Both Gs and Gi Proteins Are Critically Involved in Isoproterenol-induced Cardiomyocyte Hypertrophy*

Activation of β-adrenoreceptors induces cardiomyocyte hypertrophy. In the present study, we examined isoproterenol-evoked intracellular signal transduction pathways leading to activation of extracellular signal-regulated kinases (ERKs) and cardiomyocyte hypertrophy. Inhibitors for cAMP and protein kinase A (PKA) abolished isoproterenol-evoked ERK activation, suggesting that Gs protein is involved in the activation. Inhibition of Gi protein by pertussis toxin, however, also suppressed isoproterenol-induced ERK activation. Overexpression of the Gβγ subunit binding domain of the β-adrenoreceptor kinase 1 and of COOH-terminal Src kinase, which inhibit functions of Gβγ and the Src family tyrosine kinases, respectively, also inhibited isoproterenol-induced ERK activation. Overexpression of dominant-negative mutants of Ras and Raf-1 kinase and of the β-adrenoreceptor mutant that lacks phosphorylation sites by PKA abolished isoproterenol-stimulated ERK activation. The isoproterenol-induced increase in protein synthesis was also suppressed by inhibitors for PKA, Gi, tyrosine kinases, or Ras. These results suggest that isoproterenol induces ERK activation and cardiomyocyte hypertrophy through two different G proteins, Gs and Gi. cAMP-dependent PKA activation through Gs may phosphorylate the β-adrenoreceptor, leading to coupling of the receptor from Gs to Gi. Activation of Giactivates ERKs through Gβγ, Src family tyrosine kinases, Ras, and Raf-1 kinase.

Cardiac hypertrophy is often associated with an increase in intracardiac sympathetic nerve activity and with elevated plasma catecholamines (1). Treatment of cardiac myocytes with catecholamines not only changes their functions such as beating rates and contractile activity but also induces typical hypertrophic responses (2)(3)(4)(5)(6)(7). There are two major subtypes, ␣ and ␤, in adrenoreceptors (ARs). 1 AR agonists such as norepi-nephrine (NE) (␣ and ␤), phenylephrine (PHE) (␣), and isoproterenol (ISO) (␤) have been reported to induce cardiomyocyte hypertrophy (2)(3)(4)(5)(6)(7). Prolonged infusion of subpressor doses of NE increases the mass of the myocardium and the thickness of the left ventricular wall, suggesting that NE has direct hypertrophic effects on cardiac myocytes without affecting afterload (2). We have reported that NE induces cardiomyocyte hypertrophy through both ␣and ␤-ARs (7). It has been reported that PHE evokes hypertrophic responses in the cardiac myocytes of neonatal rats (5) and that expression of constitutively active ␣-AR induces cardiac hypertrophy in adult mice (8). Both in vivo and in vitro studies demonstrate that ISO also stimulates expression of proto-oncogenes in cardiomyocytes and induces cardiac hypertrophy (6,9,10).
Activation of each AR evokes specific intracellular signals (4,11). It has been shown that stimulation of ␣ 1 -AR activates phosphoinositide-specific phospholipase C via G q protein and hydrolyzes phosphoinositide 4,5-bisphosphate into inositol 1,4,5-triphosphate and diacylglycerol. Diacylglycerol activates protein kinase C (PKC) leading to activation of the Raf-1 kinase/extracellular signal-regulated protein kinase (ERK) cascade (12,13). There was a report indicating that PHE activates ERKs and induces cardiomyocyte hypertrophy through the Rasdependent pathway (5). Regarding ␤-AR-induced signaling pathways, stimulation of ␤-AR activates adenylyl cyclase through a different G protein, G s . Activation of adenylyl cyclase produces a second messenger cyclic adenosine monophosphate (cAMP), leading to activation of cAMP-dependent protein kinase (PKA) (14). PKA activation is important for the control of cell growth and differentiation. cAMP/PKA has been reported to have an inhibitory effect on the activation of ERKs stimulated by growth factors in many cell types such as Rat-1 cells, smooth muscle cells, Chinese hamster ovary cells, COS-7, and adipocytes (15)(16)(17)(18)(19). On the contrary, in some cell types such as PC12 cells, Swiss-3T3 cells, and S49 mouse lymphoma cells, cAMP activates ERKs and potentiates the effects of growth factors on differentiation and gene expression (20 -24). In human endothelial cells, down-regulation of the ␣ subunit of G s (G s␣ ) abolishes ␤-AR-mediated ERK activation by ISO (25), suggesting that G s␣ -dependent cAMP elevation and PKA activation are responsible for the activation of ERKs. We and others have also demonstrated that ␤-AR agonists including ISO significantly activate ERKs and increase protein synthesis through cAMP/PKA in cardiac myocytes (6,7). However, the molecular mechanisms of ␤-AR agonist-induced ERK activation remain largely unknown. In the present study, we examined the molecular mechanism of ISO-evoked activation of ERKs in the cardiomyocytes of neonatal rats. We observed that ISO-induced activation of ERKs is dependent on both G s /cAMP/ PKA and G i /Src/Ras signaling pathways. Quite recently, it has been reported that PKA activated through ␤-AR phosphorylates agonist-coupled ␤-AR, leading to a change of the receptor coupling from G s to G i (26). We examined whether phosphorylation of ␤-AR is critical to ISO-induced activation of ERKs also in cardiac myocytes.

