New clues to diabetes' cause and treatment.

Type I diabetes, the most serious form of the disease, results from an abnormality in a class of immune-system proteins, a new study involving human diabetics indicates. A second study suggests that transplants of hollow fibers containing pancreas cells may one day help diabetics maintain normal blood-sugar levels without insulin.

Usually striking by early adolescence, Type I diabetes -- the insulin-dependent form of the disease -- currently afflicts roughly 1 million people in the United States. Because the pancreatic "beta" cells no longer produce enough insulin to process diet-derived sugar, these diabetics must frequently inject themselves with supplemental insulin.

In recent years, researchers have gathered evidence implicating Type I diabetes as an autoimmune disorder, in which the immune system turns against the body. They have found antibodies to beta cells in the blood of diabetics (SN: 6/18/88, p.389). People with a particular immune-system marker also prove more vulnerable to autoimmune diseases, and therefore run a higher risk of developing diabetes (SN: 6/10/89, p.357). The marker, a cell-surface protein belonging to the "major histocompatibility complex (MHC) class II," normally recognizes foreign proteins.

Denise Faustman of Massachusetts General Hospital in Boston and her colleagues now suggest that the disease might instead be associated with MHC class I proteins, which normally help the immune system discriminate between healthy body cells and those that are precancerous or infected by viruses. Blood cells from diabetic mice and people both possess lower than normal levels of MHC class I proteins, they now report in the Dec. 20 SCIENCE.

The immune system in an individual with Type I diabetes may fail to recognize its own beta cells because they are deficient in the MHC class I proteins, which tag them as "self," Faustman says. "We think [a reduction in MHC class I] is present even before autoantibodies occur," she says. If true, her team's discovery may lead to a clinically useful way to predict which children will eventually develop Type I diabetes.

The new study "should redirect interest to MHC class I as a susceptibility market for diabetes," says Mark Atkinson, a diabetes researcher at the University of Florida in Gainesville. He adds, however, that proof of MHC class I's value as a prognostic tool must await the outcome of larger studies.

In the same issue of SCIENCE, Paul E. Lacy of Washington University in St. Louis and his co-workers report developing a potential treatment for Type I diabetes. The researchers implanted permeable plastic fibers containing healthy beta cells into the bellies of rats with diabetes symptoms. Protected from the immune system by the encasing fibers, these implanted cells thrived during the 60-day study period, the reseachers report. More important, the implants produced enough insulin to regulate the rats' blood-sugar levels. When the team removed the fibers, the rats' blood-sugar levels soared back to unregulated levels.