Evidence for a GTP-binding protein coupling thrombin receptor to PIP2-phospholipase C in membranes of hamster fibroblasts

The exocytosis induced in HL-60 cells by 4-hydroxynonenal, a lipid peroxidation product, is not prevented by reduced glutathione

The effect of reduced glutathione (GSH) was studied on exocytosis triggered by 4-hydroxynonenal in HL-60 cells induced to differentiate towards the granulocytic cell line by dimethylsulfoxide; we measured beta-glucuronidase secretion from cells incubated at 37 degrees C in the presence of 5 mM GSH. GSH addition to the cell suspensions failed to induce any significant change of the exocytosis stimulated by HNE concentrations between 10(-8) and 10(-6) M. In contrast however, 5 mM GSH was able to fully prevent the release of lactate dehydrogenase observed in the presence of 50 microM HNE, a concentration much higher than that able to stimulate the exocytotic secretion. As the activation of phosphoinositide-specific phospholipase C (PLC) has been shown to play a major role in HNE-induced exocytosis, we studied the GSH effect on the breakdown of phosphatidylinositol-4,5-bisphosphate added to plasma membranes isolated from rat neutrophils and incubated in the presence of increasing concentrations of the aldehyde. In neutrophil membranes HNE induced a significant increase of PLC activity when used in the same concentrations as those able to stimulate beta-glucuronidase secretion in DMSO-differentiated HL-60 cells; the presence of 5 mM GSH failed to prevent its action. Our results suggest that these low aldehyde concentrations, which have actually been found in exudates, may increase tissue damage in inflammation through the release of lytic enzymes by neutrophils; it seems unlikely that their effects could be influenced by the levels of -SH groups present in the exudate and by its protein concentration.

Action of 2-nonenal and 4-hydroxynonenal on phosphoinositide-specific phosopholipase C in undifferentiated and DMSO-differentiated HL-60 cells

The promyelocytic cell line HL-60 has been used as an in vitro model to study the mechanism of action of two chemotactic aldehydes, 2-nonenal and 4-hydroxynonenal. Increasing aldehyde concentrations have been added to undifferentiated and DMSO-differentiated cells incubated at 37 degrees C and their effect on phosphoinositide-specific phospholipase C has been analysed by using a specific inositol-1,4,5-tris-phosphate assay system. Concentrations of 2-nonenal between 10(-9) and 10(-7) M significantly increased the enzymatic-activity in DMSO-differentiated HL-60 cells, while 10(-9) and 10(-8) M concentrations were active in the undifferentiated cells. 4-Hydroxynonenal was able to activate phospholipase C both in undifferentiated and DMSO-differentiated cells at concentrations ranging from 10(-8) to 10(-6) M. The concentrations of both compounds active on phospholipase C displayed a good correspondence with those which had been reported to be chemotactic towards rat neutrophils. In the case of 4-hydroxynonenal, the present results confirm its ability to activate phospholipase C, which we had previously shown in isolated neutrophil plasma membranes. The comparison of the effects of 2-nonenal and 4-hydroxynonenal on chemotaxis and phospholipase C activation suggests a common mechanism of action for both aldehydes, for which the presence of the double bond seems to be required.

Activation of phosphoinositide-specific phospholipase C of rat neutrophils by the chemotactic aldehydes 4-hydroxy-2,3-trans-nonenal and 4-hydroxy-2,3-trans-octenal

A comparison has been made between the effects of 4-hydroxy-2,3-trans-nonenal (HNE) and 4-hydroxy-2,3-trans-octenal (HOE), two lipid peroxidation products, on the basal and GTPgammaS-stimulated activities of phosphoinositide-specific phospholipase C (PL-C) of rat polymorphonuclear leukocytes. PL-C activity was determined in vitro by measuring the hydrolysis of [3H] phosphatidylinositol-4,5-bis- phosphate (PtdIns-P2) added as exogenous substrate to neutrophil plasma membranes. PL-C was activated by concentrations of HNE ranging from 10(-8) to 10(-6) M both in the presence and in the absence of 2 x 10(-5) M GTPgammaS; HOE stimulated the enzymatic activity between 10(-11) and 10(-8) M; maximal stimulation was given by 10(-11) M HOE plus GTPgammaS. The aldehyde concentrations able to accelerate PtdIns-P2 breakdown displayed a good correspondence with those which have been reported to stimulate the oriented migration of rat neutrophils. Pretreatment of neutrophils with pertussis toxin prevented the stimulation of PL-C by 10(-11) M HOE and by HOE plus GTPgammaS. Our results suggest that the chemotactic action of HNE and HOE might depend on the activation of PL-C; furthermore a regulatory G protein appears to be involved in the acceleration of PtdIns-P2 turnover by HOE.