EXPERIMENTAL PROCEDURES
Materials-[␥-32 P]ATP and [ 3 H]phenylalanine were purchased from NEN Life Science Products. Dulbecco's modified Eagle's medium, fetal bovine serum, and genistein were from Life Technologies, Inc. Pertussis toxin (PTX) was from List Biological Laboratories, Inc. Calphostin C was from BIOMOL. RpcAMP and H89 were from Biolog. Polyclonal antibodies against ␤ 1 -or ␤ 2 -AR, Shc, and Grb2 were from Santa Cruz Biotechnology Inc. Anti-hemagglutinin (HA) polyclonal antibody was from Mitsubishi Biochemical Laboratories (Japan). Other reagents were purchased from Sigma.
Cell Culture-Cardiac myocytes from ventricles of 1-day-old Wistar rats were isolated as described previously (27). Cardiac myocytes were plated at a field density of 10 5 cells/cm 2 on culture dishes with culture medium (Dulbecco's modified Eagle's medium with 10% fetal bovine serum). The culture medium was changed to serum-free Dulbecco's modified Eagle's medium at 48 h before treatment.
Transfection-After 24 h of plating cardiac myocytes on culture dishes, DNA was transfected by the calcium phosphate method as described previously (27). For each dish, 2.5 g of HA-ERK2 plasmid DNA was transfected with or without 7.5 g of other relevant plasmids such as ␤ 2 -ARwt, ␤ 2 -ARmut, ␤ARK1 495-689 , Csk, D.N.Ras, or D.N.Raf-1. After transfection, the cells were maintained in serum-free Dulbecco's modified Eagle's medium for 48 h before stimulation with ISO. The transfection efficiency of each experiment was ϳ1% in cardiac myocytes as assessed by LacZ staining after transfection of a LacZcontaining expression plasmid.
[ 3 H]Phenylalanine Incorporation-Protein synthesis was determined by assessing the incorporation of labeled phenylalanine from the extracellular medium into the total trichloroacetic acid-precipitable cell protein (7, 29 -32). Cardiac myocytes were serum deprived for 2 days, pretreated with or without a variety of inhibitory agents, and then incubated for 24 h with ISO. [ 3 H]Phenylalanine (1 Ci/ml) was added 2 h before the harvest. It has been reported that continuous long term (Ͼ24 h) labeling of cultured cardiomyocytes is intrinsically nonlinear, such that labeled amino acid accumulates asymptotically into protein (33). Because it has been reported that the specific radioactivities of the aminoacyl-tRNA pool may be equilibrated after 5 min and remain so over 2 h of labeling (29 -32, 34), pulse labeling for 2 h with [ 3 H]phenylalanine is sufficient to achieve adequate equilibration of the specific activity of phenylalanyl-tRNA. At the end of the labeling incubation, the plates were placed on ice, quickly washed twice with ice-cold phosphate-buffered saline, incubated for 30 min with 10% trichloroacetic acid, and washed. Precipitates were solubilized for 30 min in 1 M NaOH and neutralized, and total radioactivity was measured by liquid scintillation spectroscopy.
