Molecular and Cellular Biochemistry 198: 19–30, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands.

19

Basic fibroblast growth factor stimulates cytosolic phospholipase A2, phospholipase C-γγ1 and phospholipase D through distinguishable signaling mechanisms Gaurisankar Sa1 and Tanya Das2 1

Animal Physiology Section; 2Immunotechnology Section, Bose Institute, P-1/12 CIT Scheme VII M, Calcutta, India

Received 16 March 1998; accepted 28 September 1998

Abstract Fibroblast growth factors (FGFs) stimulate proliferation, differentiation and motility of different cell types. The cellular effects of FGF are transduced by its interaction with any one of four members of a family of high affinity, cell surface FGF receptors (FGFRs) that have autophosphorylating tyrosine kinase activity. Activation of FGFR causes release of various low molecular weight signaling molecules which are required for the pleotropic effects of FGFs. We report here that basic FGF plays critical role in membrane phospholipid hydrolysis in NIH 3T3 cells that are stably transfected with FGFR1. Upon binding to FGFR1, basic FGF stimulates cytosolic form of phospholipase A2 (cPLA2), phospholipase C-γ1 (PLC-γ1) and phospholipase D (PLD), the key enzymes for the production of various lipid second messengers, in a tyrosine kinase-dependent manner. In addition to tyrosine phosphorylation, cPLA2 catalytic activation requires serine phosphorylation by p42 mitogen-activated protein (MAP) kinase and possibly pertussis toxin-sensitive G-protein coupling. On the other hand, phosphatidyl inositol 4,5 bisphosphate (PIP2) hydrolysis requires direct phosphorylation at tyrosine residue of the PLC-γ1 isozyme. The activation of PLD needs direct or indirect receptor tyrosine kinase and protein kinase C (PKC) activities. Additionally, it also requires botulinum toxin Csensitive Rho-like G-protein activation. All these results suggest that the pleotropic effects of FGF are exerted through its tyrosine kinase receptors and individual effectors are activated via distinguishable signaling mechanisms according to the cell’s need. (Mol Cell Biochem 198: 19–30, 1999) Key words: fibroblast growth factor receptor, cytosolic phospholipase A2, phospholipase C-γ1, phospholipase D, signal transduction Abbreviations: FGF – fibroblast growth factor; PLA2 – phospholipase A2; PLC – phospholipase C; IP3 – inositol 1,4,5 trisphosphate; PKC – protein kinase C; MAP kinase – mitogen-activated protein kinase; MBP – myelin basic protein; PMA – phorbol 12-myristate 13-acetate; PtdCho – phosphatidylcholine; PtdIns – phosphatidylinositol; PtdEtOH – phosphatidylethanol; PtdOH – phosphatidic acid; PIP2 – phosphatidylinositol 4,5-bisphosphate; DMSO – dimethyl sulfoxide; PMSF – phenylmethyl sulfonyl floride.

Introduction Basic fibroblast growth factor (FGF) is a potent mitogenic and chemotactic factor for a variety of cell types [1]. Its activity has been implicated in multiple physiological and patho-

physiological processes including differentiation, wound healing and tumor angiogenesis [2]. The pleotropic effects of FGF are transduced by its interaction with any one of four members of a family of high-affinity, cell surface FGF receptors. These receptors are ‘single pass’ transmembrane

Address for offprints: G. Sa, Animal Physiology Section, Bose Institute, P-1/12 CIT Scheme VII M, Calcutta 700 054, India

20 proteins with a kinase activity that induces phosphorylation of tyrosine residues in the receptor itself, as well as other intracellular substrate proteins [3–5]. Basic FGF also triggers a series of down-stream events including activation of p21ras [6], ras-like protein Rho [7], mitogen-activated protein (MAP) kinase [8, 9] and expression of early responsive genes [10]. In addition, basic FGF rapidly induces the release of a number of low molecular weight signaling molecules including arachidonate, inositol 1,4,5-trisphosphate (IP3), diacylglycerol (DAG), phosphatidic acid (PtdOH) [11–13], etc. But the sequential mechanisms by which this pleotropic agent induces the release of these important biomolecules are not well established. Eicosanoids derived from arachidonate elicit multiple physiological and pathophysiological responses. We have previously demonstrated that arachidonate release is necessary for FGF-stimulated endothelial cell motility [11], and that this important biomolecule is derived as a result of activation of cytosolic phospholipase A2 (cPLA2) [9]. Various studies have shown that cPLA2 is activated by phosphorylation by MAP kinase [9, 14]. The phosphorylated enzyme is translocated from cytosol to membrane in a process utilizing Ca 2+ dependent phospholipid-binding domain [15]. As with other growth factors, such as epidermal growth factor (EGF), platelet-derived growth factor (PDGF), etc., the binding of FGF to its receptors leads to phospholipase C-γ1 (PLC-γ1) activation by tyrosine phosphorylation [3]. Activated PLC-γ1 stimulates PIP2 hydrolysis, forming two second messengers, IP3, which releases sequestered calcium from a subpopulation of the endoplasmic reticulum, and diacylglycerol (DAG). DAG and Ca2+ in turn activate protein kinase C (PKC) [16]. These two signaling molecules participate in the transduction of different mitogenic signals across the plasma membrane and regulate various cellular events. Mohammadi et al. [17] and Peters et al. [18] independently demonstrated that point mutation in the FGF receptor selectively eliminates activation of PLC-γ1 and that phosphatidylinositol (PtdIns) hydrolysis is not essentially required for FGF-induced mitogenesis. Phosphatidylcholine (PtdCho) constitutes the largest fraction of total plasma membrane phospholipids and was previously thought to be metabolically rather stable compared to PtdIns. However, PtdCho has been shown to be hydrolyzed in response to various extracellular signal molecules upon receptor-mediated activation of phospholipase A2, C and D [19]. Recent evidences show that many agonists, including FGF, stimulate DAG production through the hydrolysis of PtdCho and that PLD is the major enzyme involved [13]. Phospholipase D initially produces PtdOH, which may have second messenger role, but can also be rapidly converted by PtdOHphosphohydrolase to DAG [20]. Most agonists induce a biphasic production of DAG, with PtdIns-PLC being responsible for the initial rapid increase and PtdCho-hydrolyzing

PLD for the second sustained increase. Phospholipase Dcatalyzed PtdOH formation has been shown to inhibit G2/M phase transition of a number of cultured cells in a reversible manner as a part of their biomodal function in growth control [21]. Although agonist-induced activation of PLD is a widely occurring phenomenon, the regulatory mechanisms involved are not well defined. Evidence for the involvement of PKC in the regulation of PLD has been obtained in many different cell systems with various agonists. Besides PKC, receptor tyrosine kinases themselves, Ca2+, and small GTP-binding protein (Gprotein) rho has been shown to be involved in the regulation of cellular PLD activity. However, evidence for the sequential mechanisms is limited, and there have been several reports presenting contradictory findings [13]. We designed the present study to identify possible mechanisms by which basic FGF activates PLA2, PLC-γ1 and PLD in NIH 3T3 cells stably expressing FGFR1. It is demonstrated that basal phospholipases in these cells are activated by various different mechanisms apparently involving tyrosine phosphorylation, PKC, MAP kinase and G-proteins. Evidence is provided that hydrolysis of phospholipids involved receptor tyrosine kinase as an initial step.

