Aims. Type 1 diabetes (T1D) results from the autoimmune destruction of pancreatic ß-cells. The HLA DQB1*0302 class II HLA gene carry the highest risk for T1D, but class I MHC alleles modulate the risk, in particular HLA-A*0201. Among ß-cell autoantigens, insulin has been ascribed a key role in T1D. While this knowledge has boosted efforts to prevent T1D using insulin-specific immunotherapy, these efforts have been unsuccessful when translated into clinical trials, pointing to the need for new preclinical T1D models. The aims are to: 1) characterize the immune and metabolic phenotypes of humanized mice (YES) expressing human insulin (hPPI), HLA-A*0201, DQ8 transgenes and lacking their mouse counterparts, 2) evaluate YES mice as a new preclinical model for insulitis and diabetes, 3) characterize hPPI epitopes targeted by CD4+ and CD8+ T cells and develop T-cell assays in these mice, 4) evaluate strategies to induce immune tolerance to hPPI as a therapy in T1D. Methodology. We have obtained YES mice that lack the expression of murine class I, class II and insulin genes, express HLA-A*0201, DQ8 transgenes and human insulin (hINS) transgenes. Analysis of pancreases from YES mice shows normal islet morphology, but we have preliminary evidence that YES mice develop insulitis following immunization against a human ß cell line. To characterize the immune and metabolic phenotypes of YES mice, three YES lines will be studied. As the replacement of murine class I and class II MHC molecules may interfere with the selection of class I- and class II-restricted T cells and with antigen presentation, the main immune subsets, in particular T lymphocytes, will be quantified by FACS in thymus, spleen, lymph nodes and pancreas. Cytotoxic and proliferative responses to conventional antigens will be evaluated. Insulin gene expression will be quantified in liver, pancreas, thymus, spleen, kidney, inguinal lymph node and the metabolic phenotype of YES mice further characterized and compared to that of the parental mouse lines used. To evaluate YES mice as a new preclinical model for insulitis and diabetes, attempts will be made to characterize further and optimize the induction of insulitis and immune-mediated ß cell destruction using different immunization procedures and introduction of B7.1 expression on ß-cells. To characterize hPPI epitopes targeted by autoimmune CD4+ and CD8+ T cells and apply T-cell assays to detect ß-cell autoimmunity in this model, DQ8-restricted CD4+ and A*0201-restricted CD8+ T cell responses to human insulin will be evaluated and the repertoire of epitopes recognized defined. Immunogenicity of human insulin peptides and recognition by CD4+ T cells will be studied using proliferation assays and evaluating production of cytokines. Insulin-specific CD8+ T cells will be studied – immunization against a library of insulin peptides, cytotoxicity assays, single cell PCR of tetramer-sorted cells. T cell clones will be generated and tested for recognition of P815 cells cotransfected with HLA-A*0201 and hINS. The same T cell assays will be used to study insulin-specific CD8+ and CD4+ T cells along induction of insulitis/diabetes. To evaluate the feasibility of inducing immune tolerance to insulin, and more generally ß-cells, in this model, as preventive or therapeutic strategies in T1D, we will assess whether delivery of a human insulin-Fc fusion protein can induce immune tolerance in YES mice, and identify immune correlates and mechanisms of diabetes protection. Insulin-Fc fusion proteins will be delivered to YES offsprings upon treatment of mothers during pregnancy or lactation, or to YES mice at 3 week of age or following induction of insulitis/diabetes. Mice will be evaluated for induction of T cell tolerance to human insulin and for prevention of insulitis/diabetes. Setting up a new preclinical T1D model may help the development of T-cell assays and pave the way to “vaccination” strategies in T1D.