Prostaglandin F2α (PGF2α) and the Isoprostane, 8,12-iso-Isoprostane F2α-III, Induce Cardiomyocyte Hypertrophy

Prostaglandin receptors may be activated by their cognate ligand or by free radical catalyzed isoprostanes, products of arachidonic acid peroxidation. For example, prostaglandin F2α (PGF2α) causes hypertrophy of neonatal rat ventricular myocytes, via the PGF2α receptor (FP). However, the FP may also be activated by the isoprostane, 8,12-iso-iPF2α-III (Kunapuli, P., Lawson, J. A., Rokach, J., and FitzGerald, G. A.(1997) J. Biol. Chem. 272, 27147–27154).Both ligands induce myocyte hypertrophy with overlapping potencies. Interestingly, the hypertrophic effects of these two agonists on cardiomyocytes are additive. Furthermore, the preference of these two agonists for activation of intracellular signal transduction pathways differs in several respects. Thus, PGF2α and 8,12-iso-iPF2α-III stimulate inositol phosphate formation with EC50 values of 50 ± 12 nm and 3.5 ± 0.6 μm, respectively. Moreover, PGF2α causes a robust activation (∼50-fold) of Erk2, whereas 8,12-iso-iPF2α-III has no effect. Similarly, PGF2α causes translocation of cytosolic phospholipase A2 and also results in a 7-fold increment in the formation of 6-keto-PGF1α, whereas 8,12-iso-iPF2α-III exerts no effect on this pathway. On the other hand, both agonists are equally potent in activating JNK1 and c-Jun, whereas neither activates the p38 kinase. Both PGF2α and 8,12-iso-iPF2α-III activate the p70S6 kinase (p70S6K), but not Akt, downstream of phosphatidylinositol-3-kinase (PI3K). However, both wortmannin, a PI3K inhibitor, and rapamycin, an inhibitor of p70S6K activity, inhibit 8,12-iso-iPF2α-III -induced myocyte hypertrophy, with IC50 values of 60 ± 12 and 3 ± 1.7 nm, respectively, whereas neither compound abrogates the PGF2α-mediated response. Thus, both PGF2α and 8,12-iso-iPF2α-III induce myocyte hypertrophy via discrete signaling pathways. Although both agonists signal via the JNK pathway to initiate changes in c-Jun-dependent gene transcription, PGF2αpreferentially activates the MEK-Erk2- cytosolic phospholipase A2 pathway. In contrast, the PI3K-p70S6Kpathway appears to be essential for 8,12-iso-iPF2α-III-induced myocyte hypertrophy.

Prostaglandins (PGs) 1 are arachidonic acid metabolites that are produced in a wide variety of tissues in response to mechanical and chemical stimuli. The actions of PGF 2␣ are thought to be mediated via the PGF 2␣ receptor (FP), which is a member of the G protein-coupled receptor (GPCR) superfamily (1). PGF 2␣ has diverse physiological actions, ranging from being a potent luteolytic agent (2) to causing vascular smooth muscle contraction (3). In the myocardium, the formation of PGs is induced by pressure overload (4) which can result in cardiac hypertrophy (5,6). Conversely, PG synthase inhibitors diminish the hypertrophic response induced by hypertension (7).
Prostaglandin F 2␣ (PGF 2␣ ) has recently been shown to stimulate hypertrophy of neonatal rat ventricular myocytes and to induce the expression of myofibrillar genes, independent of muscle contraction (8). Paoni and co-workers (9) have demonstrated that ventricular myocytes enlarge and produce ANF in response to PGF 2␣ and its analog, fluprostenol, in a dose-dependent manner. Furthermore, chronic administration of fluprostenol resulted in an increase in cardiac growth (heart weight-and ventricular weight-to-body weight ratios) in vivo (9). Although the actions of PGF 2␣ on cardiomyocytes suggest important roles for this eicosanoid in development, compensatory hypertrophy, and recovery of the heart from injury, the molecular mechanisms of PGF 2␣ -induced cardiac myocyte hypertrophy remain largely unknown. F 2 isoprostanes (iPs) are PGF 2␣ isomers that are generated by free radical-catalyzed peroxidation of arachidonic acid (10). We have performed total stereospecific synthesis of several iPs, including 8,12-iso-iPF 2␣ -III (previously known as 12-iso-PGF 2␣ ) (11) and iPF 2␣ -VI (previously known as IPF 2␣ -I) (12) and have recently demonstrated that the iP, 8,12-iso-iPF 2␣ -III, may activate FP in a specific and saturable manner (13). The iP, 8,12-iso-iPF 2␣ -III is a member of the group III isoprostanes, which also includes 8-iso-iPF 2␣ -III. Group III iPs are derived originally from 11-hydroperoxyeicosatetraenoic acid. The 8-isoand 8,12-iso-iPF 2␣ -III are identical in every respect with PGF 2␣ except for the stereochemistry of the side chains at C-8 and C-12. We and others (14 -16) have also shown that 8-iso-iPF 2␣ -III may activate the thromboxane receptor in a specific and * This work was supported in part by National Institutes of Health Training Grant T-32-HL07843-02 (to P. K.), by National Institutes of Health Grants DK44730 (to J. R.), DK45696 (to J. L. M.), and HL54500 (to G. A. F.), and by National Science Foundation Grant CHE-9013145. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
During embryonic development, the heart enlarges by the proliferation of cardiac myocytes. Shortly after birth, however, cardiac myocytes lose their capacity for mitogenesis, and further growth of the myocardium occurs by enlargement of the existing cardiomyocytes (hypertrophy) (20). Cardiac hypertrophy enables the heart to adapt to demands for increased cardiac output and to injury and is characterized by an increase in protein content of the individual cardiomyocytes (21). The neonatal rat ventricular myocyte culture has been developed to mimic many features of the hypertrophic response in vivo, including an increase in cell size and protein content, activation of immediate early genes, and activation of embryonic genes, such as atrial natriuretic factor (ANF) (20). Norepinephrine (22), phenylephrine (23), endothelin-I (24), and angiotensin-II (25) have been shown to induce GPCR-mediated activation and thereby phosphorylation of different intracellular signaling intermediates, resulting in myocyte hypertrophy.