Assay of ERK Activity-ERK activities were measured using myelin basic protein (MBP)-containing gel as described previously (27). In brief, cell lysates were electrophoresed on an SDS-polyacrylamide gel containing 0.5 mg/ml MBP. ERKs in the gel were denatured in guanidine HCl, renatured in Tris-HCl containing Triton X-100 and 2-mercaptoethanol, and incubated with [␥-32 P]ATP. Phosphorylation activities of ERKs were assayed by subjecting to autoradiography. The activity of transfected HA-ERK2 was assayed as described previously (27). In brief, after transfection of HA-ERK2, cell lysates were incubated with an anti-HA polyclonal antibody. The immune complex was incubated with MBP and [␥-32 P]ATP for 10 min at 30°C. The sample was subjected to SDS-PAGE, and the phosphorylated MBP band was visualized by autoradiography.
Phosphorylation of Shc and Its Association with Grb2-Tyrosine phosphorylation of Shc and the association of Shc with Grb2 were examined by Western blot analysis as described previously (28). In brief, cell lysates were incubated with an anti-Shc or an anti-Grb2 antibody, and the immune complexes were subjected to SDS-PAGE and transferred to Immobilon-P membranes (Millipore). The membranes were immunoblotted with an anti-phosphotyrosine antibody (PY20) or an anti-Shc antibody, and the immunoreactivity was detected using the enhanced chemiluminescence reaction (ECL) system (Amersham Pharmacia Biotech) according to the manufacturer's directions.
Examination of ␤-ARs in Cardiac Myocytes-The amount of ␤-ARs on the cell surface was examined by Western blotting after dividing the membrane fraction and the cytoplasmic fraction. Briefly, cells were lysed by RIPA buffer (1 ϫ phosphate-buffered saline, pH 7.4, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 30 g/ml aprotinin). Whole lysate was first centrifuged at 200 ϫ g to remove nuclei, and then the supernatant was centrifuged at 15,000 ϫ g for 30 min at 4°C to pellet cell membrane. The pelleted membrane fractions were washed, subjected to SDS-PAGE, and then transferred to Immobilon-P membranes. After blocking with 30% non-fat dry milk, the membranes were incubated with an anti-␤ 1 -AR or an anti-␤ 2 -AR polyclonal antibody. The immunoreactivity was detected using the ECL system.
Analysis of Statistics-Statistical comparison was carried out within three or more groups using one-way analysis of variance and Dunnett's t test. Values of p Ͻ 0.05 were considered statistically significant.

cAMP-dependent PKA Mediates ISO-induced Activation of
ERKs-We have demonstrated recently that ␤-AR stimulation by NE activates ERKs through cAMP/PKA-dependent pathways in cardiac myocytes (7). We therefore defined the role of cAMP/PKA in ISO-induced ERK activation. When cAMP was inhibited by a cAMP analog, RpcAMP (100 M), ISO-evoked ERK activation was completely suppressed (see Fig. 1A). Pretreatment with a PKA inhibitor, H89 (10 M), also abolished ISO-induced activation of ERKs, whereas inhibition of PKC by calphostin C (1 M) did not affect the ERK activation by ISO (Fig. 1A). On the other hand, PHE-induced activation of ERKs was suppressed by calphostin C but not by RpcAMP or H89 (Fig. 1B). These results suggest that cAMP-dependent PKA activation probably through G s is required for ISO-induced activation of ERKs in cardiac myocytes.
G i Protein Is Involved in ISO-induced Activation of ERKs-We next examined the possibility that G i protein is involved in ISO-induced ERK activation because ␤-AR was reported to bind G i (35,36), and activation of G i protein can activate ERKs in many cell types (37,38). We preincubated cardiac myocytes with 100 ng/ml PTX and then treated them with ISO or PHE. Activation of ERKs by ISO was abolished by the PTX treatment for 24 h, whereas PTX had no effect on PHE-induced ERK activation ( Fig. 2A). These results suggest that PTX-sensitive G i protein is responsible for ISO-but not PHE-induced activation of ERKs in cardiac myocytes.
It has been reported recently that stimulation of G i proteincoupled receptors activates ERKs by ␤␥ subunits (G ␤␥ ) but not by the ␣ subunit (39 -41). Therefore, the role of G ␤␥ in ISOinduced ERK activation was examined by introducing HA-ERK2 and a minigene construct encoding ␤ARK1 495-689 polypeptides, which inhibit G ␤␥ -dependent activation of a wide variety of cell regulatory processes (39,40) into cultured cardiac myocytes. ISO increased the activity of the transfected ERK2, and overexpression of ␤ARK1 495-689 completely suppressed the ISO-induced activation of ERK2 (Fig. 2B). These findings suggest that ISO activates ERKs via the G ␤␥ -depend-ent pathway in cardiac myocytes.
