Muscarinic-agonist and guanine nucleotide activation of polyphosphoinositide phosphodiesterase in isolated islet-cell membranes

Cinderella story: PI4P goes from precursor to key signaling molecule

Phosphatidylinositol lipids are signaling molecules involved in nearly all aspects of cellular regulation. Production of phosphatidylinositol 4-phosphate (PI4P) has long been recognized as one of the first steps in generating poly-phosphatidylinositol phosphates involved in actin organization, cell migration, and signal transduction. In addition, progress over the last decade has brought to light independent roles for PI4P in membrane trafficking and lipid homeostasis. Here, we describe recent advances that reveal the breadth of processes regulated by PI4P, the spectrum of PI4P effectors, and the mechanisms of spatiotemporal control that coordinate crosstalk between PI4P and cellular signaling pathways.

Delta cell secretory responses to insulin secretagogues are not mediated indirectly by insulin

AIMS/HYPOTHESIS

Somatostatin from islet delta cells inhibits insulin and glucagon secretion, but knowledge of the regulation of pancreatic somatostatin release is limited. Some insulin secretagogues stimulate somatostatin secretion, and here we investigated whether delta cell secretory responses are indirectly regulated in a paracrine manner by insulin released from beta cells.

METHODS

Hormone release from static incubations of primary mouse islets or somatostatin-secreting TGP52 cells was measured by RIA. mRNA expression was assessed by RT-PCR.

RESULTS

Glucose and a range of other physiological and pharmacological agents stimulated insulin and somatostatin release, and insulin receptor mRNA was expressed in islets, MIN6 beta cells and TGP52 cells. However, exogenous insulin did not modulate basal or glucose-induced somatostatin secretion from islets, nor did pre-incubation with an antibody against the insulin receptor or with the insulin receptor tyrosine kinase inhibitor, HNMPA(AM)(3). Glucose and tolbutamide stimulated somatostatin release from TGP52 cells, whereas a range of receptor-operating agents had no effect, the latter being consistent with a lack of corresponding receptor mRNA expression in these cells. Parasympathetic activation stimulated insulin, but inhibited somatostatin release from mouse islets in accordance with differences in muscarinic receptor mRNA expression in islets, MIN6 and TGP52 cells. The inhibitory effect on somatostatin secretion was reversed by pertussis toxin or the muscarinic receptor 2 antagonist, methoctramine.

CONCLUSIONS/INTERPRETATIONS

A number of insulin secretagogues have analogous effects on insulin and somatostatin release, but this similarity of response is not mediated by an indirect, paracrine action of insulin on delta cells.

Oscillatory control of insulin secretion

Pancreatic beta-cells possess an inherent ability to generate oscillatory signals that trigger insulin release. Coordination of the secretory activity among beta-cells results in pulsatile insulin secretion from the pancreas, which is considered important for the action of the hormone in the target tissues. This review focuses on the mechanisms underlying oscillatory control of insulin secretion at the level of the individual beta-cell. Recent studies have demonstrated that oscillations of the cytoplasmic Ca(2+) concentration are synchronized with oscillations in beta-cell metabolism, intracellular cAMP concentration, phospholipase C activity and plasma membrane phosphoinositide lipid concentrations. There are complex interdependencies between the different messengers and signalling pathways that contribute to amplitude regulation and shaping of the insulin secretory response to nutrient stimuli and neurohormonal modulators. Several of these pathways may be important pharmacological targets for improving pulsatile insulin secretion in type 2 diabetes.

Mechanisms and physiological significance of the cholinergic control of pancreatic beta-cell function

Acetylcholine (ACh), the major parasympathetic neurotransmitter, is released by intrapancreatic nerve endings during the preabsorptive and absorptive phases of feeding. In beta-cells, ACh binds to muscarinic M(3) receptors and exerts complex effects, which culminate in an increase of glucose (nutrient)-induced insulin secretion. Activation of PLC generates diacylglycerol. Activation of PLA(2) produces arachidonic acid and lysophosphatidylcholine. These phospholipid-derived messengers, particularly diacylglycerol, activate PKC, thereby increasing the efficiency of free cytosolic Ca(2+) concentration ([Ca(2+)](c)) on exocytosis of insulin granules. IP3, also produced by PLC, causes a rapid elevation of [Ca(2+)](c) by mobilizing Ca(2+) from the endoplasmic reticulum; the resulting fall in Ca(2+) in the organelle produces a small capacitative Ca(2+) entry. ACh also depolarizes the plasma membrane of beta-cells by a Na(+)- dependent mechanism. When the plasma membrane is already depolarized by secretagogues such as glucose, this additional depolarization induces a sustained increase in [Ca(2+)](c). Surprisingly, ACh can also inhibit voltage-dependent Ca(2+) channels and stimulate Ca(2+) efflux when [Ca(2+)](c) is elevated. However, under physiological conditions, the net effect of ACh on [Ca(2+)](c) is always positive. The insulinotropic effect of ACh results from two mechanisms: one involves a rise in [Ca(2+)](c) and the other involves a marked, PKC-mediated increase in the efficiency of Ca(2+) on exocytosis. The paper also discusses the mechanisms explaining the glucose dependence of the effects of ACh on insulin release.