Materials and methods Myelin basic protein (MBP), phorbol 12-myristate 13-acetate (PMA), 5'-methylthioadenosine (MTA) and general reagents for kinase assay were purchased from Sigma, USA. PKC assay kit, herbimycin A and fetal bovine serum (FBS), Dulbeco’s modified Eagle’s (DME) medium etc., were procured from Gibco BRL, USA. [γ-32P]ATP (specific activity 3000 Ci/mmole) was purchased from BRIT, India. Phosphatidyl [2-3H]inositol 4,5-bisphosphate ([3H]PIP2) (specific activity 6 Ci/mmole), [9,10(n)-3H]oleic acid (10 Ci/mmol) and L-α-1-palmitoyl-2-[14C]arachidonyl phosphatidylcholine (specific activity 53 mCi/mmol) and [125I]protein A were purchased from Dupont NEN, USA. Pertussis toxin, cholera toxin and H7 were procured from Biomol inc. USA. Botulinum toxin C3 was obtained from Calbiochem, USA. Monoclonal anti-PLC-γ1 antibody and anti-phosphotyrosine antibodies (both monoclonal and polyclonal) were obtained from SantaCruz, USA. Rabbit polyclonal anti-cPLA2 was obtained as described previously [9]. Polyclonal antibody against p42 MAP kinase (TR10) [8] was kindly supplied by Dr. MJ Weber, University of Virginia, USA and sucrose monolaurate was a gift from Dr. M-J Im, Cleveland Clinic Foundation, USA. Cell culture NIH 3T3 cells overexpressing transfected human FGFR1 or control vector (obtained from Dr. M Jaye of Ronae Polancae

21 Röarer, USA) were maintained in DME medium containing 10% FBS [3]. Cells were grown to 70–80% confluency and transferred to serum free medium containing 1 mg/ml gelatin at least 24 h before use. For studying the effect of different inhibitors/toxins, the cells were pretreated with appropriate antagonists for 90 min prior to the addition of basic FGF. For the inclusion of botulinum toxin C3 which does not readily penetrate cell membrane, cells were transiently permeabilized with saponine [11]. The monolayers were incubated at 37°C for 10 min in buffer (15 mM Hepes, 135 mM KCl, 15 mM NaCl, 1.5 mM MgCl2, 1 mM EGTA and 0.2 mM CaCl2, pH 7.0) containing 18 µg/ml saponine. In multiple repetitions, this procedure permeabilized >90% cells as determined by trypan blue inclusion. Membrane integrity was restored within 2 h and permeabilization by saponine did not alter the enzyme activities since those of treated and untreated cells were more or less similar. In case of PMA pretreatment, it was dissolved in DMSO and incubated with the cells for 16 h prior to agonist treatment. Control cells were exposed to DMSO as a carrier vehicle. All incubations were terminated by removing the incubation medium and washing the cells with Hank’s balanced salt solution (4°C). Cells were then lysed by brief sonication (50 W, 15 sec) in required buffer and used for further studies. Determination of in vitro phospholipase A2 activity Phospholipase A2 activity in lysates of control and FGFR1 transfected NIH 3T3 cells was determined by an in vitro assay by a modification of previous method [9]. In brief, L-α-1palmitoyl-2-[14C]arachidonyl PtdCho was dried under N2 and resuspended in DMSO by vigorous mixing, and dispersed in a bath sonicator. The substrate (5 µl containing 2 × 105 cpm) was incubated with cell lysates (2 × 106 cells) at 37°C for 30 min in a total volume of 0.2 ml of 25 mM Hepes (pH 7.5) containing 200 mM sucrose, protease inhibitor (10 µg/ml each of benzamidine, trypsin inhibitor, bacitracin, 5 µg/ml each of leupeptine, pepstatin A, antipain and PMSF), phosphatase inhibitor (5 µM each of o-phosphoserine, o-phosphotyrosine, o-phosphothreonine, β-glycerophosphate, p-nitrophenyl phosphate and sodium vanadate) cocktails and 0.05% sucrose monolaurate in the presence of 2 µM Ca2+ and 5 mM dithiotheritol. The reaction was stopped with 0.1 ml ice-cold ethanol containing 2% (v/v) glacial acetic acid and 100 µg/ml of unlabelled arachidonic acid. Heptane (1.5 ml) and water (1.0 ml) were added and mixed. The organic phase was dried and dissolved in chloroform, and radioactivity in the arachidonic acid band was determined after thin layer chromatography. In vitro phospholipase C assay Phospholipase C activity in control and FGFR1-transfected NIH 3T3 cell lysates (2 × 106 cells) or anti-PLC-γ1 immuno-

precipitates from untreated or basic FGF-treated cytosolic and membrane fractions of same number of cells were determined by an in vitro assay by some modification of the method [22]. In brief, [2-3H]PIP2 (200–250 cpm/pmol) was dried under N2 and resuspended in assay buffer containing 20 mM Hepes (pH 7.2), 0.1% sodium deoxycholate, 300 µM CaCl2, 100 µM EGTA and 100 mM NaCl. The reaction was started by adding substrate (5 µM) and incubating at 30°C for 10 min in a total volume of 0.2 ml assay buffer containing protease and phosphatase inhibitor cocktails. The reaction was stopped by adding 0.2 ml of 1 N HCl followed by 0.75 ml of chloroform: methanol: HCl (100:100:0.6). The mixture was then vortexed vigorously and centrifuged at 3000 rpm for 10 min. An aliquot (250 µl) of the upper phase was removed and counted for radioactivity using a β counter (Wallac).

Assay of phospholipase D For the determination of intact cell phospholipase D activity, cellular phospholipids were metabolically labeled by incubating nearly confluent monolayers of control and FGFR1 transfected NIH 3T3 cells (2 × 10 6 cells) with [3H]oleic acid (2 µCi/ml) for 20–24 h, the labeling medium was replaced and the cells were equilibrated twice for 30 min at 37°C in serum-depleted medium. Cells were then either pretreated with inhibitors or toxins for 90 min before the addition of 10 ng/ml basic FGF in the presence or absence of 400 mM ethanol. The reaction was performed at 37°C and was stopped by adding 1 ml ice-cold methanol [23]. Cells were scraped off the dishes and the phospholipids were extracted as described previously. For the phosphatidylethanol (PtdEtOH) analysis, the lipid extracts were separated by TLC using organic phase of a mixture of ethylacetate:2,2,4trimethyl pentane:acetic acid:water (13:2:3:10; v/v) as the mobile phase. Radioactivity in the PtdEtOH band was measured by liquid scintillation spectroscopy.

