MAP Kinase Mediates Epidermal Growth Factor- and Phorbol Ester-induced Prostacyclin Formation in Cardiomyocytes

J Mol Cell Cardiol 30, 933–945 (1998)

MAP Kinase Mediates Epidermal Growth Factor- and Phorbol Ester-induced Prostacyclin Formation in Cardiomyocytes Silvia Braconi Quintaje, Michela Rebsamen, Dennis J. Church1, Michel B. Vallotton and Ursula Lang Division of Endocrinology, University Hospital, CH-1211 Geneva 14, Switzerland, and 1 Serono, Geneva Pharmaceutical Research Institute, Geneva, Switzerland (Received 18 July 1997, accepted in revised form 3 February 1998) S. B. Q, M. R, D. J. C, M. B. V  U. L. MAP Kinase Mediates Epidermal Growth Factor- and Phorbol Ester-induced Prostacyclin Formation in Cardiomyocytes. Journal of Molecular and Cellular Cardiology (1998) 30, 933–945. We studied the role of protein kinase C (PKC) and mitogen-activated protein kinase (MAPK) in epidermal growth factor (EGF)-induced prostacyclin (PGI2) production in cultured, spontaneously-beating neonatal ventricular rat cardiomyocytes. To this purpose, the effect of EGF on cardiomyocyte MAPK phosphorylation, MAPK activity and PGI2-production were investigated, and compared to those induced by the PKC activator 4b phorbol 12-myristate 13-acetate (PMA). Both EGF (0.1 l) and PMA (0.1 l) induced the rapid and reversible phosphorylation of 42 KDa-MAPK in ventricular cardiomyocytes, responses that were accompanied by transient increases in MAPK activity (190–230% of control values within 5 min), and two- to three-fold increases in PGI2 formation. The tyrosine kinase inhibitors lavendustin (1 l) and genistein (10 l) strongly inhibited EGF-induced MAPK activation and PGI2-formation, but had no effect on PMA-stimulated responses. Experiments with the PKC inhibitor CGP 41251 (1 l) or with PKC-downregulated cells demonstrated that in contrast to the PMA-stimulated responses, EGF-induced MAPK activation and PGI2-production were PKCindependent processes. Investigating the role of MAPK in EGF- and in PMA-promoted PGI2-formation, we found that the MAPK-inhibitor 6-thioguanine (500 l), as well as the MAPK-kinase-inhibitor PD98059 (50 l) abolished both EGF- and PMA-stimulated PGI2-production in cardiomyocytes. Our results indicate that MAPKactivation is at the basis of both growth factor receptor and PKC-dependent eicosanoid-formation in ventricular cardiomyocytes, where EGF-induced prostaglandin-production takes place via a PKC-independent pathway.  1998 Academic Press Limited

K W: Ventricular cardiomyocyte; Epidermal growth factor; Protein kinase C; MAP kinase; Prostacyclin.

Introduction The mammalian heart synthesizes and secretes numerous cyclooxygenase products (Van Bilsen et al., 1989; Berti et al., 1990; Church et al., 1993), notably prostacyclin (PGI2), prostaglandin E2 (PGE2) and prostaglandin F2a (PGF2a ), eicosanoids that are potent auto- and paracrine modulators of biological function noted for their mitogenic

and secretory effects. At the level of the cardiomyocyte itself, it appears that both protein kinase C (PKC) and phospholipase A2 (PLA2) play major roles in eicosanoid formation induced by agonists acting at cell surface receptors functionally coupled to phospholipase C (PLC), insofar as PKC-activating phorbol diesters promote substantial prostaglandin production on their own (Church et al., 1993), while PKC inhibition, PKC

Please address all correspondence to: Dr Ursula Lang, Division of Endocrinology, University Hospital, CH-1211 Geneva 14, Switzerland.

0022–2828/98/050933+13 $25.00/0

mc980657

 1998 Academic Press Limited

934

S. B. Quintaje et al.

downregulation and PLA2 inhibition abolish the near totality of eicosanoid formation induced by the vasoactive humoral agents angiotensin II, vasopressin and substance P (Church et al., 1994a, 1996; Van der Bent et al., 1994). Relatively little is known concerning the role of PKC in myocardial prostaglandin formation induced by growth factors, another class of humoral agents noted for their PKC and eicosanoid production activating effects. Among these is the epidermal growth factor (EGF), a polypeptide that modulates cell growth and differentiation by binding to its cell surface receptor (EGF-R), which is a tyrosine kinase (Spaargaren et al., 1991; Boonstra et al., 1995) and is present in ventricular cardiomyocytes (Rabkin et al., 1987; Yu et al., 1992). There are potentially two distinct signalling pathways which may control EGF-induced eicosanoid formation (Nemenoff et al., 1993). On the one hand, the binding of EGF to EGF-R has been shown to lead to the activation of the GTP-binding protein Ras (Ullrich and Schlessinger, 1990), a response which in turn leads to the activation of mitogen-activated protein kinase (MAPK) via the intermediate actions of the kinases Raf and MAPK kinase (MEK) (Kyriakis et al., 1992; Lange-Carter et al., 1993). It is well established that MAPK represents a family of seryl/threonyl kinases that play an important role in regulating cell growth following exposure to mitogenic stimuli (Cobb et al., 1991), and it has recently been demonstrated that MAPK activation leads to the activation of cytosolic PLA2 in cultured cells, an event dependent on the MAPK-induced phosphorylation of this enzyme (Lin et al., 1992, 1993; Qiu and Leslie, 1994; Xing and Insel, 1996). In this view, it is likely that Ras-mediated MAPK activation plays a major role in the growth factor-induced generation of eicosanoids in a variety of cell types, and that a similar signalling pathway may also exist in cardiomyocytes. Nevertheless, it has also been shown that the activation of EGF-R leads to the activation of PLC c (Margolis et al., 1989), an event contributing to enhanced PKC activity induced by the production of phosphatidylinositol bisphosphate-derived diacylglycerols (Bonventre et al., 1990; Boonstra et al., 1995). Through this route, PKC activation would lead either to MAPK activation via a MEKdependent pathway (Lange-Carter et al., 1993), which subsequently leads to phospholipid hydrolysis and prostaglandin formation as per the above, or alternatively, to the direct PKC-mediated phosphorylation and activation of PLA2, as previously shown in vitro (Lin et al., 1993; Nemenoff et al., 1993).