Pancreatic beta-cell receptors and G proteins coupled to adenylyl cyclase

Glucagon and tGLP-1 receptors can be either coexpressed or selectively expressed in beta-cell models. Our results indicate that both these peptides can regulate insulin secretion from beta-cells through their own specific receptors. The finding of a selective expression of G proteins in insulin and glucagon cells indicates a clear difference in their transduction pathways. A key role of the G alpha s family in beta-cell function is further supported by its conserved cell distribution between different species. In conclusion, one could postulate that in the human beta-cells, tGLP-1 and glucagon receptors could mediate their action through different G protein alpha-subunits of the G alpha s family.

Calcium as a second messenger of the action of transforming growth factor-beta on insulin secretion

In MIN6 insulinoma cells, transforming growth factor-beta (TGF-beta) induced the oscillatory elevation of the cytoplasmic free calcium concentration, [Ca2+]c, in the presence of 5.5 mM glucose. The increase in [Ca2+]c induced by TGF-beta was totally dependent on calcium entry and attenuated by nifedipine or nickel chloride. In contrast, carbachol elevated [Ca2+]c in the presence of nickel chloride. When the plasma membrane was hyperpolarized by diazoxide, TGF-beta did not raise [Ca2+]c, whereas both carbachol and depolarizing concentration of potassium elevated [Ca2+]c under the same conditions. TGF-beta did not affect either the cellular cyclic AMP or inositol trisphosphate levels. In the presence of 5.5 mM glucose, TGF-beta induced a 3-fold increase in insulin secretion and the effect of TGF-beta was blocked by either nifedipine or nickel chloride. TGF-beta did not stimulate insulin secretion in the presence of 100 microM diazoxide, whereas both carbachol and 40 mM potassium chloride significantly increased insulin secretion. These results suggest that TGF-beta induces the oscillatory elevation of [Ca2+]c in MIN6 cells by stimulating calcium entry via voltage-dependent calcium channels. Calcium is an intracellular messenger of the action of TGF-beta on insulin secretion.

Synergistic interaction of glucose and neurohumoral agonists to stimulate islet phosphoinositide hydrolysis

The interaction between neurohumoral agonists and glucose to stimulate phosphoinositide (PI)-specific phospholipase C (PLC) and insulin release was examined. In freshly isolated rat islets, maximal glucose (40 mM), cholecystokinin (CCK; 300 nM), or carbachol (CCh; 1 mM) stimulated PI hydrolysis 6.5-, 9.8-, and 5.7-fold, respectively, above basal. The combination of glucose and CCK or of glucose and CCh, but not of CCK and CCh, synergistically increased PI hydrolysis 23.2- and 21.6-fold, respectively, indicating that these secretagogues activate PLC by distinct pathways and that there is an interaction between them. This synergy was maximal at physiological concentrations of stimulatory glucose (8-10 mM) and was paralleled by a marked synergistic stimulation of insulin secretion. The enhanced PI response was partially Ca2+ dependent and may involve the activation of distinct isozymes of PLC, which we identify in islets. These studies demonstrate for the first time a unique and highly sensitive synergistic interaction between glucose and neurohumoral agonists to stimulate PI hydrolysis, and they suggest that enhanced PI hydrolysis is important in the potentiation of glucose- and neurohumoral-stimulated insulin secretion.

G-proteins coupled to phosphoinositide hydrolysis in the cochlear and vestibular sensory epithelia of the rat are insensitive to cholera and pertussis toxins

