Oral Presentation The 2nd Prato Conference on the Pathogenesis of Bacterial Diseases of Animals 2012

Comprehensive assignment of roles to Salmonella and E. coli genes in food-producing animals (#7)

Mark Stevens 1 , Duncan Maskell 2
  1. The Roslin Institute & Royal (Dick) School of Veterinary Studies, University of Edinburgh, Edinburgh, United Kingdom
  2. Department of Veterinary Medicine, University of Cambridge, Cambridge, United Kingdom

Salmonella and E. coli are animal and zoonotic pathogens of worldwide importance. Complete genome sequences for prevalent S. enterica serovars and E. coli pathotypes have existed for a decade, however relatively few roles have been assigned to the encoded genes in food-producing animals. Toward this aim, we recently used transposon-directed insertion-site sequencing (TraDIS) to screen a library of 8550 random mutants of S. Typhimurium for their ability to colonise the intestines of chickens, pigs and calves. TraDIS relies on massively-parallel sequencing of transposon flanking regions, enabling insertion sites to be precisely mapped and the abundance of the cognate mutants to be quantified from the number of corresponding sequence reads. TraDIS simultaneously assigned the identity and phenotype of >90% of mutants screened, revealing roles for 2721 different genes in three key reservoir hosts. A core set of genes mediating intestinal colonisation of each host was identified but mutations in some loci produced host-specific phenotypes that were not observed in surrogate rodent models. Independent mutations affecting the same gene or pathway produced consistent phenotypes and the data were further validated by testing of defined null mutants in chickens. We also retrospectively applied TraDIS to a library of 1805 E. coli O157 mutants previously screened for their ability to colonise the intestines of calves by signature-tagged mutagenesis (STM). TraDIS assigned the identity and phenotype of 1645 of the E. coli O157 mutants screened, providing far richer functional annotation than previously feasible. TraDIS proved to be far more sensitive, identifying tens of novel mutations affecting Type III secretion, cytotoxins and adhesins that were missed by STM analysis of the same library. The data provide valuable data for the rational design of control strategies and establish TraDIS as the method of choice for functional annotation of microbial genomes with minimal use of target hosts.