Stimulation of ␣ 1 -adrenergic receptors (␣ 1 -AR) in ventricular myocytes triggers an increase in cell size, organization of myofilaments into sarcomeric units, and transcriptional activation of ANF. It is well established that Ras-dependent pathways involving Raf, MEK, and Erks are involved in ␣ 1 -AR-mediated hypertrophy in vivo and in vitro (26 -28). Recently, Brown and co-workers (29) have shown that ␣ 1 -AR-mediated cardiac myocyte hypertrophy involves signaling through MEK kinase, JNK kinase, and JNK, with a resultant increase in c-Jun transcriptional activity. Activation of c-Jun culminates in increased ANF gene transcription.
In this report, we establish for the first time that the biological action of an F 2 isoprostane, 8,12-iso-iPF 2␣ -III, is similar in potency to that of the cognate prostaglandin ligand, PGF 2␣ . However, despite apparently activating the same membrane receptor, PGF 2␣ and 8,12-iso-iPF 2␣ -III induce ventricular myocyte hypertrophy by preferentially activating different intracellular signaling pathways.
Cell Culture-Neonatal ventricular myocytes were cultured from 1-day-old Sprague-Dawley rats as described by Simpson and Savion (30). Briefly, 1-mm cubes of ventricles were placed in ice-cold dissociation buffer, pH 7.5 (137 mM NaCl, 5.36 mM KCl, 0.81 mM MgSO 4 ⅐7H 2 O, 5.55 mM dextrose, 0.44 mM KH 2 PO 4 , 0.34 mM Na 2 HPO 4 ⅐7H 2 O, 20 mM HEPES, 50 units/ml penicillin, 50 g/ml streptomycin), containing 0.1% trypsin and 0.002% DNase. After mechanical dissociation of the tissue by several rounds of alternate gently pipetting and centrifugation at 700 ϫ g for 10 min at 4°C, cells were preplated for 30 min in minimum Eagle's medium with 5% calf serum at 37°C to enhance the attachment of nonmyocytes, leaving the myocytes in suspension. An aliquot of the non-attached cells was counted in a hemocytometer in quadruplicate with 0.4% trypan blue to exclude dead cells. Nonmyocytes in the myocyte culture were limited to Ͻ10% of the total cell number by inclusion of 0.1 mM bromodeoxyuridine in the culture medium for 6 -8 h. Myocytes were then plated in Dulbecco's modified Eagle's medium (DMEM) with 10% calf serum and antibiotics at a density of 500 cells per mm 2 and maintained overnight at 37°C in 5% CO 2 .
Measurement of Inositol Phosphate Formation-Myocytes in 12-well dishes were serum-starved for 24 h and then labeled to equilibrium with myo-[2-3 H]inositol (2 Ci/ml) for 16 -24 h in serum-free DMEM containing 20 mM HEPES, pH 7.5, and 0.5% Albumax. Cells were preincubated with 20 mM LiCl for 15 min at 37°C and then stimulated by addition of agonist for 10 min at 37°C. Total inositol phosphates were measured as described previously (31). Briefly, InsP formation was stopped by aspiration of the medium. Formic acid (0.75 ml of 10 mM solution per well) was added, and the plates were incubated at room temperature for 30 min. The solution containing the extracted InsP was neutralized and diluted with 3 ml of 5 mM NH 4 OH (yielding a final pH of 8 -9) and then applied directly to a column containing 0.7 ml of the anion exchange resin, AG-1-X8. The column was washed with 4 ml of 40 mM ammonium formate, pH 5.0, and the total InsPs was eluted with 4 ml of 2 M ammonium formate, pH 5.0. One ml of the eluate was counted in 9 ml of scintillant. Results presented are the mean Ϯ S.E. of 4 independent experiments performed in duplicate.
Measurement of Phenylalanine Incorporation-[ 3 H]Phenylalanine incorporation was determined to estimate the relative rates of protein synthesis as described (8). Briefly, myocytes in 12-well dishes were serum-starved for 24 h and stimulated in serum-free DMEM with agonist (or vehicle, as control) for 24 h at 37°C. The medium was replaced with serum-free DMEM containing 0.36 mM L-phenylalanine and 5 Ci/ml L-[2,3,4,5,6-3 H]phenylalanine during the last 4 h of stimulation. Cells were rinsed twice with PBS and incubated in ice-cold 10% trichloroacetic acid for 30 min on ice. Cell precipitates were then washed twice with ice-cold 10% trichloroacetic acid and solubilized in 1% SDS (1 ml/well) at 37°C for 1 h. The SDS-soluble protein was counted in 9 ml of scintillant. Results presented are the mean Ϯ S.E. of 4 -6 independent experiments performed in duplicate.
