Abstract

The human herpesvirus cytomegalovirus (CMV) is a prevalent pathogen and infects a person for life. After primary infection, CMV latently resides in certain body cells to avoid clearance by the immune system. Primary infection and reactivations from latency usually go unnoticed in persons with an intact immune system. But in immunocompromised persons, such as transplant recipients, CMV can cause severe diseases, because these persons cannot keep the virus in check. Virus-specific CD8+ T cells are the key component of CMV-directed immunity and their presence is associated with protection from overt disease. Several studies have already explored the human T-cell response against the virus. However, the CMV-specific T-cell repertoire is highly complex, and it remains an enigma which CD8+ T cells with which CMV antigen specificity are the most protective and best control the virus. CD8+ T cells recognise antigens using a specialised surface protein, the heterodimeric αβ T-cell receptor (TCR). This receptor is highly variable between T cells and billions of different TCRs can be found on the T cells of a single person. In this PhD project, the CD8+ T-cell repertoire specific for CMV was assessed at high resolution: Virus-specific CD8+ T cells from healthy virus carriers were enriched through in vitro stimulation with short peptides and analysed by TCR mRNA sequencing, primarily of the β-chain. CMV-specific TCRβ sequences were identified by computational analysis via multi-sample comparisons. Subsequently, the magnitude and diversity of interindividual and intraindividual CMV-specific TCRβ repertoires was analysed. The focus was on the TCR repertoire specific for HLA-C--restricted CMV peptides, because presentation of such peptides is least affected by viral immunomodulation. In total, 1809 CMV-specific TCRβ amino acid sequences were identified in this project. The CMV-specific TCR repertoires were generally highly clonal, but repertoire diversity differed between peptides. CMV-specific TCRβ clonotypes, particularly those specific for HLA-C--restricted peptides, were highly frequent in peripheral blood of virus carriers; in 3 of 9 donors, the TCR with the highest frequency in the entire CD8+ T cell repertoire was specific for one CMV peptide. Several TCRβ clonotypes with the same CMV peptide specificity were identical or strikingly similar at the amino acid level and shared by multiple donors. The cumulative frequency of these 162 shared TCRβ clonotypes was significantly higher in CMV-positive than CMV-negative donors in this cohort and a large independent validation cohort. Consequently, CMV infection leaves a specific TCRβ signature in the T-cell repertoires of its human hosts. Signature TCRβ sequences will be valuable in disease monitoring, for instance as markers for the presence of a CMV-specific T-cell response. In addition, such TCRs hold great potential for adoptive T-cell transfer, since they are tolerant to a wide range of HLA-self peptide complexes and are therefore less likely to cause toxicity in the recipient.