Stat signaling pathways to induce proliferation while promoting differentiation in embryonic astrocytes

Int. J. Devl Neuroscience 18 (2000) 693±704

www.elsevier.com/locate/ijdevneu

Prolactin concurrently activates Src-PLD and JAK/Stat signaling pathways to induce proliferation while promoting di€erentiation in embryonic astrocytes Dimitra Mangoura a, b, c,*, Chris Pelletiere a, Soyan Leung a, Nikos Sakellaridis a, De Xin Wang a a

Department of Pediatrics, The University of Chicago, 5841 South Maryland Avenue, Chicago, IL 60637, USA Committee on Neurobiology, The University of Chicago, 5841 South Maryland Avenue, Chicago, IL 60637, USA c Committee on Cell Physiology, The University of Chicago, 5841 South Maryland Avenue, Chicago, IL 60637, USA b

Abstract In normal development, embryonic astrocytes progress through their cell lineage by acquiring di€erentiation, by apoptosis, and by proliferation. In this study, we show that embryonic astrocytes may maintain and make gains in di€erentiation as they simultaneously progress through one cell cycle when induced by prolactin (PRL). Prolactin induced the majority of astrocytes to incorporate bromodeoxyuridine (BrdU) with a four-fold increase over controls after 18 h of exposure. Investigating possible mitogenic signaling pathways we show for the ®rst time that prolactin is coupled to a sustained phospholipase D (PLD) activation, with an ecacy similar to the phorbol ester and astrocytic mitogen 12-tetradecanoylphorbol-13-acetate (TPA). Both cyclosporine and suramin abolished this activation. Staurosporine and calphostin C also inhibited the PRL e€ect by 50%, consistent with involvement of protein kinase C-(PKC)-a, the major PKC isoform in astrocytes. Genistein and PP1 blocked the activation indicating additional regulation by cytosolic tyrosine kinases. This pro®le of PLD activation was suggestive of a PLD I isoform and a mitogenic response. Upon completion of the cell cycle, analysis of glia ®brillary acidic protein (GFAP) and vimentin abundance, and glutamine synthetase (GS) activity showed that astrocytes had gained in expression of di€erentiation markers. Moreover, the intensity of GFAP immuno¯uorescence was greater per cell, as was the length of the cell processes. In exploring the signaling for prolactin-induced di€erentiation we found that prolactin activated the tyrosine kinase Janus kinase (JAK) 2 and signi®cantly stimulated tyrosine, phosphorylation of the prolactin receptor. Stat 1 and 3 were also activated presumably downstream to JAK2 activation. A rapid translocation of the cytosolic Stats over the nucleus was seen in nearly every astrocyte corresponding well with the gains in GFAP per cell. The Stats translocation did not depend on MEK-ERK inhibition by PD98059, inhibition of p38 by 1 mm SB203580, or Src kinase family inhibition by PP1. Our results demonstrate the ability of PRL to concurrently induce activation of PLD, a mitogenic signaling pathway in astrocytes, and prolonged stimulation of Stat1, compatible with the increased GFAP upregulation and cell di€erentiation. Considered together this data may provide an explanation on the fast gain in both numbers and di€erentiation in the astrocytic population during development (HD 09402, CRF). 7 2000 ISDN. Published by Elsevier Science Ltd. All rights reserved. Keywords: Tyrosine phosphorylation; Tyrosine kinases; Src; Stats; Chick embryo; Cultures; PLD

Abbreviations: SDS±PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; HPTLC, high pressure liquid chromatography. * Corresponding author: Kennedy Mental Retardation Research Center MC 5058, Division of Biological Sciences, The University of Chicago, 5841 South Maryland Avenue, Chicago, IL 60637, USA. Tel.: +1-773-702-1136; fax: +1-773-702-9234. E-mail address: [email protected] (D. Mangoura).

In normal development, embryonic astrocytes progress through their cell lineage by undergoing di€erentiation and by proliferation [15,29]. In a series of studies with A. Vernadakis, we have shown that secreted factors regulate both astrocytic proliferation and astrocytic di€erentiation in primary astrocytic cultures and in the astrocytoma cell line C62B

0736-5748/00/$20.00 7 2000 ISDN. Published by Elsevier Science Ltd. All rights reserved. PII: S 0 7 3 6 - 5 7 4 8 ( 0 0 ) 0 0 0 3 1 - 9

694

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

[13,31,42,43]. Primary cultures from either the cerebral hemispheres (CH) or the cerebellum (CB) of 15-dayold chick embryos (E15) have been thoroughly studied and carefully correlated with in vivo events by Antonia Vernadakis and coworkers [13,31,42,43]. By exploring several aspects of astrocytic development these studies have established the E15CH and E15CB cultures as models for gliogenesis in culture, featuring immature [GFAP-(ÿ)/A2B5-(+)] astrocytes which develop to fully di€erentiated astrocytes with developmental timetables precisely re¯ecting the pro®les observed in vivo [31,42,43]. In more recent studies, we have used these two in culture models to determine whether astrocytes may di€erentiate as they proliferate actively and found that cell di€erentiation may occur during progression through one cell cycle [29,33,35]. Speci®cally, we showed that the potent mitogen phorbol 12-O-tetradecanoylphorbol-13-acetate (TPA) induced one cell cycle and resulted in a concomitant dedi€erentiation, producing astrocytes with characteristics of an immature phenotype [33,35]. Thus, TPA increases both the rate of assembly of newly synthesized full-length or nascent vimentin and leads in the long term to increases in vimentin expression, while basic parameters of astrocytic di€erentiation such as GFAP expression [6] and glutamine synthetase activity [36] are downregulated. Accordingly, inhibition of the TPA signaling and e€ects with staurosporine inhibited proliferation and induced di€erentiation [29]. The signaling pathways that regulate the induction of astrocytic di€erentiation and proliferation are just beginning to be understood. The major signaling event in the TPA paradigm is PKC activation and a tyrosine kinase-regulated, sustained activation of phospholipase D (PLD) [24]. PLD catalyzes the hydrolysis of phosphatidylcholine and phosphatidylethanolamine to form phosphatidic acid [18,20,24,27]. Membrane receptor activation may stimulate PLD to hydrolyze membrane phospholipids at the terminal phosphodiester bond to yield free polar head groups and phosphatidate. Phosphatidate may then function as a second messenger for downstream signaling cascades linked to induction of mitosis. In addition, a coupled action of phosphatidate phosphohydrolase (PPH) generates diacylglycerol which can in turn serve as a source for diacylglycerol generation, leading on to protein kinase C (PKC) activation and further downstream mitogenic signaling via mitogen-activated protein kinases (MAPKs). Therefore, PLD is now viewed as a key enzyme for cell mitosis in several cell types, de®nitely inhibiting astrocytes [25,27,34,35]. PKC activation is also required by several agonists, such as ciliary neurotrophic factor (CNTF), for activation of the MAP kinase ERK and subsequent astrocytic di€erentiation [40]. Therefore, long term

