Direct identification of electrophysiologically monitored cells within intact mouse islets of Langerhans.

Cells found to be electrically active within microdissected mouse islets of Langerhans perifused with high (greater than or equal to 11.1 mM) glucose concentrations were labeled by injecting Lucifer yellow through the recording electrode. After fixation, these cells were located by fluorescence microscopy on sections serially cut throughout the islets and were subsequently identified by immunofluororescence staining with specific anti-islet hormone sera. Electrophysiologic control confirmed that the electrode tip had remained within the same cell throughout the experiment and showed that Lucifer yellow labeling did not affect the electrical activity of the impaled cell. Upon individual impalements, Lucifer yellow labeled either the impaled cell alone or this cell and some of its neighbors to which it was dye coupled. Immunofluorescence staining of the Lucifer yellow-labeled cells revealed that glucose-induced electrical activity was recorded from individual B-cells or groups of dye-coupled B-cells as well as from A-cells coupled to B-cells.

Analysis of the effects of calcium or magnesium on voltage-clamp currents in perfused squid axons bathed in solutions of high potassium.

Isolated axons from the squid, Dosidicus gigas, were internally perfused with potassium fluoride solutions. Membrane currents were measured following step changes of membrane potential in a voltage-clamp arrangement with external isosmotic solution changes in the order: potassium-free artificial seawater; potassium chloride; potassium chloride containing 10, 25, 40 or 50, mM calcium or magnesium; and potassium-free artificial seawater. The following results suggest that the currents measured under voltage clamp with potassium outside and inside can be separated into two components and that one of them, the predominant one, is carried through the potassium system. (a) Outward currents in isosmotic potassium were strongly and reversibly reduced by tetraethylammonium chloride. (b) Without calcium or magnesium a progressive increase in the nontime-dependent component of the currents (leakage) occurred. (c) The restoration of calcium or magnesium within 15-30 min decreases this leakage. (d) With 50 mM divalent ions the steady-state current-voltage curve was nonlinear with negative resistance as observed in intact axons in isosmotic potassium. (e) The time-dependent components of the membrane currents were not clearly affected by calcium or magnesium. These results show a strong dependence of the leakage currents on external calcium or magnesium concentration but provide no support for the involvement of calcium or magnesium in the kinetics of the potassium system.

Properties of the Ca-activated K+ channel in pancreatic beta-cells.

The existence of [Ca2+]i-activated K+-channels in the pancreatic beta-cell membrane is based in two observations: quinine inhibits K+-permeability and, increasing intracellular Ca2+ stimulates it. The changes in K+-permeability of the beta-cell have been monitored electrically by combining measurements of the dependence of the membrane potential on external K+ concentration and input resistance. The changes in the passive 42K and 86Rb efflux from the whole islet have been measured directly. Intracellular Ca2+ has been increased by various means, including increasing extracellular Ca2+, addition of the Ca2+-ionophore A23187 or noradrenaline and application of mitochondrial uncouplers and blockers. In addition to quinine, many other substances have been found to inhibit or modulate the [Ca2+]i-activated K+-channel. The most important of these is the natural stimulus for insulin secretion, glucose. Glucose may inhibit K+-permeability by lowering intracellular Ca2+. Glibenclamide, a hypoglycaemic sulphonylurea, is about 25 times more active than quinine in blocking the K+-channel in beta-cells. The methylxanthines, c-AMP, various calmodulin inhibitors and Ba2+ also inhibit K+-permeability. Genetically diabetic mice have been studied and show an alteration in the [Ca2+]i-activated K+-channel. It is concluded that the [Ca2+]i-activated K+-channel plays a major role in the normal function of the pancreatic beta-cell. The study of its properties should prove valuable for the understanding and treatment of diabetes.

Tetracaine stimulates insulin secretion from the pancreatic beta-cell by release of intracellular calcium.