Measurement of Total Protein Content-Total protein content of myocytes was measured as described (9). Briefly, cells were plated as described above and stimulated with agonist in serum-free DMEM for 24 h at 37°C. The cells were washed twice with PBS, and 100 l of trypsin/EDTA was added to each well. The plates were incubated at 37°C until the cells had rounded. A solution of 5% fetal calf serum in PBS was added to each well to stop the reaction. The cells were harvested, washed by centrifugation, and resuspended in 1 ml of PBS, and the cell number was determined. SDS was added to a final of concentration of 1%, and the cell suspension was vortexed and incubated at 4°C overnight. This mixture was then warmed to 37°C, and protein concentration was determined using the Pierce Micro BCA kit. Protein contents were normalized to the cell count. Results presented are the mean Ϯ S.E of four independent experiments performed in triplicate.
Measurement of ANF-Cells were plated, starved, and stimulated with agonist for 24 h at 37°C in serum-free DMEM as described above to measure ANF formation by myocytes. The culture supernatant was used to assay for ANF by RIA according to manufacturer's instructions. Results presented are the mean Ϯ S.E. of 3 independent experiments performed in triplicate. Statistical analysis of ANF under conditions of additive hypertrophy was performed with Student's unpaired t test (p Ͻ 0.01).
Measurement of cAMP-Myocytes were plated and starved as described above. Cells were pretreated with 0.5 mM isobutylmethylxanthine for 15 min at 37°C prior to the addition of 10 M PGF 2␣ or 8,12-iso-iPF 2a -III for 10 min at 37°C to quantitate the accumulation of intracellular cAMP. Cells were stimulated with 10 M forskolin and 100 M agonist for 10 min at 37°C following pretreatment with isobutylmethylxanthine to analyze the effects of PGF 2␣ and 8,12-iso-iPF 2␣ -III on inhibition of adenylyl cyclase activity. Reactions were terminated by aspiration, and cAMP was extracted with ice-cold 65% ethanol for 30 min. Samples were dried under vacuum and reconstituted in assay buffer, and cAMP was measured by radioimmunoassay (n ϭ 3), according to manufacturer's instructions.
Cell Fractionation-Myocytes were fractionated into membrane and cytosolic fractions essentially as described earlier (32) with slight modifications. Myocytes in 100-mm plates were serum-starved as described above. Cells were stimulated with vehicle (control), 1 M PGF 2␣ , or 8,12-iso-iPF 2␣ -III for 10 min at 37°C. Cells were then lysed in hypotonic lysis buffer (10 mM Tris-HCl, pH 8.0, 2 mM EDTA, 1 mM MgCl 2 , 10 mM Na 2 P 2 O 7 , 10 mM NaF, 500 M Na 3 VO 4 , 1 mM dithiothreitol, and protease inhibitors) by shearing 10 times through a 25-gauge needle. Unbroken cells were pelleted by low speed centrifugation. The supernatant was subjected to a high speed centrifugation at 50,000 ϫ g for 30 min. The resulting pellet (P) was designated the membrane fraction, and the supernatant (S) was designated the cytosolic fraction. The protein content of the membrane and pellet fractions were analyzed by Pierce Micro BCA kit.
Immunocytochemistry-For immunocytochemistry of cPLA 2 , myocytes were plated on chamber slides (Nunc, Napierville, IL), stimulated with vehicle (control) or 10 M agonist for 5 min at 37°C, and fixed with 70% methanol, 30% acetone at Ϫ20°C for 10 min followed by incubation at room temperature for 5 min. The cells were blocked with 2% bovine serum albumin/PBS and then treated with 1:25 dilution of the monoclonal human cPLA 2 antibody in 0.5% bovine serum albumin/PBS (Santa Cruz Biotechnology, Santa Cruz, CA) overnight at 4°C, followed by a 1-h incubation with fluorescein isothiocyanate-labeled anti-mouse IgG in 0.5% bovine serum albumin/PBS at room temperature. Between each step, cells were washed three times for 10 min with PBS. Slides were mounted in Vectashield (Vector Laboratories, Burlingame, CA) and examined by fluorescence confocal microscopy.