downregulation of PKC-a by TPA may disrupt this pathway and lead to de-di€erentiation. A second and better studied pathway involves activation of the signal transducer and activator of transcription (Stat) family of transcription factors. Speci®cally, genetic analysis has shown that the tyrosine kinases of the Janus kinase family (JAK) signaling pathway speci®cally induce di€erentiation in astrocytes by activating the Stats [44]. CNTF, for example, causes the vast majority of stem cells to di€erentiate into astrocytes [22,23] by speci®cally activating JAK2 and Stat3, which upregulate the GFAP promoter and gene expression [8]. Pituitary and extrapituitary prolactin has several physiological growth factor-type actions in the CNS, such as sleep±wake cycle regulation, stimulation of avian parenting and mating behavior and, at least in culture, as a mitogen for astrocytes [4,7,14,38]. The prolactin receptor (PRLR) itself does not demonstrate protein kinase activity, but signals through activation of preassociated cytoplasmic tyrosine kinases from the JAK and Src kinase families [2,5,10,12,26,41,47]. JAKs have been primarily thought of as Stat activators, considered to be the major e€ectors for prolactin-dependent cell and tissue growth [21], while Src kinase activation is associated with mitogenic signaling in most cell types [3]. Therefore, we undertook these studies to investigate the outcome of di€erentiation of astrocytes after the induction of one cell cycle by prolactin and to identify and possibly dissect the prospective signaling of prolactin, as it relates to either astrocytic proliferation or regulation of di€erentiation. 1. Experimental procedures 1.1. Materials Dulbecco's Modi®ed Eagle's Medium (DMEM) and fetal bovine serum (FBS) were purchased from Gibco/ BRL in Grand Island, NY. Ovine prolactin was purchased from Sigma and chicken prolactin was provided by the National Hormone and Pituitary Program, NIH, Rockville, MD. Staurosporine and BrdU were purchased from Sigma Chemical Company, St Louis, MO. Cyclosporine was a generous gift from Dr Peter Whitington, University of Chicago. Genistein, calphostin C, PP1, and PD98059 were purchased from LC Services Corporation, Woburn, MA. Suramin (NSC34936) was provided by the Drug Synthesis and Chemistry Branch, Developmental Therapeutics Program, Division of Cancer Treatment, National Cancer Institute, Bethesda, MD. Monoclonal antibodies (mAb) to vimentin was from Sigma, Israel; polyclonal antibodies against GFAP were from DAKO and a monoclonal to GFAP was from Sternberger-Meyer, Inc.; mAb 4G10 against phosphotyrosine, a polyclonal to JAK2, and

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

recombinant protein-agarose were purchased from Upstate Biotechnology Incorporated (UBI), Lake Placid, NY; the U5 monoclonal antibody against the prolactin receptor was from ABR Inc, Boulder, CO; mAb to Stat1 was from Transduction Laboratories, Lexington, KY; and anti-5-bromo-2 '-deoxyuridine (BrdU) antibodies from Amersham, Arlington Heights, IL. Species-speci®c antibodies conjugated to ¯uorochromes rhodamine or ¯uorescein were from Sigma, St. Louis, MO. Antibodies conjugated to peroxidase were from DAKO and unconjugated rabbit anti-mouse, goat anti-mouse, and goat anti-rabbit from Jackson Immunoresearch Laboratories, Inc, West Grove, PA. [3H]-Palmitic acid was from Amersham.

1.2. Cell culture Monolayer astrocytic cultures were established by mechanical dissociation from 15-day-old cerebelli (E15CB), as previously described [35,42]. Cultures were maintained at 378C and 10% CO2/air and the medium (DMEM+20% fetal bovine serum) was replenished every third day. Cell cultures were switched to serum-free DMEM 24 h before treatment with prolactin (ovine or chicken) or other agonists at concentrations and for times indicated per experiment, typically between culture days seven (C7) to 10.

1.3. BrdU incorporation and immunocytochemistry BrdU incorporation into the nuclei was visualized following the protocol of De Vito et al. [14] with a few modi®cations. Cultures, treated as described in Results, were ®xed in absolute methanol for 10 min at 48C. The DNA was denatured by incubation for 1 h at 378C in 1 M HCl. After neutralization with a 0.1 M borate bu€er, pH 8.5, cells were washed in phosphatebu€ered saline (PBS). Immunostaining with anti-BrdU mAb (6 mg/ml) and an anti-mouse rhodamine-conjugated IgG was done as previously described [30]. Cells were photographed using an upright Leitz microscope with an epi¯uorescence unit, using a Neo¯uar 20 lens, an Olympus camera and 800 ASA Kodak Ectachrom slide ®lms. For immunodetection of all other epitopes, astrocytes grown on glass coverslips were treated with vehicle or PRL 2 antagonists for the times and concentrations indicated in the Results and ®xed with methanol for 10 min at ÿ208C. Fluorescence immunocytochemistry was performed with indicated secondary antibodies as described [30,33]. Cells were photographed through the same microscope, using a Neo¯uar 40 lens.

