Thymic selection abnormalities in Type 1 Diabetes

Project Summary Type 1 Diabetes (T1D) is an autoimmune disease in which T cells target insulin-producing β cells. At least 60 non-HLA genetic variants each confer small increases in T1D risk. While abnormal negative selection has been implicated in NOD mice, the possible role of aberrant thymic selection in driving human T1D is unknown. Human immune system (HIS) mice in which a patient’s immune system is generated de novo from their hematopoietic stem cells (HSCs), known as the personalized immune mouse (PIM) model, provide a unique opportunity to evaluate thymic selection events that contribute to T1D, permitting assessment of immune repertoire development and experimental manipulations such as T cell receptor (TCR) transgenesis. Our preliminary data suggest that non-HLA genetic variants associated with human T1D susceptibility confer both impaired negative selection and impaired Treg diversion of autoreactive thymocytes. By comparing T1D- and healthy control (HC)-derived PI mice, we have identified abnormalities in the thymic selection of two structural varieties of TCR in T1D immune systems. One (Type I) is characterized by longer than average CDR3βs with more hydrophobic amino acids and higher self-affinity and is normally most abundant in the Treg subset. Inserting a high affinity Type I insulin-reactive TCR (Clone 5) into HC HSCs resulted in clonal deletion or Treg diversion of thymocytes with this TCR. However, both processes are defective in T1D immune systems, resulting in entry of the islet autoreactive T cells into the peripheral repertoire only in T1D-derived immune systems. The second (Type II) is characterized by shorter CDR3β regions with few hydrophobic amino acids and shows greater islet autoreactive clonal survival in T1D compared to HC immune systems. Single cell RNAseq identified a major thymocyte population undergoing negative selection that highly expresses TCR signaling and pro-apoptotic genes. This cluster was virtually absent among T1D thymocytes, indicating a profound defect in the distal signaling pathways regulating thymic deletion. This impairment was associated with T1D susceptibility single nucleotide polymorphisms in SH2B3 and Erk/MAP kinase pathway genes. The overall goal of this proposal is to better understand the genetic control and abnormalities in thymic selection involved in the development of the autoimmune repertoires that mediate T1D. We propose to: 1): Further characterize defective thymocyte selection and identify the HSC-intrinsic, non-HLA genetically- determined TCR signaling defects in T1D that result in impaired negative selection and impaired Treg differentiation. We will also define pathogenic roles of Type I and Type II TCRs in T1D immune systems with autologous iPSC-derived β-cell grafts; 2) Utilize transgenic autoreactive TCRs and specific antigen-MHC tetramers to directly assess the extent of aberrant negative selection and Treg redirection in T1D-derived immune systems. Collectively, these studies will lead to novel insights into the mechanisms by which HLA and non-HLA risk alleles predispose to thymic selection of a T cell repertoire that causes islet autoimmunity.