Banting Lecture: Glucose Turnover: A Key to Understanding the Pathogenesis of Diabetes (Indirect Effects of Insulin)

This article is divided into two parts. A retrospective overview summarizes some of the work that provided the framework and tools of the more recent studies. The five novel areas of research are related to the indirect effects of insulin. Regulation of plasma glucose is of central importance in health and diabetes. Understanding this precise regulation requires sensitive isotope dilution methods that can measure the rates at which glucose is produced by the liver and used by the tissues on a minute-to-minute basis. Validation studies indicated that the non-steady-state tracer method yields reasonable results when the specific activity of plasma glucose does not change abruptly. During hyperinsulinemic glucose clamps, the decrease in specific activity of glucose can be prevented by the MSTI. During exercise, the decrease of specific activity can be only in part ameliorated by step-tracer infusion. Depancreatized dogs are used extensively as a model of selective insulin deficiency, because dog stomach secretes physiological amounts of glucagon. This strategy can avoid injections of somatostatin, which can have other affects in addition to the suppression of insulin and glucagon. In human diabetes, in addition to an increase of glucose production, there is also an increase in glucose cycling in the liver. In animal models of diabetes, mild NIDDM, and in glucose intolerance, the percentage of increments of glucose cycling are much larger than those of glucose production. We hypothesize, therefore, that measurements of glucose cycling can be used as an early marker of glucose intolerance. Application of different tracer strategies and use of the depancreatized dog as a model of diabetes, we investigated the importance of the indirect effects of insulin in the pathogenesis of diabetes. 1) Because, in the treatment of IDDM, insulin is administered by the peripheral routes we compared the relative importance of hepatic and peripheral effects of insulin in regulating the rate of glucose production. Experiments were performed in depancreatized dogs that were initially maintained at moderate hyperglycemia (10 mM) with subbasal portal insulin infusion. During the experimental period, insulin was infused either peripherally or portally at 0.9 mU · kg⁻¹ · min⁻¹. In addition, peripheral infusions were also given at 0.45 mU · kg⁻¹ · min⁻¹ We concluded that when suprabasal insulin levels are provided to moderately hyperglycemic depancreatized dogs, the suppression of glucose production is more dependent on peripheral than portal insulin concentrations. This indirect effect of insulin may be mediated by limitation of the flow of precursors and energy substrates for gluconeogenesis and/or by suppressive effect of insulin on glucagon secretion. These results suggest that the absence of a portal-peripheral gradient in insulin-treated diabetic subjects may not be important for postprandial suppression of glucose production. 2) The glucagon-insulin ratio is an important regulator of glucose production by the liver during moderate exercise, whereas during intense exercise the catecholamines play a prominent role. Regulation of glucose uptake during exercise is very complex. In vivo, insulin can play an indirect role by inhibiting the FFA-glucose cycle and by maintaining normoglycemia; both of these factors influence glucose uptake by the muscle. 3) Streptozocin-induced diabetes in rats decreases the number of glucose transporters when measured both by cytochalasin B binding and by assessment of GLUT4 transporters. Normalization of glycemia in the diabetic rats by a 2-day phlorizin treatment, which does not affect the insulin concentration, normalizes glucose transporter number in the plasma membrane. We concluded that hyperglycemia, per se, plays an important role in regulating glucose transporter number in the muscle. 4) In alloxan-induced diabetic dogs, similarly to IDDM, glucagon's response and, therefore, the response of glucose production to declining glucose is impaired. We present a new hypothesis that an increased ratio between somatostatin and glucagon in the residual diabetic islets of diabetic dogs can, at least in part, explain the lack of glucagon response to hypoglycemia. 5) To investigate the effect of stress in diabetic animals, we used an intracerebroventricular injection of a small amount of carbachol. This compound increases the release of all counterregulatory hormones, but without affecting insulin secretion. Surprisingly, this release of counterregulatory hormones induces only a marginal change in plasma glucose, because increased glucose production is matched by a similar increase in glucose uptake. In contrast, in hyperglycemic diabetic dogs, the same carbachol injection induces a sevenfold larger increment in plasma glucose. This occurred because the metabolic clearance rate of glucose does not increase. We therefore postulated a neural mechanism that controls peripheral glucose uptake and does not require a regulatory effect of insulin. A permissive role of insulin could be to maintain normal insulin sensitivity in the muscle. From the point of view of the clinician, if diabetic neuropathy affects this pathway, this would contribute to insulin resistance. However, even more important is the possibility that this pathway may take part in the pathogenesis of diabetes.