695

1.4. Glutamine synthetase activity Glutamine synthetase activity was measured as described for primary cultures [24,42,43]. 1.5. Western blotting analysis of tissue and tissue cell culture Astrocytes were mechanically harvested in their culture medium and recovered by low-speed run centrifugation. After two rinses with ice cold PBS, astrocytes were lysed directly in 8% urea with 0.5% SDS. Equal amounts of protein were separated by SDS±PAGE, blotted onto nitrocellulose membranes, stained with the reversible Ponceau S stain to assure equal loading, and probed with antibodies against vimentin or GFAP, as described [33,35]. Puri®ed GFAP was also run in separate lanes to help quantitate densitometry [35]. Results were visualized by further incubation with HRP conjugated antibodies and the enhanced chemiluminescence detection system. Blots were scanned to produce digital images using Ofoto scanning software; densitometry analysis was performed with NIH Image 1.54, and the obtained numerical data were plotted with Microsoft Excel 4.0. 1.6. Immunoprecipitations Immunoprecipitations were performed as described previously [10,26,30]. For the prolactin receptor (PRLR), cells were harvested in 500 ml of lysis bu€er (10 mM Tris, pH 7.5, 5 mM EDTA, 50 mM NaCl, 50 mM NaF, 0.2% Nonidet P-40, 100 mM sodium orthovanadate, 1 mg/ml aprotinin, 1 mg/ml leupeptin, and 1 mM PMSF) for 30 min. After removal of insoluble material (10,000 rpm for 10 min at 48C), supernatants were assayed for protein content with the DC-Biorad kit. Equal amounts of lysates were incubated with 5 mg/ml mAb U5 for at least 2 h at 48C, prior to the addition of 2 mg/ml rabbit-anti-mouse antibodies for another hour (bridge), and 100 ml 50% protein-A-agarose for a third hour-long incubation. Immunoprecipitates were recovered by centrifugation and rinsed, before the precipitated material was eluted from the beads by heating in 25 ml of sample bu€er at 958C for 10 min. The samples were ®nally analyzed by SDS±PAGE on 8% polyacrylamide gels and by Western blotting. A similar protocol was followed for JAK2 immunoprecipitations, without the bridge step. The phosphotyrosine content of immunoprecipitated proteins was detected by incubating the membranes with the mAb 4G10. Immunoreactivity was visualized by further incubation with goat- anti-mouse-HRP conjugated antibodies and the ECL chemiluminescence procedure. To assure equal loading, the same mem-

696

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

branes were stripped and reblotted with the same antibodies that had been used for immunoprecipitations, for example immunoprecipitates with U5 mAb were reblotted also with U5, using a goat-anti-mouse alkaline phosphatase-conjugated antibody and the NBT/ BCIP colorimetric immunoreactivity detection system.

1.7. Phospholipase D (PLD) activity Primary astrocytic cultures were labeled with [3H]Palmitic acid 24 h prior to harvesting, as described [34]. Prolactin, suramin, cyclosporine, and kinase inhibitors were added at the times and concentrations

Fig. 1. Prolactin (PRL) increases BrdU uptake in chick embryo cerebellar astrocytes (E15CB). Microphotographs of BrdU incorporation into the nucleus from astrocytes treated as indicated. After 18 h, 1 mM BrdU was added for 1 h and its incorporation into DNA was visualized with indirect rhodamine immuno¯uorescence, as described in Procedures. PRL caused a signi®cant, three-fold increase over the control, and this increase was blocked by cyclosporine A (20 lens). Fig. 2. Prolactin maintains di€erentiation in E15CB. Cells were incubated for 24 h with 10 nM prolactin prior to assessments of astrocytic di€erentiation marker expression. A typical microphotograph of GFAP staining reveals a range of rhodamine immuno¯uorescence intensity in control cultures and high GFAP immunostaining and elongation of processes in PRL treated-cultures (40 lens). In addition, PRL caused signi®cant di€erences in the activity of glutamine synthetase activity (values are the mean of several cultures from three di€erent experiments 2SEM, p < 0.01). Analysis of GFAP abundance (described in procedures) also revealed signi®cant increases after 24 h of prolactin (p < 0.02) in GFAP expression. The decrease in vimentin abundance, a marker for immature astrocytes, was not signi®cant (values represent the mean of 12 culture dishes from three di€erent experiments 2SEM).

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

indicated in the results. Phospholipase D activity, measured as the formation of [3H]-phosphatidylethanol (PtdEth), was isolated by extraction and high pressure thin layer chromatography (HPTLC) [28,34]. The thin layer chromatography (TLC) plates were then placed with photographic ®lm for a 3-day exposure. Bands corresponding to the PtdEth standard were scraped, and the radioactivity was measured by liquid scintillation counting of quadruplicate samples. 1.8. Protein assay Protein content was determined by the DC Biorad kit, (Hercules, CA) using bovine serum albumin as a standard. 1.9. Statistics All experiments were performed 4±10 times. Statistical signi®cance of numerical data was analyzed by ANOVA statistics. 2. Results 2.1. Prolactin (PRL) induces proliferation in E15CB astrocytes The ability of PRL to stimulate mitosis in embryonic astrocytes was demonstrated by showing increased BrdU uptake. The average cell cycle of astrocytes in E15CB cultures lasts on the average 24±28 h [33,35]. Cell cycling was ®rst arrested by serum withdrawal for 24 h, and then cultures were incubated with PRL or vehicle. Eighteen hours later, BrdU was added at 1 mM for an additional 60 min after which cultures were ®xed and BrdU uptake was assessed immunocytochemically, as described in Procedures. As illustrated in Fig. 1, astrocytes incubated with PRL for 18 h showed a substantial three-fold increase in BrdU uptake over controls; the increase was evident with concentrations of 1, 10 or 100 nM. Treatment with 10 nM cyclosporine, a potent PRL antagonist, inhibited the PRL-induced mitogenesis as evident by an uptake of BrdU dramatically lower than the PRL-only cultures and very similar to the control cultures. Therefore, PRL induces BrdU uptake and cell proliferation in embryonic astrocytes. 2.2. Prolactin (PRL) maintains astrocytic di€erentiation after the induction of mitosis in E15CB astrocytes We subsequently examined the di€erentiation state of astrocytes after completion of one PRL-induced cell cycle, using as di€erentiation markers: (i) expression

