Abstract:
Vaccination is a great success story. In the last 200 years it has controlled
and eradicated a number of deadly diseases like small pox, measles, diphtheria,
polio and others. But still, there are many infectious diseases which kill millions of people each year that could be controlled or extirpate with a vaccine. The discovery of vaccines by classical methods is costly, time-consuming and the results are not always completely safe to use. In the post-genomic era, the availability of complete genomes of pathogenic organisms has helped in indentifying surface-exposed proteins, which are potential vaccine candidates (‘reverse vaccinology’). In parallel, immunoinformatics techniques and tools have been developed to indentify immunogenic peptide epitopes from proteins of pathogenic organisms (‘epitope mapping’), which can be used to develop peptide-based vaccines. Here I have used a clustering-based reverse vaccinology method and combined it with epitope mapping techniques to indentify peptide epitope sequences that are conserved in surface-exposed proteins among Gram-negative bacterial pathogens, and that could be used in the development of new vaccines.
In this work, I established a highly precise consensus subcellular localization
prediction pipeline for gram negative bacteria and archaea, including prominent
pathogens, based on a clustering approach. This can be used to indentify surface
exposed proteins of the pathogens, and to annotate subcellular localization of newly sequenced genomes of Gram-negative bacteria and archaea and of proteins
identified in mass spectrometry experiments. I have also established an ‘epitope
mapping’ pipeline, which can be used to identify the B cell and helper T cell epitopes conserved in different pathogenic strains of a species. As part of this work, I analyzed the influence of amino acids and their position in the C-terminal insertion signal of bacterial outer membrane proteins, revealing the presence of patterns, which are specific for both taxonomy classes and protein classes. Additionally, these results have implications for the heterologous expression of such proteins in E. coli.