Abstract
BACKGROUND: The transporter associated with antigen processing (TAP) supplies cytosolic peptides into the endoplasmic reticulum for binding to major histocompatibility complex (MHC) class I molecules. Its specificity therefore influences the repertoire of peptides presented by MHC molecules. Compared to human TAP, murine TAP's binding specificity has not been characterized as well, even though murine systems are widely used for basic studies of antigen processing and presentation. METHODOLOGY/PRINCIPAL FINDINGS: We performed a detailed experimental analysis of murine TAP binding specificity by measuring the binding affinities of 323 peptides. Based on this experimental data, a computational model of murine TAP specificity was constructed. The model was compared to previously generated data on human and murine TAP specificities. In addition, the murine TAP specificities for known epitopes and random peptides were predicted and compared to assess the impact of murine TAP selectivity on epitope selection. CONCLUSIONS/SIGNIFICANCE: Comparisons to a previously constructed model of human TAP specificity confirms the well-established differences for peptide substrates with positively charged C-termini. In addition these comparisons show that several residues at the N-terminus of peptides which strongly influence binding to human TAP showed little effect on binding to murine TAP, and that the overall influence of the aminoterminal residues on peptide affinity for murine TAP is much lower than for the human transporter. Murine TAP also partly prefers different hydrophobic amino acids than human TAP in the carboxyterminal position. These species-dependent differences in specificity determined in vitro are shown to correlate with the epitope repertoire recognized in vivo. The quantitative model of binding specificity of murine TAP developed herein should be useful for interpreting epitope mapping and immunogenicity data obtained in humanized mouse models.