697

and subcellular localization of GFAP; (ii) vimentin expression; and (iii) glutamine synthetase activity. Astrocytes were treated for 24±28 h with either vehicle or 10 nM PRL and processed for immunocytochemical and biochemical analysis. As shown in Fig. 2, under control conditions, the majority of control cells exhibited GFAP immuno¯uorescence, albeit exhibiting a range of intensity per cell. The diameter of the majority of control astrocytic bodies was larger than the length of their processes. In contrast, the intensity of immuno¯uorescence in astrocytes grown in the presence of PRL for 24 h was greater as compared to controls, with the number of low intensity GFAP astrocytes being negligible. In addition, GFAP ®laments appeared well organized, supporting processes of considerably greater length than in controls. This type of organization was even seen in astrocytes that apparently have just completed one cell cycle (Fig. 2, arrow). In sister cultures, the enzymatic activity of glutamine synthetase was measured as the formation of gglutamylhydroxamic acid, and GFAP and vimentin protein expression were quantitated by densitometry analysis of Western blots, as described [33,35]. In astrocytes that had completed one PRL-induced cell cycle, expression of GFAP was signi®cantly increased ( p < 0.01), while the level of vimentin was not a€ected (Fig. 2). Moreover, glutamine synthetase activity was also increased signi®cantly over vehicle-treated cultures ( p < 0.01) (Fig. 2). This data suggests that during the induced cell cycle PRL additionally maintained and augmented the expression of di€erentiation markers in embryonic astrocytes. 2.3. Prolactin stimulates Phospholipase D (PLD) in a time-dependent and dose-dependent manner We have shown that a multifold and sustained stimulation of PLD activity is a major action of the mitogen TPA in astrocytes [34]. To test the prediction that activation of PLD is a part of the mitogenic PRL signal cascade in astrocytes, we investigated the ability of PRL to stimulate PL D activation (measured as the formation of 3[H]palmitoylated phosphatidylethanol). Using PRL concentrations ranging from 10ÿ11 to 10ÿ5 M for times ranging from 1 to 10 min we observed drastic dose- and time-dependent PLD activation. All concentrations from 10ÿ9 to 10ÿ5 M stimulated PLD activation, with an EC50 at 10 min of exposure of 10 nM (Fig. 3, inset). Using 10 nM prolactin, we also established the time course of PLD activation (Fig. 3). Activation was apparent by 1 min, reached a peak at 10 min, and then declined, returning to basal levels by 3 h (Fig. 3). Compared to other physiological agonists [34], PRL is not only the most potent physiological agonist of PLD activation in astrocytes, but also produces the longest duration of activation.

698

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

Fig. 3. Prolactin stimulates phospholipase D activity in chick embryonic astrocytes (E15CH). Astrocytes were exposed to [3H] palmitate for 18 h, 0.2% ethanol and 10 nM prolactin for time indicated. Phospholipase D activity was measured as the formation of [3H]phosphatidylethanol (PtdEth), isolated by extraction and HPTLC; radioactivity was determined by liquid scintillation counting and counts normalized per protein content. Values (.) represent the mean 2SEM (bars) from nine experiments. Inset contains a dose response curve obtained with indicated concentrations at 10 min of exposure.

2.4. E€ects of the PKC and tyrosine kinase activity on prolactin-stimulated PLD activation To determine if PRL-stimulated PLD activation was dependent on PKC, we examined the e€ect of PKC inhibitors staurosporine (St) and calphostin C (Cal C) on phosphatidylethanol synthesis corresponding to PLD activation. Addition of 100 nM staurosporine 30 min prior to a 10 min addition of 10 nM PRL resulted in a signi®cant inhibition of PLD activation (39% of PRL control) (Fig. 4), while calphostin C, at 10 nM, inhibited PLD activation to a signi®cant but lesser degree (67% of control) (Fig. 4). Overall, the PKC inhibitors signi®cantly inhibited, but did not abolish the PRL e€ect, which suggests that there are additional, PKC-independent mechanisms of PLD regulation. The prolactin receptor (PRLR) is known to preassociate with and elicit its action through tyrosine kinases. To address the input of tyrosine kinases in the PRLR-PLD pathway, four speci®c inhibitors were tested for their e€ect on PLD activation. Both genistein, a tyrosine±kinase inhibitor, and PP1, a Src kinase inhibitor blocked the PLD activation signi®cantly, resulting in phosphatidylethanol synthesis levels only slightly higher than the basal level. Speci®cally, genistein added at 50 mg/ml 30 min prior to a 10 min treatment with 10 nM PRL caused a drastic inhibition of PLD activation (90% of PRL control) (Fig. 4). PP1 treatment also strongly inhibited the PRL-stimulated PLD activation resulting in only slightly higher than baseline activity (Fig. 4). Suramin (15 mM/ml), a direct inhibitor of the PLD-transphosphatidylation reaction

Fig. 4. Dependence of PRL-stimulated PLD activation in E15CB astrocytes on protein kinase C (PKC) and tyrosine kinases. Astrocytes were pretreated with the indicated inhibitors prior to a 10 min stimulation with 10 nM PRL (PLD activity, measured as [3H]PtdEth cpm/mg protein, after 10 min of PRL alone was acknowledged as 100%). PKC inhibitors staurosporine and calphostin C substantially inhibited the PRL-induced PLD activation. The tyrosine kinase inhibitors genistein and PP1 inhibited the PRL-stimulated PLD activation almost completely. Suramin and cyclosporine A also inhibited PLD activation signi®cantly. This experiment was repeated four times with identical results.

between ethanol and phosphatidylcholine (PC) [17], also caused similar inhibition (Fig. 4). Finally cyclosporine A (100 nM) an inhibitor of PRL-induced mitogenesis, essentially abolished the PRL-stimulated PLD activation in astrocytes (Fig. 4). 2.5. Prolactin (PRL) induces tyrosine phosphorylation of the PRL receptor and JAK2, and regulates abundance and subcellular distribution of Stat1 and Stat3 in astrocytes Activation of the JAK/Stat pathway has been documented to induce astrocytic di€erentiation, evidently by increasing the activity of the GFAP gene promoter [8]. Expression of PRLR, an established activator of this signaling pathway in several other cell types, has been strongly indicated in astrocytes but not yet demonstrated. Therefore we employed a combination of immunoprecipitation and Western blot analysis to investigate the expression and activation of the PRLR in E15CB astrocytes at 10 days in culture (C10). By immunoprecipitation of astrocytic lysates with the U5 mAb and Western blot analysis with antiphosphotyrosine antibodies, we found that the receptor was expressed as an 89 kDa protein band. Upon stimulation with chicken prolactin the receptor was tyrosine phosphorylated and it appeared as a doublet; the upper band was more immunoreactive to the antiphosphotyrosine antibodies (Fig. 5A, arrow). The increase