Action of lipid peroxidation products on phosphoinositide specific phospholipase C

4-Hydroxynonenal (HNE), a major lipid peroxidation product, displays several biological actions. Among them, the differentiation of human HL-60 cells and the stimulation of neutrophil oriented migration occur at concentrations which can be actually found in normal tissues and in body fluids. In spite of its chemotactic activity, HNE fails to increase neutrophil oxidative metabolism. The action of the aldehyde on cell migration appears to be mediated by a phosphoinositide specific phospholipase C. The acceleration of phosphatidylinositol turnover induced by 10 pM 4-hydroxyoctenal, another lipid peroxidation product, is prevented by the pretreatment of neutrophils with pertussis toxin. The mechanism of action of these 4-hydroxyalkenals appears to follow pathways common to other chemoattractants, but some differences can be found too. In particular HNE seems unable to stimulate phospholipase D activity. The action of 4-hydroxyalkenals and other lipid peroxidation products on transmembrane signalling systems and on phospholipid metabolism might regulate several cell functions, such as motility, proliferation and differentiation.

Inositol-lipid-specific phospholipase C isoenzymes and their differential regulation by receptors

Images

Experimental studies on the mechanism of action of 4-hydroxy-2,3-trans-nonenal, a lipid peroxidation product displaying chemotactic activity toward rat neutrophils

The effects of 4-hydroxy-2,3-trans-nonenal (HNE) and nonanal on the activity of phosphoinositide-specific phospholipase C of rat neutrophils have been studied in parallel with their action on neutrophil oriented migration. Concentrations of HNE ranging from 10(-7) to 10(-5) M significantly stimulated the oriented migration of rat polymorphonuclear leukocytes. HNE stimulated both the basal and GTP gamma S-induced phospholipase C activity when used at concentrations between 10(-8) and 10(-6) M. Nonanal was devoid both of chemotactic activity and of any action on phospholipase C activity. The effect of GTP gamma S on the stimulation of phospholipase C induced by HNE was higher when the lowest dose of the aldehyde was used; the finding of an additive effect between 10(-8) M HNE and 2 x 10(-5) M GTP gamma S suggests that the two compounds may share a final common pathway of action. These results suggest that the chemotactic activity of HNE might be mediated, like that of other more well-known chemoattractants, by the stimulation of phosphoinositide-specific phospholipase C.

Stimulation by epidermal growth factor of inositol phosphate production in plasma membranes from A431 cells

Plasma membranes were isolated from A431 cells previously labelled with myo-[3H]inositol during exponential growth, using a rapid procedure on Percoll gradients. They displayed a significant phospholipase (PLC) activity against phosphoinositides, which was stimulated by guanosine 5'-O-(3-thiotriphosphate) (GTP gamma S), epidermal growth factor (EGF) and fetal calf serum (FCS) (24%, 11% and 97% over controls, respectively). The effect of EGF was not significantly increased by GTP gamma S. Upon addition of cytosol, EGF promoted an almost 100% stimulation of inositol 1,4,5-trisphosphate and inositol bisphosphate generation, which displayed an absolute requirement for GTP gamma S. This dose-dependent effect of cytosol was linear until 60 micrograms/ml of cytosolic protein and decreased afterwards; it was abolished by heat treatment and trypsin hydrolysis, and it was not reproduced by an identical amount of bovine serum albumin. The same biphasic stimulation was observed with phosphotyrosyl proteins immunopurified from cytosol of A431 cells previously stimulated by EGF. Since phosphotyrosyl proteins displayed PLC activity, our data suggest that soluble protein substrates of EGF receptor tyrosine kinase, including PLC, could be involved in the regulation of phosphoinositide hydrolysis in response to EGF. Using phosphatidyl[3H]inositol 4,5-bisphosphate (PIP2) dispersed with unlabelled phosphatidylethanolamine and phosphatidylserine as an exogenous substrate, no stimulation of PLC activity by EGF could be detected, either with membranes or with membranes plus cytosol. It is concluded that EGF might stimulate hydrolysis of phosphoinositides by PLC through complex interactions between plasma membrane and cytosolic factors which still remain to be identified.