Measurement of cPLA 2 Activity-The amount of 6 keto-PGF 1␣ in the myocyte culture medium upon stimulation with vehicle (control), 10 M PGF 2␣ , or 8,12-iso-iPF 2␣ -III (2 min at 37°C) was analyzed by GC/MS as a measure of cPLA 2 activity. Briefly, 1 ng of 3,3,4,4-d4 -6-keto-PGF 1␣ was added to 1 ml of the culture medium. After 15 min at room temperature, 150 l of methoxyamine HCl in water (0.8 mg/l solution) was added, and the pH was lowered to 2.5-3.0 with formic acid. The sample was loaded on a solid phase extraction cartridge (C18 EC, 100 mg; International Sorbent Technology, Mid Glamorgan, UK) preconditioned with 0.5 ml of ethanol followed by 3 ml of water at 1 ml/min using a Rapid Trace SPE workstation (Zymark Corp., Hopkinton, MA). The cartridge was washed with 1 ml of 0.1 M phosphate buffer, pH 7.0, at 1 ml/min followed by 1 ml of 25% ethanol at 1 ml/min, 3 ϫ 5 ml of hexane at 15 ml/min and 3 ϫ 5 ml of air at 30 ml/min. The sample was eluted with 1 ml of ethyl acetate at 1 ml/min and dried under a gentle stream of nitrogen. N,N-Diisopropylethylamine, 10 l and 10% pentafluorobenzyl bromide in acetonitrile were added to the sample, which was allowed to stand for 10 min and dried under N 2 . The sample was then dissolved in 25 l of methanol, applied to a TLC plate (LK6D; Whatman), and developed with 2% methanol in ethyl acetate. A separate plate loaded with 2 g of 2,3-dinor thromboxane/methoxime/pentafluorobenzyl ester was simultaneously developed. The standard plate was dried and dipped in 8% phosphoric acid (H 3 PO 4 ) in 0.3 M cupric sulfate and heated on a hot plate to visualize the standard. The sample plates, 6-mm zones, were scraped 2 mm lower than the standard, and the scrapings were extracted from 100 l of water with 1 ml of ethyl acetate. Piridine and bis-(trimethylsilyl)trifluoroacetamide (Supelco, Bellefonte, PA), 10 l each were added to the sample, which was allowed to stand for 10 min and dried under nitrogen. The sample was dissolved in 20 l of bis-(trimethylsilyl)trifluoroacetamide in dodecane for GC/MS analysis on an MD-800 mass spectrometer (Finnigan Corp., San Jose, CA) interfaced with a GC-8000 gas chromatograph and an AS-800 autosampler. The solvents were purchased from Burdick and Jackson (Muskegon, MI). A 30-m DB5-MS (J & W Scientific, Folsom, CA) GC column was used with helium as the carrier gas. The temperature program consisted of an initial temperature of 190°C for 1 min followed by a programmed increase to 320°C at 20°C/min. The MS was operated in the negative ion (electron capture) chemical ionization mode, using ammonia as the moderating gas. The ions monitored were m/z 614.4 for the endogenous 6-keto-PGF 1␣ methoxyime pentafluorobenzyl ester trimethylsilylether derivative and m/z 618.4 for the internal standard. Results presented are the mean Ϯ S.E. of 3 independent experiments performed in triplicate. Statistical analysis was performed with Student's unpaired t test (p Ͻ 0.005).
Immunoblotting-Cells were plated in 100-mm dishes and starved as described above. Cells were stimulated with 10 M PGF 2␣ or 8,12-iso-iPF 2␣ -III for the indicated times at 37°C. In some experiments, cells were stimulated with 50 nM PGF 2␣ and 200 nM 8,12-iso-iPF 2␣ -III for the indicated times at 37°C. When the effects of protein kinase inhibitors were analyzed, the cells were preincubated with the indicated concentration of the inhibitor at 37°C for 15 min and then stimulated with agonist. The medium was quickly aspirated; cells were washed twice with PBS and then lysed in 1 ml of 20 mM Tris, pH 7.5, 150 mM NaCl, 1% Triton X-100 with protease inhibitors. Lysates were centrifuged at 14,000 rpm for 10 min at 4°C to remove cell debris. The protein concentration in the supernatant was determined by Pierce Micro BCA kit, and 30 -50 g of total protein was subjected to SDS-polyacrylamide gel electrophoresis on a 10% gel and transferred to nitrocellulose blots. The blots were blocked in 5% milk in Tris-buffered saline with 1% Triton X-100 (TBS-T) for 1 h at room temperature and then probed with various phospho-specific antibodies according to manufacturer's instructions. The blots were washed 6 times with TBS-T and probed with a 1:2000 to a 1:5000 dilution of the appropriate horseradish peroxidaseconjugated secondary antibody for one hour at room temperature. Antigen-antibody complexes were then visualized by chemiluminescence. Results presented are representative of 3-4 experiments. Wherever pertinent, densitometric analysis was performed for quantitation, and statistical analysis was performed by Student's unpaired t test.
Interestingly, in the presence of 1 M PGF 2␣ and varying concentrations of 8,12-iso-iPF 2␣ -III, the hypertrophic responses induced by PGF 2␣ and 8,12-iso-iPF 2␣ -III appear to be additive, as observed by the rate of protein synthesis (Fig. 3A) and the total protein content (Fig. 3B). Similar results were also observed in the presence of 100 nM PGF 2␣ and varying concentrations of 8,12-iso-iPF 2␣ -III (data not shown). Stimulation of myocytes with 10 nM PGF 2␣ resulted in the secretion of 3.25 Ϯ 0.13 ng/ml ANF, and stimulation with 100 nM 8,12-iso-iPF 2␣ -III resulted in the secretion of 2 Ϯ 0.18 ng/ml ANF. However, in the presence of 10 nM PGF 2␣ and 100 nM 8,12-iso-iPF 2␣ -III, there was a statistically significant increase in ANF formation (4.15 Ϯ 0.29 ng/ml; p Ͻ 0.01) above that induced by PGF 2␣ . Thus myocyte hypertrophy induced by these two agonists appears to be additive.
