A mechanistic model predicting cell surface presentation of peptides by MHC class I proteins, considering peptide competition, viral intracellular kinetics and host genotype factors

• Main Author: Eccleston, Ruth Charlotte
• Published: University College London (University of London) 2017
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.747069

Major histocompatability complex class I (MHC-I) proteins present short fragments of pathogenic or cancerous proteins (peptides) on the surface of infected cells for recognition by T lymphocytes which are stimulated upon recognition of foreign peptides. Due to the diversity of peptide sequences and the sequence-specificity of MHC-I alleles, being able to determine which peptides will be presented by which MHC-I alleles and in what proportion could be important for the development of vaccines and treatments based on the presented peptiodome. Machine learning tools, trained on experimental data, are widely used to predict immunogenic peptides. However they are unable to account for the impact the intracellular kinetics of the pathogenic or cancerous protein which will greatly influence the resultant peptidome. Here we describe a mechanistic model of peptide presentation, validated against experimental data, which accounts for intracellular peptide concentration, and can predict the relative cell surface presentation of competing peptides with varying affinities for MHC-I proteins. We demonstrate how combining this mechanistic model with the intracellular kinetics of HIV proteins can provide insight in to the experimentally reported immunogenicity of the viral protein Gag, and show how such a model can be used to predict the most abundant viral peptides presented on the cell surface. Similarly, we predict the HeLa cell peptidome and demonstrate how a simple metric can be used to approximate the abundance of a peptide based solely on protein synthesis and degradation, peptide-MHC affinity and proteasomal cleavage.