Prediction of T-cell Epitopes for Therapeutic and Prophylactic Vaccines

The spread of existing infectious diseases and the emergence of new ones call for efficient methods for vaccine development. Some of the important players in conferring immunity against pathogens are the Cytotoxic T Lymphocytes (CTL), which eliminate infected cells. Due to their deleterious effects, it is of pivotal importance that the CTLs are able to discriminate between healthy and infected cells. To accomplish this discrimination, all nucleated cells present a segment of the peptides contained in their proteins in complex with Major Histocompatibility Complex (MHC) class I molecules on the surface of the cell. The pathway leading to MHC class I presentation is initiated when the cell’s proteins are degraded by a multi-subunit, cytoplasmic protease named the proteasome. This process generates peptides, which can be transported into the Endoplasmatic Reticulum (ER) by Transporter associated with Antigen Processing (TAP) molecules. Once inside the ER, some of the peptides will bind to empty MHC class I molecules and subsequently be transported to the cell surface. Passing CTLs will recognize any non-self peptide and kill the presenting cell. The ability to predict CTL epitopes is essential for rational vaccine design as well as for diagnostic purposes and is the centre of focus of this thesis: Part I of the thesis is an introduction to the field. In part II, I describe how we generated a method, NetCTL, for predicting CTL epitopes by integrating existing methods for predicting proteasomal cleavage, TAP transport efficiency, and MHC class I affinity. Later, we generated an updated version of NetCTL and compared its performance with that of four other publicly available methods. In part III of the thesis, I describe how we applied the NetCTL method to the prediction of CTL epitopes in specific pathogens: The bacteria Mycobacterium tuberculosis, Influenza A virus, HIV, Yellow fever virus, and West Nile virus. For each of the above-mentioned viruses, a number of predicted CTL epitopes was subsequently selected in such a way that they together constitute a broad coverage of the available viral strains. Part IV consists of a review article concerning the phenomenon that some MHC class I ligands are presented independently of TAP. Lastly, part V of the thesis deals with the characterization of the proteins containing MHC class I and II ligands, e.g. their cellular localization and expression level.