Chronic hyperglycemia causes a near-total disappearance of glucose-induced insulin secretion. To determine if glucose potentiation of nonglucose secretagogues is impaired, insulin responses to 10 Ϫ 9 M glucagonlike peptide-1 (GLP-1) (7-37) were measured at 2.8, 8.3, and 16.7 mM glucose with the in vitro perfused pancreas in rats 4-6 wk after 90% pancreatectomy (Px) and sham-operated controls. In the controls, insulin output to GLP-1 was Ͼ 100-fold greater at 16.7 mM glucose versus 2.8 mM glucose. In contrast, the increase was less than threefold in Px, reaching an insulin response at 16.7 mM glucose that was 10 Ϯ 2% of the controls, well below the predicted 35-40% fractional ␤ -cell mass in these rats. Px and control rats then underwent a 40-h fast followed by pancreas perfusion using a protocol of 20 min at 16.7 mM glucose followed by 15 min at 16.7 mM glucose/ 10 Ϫ 9 M GLP-1. In control rats, fasting suppressed insulin release to high glucose (by 90%) and to GLP-1 (by 60%) without changing the pancreatic insulin content. In contrast, in Px the insulin response to GLP-1 tripled in association with a threefold increase of the insulin content, both now being twice normal when stratified for the fractional ␤ -cell mass. The mechanism of the increased pancreas insulin content was investigated by assessing islet glucose metabolism and proinsulin biosynthesis. In controls with fasting, both fell 30-50%. In Px, the degree of suppression with fasting was similar, but the attained levels both exceeded those of the controls because of higher baseline (nonfasted) values.In summary, chronic hyperglycemia is associated with a fasting-induced paradoxical increase in glucose-potentiated insulin secretion. In Px rats, the mechanism is an increase in the ␤ -cell insulin stores, which suggests a causative role for a lowered ␤ -cell insulin content in the im-

No abstract

Glucokinase is the beta-cell glucose sensor, i.e., the site in glucose metabolism that determines the glucose set-point (sensitivity) for insulin secretion. Hexokinase is also present, but it normally contributes little to glucose metabolism because of end-product inhibition by glucose 6-phosphate. There is a lowered glucose set-point for insulin secretion in 90% pancreatectomized (Px) diabetic rats. We investigated the mechanism by measuring hexokinase and glucokinase activity in islet extracts. Glucokinase activity was minimally raised in Px islets (Vmax 125% of sham-operated control rats). In contrast, hexokinase Vmax was 250% of the control value, suggesting that the increased hexokinase activity caused the beta-cell glucose hypersensitivity. Additional evidence was obtained with a 40-h fast that was performed because of a previous observation that the inhibitory effect of fasting on insulin secretion was impaired in Px rats. Glucokinase activity fell normally in the Px rats (32 +/- 4% reduction in sham vs. 37 +/- 4% in Px rats) as opposed to hexokinase activity, which was unaffected in either group. In summary, a feature of hyperglycemia is upregulated islet hexokinase activity. The result is that hexokinase assumes partial control over the glucose set-point for insulin secretion. As such, regulatory effects on insulin secretion, such as fasting, that are mediated through glucokinase activity may be altered.

The overworked-beta-cell hypothesis proposes that lowered glucose-potentiated insulin secretory responses in diabetes are secondary to hyperstimulated insulin secretion and depletion of the beta-cell insulin stores. We tested this hypothesis in normal rats using a 48-h infusion of 200 mg x kg(-1) x day(-1) tolbutamide in 20% glucose. Insulin secretion was measured by in vitro pancreas perfusion. Twice daily blood glucose values were equal in the tolbutamide-infused and control rats. Pancreas insulin content was 47 +/- 7% that of the controls (P < 0.004). Insulin responses to 16.7 mmol/l glucose, 16.7 mmol/l glucose/10 mmol/l arginine, and 5.5 mmol/l glucose/10 mmol/l arginine were reduced in parallel, except for the first phase response to 16.7 mmol/l glucose/arginine. Pancreas amylin content was unchanged in the tolbutamide-infused rats as was amylin secretion, resulting in higher than normal stored and secreted amylin-to-insulin molar ratios. Importantly, a raised amylin-to-insulin ratio and a relatively unimpaired first versus second phase insulin response for high glucose/arginine both occur in diabetic rats. Thus, our results support the overworked-beta-cell hypothesis by showing chronic beta-cell stimulation without hyperglycemia replicates part of the beta-cell dysfunction found with diabetes.

A 63-year-old womanwith pancreatic diabetes after a total pancreatectomy and splenectomy developed discitis of the L2/3 intervertebral disk. Rapidly she also developed infectious arthritis of the left knee joint and bacterial meningitis. Aspirate from the left knee contained Enterococcus faecalis. The diagnosis of discitis is generally difficult in the initial period of disease, and patients with diabetes or splenectomy are susceptible to rapid progression of the infection. Early diagnosis ofdiscitis using magnetic resonance imaging of the spine and treatment with antibiotics might have altered her clinical course. (Internal Medicine 36: 443-445, 1997)