GLUCOSE REFRACTORINESS OF PANCREATIC BETA-CELLS IN RAT MODELS OF NON-INSULIN-DEPENDENT DIABETES

A decreased insulin response, preferentially to glucose, has been cons idered a hallmark of non-insulin dependent diabetes mellitus (Type 2) in humans. Syndromes resembling human diabetes occur spontaneously in many animal species and can also be induced by treating animals with d rugs or viruses, excising their pancreases or manipulating their diet. Among these models, rat diabetes induced by neonatal streptozotocin a dministration (n-STZ models) has been first recognized as an adequate tool to study the long-term consequences of a gradually reduced beta-c ell mass. More recently, the GK (Goto Kakisaki) Wistar rat has become available and is now considered as a promising spontaneous rat model o f non-insulin dependent diabetes. We and others have found that defect s in insulin secretion and action develop in the n-STZ and the GK mode ls. which in many ways resemble those described in human noninsulin de pendent diabetes. This review is aimed to sum up with a comparative ap proach, the informations so far collected in the n-STZ and GK models c oncerning the cellular mechanisms leading to the desensitization of th eir beta-cells to glucose. Taken together, the data reinforce the view that the impairment of glucose-induced insulin release in n-STZ and G K rats is clearly related to a defect in oxidative glycolysis. This le nds to a severe decrease in the mitochondrial oxidative catabolism of glucose-derived pyruvate. its coincides with a lower ATP/ADP ratio in glucose-stimulated islets and a subsequent alteration of ionic events tightly coupled to the fuel function of the hexose in islet cells, i.e . the decrease in K+ conductance. Such a coherent conclusion is not me ant, however, to rule out other subtle anomalies in the islets of n-ST Z and/or GK rats. Indeed we have identified in the n0-STZ model a defe ctive glucose-induced hydrolysis of beta-cell membrane inositol phosph olipids which could be explained by a block in the signal that links g lucose metabolism and the activation of phospholipase C. Other hypothe sis in addition to those already discussed, have been suggested to par ticipate to the pathophysiology of insulin release in NIDDM (disturban ces in the local production of cyclic AMP, prostaglandin E(2) or opioi d peptides) and need to be tested in the n-STZ and GK models.