Effect of 4-hydroxylalkenals on hepatic phosphatidylinositol-4,5-bisphosphate-phospholipase C

The effects of some 4-hydroxyalkenals, carbonylic products of lipid peroxidation, on hepatic phosphatidylinositol-4,5-bisphosphate (PIP2)-phospholipase C (PL-C) activity were investigated. The enzymatic activity was assayed in vitro by measuring the hydrolysis of [3H]PIP2 added as exogenous substrate to liver membranes. 4-Hydroxyhexenal (HEE), 4-hydroxyoctenal (HOE) and 4-hydroxynonenal (HNE) were able to stimulate both the basal and the GTPgammaS induced PL-C activity, whereas 4-hydroxyundecenal was inactive. HOE was the most active compound, being able to accelerate PIP2 breakdown at concentrations between 10(-12) and 10(-6) M, while in the case of HEE the effective doses ranged from 10(-11) to 10(-7) M and from 10(-9) to 10(-6) M in the case of HNE. 4-Hydroxynonenal was able to increase also bombesin stimulated PL-C activity. As these aldehydes accelerated PIP2 breakdown at doses which can be actually reached in tissues, the effects shown in vitro are likely to occur in vivo.

Functional expression of Ca2(+)-mobilizing alpha-thrombin receptors in mRNA-injected Xenopus oocytes

alpha-Thrombin (TH) initiates a program of intracellular events that lead to DNA replication in quiescent CCL39 Chinese hamster lung fibroblasts via membrane receptors that have yet to be characterized at a molecular level. Functional TH receptors were expressed in Xenopus laevis oocytes following injection of poly(A)+ RNA from TH-responsive CCL39 cells; their presence was demonstrated by TH-stimulated 45Ca2+ efflux or Ca2(+)-dependent Cl- channel activation. In voltage clamp experiments on microinjected oocytes a Ca2(+)-activated Cl- current was detected in response to TH (0.2-10 U/ml). The TH response was blocked by a specific TH inhibitor, and potentiated by addition of FGF or intracellular injection of GTP-gamma-S.

Persistent activation of platelet membrane phospholipase C by proteolytic action of trypsin and thrombin

Trypsin causes rapid activation of intact platelets that mimics many actions of thrombin, including the stimulation of phospholipase C (PLC). We have examined the effects of thrombin and trypsin on PLC in a platelet membrane preparation using exogenous [3H]-phosphatidylinositol 4,5-bisphosphate (PIP2) as substrate. Trypsin induced PIP2 breakdown, which was maximal at 20 micrograms/ml, but was reduced at higher concentrations. alpha- and gamma-Thrombins also stimulated PLC-induced hydrolysis of PIP2 in membranes. This effect was inhibited by leupeptin. Exogenous [3H]phosphatidylinositol 4-monophosphate (PIP) was hydrolyzed in response to both thrombin and trypsin in the same ratio as PIP2. Activation of membrane-bound PLC persisted after removal of thrombin and trypsin. The hydrolysis of [3H]phosphatidylinositol was not activated by alpha-thrombin and trypsin. We examined the question of whether calpain was involved in the observed PLC activation by thrombin and trypsin. Although dibucaine activated a Ca2(+)-dependent protease as judged by the hydrolysis of actin-binding protein and by the activation of phosphoprotein phosphatases, it failed to stimulate the generation of phosphatidic acid in 32P-prelabeled platelets. Moreover, when PLC was assayed in the membranes, the addition of Ca2(+)-activated neutral proteinases did not increase the rate of hydrolysis of either PIP or PIP2. Our results show that proteases such as trypsin and thrombin are able to stimulate membrane-bound PLC, but this activation does not seem to be related to calpain.

Characterization of cytoplasmic and membrane-associated phosphatidylinositol 4,5-biphosphate phospholipase C activities in guinea pig ventricles

The phosphoinositide-specific phospholipase C (PLC) activity present in the soluble and sarcolemmal enriched membrane fraction from guinea pig hearts was characterized using phosphatidyl [3H]inositol 4,5-biphosphate (PIP2) or phosphatidyl [3H]inositol 4-monophosphate (PIP) as substrates. The PLC activities (cytosolic and membrane associated) were specific for polyphosphoinositides (PIP2 and PIP) since no other phospholipids were hydrolyzed at pH 7.0 under various ionic conditions. Both enzymic activities were Ca2(+)-dependent (half maximal activities were achieved around pCa 5.0). The pH, detergent (deoxycholate), divalent (Ca2+ and Mg2+), and monovalent (Na+ and K+) cation dependencies were very similar between the cytosolic and membrane-associated enzyme activities, using either PIP2 or PIP as substrate. Hydrolysis of the polyphosphoinositides was inhibited in the presence of phosphatidylethanolamine, phosphatidylserine, or phosphatidylcholine. Under optimal conditions (pH 7.0, 1 mM Ca2+, 2.5 mM Mg2+, 100 mM Na+ and 0.07% deoxycholate) the specific activities of the cytosolic and membrane-associated enzymes were 19.9 +/- 0.9 and 10.1 +/- 0.9 nmol/min/mg protein, respectively, using PIP2 as substrate. Under the same conditions these activities were 18.1 +/- 1.0 and 8.0 +/- 0.8 nmol/min/mg protein for the cytosolic and membrane fractions, respectively, using PIP as substrate. Based on the similarity of the characteristics of these two PLC enzyme activities, it is suggested that the cytosolic and membrane-associated enzyme forms may be closely related.