Further complicating the matter, previous studies have shown that the PLA2 believed to be at the basis of prostaglandin formation in the myocardium is distinctly different from the 100 KDa cytosolic PLA2 studied in other cell lines (Lin et al., 1992, 1993; Qiu and Leslie, 1994; Xing and Insel, 1996). Cardiac PLA2 is smaller (40 KDa), and is activated by intracellular ATP as opposed to increases in cytosolic free Ca2+, making it one of the few calciuminsensitive PLA2s discovered to date (Hazen et al., 1990). In this context, it is therefore difficult to foresee whether the Ras-dependent or the PKCdependent signalling pathway is involved in growth factor-induced prostaglandin formation in cardiomyocytes, as it is currently unknown whether cardiac PLA2 is even activated by MAPK. In view of the interest in understanding the biological effects of growth factors in the myocardium, where they are suspected to play an underlying role in the events leading to cardiac hypertrophy, we investigated the intracellular mechanism(s) by which EGF promotes PGI2 formation in neonatal rat ventricular cardiomyocytes. We further studied the role of MAPK in EGF- as well as in PMA-promoted myocardial PGI2 production.

Materials and Methods Materials [32P]-orthophosphate, [c-32P]-ATP, [3H]-arachidonic acid, and 6-keto-[3H] PGF1a were purchased from Amersham International (Amersham, Bucks, UK). Anti-6-keto-[3H] PGF1a antiserum was obtained from the Cayman Chemical Company (Ann Arbor, USA), 4b phorbol 12-myristate 13-acetate (PMA), 6-thioguanine, bovine serum albumin (BSA), c globulin, myelin basic protein (MBP), phenylmethylsulfonyl fluoride (PMSF), dithiothreitol (DTT), cAMP-dependent protein kinase inhibitor (PKI), leupeptin, aprotinin, pepstatin, histone IIIS (HIIIS), DNase, insulin/transferrin/sodium selenite medium supplement (ITS), and iodoacetamide were purchased from Sigma (St Louis, MO, USA). Histone IIB was from Boehringer Mannheim (Germany), while genistein and lavendustin were obtained from Anawa (Wangen. Switzerland). McCoys modified 5A medium. Hank’s Ca2+–Mg2+-free balanced salt solution (HBSS) and trypsin were purchased from Gibco (Basel, Switzerland). Penicillin and streptomycin were obtained from Hoechst (Frankfurt, Germany) and Grunenthal (Glarus, Switzerland),

MAP Kinase-mediated PGI2 Formation in Cardiomyocytes

respectively. Protein A sepharose CL-4B was obtained from Pharmacia LKB Biotechnology AB (Upsala, Sweden). Protein G-plus/protein A-agarose was purchased from Lucerna Chem AG (Oncogene Science, Lucerne, Switzerland). Fetal calf serum (FCS) was from Fakola (Basel, Switzerland), and DMEM culture medium was purchased from Life Technologies (Basel, Switzerland). Epidermal growth factor (EGF) was obtained from Inotech (Dottikon, Switzerland) and PD98059 from LC Laboratories (Woburn, MA, USA). The PKC inhibitor CGP-41251 was a kind gift from Ciba-Geigy (Basel, Switzerland). Anti-42-KDa and 44-KDa MAP kinase antisera were generously donated by Dr M. J. Dunn (Division of Nephrology, Case Western Reserve University, Cleveland, USA).

Culture of ventricular cardiomyocytes Spontaneously-beating neonatal rat ventricular cardiomyocyte cultures were obtained from 1- to 2-day-old Wistar rats using the trypsin/DNase sequential-digestion method previously described (Van der Bent et al., 1994). Briefly, 20–40 neonatal rat heart ventricles (lower two-thirds of tissue) were excised from decapitated rats and placed in 40 ml ice-cold sterile Ca2+- and Mg2+-free HBSS, containing 100 IU/ml penicillin and 10 lg/ml streptomycin. The tissue was washed with 40 ml HBSS, cut into small pieces, further washed with 40 ml of HBSS, and enzymically digested for 8 min with 10 ml of trypsin-DNase HBSS solution (2.5 and 0.03 mg/ml, respectively) at 37°C in a 50 ml sterile conical tube subjected to constant stirring. The supernatant from the first incubation was discarded, 10 ml of fresh enzyme solution were added, and the incubation procedure was repeated to total digestion of the tissue. Subsequent supernatant volumes were collected and centrifuged at 200×g for 5 min, and the resulting cell pellets were resuspended in McCoy’s modified 5A medium containing 10% FCS and 1% ITS medium supplement at 37°C. Once the sequential digestions were terminated, the cells were pooled, washed in 40 ml of McCoy’s modified 5A medium containing 10% FCS, 1% ITS, 100 IU/ml penicillin, and 10 lg/ ml streptomycin, and seeded in 90-mm plastic Petri dishes. After 3 h of incubation at 37°C, the dishes were shaken and the cardiomyocyte-containing supernatants were pooled and seeded in 90-mm Petri dishes or six-wells culture plates (Costar, Cambridge; 4–5×106 cells/Petri dish or plate). Cells were kept in culture in McCoy’s modified 5A medium containing 10% FCS, 1% ITS, 100 IU/ml

935

penicillin, 10 lg/ml streptomycin, and 0.5 lg/ml fungizone in an H2O-saturated air-CO2 (19:1) atmosphere at 37°C. The majority of cultured cells (i.e. >90%) began to contract spontaneously within 24–48 h of plating (30–60 beats/min), and exhibited positive staining for pro-ANP (Van der Bent et al., 1994), the precursor form of rat ANP ([Ile12]a-hANP) and a specific cardiomyocyte marker. Confluent, spontaneously-beating cells were used on the third day of culture for all experiments described herein.

Culture of ventricular fibroblasts Cultures of ventricular fibroblasts were prepared from 1–2-day-old Wistar rats according to the modified method of Cao and Gardner (1995). Briefly, after sequential digestions of neonatal ventricular tissue, cells were resuspended as described above and plated in 90-mm Petri dishes or six-wells culture plates for 90 min to allow for selective adhesion of cardiac fibroblasts (Sadoshima and Izumo, 1993; Cao and Gardner, 1995). Thereafter, myocardial cells were decanted from the plates and cardiac fibroblasts were cultured in McCoy’s modified 5A medium containing 10% FCS, 1% ITS, 100 IU/ml penicillin, 10 lg/ml streptomycin and 0.5 lg/ml fungizone. Confluent monolayers were used for investigating prostacyclin production, and/or were passaged using 0.05% porcine trypsin and 0.53 m EDTA. For our study, we used newly cultured monolayers of cardiac fibroblasts as well as confluent cultures from the first passage.

