We have proposed that chronic hyperglycemia alters the ability of glucose to modulate insulin secretion, and have now examined the effects of different levels of hyperglycemia on B cell function in normal rats using chronic glucose infusions. Rats weighing 220-300 g were infused with 0.45% NaCl or 20, 30, 35, or 50% glucose at 2 ml/h for 48 h, which raised the plasma glucose by 18 mg/dl in the 30% rats, 37 mg/dl in the 35% rats, and 224 mg/dl in the 50% group. Insulin secretion was then examined using the in vitro isolated perfused pancreas. Glucose-induced insulin secretion remained intact in the normoglycemic 20% glucose rats and it was potentiated in the mildly hyperglycemic 30% glucose rats. However, with even greater hyperglycemia in the 35% glucose group the insulin response to a high glucose perfusate was severely blunted, and it was totally lost in the most hyperglycemic 50% glucose rats. In a second protocol that examined glucose potentiation of arginine-stimulated insulin release, a similar impairment in the ability of glucose to modulate the insulin response to arginine was found with increasing levels of chronic hyperglycemia. On the other hand, the ability of a high glucose concentration to inhibit arginine-stimulated glucagon release was preserved in all glucose-infused rats, but the glucagon levels attained in response to the arginine at 2.8 mM glucose were much less in the 50% glucose rats than in all the other groups. These data clearly show that after 48 h of marked hyperglycemia, glucose influence upon insulin secretion in the rat is severely impaired. This model provides a relatively easy and reproducible method to study the effects of long-term hyperglycemia on B cell function.

We have previously demonstrated impaired glucose influence on insulin secretion in normal rats made overtly hyperglycemic with glucose infusions for 48 h. We examined the effects of a 96-h infusion period. Rats received 30, 35, or 50% glucose or 0.45% NaCl at 2 ml/h. The plasma glucose in the 30 and 35% rats peaked at 24 h but then returned to normal by 72 h. A peak followed by a gradual fall also occurred in the 50% rats, but significant hyperglycemia was maintained throughout. beta-Cell responsiveness to glucose was assessed with the perfused pancreas by examining the effect of altering the perfusate glucose concentration on insulin release. Insulin secretion remained intact in the 30% rats, but the response to a glucose increase was blunted in the 35% group and totally absent in the 50% rats. In a second protocol, glucose influence on arginine-stimulated insulin release was tested by adding 10 mM arginine to the perfusate and administering it with 2.8 and 16.7 mM glucose. The ability of the different glucose backgrounds to influence the insulin response decreased as the level of infused glucose rose. On the other hand, when phloridzin was added to 50% glucose during the second 48 h of infusion, glucose modulation of insulin release was completely restored. These data indicate that during a period of overt hyperglycemia, beta-cell defects can fully evolve within 48 h in the rat. Thereafter, they are maintained by the glucose infusion even though the plasma glucose level returns to near normal. Continued minor stimulation of the beta-cell may be the responsible factor.