Calcium involvement in the muscarinic response of the gastric parietal cell

The influence of extracellular Ca2+ on the mediation of carbachol stimulation in isolated rabbit gastric parietal cells was studied. Removing Ca2+ from extracellular medium caused a 42% decrease of the aminopyrine accumulation due to carbachol with the same EC50 value (approximately 5 microM). A short time depletion in extracellular calcium suppressed the carbachol-dependent Ca2+ influx without affecting Ca2+ release from internal stores (fura-2 measurements). Similarly, the production of inositol phosphates under cholinergic stimulation was reduced by 29%. A rapid increase in Ins(1,4,5)P3 was obtained 5 s after carbachol stimulation, and this increase was not changed in Ca2(+)-depleted medium. In contrast, a 20 min incubation with carbachol caused a 50% reduction in both basal and carbachol-stimulated inositol phosphate accumulations. In conclusion, phospholipase C activation, intracellular Ca2+ release and aminopyrine accumulation were sequentially observed following carbachol stimulation of the isolated gastric parietal cell and extracellular calcium contributed to sustain this acid secretory response.

GTP causes calcium release from a plant microsomal fraction

Studies on a variety of animal cell types have revealed a GTP-specific calcium-releasing mechanism in a non-mitochondrial, microsomal fraction. Here we report that GTP also induces rapid release of calcium from a zucchini (Cucurbita pepo L.) hypocotyl microsomal fraction. Maximal release occurs at 50 microM, and half-maximal release at 8 microM GTP. GTP is highly specific in its effect, and may not be replaced by UTP, ATP, CTP, TTP, GMP, or by non-hydrolysable analogues of GTP. Reuptake of calcium after release does not normally occur; however, this may be induced by non-hydrolysable GTP analogues. Calcium release is also blocked by prior treatment with these analogues.

Interaction of cerebral-cortical membranes with exogenously added phosphatidylinositol 4,5-bisphosphate. Effects on measured phospholipase C activity

Exogenously added phosphatidylinositol 4,5-bisphosphate (PtdInsP2) is rapidly associated with cerebral-cortical membranes. Substrate association with membranes was promoted by Mg2+, but inhibited by bivalent chelators. Once associated with the membrane, the PtdInsP2 was resistant to displacement by EDTA. The apparent phospholipase C activity was dependent on the degree of association of substrate with membranes. After preincubation of membranes with substrate, PtdInsP2 hydrolysis was independent of the incubation volume, indicating that substrate and membrane-associated phospholipase C were not independently diluted. Hydrolysis of the membrane-associated substrate was stimulated by Ca2+, guanosine 5'-[beta gamma-imido]triphosphate (p[NH]ppG), guanosine 5'[gamma-thio]triphosphate and carbachol in the presence of p[NH]ppG. Carbachol in the absence of guanine nucleotides, GDP, GTP, ATP and pyrophosphate was ineffective. These results demonstrate that exogenously added PtdInsP2 substrate is rapidly associated with membranes and hydrolysed by a phospholipase C whose activity is regulated by guanine nucleotides and agonist in the presence of guanine nucleotides. Use of exogenously added substrate for studies on the regulation of membrane phospholipase C requires consideration as to possible effects of incubation conditions on the partitioning of substrate into membranes.

Peptide hormones, cytosolic calcium and renal epithelial response

We have reviewed the evidence that a number of hormones interact with renal tubular epithelial cells. The evidence suggests that in the mammalian renal tubule bradykinin and parathyroid hormone interact with cell surface receptors to initiate the hydrolysis of PIP2 leading to the formation of I 1,4,5P3 and diacylglycerol in the distal and proximal tubule, respectively. The activation of this second messenger system leads to the mobilization of Ca2+ from intracellular stores. Vasopressin does not activate this second messenger system in mammalian renal epithelial cells, and we cannot demonstrate I 1,4,5P3 formation and Ca2+ mobilization either in the rabbit papillary collecting tubules or in MDCK cells. There is evidence emerging, but not discussed here, that angiotensin II may also mediate some of its effects on the mammalian proximal tubule via the inositol polyphosphate second messenger system.

