of Bordetella pertussis BrkA. Two mutant proteins (IcsA [subscript i532] and IcsA [subscript i563]) exhibited disrupted polar targeting, enabling refinement of the polar targeting region to aa 532-563. Twenty-two of the S. flexneri strains expressing IcsA [subscript i] mutants were unable to spread from cell-to-cell; further characterisation revealed that nineteen strains were unable to form either F-actin comet tails or recruit N-WASP. Since lipopolysaccharide O-antigen (LPS-Oag) on the bacterial surface has been shown to mask IcsA function, IcsA [subscript i] mutants were expressed in rough LPS (R-LPS) strains (that lack the Oag component) to investigate the effect of LPS-Oag on IcsA:N-WASP interactions. Mutants were identified that were unable to recruit N-WASP and induce F-actin comet tails when expressed in smooth LPS (S-LPS) S. flexneri strains but able to recruit N-WASP and form F-actin comet tails in a R-LPS background. These studies enabled identification of two novel functional regions (aa 330-381 and aa 508-730) involved in N-WASP interaction. For the first time, a structural model of the IcsA passenger domain was created using the Robetta protein prediction server and IcsA was predicted to form a β-helical structure. However, not all IcsA [subscript i] mutant phenotypes could not be clearly correlated to the model. As LPS-Oag had been shown to mask IcsA function, LPS profiles at the bacterial pole (where IcsA is predominantly located) were investigated. A comparison of the LPS profiles of purified minicells (derived from the bacterial cell pole) and purified whole cells, indicated that LPS populations are uniformly distributed on the polar and lateral regions of the bacterium. IcsA is a member of the autotransporter (AT) family of proteins. Another AT protein, IgA protease of Neisseria gonorrhoea forms oligomeric structures in the OM. In situ chemical cross-linking revealed that IcsA is able to form high molecular weight complexes. Moreover, IcsA [subscript i] mutants were shown to exert a negative dominant effect on WT IcsA, providing evidence for IcsA:IcsA interactions in the OM. In conclusion, studies conducted in this thesis revealed that multiple regions of IcsA interact with N-WASP, suggesting that IcsA has evolved to activate N-WASP in the presence of LPS-Oag and host actin regulatory proteins.