Type 1 diabetic serum interferes with pancreatic beta-cell Ca2+-handling.

The aim of this study was to clarify the frequency of patients with type 1 diabetes that have serum that increases pancreatic beta-cell cytoplasmic free Ca(2+) concentration, [Ca(2+)](i), and if such an effect is also present in serum from first-degree relatives. We also studied a possible link between the serum effect and ethnic background as well as presence of autoantibodies. Sera obtained from three different countries were investigated as follows: 82 Swedish Caucasians with newly diagnosed type 1 diabetes, 56 Americans with different duration of type 1 diabetes, 117 American first-degree relatives of type 1 diabetic patients with a mixed ethnic background and 31 Caucasian Finnish children with newly diagnosed type 1 diabetes. Changes in [Ca(2+)](i) , upon depolarization, were measured in beta-cells incubated overnight with sera from type 1 diabetic patients, first-degree relatives or healthy controls. Our data show that there is a group constituting approximately 30% of type 1 diabetic patients of different gender, age, ethnic background and duration of the disease, as well as first-degree relatives of type 1 diabetic patients, that have sera that interfere with pancreatic beta-cell Ca(2+)-handling. This effect on beta-cell [Ca(2+)](i) could not be correlated to the presence of autoantibodies. In a defined subgroup of patients with type 1 diabetes and first-degree relatives a defect Ca(2+)-handling may aggravate development of beta-cell destruction.

Proinsulin C-peptide and its analogues induce intracellular Ca2+ increases in human renal tubular cells.

Based on the findings that proinsulin C-peptide binds specifically to cell membranes, we investigated the effects of C-peptide and related molecules on the intracellular Ca2+ concentration ([Ca2+]i) in human renal tubular cells using the indicator fura-2/AM. The results show that human C-peptide and its C-terminal pentapeptide (positions 27-31, EGSLQ), but not the des (27-31) C-peptide or randomly scrambled C-peptide, elicit a transient increase in [Ca2+]i. Rat C-peptide and rat C-terminal pentapeptide also induce a [Ca2+]i response in human tubular cells, while a human pentapeptide analogue with Ala at position 1 gives no [Ca2+]i response, and those with Ala at positions 2-5 induce responses with different amplitudes. These results define a species cross-reactivity for C-peptide and demonstrate the importance of Glu at position 1 of the pentapeptide. Preincubation of cells with pertussis toxin abolishes the effect on [Ca2+]i by both C-peptide and the pentapeptide. These results are compatible with previous data on C-peptide binding to cells and activation of Na-,K+ATPase. Combined, all data show that C-peptide is a bioactive peptide and suggest that it elicits changes in [Ca2+]i via G-protein-coupled pathways, giving downstream enzyme effects.

Cyclin-dependent kinase 5 promotes insulin exocytosis.

Cyclin-dependent kinase 5 (Cdk5) is widely expressed although kinase activity has been described preferentially in neuronal systems. Cdk5 has an impact on actin polymerization during neuronal migration and neurite outgrowth and deregulation of the kinase has been implicated in the promotion of neurodegeneration. Recently it was shown that Cdk5 modulates dopamine signaling in neurons by regulating DARPP-32 function. In addition, Cdk5 phosphorylates munc-18 and synapsin I, two essential components of the exocytotic machinery. We have shown by reverse transcriptase-polymerase chain reaction, immunocytochemistry, and Western blotting that Cdk5 is present in the insulin-secreting pancreatic beta-cell. Subcellular fractionation of isolated beta-cells revealed a glucose-induced translocation of membrane-bound Cdk5 protein to lower density fractions. Inhibition of Cdk5 with roscovitine reduced insulin secretion with approximately 35% compared with control after glucose stimulation and with approximately 65% after depolarization with glucose and KCl. Capacitance measurements performed on single beta-cells that expressed a dominant-negative Cdk5 mutant showed impaired exocytosis. The effect on exocytosis by Cdk5 appeared to be independent of changes in free cytoplasmic Ca(2+) concentration. Taken together these results show that Cdk5 is present in beta-cells and acts as a positive regulator of insulin exocytosis.

Role of apoptosis in pancreatic beta-cell death in diabetes.