Fibroblast growth factor potentiates receptor- and nonreceptor-mediated stimulation of adenylate cyclase in hamster fibroblasts

Basic fibroblast growth factor (FGF) has no effect alone on the basal cAMP synthesis in Chinese hamster fibroblasts (CCL39) but it potentiates (by up to 50%) the stimulation of adenylate cyclase by prostaglandin E1, cholera toxin or forskolin. This potentiating effect is not abolished by pretreatment of the cells with pertussis toxin, which indicates that it is not due to the withdrawal of a tonic inhibition of adenylate cyclase by the pertussis toxin-sensitive inhibitory GTP-binding protein (Gi). Therefore, we conclude that FGF enhances the activation of adenylate cyclase by the stimulatory GTP-binding protein (Gs). Although activation of protein kinase C in CCL39 cells results in a similar potentiation of cAMP production, we provide evidence that the effect of FGF is not mediated by protein kinase C, since (1) the potentiating effects of FGF and phorbol esters are additive and (2) FGF effect persists after down-regulation of protein kinase C. A role of FGF-induced rise in cytoplasmic Ca2+ can also be ruled out because the FGF effect is not mimicked by a Ca2+ ionophore and it persists in Ca2(+)-free medium. Since a similar potentiating effect on cAMP production is elicited by epidermal growth factor, a mitogen known to activate a receptor tyrosine kinase, we suggest that the FGF effect on adenylate cyclase might be mediated by the tyrosine kinase activity that is very likely to be associated with FGF receptors.

Properties of membranous phospholipase C from WRK1 cell: sensitivity to guanylnucleotides and bacterial toxins

As previously described, WRK1 plasma membrane possesses a vasopressin-sensitive phospholipase C [G. Guillon et al., 1986, FEBS Lett. 196, 155-159]. In the present study, we examined the sensitivity of this enzyme to guanylnucleotides. GTP gamma S induces a time- and dose-dependent stimulation of Ins(1,4,5)P3 and Ins(1,4)P2 accumulation. No accumulation of InsP1, Ins(1,3,4)P3 or Ins(1,3,4,5)P4 occurred under similar conditions. Gpp(NH)p produced the same effect but was less potent. GTP and a nonhydrolyzable analogue of ATP, App(NH)p, were without effect. Calcium also stimulated the phospholipase C activity in a time- and dose-dependent manner. In the absence of calcium, the activity of GTP gamma S was considerably reduced. Physiological calcium concentrations (between 10(-8) and 10(-7) M), allowed maximal GTP gamma S stimulation of phospholipase C activity. In this system, the presence of vasopressin alone did not generate inositol phosphate accumulation. However, this hormone: (i) reduced the lag-time observed during GTP gamma S stimulation, (ii) increased the sensitivity of phospholipase C to GTP and to GTP gamma S, and (iii) did not modify the stimulation of phospholipase C induced by maximal doses of GTP gamma S. Unlike sodium fluoride, GTP gamma S elicited an irreversible activation of phospholipase C. Calcium, GTP gamma S and sodium fluoride stimulated the phospholipase C activity via mechanisms sharing a common step, since their maximal effects were not additive. Cholera toxin treatment, known to produce complete ADP-ribosylation of 'alpha s' subunits, partially reduced the basal and the maximal GTP gamma S-mediated stimulation of phospholipase C activity as well as that caused by vasopressin. This inhibition was not mimicked by treatment with either forskolin or pertussis toxin.

Insulin and IGF-1 receptors contain covalently bound palmitic acid

We have studied the biosynthesis of the insulin receptor in a human hepatoma cell line, HepG2. As previously reported, these cells synthesize a disulphide-bonded alpha 2 beta 2 tetrameric insulin receptor. Labelling of HepG2 cells with [3H]palmitate or [3H]myristate followed by immunoprecipitation with a polyclonal antireceptor antibody revealed the incorporation of palmitate, but not myristate, into the beta-subunit and alpha beta-precursor of the receptor in a hydroxylamine-sensitive linkage. The extracellular alpha-subunit was not labelled, demonstrating the specificity of incorporation. Acylation of the insulin receptor was an early event as judged by fatty acid incorporation into the alpha beta-precursor and prevention by protein synthesis inhibitors. Pulse-chase studies demonstrated the expected processing of the alpha beta-precursor to mature alpha- and beta-subunits, but no evidence for preferential turnover of the fatty acid moiety was found. The site of acylation appears to be in the transmembrane or cytoplasmic domain since proteolytic treatment of intact cells produced a truncated beta-subunit still containing label. Binding studies showed that HepG2 cells contain approximately half as many insulin-like growth factor-1 receptors as insulin receptors, raising the possibility that this receptor may also be acylated. Indeed, immunoprecipitation with the antiinsulin receptor serum of MDCK cells expressing IGF-1 receptors, but not insulin receptors, revealed bands corresponding to the alpha beta-precursor, alpha- and beta-subunits, of which the alpha beta-precursor and beta-subunits incorporated [3H]palmitate but the alpha-subunit did not.