The role of intracellular calcium stores in stimulus-secretion coupling in the pancreatic beta-cell is largely unknown. We report here that tetracaine stimulates insulin secretion from collagenase-isolated mouse islets of Langerhans in the absence of glucose or extracellular calcium. We also found that the anesthetic evokes a dose-dependent rise of the intracellular free-calcium concentration ([Ca2+]i) in cultured rat and mouse beta-cells. The tetracaine-specific [Ca2+]i rise also occurs in the absence of glucose, or in beta-cells depolarized by exposure to a Ca(2+)-deficient medium (< 1 microM) or elevated [K+]o. Furthermore, tetracaine (> or = 300 microM) depolarized the beta-cell membrane in mouse pancreatic islets, but inhibited Ca2+ entry through voltage-gated Ca2+ channels in HIT cells, an insulin-secreting cell line. From these data we conclude that tetracaine-enhancement of insulin release occurs by mechanisms that are independent of Ca2+ entry across the cell membrane. The tetracaine-induced [Ca2+]i rise in cultured rat beta-cells and insulin secretion from mouse islets is insensitive to dantrolene (20 microM), a drug that inhibits Ca2+ release evoked by cholinergic agonists in the pancreatic beta-cell, and thapsigargin (3 microM), a blocker of the endoplasmic reticulum (ER) Ca2+ pump. We conclude that the Ca2+ required for tetracaine-potentiated insulin secretion is released from intracellular Ca2+ stores other than the ER. Furthermore, tetracaine-induced Ca2+ release was unaffected by the mitochondrial electron transfer inhibitors NaN3 and rotenone. Taken together, these data show that a calcium source other than the ER and mitochondria can affect beta-cell insulin secretion.

Control of cytosolic free calcium in cultured human pancreatic beta-cells occurs by external calcium-dependent and independent mechanisms.

Changes in cytosolic intracellular free Ca2+ ([Ca2+]i) in response to glucose, glyburide, cholinergic agonists, and elevated [K+]o (external potassium concentration) were measured in cultured human islet beta-cells. In the absence of glucose, the mean resting [Ca2+]i in single beta-cells was 84.5 +/- 4.7 nM (n = 86) and remained unchanged in low external [Ca2+]o (Ca2+ concentration) (< 0.2 microM) at 23-25 C. Glucose (5.6-33 mM) induced a slow dose-related [Ca2+]i rise up to 300.0 +/- 50.6 nM (n = 19). This [Ca2+]i rise always occurred with a delay that varied from cell to cell (approximately 10-120 sec), and the steady state [Ca2+]i exhibited a sigmoidal dependence on glucose concentration (midpoint at 14.9 mM). The glucose-induced rise in [Ca2+]i was attenuated by about 62% in low external [Ca2+]o and was not affected by dantrolene, a drug that inhibits Ca2+ release from the endoplasmic reticulum. In the absence or presence of glucose, cholinergic receptor agonists evoked a biphasic increase in [Ca2+]i up to 350 nM; the delayed component of the [Ca2+]i rise was blocked by dantrolene. A rapid elevation of [K+]o to 40 mM also elicited a biphasic rise in [Ca2+]i, which peaked at about 250 nM and was inhibited by the Ca2+ channel antagonist nifedipine. Glyburide (4 microM) in the absence of glucose also induced a [Ca2+]o-dependent rise in [Ca2+]i. Increasing the concentration of glucose from 4 to 16.7 mM evoked a biphasic pattern of insulin secretion from perifused isolated islets at 37 C. Finally, in the presence of 4 mM glucose, a cholinergic muscarinic receptor agonist stimulated insulin secretion. A glucose-stimulated [Ca2+]i rise was also studied at 24 and 37 C in cultured rat islet cells. Our results suggest that the Ca2+ required for glucose-induced and muscarinic agonist-potentiated insulin release enters the cytosol from both extracellular and intracellular Ca2+ stores.

Prolactin induces maturation of glucose sensing mechanisms in cultured neonatal rat islets.