In the cochlear (CSE) and vestibular sensory epithelia (VSE), phosphoinositides are hydrolyzed in response to stimulation of phospholipase C (PLC) by cholinergic muscarinic and purinergic P2y agonists. Such receptor-mediated activation of PLC is expected to be coupled through guanine nucleotide-binding proteins (G-proteins). Although several classes of G-proteins have been identified in the inner ear, nothing is known about the type of G-proteins associated with the phosphoinositide second messenger system in CSE and VSE. Phosphoinositide hydrolysis was determined by the release of radiolabeled inositol phosphates (InsPs). Ten mM NaF plus 10 microM AlCl3 increased basal InsPs accumulation 2-fold in both CSE and VSE of the rat. Release of InsPs was also enhanced by guanosine 5'-O-(3-thiotriphosphate) (GTP-gamma-S) in saponin-permeabilized tissues. Furthermore, release of InsPs stimulated by both carbamylcholine (CCh) and adenosine 5'-O-[3-thiotriphosphate] (ATP-gamma-S) was significantly inhibited by 100 microM guanosine 5'-O-[2-thiodiphosphate] (GDP-beta-S). These results strongly suggest the involvement of G-proteins in the receptor-PLC coupling in CSE and VSE. ADP-ribosylation in membrane fractions of CSE and VSE in the presence of cholera toxin (CTX) or pertussis toxin (PTX) indicated the existence of Gs- and G(i)-type G-proteins. However, neither CTX nor PTX affected basal or agonist-stimulated release of InsPs. These observations suggest that muscarinic and P2y purinergic receptors are coupled to PLC via CTX- and PTX-insensitive G-proteins in CSE and VSE.

Stimulation of phosphoinositide hydrolysis by glyceraldehyde-3-phosphate in electroporated HIT-T15 cells

HIT-T15 cells labeled with myo-[3H] inositol were permeabilized by electroporation and subsequently stimulated with various glycolytic intermediates in the presence of 20 mM LiCl in a buffer mimicking cytosolic ionic composition. Of the various glycolytic intermediates tested, only D-glyceraldehyde-3-phosphate (G3-P) stimulated the formation of labeled inositol phosphates. The half-maximal response to G3-P occurred at a concentration of 0.75 mM. Formation of inositol phosphates in electroporated cells was also observed in response to GTP. G3-P further potentiated the formation of inositol phosphates in response to GTP, however, the interaction between G3-P and GTP was additive rather than synergistic, indicating that G3-P stimulates phosphoinositide hydrolysis in a manner different than the receptor mediated GTP-dependent activation of phospholipase C. The potentiation of the GTP response by G3-P did not appear to involve inhibition of the GTPase activity of a phosphoinositide-specific G protein, since G3-P also potentiated the formation of inositol phosphates in response to GTP-gamma-S or NaF in a nearly additive manner.

Effects of cholinergic agonists on diacylglycerol and intracellular calcium levels in pancreatic beta-cells

We have studied the effects of cholinergic agonists on the rates of insulin release and the concentrations of diacylglycerol (DAG) and intracellular free Ca2+ ([Ca2+]i) in the beta-cell line MIN6. Insulin secretion was stimulated by glucose, by glibenclamide and by bombesin. In the presence of glucose, both acetylcholine (ACh) and carbachol (CCh) produced a sustained increase in the rate of insulin release which was blocked by EGTA or verapamil. The DAG content of MIN6 beta-cells was not affected by glucose. Both CCh and ACh evoked an increase in DAG which was maximal after 5 min and returned to basal after 30 min; EGTA abolished the cholinergic-induced increase in DAG. ACh caused a transient rise in [Ca2+]i which was abolished by omission of Ca2+ or by addition of devapamil. Thus, cholinergic stimulation of beta-cell insulin release is associated with changes in both [Ca2+]i and DAG. The latter change persists longer than the former and activation of protein kinase C and sensitization of the secretory process to Ca2+ may underlie the prolonged effects of cholinergic agonists on insulin release. However, a secretory response to CCh was still evident after both [Ca2+]i and DAG had returned to control values suggesting that additional mechanisms may be involved.

A Gq-type G protein couples muscarinic receptors to inositol phosphate and calcium signaling in exocrine cells from the avian salt gland

Muscarinic acetylcholine receptor (mAChR) activation in isolated cells from the nasal salt gland of the domestic duck (Anas platyrhynchos) results in a rapid increase in the rate of phosphatidylinositol hydrolysis and pronounced intracellular calcium signals. Both responses can be elicited by treating these cells with fluoroaluminate (AlF4-) indicating the involvement of a heterotrimeric G protein in the transmembrane signaling process. To characterize this G protein, electrophoretically separated membrane proteins were blotted onto nitrocellulose filters and probed with peptide-antibodies raised against portions of different alpha-subunits of mammalian G proteins. We could demonstrate the presence of at least four different G proteins in salt gland cell membranes. Two of these proteins (40 and 41 kD) were ADP-ribosylated by pertussis toxin and were recognized by an antiserum against a common sequence in all G protein alpha-subunits. One protein (46 kD) was a cholera toxin-substrate and was recognized by a Gs-specific antiserum; the other (42 kD) was recognized by Gq-specific antisera and was resistant to ADP-ribosylation. Since the initial inositol phosphate production upon receptor activation with carbachol and the resulting calcium signals were not affected by pertussis toxin-pretreatment of salt gland cells, we conclude that muscarinic receptors are coupled to phospholipase C by a Gq-type G protein.

