Chronic hyperglycemia as a cause of impaired insulin secretion
In 1948, Lukens and Dohan administered large doses of glucose to normal cats and induced permanent hyperglycemia, hydropic degeneration of the islet of Langerhans, and ketonuria in 4 out of 35 cats studied . These investigators proposed that hyperglycemia could play a role in the pathogenesis of diabetes. In recent years, several approaches have been used to directly examine the harmful effects of chronic hyperglycemia on insulin secretion. These studies have established that chronic hyperglycemia impairs β-cell responsiveness to glucose although the exact biochemical mechanism(s) mediating this effect remain to be clearly defined.
Insulin secretion during experimental hyperglycemia
In normal subjects, β-cell hyperresponsiveness and α-cell hyporesponsiveness is observed after mild hyperglycemia induced by glucose infusions . However, exposure to higher glucose concentrations (12.6mmol L−1) for 68 hours has been reported to be associated with significant reduction in insulin secretion in humans . As discussed under section “Clinical significance” later, treatment studies in patients with type 2 diabetes have provided additional support for the idea that chronic hyperglycemia impairs insulin secretion also in humans. Since the early experiments in cats , the ability of chronic hyperglycemia to desensitize the β cell to glucose has been convincingly documented in animal models. In the partially pancreatectomized dog, maintenance of plasma glucose at greater or equal to 14 mmol L−1 for two weeks induces persistent hyperglycemia, loss of glucose induced insulin secretion, ketonuria, and weight loss . Morphometric analysis of the endocrine pancreas revealed a profound reduction in the number and size of islets. These changes were not observed in portions of pancreases removed prior to the glucose infusions or in pancreases of similarly pancreatectomized dogs not subjected to hyperglycemia. Leahy andWeir infused normal rats in vivo with various concentrations of glucose and measured the insulin response thereafter in vitro using the isolated perfused pancreas . Rats infused with 35% glucose showed a severely blunted insulin response to glucose after 48 hours, and in rats infused with 50% glucose for 48 hours, the glucose-induced insulin response was totally lost. Addition of phlorizin to the 50% glucose infusion after 48 hours for an additional 48-hour period completely restored the insulin response . Similar results were obtained by Rossetti et al. who treated mildly hyperglycemic rats with a continuous infusion of phlorizin for 4 weeks . The insulin responses to arginine, glucose, and arginine plus glucose were determined in vitro using the hyperglycemic clamp technique. Normalization of glycemia by phlorizin normalized all aspects of abnormal insulin secretion, that is, the subnormal first and second phase insulin responses to glucose, the increased insulin response to arginine, and the potentiation of insulin secretion by hyperglycemia during infusion of arginine  (Figure 27.4). Correction of hyperglycemia with phlorizin has also been shown to restore glucose-induced suppression of glucagon release in insulin-deficient alloxan diabetic dogs .
Molecular mechanisms of glucose toxicity
Carbons from ingested glucose are incorporated into nearly every class of biologic metabolite and macromolecule, and the metabolism of those molecules generates energy, oxidant byproducts, and inflammatory stress, all of which could contribute to glucose-induced insulin resistance. The complexity of the pathways of insulin signal transduction and its interactions with numerous other regulatory networks suggests that glucose toxicity is unlikely to operate through a single mechanism, and the data are consistent with that presumption. Two leading candidates for the molecular mediators of glucose toxicity are the hexosamine/O-linked glycosylation pathway and redox signaling pathways, and both of these are supported by extensive experimental data that will be summarized in this section. These are unlikely to be the only mechanisms for glucose toxicity, however. Excess glucose is metabolized to lipid, for example, and there is no question that excess lipid fluxes through mechanisms involving activation of protein kinase C, ceramide synthesis, accumulation of incompletely metabolized intermediates, and others can contribute to insulin resistance and β-cell failure in diabetes. For consideration of these pathways and mechanisms of vascular complications of diabetes the reader is referred to the other chapters in this volume and several excellent recent reviews [39,40]. We will concentrate this review on the more direct products of glucose metabolism that contribute specifically to the metabolic abnormalities of diabetes.