Apoptosis is a physiological form of cell death that occurs during normal development, and critical mediators of this process include caspases, reactive oxygen species, and Ca2+. Excessive apoptosis of the pancreatic beta-cell has been associated with diabetes. Consequently, apoptosis research has focused on how infiltrating macrophages or cytotoxic T-cells might kill pancreatic beta-cells using cytokines or death receptor triggering. Meanwhile, the intracellular events in the target beta-cell have been largely ignored. Elucidation of such targets might help develop improved treatment strategies for diabetes. This article will outline recent developments in apoptosis research, with emphasis on mechanisms that may be relevant to beta-cell death in type 1 and type 2 diabetes. Several of the models proposed in beta-cell killing converge on Ca2+ signaling, indicating that the pancreatic beta-cell may be an ideal system in which to carefully dissect the role of Ca2+ during apoptosis.

Effects of serum from patients with type 1 diabetes on primary cerebellar granule cells.

Type 1 diabetes is an autoimmune disease of unknown etiology. Our previous work has shown that a factor present in serum from type 1 diabetic patients causes increased Ca2+ channel activity and apoptotic DNA fragmentation in pancreatic beta-cells. Here we examined the effects of type 1 diabetic serum on primary cerebellar granule cells (CGCs). In CGCs, exposure to type 1 diabetic serum did not cause increased apoptosis or changes in Ca2+ channel activity. However, patient serum did cause modulation of Ca2+ signals in a cell type with triangular soma that exhibited low voltage-gated Ca2+ currents. This cell was present primarily in cultures exposed to type 1 diabetic serum. The presence of low voltage-gated Ca2+ currents and long neuronal dendrites indicated that this unique cell was of neuronal origin and not of glial origin.

7beta-hydroxycholesterol induces Ca(2+) oscillations, MAP kinase activation and apoptosis in human aortic smooth muscle cells.

In the present study, we characterize the early cytotoxic effects of 7beta-hydroxycholesterol, a major cytotoxin in oxidized LDL, in human aortic smooth muscle cells. Within a few minutes after addition, 7beta-hydroxycholesterol induced Ca(2+) oscillations with a frequency of approximately 0.3-0.4 min(-1). A few hours later, thapsigargin-sensitive Ca(2+) pools were depleted, indicating that 7beta-hydroxycholesterol perturbs intracellular Ca(2+) homeostasis. The mitogen-activated protein kinases (MAPKs) ERK1 and ERK2 (but not JNK) were activated within 5 min after addition of 7beta-hydroxycholesterol. The side-chain hydroxylated oxysterols 25-hydroxycholesterol and 27-hydroxycholesterol were more potent in inducing apoptosis than 7beta-hydroxycholesterol and cholesterol-5alpha,6alpha-epoxide, as determined by TUNEL staining. Addition of TNFalpha (10 ng/ml) and IFNgamma (20 ng/ml) enhanced the cytotoxicity of oxysterols and potentiated apoptosis. The cytokines alone were not toxic to smooth muscle cells at these concentrations. 25-Hydroxycholesterol and 7beta-hydroxycholesterol but not cholesterol inhibited protein synthesis at 4-8 h as determined by [35S]methionine incorporation assay. Morphologically, oxysterol-induced cell death was characterized by disorganization of the ER and Golgi membranes. The Ca(2+) and ERK signals preceded the ultrastructural changes induced by 7beta-hydroxycholesterol.

Influence of acarbose on post-prandial insulin requirements in patients with Type 1 diabetes.

The primary objective of this double-blind, placebo-controlled, randomised cross-over study was to investigate the influence of acarbose on insulin requirement in patients with Type 1 diabetes (T1DM) following a standardised meal. In addition, the study assessed the effects of acarbose on post-prandial triglyceride, glucagon and gastrointestinal peptide levels, gastric emptying, and oxidative glucose metabolism. Following normalisation of their blood glucose, 10 patients received a standardised meal together with acarbose (100 mg) or placebo. Each patient was evaluated twice (separated by 10+/-3 days), and the cross-over study design ensured that they received both acarbose and placebo. The insulin requirement for maintenance of normoglycaemia was assessed using a closed-loop insulin infusion system (artificial pancreas, Biostator). Acarbose produced a statistically significant reduction in mean insulin requirement over a 3-hr period following the meal compared with placebo (5171.7+/-2282.6 mU vs 8074.5+/-3045.4 mU; p=0.003). The level of blood glucose control over the same period was similar in the two groups. Gastric inhibitory polypeptide levels also showed a statistically significant decrease with acarbose treatment compared with placebo for AUC (area under the curve; p=0.006) and Cmax (maximum plasma concentration; p=0.022), but not tmax (time to reach Cmax from the start of the standardised meal; p>0.05). Analysis of the other efficacy parameters revealed no statistically significant differences between acarbose treatment and placebo (p>0.05). These results indicate that acarbose decreases insulin requirement in patients with T1DM without affecting gastric emptying.