The effects of PRL treatment on insulin content and secretion, and 86Rb and 45Ca fluxes from neonatal rat islets maintained in culture for 7-9 days were studied. PRL treatment enhanced islet insulin content by 40% and enhanced early insulin secretion evoked by 16.7 mM glucose. Insulin release stimulated by oxotremorine-M, a muscarinic agonist, in the presence of glucose (8.3 or 16.7 mM) was unchanged by PRL treatment. However, PRL treatment potentiated phorbol 12,13-dibutyrate-stimulated insulin secretion in the presence of the above glucose concentrations. PRL treatment potentiated the reduction in 86Rb efflux induced by glucose or tolbutamide and enhanced the increase in 86Rb efflux evoked by diazoxide. PRL treatment slightly potentiated the increment in 45Ca uptake induced by high concentrations of K+, but failed to affect the increment evoked by 16.7 mM glucose. Since glucose-induced 45Ca uptake was not affected by PRL, we suggest that the enhancement in first phase insulin secretion evoked by glucose in the PRL-treated islets occurs at a step in the secretory process that may involve protein kinase-C. These data further support observations that PRL treatment increases islet sensitivity to glucose.

Similarities in the stimulus-secretion coupling mechanisms of glucose- and 2-keto acid-induced insulin release.

The stimulus-secretion coupling of 2-keto acid-induced insulin release was investigated using 2-ketoisocaproate (4-methyl-2-oxopentanoate) as the principal model secretagogue. 2-Ketoisocaproate and 2-ketocaproate (2-oxo-, hexanoate) provoked changes in B cell electrical behavior characterized by an initial depolarization of the membrane potential, followed by rapid spike activity, which appeared either in a bursting pattern or as continuous activity. The onset of spike activity induced by 2-ketoisocaproate (5 mM) was biphasic in nature. The dynamic pattern of 2-ketoisocaproate-induced insulin release was also biphasic. 2-[U-14C]Ketoisocaproate (10 mM) was oxidized in islet tissue at a rate equivalent to that of [U-14C]glucose (17 mM) and a t a higher rate than 2-ketoisovalerate (3-methyl-2-oxobutyrate) and 2-keto-3-methyl-valerate, which were poor secretagogues. Like glucose, 2-ketoisocaproate provoked characteristic changes in 86Rb and 45Ca efflux from prelabeled islets and stimulated 45Ca net uptake. Proinsulin synthesis was stimulated by 2-ketoisocaproate through both a general effect on protein synthesis and a specific effect on hormonal biosynthesis. 2-Ketoisocaproate and 2-ketocaproate reproduced the effect of glucose on the islet content of ATP, ADP, AMP, NAD+, NADH, NADP+, and NADPH. These findings together with a series of observations on the effects upon the above parameters of site-specific inhibitors, e.g. respiratory inhibitors, suloctidil, theophylline, and epinephrine, suggested that the stimulus-secretion-coupling mechanisms for 2-ketoisocaproate- and glucose-induced release are similar. It is postulated that glucose- and 2-keto acid-induced insulin release may be initiated by a common signal.

Effects of glucose on insulin release and 86Rb permeability in cultured neonatal and adult rat islets.

Glucose-induced insulin release and modifications in 86Rb outflow were studied in cultured neonatal and adult rat islets. The dose-response curve for neonatal islets was steeper than for adult islets and the maximal response was clearly shifted towards lower glucose concentrations. In neonatal islets, glucose-induced insulin release was inhibited by the Ca2+-channel blocker, nifedipine. In the absence of glucose, the 86Rb outflow from neonatal islets was lower than from adult islets. Also, the glucose-induced reduction in 86Rb outflow was less pronounced in neonatal islets. Altered K+ permeability in the B-cell membrane could explain the change in glucose sensitivity of neonatal islets.

Characterization of potassium channels in pancreatic beta cells from ob/ob mice.

The patch-clamp technique in the cell-attached mode was used to study the K channels present in the membrane of cultured pancreatic beta cells from ob/ob mice. Three types of K+ channels were regularly observed, with conductances of 64, 20 and 146 pS. The conduction and kinetic properties of the 64 pS channel were similar to those of the ATP-sensitive potassium channel from normal beta cells. Furthermore, glucose blocked the activity of this channel at the same concentrations as that reported for normal cells. The 20 pS and the 146 pS were insensitive to glucose. The latter K+ channel appears to be similar to the large conductance voltage-activated potassium channels described in normal rodent beta cells. Thus, potassium channels in ob/ob pancreatic beta cells in culture are in most respects normal. Other factors may account for the abnormal electrical response to glucose of ob/ob pancreatic islets, such as reversible impairment of their function in vivo or defects not related to potassium permeability.

