This project continues the general theme of the program project and contributes to the program by 1. identifying DQ8 binding peptides and 2. examining the peptide specificity of islet infiltrating T cells. The project has strong collaborative interactions with Project 1 (Structure of DQ8/peptide complexes) and with Project 2 (DQ8 transgenic mouse model). It has been very difficult to identify the MHC/peptide specificity of tissue infiltrating T cells because specific probes have not been available. The goal of this project is to generate such probes and to define the MHC/peptide specificity of T cells in early islet infiltrates. The approach is to express multivalent, soluble MHC class II molecules and to load these molecules with defined peptides. Multivalent molecules are used because the affinity of monovalent MHC/peptide complexes for the TCR is too low. Classical studies with IgG and IgM antibodies have demonstrated that the 'functional affinity' of antibody binding to antigen is greatly enhanced by multivalent attachment. HLA-DQ8 is associated with the highest risk for the development of insulin dependent diabetes. Multivalent DQ8 molecules are expressed in the Drosophila Schneider cell system using three different designs (DQ8-IgG and DQ8-IgM fusion proteins, DQ8/streptavidin tetramers). These designs differ by the number of MHC/peptide arms per complex (2, 4 and 10) and by the presence of effector domains (Fc portion of IgG or IgM which fix complement). We will evaluate which design is most effective 1. for the in situ detection of cells and 2. for the depletion of peptide specific T cells. Experiments with a human DQ8 restricted T cell clone will examine the affinity of these probes for the TCR as well as the mechanism(s) by which target cells are killed following attachment of multivalent DQ8/peptide complexes. Little is presently known about peptide binding to DQ8, despite its importance in the pathogenesis of diabetes. DQ8 binding peptides will be identified based on an analysis of the DQ8 peptide binding motif. T cell responses to these peptides will be examined in collaboration with Dr. Lipes. The peptide specificity of T cells in early islet infiltrates will be examined in NOD and in DQ8-transgenic NOD mice, using multivalent MHC/peptide complexes for in situ detection. In situ analysis with multivalent MHC/peptide complexes may prove to be a powerful approach for characterizing the specificity of T cells in early autoimmune lesions.