An endogenous peptide isolated from the gut, NK-lysin, stimulates insulin secretion without changes in cytosolic free Ca2+ concentration.

We have recently isolated and cloned a novel endogenous peptide from pig intestine, NK-lysin (NKL). In the present study we show that NKL (1-100 nM) potently and reversibly stimulates insulin secretion in rat pancreatic islets and in the beta-cell line HIT T15. This effect of NKL was not accompanied by changes in cytoplasmic free calcium concentration. The stimulatory activity of NKL on insulin release was also observed in permeabilized islets under Ca2+-clamped conditions. Preincubation of HIT T15 cells with NKL for 1 h or 24 h did not influence cell viability. Possible mechanisms of insulinotropic activity of NKL are discussed.

X-ray microanalysis of in situ and isolated pancreatic islets.

The effect of stimulation of insulin secretion in pancreatic beta cells on the elemental composition of these cells was investigated by x-ray microanalysis. In vitro experiments on isolated islets of Langerhans from ob/ob mice were compared to in situ experiments. The only significant difference in the elemental composition of beta cells from ob/ob mice versus their lean counterparts is a lower Ca concentration in the ob/ob animals. The nucleus of the beta cells has a higher concentration of P, K, and Na than the cytoplasm, which has a higher concentration of S and Cl. No polarized ion distribution in the cytoplasm of the beta cells was observed. Isolated beta cells show a higher concentration of Na and Cl and a lower concentration of K than their in situ counterparts. Stimulation of insulin secretion with glucose both in situ and in vitro showed only very small effects on the elemental composition of the beta cells: a tendency to a decreased P content was noted. In vitro experiments using stimulation with high extracellular K+ showed, in addition, a small increase in the intracellular K concentration. In conclusion, while the elemental content of beta cells in vitro differs from that in situ, the response to glucose stimulation appears to be similar in both systems.

Ca2+ channel blockers verapamil and nifedipine inhibit apoptosis induced by 25-hydroxycholesterol in human aortic smooth muscle cells.

We have characterized the death of human aortic smooth muscle cells induced by 25-hydroxycholesterol, an oxidation product of cholesterol. Chromatin condensation characteristic of apoptosis was observed by enzymatic (TUNEL) staining of chromatin, and by electron microscopy. Fourteen percent of cells treated with 5 microg/ml of 25-hydroxycholesterol for 24 h displayed chromatin degradation as determined by positive TUNEL staining. Addition of TNF alpha (10 ng/ml) and IFN gamma (20 ng/ml) increased the proportion of TUNEL positive cells to 30%, whereas the cytokines alone were without effect. After 48 h, 40% of the cells treated with 5 microg/ml of 25-hydroxycholesterol were TUNEL positive, and 21% of the cells displayed chromatin condensation. Oligonucleosomal DNA fragmentation typical of apoptosis was demonstrated by agarose gel electrophoresis. Furthermore, activation of the ICE-like protease caspase 3 (CPP32) was observed in cells treated with 25-hydroxycholesterol. Addition of the Ca2+ entry blockers verapamil or nifedipine to the culture medium inhibited apoptosis by more than 70% and reduced cytotoxicity, while removal of Ca2+ from culture medium reduced apoptosis by 42%. Within a few minutes after addition, 25-hydroxycholesterol induced intracellular Ca2+ oscillations with a frequency of approximately 0.3-0.4 min(-1). Thus it appears that Ca2+ influx through plasma membrane channels is an important signal in oxysterol-induced apoptosis. Addition of TNF alpha and IFN gamma enhanced cytotoxicity and resulted in a higher proportion of apoptotic cells, suggesting that inflammatory cytokines can increase the cytotoxicity of lipid oxidation products.

The antidiabetogenic effect of GLP-1 is maintained during a 7-day treatment period and improves diabetic dyslipoproteinemia in NIDDM patients.

