No abstract

Our previous finding that about 15% of newly diagnosed patients with Type 1 (insulin-dependent) diabetes mellitus had human cytomegalovirus genome in their lymphocytes and islet cell autoantibodies in their sera, suggests that autoimmune Type 1 diabetes is associated with persistent cytomegalovirus infection under certain circumstances. This investigation was initiated to see if cytomegalovirus can induce islet cell autoantibodies and if the autoantibodies react with any specific islet protein(s). Monoclonal antibodies were generated after immunizing Balb/c mice with human cytomegalovirus. When these monoclonal antibodies were tested for the presence of islet cell antibodies were tested for the presence of islet cell antibodies, one (MCMVA-51) of 13 monoclonal antibodies reacted strongly with the islets. The titer of islet cell antibodies was 1:2000. When this monoclonal antibody was reacted with the proteins from the solubilized fraction of human pancreatic islets using the western immunoblotting technique, a band with a molecular weight of 38 kilodalton was detected. The 38 kilodalton band was not observed when the monoclonal antibody was reacted with the proteins prepared from pancreatic islet tissues of rats and mice or from other human organs including stomach, liver, spleen and brain, indicating that the 38 kilodalton protein is human islet cell-specific. It is concluded that human cytomegalovirus can induce islet cell antibodies that react with a 38 kilodalton human islet cell protein and that this protein component may represent islet cell-specific target antigens associated with persistent cytomegalovirus infection.

Diabetes-prone biobreeding (BB) rats often develop lymphocytic thyroiditis. Intraperitoneal administration of silica to young BB rats (40-days-old) nearly completely prevented the development of lymphocytic thyroiditis as well as insulitis. Since silica is known to be toxic to macrophages, these data suggest that the presentation of autoantigen(s) on the specific target cells such as thyroid and pancreatic B cells by antigen-presenting cells (e.g., macrophages) would be the initial step in the development of organ-specific autoimmune diseases in diabetes-prone BB rats.

The subpopulation of lymphoid cells at the different stages of insulitis in BB rats was determined by immunohistochemical techniques with various monoclonal antibodies, including the recently developed OX41, which distinguishes macrophages from T-lymphocytes, OX19 for pan-T-lymphocytes, W3/25 for both helper T-lymphocytes and macrophages, OX8 for cytotoxic T-lymphocytes and natural killer cells, and OX12 for B-lymphocytes. The major population of infiltrated cells found during the early stages of insulitis appeared to be macrophages. This preceded invasion by a mixed population of cells, including both T- and B-lymphocytes and/or natural killer cells. The preferential infiltration of macrophages during the early stages of insulitis strongly suggested that there might be an initial change in the target beta-cells that precedes their immune destruction, although the amplification of immune response by activated T-lymphocytes and natural killer cells at a later stage seemed to be required for the clinical expression of the disease.