P2-purinoceptor-stimulated phosphoinositide turnover in chick myotubes. Calcium mobilization and the role of guanyl nucleotide-binding proteins

ATP, a trigger of P2-purinoceptor-mediated polyphosphoinositide (PI) turnover in cultured myotubes, increased cytosolic calcium levels in a time- and dose-dependent manner (quin2 fluorescence). The calcium was released from intracellular stores, as acute addition of 5 mM EGTA was without significant effect. Adenosine 5'-(3-thiotriphosphate) and 5'-adenylyl imidodiphosphate also increased intracellular levels of inositol phosphates (InsP) and cytosolic calcium levels. Treatment with cholera or pertussis toxin of myotube cultures did not affect the P2-purinoceptor-mediated InsP increase although PI turnover in permeabilized myotubes was stimulated by guanosine 5'-(3-thiotriphosphate). The results suggest that myotube P2-purinoceptors trigger PI turnover and increase intracellular free calcium levels, via a mechanism insensitive to ADP-ribosylation, by cholera or pertussis toxin of guanyl nucleotide-binding (G) proteins. However, the presence of a phospholipase C-coupled G-protein was otherwise demonstrated.

Thrombin exerts a dual effect on stimulated adenylate cyclase in hamster fibroblasts, an inhibition via a GTP-binding protein and a potentiation via activation of protein kinase C

Previous studies in Chinese-hamster fibroblasts (CCL39 line) indicate that an important signalling pathway involved in thrombin's mitogenicity is the activation of a phosphoinositide-specific phospholipase C, mediated by a pertussis-toxin-sensitive GTP-binding protein (Gp). The present studies examine the effects of thrombin on the adenylate cyclase system and the interactions between the two signal transduction pathways. We report that thrombin exerts two opposite effects on cyclic AMP accumulation stimulated by cholera toxin, forskolin or prostaglandin E1. (1) Low thrombin concentrations (below 0.1 nM) decrease cyclic AMP formation. A similar inhibition is induced by A1F4-, and both thrombin- and A1F4- -induced inhibitions are abolished by pertussis toxin. (2) Increasing thrombin concentration from 0.1 to 10 nM results in a progressive suppression of adenylate cyclase inhibition and in a marked enhancement of cyclic AMP formation in pertussis-toxin-treated cells. A similar stimulation is induced by an active phorbol ester, and thrombin-induced potentiation of adenylate cyclase is suppressed by down-regulation of protein kinase C. Therefore, we conclude that (1) the inhibitory effect of thrombin on adenylate cyclase is the direct consequence of the activation of a pertussis-toxin-sensitive inhibitory GTP-binding protein (Gi) possibly identical with Gp, and (2) the potentiating effect of thrombin on cyclic AMP formation is due to stimulation of protein kinase C, as an indirect consequence of Gp activation. Our results suggest that the target of protein kinase C is an element of the adenylate cyclase-stimulatory GTP-binding protein (Gs) complex. At low thrombin concentrations, activation of phospholipase C is greatly attenuated by increased cyclic AMP, leading to predominance of the Gi-mediated inhibition.

G-proteins, the inositol lipid signalling pathway, and secretion

The formation of the second messenger cyclic AMP (cAMP) is known to be coupled to its receptor via a guanine nucleotide regulatory protein, GS. Ca2+-mobilizing receptors stimulate the hydrolysis of phosphatidylinositol bisphosphate (PtdIns(4,5)P2), which generates two intracellular signals Ins(1,4,5)P3 and diacylglycerol. We review the evidence that this signalling system is also composed of three types of proteins: receptor, G-protein and effector. The G-protein that couples to the effector, polyphosphoinositide phosphodiesterase (PPI-PDE), is a novel G-protein, GP, which is a substrate for pertussis toxin in some cells (e.g. neutrophils and platelets) but not others (e.g. pancreatic acinar cells and GH3 cells). This implies that GP is not a single G-protein but encompasses a family of proteins that can activate PPI-PDE. We have also identified a role for another G-protein, GE, which is involved in the secretory process in mast cells and neutrophils. In this case, neither the receptor nor effector has been identified and the main evidence for proposing this second G-protein is based on the ability of guanine nucleotide analogues (e.g. GTP gamma S) to stimulate secretion independently of PPI-PDE activation.