Determination of prostacyclin production and arachidonate release For assessment of PGI2 formation, six-well tissue culture plates containing spontaneously-contracting cardiomyocyte monolayers were washed with 2 ml Krebs–Ringer buffer containing 0.2% bovine serum albumin (BSA) and 0.2% glucose as previously described (Church et al., 1993). After replacing the supernatant with 1ml fresh buffer, the cells were incubated at 37°C for 60 min in the presence of various pharmacological agents, and 50 ll aliquots of the supernatants were collected and assayed for 6-keto-PGF1a, the stable metabolite of PGI2, according to a radio-immunological method already described (Church et al., 1993). Prior to stimulation, cardiomyocyte monolayers were kept in culture medium containing 10% FCS, since in cells which had been kept for 24 h in serum-free

936

S. B. Quintaje et al.

medium, basal PGI2-production was found to be below the detection limit and agonist-induced PGI2 responses to be very weak when compared to cells kept in serum-containing medium (data not shown). Agents used in these studies were dissolved in either water, DMSO, or mixes thereof, such that the final concentration of DMSO did not exceed 0.5%. When tested for 6-keto-PGF1a inhibiting or stimulating properties, 1% concentrations of DMSO induced no response in these preparations (data not shown). Relative affinity of the 6-keto-PGF1a antiserum was determined at 14% for PGF1a, 2% for PGE1, 1.7% for PGF2a, 0.5% for 6-keto-PGE1 and PGE2, and less than 0.05% for other major arachidonic acid metabolites. Non-specific binding was estimated at 2.9±0.11% (n=10). The detection limit for 6-keto-PGF1a determinations was 3 pg/ml of incubation medium, while the intra- and interassay coefficients of variation were estimated at 4 and 6%, respectively (n=10). Arachidonic acid (AA) release was assessed in sixwell tissue culture plates containing spontaneouslycontracting cardiomyocyte monolayers that were incubated for 16 h with 0.3–0.4 lCi/ml [3H]-AA in culture medium containing 1% FCS. At the end of the labelling period, cells were washed three times with Krebs–Ringer buffer containing 0.2% bovine serum albumin (BSA) and 0.2% glucose. After replacing the supernatant with 1 ml fresh buffer, the cells were incubated at 37°C for various times in the presence or absence of PLA2 or MAPK inhibitors, and aliquots of the incubation media were counted in a liquid scintillation counter as an index of arachidonate release.

Determination of MAPK, PKC and PKA activities In order to study MAPK activation under the same experimental conditions as PGI2 formation, ventricular cardiomyocytes were cultured and stimulated as described in the section determination of PGI2-production. MAPK activity determinations were carried out by homogenizing treated cardiomyocyte monolayers in lysis buffer containing 50 m Tris/HCl, 1% Triton X-100, 150 m NaCl, 10% glycerol, 2 m EDTA, 2 m EGTA, 40 m b-glycerophosphate, 50 m NaF, 10 m sodium pyrophosphate, 200 l Na3VO4, 0.3 m leupeptin, 1 l pepstatin A, 1 m PMSF and 100 n okadaic acid at pH 7.4. The homogenates were centrifuged at 100 000×g for 1 h at 4°C, and the resulting supernatants were assayed for MAP kinase activity by measuring the transfer of 32P from [c-32P] ATP to myelin basic protein (MBP) as described by Wang

et al. (1992). The final reaction mixture (160 ll) contained 1–2×106 ct/min of [c-32P] ATP, 8 l ATP, 0.25 m EDTA, 1.25 m MgCl2, 0.125 m DTT, 3.12 m b-glycerophosphate, 10 m Tris/ HCl, 215 lg/ml MBP and 3.12 lg/ml cAMP-dependent protein kinase inhibitor (PKI). Following a 10-min incubation at 30°C, 120 ll aliquots of each reaction mixture were spotted onto 2.5-cm2 Whatman P18 phosphocellulose paper filters, followed by immediate addition of 4 ml washing buffer (20% w/v TCA and 75 m pyrophosphate). After 5–10 min, free [c-32P]-ATP was removed by four washes with the same solution. 32P-incorporation was determined by liquid scintillation counting. PKC activity in ventricular heart tissue was determined by semipurifying cytosolic PKC using DEAE cellulose chromatography, as described by Braconi et al. (1992). Eluates were assessed for PKC activity by measuring the incorporation of 32P from [c-32P]-ATP into histone III-S in the absence and in the presence of lipid activators, as previously indicated (Braconi et al., 1992). PKA activity was assessed the same way, except that histone IIB was used as substrate, and the phosphorylation reaction was performed in the absence and in the presence of 10 l cAMP.

MAP kinase phosphorylation and immunoprecipitation Spontaneously-beating neonatal rat cardiomyocytes were washed and incubated for 2 h in phosphate-free DMEM culture medium in the presence of [32P]-orthophosphate (0.5–1 mCi/Petri), providing steady-state labelling of endogenous phosphopeptides. Thereafter, the cells were exposed to agonists for the indicated periods of time. The reaction was stopped by removing the medium from the cells, and by washing the monolayers twice with buffer containing 50 m Tris/HCl and 150 m NaCl at pH 7.5. Cardiomyocytes were subsequently scraped into 1 ml of MAP kinase lysis buffer containing 1% Triton X-100 (see above), and homogenized by 30 passes through a 26-gauge needle fitted to a 1-ml syringe. The homogenates were centrifuged for 10 min at 14 000×g, and the resulting supernatants used for immunoprecipitation. Immunoprecipitation of phosphorylated MAPK was achieved by incubating cell extracts containing 150×106 ct/min of labelled protein, 5% of fetal calf serum, and 1.5 m iodoacetamide with either anti-42-KDa or 44-KDa MAP kinase antisera (Wang et al., 1992) for 3 h at room temperature, or overnight at 4°C. Thereafter, 100 ll of a 50% slurry of protein A-Sepharose CL-4B and/or 15 ll protein G

937

MAP Kinase-mediated PGI2 Formation in Cardiomyocytes

Statistical analysis Student–Fisher unpaired bilateral t-tests and/or ANOVA using the Sheffe F-test criterion for unbalanced groups were used where applicable. A value of P<0.05 was accepted as statistically significant. Results represent the means±.. of at least three experiments performed in duplicate or triplicate determinations.

(a)

6-keto-PGF1 (% of control)

350

EGF- and/or PMA-induced PGI2 formation in ventricular cardiomyocytes Incubation of spontaneously-beating neonatal rat ventricular cardiomyocytes with 0.1 l EGF or 0.1 l PMA led to marked increases in cellular PGI2-formation, as assessed by the determination of its stable metabolite 6-keto-PGF1a. The effects were apparent within 15 min of applying EGF or PMA to the cell monolayers [Fig. 1(b)], with EGF and PMA inducing maximal PGI2-production at 60 min (from 205±29 to 623±39 and to 568±58 pg/mg cell protein, respectively, n=4–9), and exhibiting half maximal effective concentrations (EC50) in the range of 4–8 n [n=5–9, Fig. 1(a)]. Interestingly, no significant additive effect was observed when cells were incubated with both stimuli at the same time for 60 min at various concentrations [Fig. 1(a)] or with maximal concentrations (0.1 l) for different times [Fig. 1(b)].

