Sulphonylurea receptor-1, sulphonylureas and amplification of insulin secretion by Epac activation in ß cells.

Amplification of insulin secretion by cyclic AMP involves activation of protein kinase A (PKA) and Epac2 in pancreatic ß cells. Recent hypotheses suggest that sulphonylurea receptor-1 (SUR1), the regulatory subunit of ATP-sensitive potassium channels, is implicated in Epac2 effects and that direct activation of Epac2 by hypoglycaemic sulphonylureas contributes to the stimulation of insulin secretion by these drugs. In the present experiments, using islets from Sur1KO mice, we show that dibutyryl-cAMP and membrane-permeant selective activators of Epac or PKA normally amplify insulin secretion in ß cells lacking SUR1. In contrast to Epac activator, sulphonylureas (glibenclamide and tolbutamide) did not increase insulin secretion in Sur1KO islets, as would be expected if they were activating Epac2 directly. Furthermore, glibenclamide and tolbutamide did not augment the amplification of insulin secretion produced by Epac activator or dibutyryl-cAMP. Collectively, the results show that SUR1 is dispensable for amplification of insulin secretion by Epac2 activation and that direct activation of Epac2 is unimportant for the action of therapeutic concentrations of sulphonylureas in ß cells.

Tolbutamide and diazoxide influence insulin secretion by changing the concentration but not the action of cytoplasmic Ca2+ in beta-cells.

Sulfonylureas stimulate insulin secretion by blocking ATP-sensitive K+ channels (K+-ATP channels) of the beta-cell membrane, thereby causing depolarization, Ca2+ influx, and rise in cytoplasmic Ca2+ concentration ([Ca2+]i), whereas diazoxide inhibits insulin secretion by opening K+-ATP channels. It has been suggested recently that these drugs also respectively increase and decrease the efficacy of Ca2+ on exocytosis. This hypothesis was tested here with intact islets or single beta-cells from normal mice. Depolarizing islet cells by raising extracellular K+ from 4.8 to 15, 30, and 60 mmol/l progressively raised [Ca2+]i and stimulated insulin secretion. The magnitude of the [Ca2+]i rise produced by a subsequent addition of 100 micromol/l tolbutamide decreased as the concentration of K+ was increased. The effect on insulin secretion paralleled that on [Ca2+]i. Similarly, the magnitudes of the [Ca2+]i drop and of the inhibition of insulin secretion produced by 250 micromol/l diazoxide were inversely related to the concentration of K+. Either drug was effective on secretion only when it increased or decreased [Ca2+]i. Exocytosis of insulin granules from single, voltage-clamped beta-cells was also studied by measuring cell capacitance changes. In the perforated patch configuration, exocytosis was evoked by depolarizing pulses. Addition of tolbutamide to the extracellular medium did not affect the Ca2+ current and the resulting change in cell capacitance. In the whole-cell configuration, cell capacitance increased with the concentration of free Ca2+ in the solution diffusing from the pipette into the cell. It was markedly potentiated by cAMP, was inhibited by activation of alpha2-adrenoceptors with clonidine, and was strongly augmented by acetylcholine. In contrast, tolbutamide was ineffective whether applied intra- or extracellularly, at low or high free Ca2+, and with or without cAMP. Diazoxide also failed to interfere directly with exocytosis. These results indicate that tolbutamide and diazoxide affect insulin secretion by changing the concentration, not the action, of Ca2+ in beta-cells.

Relative contribution of Ca2+-dependent and Ca2+-independent mechanisms to the regulation of insulin secretion by glucose.

Although insulin secretion is usually regarded as a Ca2+-dependent mechanism, recent studies have suggested the existence of a Ca2+-independent pathway of regulation by glucose. Here, mouse islets were used to compare the contribution of Ca2+-dependent and -independent pathways. Glucose increased insulin release in a concentration-dependent manner both in a control medium, when it depolarizes beta cells and raises [Ca2+]i (triggering signal), and in the presence of 30 mM K+ and diazoxide, when it does not further raise [Ca2+]i but increases its efficacy on exocytosis. Both Ca2+-dependent responses were amplified by glucagon-like peptide-1+acetylcholine, and were strongly potentiated by forskolin+PMA. Under conditions of mild or stringent Ca2+ deprivation, glucose had no effect either alone or with GLP-1 and acetylcholine, and was poorly effective even during pharmacological activation of protein kinases A and C. Similar results were obtained with rat islets. It is concluded that physiological regulation of insulin release by glucose is essentially achieved through the two Ca2+-dependent pathways without significant contribution of a Ca2+-independent mechanism.

Multiple effects and stimulation of insulin secretion by the tyrosine kinase inhibitor genistein in normal mouse islets.