Western blot detection of phosphotyrosyl proteins Confluent and quiescent control NIH 3T3 cells and cells (2 × 106 cells) over-expressing FGFR1 were treated with basic FGF (10 ng/ml) for 10 min at 37oC. The medium was aspirated, washed, and the cells were solubilized in 1% NP40, 0.5% SDS containing phosphate-buffered saline (PBS) and phosphatase inhibitors. The lysates were cleared by centrifugation (14,000 × g, 5 min) and then with pre-immune sera. The phosphotyrosyl proteins were immunoprecipitated with purified monoclonal anti-phosphotyrosine antibody coupled with protein A-Sepharose. The immunoprecipitates were centrifuged at 12,000 rpm for 2 min at 4°C and the pellets

22 were washed with ice-cold PBS containing phosphatase inhibitors. The immunopurified proteins were treated with 2 × Laemmli sample buffer, boiled for 5 min and subjected to SDS-PAGE (8% gel). The samples were transferred to nitrocellulose paper at 2.5 mA/cm2 for 45–60 min. Blocking was done for 2 h at 45°C in 5% non-fat dry milk containing 1% FBS in PBS-Tween 20 (0.1%). The blot was incubated for 2 h with rabbit anti-phosphotyrosine antibody (1:1,000) and then with [125I]protein A (0.25 µCi/ml). The signal was detected by exposure of Kodak XAR film. The amount of cell extracts applied to gels were normalized by protein concentration measured by a BioRad DC assay using BSA as standard.

Determination of MAP kinase activity To immunoprecipitate p42 MAP kinase from the cytosols of untreated and basic FGF treated cells (2 × 106 cells in each case), the cytosols were first precleared with rabbit IgG and then incubated with rabbit polyclonal anti-p42 MAP kinase antibody for 4 h at 4°C. The samples were then treated with protein A-Sepharose beads (precoated with 3% non-fat dry milk for 1 h and washed by centrifugation). The immunoprecipitates were centrifuged at 12,000 rpm for 2 min at 4°C and the pellets were washed with ice-cold PBS containing phosphatase inhibitors. The pellets were then assayed for MAP kinase activity [9] using MBP (10 µM) as substrate. Control substrates were either dephosphorylated casein (2 µg) or histone H1 (2 µg). Incorporation of 32Pi from 5 µM [γ32 P]ATP (5 µCi) into the substrates by the immunoprecipitate was monitored. Assays were performed in a buffer containing 10 mM Tris-HCl, pH 7.4, 250 mM sucrose, 5 mM MgCl2, 0.5 mM EDTA in a total volume of 50 µl at 25°C for 10 min and reactions were terminated with 5% orthophosphoric acid and the reaction mixture was spotted onto phosphocellulose filter paper followed by extensive washing with 0.5% orthophosphoric acid. Radioactivity was determined by liquid scintillation spectroscopy.

In vitro protein kinase C (PKC) assay FGFR1 transfected cells (2 × 106 cells) were treated with basic FGF (10 ng/ml) or medium alone for 10 min at 37°C. The cells were then lysed by sonication in 20 mM Tris-HCl buffer, pH 7.4 containing 5 mM EDTA, protease and phosphatase inhibitor cocktails as mentioned above. The lysates were centrifuged at 900 × g for 10 min. The low speed supernatants were centrifuged at 40,000 × g for 1 h and the pellets were harvested and washed with the same buffer. The membranes were purified by percoll gradient centrifugation [9] and suspended at 10 mg protein/ml. To obtain the cytosolic fraction,

the supernatants were recentrifuged at 105,000 × g for 1 h, and the supernatants were adjusted to 10 mg protein/ml. PKC activity in both the cytosolic and membrane fractions was assayed using PKC assay kit (Gibco BRL) according to the supplier’s instructions. In brief, incorporation of 32Pi from [γ-32P]ATP into the specific PKC substrate peptide (QKRPSQRSKYL) by the cytosolic or membrane fractions was monitored. Assays were performed in a buffer containing 10 mM TrisHCl, pH 7.4, containing 100 mM NaCl, 5 mM MgCl2, 5 mM CaCl2, 1% Triton X100, 20 µM ATP, 5 µM [γ-32P]ATP, in the presence of protease and phosphatase inhibitor cocktails. Incubations were performed under linear assay condition at 30°C for 10 min. The reaction mixture was spotted onto phosphocellulose paper followed by extensive washing with 0.5% phosphoric acid. Radioactivity was determined by liquid scintillation spectroscopy.

Western blot analysis of cPLA2, p42 MAP kinase and ‘mobility shift’ assay of phosphorylation Semiconfluent and quiescent transfected cells (2 × 106 cells) after respective treatments were lysed by sonication in PBS containing protease and phosphatase inhibitors. The samples were treated with 2 × Laemmli sample buffer and separated on 10% SDS-PAGE for p42 MAP kinase and on 8% SDSPAGE for cPLA2, using acrylamide: bisacrylamide (60:1, w/ w and 20 mA/gel) as described earlier [9]. The samples were transferred to nitrocellulose paper. The native and phosphorylated proteins were detected with rabbit anti-p42 MAP kinase antibody (1:5,000) or rabbit anti-cPLA2 antibody (1:1,000). For the detection of phosphotyrosine moiety of cPLA2, the protein was immunoisolated from treated or untreated cell lysates using anti-cPLA2 antibody and protein A-Sepharose as described earlier and blotted with antiphosphotyrosine antibody. The second antibody used for the detection of signals was goat anti-rabbit antibody conjugated with horseradish peroxidase (HRP) (1:1,000). The signal was detected by color reaction using diaminobenzene (DAB) as substrate. Tyrosine phosphorylation of PLC-γ1 Semiconfluent and quiescent FGFR1 overexpressed cells (2 × 10 6 cells) after respective treatments were lysed and solubilized in 1% NP-40, 0.5% SDS containing PBS with protease and phosphatase inhibitors and subjected to immunoprecipitation with monoclonal anit-PLC-γ1 antibody for 4 h at 4°C. The immunocomplexes were harvested with precoated protein A-Sepharose. In separate experiments PLC-γ1 was immunoisolated from untreated and basic FGF-treated cytosolic and membrane fractions of

23 cells. Immunoisolated proteins were solubilized with gel soluble buffer and separated on 8% SDS-PAGE. For western blotting, proteins were transferred from the gel to the nitrocellulose paper and one part was blotted with monoclonal anti-PLC-γ1 antibody (1:500) and the other part was probed with monoclonal anti-phosphotyrosine antibody (1:1,000) and then with horseradish peroxidase conjugated goat anti-mouse antibody (1:1,000) for the identification of PLC-γ1 and phosphotyrosine moiety in PLC-γ1 respectively. The signals were detected by color reaction using DAB as substrate. Total amount of protein applied to gel was normalized by the total protein in the aliquot from which the PLC-γ1 was immunoprecipated rather than the actual amount of PLC-γ1 present. For the quantitation of the PLCγ1 and phosphotyrosine moiety in PLC-γ1, the bands were densitometrically scanned using Imaging Densitometer and Multi-Analyst software (BioRad).