*

300

* 250 200

*

*

*

*

*

*

*

*

150

50

0

10–8 10–9 Agonist concentration (M)

10–7

(b)

0.80

* * *

0.60

* *

*

* 0.40

* *

0.20

0

Results

*

100

6-keto-PGF1 (ng/mg cell prot.)

plus protein A agarose in lysis buffer were added to each sample, and the mixture was incubated for 90 min at room termperature under mild shaking. After centrifugation for 5 min at 3000×g, immunocomplexes were washed three times with 1 ml of low salt buffer (50 m Tris/HCl, 150 m NaCl, 0.2% Triton X-100, 2 m EDTA, 2 m EGTA and 0.1% SDS at pH 7.5), three times with 1 ml of a high salt buffer (50 m Tris/HCl, 500 m NaCl, 0.2% Triton X-100, 2 m EDTA, 2 m EGTA and 0.1% SDS at pH 7.5), and once with 1 ml of 10 m Tris/HCl at pH 7.5. Pellets were boiled for 5 min in 50 ll Laemmli dissociation buffer, and subjected to SDS-PAGE (10%). After fixing in 25% isopropanol and 10% acetic acid, the gels were dried and exposed to XOMat AR films at −70°C. Analysis of the radiographic results were performed at non-saturing exposures by means of a densitometer (Molecular Dynamics).

30

60 Time (min)

90

120

Figure 1 Concentration- and time-dependent effects of EGF and/or PMA on PGI2 production in ventricular rat cardiomyocytes. Spontaneously-beating cultured cardiomyocytes were incubated at 37°C with varying concentrations of EGF and/or PMA for 60 min (a) or with 0.1 l EGF and/or 0.1 l PMA for the indicated times (b), and 6-keto-PGF1a release was determined as described in the Materials and Methods section. (Β) Control; (Ε) EGF; (Χ) PMA; (Μ) EGF plus PMA. Results represent the means±.. of 4–9 experiments performed in triplicate determinations. ∗P<0.05 v corresponding control.

Likewise, the effects of EGF and PMA were not additive when cardiomyocytes were exposed for only 30 min at submaximal concentrations (0.01 l) of EGF and PMA [Fig. 2(a), n=5–8]. Although previous work from our laboratory indicates that cultures of spontaneously-beating neonatal ventricular cardiomyocytes consist to approximately 90% of cardiomyocytes (Church et al., 1994b; Van der Bent et al., 1994), purity of cell monolayers remains a problem in studies using

938

S. B. Quintaje et al.

0.8

*

(a)

(b)

(a) *

0.6

0.4 * *

6-keto-PGF1 (ng/mg cell prot.)

*

*

0.2

0.0

Cardiomyocytes Cardiomyocytes

Cardiac fibroblasts

Figure 2 EGF- and/or PMA-induced PGI2 production in ventricular fibroblasts and cardiomyocytes. (a) Cultured cardiomyocytes were incubated with 0.01 l EGF (∆), with 0.01 l PMA (Γ) or with both (Ε) at 37°C for 30 min (n=5–8). (b) Cultured ventricular fibroblasts and cardiomyocytes were exposed to 0.1 l EGF, and/or 0.1 l PMA at 37°C for 60 min and 6-keto-PGF1a formation was determined as described in the Materials and Methods section. Results represent the means±.. of 4–6 experiments performed in triplicate determinations. ∗P<0.05 v control (Φ).

primary cultures. For this reason, we also tested the effects of EGF and PMA on PGI2-production in cardiac fibroblasts. As shown in Figure 2(b), PGI2formation in cardiac fibroblasts was much lower than that in cardiomyocytes. In contrast to cardiomyocytes, incubation of cardiac fibroblasts with 0.1 l EGF and/or 0.1 l PMA for 60 min, did not affect basal PGI2-production [Fig. 2(b)], indicating that EGF- and PMA-induced PGI2-production in monolayers of cardiomyocytes are indeed due to cardiomyocytes and not to cardiac fibroblasts. Our observations suggest that EGF- and PMAinduced signalling cascades share a common step in prostacyclin-formation in cardiomyocytes, perhaps that of PKC-activation. We investigated this hypothesis by measuring EGF- and PMA-induced PGI2formation in the presence of the tyrosine kinaseinhibitors lavendustin and genistein, and compared results with those obtained either in the presence of the PKC-inhibitor CGP 41251 (Meyer et al., 1989), or in PKC downregulated cells. As shown in Figure 3(a), incubation of cardiomyocytes with either 1 l lavendustin or 10l genistein had no effect on basal PGI2-production, yet fully abolished PGI2-formation induced by 0.1 l EGF (86±7 and 93±9% inhibition, respectively, n=5), and this without affecting the response induced by 0.1 l PMA.

* 0.6

* *

0.4

0.2

0

Basal

(b) 0.8 6-keto-PGF1 (ng/mg cell prot.)

6-keto-PGF1 (ng/mg cell prot.)

*

0.6

Lavendustin

Genistein

*

*

CGP41251

24 h PMA

* *

0.4

0.2

0

Basal

Figure 3 Implication of tyrosine kinase and PKC in EGFand PMA-induced PGI2 production in cardiomyocytes. (a) Cultured cardiomyocytes were incubated with 0.1 l EGF (∆) or PMA (Ε) in the presence of either 1 l lavendustin or 10 l genistein at 37°C for 60 min, and 6-keto-PGF1a formation was determined as described in Materials and Methods. Results represent the means±.. of five experiments performed in triplicate determinations. ∗P<0.05 v control (Φ). (b) Cardiomyocytes were incubated wth 0.1 l EGF or PMA in the presence or absence of CGP 41251 (1 l) and 6-keto-PGF1a formation was determined. Alternatively, cells were pretreated with 0.1 l PMA for 24 h prior to stimulation with 0.1 l EGF or PMA for 60 min. Results represent the means±.. of 4–6 experiments performed in triplicate determinations. ∗P<0.05 v control.

Surprisingly, the presence of the PKC inhibitor CGP 41251 (1 l) as well as downregulation of cardiomyocyte PKC by prolonged pretreatment of cells with 0.1 l PMA, had no effect on EGFinduced PGI2-production, while the PMA-stimulated response was inhibited by 85±7 and 97±6%, respectively [Fig. 3(b), n=4–6]. Taken together, these results suggest that EGF-induced