OBJECTIVE: To investigate the long-term antidiabetogenic effect of glucagon-like peptide 1 (GLP-1) and its influence on diabetic dyslipoproteinemia, patients with NIDDM were treated with GLP-1 subcutaneously for 1 week. RESEARCH DESIGN AND METHODS: Twelve patients participated in the study. The 1st week of the study, all of them were on intensive insulin treatment and from day 8, four were randomized to a control group continuing with insulin, and eight to a treatment group where GLP-1 was given at meals together with regular insulin from day 8 to 12. On days 13 and 14, they were only given GLP-1 at meals. NPH insulin at bedtime was given throughout the study. RESULTS: In the GLP-1-treated patients, the doses of regular insulin, given to keep a satisfactory blood glucose control, were reduced compared with treatment with insulin only. GLP-1 virtually inhibited the early increase in blood glucose after the meals, whereas an increase of approximately 2 mmol was seen during an optimized insulin treatment. In agreement with the short half-life of the peptide, 2-h postprandial plasma insulin levels were significantly decreased both at day 12 and 14, suggesting that there was not enough GLP-1 left to stimulate endogenous insulin release and compensate for the decrease in the dose of exogenous insulin. Therefore, the effect of GLP-1 was lost before the next meal, resulting in increased preprandial blood glucose values at lunch and dinner. The concentration of VLDL triglycerides decreased already during the 1st week. This decrease persisted during the 2nd week when GLP-1 was included in the treatment. No changes were observed in the levels of LDL and HDL cholesterol. The LDL particle diameter increased from a mean of 22.3 to 22.6 nm (P < 0.01) in response to insulin treatment. A further increment to 22.9 nm (P < 0.05) was seen after GLP-1 treatment. The LDL particle size did not change in the control group. Lipoprotein lipase activity was decreased by 27% and hepatic lipase was reduced by 13% in the GLP-1-treated group. CONCLUSIONS: We confirm the antidiabetogenic effect of GLP-1 in NIDDM patients. This effect was maintained during 7 days, which implies that the patients did not develop tolerance during this treatment period. Intensive insulin treatment, leading to normotriglyceridemia, increased the mean LDL particle diameter, which is likely to lower the risk of future coronary heart disease in patients with NIDDM. Furthermore, an additive effect of GLP-1 is indicated. Hence, this study gives additional evidence that GLP-1 may be useful as an agent for treating NIDDM.

Glucagon-like peptide I enhances the insulinotropic effect of glibenclamide in NIDDM patients and in the perfused rat pancreas.

OBJECTIVE: To investigate the acute effects of glibenclamide and glucagon-like peptide I (GLP-I) and their combination in perfused isolated rat pancreas and in patients with secondary failure to sulfonylureas. RESEARCH DESIGN AND METHODS: Rat islets were perfused with 10 nmol/l GLP-I in combination with 2 mumol/l glibenclamide. In human experiments, GLP-I (0.75 pmol. kg-1.min-1) was given as a continuous infusion during 240 min, while glibenclamide (3.5 mg) was administered orally. Eight patients participated in the study (age 57.6 +/- 2.7 years, BMI 28.7 +/- 1.5 kg/m2, mean +/- SE). In all subjects, blood glucose was first normalized by insulin infusion administered by an artificial pancreas (Biostator). RESULTS: GLP-I increased the insulinotropic effect of glibenclamide fourfold in the perfused rat pancreas. In human experiments, treatment with GLP-I alone and in combination with glibenclamide significantly decreased basal glucose levels (5.1 +/- 0.4 and 4.5 +/- 0.1 vs. 6.0 +/- 0.3 mmol/l, P < 0.01), while with only glibenclamide, glucose concentrations remained unchanged. GLP-I markedly decreased total integrated glucose response to the meal (353 +/- 60 vs. 724 +/- 91 mmol.l-1. min-1, area under the curve [AUC] [-30-180 min], P < 0.02), whereas glibenclamide had no effect (598 +/- 101 mmol.l-1. min-1, AUC [-30-180 min], NS). The combined treatment further enhanced the glucose lowering effect of GLP-I (138 +/- 24 mmol. l-1.min, AUC [-30-180 min], P < 0.001). GLP-I, glibenclamide, and combined treat-stimulated meal-induced insulin release as reflected by insulinogenic indexes (control 1.44 +/- 0.4; GLP-I 6.3 +/- 1.6, P < 0.01; glibenclamide 6.8 +/- 2.1, P < 0.01; combination 20.7 +/- 5.0, P < 0.001). GLP-I inhibited basal but not postprandial glucagon responses. Using paracetamol as a marker for gastric emptying rate of the test meal, treatment with GLP-I decreased gastric emptying at 180 min by approximately 50% compared with the control subjects (P < 0.01). CONCLUSIONS: In acute experiments on overweight patients with NIDDM, GLP-I exerted a marked antidiabetogenic action on the basal and postprandial state. The peptide stimulated insulin, suppressed basal glucagon release, and prolonged gastric emptying. The glucose-lowering effect of GLP-I was further enhanced by glibenclamide. This action may be at least partially accounted for by a synergistic effect of these two compounds on insulin release. Glibenclamide per se enhanced postprandial but not basal insulin release and exerted a less pronounced antidiabetogenic effect compared with GLP-I.