Results Involvement of tyrosine kinase on basic FGF-stimulated PLA2, PLC and PLD activities Basic FGF has been shown to stimulate significantly the catalytic activities of PLA2 (p < 0.005), PLC (p < 0.001) and PLD (p < 0.002) in NIH 3T3 control cells (Fig. 1A) as well as cells overexpressing FGFR1 (Fig. 1B). The maximal activation of PLC is at 10 min of basic FGF stimulation, whereas PLA2 and PLD activities showed peak at 15 min of growth factor stimulation (data not shown). The activation of all the three enzymes were predominant in FGFR1 expressing cells than that of the control cells. In the cell lysates basic FGF stimulated arachidonic acid release from L-α-1-palmitoly-2-[1-14C]arachidonyl PtdCho. Since various signal transduction pathways share with basic FGF, the common feature of tyrosine kinase activity of the receptor. Accordingly, we examined whether tyrosine kinase inhibitors can block PLA2 activity or not. In these studies, we found that herbimycin A, a specific tyrosine kinase inhibitor, blocked basic FGF-stimulated PLA2 activity in control (Fig. 1A) as well as in FGFR1 transfected cells (Fig. 1B), indicating the involvement of tyrosine phosphorylation as an essential event. Treatment of cells with basic FGF yielded a detectable increase in the formation of inositol triphosphate (IP3) from phosphatidyl [2-3H] inositol 4,5-bisphosphate. Basic FGF increased PLC catalytic activity approximately 3.5 fold in FGFR1 overexpressing cells (Fig. 1B) which is higher than the control cells (approximately two-fold; Fig. 1A). It is well established that growth factors stimulate tyrosine phosphorylation of PLC-γ1 in intact cells [3]. To determine whether our observation of basic FGF-stimulated PLC

Fig. 1. Involvement of tyrosine kinase on basic FGF-stimulated PLA2, PLC and PLD activities. Semi-confluent and quiescent control (A) or FGFR1 transfected (B) cells were pretreated for 90 min with medium alone (open bars & solid bars) or with 100 nM herbimycin A (light striped bars & dark striped bars). Cells were then incubated with (solid bars & dark striped bars) or without (open bars & light striped bars) 10 ng/ml basic FGF as indicated in the text. Cellular PLA2, PLC and PLD activities were measured as described in ‘Materials and methods’. Data are expressed as the mean ± S.E.M. of triplicate cultures.

catalytic activation in intact cells requires tyrosine phosphorylation, we evaluated the capacity of tyrosine kinase inhibitors to influence the in vitro catalytic activity of PLC that hydrolyzes PIP2 in control or FGFR1 transfected NIH 3T3 cells. Herbimycin A had been shown to block this stimulation below the normal level in both the cell lines (Fig. 1A and B). Both the control and the FGFR1 transfected cells were labeled with [3H]oleic acid in order to measure PLD activity. [3H]PtdEtOH formed in the presence of ethanol is a specific product of phospholipase D-catalyzed transphosphatidylation reaction and is metabolically relatively stable. Stimulation of the cells with basic FGF caused an increase in PtdEtOH formation. Pretreatment of the cells with herbimycin A markedly reduced the PLD activity in both basic FGF-stimulated and unstimulated cells (Fig. 1A and B). Pretreatment of cells for 90 min MTA (500 µM), an inhibitor of specific

24 FGF receptor tyrosine kinase [24], completely blocked basic FGF-stimulated activities of all the three enzyme activities (data not shown), indicating that the early tyrosine kinase activity is due to receptor tyrosine kinase activation.

Phosphotyrosyl proteins induced by basic FGF Reports of FGF-induced tyrosine kinase activity in different cell lines prompted us to investigate the status of tyrosine phosphorylation of cellular proteins in FGFR1 transfected cells in response to basic FGF and the effect of tyrosine kinase inhibitor on that phosphorylation. To enrich the phosphoproteins, cell lysates were sequentially immunoprecipitated and immunoblotted with purified anti-phosphotyrosine antibodies. By these methods several phosphotyrosyl proteins were observed in quiescent cells (Fig. 2). Basic FGF-stimulated rapid phosphorylation of several proteins particularly pp150, pp130, pp90, pp66, pp45 etc. Herbimycin A specifically inhibits tyrosine phosphorylation of these proteins indicating the specificity of the inhibitor.

Fig. 2. Basic FGF-induced phosphotyrosyl pattern of cellular proteins. Semi-confluent and quiescent control (lanes 1 & 2) or FGFR1 transfected (lanes 3–7) cells, preincubated with herbimycin A or anti-basic FGF for 90 min, were treated with (+) or without (–) 10 ng/ml basic FGF for 10 min. Cells were immediately lysed, immunoprecipitated with monoclonal antiphosphotyrosine antibody, subjected to SDS-PAGE and immunoblotted with rabbit anti-phosphotyrosine antibody as outlined under ‘Materials and methods’.

Effect of different bacterial toxins on basic FGFstimulated PLA2, PLC and PLD activities The involvement of GTP-binding regulatory proteins in basic FGF-stimulated activation of different phospholipases were tested using various bacterial endotoxins which were known to influence both heterotrimeric G-proteins (Gi- and Gs-like G-proteins) as well as ras-related small molecular weight Gproteins (Rho, ARF etc.). Pertussis toxin, which inhibits Gilike G-proteins by ADP ribosylation, completely blocked basic FGF-stimulated PLA2 activity (Fig. 3A), whereas it can not affect either PLC or PLD activities (Fig. 3B and C). Cholera toxin, which inhibits GTPase activity of Gs-like Gproteins, had no effect in any of the phospholipases indicating that G s-like G-proteins are not involved in basic FGF-

Fig. 3. Effect of bacterial toxins on basic FGF-stimulated PLA2, PLC and PLD activities. Semi-confluent and quiescent FGFR1 transfected cells were treated for 90 min with medium alone or with pertussis toxin (100 ng/ml) or cholera toxin (250 ng/ml). In parallel experiments cells were permeabilized and incubated with botulinum toxin C3 (100 ng/ml). Cells were then incubated with (dark striped bars) or without (light striped bars) basic FGF as indicated in the text. (A) PLA2, (B) PLC, and (C) PLD activities were measured in the cell lysates as described in the ‘Materials and methods’. Data represent mean ± S.E.M. for typical experiment performed in triplicate and expressed as percentage of control.

25 mediated PLA2, PLC or PLD activation (Fig. 3A, B and C). In case of botulinum toxin C3, which blocks rho, a small GTP-binding protein, complete inhibition of basic FGFinduced PLD activation was observed (Fig. 3C). On the other hand it had no effect on either PLA2 or PLC activity (Fig. 3A and B). All these data suggest that pertussis toxin-sensitive G-protein may be involved in basic FGF-mediated PLA2 activation and botulinum toxin-sensitive rho-like G-protein may play a critical role in basic FGF-stimulated PLD activation. However, PLC catalytic activation does not require any toxin-sensitive G-protein activation.