PGF 2␣ and 8,12-Iso-iPF 2␣ -III -induced Second Messenger Formation in Myocytes-Stimulation of ventricular myocytes with PGF 2␣ results in a dose-dependent increase in InsP formation. Maximal levels of InsP formation were observed at 1 M, with an EC 50 of 50 Ϯ 12 nM (Fig. 4), both similar to the values reported in cells expressing FP (13). On the other hand, although the iP, 8,12-iso-iPF 2␣ -III also stimulated a dose-dependent increase in total inositol phosphate formation in ventricular myocytes, the EC 50 of InsP formation was significantly higher (3.5 Ϯ 0.6 M), also similar to that reported in cells expressing FP (13). Comparison of the EC 50 values for InsP formation by the two agonists revealed a statistically significant difference (p Ͻ 0.005). We also analyzed the ability of these two agonists to modulate other second messengers in myocytes. Neither agonist increased nor decreased the level of cAMP in ventricular myocytes (data not shown). Intracellular Signaling Pathways Activated by PGF 2␣ and 8,12-Iso-iPF 2␣ -III in Myocytes-Hypertrophy of neonatal rat ventricular myocytes induced by agonists that activate GPCRs has been shown to involve the activation of several intracellular signal transduction pathways (20,33,34). Phosphorylation is a common mechanism of activation of these intracellular signaling intermediates (35,36). We therefore initiated studies to delineate further the major signaling pathways involved in PGF 2␣ and/or 8,12-iso-iPF 2␣ -III -induced ventricular myocyte hypertrophy.
Stimulation of ventricular myocytes with PGF 2␣ caused a robust stimulation of the 42-kDa Erk2 (Fig. 5A, top left), as measured by an increase in Erk2 phosphorylation. Maximal phosphorylation was observed at 5 min, after which phosphorylation slowly declined to a low level by 60 min. This effect of PGF 2␣ on Erk2 phosphorylation was MEK-dependent, since it was inhibited by the MEK inhibitor, PD98059 (37) (Fig. 5A, top  middle). Densitometric analysis revealed that PGF 2␣ increased Erk2 phosphorylation by 50-fold (Fig. 5B). In contrast, 8,12-iso-iPF 2␣ -III caused a very minimal increase in Erk2 phosphoryl-ation (Fig. 5, A bottom and B). This difference in Erk2 phosphorylation by the two agonists was also observed when myocytes were stimulated with 50 nM PGF 2␣ or 200 nM 8,12iso-iPF 2␣ -III (Fig. 5, A top right).
Cytosolic PLA 2 is a substrate of Erks in some cells, resulting in the phosphorylation of cPLA 2 at Ser-505 (38). However, cPLA 2 phosphorylation by Erk in itself may not be sufficient for cPLA 2 activation or arachidonic acid release (39). When activated, cPLA 2 undergoes Ca 2ϩ -dependent translocation from the soluble to the membrane fraction in the nuclear envelope and endoplasmic reticulum, thereby gaining access to its arachidonoyl-containing phospholipid substrate (40). We analyzed therefore the effect the two agonists on cPLA 2 localization in myocytes. Western blot analysis of cytosolic and membrane fractions revealed that in unstimulated cells, cPLA 2 appears to be mostly cytosolic (Fig. 6A, top). When stimulated with 8,12iso-iPF 2␣ -III, there is minimal change in this pattern of cPLA 2 localization. However, when stimulated with PGF 2␣ ,the majority of the cPLA 2 appears to be membrane-associated, with minimal amounts in the cytosolic fraction. Confocal microscopic analysis (Fig. 6A, bottom) of immunofluorescence staining of cPLA 2 confirms that upon stimulation of ventricular myocytes with PGF 2␣ , cPLA 2 translocates from the cytosol to the nuclear membrane. We also analyzed the effect of the two agonists on cPLA 2 activity by measuring the levels of 6-keto-PGF 1␣ , the hydrolysis product of prostacyclin, which is the predominant arachidonic acid metabolite formed by ventricular myocytes (41). Concentrations of 6-keto-PGF 1␣ in the culture supernatant increased significantly (p Ͻ 0.005) when cells were stimulated with PGF 2␣ (7-fold increase) but not with 8,12-iso-iPF 2␣ -III (Fig. 6B). This increment in product formation was sustained from 2 min to 2.5 h (data not shown). These observations are consistent with a difference in PGF 2␣ versus 8,12iso-iPF 2␣ -III induced activation of Erk. These results indicate that PGF 2␣ stimulates the MEK-Erk2-cPLA 2 pathway,  whereas 8,12-iso-iPF 2␣ -III does not.
Since ␣ 1 -adrenergic agonists stimulate JNK activity and hypertrophy in ventricular myocytes (29), the ability of PGF 2␣ and 8,12-iso-iPF 2␣ -III to stimulate other mitogen-activated protein kinase (MAPK) isoforms was examined. Both agonists stimulated phosphorylation of the 46-kDa JNK1 (Fig. 7A) but not the 52-kDa JNK2 with equal potency. In both cases, JNK1 was activated at 5 min and retained the same activity at 30 min. Consistent with these results, both agonists also stimulated the phosphorylation of the transcription factor c-Jun, a substrate of JNK1 with equal potency (Fig. 7B). The level of phosphorylated c-Jun was maximal at 30 min. In contrast, neither agonist stimulated the phosphorylation of the p38 kinase (Fig. 6C).
