Abstract
Background. Inclusion body myositis (IBM) is the most common inflammatory muscle diseases that primarily affects the elderly. It is characterised by autoim¬mune aggression and degeneration of skeletal muscles, which leads to severe disability over time. Although the aetiology of IBM is uncertain, numerous lines of evidence point to T cells playing a pathogenic key role. One of the major ques¬tions that remains unsolved regarding IBM muscle-invading T cells is the nature of the antigen that drives their autoreactivity. We hypothesised that the analysis of T cells’ T cell receptor (TCR) repertoire will give insights into their specific¬ity. We used high-throughput sequencing of TCRβ chains in T cells isolated from muscle biopsies and matched blood samples to compare T cell clonal profiles within the inflammation-affected muscle tissues and systemically.
Methods. Blood and muscle samples were collected on the same days for each of the four donors. Muscle-invading T cells were isolated as bulk while peripheral blood mononuclear cells were subjected to CD4+ and CD8+ T cell separation. The TCRβ sequencing was performed using the high-throughput Illumina Miseq platform. Downstream data analysis of TCRβ repertoires was performed using in-house VGAS tool, Immunarch R package and VDJtools. HLA haplotype of each donor was determined by high resolution typing of HLA class I and II al¬leles using Illumina Miseq platform in an American Society for Histocompat¬ibility and Immunogenetics (ASHI)-accredited laboratory.
Results. Analysis of muscle-T cells for TCR repertoire overlap revealed shared clonotypes between patients. A unique TCRβ sequence was shared between pa¬tients 1, 3, and 4, while in patient 2, although this sequence was not found in the muscle, it ranked as a predominant clone in the blood. The patients 1 and 3 displayed nine sequences in common, as well as their top three TCRβ sequences. Furthermore, analysis of TCR repertoire usage in the corresponding blood sam¬ples showed common clonotypes between muscle and blood, but at higher fre¬quency in muscles indicating that a preferential expansion occurred in this tissue. Moreover, querying the top five dominant TCRβ CDR3 sequences of muscle from each patient in a curated database of TCR identified multiple highly similar sequences with known specificity for antigens derived from virus and muscle proteins, especially in the CD8+ T cell subset. In patient 1 and 3, the topmost clone showed a high level of similarity with a CDR3 specific for CMV-derived antigen as well as self-antigens derived from protein phosphatase 1F (PPM1F) and A-kinase anchor protein 9 (AKAP9). The HLA alleles reported to present these autoantigens matched with the HLA haplotype of patient 3. On the other hand, for patient 2, we identified a clone bearing high CDR3 similarity with known sequences documented to be associated with HIV-1 and EBV-derived an¬tigens, and another CDR3 that exactly matched a sequence specific for a gluten-derived peptide. Interestingly, this patient carries HLA-DQ9 which is a reported risk factor for celiac disease.
Conclusions. Our findings identified public TCRs in IBM muscle, and the pres¬ence of expanded T cell clones harbouring TCR sequences with striking similari¬ties between virus and muscle-derived antigens, suggesting that an underlying molecular mimicry mechanism may generate autoreactive T cells.