Potential mechanisms by which certain foods promote or inhibit the development of spontaneous diabetes in BB rats: dose, timing, early effect on islet area, and switch in infiltrate from Th1 to Th2 cells.

Certain diets can have major effects on the development of IDDM in DP-BB rats, but data are scant on the timing, dose, and mechanisms involved. We therefore determined the dose response, timing, and duration of exposure required to induce diabetes, and characterized the effects of nutritionally adequate diets with widely different diabetogenicity on the pancreatic islet area and cytokines. DP-BB rats were fed a diabetogenic, cereal-based, NIH-07 (NIH) diet or a protective, casein or hydrolyzed casein (HC)-based, semipurified diet. Rats were fed from weaning to 50 or 100 days with the HC diet and then switched to the NIH diet, or fed the NIH diet from weaning to 50 days and switched to the HC diet. Pancreas histology and diabetes outcome were determined. Semiquantitative morphometric analyses of hematoxylin and eosin-stained sections of pancreas from 41-day-old rats were also carried out. Diet-induced effects on pancreatic cytokine levels were measured at 70 days using reverse transcriptase-polymerase chain reaction analysis of gamma-interferon (IFN-gamma), interleukin-10 (IL-10), and transforming growth factor-beta (TGF-beta). Long-term daily exposure, particularly around the beginning of puberty to late adolescence (50-100 days), was important for development of diabetes. DP-BB rats could be rescued from diabetes development by feeding them a low-diabetogen HC diet as late as 50 days. Diabetes frequency was highest in rats fed 70% and 100% NIH diets. By age 41 days, before classic insulitis, the islet area in HC-fed DP-BB rats was 65% greater than in NIH-fed rats. By 70 days, when mononuclear cells were visible in the islets of most NIH-fed, but not HC-fed rats, the more pronounced inflammatory process in NIH-fed rats was associated with a Th1 cytokine pattern (high IFN-gamma and low IL-10 and TGF-beta), whereas the pancreases of HC-fed rats showed fewer infiltrating cells, low levels of IFN-gamma, and high levels of TGF-beta, typical of a Th2 cytokine pattern. Thus dietary modification can occur as late as puberty. Further, long-term exposure to sufficient amounts of food diabetogens between 50 and 100 days was required for maximum diabetes induction. The islet area was modified by diet before signs of classic insulitis. Pancreatic inflammation in NIH-fed animals is a Th1-dependent phenomenon. The HC diet inhibited insulitis and was associated with a Th2 cytokine pattern in the pancreas, protecting diabetes-prone rats from developing diabetes.