Taken together, the data support a model where hypertrophy in response to PGF 2␣ and 8,12-iso-iPF 2␣ -III proceeds through both overlapping and distinct signaling pathways. FIG. 5. Effect of PGF 2␣ and 8,12-iso-iPF 2␣ -III on the MEK-Erk-cPLA 2 cascade. A, Western blot analysis of Erk phosphorylation. Crude cell lysates (30 g/lane) prepared from myocytes stimulated with 10 M PGF 2␣ or 8,12-iso-iPF 2␣ -III for 2-60 min at 37°C were electrophoresed on a 10% SDS-polyacrylamide gel, transferred to a nitrocellulose membrane, and probed with a phospho-specific Erk antibody as described under "Experimental Procedures." To analyze the effect of the MEK inhibitor, PD98059 on Erk-2 activation, myocytes were preincubated with 10 M PD98059 for 15 min at 37°C, stimulated with 10 M PGF 2␣ for 5 min at 37°C, and processed for Western blot analysis as described above. The top right panel represents lysates from myocytes stimulated with 50 nM PGF 2␣ or 200 nM 8,12-iso-iPF 2␣ -III for 5 min at 37°C. B, quantitation of Erk2 phosphorylation. The Erk-2 signal observed in Western blots was quantitated by densitometric analysis. Statistical analysis was performed by Student's unpaired t test (p Ͻ 0.005).

DISCUSSION
PGF 2␣ has been shown to cause hypertrophy of neonatal rat ventricular myocytes in vitro and to stimulate cardiac growth in vivo (8,9). These actions of PGF 2␣ are specific and dose-dependent and presumably mediated by its actions on the PGF 2␣ receptor, FP, which has been reported to couple to G␣ q (44) to result in InsP formation. PGF 2␣ -induced ventricular myocyte hypertrophy is independent of muscle contraction and does not affect myocyte proliferation or [ 3 H]thymidine incorporation. However, it involves the induction of myofibrillar genes like myosin light chain-2 (8), c-fos, ANF and ␣-skeletal actin (9). The precise functional importance of these events and. indeed, the molecular events downstream of FP activation are largely unknown.
We have recently demonstrated that FP may also be activated by the free radical catalyzed peroxidation product of arachidonic acid, 8,12-iso-iPF 2␣ -III, in a stable expression system, albeit less potently than by its cognate ligand, PGF 2␣ (13). We now report that 8,12-iso-iPF 2␣ -III is a potent inducer of myocyte hypertrophy. Importantly, this is the first demonstration in which a physiological response mediated by an F 2 isoprostane approximates the potency of the cognate prostaglandin ligand. Although iPs have been speculated to act as incidental ligands at prostanoid receptors, they appear to be weak agonists of prostanoid receptors in vitro. In addition to acting on prostanoid receptors, iPs have been speculated to exert their biological actions by acting on specific isoprostane receptors, although none have been cloned to date. The potency of 8,12-iso-iPF 2␣ -III-induced physiological response (myocyte hypertrophy) is consistent with this notion.
Both PGF 2␣ and 8,12-iso-iPF 2␣ -III stimulate some parameters of myocyte hypertrophy with similar potencies. Thus, both agonists stimulate the JNK-c-Jun pathway, resulting in increased ANF production. This is reminiscent of the activation of ␣ 1 -adrenergic receptors in ventricular myocytes, which results in myocyte hypertrophy involving transcriptional activation of the ANF gene via a Ras-MEK kinase-JNK kinase-Jnkc-Jun pathway (29).
Our results demonstrate that 8,12-iso-iPF 2␣ -III -induced myocyte hypertrophy also requires signaling via the PI3K-p70 S6K pathway. In addition to agonist-induced changes in gene transcription, ventricular myocyte hypertrophy also involves an increase in protein content of the cells and reorganization of contractile proteins into sarcomeric units. The ribosomal S6 protein regulates translation of some mRNAs into proteins and thereby plays an important role in the activation of protein synthesis (42,45). p70 S6K phosphorylates and activates the ribosomal S6 protein, resulting in an increased rate of protein synthesis. p70 S6K functions in a signaling pathway downstream of PI3K (42). Several lines of evidence suggest that the PI3K-p70 S6K pathway is essential for 8,12-iso-iPF 2␣ -IIIinduced myocyte hypertrophy. First, 8,12-iso-iPF 2␣ -III induces the phosphorylation of p70 S6K . Second, wortmannin, a PI3K inhibitor, inhibits 8,12-iso-iPF 2␣ -III-induced myocyte hypertrophy. The IC 50 of wortmannin for inhibition of 8,12-iso-iPF 2␣ -III-induced myocyte hypertrophy suggests that this agonist may be signaling through PI3K␥ (46). Finally, consistent with this observation, rapamycin, a specific inhibitor of p70 S6K activity (42), also inhibits 8,12-iso-iPF 2␣ -III-induced myocyte hypertrophy. Rapamycin has previously been shown to inhibit angiotensin II and ␣ 1 -AR-mediated