Immunoprecipitation of a pertussis toxin substrate of the G(o) family from rat islets of Langerhans

Rat islets express a pertussis toxin sensitive G-protein involved in receptor-mediated inhibition of insulin secretion. This has been assumed previously to represent "G(i)" which couples inhibitory receptors to adenylate cyclase. Incubation of islet G-proteins with 32P-NAD and pertussis toxin resulted in the labelling of a band of molecular weight 40,000. This band was very broad and did not allow resolution of individual components. Incubation of the radiolabelled proteins with an anti-G(o) antiserum resulted in specific immunoprecipitation of a 32P-labelled band. These results demonstrate that the complement of pertussis toxin sensitive G-proteins in rat islets includes G(o).

Carbachol stimulation of phospholipase A2 and insulin secretion in pancreatic islets

Arachidonic acid has been implicated as a second messenger in insulin secretion by islets of Langerhans. D-Glucose, the major physiological stimulus, increases unesterified arachidonate accumulation in islets. We now show, for the first time, that the muscarinic agonist carbachol, at concentrations which stimulate insulin secretion, causes a rapid and nearly 3-fold increase in arachidonic acid accumulation in islets. The combination of glucose and carbachol has an additive effect on unesterified arachidonate release. There is a large component of secretagogue-induced arachidonate accumulation that is independent of extracellular Ca2+. Carbachol stimulation of arachidonic acid release is mediated by activation of phospholipase A2, as demonstrated by early increases in endogenous lysophosphatidylcholine. In addition to phospholipase A2 activation, carbachol-induced arachidonic acid accumulation also appears to involve diacylglycerol hydrolysis, since the diacylglycerol lipase inhibitor RG80267 partly inhibited arachidonic acid accumulation. In contrast, glucose-induced arachidonic acid accumulation appears to reflect diacylglycerol hydrolysis entirely. Our observations indicate that phospholipase A2 has an important role in muscarinic-induced insulin secretion.

Regulation of phospholipase C delta activity by sphingomyelin and sphingosine

Phospholipase C delta (PLC delta) is strongly inhibited by sphingomyelin (SM). The inhibition occurs in both the presence and the absence of spermine, an activator of PLC delta. Phosphatidylethanolamine (PE), phosphatidylcholine (PC), phosphatidylserine (PS), and phosphatidylinositol (PI) also inhibit PLC delta in the presence of spermine but are much less effective than SM. PE and PC activate and PS and PI inhibit PLC delta in the absence of spermine. Again, the inhibition by PS and PI is much weaker than the inhibition observed with SM. Similar or identical effects are observed in detergent micelle and liposome assays. Comparisons of physiological concentrations of SM with concentrations yielding 50% inhibition of PLC delta in vitro indicate that SM is likely to be a major factor in regulating the activity of PLC delta by inhibition. It is proposed that, in vivo, sphingomyelin acts as an inhibitor of PLC delta, which enables the enzyme to be regulated by activation. In certain circumstances, there is a substantial decline in SM and this may lead to a partial relief of the inhibition. PLC delta is activated by sphingosine in the absence of spermine. However, this activation occurs at unphysiologically high concentrations of sphingosine. The effects of SM and sphingosine on PLC delta in marked contrast to those observed with protein kinase C, which is unaffected by sphingomyelin and inhibited by sphingosine.

Phosphatidylinositol availability and polyphosphoinositide synthesis in pancreatic islet cell membranes

Polyphosphoinositide synthesis in isolated islets of the rat was determined by the phosphorylation of endogenous phosphatidylinositol (PtdIns) by PtdIns kinase and [gamma-32P]ATP to form [32P]-phosphatidylinositol 4-phosphate (PtdInsP) in cell homogenates. Glucose stimulation of intact islets resulted in a time- and concentration-dependent reduction in PtdInsP synthesis. Similarly, the stimulation of intact islets with carbachol (CCh), cholecystokinin (CCK-8S), or tolbutamide for 15 min reduced PtdInsP production in a concentration-dependent manner. The effects of glucose, tolbutamide and CCh were reversible. PtdInsP hydrolysis did not account for the reduction in PtdInsP recovery. The addition of exogenous PtdIns to the PtdIns kinase assay significantly increased basal PtdInsP levels. In addition, exogenous PtdIns completely reversed the inhibitory effects of glucose and increased PtdIns kinase activity in homogenates of glucose-stimulated islets to levels found in control homogenate with PtdIns. Exogenous PtdIns also increased PtdIns kinase activity in CCK-8S-treated islets, although exogenous PtdIns did not overcome the tolbutamide-induced inhibition of PtdIns kinase. The Vmax of PtdIns kinase in homogenates of islets treated with tolbutamide was reduced significantly, although glucose did not affect the Vmax. In addition, the Km values for ATP and PtdIns were not altered by exposure of the islets to cell stimuli. The results suggest that the level of PtdIns in islet cell membranes is rate limiting for PtdInsP synthesis, and that tolbutamide is a noncompetitive inhibitor of PtdIns kinase.