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

was evident at 5 min, and continued to reach its maximum at 30 min of incubation with 10 nM PRL. Next, we examined the tyrosine phosphorylation of the tyrosine kinase JAK2 with time of exposure to 10 nM PRL. Cell lysates from astrocytes treated with prolactin for the indicated times were immunoprecipitated with a polyclonal antibody to JAK2 and the immunoprecipitates were processed for Western blot analysis with antiphosphotyrosine antibodies. JAK2 was acutely and maximally tyrosine phosphorylated by PRL at 1 min of incubation, but not in the presence of cyclosporin (Fig. 5B). By 30 min, the levels of tyrosine phosphorylation of JAK2 had signi®cantly declined (Fig. 5B). We further investigated the JAK-Stat pathway activation by PRL in astrocytes. After verifying by immunoprecipitations and Western blot analysis that Stat1 phosphorylation was indeed induced by PRL (data not shown), we proceeded to investigate the activation at the single cell level, by analyzing the expression and subcellular localization of Stat1 in astrocytes in response to PRL. Immunocytochemistry with a mAb against Stat1 revealed an increasing translocation of Stat1 staining in the nucleus of virtually all astrocytes at time of PRL treatment (Fig. 5C). The baseline levels were relatively low (Fig. 5C, vehicle), however, as early as 5 min, concentration of Stat1 immunoreactivity towards the perinuclear area was apparent, became prominent by 15 min, and continued increasing for over 2 h (Fig. 5C). PRL also activated Stat3 with a time course similar to Stat1. The regulation of the phosphorylation of both Stats by speci®c signaling kinase inhibitors was also very similar as shown representatively in Fig. 5D. Stat3 immunoreactivity (red) was low in unstimulated cultures and mostly cytosolic as compared to GFAP immunoreactivity (green) in double staining experiments (arrows in upper row point to the same cell). Upon activation with PRL Stat3 immunoreactivity became increasingly concentrated perinuclearly. This translocation (seen as red throughout Fig. 5D) over the nucleus (seen as blue with Hoechst counter staining) was not inhibited by PD98057 [1], an inhibitor of the ERK kinase MEK (arrows in 2nd row point to a typical image of activated Stat3 and the same spot under the UV ®lter), 1 mM PP1, an inhibitor of the Src family of tyrosine kinases (arrows in the 3rd point to the same cell), or SB 203580 at 1 mM (4th row), a concentration that inhibits the MAP kinase p38 [20]. 3. Discussion These studies were undertaken to investigate the coupling of the prolactin receptor to signaling pathways as it relates to the induction of one cell cycle and

699

regulation of astrocytic di€erentiation-speci®c genes. Our results demonstrate the ability of PRL to concurrently induce activation of PLD, a mitogenic signaling pathway in astrocytes, and produce a prolonged stimulation of Stat1, leading to increased GFAP upregulation and cell di€erentiation. It is known that during development astrocytes both proliferate rapidly and at the same time di€erentiate, progressing along their cell lineage, one of the more complicated among tissues [9]. It is not clear however when the advance in di€erentiation occurs, i.e., during mitosis or between mitoses. We examined whether the known mitogen prolactin may have concomitant e€ects on proliferation and di€erentiation in astrocytes, by assessing mitosis and then di€erentiation of astrocytes at the completion of one cell cycle. Prolactin caused the majority of cells (>60%) to enter the cell cycle, as shown by the numbers of cells synthesizing DNA (Fig. 1A). Cell proliferation was dependent upon PRL binding to the PRL receptor as cyclosporine, which removes the PRL molecule from its receptor, resulted in a level of BrdU uptake similar to the levels of uptake in untreated astrocytes. Our results support previous studies indicating PRL as a potent mitogen in astrocytic population [14] by extending the observation both to di€erent species and to embryonic stages. Moreover, all parameters of di€erentiation that we examined were enhanced after completion of a single, PRL-induced cell cycle. Total GFAP abundance and glutamine synthetase activity was signi®cantly higher, while vimentin abundance was slightly decreased. The increases in length of astrocytic processes and in GFAP immunostaining per cell were consistent with progress in di€erentiation per astrocyte after exposure to PRL. This result furthers the notion that di€erentiation may be speci®cally regulated during the cell cycle [37]. As we have shown previously, TPA, a potent mitogen in astrocytes, downregulates several of the astrocytic di€erentiation markers towards levels of expression found in immature astrocytes [33]. Speci®cally, TPA increases both the post-translational assembly (rate of assembly of newly synthesized full-length protein) and the co-translational assembly (rate of assembly of nascent protein) of vimentin [33], and leads to long term increases in vimentin abundance. Vimentin abundance decreases with astrocytic maturation as well as maturation in other lineages [39]. TPA also decreased GFAP abundance and glutamine synthetase activity, which both increase with astrocytic maturation and are thus the overall indicators of the astrocytic phenotype [6]. In contrast, none of these parameters were downregulated as the outcome of one prolactin-induced cell cycle. These ®ndings strongly indicated that PRL signaling induced proliferation, while it maintained and enhanced the di€erentiation of the astrocytes.

700

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

This work is also a ®rst report on the coupling of PRLR to PLD activation, a major component of mitogenic signaling in astrocyte [25,34]. The activation peak was three-fold higher than the control levels, making PRL the most potent physiological stimulator of PLD activation yet described in chick embryo astrocytes. Once activated, PLD produces two important second messengers, phosphatidic acid (PA) and diacylglycerol (DAG). PA has been shown to bring about mitogenic responses [16,48], while DAG can activate PKC. Once activated, PKC is then able to activate PLD directly, producing a feedback loop, until DAG production ceases. Recent site-speci®c mutagenesis studies have shown that the alpha isoform of PKC interacts with a site in the amino terminus of PLD1 to confer PKC-a-dependent PLD activation in vitro [19,46]. PKC inhibitors staurosporine and the more speci®c Calphostin C inhibited 40±60% of the PRLstimulated PLD activation. Taking into consideration that PKC-a is the major PKC isoform in chick embryo astrocytes [32], PLD1 is most likely the PLD isoform activated by prolactin in the same cells. Another class of e€ectors that may act either alone or synergistically with PKC-a to stimulate PLD1 in vitro, are small GTP-binding protein members of the Arf, Rac, or Rho families; Src may facilitate the interactions. Constitutively active tyrosine kinase v-Src, is known to stimulate the generation of PA and LPA from PtdCho breakdown, primarily via a Ras-dependent PLD activation. Based on site speci®c mutagenesis studies, the Src-dependent PLD activation is mediated by the exchange factor RalA. A Src-activated Ras brings Ral-GDS to the plasma membrane, where Ral-GDS brings a RalA-PLD complex together, generating further signaling. The model suggests that Ral is bound to PLD but cannot directly activate the enzyme; therefore other molecules, such as Arf, Rac, or Rho are necessary for a fully activated signaling complex. The tyrosine kinase inhibitor genistein and the Src speci®c inhibitor PP1 both abolished the PRL-stimulated activation peak, yielding phosphatidylethanol synthesis almost identical to the basal levels. This evidence suggests that PRL-stimulated PLD activation is regulated by tyrosine kinases. Furthermore, both the onset timing of tyrosine kinase Src (DM, personal data) and PLD activation and the >95% inhibition by the Src inhibitor PP1 are consistent with Src activation being upstream to PLD activation in E15CB astrocytes. While our ®ndings do not exclude the possibility that PLD2 is expressed in embryonic astrocytes or that it is activated by prolactin, they suggest that the observed PLD activity is consistent with the PLD1 type. Similarities, such as the sensitivity to inhibition of tyrosine kinases, in the regulation of PLD activation by TPA with