939

MAP Kinase-mediated PGI2 Formation in Cardiomyocytes

EGF-induced activation and phosphorylation of MAP kinase We further investigated the effect of EGF on cardiac MAPK in comparison to that induced by PKCactivation. To this purpose, we incubated spontaneously-beating ventricular cardiomyocytes with 0.1 l EGF and 0.1 l PMA for 5–15 min, and tested for increased MAPK activity in cytosolic extracts by measuring phosphorylation of myelin basic protein in the presence of cAMP-dependent protein kinase inhibitor. As is shown in Figure 4(a), exposure of cardiomyocytes to 0.1 l EGF led to a marked transient increase in cytosolic MAPK activity, an effect which was found to be maximal at 5 min (205±15% of basal MAPK activity, n=6), yet which decreased to control levels thereafter (125±8% of basal values at 15 min, n=5). Incubation of cardiomyocytes with 0.1 l PMA had a similar effect, the phorbol ester transiently increasing cytosolic MAPK activity to 217±16% of basal cytosolic activity within 5 min of application [Fig. 4(a), n=8). These transient increases in MAPK activity induced by 0.1 l EGF and 0.1 l PMA were accompanied by transient increases in the phosphorylation level of cardiomyocyte 42-KDa MAPK. As shown by the autoradiograms in Figure 4(b), densitometric scanning of [32P]-orthophosphate-labelled, SDS PAGE-resolved cardiomyocyte extracts revealed that 5-min exposures of cells to either 0.1 l EGF or 0.1 l PMA increased the phosphorylation level of cellular 42-KDa MAPK to 345±50 and 474±65% of basal MAPK phosphorylation, respectively (n=3), responses that were transient in nature insofar as these signals respectively decreased to 228±36 and 316±44% of control values after 15 min of incubation (n= 3–4). Western blotting analysis and immunoprecipitation using an antibody directed against 44-KDa MAPK indicated that neonatal rat ventricular cardiomyocytes contain only very small amounts of 44-KDa MAPK, whose phosphorylation level(s) were practically undetectable under the experimental conditions used (data not shown). In view of the observation that tyrosine kinase inhibition abolishes EGF-induced PGI2 formation without affecting the PMA-induced response [Fig. 3(a)], and that EGF-stimulated PGI2-production is a PKC-independent process in cardiomyocytes [Fig.

(a) MAP kinase activity (% of control)

PGI2-formation is a tyrosine kinase-dependent process in ventricular cardiomyocytes, an event which does not require the concomitant activation of PKC.

225

* *

200

*

175 150 125 100 0

0

5

15

10 Time (min)

(b) EGF 42 kDa

PMA

42 kDa

0

5

15

Time (min)

Figure 4 Effect of EGF and PMA on cytosolic MAPK activity and on phosphorylation levels of 42-kDa-MAP kinase in cardiomyocytes. (a) Cells were stimulated with 0.1 l EGF or PMA for the indicated periods of time and MAPK activity was determined as described in experimental procedures. Addition of EGF (Ε); addition of PMA (Χ). The figure shows data pooled from 4–8 experiments. Mean values±.. (bars) were expressed as percentage of control values. ∗P<0.05 v control. (b) [32P]Orthophosphate-labelled cardiomyocytes were incubated for the indicated time periods with 0.1 l EGF or PMA. After stimulation cells were lysed and immunoprecipitated with a specific antibody against 42-kDa-MAP kinase, as described in experimental procedures. The immunoprecipitates were washed and subjected to SDSpolyacrylamide gel electrophoresis followed by autoradiography. Autoradiograms of dried gels are shown. Similar results were obtained in 3–4 separate experiments.

3(b)], we further tested the effect of both tyrosine kinase-inhibition and PKC downregulation on EGFinduced MAPK phosphorylation in these cells. As shown in Figure 5(a) incubation of [32P]-orthophosphate-labelled ventricular cardiomyocytes with 0.1 l EGF in the presence of 1 l lavendustin led to a marked inhibition of EGF-induced 42KDa MAPK phosphorylation. Interestingly, and as

940

S. B. Quintaje et al.

(a) 42 kDa

C

E E+L P P+L

(b)

42 kDa

C

P NT

C

P

24 hP

C

E NT

C

E

24 hP

Figure 5 Effects of tyrosine kinase inhibition and PKC downregulation on phosphorylation levels of 42-kDaMAP kinase in cardiomyocytes. (a) [32P]-Orthophosphatelabelled cardiomyocytes (see Materials and Methods section) were incubated for 10 min at 37°C with 0.1 l EGF (E) or PMA (P) in the presence or absence of 1 l lavendustin (L). (b) Cultured cells were pre-incubated with or without 0.1 l PMA for 24 h (24 h P or nontreated, NT), labelled with [32P]-orthophosphate and further stimulated with 0.1 l EGF (E) or PMA (P) for 10 min. After stimulation, cells were lysed and immunoprecipitated with a specific antibody against 42kDa-MAP kinase, as described in experimental procedures. The immunoprecipitates were washed and subjected to SDS-polyacrylamide gel electrophoresis followed by autoradiography. Autoradiograms of dried gels are shown. Similar results were obtained in 2–3 separate experiments.

previously seen for prostacyclin release, lavendustin had no effect on PMA-induced 42-KDa MAPK phosphorylation under the same conditions [Fig. 5(a)], while downregulation of cardiomyocyte PKC led to a substantial reduction in PMA-induced 42-KDa MAPK phosphorylation without affecting EGF-induced 42-KDa MAPK phosphorylation [Fig. 5(b)].

MAPK inhibition and PGI2-production in ventricular cardiomyocytes The high degree of correlation between EGF- and PMA-induced PGI2-formation and increases in MAPK-activity and -phosphorylation suggests that MAPK activation is at the basis of EGF- and PMAinduced PGI2-production in cardiomyocytes. In order to investigate this hypothesis, we incubated ventricular cardiomyocytes with 0.1 l EGF in the presence of the MAPK-inhibiting purine analogs 6thioguanine and 2-aminopurine, compounds that

MAP kinase activity (% of control)

300 * 200

*

100 * 0

Basal

*

*

6-Thioguanine

Figure 6 Inhibitory effect of 6-thioguanine on MAPK activity in cultured cardiomyocytes. Cultured cardiomyocytes were pre-incubated for 30 min with or without 0.5 m 6-thioguanine and further stimulated with 0.1 l EGF (∆) or PMA (Ε) for 5 min at 37°C in the presence or absence of 0.5 m 6-thioguanine. MAPK activity was determined as described in experimental procedures. The figure shows data pooled from 3–4 experiments. Mean values±.. (bars) were expressed as percentage of control (Φ) values. ∗P<0.05 v basal values in untreated cells.

have previously been shown to suppress phenylepinephrine-induced MAPK activation and gene expression in these same cells (Thorburn et al., 1994). As shown in Figure 6, incubation of cardiomyocytes with 500 l 6-thioguanine reduced basal cytosolic MAPK activity by 63±7%, and fully abolished the increases in cardiomyocyte MAPK activity induced by exposing cells to either 0.1 l EGF or 0.1 l PMA for 5 min (n=3–4). Similar results were obtained in some instances with 2aminopurine, although this compound’s effects varied considerably from one experiment to another (data not shown). As shown in Figure 7(a), the inhibitory effect of 6-thioguanine was found to be selective for MAPK over both PKC and cAMP-dependent kinase (PKA), insofar as a concentration which fully-inhibited cardiomyocyte MAPK activity (1 m) had no effect on cardiomyocyte PKC or PKA activity. Along the same lines, and contrary to the documented phospholipase A2 and cyclooxygenase inhibitors quinacrine (10 l) and indomethacin (10 l), 500 l 6-thioguanine was ineffective at inhibiting either cardiomyocyte phospholipase A2 [Fig. 7(b)], or the conversion of arachidonic acid into PGI2 in these preparations [Fig. 7(c)]. In view of the apparent specific inhibitory action of 6-thioguanine on MAPK activity, we finally tested

941

MAP Kinase-mediated PGI2 Formation in Cardiomyocytes 1.00 6-keto-PGF1 (ng/mg cell prot.)

