In pancreatic β cells, the increase in the ATP/ADP ratio that follows a stimulation by glucose is thought to play an important role in the Ca 2+ -dependent increase in insulin secretion. Here we have investigated the possible interactions between Ca 2+ and adenine nucleotides in mouse islets. Measurements of both parameters in the same single islet showed that the rise in the ATP/ADP ratio precedes any rise in the cytoplasmic free-Ca 2+ concentration ([Ca 2+ ] i ) and is already present during the initial transient lowering of [Ca 2+ ] i produced by the sugar. Blockade of Ca 2+ influx with nimodipine did not prevent the concentration-dependent increase in the ATP/ADP ratio produced by glucose and even augmented the ratio at all glucose concentrations which normally stimulate Ca 2+ influx. In contrast, stimulation of Ca 2+ influx by 30 mM K + or 100 µ M tolbutamide lowered the ATP/ADP ratio. This lowering was of rapid onset and reversibility, sustained and prevented by nimodipine or omission of extracellular Ca 2+ . It was, however, not attenuated after blockade of secretion by activation of α 2 -adrenoceptors. The difference in islet ATP/ADP ratio during blockade and stimulation of Ca 2+ influx was similar to that observed between threshold and submaximal glucose concentrations. The results suggest that the following feedback loop could control the oscillations of membrane potential and [Ca 2+ ] i in β cells. Glucose metabolism increases the ATP/ADP ratio in a Ca 2+ -independent manner, which leads to closure of ATP-sensitive K + channels, depolarization and stimulation of Ca 2+ influx. The resulting increase in [Ca 2+ ] i causes a larger consumption than production of ATP, which induces reopening of ATP-sensitive K + channels and arrest of Ca 2+ influx. Upon lowering of [Ca 2+ ] i the ATP/ADP ratio increases again and a new cycle may start.

The ob gene, specifically expressed in fat cells, encodes leptin, a hormone that induces satiety and increases energy expenditure. In this study, we investigated the interactions between glucocorticoids and insulin on ob gene expression in cultured explants of rat adipose tissue. Only low levels of ob mRNA were detected when adipose tissue from fasted rats was cultured for 12–24 h in minimal essential medium. However, the addition of dexamethasone to the medium increased ob gene expression in a concentration-dependent manner (EC 50 10 nM). With 1 μ M dexamethasone, ob mRNA levels were similar to those in fresh fat pads from fed rats, reaching a maximum after 12 h. The effect of dexamethasone was blocked by actinomycin D, which indicates an action on transcription. This effect was increased when a minimum amount of fuel (glucose or a mixture of lactate and pyruvate) was supplied in the medium. Unlike dexamethasone, insulin, even when combined with high glucose concentrations, did not induce ob expression, although it strongly increased the accumulation of mRNA species for fatty acid synthase (FAS), the insulin-sensitive glucose transporter GLUT4 and the γ isoform of peroxisome proliferator-activated receptor (PPARγ). Unexpectedly, insulin dose-dependently inhibited dexamethasone-induced ob mRNA accumulation. This effect was observed at low concentrations of insulin (IC 50 1 nM) and was delayed in onset, beginning after 6–9 h of culture. It was mimicked by insulin-like growth factor 1 (IGF-1) (100 nM). The inhibition by insulin was only detectable when fuels were present and/or when a critical level of ob expression was reached. As this inhibitory effect was reversed by cycloheximide, this suggests that it required ongoing protein synthesis. In conclusion, unlike dexamethasone, insulin had no direct stimulatory effect on ob gene expression. On the other hand, insulin (and IGF-1) even inhibited the dexamethasone-induced accumulation of ob mRNA. The underlying mechanism involved ongoing synthesis of an inhibitory protein by insulin, which is in keeping with its delayed effect. Moreover, the expression of genes for FAS, GLUT4 and PPARγ may be inversely related to that of ob .

Glucose metabolism in pancreatic B-cells leads to an increase in the ATP/ADP ratio that might participate in the regulation of insulin secretion. Good correlations have also been observed between guanine nucleotide levels in isolated pancreatic islets and insulin secretion. To assess whether guanine nucleotides have a specific role in stimulus–secretion coupling, their concentration should be modified selectively. This was attempted by culturing mouse islets overnight in the presence of mycophenolic acid (MPA), an inhibitor of GMP synthesis at the level of IMP dehydrogenase. The drug (25–50 μ g/ml) did not affect the insulin content but decreased the GTP content of the islets and inhibited insulin secretion during subsequent incubation in the presence of 15 mM glucose. However, MPA also decreased the ATP/ADP ratio in the islets. The addition of guanine to the culture medium (to stimulate the salvage pathway of GTP synthesis) restored normal GTP levels, corrected the ATP/ADP ratio and partly prevented the inhibition of insulin release. In contrast, attempts to stimulate ATP synthesis specifically (by provision of adenine or adenosine) failed to reverse any of the effects of MPA. It is concluded that guanine and adenine nucleotide pools are tightly linked and cannot be specifically affected by MPA in pancreatic islet cells, probably because of the activity of nucleoside diphosphate kinase and because of the role of GTP in several reactions leading to adenine nucleotide generation. Contrary to previous claims, MPA is not an adequate tool for evaluating a specific role of guanine nucleotides in the control of insulin secretion.