Basic FGF stimulates phosphorylation of cPLA2 and p42 MAP kinase We showed previously that basic FGF stimulates cytosolic forms of PLA2 (cPLA2) in bovine aortic endothelial cells [9]. Activation of cPLA2 by growth factors is thought to require phosphorylation reactions involving MAP kinase and/or tyrosine kinase [9, 13, 25]. The decrease in electrophoretic mobility or ‘mobility shift’ of the phosphorylated protein during SDS-PAGE [9] was used to investigate phosphorylation of cPLA2 by basic FGF. Lysates of FGFR1 overexpressed NIH 3T3 cells, pretreated with basic FGF, were prepared for immunoblot analysis using anti-cPLA2 antibody. Results indicated that cell lysates contained cPLA2 and that most of the enzyme was unphosphorylated in unstimulated cells (Fig. 4A). Within 15 min of agonist stimulation more than 95% of the cPLA2 became phosphorylated. Although it has been established clearly that phosphorylation of cPLA2 on serine residue by p42 MAP kinase increases its activity [14], exposure of keratinocytes to transforming growth factor (TGF)-α [26] or interferon (IFN)-α [25] results in tyrosine phosphorylation of cPLA2. Since tyrosine phosphorylations of intracellular substrate proteins are common feature of growth factor action [27], an obvious question to us was whether cPLA2 is phosphorylated on tyrosine in addition to serine residue, following exposure of cells to basic FGF. Results of Fig. 4B indicate that, although a tyrosine phosphorylated component was present in control cell, treatment with basic FGF induced an increase in phosphorylation of cPLA2 on tyrosine residue. Activation of cPLA2 requires not only phosphorylation but also Ca2+-dependent translocation of the enzyme from cytosol to membrane [15]. To specifically demonstrate cPLA2 translocation, membrane and cytosolic fractions from basic FGFtreated cells were seperated and subjected to immunoblotting with anti-cPLA2 antibody. Densitometric scanning of cPLA2 bands suggests that most of the enzyme remained in cytosolic fraction in unstimulated cells. Basic FGF-treatment stimulated translocation of the enzyme to the membrane fraction (Fig. 4C). To determine intracellular location of the cPLA2 activity,

Fig. 4. Basic FGF-stimulated phosphorylation and activation of cPLA2. (A) FGFR1 transfected (2 × 106) cells were incubated with or without 10 ng/ml of basic FGF at 37oC for 15 min. Cell lysates were subjected to SDS-PAGE and immunoblotted with rabbit anti-cPLA2. The phosphorylated (cPLA2-P) and native cPLA2 isoforms are separated due to reduction of electrophoretic mobility. (B) Cells after basic FGF treatment were lysed, cPLA2 were immunopurified with anti-cPLA2 antibody, processed for SDSPAGE and Western blotted with anti-phosphotyrosine antibody as described in the ‘Materials and methods’. (C) Cytosolic and membrane fractions prepared from FGFR1 transfected (2 × 106) cells pretreated with or without basic FGF. The fractions were subjected to SDS-PAGE and immunoblotted with rabbit anti-cPLA2. The cPLA2 bands were scanned and the densitometric values were plotted as relative partition of the enzyme content (light solid bars) in cytosolic and membrane fractions. cPLA2 enzyme activities were assayed in vitro in each fractions and the values were represented as relative enzyme activity (dark solid bars) in different fractions.

in vitro PLA2 activity was measured into membrane and cytosolic fractions. PLA 2 activity in unstimulated cell membrane was low, but stimulated by basic FGF by about 3.5 fold (Fig. 4C). These results also suggest that though in unstimulated cells the enzyme mainly remained in cytosolic fractions as an inactive form. Upon basic FGF stimulation the enzyme translocated into membrane fraction in active form. Since we and others [9, 14] had already established the necessity of MAP kinase activation for the induction of cPLA2 activity by growth factors, we determined whether

26 MAP kinase is also activated in response to basic FGF treatment in our assay system, by using p42 MAP kinase ‘mobility shift’ assay. When cell cytosols were immunoblotted with anti-p42 MAP kinase antibody, basic FGF stimulated phosphorylation of p42 MAP kinase was observed, whereby, reduction of electrophoretic mobility in SDS-PAGE was obtained (Fig. 5A). This result was further confirmed by in vitro myelin basic protein kinase assay using p42 MAP

kinase immunoprecipitate (Fig. 5B). To test the requirement of p42 MAP kinase for the activation of cPLA2, cell lysates were depleted of p42 MAP kinase by repeated immunoprecipitation and PLA2 activity was measured as release of [1-14C]arachidonate from its PtdCho precursor. Basic FGFstimulated PLA2 activity was essentially blocked by this treatment (Fig. 5C), demonstrating the critical regulatory role of p42 MAP kinase. All these results indicate that basic FGF stimulates cPLA2 activation by tyrosine phosphorylation as well as serine phosphorylation via p42 MAP kinase.

Basic FGF stimulated tyrosine phosphorylation and translocation of PLC-γ1 We have already showed that lysates from basic FGFstimulated cells had elevated PLC activity and specific tyrosine kinase inhibitors blocked the activation. And it is well established that catalytic activation of PLC-γ1 isoform is regulated by tyrosine kinases. So our next aim was to find out whether basic FGF can tyrosine phosphorylate PLC-γ1 in FGFR1 overexpressed cells. When basic FGF-treated cell lysate was subjected to immunoprecipitation with anti-PLCγ1 antibody and immunoblotted with anti-phosphotyrosine antibody, it was observed that in deed PLC-γ1 is tyrosine phosphorylated as a result of basic FGF stimulation (Fig. 6A and B). In untreated cells most of the PLC-γ1 resides in the cytosol in unphosphorylated forms. Basic FGF treatment caused tyrosine phosphorylation, association of the enzyme to the membrane and catalytic activation of the enzyme (Fig. 6C). This result is in conformity with the findings that PDGF and EGF also tyrosine phosphorylate as well as activate PLCγ1 isoform in different cell lines [3, 17, 18].