Evidence against a role for a pertussis toxin-sensitive G protein in Ca2+ mobilization in rat parotid acinar cells

Hormone-induced Ca2+ mobilization in rat parotid acinar cells is reportedly mediated via an as yet uncharacterized G protein. We have studied the sensitivity to pertussis toxin (PTx) of this signal transduction mechanism. When rats were treated with Ptx (1.3-1.5 micrograms per animal) for 72 h, a 41 kDa membrane protein was ADP-ribosylated. This PTx treatment regimen, also, resulted in a more than 80% block of the ability of the muscarinic agonist carbachol to inhibit beta-adrenergic receptor-stimulated parotid adenylyl cyclase activity. However, cytosolic Ca2+ levels, in response to either carbachol or AIF-4, were comparable in cells prepared from both untreated or PTx-treated rats, when incubated either in the absence or presence of extracellular Ca2+. Further, both the sensitivity of the Ca2+ response to carbachol and the ability of the agonist-sensitive intracellular Ca2+ stores to be refilled by extracellular Ca2+ were unaffected by PTx treatment. Parotid membranes also contained three low-molecular-weight GTP-binding proteins (25, 22 and 18 kDa) which were unaffected by PTx. These results show that there is only one detectable substrate in parotid membranes for a PTx-catalyzed ADP-ribosylation and that hormone-induced Ca2+ mobilization events in parotid acinar cells are not mediated via PTx-sensitive components.

Muscarinic agonists cause calcium influx and calcium mobilization in forebrain neurons in vitro

We have examined the effects of the muscarinic agonist carbachol on the intracellular free Ca2+ concentration ([Ca2+]i) in primary cultures of neurons from rat forebrain using the Ca2+-sensitive fluorescent dye fura-2. Addition of carbachol increased the [Ca2+]i in approximately 60% of cells studied. Oxotremorine-M, but not pilocarpine, mimicked the effects of carbachol. The response was reduced by 60% on removal of extracellular Ca2+, a finding suggesting that muscarinic receptor activation causes Ca2+ influx in addition to intracellular Ca2+ mobilization. Tetrodotoxin and nitrendipine also significantly reduced the response to carbachol. These studies suggest that the changes in [Ca2+]i produced by activation of muscarinic receptors result in part from mobilization of intracellular Ca2+ and that influx through voltage-sensitive Ca2+ channels also provides a significant contribution to the net [Ca2+]i change observed.

On the transduction mechanism for muscarine-induced inhibition of M-current in cultured rat sympathetic neurones

1. Dissociated adult or fetal rat superior cervical ganglion cells were voltage-clamped through a single patch pipette. The voltage-dependent K+ current, IM (M-current), was maintained by including MgATP in the pipette solution and by buffering the solution pH to 6.7. 2. Bath-applied muscarine (0.4 microM) produced a reversible inhibition of IM. 3. Addition of Gpp(NH)p (200 microM) or GTP-gamma-S (500 microM) to the pipette solution induced a slowly developing inhibition of IM and prevented recovery from subsequent muscarine-induced inhibition. 4. Addition of GDP-beta-S (500 microM) to the pipette solution reduced the amount of IM inhibition produced by 0.4 microM-muscarine by 42% and reduced the associated inward shift of the holding current by 56%. 5. Cells responded normally to muscarine after pre-treatment for 4-27 h with 500 ng ml-1 pertussis toxin (PTx). 6. IM was not diminished by extracellular addition of 1 mM-dibutyryl cyclic AMP, 8-bromo-cyclic AMP or dibutyryl cyclic GMP, or of 10 microM-forskolin. 7. IM was not reduced by inclusion of Li+ (2 mM) or inositol 1,4,5-trisphosphate (IP3, 100 microM) in the patch pipette, nor by ionophoretic injection of IP3 from an inserted micropipette. 8. Addition of 4-beta-phorbol 12,13-dibutyrate (PDBu, 0.5-2 microM) to the extracellular medium partly inhibited IM and reduced an additional component of resting membrane current. This effect was not replicated by 4-alpha-phorbol 12,13-didecanoate. 9. It is concluded that the inhibition of IM by muscarine is mediated through activation of a PTx-insensitive GTP-binding protein. The effect of muscarine appears not to be mediated by cyclic nucleotides or IP3 but may possibly involve the generation of diacylglycerols and activation of protein kinase C.