Transmembrane signalling pathways initiating cell growth in fibroblasts

The mechanisms of growth factor action were studied in a fibroblastic cell line capable of reversible growth arrest in G0-G1. This cell line, derived from Chinese hamster lung, can be stimulated to divide by a limited set of purified growth factors, including EGF, FGF, PDGF, alpha-thrombin (THR), serotonin (5-HT) and insulin. THR and 5-HT stimulate, via a G-protein (Gp), a polyphosphoinositide-specific phospholipase C (PtdIns(4,5)P2-PLC). In contrast, the mitogens EGF, FGF, PDGF, and insulin do not stimulate PtdIns(4,5)P2-PLC unless this pathway has been preactivated by THR or AlF-4. Finally, from the specific inhibitory action of pertussis toxin on THR- and 5-HT-induced DNA synthesis, and from the exploitation of the 5-HT pharmacological tools, we conclude that: (i) there are at least two distinct G-proteins involved in signalling growth: Gp, coupling receptors to PtdIns(4,5)P2-PLC, and Gi, coupling receptors negatively to adenylyl cyclase and probably to other unknown effector(s); (ii) activation of receptor-tyrosine kinases provides an alternate growth factor signalling pathway, independent of Gp- and Gi-mediated actions; and (iii) tyrosine kinases positively 'cross-communicate' with the inositol-lipid pathway (phosphorylation of Gp, PLC, PtdIns kinases...?).

Bradykinin-activated membrane-associated phospholipase C in Madin-Darby canine kidney cells

Previous studies have demonstrated that bradykinin stimulates the rapid release of inositol 1,4,5 trisphosphate (IP3) from membrane phosphatidylinositol 4,5 bisphosphate (PIP2) in Madin-Darby canine kidney (MDCK) cells. Since current evidence would suggest that the activation of phospholipase C (PLC) is mediated through a guanine nucleotide-binding protein in receptor-mediated activation of PLC, we evaluated the role of guanine nucleotide proteins in receptor-mediated (bradykinin-stimulated) activation of PLC in MDCK cells. Bradykinin at 10(-7) M produced a marked increase in IP3 formation within 10 s increasing from a basal level of 46.2 to 686.6 pmol/mg cell protein a 15-fold increase. Pretreatment of MDCK cells in culture with 200 ng/ml of pertussis toxin for 4 h reduced the bradykinin-stimulated response to 205.8 pmol/mg protein. A 41-kD protein substrate in MDCK membranes was ADP ribosylated in vitro in the presence of pertussis toxin. The ADP ribosylation in vitro was inhibited by pretreatment of the cells in culture with pertussis toxin. Membranes from MDCK cells incubated in the presence of [3H]PIP2/phosphatidyl ethanolamine liposomes demonstrated hydrolysis of [3H]PIP2 with release of [3H]IP3 when GTP 100 microM or GTP gamma S 10 microM was added. Bradykinin 10(-7) M added with GTP 100 microM markedly increased the rate of hydrolysis within 10 s, thus demonstrating a similar time course of PLC activation as intact cells. These results demonstrate that bradykinin binds to its receptor and activates a membrane-associated PLC through a pertussis toxin-sensitive, guanine nucleotide protein.

GTP gamma S activation of proto-oncogene expression in transiently permeabilised Swiss 3T3 fibroblasts

A technique of transient permeabilisation has been used to show that the introduction of guanosine 5'-O-(3-thiotriphosphate) (GTP gamma S), a non-hydrolysable analogue of GTP, into intact Swiss 3T3 fibroblasts stimulates phosphoinositide hydrolysis, cyclic AMP accumulation and the activation of c-fos and c-myc proto-oncogenes. Of a number of nucleotide triphosphates introduced into the cells, only GTP and its non-hydrolysable analogues activated inositol phosphate release, suggesting that this response is mediated by guanine nucleotide regulatory (G) protein(s). The data demonstrate that transient permeabilisation provides a method of examining the involvement of G-proteins in nuclear activation.

Phospholipase C activity and substrate specificity in frog photoreceptors

The photoreceptor, like many other cells, undergoes receptor-mediated breakdown of phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P2]. The lack of phosphatidylinositol (PdtIns) breakdown during receptor stimulation suggests the existence of a phospholipase C specific for polyphosphoinositides. Phospholipase C activity in frog rod outer segments was assayed with several substrates. The activity was selective for the polyphosphoinositides, phosphatidylinositol 4-phosphate [PtdIns(4)P] and PtdIns(4,5)P2. PtdIns was hydrolysed at 2% of the rate of hydrolysis of PtdIns(4,5)P2. No activity was detected when phosphatidylcholine, phosphatidylserine, or phosphatidylethanolamine were used as substrates. The enzymatic activity was optimal at neutral pH. These findings suggest, but do not prove, that this activity might contain the light-regulated phospholipase C of the photoreceptor.