myocyte hypertrophy (45,47). In contrast, although PGF 2␣ also stimulates p70 S6K in these cells, the contribution of this pathway to PGF 2␣ -stimulated myocyte hypertrophy remains obscure, since neither wortmannin nor rapamycin have any significant effect on PGF 2␣ -induced hypertrophy. Thus, although both agonists cause a similar level of induction of p70 S6K phosphorylation, the contribution of the PI3K-p70 S6K pathway to myocyte hypertrophy appears to differ. The role of PGF 2␣ -mediated signaling through the PI3K-p70 S6K pathway appears to be similar to that previously observed with the ␣ 1 -AR, in which rapamycin does not inhibit phenylephrine-stimulated induction of ANF and ␣-skeletal actin genes (comparable to PGF 2␣ -induced hypertrophy), although LY294002, another PI3K inhibitor, inhibits phenylephrine-stimulated p70 S6K activity. This is comparable to wortmannin, which inhibits PGF 2␣ -induced p70 S6K activity (45). Our results are thus consistent with the notion that intracellular signaling pathways responsible for transcrip-FIG. 6. Effect of PGF 2␣ and 8,12-iso-iPF 2␣ -III on cPLA 2 activity. A, localization of cPLA 2 . Top, Western blot analysis. Myocytes were stimulated with vehicle (control), 10 M PGF 2␣ , or 8,12-iso-iPF 2␣ -III for 10 min at 37°C. Cells were then lysed and fractionated as described under "Experimental Procedures." 50 g of total protein from the membrane fraction (P) or cytosolic fraction (S) was subjected to Western blot analysis and probed with a monoclonal antibody to human cPLA 2 . Bottom, immunocytochemistry: myocytes were plated on chamber slides, stimulated with vehicle (control), 10 M PGF 2␣ , or 8,12-iso-iPF 2␣ -III for 5 min at 37°C, fixed, and processed with monoclonal cPLA 2 antibody as described under "Experimental Procedures." Immunofluorescent staining was visualized by confocal microscopy (magnification, ϫ 100). B, measurement of 6-keto-PGF1␣. Myocytes were stimulated with 10 M PGF 2␣ or 8,12-iso-iPF 2␣ -III for 2 min at 37°C, and the amount of 6-keto-PGF 1␣ in culture supernatant was determined by GC/MS as described under "Experimental Procedures." Statistical analysis was performed with Student's unpaired t test (p Ͻ 0.005). tional and translational responses diverge early after receptor stimulation in ventricular myocytes (45). Interestingly, neither agonist phosphorylates Akt, another downstream effector of PI3K. Although p70 S6K has been proposed to function downstream from Akt, there are examples where Akt and p70 S6K exert differential effects. For example, in some cells, Akt functions in pathways that protect cells from apoptosis, whereas p70 S6K does not (48). Our results suggest that eicosanoid-induced myocyte hypertrophy may be mediated by signaling pathways involving p70 S6K but not necessarily Akt.
There are three classes of mitogen-activated protein kinases (MAPKs) that are responsive to extracellular stimuli in many cells: extracellular signal regulated kinases (Erks), c-Jun Nterminal kinase (JNK), and the p38 class of kinases. Although both PGF 2␣ and 8,12-iso-iPF 2␣ -III are equally potent activators of JNK, only PGF 2␣ appears to signal through the MEK-Erk-cPLA 2 pathway. This is supported by the level of Erk2 phosphorylation and cPLA 2 activation in addition to the fact that the MEK inhibitor, PD98059, inhibits PGF 2␣ -induced myocyte hypertrophy. In an analogous scenario, the mitogenic effects of PGF 2␣ in NIH3T3 cells have been attributed to G␣ q -mediated activation of the Ras-Erk pathway (49). Stimulation of oxidative stress by the addition of hydrogen peroxide in vitro has been shown to activate the Ras-Raf-Erk mediated signaling pathway in ventricular myocytes (50). However, our results demonstrate that the F 2 isoprostane, 8,12-iso-iPF 2␣ -III, a prod-uct of free radical-catalyzed peroxidation of arachidonic acid, fails to cause significant activation of the MEK-Erk-cPLA 2 pathway in these cells. Although the level of activation of the MEK-Erk-cPLA 2 pathway by the two agonists is significantly different, PD98059 inhibits myocyte hypertrophy induced by both agonists. These results may indicate that basal levels of MEK activity are required for myocyte hypertrophy, and any decrease in this level interferes with the physiological response. Alternatively, PD98059 may be acting on other cellular targets (37), although this is unlikely at the concentrations used in this study.
Although the p38 MAPKs have been shown to be involved in cardiac muscle cell hypertrophy (51), our results reveal that neither PGF 2␣ nor 8,12-iso-iPF 2␣ -III activates p38 in ventricular myocytes. The dissociation of JNK and p38 activation further supports the notion that these closely related kinases perform different physiological roles.