Protein kinase activity (% of control)

(a) 120 100 80 60 *

40

*

20 0 0.0001

* * *

0.001

0.01

0.1

10

1

*

0.80 * 0.60

0.40

0.20

0.00

Basal

6-Thioguanine

PD98059

6-Thioguanine (mM)

3

[ H] Arachidonic acid release (ct/min/well)

(b)

Figure 8. Effect of MAPK inhibition on EGF- and PMAinduced PGI2 production. Cultured cardiomyocytes were pre-incubated for 30 min with or without 500 l 6thioguanine or 50 l PD98059. Thereafter, cells were stimulated with 0.1 l EGF (∆) or PMA (Ε) for 60 min at 37°C in the presence or absence of 6-thioguanine (500 l) or PD98059 (50 l), and 6-keto-PGF1a formation was determined as described in the Materials and Methods section. Results represent the means±.. of four experiments performed in triplicate determinations. ∗P<0.05 v control (Φ).

2000

1000 *

0.0

Control

Quin

6-TG

(c)

6-keto-PGF1 (ng/mg cell prot.)

1.2 * * 0.8

0.4

0.0

Basal

Indo

6-TG

Figure 7 6-Thioguanine specifically inhibits MAPK. (a) Effects of 6-thioguanine on MAPK, PKC and PKA activities. Cytosolic cell extracts from cardiomyocytes were incubated with myelin basic protein (MBP) as substrate in the presence of 0.5 l to 1 m 6-thioguanine for 10 min at 30°C, and MAPK activity (Χ) was determined. Cytosolic, semi-purified PKC and PKA preparations from ventricular rat heart tissue (see experimental procedures) were incubated with 0.1–1.0 m 6-thioguanine for 10 min at 30° C, and PKC (Φ) and PKA (Μ) activities were assessed. Results represent the means±.. of 4–7

the effect of this purine analog on EGF- and PMAinduced PGI2-formation. As shown in Figure 8, inhibition of cardiomyocyte MAPK by 500 l 6thioguanine had no effect on basal PGI2-formation, but completely suppressed PGI2-production induced by both 0.1 l EGF and 0.1 l PMA (100 and 96% inhibition, respectively, n=4). To further confirm the role of MAPK in EGF- and PMA-induced PGI2-formation, we also tested the effect of the recently described MAPK kinase (MEK) inhibitor PD98059 (Alessi et al., 1995) in myocardial PGI2production. As shown in Figure 8, 50 l PD98059 fully abolished the EGF- and PMA-induced PGI2 response, while inhibiting basal PGI2-production by experiments performed in duplicate determinations. ∗P<0.05 v control. (b) Effect of 6-thiogaunine on PLA2 activity in cardiomyocytes. [3H]-AA-labelled cardiomyocytes were incubated for 30 min with 10 l quinacrine (Quin) or 0.5 m 6-thioguanine (6-TG) and AAliberation was measured in the supernatant (see experimental procedures). Results represent the means±.. of three experiments performed in triplicate determinations. ∗P<0.05 v control. (c) Effect of 6-thioguanine on cyclooxygenase activity in cardiomyocytes. Cardiomyocytes were incubated for 60 min with 5 l AA (Γ) in the presence or absence of 10 l indomethacin (Indo) or 0.5 m 6-thioguanine (6-TG), and PGI2 production was determined. Results represent the means±.. of four experiments performed in triplicate determinations. ∗P<0.05 v control (Φ).

942

S. B. Quintaje et al.

48±6% (n=4). In connection with the above, these last results indicate that both EGF and PMA promote PGI2-formation in cardiomyocytes via the MAPK-dependent activation of a cellular phospholipase, a process which occurs independently of PKC-activation for EGF.

Discussion The present study demonstrates that ventricular cardiomyocytes clearly represent a target cell type for EGF action as far as MAPK activation and prostacyclin formation are concerned. Sustaining this observation, EGF promotes transient MAPK activation in these cells, a response accompanied by tyrosine kinase-dependent MAPK phosphorylation and enhanced prostacyclin production. While the PKC activating phorbol diester PMA superficially mimicks EGF action by promoting similar MAPK and prostaglandin-production-activating effects in this system, experiments conducted in the presence of the PKC inhibitor CGP41251 (Meyer et al., 1989) and in PKC-downregulated cells indicate that EGF induces PGI2-formation in a PKC-independent manner in these cells. However, it has previously been shown that PKC activation is a potent stimulus for PGI2-production in these cells, to the point that the activation of PKC appears to be the key step in prostaglandin-generation induced by the PLC-activating agonists angiotensin II, Substance P and vasopressin (Church et al., 1994a, 1996; Van der Bent et al., 1994). In connection with the observation that PKC-activation can be triggered by growth factors acting through tyrosine kinase-receptors that are functionally coupled to PLCc (Margolis et al., 1989; Lange-Carter et al., 1993), it could readily be surmised that EGF-induced PGI2formation is a PKC-dependent event in ventricular myocytes. Arguing against this hypothesis, however, we found that EGF-induced PGI2-response was unaffected by PKC inhibition, in these cells, clearly demonstrating that PKC-activation is not required for EGF-induced prostaglandin-formation in ventricular cardiomyocytes, which is in agreement with the observation that EGF activates PLA2 through a PKC-independent pathway in mesangial cells (Bonventre et al., 1990). Indeed, recent reports have shown that while PLA2 is a direct substrate for MAPK and PKC in vitro (Lin et al., 1993; Nemenoff et al., 1993), there is reason to believe that MAPK is the endogenous activator of PLA2 in vivo, as the phosphorylation of PLA2 by PKC only results in a