The ATP-sensitive potassium channel in pancreatic B-cells is inhibited in physiological bicarbonate buffer.

The effects of bicarbonate buffer (HCO3-/CO2) on the activity of the two K+ channels proposed by some to control the pancreatic B-cell membrane response to glucose were studied. Single K+-channel records from membrane patches of cultured B-cells dissociated from adult rat islets exposed to a glucose- and bicarbonate-free medium (Na-Hepes in place of bicarbonate) exhibit the activity of both the ATP-sensitive as well as the [Ca2+]i-activated K+ channels. However, in the presence of bicarbonate-buffered Krebs solution, the activity of the ATP-sensitive K+ channel is inhibited leaving the activity of the K+ channel activated by intracellular [Ca2+]i unaffected. In the absence of bicarbonate (Hepes/NaOH in place of bicarbonate), lowering the external pH from 7.4 to 7.0 also has differential effects on the two K+ channels. While the K+ channel sensitive to ATP is inhibited, the K+ channel activated by a rise in [Ca2+]i is not affected. To determine whether the response of the B-cell in culture to bicarbonate is also present when the B-cell is functioning within the islet syncytium, the effects of bicarbonate removal on membrane potential of B-cells from intact mouse islets were compared. These studies showed that glucose-evoked electrical activity is also blocked in bicarbonate-free Krebs solution. Furthermore, in the absence of bicarbonate and presence of glucose (11 mM), electrical activity was recovered by lowering the pHo from 7.4 to 7.0. The ATP-sensitive K+-channel activity is greatly reduced by physiologically buffered solutions in pancreatic B-cells in culture. The most likely explanation for the bicarbonate effects is that they are mediated by cytosolic pH changes. Removal of bicarbonate (keeping the external pH at 7.4 with Hepes/NaOH as buffer) would increase the pHi. Since the activity of the [Ca2+]i-dependent K+ channels is not affected by the removal of the bicarbonate buffer, our patch-clamp data in cultured B-cells indicate an involvement of [Ca2+]i-activated K+ channels in the control of the membrane potential. For the B-cell in the islet, we propose that the burst pattern of electrical activity (Ca2+ entry) is controlled, at least in part, by the [Ca2+]i-activated K+ channel.

Prolactin treatment increases GLUT2 but not the G protein subunit content in cell membranes from cultured neonatal rat islets.

Neonatal rat islets exhibit a reduced secretory response to glucose, compared to adult rat islets. The maturation of the secretory response is stimulated by prolactin (PRL). We show here by immunoblot analysis that PRL increases the beta-cell/liver glucose transporter GLUT2 in membrane fractions from cultured neonatal rat islets. This increase (+86%) may explain, at least in part, the development of a mature glucose response. G proteins modulate insulin secretion from pancreatic beta-cells. We show here by immunoblot analysis that, in contrast to the effect on GLUT2, PRL treatment does not modify the G protein subunits alpha i2, alpha i3, alpha o, alpha s, alpha q and beta 35 and beta 36, in cultured neonatal islets.

Resistance to apamin of the Ca2+-activated K+ permeability in pancreatic B-cells.

The bee venom neurotoxin apamin failed to affect 86Rb outflow and insulin release from rat pancreatic islets stimulated by D-glucose or the Ca2+-ionophore A23187. Apamin, in contrast to quinine or A23187, also failed to affect bioelectrical activity in mouse islet cells. These findings suggest that, like in erythrocytes, and at variance with the situation found in smooth muscle, liver or neuroblastoma cells, the Ca2+-activated K+ permeability in the pancreatic B-cell is resistant to apamin.

A new class of calcium channels activated by glucose in human pancreatic beta-cells.