that by prolactin, indicate that this PRL signaling has most likely, the same biological outcome, namely mitosis. Tyrosine kinases physically preassociate with PRLR and, upon ligand binding and conformation changes, become activated to transduce the biological e€ects of PRL. Activation of JAK2, in particular, is considered the ®rst event following ligand stimulation and receptor dimerization. JAK2 may then phosphorylate the receptor either directly as its own substrate, or via an intermediate tyrosine kinase with speci®city for the PRL receptor. Our results support this timetable in astrocytes, since JAK2 reached is maximum tyrosine phosphorylation acutely (1 min), while the phosphotyrosine content of the receptor kept increasing for relatively longer. In the E15CB system of chick cerebellar astrocytes, the receptor appears with an apparent Mr of 89 kDa. This is compatible with the reported size of a chicken PRLR cDNA, which consists of 222 bp of the 5'-noncoding region, 2496 bp of open reading frame, and 165 bp of 3 '-noncoding region [45]. Activation of Stat1 was as acute as the JAK2 activation, however its duration was longer. Increased phosphorylation and translocation to the nucleus of Stat1 was evident for up to 6 h after the introduction of PRL into the culture media; activation of Stat3 was somewhat shorter. However, the time course of phosphorylation alone does not de®ne the nature of the signaling as mitogenic or di€erentiation related. More enlightening was the time course and the extent of nuclear translocation of the activated Stats, which were compatible with Stats being transcriptionally active. Stat1 in particular was active in virtually every astrocyte (Fig. 5C), in a pattern consistent with the upregulation of GFAP immunocytochemistry and processes formation (Fig. 2), in contrast to the smaller percentage of astrocytes which incorporated BrdU to undergo proliferation. Therefore, it appears that the activation of Stat1 correlated tightly with astrocytic di€erentiation. Previous studies addressing whether the JAK/Stat pathway alone is ecient for astrocytic di€erentiation have indicated that ERK activation [40] may be additionally required. These two pathways may also crosstalk; in some cases Stats may be activated in an ERK-dependent pathway [11] and inhibition of ERK activation delays CNTF (and JAK/Stat-dependent)dependent astrocytic di€erentiation, while in the case of bone morphogenetic proteins ERK was solely required [40]. However, these studies were focused on stem cell long-term di€erentiation to astrocyte [22]. In our system, which consists of committed astrocytes, it appears that both pathways function concurrently and independently towards the same target, astrocytic di€erentiation. Speci®cally, PRL activated ERK, but inhibition of ERK did not inhibit astrocytic prolifer-

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

701

Fig. 5. Activation of the JAK/Stat pathways by the prolactin receptor (PRLR) in astrocytes. A, prolactin-dependent induction of tyrosine phosphorylation of PRLR; and B, JAK2. Astrocytes were stimulated with control or 10 nM PRL for various time intervals and lysed as described in Procedures. Lysates were split in half and immunoprecipitated with U5 anti-grolactin receptor (PRLR) mAb (A), or with polyclonal antibodies against human JAK2 (B). Recovered proteins were analyzed by SDS±PAGE and Western blots with antiphosphotyrosine antibodies. (A) Astrocytes express PRLR, with an apparent Mr of 89 kDa. PRL increases the levels of tyrosine phosphorylation of the receptor (arrow). (B) Western blot analysis of the immunoprecipitates with anti-phosphotyrosine antibodies revealed that the tyrosine phosphorylation of JAK2 occurs prior to PRLR phosphorylation (MWM are the Mr markers; IgG (H) indicates the heavy chain of the immunoprecipitating mouse monoclonal antibodies). C, Prolactin-dependent translocation of Stat1. Cultured astrocytes were treated, ®xed, and immunostained with a monoclonal antibody to Stat1 and a secondary antibody, conjugated to ¯uorescein. Microphotographs show that PRL increased the intensity of the staining, plus the subcellular localization of the transcription factor Stat1 became primarily nuclear (40 lens). D, regulation of prolactin-dependent translocation of Stat3. Cultured astrocytes were treated, ®xed, and immunostained with a polyclonal antibody to Stat3 and a rhodamineconjugated secondary antibody (left column); some cultures were double stained with an mAb to GFAP and a ¯uorescein-conjugated secondary antibody (upper right panel) or with the Hoechst counter stain (remaining panels in right column). Microphotographs show that under control conditions Stat3 immunoreactivity was low and mostly cytosolic, as compared to GFAP (arrows top row). The PRL-induced perinuclear concentration of Stat3 immunoreactivity was not inhibited by the MEK inhibitor of the PD98057, the Src inhibitor PP1, or the p38 inhibitor SB 203580 (arrows in 2nd , 3rd and 4th rows point to the same point) (40 lens).

