In silico inference of immunological relationship between protein antigens based on their cytotoxic T-lymphocyte epitope repertoires

Abstract

The importance of Cytotoxic T-Cell (CTL) reponses during the course of intracellular infections has received a lot of attention during the past few decades. CTLs respond to epitopes presented by the Major Histocompatibility Complex (MHC) originating from intracellular proteins for which they have an appropriate T-Cell Receptor (TCR) for. This response is crucial for the control of pathogens such as Influenza, Hepatitis, HIV and others by destroying the cell in which the pathogen replicates. Due to the extreme polymorphism of MHC molecules, Computational Immunology techniques have been developed to detect potential MHC ligands and as a consequence, potential CTL epitopes. The polymorphism factor needs to be taken into account especially when concerning the design of vaccines with a CTL response component to maximize population coverage. Tools have been constructed that combine the predictions tools concerning major steps in this pathway, that is, proteasomal cleavage, Transporter associated with Antigen Presentation (TAP) affinity, Major Histocompatibility Complex (MHC) affinity and Immunogenicity. In this study, a novel method is developed to combine the different steps in the pathway, which includes the development of a novel TAP predictor. Furthermore, by using a BLOSUM-based score in conjunction with the epitope prediction results, a novel CTL epitopebased clustering method was developed. Two pathogens with major CTL epitope components, but vastly different mutation rates were chosen to infer whether the aforementioned methods can be used to detect potential CTL epitopes and group sequences together based on shared immunogenicity.