The adverse effects of long-term exposure to NEFA on glucose-stimulated insulin secretion are well established . Much less is known about the consequences of a chronic elevation of NEFA on the release of somatostatin and glucagon. Here we demonstrate a dichotomy in the effects of long-term exposure to NEFA. Thus, whereas glucose-induced release of somatostatin, like that of insulin, was reduced following 72 h exposure to the NEFA, glucagon secretion was stimulated. In addition, the ability of high glucose concentrations to suppress glucagon release was abolished.
Following long-term exposure to the lipids, 80% to 90% of beta cells displayed signs of lipid deposition inside the cells in the form of lipid droplets and crescent-like structures, respectively. Many of the crescent-like structures were (at least partially) membrane-delimited, whereas the lipid droplets observed following exposure to oleate were without boundary membrane. The crescent-like structures might represent expansions of the endoplasmic reticulum , possibly an early sign of endoplasmic reticulum stress. Importantly, the ultrastructure of the non-beta cells was much less affected by long-term exposure to lipids, with the crescent-like structures being confined to the beta cells. Thus, alpha and delta cells would be expected to be less sensitive to lipids than beta cells and this may contribute to the hormonal abnormalities associated with obesity and diabetes.
The regulation of somatostatin secretion has traditionally been regarded to be similar to that of insulin secretion. Recently, however, it has become apparent that the mechanisms controlling glucose-induced insulin and somatostatin secretion are quite different. Thus, somatostatin release is highly dependent on Ca2+-induced Ca2+ release (CICR) . It is possible that long-term exposure to NEFA might interfere with CICR and thus lower glucose-induced somatostatin secretion.
There have been few reports on the long-term effects of glucose and NEFA on glucagon secretion. Rat islets exposed for 24 h to 0.6 mmol/l palmitate have been reported to have 50% lower glucagon content than controls . It has been argued that this effect is the result of insufficient biosynthesis in the face of higher secretion . This may explain why glucagon content decreased following long-term exposure to NEFA (30–40%), a condition that led to stimulation of glucagon secretion. Conversely, the observed oversecretion of glucagon in islets cultured in the presence of NEFA cannot be accounted for by increased islet glucagon content. Thus, it seems justifiable to conclude that this stimulation results from a genuine effect on intracellular signalling within the alpha cell.
Glucagon secretion has been proposed to be under paracrine control. According to this concept, the reduction of insulin (and other factors co-released with insulin such as Zn2+) and somatostatin may underlie the hypersecretion of glucagon. However, it should be noted that culture at high glucose alone has only a marginal effect on glucose-induced insulin and somatostatin secretion and yet the suppressor effect of high glucose on glucagon secretion is abolished. Thus, the observed stimulation of glucagon release cannot be explained in terms of defective paracrine regulation and rather suggests that the defect occurs within the alpha cell itself. In this context it may be of relevance that NEFA metabolites like palmitoyl-CoA activates KATP-channels . Recent data indicate that low concentrations of the KATP-channel opener diazoxide stimulate glucagon secretion and antagonise the inhibitory effect of glucose . If enough acyl-CoAs are generated in alpha cells to exert a diazoxide-like effect, it may contribute to the loss of glucose regulation of glucagon secretion. The long-term effects of NEFA are rather similar to the acute actions previously reported . Under acute conditions, application of NEFA did in fact exert a diazoxide-like effect on alpha cell [Ca2+]i and lead to the reappearance of [Ca2+]i oscillations in cells where they had been suppressed by high glucose.
In mouse pancreatic beta cells, long-term exposure to lipids does not result in any obvious functional impairment that explains the strong reduction of glucose-induced insulin secretion. Thus, metabolism, electrical activity and intracellular [Ca2+]i signalling all proceed as well as or better than in control cells and much of the effect on glucose induced insulin secretion is independent of cytotoxicity . Instead, we proposed that lipids may interfere with insulin release at the level of fusion pore expansion (i.e. distal to the fusion of the granules with the plasma membrane), resulting in failure of the fusion pore to expand sufficiently to allow macromolecules like insulin to exit. The fact that glucagon secretion is enhanced rather than inhibited following lipid exposure suggests that there is some cell-specificity of the effect.
Oversecretion of glucagon and insufficient release of insulin are part of the aetiology of type 2 diabetes. The data presented here illustrate that the two defects may be interrelated. Thus, mild disturbances of insulin and glucagon secretion and/or a slight elevation of plasma NEFA levels can be envisaged to result in a vicious cycle of progressively impaired glucose regulation of insulin and glucagon release that eventually culminates in glucose intolerance and overt diabetes.