Single calcium-channel currents were recorded from membrane patches of cultured beta-cells dissociated from human islets of Langerhans. In the absence of exogenous glucose, low frequency spontaneous calcium-channel openings of small amplitude (-0.34 +/- 0.02 pA at 0 mV pipet potential) were observed in all membrane patches examined (25 mM Ca2+ in the patch pipet). The frequency of channel openings was rather insensitive to the membrane potential across the patch (range from ca 0 to 60 mV pipet potential; chord conductance 4.9 +/- 0.2 pS). Addition of glucose induced a dose-dependent increase in the frequency of openings of the Ca2(+)-channel (from now on referred to as the CaG-channel). A few minutes after the addition of glucose (greater than or equal to 11 mM), bursts of action potentials were often observed which were elicited only if Ca2+ was present in the solution bathing the beta-cells. Application of glucose in the presence of mannoheptulose (11 mM), a blocker of the hexokinase controlling the first stage of glycolysis, had no effect and the activity of the CaG-channel remained at its resting level. The readily permeant mitochondrial substrate 2-keto-isocaproate (KIC, 10 mM) was as effective as glucose in eliciting action potentials from cells forming part of cell aggregates. The activity of the CaG-channel was significantly increased by KIC (11 mM). Although spike and Ca2(+)-channel activity were markedly stimulated by glucose or KIC in all cells examined, regular bursts of action potentials were seen only if the patch was formed on beta-cells which were part of a cell aggregate. Mannoheptulose (11 mM) prevented the activation of the CaG-channel by glucose (11 mM) but not by KIC (11 mM). Once activated, the CaG-channel remained active even after excision of the patch. We propose that the physiological control of this Ca2(+)-channel is mediated by one or more products of glucose metabolism.

Quantitative immunocytochemical study of islet cell populations in diabetic calmodulin-transgenic mice.

The present study describes the changes in the endocrine pancreas of severely diabetic calmodulin-transgenic mice using light microscopic immunocytochemical and morphometric techniques. A marked reduction in the number and volume of islets, together with distortion of their normal architecture, was found in diabetic mice. In addition, the volume density of both endocrine tissue and B-cells was decreased. An irregular distribution of non-B-cells was also observed in diabetic animals. The volume density and the percentage of A-cells appeared increased. However, when quantified per area unit, the number of all the islet cell types diminished, although only the decrease in B-cell number was statistically significant. The decrease in B-cell mass might account for the diabetic state developed in this animal model.

Dissociation by methylamine of insulin release from glucose-induced electrical activity in isolated mouse islets of Langerhans.

The effect of methylamine on electrical activity and simultaneously measured insulin release was investigated in single perifused islets of normal mice. Methylamine, (2 mmol/L or 6 mmol/L) failed to affect beta-cell input resistance and only caused a modest and transient inhibition of electrical activity of islets exposed to 11.1 mmol/L glucose. Methylamine (2 mmol/L) inhibited insulin release evoked by a five-minute rise in glucose concentration from 5.6 to 22.2 mmol/L, even when the glucose-induced electrical activity remained unaltered. Methylamine, at 2 or 5 mmol/L, partially inhibited insulin release but failed to affect the continuous electrical activity in islets exposed throughout to 22.2 mmol/L glucose. At 10 mmol/L, methylamine reduced both insulin release and electrical activity. These data reinforce the idea that the glucose-induced changes in beta-cell membrane potential represent an early event in the process of stimulus-secretion coupling and can be dissociated from the subsequent process of insulin release.

Characterization and control of pulsatile secretion of insulin and glucagon.

Periodic oscillation of insulin and glucagon by isolated mice islets has been studied. Pulsatile secretion of insulin and glucagon was observed at all glucose concentrations tested. The frequency of oscillation per 20 min for glucagon was 5.0 +/- 0.26 and for insulin 4.0 +/- 0.26 (n = 6), approximating to periodicities of 4 and 5 min, respectively. These did not change by increasing the glucose concentration to 11.1 or 22.2 mM from 5.5 mM (basal). The maximal amplitude of glucagon secretion was not altered by raising the glucose concentration to 11.1 mM from basal. However, 22.2 mM glucose significantly suppressed the amount of glucagon released when compared with glucagon secretion in the presence of 5.5 mM glucose. In contrast, the maximal amplitude of insulin increased from 444.2 +/- 37.7 to 777.2 +/- 61.4 and from 271.8 +/- 35 to 701 +/- 26.5 pg/min (p less than 0.01, n = 6) by switching from basal to 11.1 and 22.2 mM glucose, respectively. We conclude from this study that the pacemaker controlling pulsatile secretion of insulin and glucagon is within the islet. Although the amplitude of secretion of these hormones is regulated by the ambient glucose concentration, the frequency of their pulsatile secretion is not.