The ob gene encodes leptin, a hormone which induces satiety and increases energy expenditure. The peroxisome proliferator-activated receptor γ2 isoform (PPARγ2) gene encodes a transcription factor which controls adipocyte differentiation and expression of fat-specific genes. We have studied the regulation of these two genes in white adipose tissue (WAT) during the sucklingŐweaning transition. Suckling rats ingest a high-fat diet (milk). Fat-pad weight barely varied during the last week of suckling. ob mRNA levels, which were very low in 15-day-old rats, rose ∼ 6-fold until weaning at 21 days. When the rats were weaned on to a standard (high-carbohydrate) laboratory chow, epididymal WAT enlarged ∼ 7-fold, and ob mRNA kept increasing progressively and doubled between 21 and 30 days. This evolution contrasted with that of fatty acid synthase (FAS) mRNA, which increased sharply, but only after weaning. To distinguish between the influence of developmental and nutritional factors on ob expression, a group of rats was weaned on to a high-fat diet. This prevented the rise in glycaemia and insulinaemia and the decrease in plasma non-esterified fatty acids which otherwise occurred at weaning. This also resulted in a slight (10Ő15%) decrease in food intake and body weight gain. Under this high-fat diet, the rise of ob mRNA in WAT was augmented (3.7-fold in 30- versus 21-day-old pups), whereas the normal rise in FAS mRNA levels was attenuated. Fat-pad weights and adipocyte cell size and number were roughly similar in high-carbohydrate- and high-fat-weaned pups. mRNA levels of PPARγ2, like those of ob , were low in the WAT of 15-day-old suckling pups, doubled at 21 days, and reached a maximum as soon as 23 days. This evolution further differed from that of ob mRNA in not being influenced by diet composition. In conclusion, ob expression markedly increases during the sucklingŐweaning transition, and this effect is accentuated by a high-fat diet. Qualitative nutritional changes in ob mRNA were correlated with neither acute changes in adipose-tissue mass, nor cell size/number, nor variations in insulinaemia. PPARγ2 also increased during suckling, but rapidly reached a plateau after weaning and no longer changed thereafter. Unlike ob , PPARγ2 was not influenced by the diet composition.

The potential roles of protein tyrosine kinases (TKs) and of phosphotyrosine phosphatases (PTPs) in pancreatic islet function are not known. In this study, we investigated whether vanadate, a potent PTP inhibitor, affects phosphoinositide (PI) metabolism by a TK-dependent pathway in isolated mouse islets. To avoid the confounding effects of changes in Ca 2+ influx, all experiments were performed in the absence of Ca 2+ . In the presence of 15 mM glucose, vanadate, acetylcholine (ACh) or [Arg]vasopressin (AVP) strongly stimulated Ins P production. Vanadate also increased PtdIns P levels in membranes. The TK inhibitor genistein (not its inactive analogues genistin and daidzein) significantly reduced vanadate effects, but was without effect in the absence of stimulation or in the presence of ACh or AVP. Islet proteins resolved by SDS/PAGE were analysed by immunoblotting with anti-phosphotyrosine antibody. Under control conditions, several phosphotyrosyl-proteins (PYPs) were present. Vanadate increased phosphotyrosine residues on several PYPs, notably two proteins of 145 and 85 kDa. This effect was prevented by genistein. p145 and p85 could correspond to phospholipase Cγ (PLCγ) and the regulatory subunit of PtdIns-3-kinase (PtdIns-3K) respectively. Both proteins are expressed in islets, as revealed by immunoblots with specific antibodies. Tungstate, another PTP inhibitor, reproduced vanadate effects, but inhibition of PtdIns-3K by wortmannin failed to affect vanadate-increased PtdIns P levels. Incubation of the islets in the presence of 10% (v/v) fetal calf serum instead of BSA increased Ins P production and this effect was prevented by genistein. These results suggest that inhibition of PTP increases Ins P production in mouse islets by a TK-dependent pathway. They also provide evidence for a potential role of TK and PTP in pancreatic B-cell function.

