Salmonella is an important food borne pathogen. Over 100,000 cases of human Salmonella infection are reported in the European Union each year, resulting in an economic burden estimated to be around 3 billion Euros per year (EFSA, 2012). In a European Food Safety Authority (EFSA) survey between 2006 and 2007 S. Typhimurium was the most common serovar of Salmonella isolated from pig carcasses (EFSA, 2008a). Pigs can be asymptomatic carriers of S. Typhimurium (Berends et al., 1996) and contaminated pork contributes significantly to the number of human infections. It has been estimated that the porcine Salmonella reservoir contributes between 10-20% of human salmonellosis cases per year (VLA, 2010). In addition to improvements in biosecurity and husbandry practices, immune-prophylaxis is an important method to reduce the prevalence of food borne pathogens such as Salmonella in reservoir species. An understanding of the molecular basis of bacterial colonisation and persistence in the reservoir host is crucial to rational vaccine design and targeting relevant species. S. Typhimurium expresses multiple surface factors involved in adherence and colonisation of gut epithelium in several host species. The aim of this project was to identify factors involved in S. Typhimurium colonisation of the porcine gut. The work presented here specifically focuses on the role of flagella in the colonisation of porcine gut epithelium. Flagella are motility organelles possessed by many bacterial species. Flagella can also function as surface adhesins, shown in Escherichia coli O157:H7 (Mahajan et al., 2009), and Pseudomonas (De Bentzmann et al., 1996, Lillehoj et al., 2002). Flagellin is the major flagellar filament structural protein approximately 50kDa in size. Salmonella enterica has the ability to switch between two alternate, antigenic forms of its flagellin filament protein, expressing either FliC or FljB (Macnab, 1996). The biological relevance of these two types of flagella filament protein is still not understood. It has been postulated that the presence of a second phase type of flagella may offer an advantage to the bacteria by avoiding recognition by the immune system. However, studies have shown that both FliC and FljB flagella activate Toll-like receptor-5 (TLR-5) mediated by nuclear factor (NF)-κB signalling (Simon and Samuel, 2007b). One specific objective of this research was to compare the role of flagellar phase types in S. Typhimurium adherence and colonisation of porcine gut. To this end a porcine colonic primary epithelial cell culture and ex vivo tissue explants were developed as in vitro infection models. Primary colonic cell cultures were phenotypically characterised using specific markers for epithelial and M cells. In addition to primary epithelial cell culture, porcine intestinal epithelial cell line, IPEC-J2, was also used for specific flagellar interaction studies. The role of flagella in interaction of S. Typhimurium to porcine intestinal epithelium was tested using S. Typhimurium strain SL1344 and flagella mutant derivative strains. Flagella mutant strains exhibited reduced binding to porcine intestinal epithelial cells. Purified flagella proteins were also shown to bind porcine intestinal epithelial cells. Moreover, flagella specific anti-sera suppressed S. Typhimurium adherence to both porcine intestinal epithelial cells as well as porcine colonic explants. The immuno-protective role of flagella as a potential S. Typhimurium vaccine candidate was tested during vaccine efficacy studies in pigs. Parenteral immunisation of pigs with purified FliC and FljB flagella proteins induced production of both IgG and IgA antibodies. The vaccination of pigs with Salmonella flagella provided some protection against challenge as fewer ileum tissue samples from the pigs in the vaccinated group tested positive for Salmonella. The intestinal contents from the vaccinated pigs tested for Salmonella post mortem appeared to also have lower levels of Salmonella compared to un-vaccinated controls, though these were not significantly different between groups. This project has identified flagella as one potential subunit of a multivalent subunit vaccine to help control salmonellosis in the porcine reservoir.

Flagella, the helical propellers that extend from the bacterial cell surface, illustrate how complex nanomachines assemble outside the cell. The sequential construction of the flagellar rod, hook, and filament requires export of thousands of structural subunits across the cell membrane and this is achieved by a specialised flagellar Type III Secretion System (fT3SS) located at the base of each flagellum. The fT3SS imposes a crude ordering of subunits, with filament subunits only exported once the rod and hook are complete. This “export specificity switch” is controlled by the FlhB component of the fT3SS export gate in response to a signal from the exported molecular ruler FliK, which monitors the length of the growing hook. This study seeks to clarify how rod and hook subunits interact with FlhB, and how FlhB switches export specificity. Rod and hook subunits possess a conserved gate recognition motif (GRM; Fxxxφ, with φ being any hydrophobic residue) that is proposed to bind a surface-exposed hydrophobic patch on the FlhB cytosolic domain. Mutation of the GRM phenylalanine and the final hydrophobic residue resulted in impaired subunit export and decreased cell motility. Isothermal titration calorimetry was performed to assess whether subunit export order is imposed at FlhB. These experiments showed that rod and hook subunits bind to FlhB with micromolar dissociation constants (5-45 μM), suggesting transient interactions. There was no clear correlation between subunit affinity for FlhB and the order of subunit assembly in the nascent flagellum. Solution-state nuclear magnetic resonance (NMR) spectroscopy supported prior data showing that rod and hook subunits interact with FlhB’s surface-exposed hydrophobic patch. NMR also indicated that residues away from the patch undergo a conformational change on subunit binding. FlhB autocleaves rapidly in its cytosolic domain, and the resulting polypeptides (FlhBCN and FlhBCC) are held together by non-covalent interactions between b-strands that encompass the autocleavage site. The autocleavage event is a prerequisite for the export specificity switch, but its function is unclear. Analysis of the cellular localization of FlhBCN and FlhBCC revealed that FlhBCC dissociated from the membrane export machinery, but only in the presence of FliK. Biochemical and biophysical studies of FlhB variants that undergo export specificity switching in the absence of FliK showed that these FlhB “autonomous switchers” were less stable than wildtype FlhB and their FlhBCC domain could dissociate from the export machinery in the absence of FliK. The results suggest that the export specificity switch involves a FliK-dependent loss of FlhBCC from the export machinery, eliminating the binding site for rod and hook subunits.