Autoantibody based prediction of type 1A diabetes provides the basis for studies of the development of type 1 diabetes, the search for initiating environmental factors, and for preventive trials. Many children, depending upon HLA type, develop autoantibodies within the first 9 months of life and we have recent evidence that transient expression of """"""""autoantibodies"""""""" can occur but is infrequent. The best current predictor of type 1A diabetes is the expression of multiple anti-islet autoantibodies (""""""""combinatorial prediction""""""""). Though autoantibody assays have improved we have preliminary data that their diagnostic, prognostic, and pathogenic information can be further improved and improvements will be important for disease prediction and studies of disease pathogenesis. In the present application we will validate the utility of ICA5I2 epitope fragment assays, develop similar assays for GAD65 autoantibodies, develop rapid (less than 2 hours) autoantibody assays, attempt to develop higher sensitivity anti-insulin autoantibody assays and test the hypothesis that a subset of anti-insulin autoantibody negative patients express anti-insulin autoantibodies, and further evaluate genetic determinants of antibody expression. In particular we have combined in what we believe to be a relatively unique set of studies, a phage display library to define non-overlapping epitopes of 1CA512 with in vitro transcription and translation of identified epitope fragments, to develop epitope fragment specific radioassays for anti-islet autoantibodies. Our preliminary data suggests that using epitope information we can distinguish autoantibodies of normal controls and relatives with transient expression of autoantibodies from ICA512 autoantibodies associated with disease. Though """"""""non-diabetes"""""""" ICA5I 2 autoantibodies are infrequent, given the low risk of progression to diabetes, by Bayes' theorem identification of false positives at a potential 1 percent level can have a major influence on positive predictive value and diagnostic accuracy. We will apply the phage strategy used successfully with 1CA512 to the GAD65 molecule. In addition we believe that the insulin autoantibody assay can be improved and assays with increased sensitivity may provide particularly important information. We will utilize affinity purification of anti-insulin autoantibodies to identify sera that though negative with standard assays may be anti-insulin autoantibody positive and attempt to develop an oligonucleotide insulin hybrid PCR based assay. We have preliminary data that rapid anti-islet autoantibody assays are possible and in the current application will validate such assays and then obtain preliminary data from stored sera of past organ donors relative to the percentage of organ donors who are autoantibody positive. A long-term goal is to aid in establishing a national resource to obtain pancreas from individuals who had a high risk of developing type 1 diabetes, but became an organ donor. Pancreas from such individuals would allow histologic analysis, studies of disease pathogenesis (e.g. TCR usage and virology), and we believe most importantly T cell cloning from diseased islets.
Showing the most recent 10 out of 234 publications