Fig. 5. Role of p42 MAP kinase-mediated phosphorylation of cPLA2 activation. NIH 3T3 cells transfected with FGFR1 were incubate with or without 10 ng/ml basic FGF at 37°C. (A) Cells were lysed and subjected to Western blot analysis using anti-p42 MAP kinase antibody. The phosphorylated (MAPK-P) and native (MAPK) p42 MAP kinase isoforms are separated by virtue of the ‘mobility shift’ due to phosphorylation. (B) p42 MAP kinase was immunoisolated from the soluble supernatant with rabbit anti-p42 MAP kinase antiserum. MAP kinase activities in the immunoisolated proteins were measures as phosphorylation of myelin basic protein. The results are shown as fold stimulation as compared to unstimulated cells. (C) Cytosolic and membrane fractions were prepared from FGFR1 transfected (2 × 106) cells. p42 MAP kinase was immunodepleted from the cytosolic fraction with anti-p42 MAP kinase IgG (or control IgG). Immunodepleted cytosol was combined with the membrane fraction in the presence (dark solid bars) or absence (light solid bars) of basic FGF. PLA2 activity of the combined fractions was determined by hydrolysis of L-α-1palmitoyl-2-[14C]arachidonyl PtdCho. The combined fractions were also incubated with calf intestinal alkaline phosphatase (20 µg/ml) for 30 min before measurement of cPLA2 activity. Data represent mean ± S.E.M. for typical experiment performed in triplicate.

Involvement of PKC in the regulation of phospholipase D activity Several observations suggest that in addition to the involvement of botulinum toxin-sensitive rho like G-protein, protein kinase C activation also plays crucial role in PLD catalytic activation. To investigate the role of PKC in basic FGF-stimulated PLD activation, we applied two different approaches. In one approach involvement of PKC in basic FGF-induced signaling pathway was studied by monitoring direct translocation and activation of PKC from cytosol to membrane as a result of agonist stimulation. In control cells, PKC resided mainly in the cytosol and less in membrane. Basic FGF stimulated the translocation of PKC from cytosol to membrane (p < 0.001). As shown in the Fig. 7A, basic FGF could increase the membrane activity by more than four fold in comparison to control. Activity of this PKC was completely blocked by the PKC peptide inhibitor when added in the assay

27

Fig. 6. Role of basic FGF on tyrosine phosphorylation of PLC-γ1. FGFR1 transfected cells were either unstimulated or stimulated with 10 ng/ml basic FGF for 10 min at 37°C. Cells were lysed and immunoprecipitated with anti-PLC-γ1 antibody as described in ‘Materials and methods’. (A) Half of the immunoprecipitate was resolved by SDS-PAGE and immunoblotted with anti-PLC-γ1 antibody. (B) The other half was blotted with antiphosphotyrosine antibodies. (C) Cytosolic and membrane fractions were prepared from FGFR1 transfected cells pretreated with or without basic FGF. PLC-γ1 was immunoisolated from each fraction and half of the immunoisolated protein was immunoblotted with anti-PLC-γ1 antibody. The other half was blotted with anti-phosphotyrosine antibody. PLC-γ1 bands (open bars) as well as the phosphotyrosine bands (light solid bars) of PLC-γ1 were scanned and densitometric values were plotted as relative partition of the enzyme content in cytosolic and membrane fractions. PLCγ1 enzyme activity was assayed in vitro in each fraction and the values were represented as relative enzyme activity (dark solid bars) in different fractions. Data represent mean ± S.E.M. for typical experiment performed in triplicate.

system. In the second approach, we investigated the role of PKC in basic FGF-stimulated PLD by down regulating the enzyme by prolong exposure of the cells to PMA. It has been observed that treatment of cells with 500 nM PMA for more than 16 h almost completely down regulate α, β and ε isoforms of PKC. We tested the effect of PKC down regulation on basic FGF-stimulated PtdEtOH formation. As shown in the Fig. 7B, the agonist treatment itself significantly (p < 0.002) enhanced the radioactivity in PtdEtOH suggesting that

Fig. 7. Involvement of basic FGF-induced PKC activation in phospholipase D stimulation. (A) FGFR1 transfected cells were treated with or without 10 ng/ml basic FGF for 15 min. Cytosolic (light striped bars) and membrane (dark striped bars) fractions were isolated by ultracentrifugation and Percoll gradient centrifugation, respectively. Protein kinase C activities in individual fractions were assayed as described in ‘Materials and methods’. (B) Cells were preincubated for 16 h with 500 nM PMA/ DMSO or for 90 min with 2.5 µM H7 and then with medium alone (light striped bars) or 10 ng/ ml basic FGF (dark striped bars) for 15 min. PLD activities were measured as described under ‘Materials and methods’. Data are mean ± S.E.M. of triplicate experiments.

the activity of PLD was increased. However, the ability of basic FGF to promote PtdEtOH accumulation was almost completely abolished in the PKC-down regulated cells. These results were further tested using PKC inhibitor, H7. This specific PKC inhibitor could block basic FGF-stimulated PLD activation at its IC 50 concentration. These results strongly support the hypothesis that in addition to the involvement of tyrosine kinase and rho-like G-protein, PKC is critically involved in PLD activation by basic FGF.

Discussion Growth factors stimulate a variety of cellular responses, including the generation of lipid second messengers. Some of these responses are rapid and transient, while others are

28 slower and sustained. The results presented here have demonstrated that arachidonate, IP3, DAG and PtdOH are generated as a result of cPLA2, PLC-γ1 and PLD activation in distinguishable signaling mechanisms in FGFR1 overexpressed cells treated with basic FGF and that, in common, receptor tyrosine kinase activation plays a key role in the enzymatic activation of these three phospholipases. Agonist-induced release of arachidonic acid by PLA2 and its subsequent conversion of bioactive eicosanoids, regulate multiple normal processes. In this report we provided evidences that basic FGF stimulates the activity of the cytosolic form of PLA2. It is clear that cPLA2 enzyme activity is regulated by serine phosphorylation by p42 MAP kinase and translocation to the membrane by submicromolar concentration of Ca2+. Several other mechanisms have been postulated to be involved in cPLA2 activation. These include activation of PKC [14, 28], activation of tyrosine kinase [25, 26] and coupling through G-protein [9, 29, 30] etc. The observation that FGF receptor stimulation is sufficient to activate cPLA2 and to cause intracellular release of arachidonic acid even in the absence of auxiliary factors is consistent with the pluripotent activity of the receptor [2]. Our data and that from the other laboratories [9, 14, 30] also indicate the requirement of p42 MAP kinase-mediated serine phosphorylation for the activation of cPLA2. This observation is also consistent with known activation of MAP kinase by basic FGF [8, 9]. However, activation of MAP kinase alone may not be sufficient to fully activate cPLA2, since granulocyte/ macrophage colony-stimulating factor activates MAP kinase, but is rather a weak activator of cPLA2 [30]. An additional activity is required for full activation of cPLA2. Tyrosine kinase may be the candidate which activates cPLA2 by MAP kinase-dependent and -independent mechanisms [9, 14, 25]. Activation of this kinase by basic FGF and its role in various cellular functions has been well documented [2]. Our data also indicate that basic FGF induces phosphorylation at tyrosine residue of cPLA2. The role of basic FGF on calciummediated translocation of cPLA2 has also been clearly shown [15]. In our system IP3-mediated mobilization of intracellular Ca 2+ may result from activation of PLC-γ1 by the FGF receptor indicating the cross-talk between PLC-γ1 and PLA2 pathways. This result may also explain why cPLA2 enzymatic activation was slightly delayed (peak at 15 min after basic FGF stimulation) when compared with that of PLC-γ1 (peak at 10 min, data not shown). In addition, there is evidence for a role of G-protein in activation of PLA2 [29], and most likely cPLA2 [9, 29, 31] (however, G-protein-independent activation of cPLA2 has also been reported [32]). We have shown that basic FGF-dependent activation of cPLA2 in FGFR1 overexpressed cell is completely blocked by pertussis toxin, indicating the involvement of pertussis toxin-sensitive G-protein in this mechanism. All these observations led to the conclusion that, the activation of multiple signal transduction