These results also present several lines of evidence compatible with the possibility that 8,12-iso-iPF 2␣ -III may induce ventricular myocyte hypertrophy by activating receptor(s) in addition to FP. First, the functional response induced in myocytes by PGF 2␣ and 8,12-iso-iPF 2␣ -III is additive, as judged by three different parameters of myocyte hypertrophy. Indeed, it is possible that these two agonists may interact in a synergistic manner. However, limitations of the system permit us to conclude only that these effects are at least additive. Second, FIG. 7. Effect of PGF 2␣ and 8,12-iso-iPF 2␣ -III on JNK and p38. Crude cell lysates (30 g/lane) prepared from ventricular myocytes stimulated with 10 M PGF 2␣ or 8,12-iso-iPF 2␣ -III for various times were subjected to Western blot analysis and probed with a phospho-specific JNK antibody (A), phospho-specific c-Jun antibody (B), and phospho-specific p38 antibody (30 min stimulation) (C), as described under "Experimental Procedures." Vehicle-stimulated myocytes served as a negative control. A and B, the right panel represents lysates from myocytes stimulated with 50 nM PGF 2␣ or 200 nM 8,12-iso-iPF 2␣ -III for 30 min at 37°C. 8,12-iso-iPF 2␣ -III is significantly weaker than PGF 2␣ in stimulating inositol phosphate formation, presumably via FP. Third, the intracellular signaling pathways activated by these two agonists are clearly different, albeit overlapping. In addition to activating PG receptors, F 2 isoprostanes have been speculated to exert their biological actions by acting on specific isoprostane receptors (15). However, none have been cloned or identified unambiguously to date. In the absence of specific FP antagonists, the extent to which the FP mediates 8,12-iso-iPF 2␣ -III-induced myocyte hypertrophy remains unclear. Investigations on ventricular myocytes from FP Ϫ/Ϫ mice (52) may help clarify this issue.
Although 8,12-iso-iPF 2␣ -III can activate FP, its relative potency for FP activation, reflected by InsP formation, is clearly weaker than its comparative ability to evoke a hypertrophic response. Another possible explanation for this observation could be that whereas PGF 2␣ and 8,12-iso-iPF 2␣ -III both activate the FP, they may transduce signals preferentially through different G proteins. GPCRs are capable of coupling to multiple G proteins (53)(54)(55)(56). Indeed, both thromboxane A 2 and 8-iso-iPF 2␣ -III may activate both G q and G 11 (57). To date, FP has only been shown to couple to G q in Chinese hamster cells (44). FP does not couple to G s or G i in ventricular myocytes. However, the ability of FP to couple to other G proteins remains to FIG. 8. Effect of PGF 2␣ and 8,12-iso-iPF 2␣ -III on p70 S6K and Akt phosphorylation. A, phosphorylation of p70 S6K . Crude cell lysates (50 g/lane) prepared from ventricular myocytes stimulated with 10 M PGF 2␣ or 8,12-iso-iPF 2␣ -III for various times were subjected to Western blot analysis and probed with a phospho-specific p70 S6K antibody as described under "Experimental Procedures." Crude cell lysates from vehicle-treated (6 h) myocytes served as a negative control. The right panel represents lysates from myocytes stimulated with 50 nM PGF 2␣ or 200 nM 8,12-iso-iPF 2␣ -III for 1 h at 37°C. B, effect of wortmannin and rapamycin on p70 S6K phosphorylation. Ventricular myocytes preincubated with or without 1 nM wortmannin or rapamycin for 10 min at 37°C were stimulated with 10 M PGF 2␣ or 8,12-iso-iPF 2␣ -III for 30 min at 37°C. Crude cell lysates were prepared and analyzed (50 g/lane) for the presence of phosphorylated p70 S6K by a gel shift assay on a 7.5% SDS-polyacrylamide gel, followed by Western blot analysis using a p70 S6K antibody. Phosphorylated p70 S6K appears as a higher molecular weight band on the gel. C, effect of PGF2␣ and 8,12-iso-iPF2␣-III on Akt. Crude cell lysates (50 g/lane) prepared from ventricular myocytes stimulated with 10 M PGF 2␣ or 8,12-iso-iPF 2␣ -III for 30 min at 37°C were subjected to Western blot analysis and probed with a Akt antibody (left) or a phospho-specific Akt antibody (right) as described under "Experimental Procedures." Crude cell lysates from vehicle-treated myocytes (control) served as a positive control for the Akt antibody. NIH3T3 cells treated with platelet-derived growth factor and vehicle-treated myocytes (control) served as positive and negative controls, respectively, for the phospho-specific Akt antibody. be investigated. Activation of G q has been shown to play a central role in GPCR-mediated cardiac myocyte hypertrophy (58). The ability of PGF 2␣ and 8,12-iso-iPF 2␣ -III to cause hypertrophy in ventricular myocytes from mice deficient in G q (59) would elucidate the ability of FP to couple to other heterotrimeric G proteins in vivo.
Activation of the FP is likely to be of importance in cardiac development (8) and may condition the response of the failing heart, a syndrome characterized by oxidative stress in vivo (60). Although iPs may act as incidental ligands at membrane receptors for PGs, they are commonly much less potent than the cognate ligand in evoking functional responses, such as vasoconstriction (61) or mitogenesis (13,62). In this regard, the potency with which PGF 2␣ and 8,12-iso-iPF 2␣ -III evoke myo-cyte hypertrophy, but not activation of the FP, as reflected by InsP formation, is unusual. Although differential utilization of G proteins or perhaps activation of distinct receptors by iPs remain formal possibilities, we have demonstrated that the two ligands also appear to utilize distinct but overlapping intracellular signaling pathways to mediate this response. Insight into the mechanisms of cellular activation by iPs may clarify their functional importance in vivo in syndromes such as heart failure and ischemia-reperfusion injury (17,63), in which oxidant stress and augmented PG synthesis coincide.