702

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

Fig. 5 (Continued)

Fig. 6

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704

ation (DM, submitted), suggesting that ERK activation may also participate in di€erentiation by regulating other genes of the astrocytic maturation program. Furthermore, it appears that Stat activation is independent from ERK activation because the ERK kinase (MEK) inhibitor PD95087 did not have any e€ect on the nuclear translocation of either Stat in response to PRL. Similarly, low concentration of SB203580, which speci®cally inhibits the MAP kinase p38, had no e€ect on the translocation. The Src inhibitor PP1 had no e€ect either, which indicates that JAK/ Stat is independent from Src activation and further supports the candidacy of the Src-PLD pathway as the mitogenic signaling for PRL. In conclusion, our ®ndings are consistent with a model, depicted in Fig. 6, where prolactin activates a mitogenic signaling pathway in parallel with di€erentiation pathways. The multiplicity of signaling as in the case of PRL argues for a need to generate astrocytes that may undergo proliferation in a regulated manner without loss of function during development. Considered together, this data may provide an explanation of the fast gain in both numbers and extent of di€erentiation in the astrocytic population during development.

Acknowledgements This work was supported by The National Institutes Grant HD-09402 and Brain Research Foundation Award to DM.

[8]

[9] [10]

[11]

[12]

[13]

[14]

[15] [16]

[17]

[18]

[19]

References [1] Alessi, D. R., Cuenda, A., Cohen, P., Dudley, D. T. and Saltiel, A. R., PD098059 is a speci®c inhibitor of the activation of mitogen-activated protein kinase kinase in vitro and in vivo. J. Biol. Chem., 1995, 270, 27489±27494. [2] Bazan, J. F., A novel family of growth factor receptors: a common binding domain in the growth hormone, prolactin, erythropoietin and IL-6 receptors, and the p75 IL-2 receptor betachain. Biochem. Biophys. Res. Commun., 1989, 164, 788±795. [3] Belsches, A. P., Haskell, M. D. and Parsons, S. J., Role of c-Src tyrosine kinase in EGF-induced mitogenesis. Front. Biosci., 1997, 2, 501±518. [4] Ben-Jonathan, N., Mershon, J. L., Allen, D. L. and Steinmetz, R. W., Extrapituitary prolactin: distribution, regulation, functions, and clinical aspects. Endocr. Rev., 1996, 17, 639±669. [5] Berlanga, J. J., Fresno, Vara J. A., Martin-Perez, J. and GarciaRuiz, J. P., Prolactin receptor is associated with c-src kinase in rat liver. Mol. Endocrinol., 1995, 9, 1461±1467. [6] Bignami, A. and Dahl, D., The astroglial response to stabbing. Immuno¯uorescence studies with antibodies to astrocyte-speci®c protein (GFA) in mammalian and submammalian vertebrates. Neuropathol. Appl. Neuroblol., 1976, 2, 99±111. [7] Bole-Feysot, C., Gon, V., Edery, M., Binart, N. and Kelly, P. A., Prolactin (PRL) and its receptor: actions, signal transduc-

[20]

[21]

[22]

[23]

[24] [25]

[26]

703

tion pathways and phenotypes observed in PRL receptor knockout mice. Endocr. Rev., 1998, 19, 225±268. Bonni, A., Sun, Y., Nadal-Vicens, M., Bhatt, A., Frank, D. A., Rozovsky, I., Stahl, N., Yancopoulos, G. D. and Greenberg, M. E., Regulation of gliogenesis in the central nervous system by the JAK-STAT signaling pathway. Science, 1997, 278, 477± 483. Cameron, R. S. and Rakic, P., Glial cell lineage in the cerebral cortex: a review and synthesis. Glia, 1991, 4, 124±137. Cheng Y., Zhizhin I., Periman R. L. and Mangoura D., Prolactin-induced cell proliferation in PC12 cells depends on JNK but not ERK activation. J. Biol. Chem., 2000, 275, 23326± 23332. Chung, J., Uchida, E., Grammar, T. and Blenis, J., STAT3 serine phosphorylation by ERK-dependent and -independent pathways negatively modulates its tyrosine phosphorylation. Mol. Cell. Biol., 1997, 17, 6508±6516. DaSilva, L., Howard, O. M., Rui, H., Kirken, R. A. and Farrar, W. L., Growth signaling and JAK2 association mediated by membrane-proximal cytoplasmic regions of prolactin. J. Biol. Chem., 1994, 269, 18267±18270. Davies D. L. and Vernadakis A., Responses in astrocytic C6 glioma cells to ethanol and dibutyryl cyclic AMP. Brain Res., 389, 1986, 253±260. DeVito, W. J., Avakian, C., Stone, S. and Okulicz, W. C., Prolactin-stimulated mitogenesis of cultured astrocytes. Endocrinol., 1992, 130, 2549±2556. Fedoro€, S. and Vernadakis, A. (eds), Astrocytes, Vols 1. Academic Press, New York, 1986. Fukami, K. and Takenawa, T., Phosphatidic acid that accumulates in platelet-derived growth factor-stimulated Balb/c 3T3 cells is a potential mitogenic signal. J. Biol. Chem., 1992, 297, 10988±10993. Gratas, C. and Powis, G., Inhibition of phospholipase D by agents that inhibit cell growth. Anticancer Res., 1993, 13, 1239± 1244. Ha, K. S. and Exton, J. H., Activation of actin polymerization by phosphatidic acid derived from phosphatidylcholine in IIC9 ®broblasts. J. Cell Biol., 1993, 123, 1789±1796. Hammond, S. M., Autschuller, Y. M., Tsung, T.-C., Rudge, S. A., Rose, K., Engebrecht, J., Morris, A. J. and Frohman, M. A., Human ADP-ribosylation factor-activated phosphatidylcholine-speci®c phospholipase D de®nes a new and highly conserved gene family. J. Biol. Chem., 1995, 270, 29640±29643. Han, Z., Boyle, D. L., Aupperle, K. R., Bennett, B., Manning, A. M. and Firestein, G. S., Jun N-terminal kinase in rheumatoid arthritis. J. Pharmacol. Exp. Ther., 1999, 291, 124±130. Hennighausen, L., Robinson, G. W., Wagner, K. U. and Liu, W., Prolactin signaling in mammary gland development. J. Biol. Chem., 1997, 272, 7567±7569. Johe, K. K., Hazel, T. G., Muller, T., Dugich-Djordjevic, M. M. and McKay, R. D. G., Single factors direct the di€erentiation of stem cells from fetal and adult central nervous system. Genes Dev., 1996, 10, 3129±3140. Kahn, M. A., Huang, C. J., Caruso, A., Barresi, V., Nazarian, R., Condorelli, D. F. and de Vellis, J., Ciliary neurotrophic factor activates JAK/Stat signal transduction cascade and induces transcriptional expression of glial ®brillary acidic protein in glial cells. J. Neurochem., 1997, 68, 1413±1423. Kanfer, J. N., The base exchange enzymes and phospholipase D of mammalian tissues. Can. J. Biochem., 1980, 58, 1370±1380. KoÈtter, K. and Klein, J., Adrenergic modulation of astroglial phospholipase D activity and cell proliferation. Brain Res., 1999, 830, 138±145. Lebrun, J. J., Ali, S., Sofer, L., Ullrich, A. and Kelly, P. A., Prolactin-induced proliferation of Nb2 cells involves tyrosine