1. Islets from normal mice were used to test the acute effects of genistein, a potent tyrosine kinase inhibitor, on stimulus-secretion coupling in pancreatic beta-cells. 2. Genistein produced a concentration-dependent (10-100 microM), reversible, increase of insulin release. This effect was marginal on basal release or in the presence of non-metabolized secretagogues, and much larger in the presence of glucose or other nutrients. The increase in insulin release caused by 100 microM genistein was abolished by adrenaline or omission of extracellular Ca2+. It was not accompanied by any rise of cyclic AMP, inositol phosphate or adenine nucleotide levels. 3. Although genistein slightly inhibited ATP-sensitive K+ channels, as shown by 86Rb efflux and patch-clamp experiments, this effect could not explain the action of the drug on insulin release because the latter persisted when ATP-sensitive K+ channels were all blocked by maximally effective concentrations of glucose and tolbutamide. Genistein was also effective when ATP-sensitive K+ channels were opened by diazoxide and the beta-cell membrane depolarized by 30 mM K, but ineffective in the presence of diazoxide and normal extracellular K. 4. Genistein paradoxically decreased Ca2+ influx in beta-cells, as shown by the inhibition of glucose-induced electrical activity, by the inhibition of Ca2+ currents (perforated patches) and by the lowering of cytosolic [Ca2+]i (fura-2 technique). Genistein thus increases insulin release in spite of a lowering of [Ca2+]i in beta-cells. 5. Daidzein, an analogue of genistein reported not to affect tyrosine kinases, was slightly less potent than genistein on K+ and Ca2+ channels, but increased insulin secretion in a similar way. Three other tyrosine kinase inhibitors, tyrphostin A47, herbimycin A and an analogue of erbstatin variably affected insulin secretion.6. Genistein exerts a number of heretofore unrecognized effects. The unusual mechanisms, by which genistein increases insulin release in spite of a decrease in beta-cell [Ca2+]i and without activating known signalling pathways, do not seem to result from an inhibition of tyrosine kinases.

Mechanisms of the stimulation of insulin release by saturated fatty acids. A study of palmitate effects in mouse beta-cells.

The mechanisms by which fatty acids increase insulin release are not known. In this study, mouse islets were used as a model and palmitate as a reference compound to study in vitro how saturated fatty acids influence pancreatic beta-cells. Palmitate (625 microM) was bound to albumin. It did not affect basal insulin release (3 mM glucose) but increased the release induced by 10-15 mM glucose. This effect was dependent on the concentration of free rather than total palmitate. It was reversible and abolished by epinephrine, diazoxide, nimodipine, or omission of extracellular Ca. Bromopalmitate and methyl palmoxirate, two inhibitors of fatty acid oxidation, were ineffective alone, and only bromopalmitate partially inhibited the effects of palmitate on insulin release. The increase in insulin release produced by palmitate could not be ascribed to a blockade of ATP-sensitive K(+)-channels because the fatty acid only barely decreased 86Rb efflux and did not depolarize beta-cells in 3 mM glucose. The small effect on 86Rb efflux might be attributed to a slight rise in the ATP/ADP ratio. No such rise occurred when palmitate was tested in 15 mM glucose, and the fatty acid consistently accelerated 86Rb efflux under these conditions. Measurements of beta-cell membrane potential (intracellular microelectrodes) and of free cytoplasmic calcium (Cai2+) in beta-cells (Fura 2 technique) showed that palmitate increases Ca2+ influx; it also caused a very small mobilization of intracellular Ca. The persistence of this stimulation of Ca2+ influx in the presence of diazoxide and high K+ suggests that palmitate might act on Ca2+ channels.(ABSTRACT TRUNCATED AT 250 WORDS)

Two sites of glucose control of insulin release with distinct dependence on the energy state in pancreatic B-cells.

The energy state of pancreatic B-cells may influence insulin release at several steps of stimulus-secretion coupling. By closing ATP-sensitive K+ channels (K(+)-ATP channels), a rise in the ATP/ADP ratio may regulate the membrane potential, and hence Ca2+ influx. It may also modulate the effectiveness of Ca2+ on its intracellular targets. To assess the existence of these two roles and determine their relative importance for insulin release, we tested the effects of azide, a mitochondrial poison, on mouse B-cell function under various conditions. During stimulation by glucose alone, when K(+)-ATP channels are controlled by cellular metabolism, azide caused parallel, concentration-dependent (0.5-5 mM), membrane repolarization, decrease in cytosolic Ca2+ concentration [Ca2+]i and inhibition of insulin release. When K(+)-ATP channels were closed pharmacologically (by tolbutamide in high glucose), azide did not repolarize the membrane or decrease [Ca2+]i, and was much less effective in inhibiting insulin release. A similar resistance to azide was observed when K(+)-ATP channels were opened by diazoxide, and high K+ was used to depolarize the membrane and raise [Ca2+]i. In contrast, azide similarly decreased ATP levels and increased ADP levels, thereby lowering the ATP/ADP ratio under all conditions. In conclusion, lowering the ATP/ADP ratio in B-cells can inhibit insulin release even when [Ca2+]i remains high. However, this distal step is much more resistant to a decrease in the energy state of B-cells than is the control of membrane potential by K(+)-ATP channels. Generation of the signal triggering insulin release, high [Ca2+]i, through metabolic control of membrane potential requires a higher global ATP/ADP ratio than does activation of the secretory process itself.