Prostaglandin E2 inhibits phosphoinositide metabolism in isolated pancreatic islets

Isolated islets of the rat labelled with myo-[3H]inositol showed decreased accumulation of total inositol phosphates (InsPs) and [3H]polyphosphoinositide hydrolysis in response to glucose after preincubation with prostaglandin E2 (PGE2). The response was concentration-dependent and specific for PGE2. PGE2 did not affect basal [3H]phosphoinositide hydrolysis or InsPs accumulation. Pertussis-toxin pretreatment antagonized the response to PGE2, whereas 8-bromo cyclic AMP was without effect. The PGE2-induced decrease in InsPs may contribute to the suppression of insulin secretion.

Effects of Bordetella pertussis toxin on catecholamine inhibition of insulin release from intact and electrically permeabilized rat islets

Noradrenaline- and clonidine-induced inhibition of insulin release from intact and electrically permeabilized rat islets was markedly relieved by prior exposure to 100 ng of Bordetella pertussis toxin/ml. The reversal of catecholamine inhibition of insulin secretion by this toxin was not associated with a decrease in specific binding of the alpha 2-adrenergic ligand [3H]yohimbine, and could not be fully explained by an increase in intracellular cyclic AMP. Exposure of intact islets to 1 microgram of pertussis toxin/ml for 2 h, followed by electrical permeabilization and incubation with 5 microCi of [alpha-32P]NAD+, resulted in the ADP-ribosylation in situ of a protein of molecular mass approx. 41 kDa. These results suggest that pertussis toxin alleviates catecholamine inhibition of beta-cell secretory responses by ADP-ribosylating at least one protein of molecular mass 41 kDa. In analogous systems the 41 kDa substrate of pertussis toxin has been shown to be the alpha subunit of Gi, but catecholamine-activated G proteins linked to effector systems other than adenylate cyclase might also be modified by this toxin in pancreatic beta-cells.

Functionally distinct G proteins selectively couple different receptors to PI hydrolysis in the same cell

The number of G proteins identified by molecular cloning exceeds the number of known G protein functions. Here we show that a cell can possess multiple G proteins that carry out a similar function, the activation of phospholipase C, but couple selectively to different receptors, which are endogenous to the cell or introduced by DNA transfection. These G proteins (termed Gp) can be distinguished by their sensitivity to pertussis toxin. The assignment of a given Gp pathway to specific receptors is confirmed by the additivity relationships of the PI hydrolysis response mediated by the different receptors. Significantly different amounts of PI hydrolysis are activated through each Gp pathway, suggesting that Gp proteins also differ in their coupling to phospholipase C. These results indicate that distinct Gp pathways in a given cell exist to couple different receptors to PI hydrolysis selectively, and may specify the nature of the cellular response to different receptors by determining the magnitude of PI hydrolysis.

Membrane phospholipid turnover as an intermediary step in insulin secretion. Putative roles of phospholipases in cell signaling

One or more phospholipases of the C and A2 types exist in rodent islets and may play a pivotal role in the cell signaling cascade culminating in exocytotic insulin release. Phospholipase C generates myo-inositol-1,4,5-trisphosphate, which mobilizes a "pool" of calcium in the endoplasmic reticulum and which may also secondarily facilitate calcium (Ca++) influx from the extracellular space to replenish that pool. Diacylglycerol is also generated by phospholipase C action and activates protein kinase C; it may thereby potentiate the cellular response to elevations in cytosolic free Ca++ concentration. Arachidonic acid may be released during the degradation of diacylglycerol and may also contribute to islet activation. Phospholipase C is activated by glucose, cholinergic agonists, and probably by Ca++ fluxes. Phospholipase A2 action generates arachidonic acid and lysophospholipids. Certain lysophospholipids mobilize cellular Ca++, at least in part from superficial, plasmalemmal stores. Native (unoxygenated) arachidonic acid also has the capability of mobilizing cellular Ca++ from membrane-bound stores; it may, in addition, activate protein kinase C, as suggested by recent indirect studies. The further metabolism of arachidonic acid via lipoxygenase and cyclo-oxygenase appears to provide positive and negative modulation, respectively, of stimulated insulin secretion. Many pieces of the puzzle remain, however, to be supplied. For example, it has not yet been unequivocally demonstrated that phospholipase A2 is activated by physiologic stimuli in intact islets. Furthermore, the absence of truly specific pharmacologic stimulators or inhibitors of these processes currently precludes precise delineation of the respective physiologic roles of each potential mediator in stimulus-secretion coupling. When such roles are elucidated, it can be asked whether the defects in insulin secretion in diabetes mellitus may be due in part to abnormalities in the turnover of beta-cell membrane phospholipids and the generation of intracellular lipid-derived signals.