The mechanisms by which cationic amino acids influence pancreatic B-cell function have been studied by monitoring simultaneously 86 Rb + efflux and insulin release from perifused rat islets. The effects of two reference amino acids arginine and lysine were compared with those of closely related substances to define the structural requirements for recognition of these molecules as secretagogues. Arginine accelerated 86 Rb + efflux and increased insulin release in the absence or in the presence of 7m m -glucose. Its effects on efflux did not require the presence of extracellular Ca 2+ or Na + , but its insulinotropic effects were suppressed in a Ca 2+ -free medium and inhibited in an Na + -free medium. Among arginine derivatives, only 2-amino-3-guanidinopropionic acid mimicked its effects on 86 Rb + efflux and insulin release; citrulline, guanidinoacetic acid, 3-guanidinopropionic acid and guanidine were inactive. Norvaline and valine also increased 86 Rb + efflux, but their effect required the presence of extracellular Na + ; they did not stimulate insulin release. Lysine as well as the shorter-chain cationic amino acids ornithine and 2,4-diaminobutyric acid accelerated 86 Rb + efflux in a Ca 2+ - and Na + -independent manner. Their stimulation of insulin release was suppressed by Ca 2+ omission, but only partially inhibited in an Na + -free medium. The uncharged glutamine and norleucine increased the rate of 86 Rb + efflux in the presence of glucose, only if extracellular Na + was present. Norleucine slightly increased release in a Ca 2+ - and Na + -dependent manner. The effects of lysine on efflux and release were not mimicked by other related substances such as 1,5-diaminopentane and 6-aminohexanoic acid. The results suggest that the depolarizing effect of cationic amino acids is due to accumulation of these positively charged molecules in B-cells. This causes acceleration of the efflux of K + ( 86 Rb + ) and activation of the influx of Ca 2+ (which triggers insulin release). The prerequisite for the stimulation of B-cells by this mechanism appears to be the presence of a positive charge on the side chain of the amino acid, rather than a specific group.

The possible involvement of calmodulin in insulin release was evaluated by studying the effects on intact islets of trifluoperazine and pimozide, two antipsychotic agents known to bind strongly to calmodulin in cell-free systems. Trifluoperazine (10–100μ m ) produced a dose- and time-dependent inhibition of the two phases of glucose-stimulated insulin release. The effect was not reversible by simple washing of the drug, but could be prevented by cytochalasin B or theophylline. Trifluoperazine also inhibited the release induced by glyceraldehyde, oxoisocaproate, tolbutamide or barium, but not that stimulated by 10m m -theophylline or 1m m -3-isobutyl-1-methylxanthine. Pimozide (0.5–10μ m ) also produced a dose-dependent inhibition of insulin release triggered by glucose, leucine or barium, but did not affect the release induced by methylxanthines. Glucose utilization by islet cells was not modified by trifluoperazine (25μ m ), which slightly increased cyclic AMP concentration in islets incubated without glucose. The drug did not prevent the increase in cyclic AMP concentration observed after 10min of glucose stimulation, but suppressed it after 60min. Basal or glucose-stimulated Ca 2+ influx (5min) was unaffected by 25μ m -trifluoperazine, whereas Ca 2+ net uptake (60min) was inhibited by 20%. Glucose-stimulated Ca 2+ uptake was almost unaffected by pimozide. In a Ca 2+ -free medium, trifluoperazine decreased Ca 2+ efflux from the islets and did not prevent the further decrease by glucose; in the presence of Ca 2+ , the drug again decreased Ca 2+ efflux and inhibited the stimulation normally produced by glucose. In the absence of glucose, trifluoperazine lowered the rate of Rb + efflux from the islets, decreased Rb + influx (10min), but did not affect Rb + net uptake (60min). It did not interfere with the ability of glucose to decrease Rb + efflux rate further and to increase Rb + net uptake. The results show thus that trifluoperazine does not alter the initial key events of the stimulus–secretion coupling. Its inhibition of insulin release suggests a role of calmodulin at late stages of the secretory process.

The K + permeability of pancreatic islet cells was studied by monitoring the efflux of 86 Rb + (used as tracer for K + ) from perifused rat islets and measuring the uptake of 42 K + . Glucose markedly and reversibly decreased 86 Rb + efflux from islet cells and this effect was antagonized by inhibitors of the metabolic degradation of the sugar, i.e. mannoheptulose, iodoacetate, glucosamine and 2-deoxyglucose. Among glucose metabolites, glyceraldehyde reduced the K + permeability even more potently than did glucose itself; pyruvate and lactate alone exhibited only a small effect, but potentiated that of glucose. Other metabolized sugars, like mannose, glucosamine and N -acetylglucosamine, also decreased 86 Rb + efflux from islet cells. Fructose was effective only in the presence of glucose. Non-metabolized sugars like galactose, 2-deoxyglucose and 3- O -methylglucose had no effect. The changes in K + permeability by agents known to modify the concentrations of nicotinamide nucleotides, glutathione or ATP in islet cells were also studied. Increasing NAD(P)H concentrations in islet cells by pentobarbital rapidly and reversibly reduced 86 Rb + efflux; exogenous reduced glutathione produced a similar though weaker effect. By contrast, oxidizing nicotinamide nucleotides with phenazine methosulphate or Methylene Blue, or oxidizing glutathione by t-butyl hydroperoxide increased the K + permeability of islet cells. Uncoupling the oxidative phosphorylations with dicumarol also augmented 86 Rb + efflux markedly. In the absence of glucose, but not in its presence, methylxanthines reduced 86 Rb + efflux from the islets; such was not the case for cholera toxin or dibutyryl cyclic AMP. Glucose and glyceraldehyde had no effect on 42 K + uptake after a short incubation (10min), but augmented it after 60min; the effect of glucose was suppressed by mannoheptulose and not mimicked by 3- O -methylglucose. The results clearly establish the importance of the metabolic degradation of glucose and other substrates for the control of the K + permeability in pancreatic islet cells and support the concept that a decrease in the K + permeability represents a major step of the B-cell response to physiological stimulation.