Vanadate stimulation of insulin release in normal mouse islets.

The effects of vanadate (Na3VO4) on pancreatic B-cell function were studied in normal mouse islets. Vanadate did not affect basal insulin release but potentiated the effect of 7-30 mM glucose at concentrations of 0.1-1 mM. This effect was progressive and slowly reversible. It was abolished by omission of extracellular Ca2+ but unaffected by blockers of adrenergic or muscarinic receptors. Comparison of the changes in membrane potential, 86Rb efflux and 45Ca efflux that vanadate and ouabain produced in B-cells made it possible to exclude the hypothesis that vanadate increases insulin release by blocking the sodium pump. Vanadate was also without effect on cAMP levels. On the other hand, it markedly changed the characteristics of the Ca(2+)-dependent electrical activity and of the oscillations of cytoplasmic Ca2+ recorded in B-cells stimulated by 15 mM glucose. In the steady state, Ca2+ influx was increased by vanadate, and this resulted in a rise in cytoplasmic Ca2+. The exact mechanisms underlying these changes could not be established but a blockade of K channels was excluded. In the presence of LiCl, vanadate markedly increased inositol phosphate levels in islet cells. This effect was attenuated but not suppressed by omission of Ca2+. A small increase in inositol bisphosphate was still produced by vanadate in the absence of LiCl. These results suggest that vanadate both stimulates phosphoinositide breakdown and inhibits inositol phosphate degradation. In conclusion, vanadate does not induce insulin release, but markedly potentiates the stimulation by glucose. This property is not due to an inhibition of the sodium pump or to a rise in cAMP concentration. It results from a complex interplay between changes in B-cell membrane potential, phosphoinositide metabolism and Ca2+ handling.

Mechanisms of the stimulation of insulin release by arginine-vasopressin in normal mouse islets.

The mechanisms by which arginine-vasopressin (AVP) affects pancreatic B-cell function were studied in normal mouse islets. AVP produced a dose-dependent (0.1-1000 nM; EC50 approximately 1-2 nM) amplification of glucose-induced insulin release. This amplification was of slow onset and reversibility. AVP was ineffective when the concentration of glucose was less than 7 mM, but was still very effective in 30 mM glucose. The increase in insulin release produced by AVP was accompanied by small accelerations of 86Rb and 45Ca efflux from islet cells. Omission of extracellular Ca2+ accentuated the effect of AVP on 86Rb efflux, attenuated that on 45Ca efflux, and abolished that on release. Under no condition did AVP inhibit 86Rb efflux. AVP did not significantly affect cAMP levels, but increased inositol phosphate levels in islet cells, even in the absence of extracellular Ca2+. AVP did not affect the membrane potential in unstimulated B-cells and augmented glucose-induced electrical activity only slightly. This was not due to a direct action on ATP-sensitive K+ channels as revealed by patch-clamp recordings (whole cell and outside-out patches). In conclusion, AVP is not an initiator of insulin release, but it potently amplifies glucose-induced insulin release in normal mouse B-cells. This effect involves a stimulation of phosphoinositide metabolism, and presumably an activation of protein kinase C, rather than a change in cAMP levels or a direct control of the membrane potential.

Galanin and epinephrine act on distinct receptors to inhibit insulin release by the same mechanisms including an increase in K+ permeability of the B-cell membrane.

The mechanisms by which galanin and epinephrine affect pancreatic B-cell function were studied in normal mouse islets. In the presence of 15 mM glucose and 2.5 mM Ca2+, galanin (50 nM) and epinephrine (100 nM) hyperpolarized the B-cell membrane and suppressed electrical activity only transiently. These changes were accompanied by a decrease in 86Rb+ efflux from islet cells and nearly complete inhibition of insulin release. Both agents also decreased 86Rb+ efflux in the absence of Ca2+. Low concentrations (10-15 microM) of diazoxide, an activator of ATP-sensitive K+ channels, mimicked some effects of galanin and epinephrine. However, insulin release was more markedly inhibited by galanin or epinephrine than by diazoxide when electrical activity was similarly decreased, and diazoxide had no effect on 86Rb+ efflux in the absence of Ca2+. When the permeability to K+ was increased by 100 microM diazoxide and the hyperpolarization reversed by high extracellular K+, galanin and epinephrine still inhibited insulin release, but did not affect the membrane potential or 86Rb+ efflux. Galanin and epinephrine decreased glucose utilization and oxidation in islet cells by about 10%, whereas diazoxide had no effect. Blockade of alpha 2-adrenoceptors by yohimbine suppressed the effects of epinephrine, but not those of galanin. It is concluded that activation of galanin and alpha2-adrenergic receptors inhibits insulin release by the same mechanisms. These may involve an increase in K+ permeability of the B-cell membrane by opening ATP-sensitive K+ channels and an additional effect independent of the membrane potential.

