In the current proposal we will evaluate two hypotheses; that the major epitope of the insulin molecule recognized by """"""""all"""""""" prediabetic anti-insulin autoantibodies is critically dependent upon residues Bl to B3 and A11 to A13, and that the prevalence and level of anti-insulin autoantibodies is specifically determined by a gene or genes of HLA-DR4 haplotypes. The proposal is based upon preliminary results indicating that insulin autoantibodies correlate with the rate of progression to overt diabetes amongst ICA-positive first degree relatives and with expression of HLA-DR4 and studies indicating that the epitope of the insulin molecule recognized by insulin autoantibodies of high risk relatives is distinct from the receptor binding domain and is not influenced by removal of the B terminal 8 amino acids. The epitope is dependent on both A and B chain residues and in particular those residues altered in fish and guinea pig insulin and can be markedly altered by a single amino acid substitution (Al3 leucine to tryptophan) . In addition all insulin autoantibody positive sera we have studied react as well if not better with human proinsulin as compared to insulin. Modification of proinsulin by site directed mutagenesis will be utilized to define the exact epitope responsible for autoantibody binding. A larger series of relatives of patients with Type I diabetes will be typed for class II alleles and have insulin autoantibody and ICA analysis. The HLA haplotypes associated with Type I diabetes will be further defined with analysis of DQ and DR subtypes (and depending on results DP subtyping) and sequencing of class II genes of specific relatives.
Showing the most recent 10 out of 20 publications
