Features that shape CD8+ T-cell responses to viruses

Abstract

CD8+ T cells, via their specific T-cell receptor (TCR), target infected cells when recognizing pathogen-derived peptides (epitopes) bound to class I major histocompatibility complex molecules (MHCI, or HLA-class I in humans). HLA-B is the most polymorphic of the classical HLA-I molecules. HLA-B-restricted CD8+ T-cell responses have been most often linked to the outcome of infectious diseases, and are frequently of greater targeting frequency and/or higher magnitude – immunodominant responses. The breadth and impact of a CD8+ T-cell response to a given pathogen will likely be strongly influenced not only by the diversity of epitopes that the HLA molecules of the infected host can present – epitope repertoire -, but also by the multitude of TCRs (TCR repertoire) that can recognize the peptide-HLA complex (pHLA). The aim of this research was i) to examine features of the epitope repertoire presented by class I HLA molecules (mainly HLA-A and HLA-B) and their potential contribution to immunodominance, and ii) to analyze frequently targeted, virus-specific CD8+ T-cell responses, namely their composition. In the first part, we made use of both published (database-curated) experimental data and in silico-predicted data, and probed the ligand/peptide repertoire of HLA class I molecules in terms of repertoire diversity and binding affinity, conservation and promiscuity. In the second part, we characterized virus-specific, CD8+ T-cell responses ex vivo from blood samples of healthy (EBV- and/or CMV-positive) or infected (HIV; influenza) subjects, mainly regarding their TCR repertoire landscape. We found that the peptide repertoire presentable by HLA-B molecules is not more numerous, nor of a higher binding affinity, nor does it bind more promiscuously than the repertoire with potential to bind to HLA-A molecules – these features do not solely determine the reported dominance of HLA-B-restricted responses. We also confirmed that the degree of intra- and inter-locus promiscuity is quite high. In contrast, we found indications that, in the case of HIV-1, more conserved and potentially more constrained regions of the proteome might be the focus of HLA-B targeting. Moreover, immunodominant T-cell responses do not necessarily possess a higher TCR diversity, nor is the greater diversity a property of a particular HLA locus (HLA-A vs B) or molecule; in fact, the presented peptide seems to have a central role in favoring either the establishment of a highly diverse or narrow cognate T-cell response. Finally, persistence of a rapidly evolving virus and mutation of targeted peptide(s), or infection with a previously encountered virus (HIV-1 and influenza A, respectively), can leave its mark on the epitope-specific TCR repertoire. Particularly in HIV-1 infection, we observed parallel changes in epitope sequence and contribution of particular clonotypes to the cognate TCR repertoire, accompanied by magnitude variation of the respective T-cell response. The results described provide insights into the complex themes of HLA class I epitope repertoire and elicited T-cell responses. Bringing together information on the targets (epitopes) and ‘effects’ (T-cell repertoire and its changes, T-cell function, etc.) of T-cell responses elicited upon infection is crucial, in particular for vaccine design.