Comparison of the inhibition of insulin release by activation of adenosine and alpha 2-adrenergic receptors in rat beta-cells.

Rat islets were used to compare the mechanisms whereby adenosine and adrenaline inhibit insulin release. Adenosine (1 microM-2.5 mM) and its analogue N6(-)-phenylisopropyladenosine (L-PIA) (1 nM-10 microM) caused a concentration-dependent but incomplete (45-60%) inhibition of glucose-stimulated release. L-PIA was more potent than D-PIA [the N6(+) analogue], but much less than adrenaline, which caused nearly complete inhibition (85% at 0.1 microM). 8-Phenyltheophylline prevented the inhibitory effect of L-PIA and 50 microM-adenosine, but not that of 500 microM-adenosine or of adrenaline. In contrast, yohimbine selectively prevented the inhibition by adrenaline. Adenosine and L-PIA thus appear to exert their effects by activating membrane A1 receptors, whereas adrenaline acts on alpha 2-adrenergic receptors. Adenosine, L-PIA and adrenaline slightly inhibited 45Ca2+ efflux, 86Rb+ efflux and 45Ca2+ influx in glucose-stimulated islets. The inhibition of insulin release by adenosine or L-PIA was totally prevented by dibutyryl cyclic AMP, but was only attenuated when adenylate cyclase was activated by forskolin or when protein kinase C was stimulated by a phorbol ester. Adrenaline, on the other hand, inhibited release under these conditions. It is concluded that inhibition of adenylate cyclase, rather than direct changes in membrane K+ and Ca2+ permeabilities, underlies the inhibition of insulin release induced by activation of A1-receptors. The more complete inhibition mediated by alpha 2-adrenergic receptors appears to result from a second mechanism not triggered by adenosine.

Distinct mechanisms for two amplification systems of insulin release.

The mechanisms whereby activation of the cyclic AMP-dependent protein kinase A or the Ca2+-phospholipid-dependent protein kinase C amplifies insulin release were studied with mouse islets. Forskolin and the phorbol ester 12-O-tetradecanoylphorbol 13-acetate (TPA) were used to stimulate adenylate cyclase and protein kinase C respectively. The sulphonylurea tolbutamide was used to initiate insulin release in the presence of 3 mM-glucose. Tolbutamide alone inhibited 86Rb+ efflux, depolarized beta-cell membrane, triggered electrical activity, accelerated 45Ca2+ influx and efflux and stimulated insulin release. Forskolin alone only slightly inhibited 86Rb+ efflux, but markedly increased the effects of tolbutamide on electrical activity, 45Ca2+ influx and efflux, and insulin release. In the absence of Ca2+, only the inhibition of 86Rb+ efflux persisted. TPA (100 nM) alone slightly accelerated 45Ca2+ efflux and insulin release without affecting 45Ca2+ influx or beta-cell membrane potential. It increased the effects of tolbutamide on 45Ca2+ efflux and insulin release without changing 86Rb+ efflux, 45Ca2+ influx or electrical activity. Omission of extracellular Ca2+ suppressed all effects due to the combination of TPA and tolbutamide, but not those of TPA alone. Though ineffective alone, 10 nM-TPA amplified the releasing action of tolbutamide without affecting its ionic and electrical effects. In conclusion, the two amplification systems of insulin release involve at least partially distinct mechanisms. The cyclic AMP but not the protein kinase C system initiating signal (Ca2+ influx) triggered by the primary secretagogue.

Forskolin suppresses the slow cyclic variations of glucose-induced electrical activity in pancreatic B cells.

The membrane potential of mouse pancreatic B cells was recorded with microelectrodes. In certain cells, both the slow waves of depolarization and the intervals of repolarization triggered by glucose (10 or 15 mM) displayed regular oscillations in their duration, though the concentration of the sugar remained constant. When forskolin (0.2 microM), an activator of adenylate cyclase, was added to the medium, the electrical activity rapidly became very regular, with slow waves and intervals of constant duration. This effect was unrelated to the overall increase in activity also brought about by forskolin. The oscillations resumed in 75% of the cells after withdrawal of the drug. Under similar conditions, forskolin rapidly and reversibly raised the cAMP concentration in the islets. The data suggest that cAMP is an important modulator of the electrical activity triggered by glucose in insulin-secreting cells.