modest activation of the enzyme (Nemenoff et al., 1993). In this study, we showed that in cardiomyocytes both EGF and PMA induce rapid transient increases in MAPK activity. While other agents, such as endothelin-1, fibroblast growth factor and angiotensin II, have been shown to induce similar transient increases in cardiac MAPK activity (Bogoyevitch et al., 1994; Sadoshima et al., 1995), PMA was found to cause a more sustained increase in MAPK activity in ventricular cardiomyocytes which had been cultured for 24–48 h in serumfree conditions (Bogoyevitch et al., 1994; Sadoshima et al., 1995). This discrepancy in the kinetics of PMA-stimulated MAPK activity could be explained by different experimental conditions. Indeed, we found that in cells which had been kept for 24 h in serum-free conditions, basal PGI2-production was below the detection limit and agonist-induced PGI2responses were very low when compared to cells kept in culture medium containing 10% FCS. For this reason, and in order to study PGI2-production and MAPK-activation under the same experimental conditions, we kept ventricular cardiomyocytes in culture medium containing 10% FCS before studying EGF- or PMA-induced MAPK activation. Interestingly, EGF- or PMA-induced phosphorylation of 42-KDa MAPK was more marked than stimulation of MAPK activity, and after 15 min of exposure to EGF or PMA, phosphorylation of 42KDa MAPK was considerably less decreased than MAPK activity. In view of the fact that activation of MAPK involves phosphorylation of a threonine and a tyrosine residue (Anderson et al., 1990), our observations suggest that partial phosphorylation of the enzyme on either the threonine or the tyrosine residue is not sufficient to induce a significant increase in MAPK activity. Confirming the hypothesis that MAPK is the endogenous activator of PLA2 in vivo, EGF-induced prostacyclin-formation appears to be a MAPK-dependent event in cardiomyocytes, as the MAPKinhibiting purine 6-thioguanine abolished EGF-induced MAPK-activation and PGI2-formation in these cells without demonstrating non-selective inhibitory effects on cardiomyocyte PKC, PKA, phospholipase A2 and cyclooxygenase activities. Likewise, the specific MAPK kinase-inhibitor PD98059 (Alessi et al., 1995) completely suppressed the EGF-stimulated PGI2 response in cardiomyocytes, an observation which is in agreement with that of Patel et al. (1996) on ATP-stimulated endothelial PGI2-formation. Thus, our experiments suggest that stimulation of MAPK activity also underlies activation of the

MAP Kinase-mediated PGI2 Formation in Cardiomyocytes

40 kDa myocardial cPLA2, which has been found to be distinctly different from the 100 kDa cPLA2 characterized in other cells (Hazen et al., 1990). Further upholding the key role of MAPK in the stimulation of myocardial cPLA2, we found that the suppression of PMA-induced MAPK-activation in 6thioguanine- or PD98059-treated cardiomyocytes also abolished the PMA-induced PGI2 response, indicating that while activation of PKC is not sufficient to induce PGI2-formation in these cells, activation of MAPK is necessary for the PMA-induced PGI2 response. In summary, it appears that although MAPKdependent prostaglandin-production can be activated through PKC-dependent and -independent pathways in other systems (Qiu and Leslie, 1994; Li et al., 1995), it is the tyrosine kinase-dependent activation of MAPK that is at the basis of EGFinduced prostaglandin-release in cardiomyocytes. These findings are also consistent with studies reporting that ventricular cardiomyocytes express immunoreactive, functional EGF receptors whose activation leads to the stimulation of adenylate cyclase in rat myocytes (Nair et al., 1990; Yu et al., 1992), and promote increases in the contractile rate of chicken embryonic ventricular myocyte aggregates (Rabkin et al., 1987). In this context, the observation that a growth factor can promote MAPK-activation and prostaglandin-formation in ventricular cardiomyocytes is interesting from the physiological point of view, particularly with regard to the mechanisms underlying the pathogenesis of myocardial hypertrophy. Ventricular cardiomyocytes respond to mitogenic stimuli by increasing cell size as opposed to cell number (hypertrophy v hyperplasia, see Shubeita et al., 1990; Sadoshima et al., 1995). Interestingly, the activation of MAPK appears to be involved in the hypertrophic response(s) of cardiomyocytes to a range of physiological stimuli, including fibroblast growth factor, endothelin and angiotensin II (Bogoyevitch et al., 1994; Sadoshima et al., 1995). In view of the observation that arachidonic acid metabolites are clearly involved in MAPK-mediated mitogenic responses in other cell types, where they play the role of second messengers regulating the transcription of the early-immediate genes c-fos and c-myc (Handler et al., 1990; Sellmayer et al., 1991), it is conceivable that MAPK-dependent prostaglandin-formation constitutes a first step in the signalling cascade leading to cardiac hypertrophy. In line with this hypothesis, arachidonic acid metabolites have been shown to promote a host of different effects in cardiomyocytes, including the activation of adenylyl cyclase, increases in cellular

943

contraction frequence, the mobilization of calcium from intracellular stores, and the synthesis and release of atrial natriuretic peptide (Kovacic-Milivojevic et al., 1990; Church et al., 1994b, 1996). In conclusion, the current report presents evidence indicating that while the activation of MAPK underlies both PKC-dependent and -independent prostacyclin formation in cultured spontaneouslybeating neonatal rat ventricular cardiomyocytes, EGF promotes prostaglandin-formation through a tyrosine kinase-dependent, but PKC-independent pathway in these cells. In view of the obvious role of PKC in cardiomyocyte prostaglandin-formation induced by angiotensin II, substance P and vasopressin (Church et al., 1994a, 1996; Van der Bent et al., 1994), MAPK thus appears to serve as a convergence point integrating diverse signal transduction pathways leading to the formation of prostaglandins in these cells.

Acknowledgements We thank Mrs C. Gerber-Wicht, M. Rey and M. Klein for their excellent technical assistance. We are grateful to Dr J. M. Dunn for the generous gift of MAPK antisera, and to Ciba-Geigy (Novartis) for providing a specific PKC inhibitor. This study was supported by a grant from the Swiss National Science Foundation (No. 31-42295.94), a grant from the Swiss Foundation of Cardiology, and a grant from the Geneva Cancer League.

References A DR, C A, C P, D DT, S AR, 1995. PD 098059 is a specific inhibitor of the activation of mitogen-activated protein kinase kinase in vitro and in vivo. J Biol Chem 270: 27489–27494. A NG, M J, T NK, S TW, 1990. Requirement for integration of signals from two distinct phosphorylation pathways for activation of MAP kinase. Nature 343: 651–653. B F, M F, R G, D A L, G G, 1990. Production and biological interactions of prostacyclin and platelet-activating factor in acute myocardial ischemia in the perfused rabbit heart. Cardiovasc Pharmacol 16: 727–732. B MA, G PE, A MB, C A, L A, M CJ, P PJ, S PH, 1994. Endothelin-1 and fibroblast growth factors stimulate the mitogen-activated protein kinase signaling cascade in cardiac myocytes. J Biol Chem 269: 1110–1119. B JV, G JH, N RA, 1990. Epidermal growth factor enhances glomerulosa mesangial cell soluble PLA2 activity. J Biol Chem 265: 4934–4938. B J, R P, H B, C F, V A,