Reversible Ca2+-dependent translocation of protein kinase C and glucose-induced insulin release.

It has been reported that protein kinase C (PKC) interacts at multiple sites in beta-cell stimulus-secretion coupling. Nevertheless, there is still controversy concerning the importance of this enzyme in glucose-induced insulin release. The present study was undertaken to clarify whether glucose, directly, or through changes in cytoplasmic free Ca2+ concentration, [Ca2+]i, could promote translocation of PKC from the soluble to the membrane compartment. Whereas glucose, which increases [Ca2+]i, did not affect long-term distribution of PKC activity between soluble and membrane fractions, this distribution was reversibly affected acutely by the Ca2+ concentration in the extraction media. Translocation of PKC to the membrane by incubation of HIT cells for 10 min in the presence of 20 nM phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) resulted in a 5-fold increase in glucose-induced insulin release. This was prevented by 50 nM concentration of the PKC inhibitor staurosporine, provided that the cells were exposed to the inhibitor before the phorbol ester. Cells pretreated with TPA demonstrated increased insulin secretion in response to glucose for several hours. This time course extended beyond the disappearance of [3H]TPA from the cells, which was complete after 1 h. Activation of PKC increased both average insulin release and the amplitude of oscillations 2-fold, but did not affect oscillation frequency. The stimulatory effect of increased PKC activity on insulin release was not matched by changes in [Ca2+]i. We suggest that stimulation of the pancreatic beta-cell with glucose promotes transient translocation of certain PKC isoforms from the cytoplasm to the plasma membrane as a direct consequence of the increase in [Ca2+]i. Such a translocation may promote phosphorylation of one or several proteins involved in the regulation of the beta-cell stimulus-secretion coupling. This results in potentiation of glucose-induced activation of insulin exocytosis, an effect then not mediated by an increase in [Ca2+]i per se. Hence, pulsatile insulin release can be obtained under conditions where overall [Ca2+]i does not change, challenging the view that oscillations in [Ca2+ ]i are indeed driving the oscillations in hormone release.

Interleukin-1 beta-induced stimulation of insulin release in mouse pancreatic islets is related to diacylglycerol production and protein kinase C activation.

The aim of the present study was to investigate the mechanisms responsible for the acute, stimulatory effects of interleukin-1 beta (rIL-1 beta; 1 ng/ml) on insulin release from mouse pancreatic islets. For this purpose, mouse islets were exposed for 60-120 min to rIL-1 beta and their function and metabolism characterized during this period. The cytokine did not increase insulin release in the presence of 1.7 mM glucose, but both in the presence of 5.6 or 16.7 mM glucose, or 10 mM leucine + 2 mM glutamine, it induced a 60-100% increase in insulin release. Moreover, rIL-1 beta also enhanced the effects of 1 mu/ml glipizide on insulin release, but failed to increase insulin release induced by 30 mM KCl or by glucose plus phorbol ester (TPA; 100 nM). These early stimulatory effects of rIL-1 beta on insulin release were neither accompanied by major increases in glucose or amino acid metabolism, nor by modifications in islet cAMP content, and they were prevented by mannoheptulose, diazoxide or verapamil. rIL-1 beta potentiation of glucose-induced insulin release was not accompanied by modifications in [Ca2+]i, but the cytokine increased diacylglycerol production and induced protein kinase C (PKC) activation. Down-regulation of PKC completely prevented the stimulatory effects of rIL-1 beta on glucose-induced insulin release. In conclusion, rIL-1 beta induces an early stimulation of insulin release in mouse beta-cells by a mechanism independent of glucose metabolism, cAMP generation or modifications in [Ca2+]i. This effect is probably related to diacylglycerol formation and stimulation of PKC.