pathways by basic FGF receptor is most likely responsible for its sufficiency to stimulate functional cPLA2 activity. Several important components of intracellular signaling pathway as activated by receptor tyrosine kinases have been identified and their functions are being gradually unveiled. It is already known that a tyrosine autophosphorylation site of FGFR1 at residue 766 was demonstrated to bind the SH2 domain of PLC-γ1, causing membrane association of PLCγ1 [17, 18], which generally resides in cytosol in inactive form. Our studies showed that basic FGF treatment causes tyrosine phosphorylation, association of PLC-γ1 to the membrane and catalytic activation of the enzyme. It is likely that multiple pathways exist to transduce signals from receptor tyrosine kinase to the nucleus. One of the pathways may be through PLC-γ1 activation that ultimately leads to PKC activation via DAG production and Ca2+ mobilization through IP3. Increased in intracellular Ca2+ and activated PKC control variety of intracellular messages including activation of Raf-1, MAP kinase, cPLA2, PLD etc. So catalytic activation of PLC-γ1 by basic FGF plays key role in different cellular events. It is well accepted that PtdCho hydrolysis by PLD is a widespread cellular response to many agonists [13, 23]. Through this pathway, many potential signaling molecules such as PtdOH, lyso-PtdOH, DAG, arachidonate and its metabolites can be produced. The extracellular ligands EGF and PMA, each working via distinct physiological signal pathway, have been shown to inhibit the G2/M transition of a number of cultured cells in a reversible manner [21]. The restriction in G2 phase by both ligands occurs so rapidly that a search for inhibitory mediator(s) was promising approach and has already been started. Phosphatidic acid turned out to be a possible candidate for such inhibitory mediator [21]. In recent years increasing evidences have been accumulated that PtdOH may have biological functions of its own in a number of ligand-induced processes in mammalian cells, particularly in growth control [33]. Most G-protein-coupled receptors and receptor tyrosine kinase-coupled agonists have been shown to stimulate PtdCho hydrolysis by PLD by a process that appears to be down stream of the initial IP3 and DAG formation by PtdIns-PLC and to involve PKC activation [13, 20, 23, 26]. However, their detailed mechanisms are not fully understood. Because of our evidence that basic FGF is coupled to PtdIns-PLC signaling pathway in our system, we hypothesized that the activation of PLD by basic FGF would be PKCdependent. Our results show that the effect of basic FGF on the hydrolysis of PtdCho appears to involve PKC. However, it should be stressed that this assumption is based solely on the inhibition of the basic FGF effect by prolonged treatment of cell with PMA. In addition to down-regulation of PKC [13], prolonged treatment of cells with PMA may also cause changes both in the level of receptors and in the ability of receptors to bind basic FGF. Accordingly, further experiments were designed to test the involvement of PKC in basic FGF-

29 induced PLD activation. The specific PKC inhibitor H7 was unable to completely abolish the stimulation of PLD by basic FGF. Although this could merely be due to an inability of the inhibitor to render full inhibition of PKC isozyme at non-toxic dose, it is possible that PKC-independent mechanisms of PLD activation also exists. However, the possibility of a cross-talk between PLC-γ1-PKC and PLD pathways can be suggested. Recently cytosolic factors including ARF and rho proteins are identified as regulatory factors for PLD activity [34, 35]. In this study we also examined the involvement of rho protein in regulation of PLD activity. To this end, we found that treatment of cells with botulinum toxin C3, which inactivate rho proteins (by ADP-ribosylation by C3 transferase), potentially and efficiently blocks basic FGF-stimulated PLD activity. However, other toxin-sensitive G-protein inactivators can not significantly alter basic FGF-induced PLD activation, indicating an essential role of rho protein in this receptor action. All these results together suggest the presence of distinct regulatory mechanisms and /or involvement of different PLD isoforms. In conclusion, we demonstrate that the pleotropic effects of basic FGF are transduced through its receptors by diverse types of signaling. One of the signaling is the hydrolysis of membrane phospholipids. Basic FGF stimulates catalytic activation of cPLA2, PLC-γ1 and PLD through individual and distinguishable signaling mechanisms. These lipid second messengers control diverse types of cellular functions either directly or by modulating different signaling proteins or enzymes and hence transcriptionally activate various gene products. Cross-interaction of individual pathways also plays critical role in transducing a complex signal. Since signals are affecting almost all biochemical processes inside cells, there are, not one but many pathways for signal transduction, with various regulatory or checking points. And to avoid ‘traffic jams’, there are many alternatives or by-pass pathways also.

Acknowledgements We gratefully acknowledge Dr. MJ Weber for providing antip42 MAP kinase antibody, Dr. M Jaye for FGFR1 overexpressing cells and Dr. M-J Im for generous gift of sucrose monolaurate. Distributed Information Center of the Institute is gratefully acknowledged for general assistance. This work was supported partly by Department of Science and Technology, and Council of Scientific and Industrial Research, Government of India.

References 1.

Burgess WH, Maciag T: Heparin binding (fibroblast) growth factor family of proteins. Annu Rev Biochem 58: 575–606, 1989