GTP- and inositol 1,4,5-trisphosphate-induced release of 45Ca2+ from a membrane store co-localized with pancreatic-islet-cell plasma membrane

Inositol 1,4,5-trisphosphate [Ins(1,4,5)P3], arising from hydrolysis of phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P2], is proposed as the link between membrane-receptor activation and mobilization of Ca2+ from intracellular sites in hormone-secreting cells. The location of Ins(1,4,5)P3-sensitive membranes was investigated in cultured neonatal beta-cells. Membranes were obtained after lysis of cells attached to positively charged Sephadex. After lysis the presence of the enzyme markers 5'-nucleotidase, glucose-6-phosphatase, NADH-cytochrome c reductase, UDP-galactosyltransferase and succinate dehydrogenase indicated the mixed nature of the preparation. After sonication, however, UDP-galactosyltransferase and succinate dehydrogenase activities were undetectable, but 4.8% of total cellular glucose-6-phosphatase and 3.4% of total cellular NADH-cytochrome c reductase remained with 5'-nucleotidase in the preparation, indicating endoplasmic-reticulum association. ATP-dependent 45Ca2+ accumulation was shown in this preparation (410 +/- 24 pmol/mg of protein at 150 nM free Ca2+) and was inhibited by vanadate (100 microM). Ca2+ release was effected by Ins(1,4,5)P3, with half-maximal release at 0.5 +/- 0.14 microM-Ins(1,4,5)P3, t1/2 11.2 +/- 1.1 s. GTP- and guanosine 5'-[beta gamma-imido]triphosphate (p[NH]ppG)-promoted release of 45Ca2+ was demonstrated in this preparation, but the kinetics of release (half-maximal Ca2+ release at 5.4 +/- 0.7 microM, with t1/2 77.3 +/- 6.9 s, and at 51.1 +/- 4.2 microM, with t1/2 19.0 +/- 2.2 s, for GTP and p[NH]ppG respectively), and the ability of neomycin sulphate to block p[NH]ppG-induced release only, are indicative of separate release mechanisms after treatment with these agents. A close association between plasma membrane and elements of the endoplasmic reticulum is indicated in this model, providing a possible mechanism for local alterations in free Ca2+ in the sub-plasma-membrane region.

Glucose and carbachol generate 1,2-diacylglycerols by different mechanisms in pancreatic islets

Diacylglycerols (DAG) modulate secretory responses by the activation of protein kinase C. Early changes in DAG formation induced by the muscarinic receptor agonist carbachol were compared to those caused by the nutrient secretagogue glucose in pancreatic islets. Turnover rates of DAG were investigated in radiolabeling experiments, whereas changes in total mass and fatty acid composition of DAG were assessed by gas-liquid chromatography. When islet lipids were labeled to steady state in tissue culture with [3H]glycerol, carbachol induced a rapid (10 s) and sustained increase of [3H]DAG generation. In contrast, glucose stimulation failed to increase [3H]glycerol containing DAG, and this was probably due to the isotopic dilution of the label secondary to enhanced glycolysis. This was substantiated by following the transfer of 14C from glucose into DAG. Within 1 min of acute exposure of islets to D-[U-14C]-glucose at stimulatory concentrations, DAG labeling increased fivefold representing up to 2% of total glucose usage. Similar stimulation of 14C incorporation into other neutral lipids and inositol phospholipids was observed, suggesting the enhanced de novo synthesis of phosphatidic acid, the common precursor for DAG, and inositol phospholipids from glycolytic intermediates. Transfer of 14C from glucose was not stimulated by agents such as carbachol and exogenous phospholipase C that act primarily on inositol phospholipid breakdown. The total mass of islet DAG was increased by 60% after both carbachol and glucose stimulation. However, analysis of the fatty acid composition of carbachol-generated DAG revealed at the early time point (10 s) a prevalent stearoyl-arachidonoyl configuration similar to that reported for inositol phospholipids. This pattern shifted to a DAG enriched in palmitic acid at a later time point. Glucose-stimulated islets displayed a predominance of palmitic acid containing DAG, indicating increased de novo synthesis of the putative second messenger rather than its formation by inositol phospholipid hydrolysis. Indeed, steady-state labeling of these phospholipids with [3H]inositol confirmed this idea since only carbachol caused detectable inositol phospholipid hydrolysis. Thus, although protein kinase C may be activated by both carbachol and glucose, the two secretagogues generate diacylglycerols through different mechanisms.

