Decoding Peptide-MHC Recognition to unravel T cell involvement in Type 1 Diabetes and allogeneic hematopoietic stem cell transplantation

Technologies to identify T cell specificities have rapidly improved in the last decade, providing a rigorous platform to explore T cell-related immunological processes and diseases. With therapeutics progressing towards a more personalized approach, high-throughput exploration of T cell specificities is key to identifying novel disease markers and improve immunotherapies. In this thesis, we explore T cell specificity identification in the context of type 1 diabetes (T1D), graft versus leukemia (GvL), and graft versus host disease (GvHD).
The autoimmune disease T1D is the consequence of an unknown trigger event initiating an auto-reactive immune response towards β cells in the pancreatic islets of Langerhans. This inevitably results in β cell destruction, insulin deficiency, dysglycemia, and a life-long dependency on insulin-substitution. No cure for T1D exists, however, in the recent decade, tremendous efforts have focused on therapeutic interventions prior to clinical onset of T1D. To improve on these efforts, immunological markers of progression, risk factors, endotype heterogeneity classification, and therapeutic targets must be improved. The consensus in the field of T1D describes cytotoxic T cells as the main effector cells involved in β cell destruction. Thus, identification of β cell-reactive T cells in the context of T1D could establish a set of novel markers or therapeutic targets, to halt the progression of T1D. Therefore, in manuscript I, we utilized the high-throughput screening technique of DNA barcode-labeled peptide:MHC multimers, to identify T cell recognition towards a large set of 599 β cell-peptides. Here, we identified a novel set of β cell epitopes recognized by CD8+ T cells, not previously described in T1D. Prevalent epitopes were predominantly shared between T1D patients and non-diabetic healthy donors, whilst a set of low prevalent epitopes were uniquely found in T1D patients. Furthermore, phenotypical analysis of β cell-recognizing CD8+ T cells suggests a naïve phenotype regardless of disease status, except for adolescent T1D patients, displaying an elevated fraction of antigen-experienced β cell-reactive CD8+ T cells compared to the non-diabetic healthy counterpart. These data suggest that a benign autoimmune repertoire of β cell-recognizing CD8+ T cells exist in T1D patients and non-diabetic healthy individuals, whereas a set of β cell-recognizing CD8+ T cells unique to T1D patients are of relevance for future investigation in the context of T1D.
Allogeneic stem cell transplantation in the context of hematological malignancies is associated with a beneficial GvL response mediated by donor T cells. Unfortunately, these desired T cell responses are often accompanied with the undesired T cell responses towards healthy tissues, GvHD. Allo-reactive donor T cells recognize polymorphic peptides presented on HLA molecules on the cell surface of patient cells. These polymorphic peptides, or minor histocompatibility antigens (MiHAs), are produced by genetic differences between the donor and patient. As the allo-reactive donor T cells have not been negatively selected against these MiHAs in tolerogenic mechanisms prior to infusion, these MiHAs are seen as “non-self”. As a pre-emptive measure, the allogeneic stem cell transplant is depleted of lymphocytes, to avoid GvHD. This, followed by a prophylactic donor lymphocyte infusion to regain the beneficial GvL effect. However, GvHD still remains a serious side effect. High-throughput identification and monitoring of allo-reactive donor T cell responses in patients, has previously been challenging. Therefore, in manuscript II, we established a screening tool for monitoring allo-reactive CD8+ T cells and evaluated the frequencies and distribution of MiHA targets in regard to the tissue expression. Using the DNA barcode-labeled pMHC multimer strategy, we find that the amount of T cell recognition towards MiHA with a broad tissue distribution correlate to GvHD severity. This enables further investigation of distinct MiHAs impact on GvHD and GvL outcome. Our tool enables future studies to improve on donor-patient pairing and MiHA-specific immunotherapeutic interventions.