2. Jaye M, Schlessinger J, Dionne CA: Fibroblast growth factor receptor tyrosine kinase: Molecular analysis and signal transduction. Biochim Biophys Acta 1135: 185–199, 1992 3. Mohammadi M, Honegger AM, Rotin D, Fischer R, Bellot F, Li W, Dionne CA, Jaye M, Rubinstein M, Schlessinger J: A tyrosinephosphorylated carboxyterminal peptide of the fibroblast growth factor receptor (flg) is a binding site for SH2 domain of Phospholipase C-γ1. Mol Cell Biol 11: 5068–5078, 1991 4. Zhang X, Hu X, Hampton B, Burgess WH, Friesel R, Maciag T: Murine cortactine is phosphorylated in response to fibroblast growth factor-1 on tyrosine residues late in the G1 phase of the BALB/c cell cycle. J Biol Chem 268: 24427–24431, 1993 5. Spivak-Kroizman T, Mohammadi M, Hu P, Jaye M, Schlessinger J, Lax I: Point mutation in the fibroblast growth factor receptor eliminate phosphatidylinositol hydrolysis without affecting neuronal differentiation of PC12 cells. J Biol Chem 269: 14419–14423, 1994 6. Kremer NE, D’Arcangelo G, Thomas SM, DeMarco M, Bugge JS, Halegousa S: Signal transduction by nerve growth factor and fibroblast growth factor in PC 12 cells requires a sequence of Src and Ras action. J Cell Biol 115: 809–819, 1991 7. Fox PL, Sa G, Dobrowolski SF, Stacey DW: The regulation of endothelial cell motility by p21ras. Oncogene 9: 3519–3526, 1994 8. L’Allemain G, Pouyssegur J, Weber MJ: p42/mitogen-activated protein kinase as a converging target for different growth factor signaling pathways: Use of pertussis toxin as a discrimination factor. Cell Reg 2: 675–684, 1991 9. Sa G, Murugesan G, Jaye M, Ivashchenko Y, Fox PL: Activation of cytosolic phospholipase A2 by basic fibroblast growth factor via a p42 mitogen-activated protein kinase-dependent phosphorylation pathway in endothelial cells. J Biol Chem 270: 2360–2366, 1995 10. Boutherin-Falson O, Reuse S, Dunont JE, Boeynaems JM: Increased levels of c-fos and c-myc mRNA in ATP-stimulated endothelial cells. Biochem Biophys Res Commun 172: 306–312, 1990 11. Sa G, Fox PL: Basic fibroblast growth factor-stimulated endothelial cell movement is mediated by a pertussis toxin-sensitive pathway regulating phospholipase A2. J Biol Chem 269: 3219–3225, 1994 12. Wahl MI, Jones GA, Nishibe S, Rhee SG, Carpenter G: Growth factor stimulation of phospholipase C-γ1 activity: Comparative properties of control and activated enzymes. J Biol Chem 267: 10447–10456, 1992 13. Yeo EJ, Exton JH: Stimulation of phospholipase D by epidermal growth factor requires protein kinase C activation in swiss 3T3 cells. J Biol Chem 270: 3980–3988, 1995 14. Lin L-L, Wartmann M, Lin AY, Knoff JL, Seth A, Davis RJ: cPLA2 is phosphorylated and activated by MAP kinase. Cell 72: 269–278, 1993 15. Clark JD, Milona N, Knoff JL: Purification of a 110-kilodalton cytosolic phospholipase A2 from human monocytic cell line U937. Proc Natl Acad Sci USA 87: 7708–7712, 1990 16. Hug H, Sarre TF: Protein kinase C isoenzymes: Divergence in signal transduction? Biochem J 291: 329–343, 1993 17. Mohammadi M, Dionne CA, Li W, Li N, Spivak T, Honegger AM, Jaye M, Schlessinger J: Point mutation in FGF receptor elements phosphatidylinositol hydrolysis without affecting mitogenesis. Nature 358: 681–684, 1992 18. Peters KG, Marie J, Wilson E, Ives HE, Escobedo J, Del-Rosario M, Mirda D, Williams LT: Point mutation of an FGF receptor abolishes phosphatidylinositol turnover and Ca2+ flux but not mitogenesis. Nature 358: 678–681, 1992 19. Billah MH, Anthes JC: The regulation and cellular function of phosphatidylcholine hydrolysis. Biochem J 269: 281–291, 1990 20. Exton JH: Phosphatidylcholine breakdown and signal transduction. Biochim Biophys Acta 1212: 26–42, 1994

30 21. Kaszkin M, Richards J, Kinzel V: Phosphatidic acid mobilized by phospholipase D is involved in the phorbol 12-myristate 13-acetateinduced G2 delay of A431 cells. Biochem J 314: 129–138, 1996 22. Das T, Baek KJ, Gray CD, Im M-J: Evidence that Gh protein is a signal mediator from α1-adrenoceptor to a phospholipase C: II. Purification and characterization of a Gh-coupled 69 kDa phospholipase C and reconstitution of α1-adrenoceptor, Gh family and phospholipase C. J Biol Chem 268: 27398–27405, 1993 23. Schmidt M, Howe SM, Fasselt B, Homann D, Rumenapp U, Sandmann J, Jakobs KH: Mechanisms of phospholipase D stimulation by m3 muscarinic acetylcholine receptor. Evidence for involvement of tyrosine phosphorylation. Eur J Biochem 225: 667–675, 1994 24. Maher PA: Inhibition of the tyrosine kinase activity of the fibroblast growth factor receptor by the methyltransferase inhibitor 5′methylthioadenosine. J Biol Chem 268: 4244–4249, 1993 25. Flati V, Haque SJ, Williams BRG: Interferon-α-induced phosphorylation and activation of cytosolic phospholipase A2 is required for the formation of interferon-stimulated gene factor three. EMBO J 15: 1566–1577, 1996 26. Kast R, Furstenberger G, Marks F: Activation of cytosolic phospholipase A2 by transforming growth factor-α in HEL-30 keratinocytes. J Biol Chem 268: 16795–16802, 1993 27. Schlessinger J, Ullrich A: Growth factor signaling by receptor tyrosine kinase. Neuron 9: 383–391, 1992 28. Qiu Z-H, DeCarvalho M, Leslie CC: Regulation of phospholipase A2 activation by phosphorylation in mouse peritoneal macrophages. J Biol Chem 268: 24506–24513, 1993

29. Xing M, Wilkins PL, McConnell BK, Mattera R: Regulation of phospholipase A2 activity in undifferentiated and neutrophil-like HL 60 cells. Linkage between impaired responses to agonists and absence of protein kinase C-dependent phosphorylation of cytosolic phospholipase A2. J Biol Chem 269: 3117–3124, 1994 30. Durstin M, Durstin S, Molski TFP, Becker EL, Sha’affi RI: Cytoplasmic phospholipase A 2 translocates to membrane fraction in human neutrophils activated by stimuli that phosphorylate mitogen-activated protein kinase. Proc Natl Acad Sci USA 91: 3142–3146, 1994 31. Winitz S, Gupta SK, Qian N-X, Heasley LE, Nemenoff RA, Johnson GL: Expression of mutant Gi2α subunit inhibits ATP and thrombin stimulation of cytoplasmic phospholipase A2-mediated arachidonic acid release independent of Ca2+ and mitogen-activated protein kinase activation. J Biol Chem 269: 1889–1895, 1994 32. Wu T, Levine SJ, Lawrence MG, Logun C, Angus CW, Shelhamer JH: Interferon-γ induces the synthesis and activation of cytosolic phospholipase A2. J Clin Invest 93: 571–577, 1994 33. Sporn MB, Roberts AB: Peptide growth factors are multifunctional. Nature 332: 217–219, 1988 34. Cockcroft S, Thomas GMH, Fensome A, Geny B, Cunningham E, Gout I, Hiles I, Totty NF, Troung O, Hsuan JJ: Phospholipase D: A downstream effector of ARF in granulocytes. Science 263: 523–526, 1994 35. Ohguchi K, Banno Y, Nakashima S, Nozawa Y: Regulation of membrane-bound phospholipase D by protein kinase C in HL 60 cells, Synergistic action of small GTP-binding protein RhoA. J Biol Chem 271: 4366–4372, 1996