Dissociation between intracellular calcium mobilization and insulin secretion in isolated rat islets of Langerhans

The neuropeptide bombesin provoked a dose-dependent stimulation of 45Ca2+ efflux from pre-loaded islets of Langerhans. This response occurred rapidly, was not sustained and did not depend on the presence of extracellular calcium, suggesting that it resulted from the mobilization of intracellular calcium stores. Under conditions when large increases in 45Ca2+ efflux were observed, bombesin completely failed to stimulate the rate of insulin secretion. Similar results were also obtained with the muscarinic cholinergic agonist, carbachol. The data suggest that the release of calcium from intracellular pools is not sufficient to induce an increase in insulin secretion in normal islet cells.

An M2 muscarinic receptor subtype coupled to both adenylyl cyclase and phosphoinositide turnover

To investigate whether a particular receptor subtype can be coupled to multiple effector systems, recombinant M2 muscarinic receptors were expressed in cells lacking endogenous receptor. The muscarinic agonist carbachol both inhibited adenylyl cyclase and stimulated phosphoinositide hydrolysis. The stimulation of phosphoinositide hydrolysis was significantly less efficient and more dependent on receptor levels than the inhibition of adenylyl cyclase. Both responses were mediated by guanine nucleotide binding proteins, as evidenced by their inhibition by pertussis toxin; the more efficiently coupled adenylyl cyclase response was significantly more sensitive. Thus, individual subtypes of a given receptor are capable of regulating multiple effector pathways.

Neomycin inhibits K+- and veratridine-stimulated noradrenaline release in rat brain slices and rat brain synaptosomes

The possible involvement of phosphoinositides' turnover in the process of neurotransmitter release in the central nervous system (CNS) was studied using rat brain slices and synaptosomes. A depolarizing concentration of potassium chloride (25 mM) induces an 8.6 +/- 0.4% increase of [3H]noradrenaline [( 3H]NA) fractional release in cerebral cortical slices above spontaneous release, and 15 mM KCl induces a 3-fold increase of [3H]NA release in rat brain synaptosomes. Neomycin, an aminoglycoside which binds phosphoinositides, inhibits the potassium-induced release in cortical slices with an IC50 = 0.5 +/- 0.07 mM and with IC50 = 0.2 +/- 0.03 mM in synaptosomes. Veratridine, a veratrum alkaloid which increases membrane permeability to sodium ions and causes depolarization of neuronal cells, induces a net 13.4 +/- 0.3% increase of [3H]NA fractional release above spontaneous release in cortical slices. In analogy to K+ stimulation, neomycin inhibits the veratridine-stimulated release in cortical slices with an IC50 = 0.65 +/- 0.1 mM. It appears that the recycling of phosphoinositides, which is necessary for Ca2+ mobilization, participates in the Ca2+-dependent induced neurotransmitter release in the central nervous system.

Regulation of exocytosis in electrically permeabilized insulin-secreting cells. Evidence for Ca2+ dependent and independent secretion

The regulation of insulin secretion from RINm5F cells exposed to high voltage discharge has been investigated. Electron microscopy revealed that the overall structure of the cells was preserved after permeabilization. In this preparation insulin release was stimulated by Ca2+ (EC50 = 2.4 microM). The stable GTP analogue GTP gamma S enhanced secretion both at intermediate (nano- to micromolar) and vanishingly low (less than 10 pM) Ca2+ concentrations. At optimal Ca2+ (10 microM) the effect of GTP gamma S was greatly reduced. We investigated whether the secretory response to GTP analogues was mediated by any of three enzyme systems regulated by GTP-binding proteins, i.e. generation of cyclic AMP by adenylate cyclase, of diacylglycerol by phospholipase C and of arachidonic acid by phospholipase A2. The involvement of these messenger systems could be excluded as (i) cyclic AMP only had minor, Ca2+ dependent effects, (ii) phospholipase C was not activated in the absence of Ca2+ and insulin secretion due to the phorbol ester TPA displayed a different Ca2+ dependency, (iii) arachidonic acid did not elicit Ca2+ independent insulin secretion. These results, taken together with the finding that insulin secretion due to Ca2+ or TPA is attenuated by the inhibitory guanine nucleotide GDP beta S, suggest the existence of a regulatory site in exocytosis which is sensitive to guanine nucleotides.