Guanine-nucleotide and hormone regulation of polyphosphoinositide phospholipase C activity of rat liver plasma membranes. Bivalent-cation and phospholipid requirements

The effect of the GTP analogue guanosine 5'-[gamma-thio]triphosphate (GTP[S]) on the polyphosphoinositide phospholipase C (PLC) of rat liver was examined by using exogenous [3H]phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P2]. GTP[S] stimulated the membrane-bound PLC up to 20-fold, with a half-maximal effect at approx. 100 nM. Stimulation was also observed with guanosine 5'-[beta gamma-imido]triphosphate, but not with adenosine 5'-[gamma-thio]triphosphate, and was inhibited by guanosine 5'-[beta-thio]diphosphate. Membrane-bound PLC was entirely Ca2+-dependent, and GTP[S] produced both a decrease in the Ca2+ requirement and an increase in activity at saturating [Ca2+]. The stimulatory action of GTP[S] required millimolar Mg2+. [8-arginine]Vasopressin (100 nM) stimulated the PLC activity approx. 2-fold in the presence of 10 nM-GTP[S], but had no effect in the absence of GTP[S] or at 1 microM-GTP[S]. The hydrolysis of PtdIns(4,5)P2 by membrane-bound PLC was increased when the substrate was mixed with phosphatidylethanolamine, phosphatidylcholine or various combinations of these with phosphatidylserine. With PtdIns(4,5)P2, alone or mixed with phosphatidylcholine, GTP[S] evoked little or no stimulation of the PLC activity. However, maximal stimulation by GTP[S] was observed in the presence of a 2-fold molar excess of phosphatidylserine or various combinations of phosphatidylethanolamine and phosphatidylserine. Hydrolysis of [3H]phosphatidylinositol 4-phosphate by membrane-bound PLC was also increased by GTP[S]. However, [3H]phosphatidylinositol was a poor substrate, and its hydrolysis was barely affected by GTP[S]. Cytosolic PtdIns(4,5)P2-PLC exhibited a Ca2+-dependence similar to that of the membrane-bound activity, but was unaffected by GTP[S]. It is concluded that rat liver plasma membranes possess a Ca2+-dependent polyphosphoinositide PLC that is activated by hormones and GTP analogues, depending on the Mg2+ concentration and phospholipid environment. It is proposed that GTP analogues and hormones, acting through a guanine nucleotide-binding protein, activate the enzyme mainly by lowering its Ca2+ requirement.

In vitro synthesis of 32P-labelled phosphatidylinositol 4,5-bisphosphate and its hydrolysis by smooth muscle membrane-bound phospholipase C

For studies of phospholipase C (PLC) activity in cell-free systems, 32P-labelled phosphatidylinositol 4,5-bisphosphate (PIP2) was prepared enzymatically by phosphorylating phosphatidylinositol 4-phosphate (PIP) in the presence of [gamma-32P]ATP using a PIP kinase partially purified from bovine retinae. PLC activity was determined by incubating membranes of DDT1 MF-2 cells with 32P-PIP2 and measuring remaining non-hydrolyzed substrate as well as accumulation of the hydrolysis product, inositol trisphosphate (IP3). Guanine nucleotides stimulated PIP2 hydrolysis and IP3 release. Additional increase in IP3 accumulation was observed with adrenaline plus guanine nucleotides.

Two growth factor signalling pathways in fibroblasts distinguished by pertussis toxin

The primary action of a family of mitogens including bombesin, bradykinin, vasopressin and alpha-thrombin is to activate the hydrolysis of polyphosphoinositides. Hydrolysis of phosphatidylinositol 4,5-bisphosphate (PtdIns(4,5)P2) by phospholipase C is mediated through coupling of surface receptors to a GTP-binding protein (Gp protein) which, in some cells, is inactivated by the toxin of Bordetella pertussis. It is not known whether this signalling pathway is involved in initiating DNA replication, whereas it has been firmly established that reinitiation of DNA synthesis can be triggered without activation of PtdIns(4,5)P2 hydrolysis by, for example, EGF (epidermal growth factor), FGF (fibroblast growth factor) and insulin/IGF-I (insulin-like growth factor-I), members of a class of mitogens known to activate receptor tyrosine kinases. Taking advantage of the fact that Chinese hamster lung fibroblasts respond to either class of mitogens and that their Gp protein appears to be sensitive to pertussis toxin, we have now analysed the toxin's effect on reinitiation of DNA synthesis and find that it inhibits up to 95% of thrombin-induced mitogenicity without affecting EGF- or FGF-induced DNA synthesis and proliferation. These findings strongly suggest that activation of PtdIns(4,5)P2-phospholipase C has a determinant function in growth control, and confirm the existence of alternative growth factor-signalling pathways independent of polyphosphoinositide breakdown.