Application of sequence typing to the epidemiology of Campylobacter jejuni

The bacterium Campylobacter jejuni is responsible for a substantial burden of human disease. It has a wide host and environmental distribution. The ecology in non-human hosts has been studied, in particular for farm animals, but the relationship between different host and environmental niches, including the extent to which host and geography are associated with genetic differentiation, is uncertain. Open epidemiological questions include the largely unexplained summer peaks in many temperate countries, apparent paucity of outbreaks despite a high incidence of foodborne infection, and uncertainty in the quantitative contributions of different sources to human infection. Bacterial subtyping has not made a substantial contribution to the ecology or epidemiology of this species. This thesis applied a housekeeping gene multilocus sequence typing scheme, complemented with more variable antigen genes, in epidemiological and ecological studies of C. jejuni. These studies identified a large contribution of one clonal group to the high level of human infection during summer along with a similar seasonal rise in relative prevalence for this Glade in poultry meat, and the association of other phenotypic characteristics with the Glade. A lack of geographical variation between distant populations within England contrasted with substantial differences at international level. Genetic differentiation by host species exceeded geographic and temporal effects and showed the potential of using multilocus genotype to attribute human infection to animal host sources. Recombination played a major role in the generation of this genetic differentiation, which finding informed the use of alleles rather than summary measures of genotype in population assignment of C. jejuni. Using fine typing with porA and flaA gene fragments to produce higher discrimination than previously applied allowed the identification of genotypic clusters, demonstrated emergent clades within the poultry industry and human disease, and confirmed empirically the predictions from theoretical work that even more highly discriminatory methods will be needed to reliably identify outbreaks and their specific source using genotype. This work has thus exploited the of integration multilocus sequencing in studies of the epidemiology and ecology of C. jejuni, developed analytical approaches for this application, and identified some limitations and the extent to which these may be tractable with more extensive sequence data and further development of analytical approaches.