704

[27] [28] [29]

[30] [31]

[32]

[33]

[34]

[35]

[36] [37]

D. Mangoura et al. / Int. J. Devl Neuroscience 18 (2000) 693±704 phosphorylation of the prolactin receptor and its associated tyrosine kinase JAK2. J. Biol. Chem., 1994, 269, 14021±14026. Liscovitch, M., Phospholipase D: role in signal transduction and membrane trac. J. Lipid Mediat. Cell Signal., 1996, 14(13), 215±221. Mangoura, D. and Dawson, G., Opioid peptides activate Phospholipase D and protein Kinase C±E in chick embryo neuronal cultures. Proc. Nat. Acad. Sci., 1993, 90, 2915±2919. Mangoura, D. and Dawson, G., Programmed cell death in cortical chick embryo astrocytes is associated with activation of protein kinase PK60 and ceramide formation. J. Neurochem., 1998, 70, 130±138. Mangoura, D., m-Opioid peptides activate FAK and cortical cytoskeleton proteins cortactin and vinculin in chick embryonic neurons. J. Neurosci. Res., 1997, 50, 391±401. Mangoura, D., Jones, J., Sakellaridis, N. and Vernadakis, A., Early and late passage C-6 glial cell growth: similarities with primary glial cells in culture. Neurochem. Res., 1989, 14, 941± 947. Mangoura, D., Sogos, V. and Dawson, G., Protein kinase C±E is a developmentally regulated neuronal isoform in the chick embryo central nervous system. J. Neurosci Res., 1993, 35, 488± 498. Mangoura, D., Sogos, V. and Dawson, G., Phorbol ester and PKC signaling regulate proliferation vimentin cytoskeleton assembly and glutamine synthetase activity of chick embryo cerebrum astrocytes in culture. Brain Res., 1995, 87, 1±11. Mangoura, D., Sogos, V., Pelletiere, C. and Dawson, G., Di€erential regulation of phospholipases C and D by phorbol esters and the physiological activators carbachol and glutamate in astrocytes from chick embryo cerebrum and cerebellum. Brain Res., 1995, 87, 12±21. Mangoura, D., Pelletiere, C., Wang, D., Sakellaridis, N. and Sogos, V., Plasticity in astrocytic phenotypes. A role for protein kinase C, tyrosine kinases, and cytoskeleton signaling. Adv. Exp. Med. Biol., 1997, 429, 39±53. Martinez-Hemandez, A., Bell, K. P. and Norenberg, M. D., Glutamine±synthetase±glial localization in the brain. Science, 1977, 195, 1356±1358. Nishizawa, K., Yano, T., Shibata, M., Ando, S., Saga, S., Takahashi, T. and Inagaki, M., Speci®c localization of phos-

[38]

[39] [40] [41]

[42] [43]

[44] [45]

[46] [47]

[48]

phointermediate ®lament protein in the constricted area of dividing cells. J. Biol. Chem., 1991, 15;266(5), 3074±3079. Ohkubo, T., Tanaka, M., Nakashima, K., Talbot, R. T. and Sharp, P. J., Prolactin receptor gene expression in the brain and peripheral tissues in broody and nonbroody breeds of domestic hen. Gen. Comp. Endocrino., 1998, 109(1), 60±68. Pixley, S. R. K. and De Vellis, J., Transition between immature glia and mature astrocytes studied with a monoclonal antibody to vimentin. Dev. Brain Res., 1984, 15, 201±209. Rajan, P. and McKay, R. D., Multiple routes to astrocytic di€erentiation in the CNS. J. Neurosci., 1998, 18, 3620±3629. Rillema, J. A., Campbell, G. S., Lawson, D. M. and Carter-Su, C., Evidence for a rapid stimulation of tyrosine kinase activity by prolactin in Nb2 rat lymphoma cells. Endocrinology, 1992, 131, 973±975. Sakellaridis, N., Bau, D., Mangoura, D. and Vernadakis, A., Developmental pro®les of glial enzymes in the chick embryo: in vivo and in culture. Neurochem. Intern., 1983, 5, 685±689. Sakellaridis, N., Mangoura, D. and Vernadakis, A., E€ects of neuron conditioned medium and fetal calf serum content on glial growth in dissociated cultures. Devl. Brain Res., 1986, 27, 31±41. Shuai, K., Stark, G. R., Kerr, I. M. and Darnell, J. E. Jr, A single phosphotyrosine residue of Stat91 required for gene activation by interferon-gamma. Science, 1993, 261, 1744±1746. Tanaka, M., Maeda, K., Okubo, T. and Nakashima, K., Double antenna structure of chicken prolactin receptor deduced from the cDNA sequence. Biochem. Biophys. Res. Commun., 1992, 30;188(2), 490±496. Tsung-Chang, S., Yelena, M., Altshuller, A. J., Morris, J. and Frohman, M. A., Molecular analysis of mammalian phospholipase D2. J. Biol. Chem., 1999, 274(Issue 1), 494±502. Waters, M. J., Daniel, N., Bignon, C. and Djiane, J., The rabbit mammary gland prolactin receptor is tyrosine-phosphorylated in response to prolactin in vivo and in vitro. J. Biol. Chem., 1995, 270, 5136±5143. Wilkes, L. C., Patel, V., Purkiss, J. R. and Boarder, M. R., Endothelin-1 stimulated phospholipase D in A10 vascular smooth muscle derived cells is dependent on tyrosine kinase. Evidence for involvement in stimulation of mitogenesis. FEBS Lett., 1993, 322, 147±150.