Rapid bacterial Virulence Annotation for the post genomic era

Enormous amounts of money have been and are continuing to be spent on the genome sequencing of many different pathogenic bacteria. Despite this most of this sequence remains with little or no biological annotation. In order to achieve this biological annotation it would require extensive animal experimentation, which is both expensive and ethically problematic. Indeed it is widely recognized that there is a need to reduce the amount of animal experimentation currently conducted. With this in mind we have developed a technology for whole-genome mapping of bacterial virulence factors using invertebrate hosts that negates the need for mammalian testing. Simply we challenge three different invertebrate animals (insect, neamtode and amoeba) with cosmid libraries from a fully sequenced pathogen. End sequencing of cosmids that adversly effect any of these hosts are aligned with the genome sequence thus identifying pathogenic genes as bounded by clusters of cosmids. This fundamentally new approach allows us to rapidly and inexpensively bridge the enormous knowledge gap that has arisen through intensive genome sequencing. The current grant does not propose to follow-up on each of the virulence factors identified in detail but to provide access to the results as a community wide resource for workers studying each of the three bacteria. Details of the screens will be posted on the web with open access and other researchers will be able to choose cosmids for further research. Alternatively, we will also enter into collaborations with such workers to investigate genes in specific cosmids using the mutants generated.

This project plans to bridge the gap between bacterial genome sequences and their functional annotation. We have set up a number of parallel screens against different organisms that allow for the rapid identification of putative virulence factors from end-sequenced cosmid libraries. Briefly, we will prepare cosmid libraries from three important pathogenic bacteria; Yersinia pseudotuberculosis strain IP32953, Francisella tularensis subspecies tularensis strain SCHU S4 and Campylobacter jejuni strain 81-176 and end-sequence these entire libraries. We will then ordinate these end sequences onto the published genomes to validate the quality and coverage of the libraries. The cosmid screens rely on the recovery of individual comids (when screened individually, for example via insect injection) or groups of cosmids (when screened as pools of clones or whole libraries) that map together in different screens. Preliminary data from the emerging human pathogen Photorhabdus asymbiotica shows that cosmids cluster in different screens depending on the apparent mode of action of the different virulence factors they contain. For example, cell toxicity or the promotion of intracellular survival. We will analyse candidates from these clusters of cosmids via transposon mutatagenesis in order to identifiy the genes responsible and post this functional annotation on a web site. This will allow other workers in the field to access the data and request specific cosmids. Alternatively, we will enter into specific collaborations to further test a range of cosmids and their associated transposon mutants with groups interested in a specific virulence factor. In this manner we will develop a community based resource designed to reduce the use of costly animal testing and to promote the functional annotation of novel virulence factors in bacterial genomes.