Effects of K(+)-induced depolarization and purinergic receptor activation on elemental content in insulin-producing RINm5F-cells.

X-ray microanalysis was used to detect elemental changes in the insulin-producing tumor cell-line RINm5F. To improve discrimination between mobile ions and ions bound to macromolecules a new approach was employed, consisting of multivariate statistical analysis of correlations between the concentrations of Na, Mg, P, S, Cl, K, and Ca. RINm5F cells, cultured on Formvar-coated titanium grids, were stimulated with high K+ or ATP, that are both known to stimulate insulin release. The buffers used contained Ca2+ or one of the Ca(2+)-analogues Sr2+ and Ba2+, to represent Ca2+ uptake in response to stimulation. After stimulation the cells were shock-frozen and freeze-dried overnight. Incubation for 10-20 seconds in a Ca(2+)-containing buffer did not significantly affect elemental composition, whereas cellular Mg, P and K decreased in a Sr(2+)-containing buffer. Depolarization with high K+ concentration caused an increase in the cellular Na content, both in Ca(2+)- and Sr(2+)-containing buffers, but not in the buffer where Ca2+ had been replaced by Ba2+. X-ray microanalysis is useful for detection of elemental changes subsequent to stimulation of cultured cells. Moreover, multivariate statistical analysis strengthens the idea that stimulation of RINm5F cells causes redistribution of ions possibly due to changes in the state of binding of some elements to cellular proteins.

Identification of synaptic proteins and their isoform mRNAs in compartments of pancreatic endocrine cells.

Several proteins that are of importance for membrane trafficking in the nerve terminal have recently been characterized. We have used Western blot and immunohistochemistry to show that synaptotagmin, synaptobrevin/VAMP (vesicle-associated membrane protein), SNAP-25 (synaptosomal-associated protein of 25 kDa), and syntaxin proteins are present in cells of the islets of Langerhans in the endocrine pancreas. Synaptotagmin-like immunoreactivity (-LI) was localized to granules within the cytoplasm of a few endocrine cells located in the periphery of the islets, identified as somatostatin-containing cells, and in many nerve fibers within the islets. VAMP-LI was seen in granules of virtually all pancreatic islet cells and also in nerve fibers. SNAP-25-LI and syntaxin-LI were predominantly present in the plasma membrane of the endocrine cells, including insulin-producing beta cells. In situ hybridization, using isoform-specific oligonucleotide probes, detected VAMP-2, cellubrevin, SNAP-25, syntaxin 1A, 4, and 5, and munc-18 mRNAs in isolated pancreatic islets and in insulin-producing cells. The results show the presence of several synaptic proteins at protein and mRNA levels in pancreatic islet cells, suggesting that they may have specific roles in the molecular regulation of exocytosis also in insulin-secreting cells.

Glucose-stimulated increase in cytoplasmic pH precedes increase in free Ca2+ in pancreatic beta-cells. A possible role for pyruvate.

The temporal relationship of glucose-induced increases in cytoplasmic pH (pHi) and cytoplasmic free Ca2+ was studied in single mouse pancreatic beta-cells and suspensions of clonal beta-cells (HIT). In both preparations of cells the increase in pHi preceded the cytoplasmic free Ca2+ increase. Therefore the alkalinization cannot be a consequence of the Ca2+ influx. A potential metabolic mechanism for the increase in pHi, involving stimulation of pyruvate transport and oxidation, was demonstrated in a model system of liver mitochondria incubated with pyruvate, ATP, and hexokinase to which glucose was then added to initiate ATP use. The involvement of this mechanism in beta-cells is suggested by the observation that the alkalinization was prevented in most cells by incubation with 3-hydroxycyanocinnamate, a mitochondrial pyruvate transport inhibitor. On the other hand, the inhibited cells exhibited normal Ca2+ responses to glucose stimulation. This indicates that neither pyruvate metabolism nor the alkalinization is of critical importance for the Ca2+ signal, though pyruvate oxidation or its metabolites may be important in downstream regulation of secretion.