944

S. B. Quintaje et al.

 B  H PMP, 1995. The epidermal growth factor. Cell Biol Int 19: 413–430. B S, C DJ, V MB, L U, 1992. Functional inhibition of protein kinase-mediated effects in myocardial tissue is due to the phosphatase 2A. Biochem J 286: 851–855. C L, G DG, 1995. Natriuretic peptides inhibit DNA synthesis in cardiac fibroblasts. Hypertension 25: 227–235. C DJ, B S, V MB, L U, 1993. Protein kinase C-mediated phospholipase A2 activation, platelet-activating factor generation and prostacyclin release in spontaneously beating rat cardiomyocytes. Biochem J 290: 477–482. C DJ, B-Q S,   B V, V MB, L U, 1994a. Protein kinase C-dependent prostaglandin production mediates angiotensin II-induced atrial natriuretic peptide release. Biochem J 298: 451–456. C DJ, V  B V, V MB, L U, 1994b. Role of prostaglandin-mediated cAMP formation in protein kinase C-dependent secretion of atrial natriuretic peptide in rat cardiomyocytes. Biochem J 303: 217–225. C DJ, A SJ, V MB, C A, K E, L U, 1996. Stimulation of atrial natriuretic peptide release by neurokinins in rat ventricular cardiomyocytes. Am J Physiol 270: H935– H944. C MH, B TG, R DJ, 1991. Extracellular signal-regulated kinases: ERK’s in progress. Cell Regul 2: 965–978. H JA, D RM, E TE, 1990. Mitogenic signaling by epidermal growth factor (EGF), but not platelet-derived growth factors, requires arachidonic acid metabolism in BALB/c 3T3 cells. J Biol Chem 265: 3669–3673. H SL, S RJ, G RW, 1990. Purification and characterization of canine myocardial cytosolic phospholipase A2. J Biol Chem 265: 10622–10630. K-M B, S HD, G DG, 1990. Arachidonic acid metabolites regulate the secretion of atrial natriuretic peptide in cultured rat atrial cardiocytes. J Physiol Pharmacol 693: 1493–1499. K JM, A H, Z XF, B P, B DL, R UR, A J, 1992. Raf-1 activates MAP kinase-kinase. Nature 358: 417–421. L-C CA, P CM, G AM, B KJ, J GI, 1993. A divergence in the MAP kinase regulatory network defined by MEK kinase and Raf. Science 260: 315–319. L X, Z F, S RW, N RA, 1995. Inhibition of MAP kinase by prostaglandin E2 and forskolin in rat mesangial cells. Am J Physiol 269: C986–C991. L LL, L AY, K JL, 1992. Cytosolic phospholipase A2 is coupled to hormonally regulated release of arachidonic acid. Proc Natl Acad Sci 89: 6147–6151. L LL, W M, L AY, K JL, S A, D RJ, 1993. cPLA2 is phosphorylated and activated by MAP kinase. Cell 72: 69–278. M BL, R SJ, F S, M M, L R, L A, S J, 1989. EGF induces tyrosine phosphorylation of phospholipase Cc, a potential mechanism for receptor signaling. Cell 57: 1101–1107. M T, R U, F D, A E, R¨  J,

M¨  M, C G, M A, 1989. A derivative of staurosporine (CGP 41251) shows selectivity for protein kinase C inhibition and in vitro anti-proliferative as well as in vivo anti-tumor activity. Int J Cancer 43: 851–856. N BG, P B, M G, P TB, 1990. Gaa mediates epidermal growth factor-elicited stimulation of rat cardiac adenylate cyclase. J Biol Chem 265: 21317–21322. N RA, W S, Q NX, V P V, J GL, H LE, 1993. Phosphorylation and activation of a high molecular weight form of phospholipase A2 by p42 microtubule-associated protein 2 kinase and protein kinase C. J Biol Chem 268: 1960–1964. P V, B C, G A, W N, B MR, 1996. Phosphorylation and activation of p42 and p44 mitogen-activated protein kinase are required for the P2 purinoreceptor stimulation of endothelial prostacyclin production. Biochem J 320: 221–226. Q ZH, L CC, 1994. Protein kinase C-dependent and -independent pathways of mitogen-activated protein kinase activation in macrophages by stimuli that activate phospholipase A2. J Biol Chem 269: 19480– 19487. R SW, S P, M S, 1987. The effect of epidermal growth factor on chronotropic response in cardiac cells in culture. Biochem Biophys Res Commun 146: 889–897. S J, I S, 1993. Molecular characterization of angiotensin II-induced hypertrophy of cardiac myocytes and hyperplasia of cardiac fibroblasts. Circ Res 73: 413–423. S J, Q Z, M JP, I S, 1995. Angiotensin II and other hypertrophic stimuli mediated by G protein-coupled receptors activate tyrosine kinase, mitogen-activated protein kinase, and 90-kD S6 kinase in cardiac myocytes. Circ Res 76: 1–15. S A, U WM, W PC, B JV, 1991. Endogenous non-cyclooxygenase metabolites of arachidonic acid modulate growth and mRNA levels of immediate-early response genes in rat mesangial cells. J Biol Chem 266: 3800–3807. S HE, MD PM, H AN, K KU, G CC, B JH, C KR, 1990. Endothelin induction of inositol phospholipid hydrolysis, sarcomere assembly, and cardiac gene expression in ventricular myocytes. J Biol Chem 265: 20555–20562. S M, D LHK, B J,  L SW, 1991. Antibody-induced dimerization activates the epidermal growth factor receptor tyrosine kinase. J Biol Chem 266: 1733–1739. T J, F JA, T A, 1994. Mitogenactivated protein kinases mediate changes in gene expression, but not cytoskeletal organization associated with cardiac muscle cell hypertrophy. J Cell Biology 126: 1565–1571. U A, S J, 1990. Signal transduction by receptors with tyrosine kinase activity. Cell 61: 203–212. V B M, E W,   V GJ, R RS, 1989. The significance of myocardial eicosanoid production. Mol Cell Biochem 88: 113–121. V  B V, C DJ, V MB, M P, K DC, C AM, L U, 1994. [Ca2+]i and protein kinase C in vasopressin-induced prosta-

MAP Kinase-mediated PGI2 Formation in Cardiomyocytes cyclin and ANP release in rat cardiomyocytes. Am J Physiol 266: H597–H605. W Y, S MS, P J, D MJ, 1992. Endothelin rapidly stimulates mitogen-activated protein kinase activity in rat mesangial cells. Biochem J 287: 589–594. X M, I PA, 1996. Protein kinase C-dependent

945

activation of cytosolic phospholipase A2 and mitogenactivated protein kinase by a1-adrenergic receptors in Madin–Darby canine kidney cells. J Clin Invest 97: 1302–1310. Y Y, N BP, P TP, 1992. Epidermal growth factor stimulates cAMP accumulation in cultured rat cardiac myocytes. J Cell Physiol 50: 559–567.