| There is a need to identify genes important to infectious diseases caused by bacteria, in particular, for genes essential to the infectious process itself and not simply the life of the bacterium. The goal of this thesis is to present and validate a novel computational approach to the identification of genes required for human respiratory tract infection caused by bacteria. This approach should afford a more detailed understanding of the pathogenic process, specifically, the mechanisms by which Haemophilus influenzae and Streptococcus pneumoniae cause infection. The methodology used to address this question was the development of a novel bioinformatics algorithm to determine if there is a unique set of genes which are shared between the bacteria that cause respiratory tract infections in humans. A set of 8 genes were identified which are unique to the unique to the organisms that reside and cause disease in the human respiratory tract, 4 of which have previously been shown to be involved in phase-variation in the pneumococcus. Mutants of the remaining 4 previously uncharacterized genes were created in S. pneumoniae and H. influenzae, and mutant strains were analyzed for their inability to survive and replicate in the infant rat model of nasopharyngeal colonization. As is standard to the field of study of microbial pathogenesis, insertion-duplication mutagenesis and animal testing were used to determine if the unique respiratory tract genes played a role in in vivo infection. The results of the in vitro and in vivo studies showed that there exists a unique set of genes shared among pathogens that cause the same disease in humans, e.g. respiratory tract infection. This study also showed that several of these unique genes are required for infection. Therefore, the computational method correctly predicted novel virulence factors for human respiratory infections caused by bacteria. We can conclude from this work that a fully computational approach can be used to predict genes important in infection in vivo , since the majority of the genes we identified are important in infection. The validation of this in silico approach warrants application to the study of other infectious diseases caused by bacteria. |