Effects of leucine on insulin secretion and beta cell membrane potential in mouse islets of Langerhans.

Leucine is known to enhance insulin secretion from islets of Langerhans, and insulin promotes leucine uptake in peripheral tissues. The present studies were designed to elucidate the effects of leucine on glucose responsiveness and stimulus secretion coupling in mouse islets of Langerhans. The effects of 20 mM leucine on insulin secretion and membrane potential were studied over a range of glucose concentrations (0-27.7 mM). Microdissected, perifused pancreatic islets from normal adult mice were used for both studies of insulin secretion and electrophysiology in order to make a close comparison between these measurements. Leucine enhanced the insulin secretion in the presence of 5.6, 11.1, and 22.2 mM glucose. In the presence of leucine, 27 mM glucose inhibited insulin secretion. In the absence of glucose-leucine did not induce electrical activity of the beta cell membrane, whereas in the presence of 5.6, 11.1, and 22.2 mM glucose leucine increased spike frequency. Thus, leucine shifts both the glucose-dependent insulin secretion and electrical activity toward lower glucose concentrations. It is concluded that leucine and glucose share a common metabolic pathway (citric acid cycle) for stimulatory effects. Leucine is deaminated to form 2-ketoisocaproic acid (KIC) and produce NH4+. We propose that in the absence of glucose this increases cytosolic pH, which in turn increases K+ permeability, and inhibits electrical activity and insulin secretion.

Differences in K+ permeability between cultured adult and neonatal rat islets of Langerhans in response to glucose, tolbutamide, diazoxide, and theophylline.

The effects of glucose, tolbutamide, and diazoxide on K+ permeability in neonatal and adult rat pancreatic islets, maintained in culture 1 week, were investigated by measuring the 86Rb outflow rate from prelabeled islets. In the absence of glucose, the 86Rb efflux was significantly lower in neonatal than adult islets. Raising the glucose concentration to 2.8, 5.6, 8.3, and 11.1 mM produced a marked reduction in the 86Rb efflux in adult islets but only a minor reduction in neonatal islets. The effect of tolbutamide to reduce, and diazoxide to increase, the 86Rb efflux was also less in neonatal islets. These results are discussed with respect to previously reported differences in insulin secretion from neonatal and adult islets in culture.

Effects of pulsatile glucose stimuli on long-term insulin secretory patterns in islets of Langerhans microdissected from Syrian hamsters.

The long-term effects of continuous and pulsatile glucose stimulation of islets of Langerhans microdissected from Syrian hamsters were examined. In the presence of a continuous glucose stimulus insulin secretion peaked during the first 3 h of stimulation followed by a decrease. In the presence of 11.2 mM glucose a second smaller peak of insulin secretion was observed 14-16 h after the perifusion started. Irrespective of the glucose concentration, insulin secretion then steadily decreased and reached very low levels by the end of the 48-h perifusion. However, glucose stimulus provided in a pulsatile manner appeared to reduce this rate of decrease in insulin secretion. Thus, after 48 h, islets exposed to the pulsatile glucose stimulus showed greater insulin responsiveness to glucose than those exposed to a constant glucose stimulus.

Effects of triiodothyronine on insulin release from cultured neonatal and adult rat islets.

Insulin secretion from neonatal and adult rat islets maintained in culture for 7-9 days in the absence or in the presence of 10 nM T3 was measured. In both neonatal and adult islets T3 treatment tends to inhibit insulin secretion only in the absence or in the presence of low glucose concentrations.