Inhibition of G i by PTX Does Not Affect Abundance of ␤-ARs-There is a possibility that blockade of G i may lead to sustained stimulation of adenylyl cyclase and may induce sustained phosphorylation and desensitization (and perhaps down-regulation) of ␤-ARs (35). It has also been reported that In brief, cell lysates were electrophoresed on an SDS-polyacrylamide gel containing 0.5 mg/ml MBP. ERKs in the gel were denatured in guanidine HCl and renatured in Tris-HCl containing Triton X-100 and 2-mercaptoethanol. Phosphorylation activities of ERKs were assayed by incubating the gel with [␥-32 P]ATP. After incubation, the gel was washed extensively and subjected to autoradiography. Representative autoradiograms of MBP phosphorylation are shown. The intensities of 42-kDa ERK band were measured by densitometric scanning of the autoradiograms. The activity was expressed relative to that of 42-kDa ERK obtained from unstimulated cardiomyocytes. The data are indicated as the mean Ϯ S.E. of three independent experiments. *p Ͻ 0.01 versus control.
PTX increases the sensitivity of ␤-AR to ISO (42). To test whether pretreatment of cardiac myocytes with PTX might affect the density of ␤-ARs on the cell membrane, we examined the amount of cell surface ␤-ARs, ␤ 1 -AR and ␤ 2 -AR, by Western blot analysis after long term treatment with PTX. As shown in Fig. 3, PTX did not affect the protein levels of ␤ 1 -AR and ␤ 2 -AR in membrane fractions. These results suggest that the inhibition of ISO-induced ERK activation by PTX is not caused by down-regulation of ␤-ARs.
Tyrosine Kinases Including Src Family Tyrosine Kinases Modulate ISO-induced Activation of ERKs-G ␤␥ has been reported to activate ERKs through non-receptor-type tyrosine Ci of [␥-32 P]ATP for 10 min. Aliquots of the reaction mixture were subjected to SDS-PAGE, and the gel was washed, dried, and subjected to autoradiography. A representative autoradiogram is shown. Relative kinase activities of 42-kDa ERK were determined by scanning each band with a densitometer. Activities were expressed relative to those of 42-kDa ERK obtained from unstimulated cardiomyocytes. The results are presented as the mean Ϯ S.E. from three independent experiments. *p Ͻ 0.01 versus control. kinases including Src family tyrosine kinases in many types of cells (28,43). We examined whether tyrosine kinases are responsible for activation of ERKs induced by ISO in cardiac myocytes. Activation of ERKs by ISO was suppressed completely when cardiomyocytes were pretreated with a broad spectrum tyrosine kinase inhibitor, genistein, whereas the activation of ERKs by PHE was not affected by the same pretreatment (Fig. 4A). These suggest that tyrosine kinases are involved in ISO-induced activation of ERKs in cardiac myocytes, whereas PHE-induced activation of ERKs is not dependent on the tyrosine kinase pathway.
We examined further the role of Src family tyrosine kinases in ISO-induced ERK activation in cardiac myocytes. Csk has been reported to phosphorylate the tyrosine residue in the carboxyl terminus of Src family protein kinases and thereby inactivate their function (44,45). We cotransfected Csk with ERK2 into cardiomyocytes and examined the activity of the transfected ERK2 after stimulation with ISO or PHE. Although cotransfection of Csk did not affect PHE-induced ERK2 activation, overexpression of Csk completely inhibited activation of ERK2 by ISO (Fig. 4B), suggesting that Src family tyrosine kinases are also involved in ISO-induced ERK activation in cardiac myocytes.
ISO Enhances Tyrosine Phosphorylation of Shc and Association of Shc with Grb2-Adapter proteins containing Src homology 2 domains such as Shc and Grb2 transduce activation of tyrosine kinases to the Ras/ERK pathway via the guanine nucleotide exchange factor Sos (46 -48). We therefore examined whether Shc is activated by ISO in cardiac myocytes. ISO rapidly (within 30 s) increased levels of tyrosine phosphorylation of 52-kDa Shc (Fig. 5A). Phosphorylation levels decreased from 5 min and returned to the basal levels by 15 min. A faint band corresponding to 46-kDa Shc was also observed with long exposure (data not shown). Next, association of Grb2 with Shc was examined by immunoprecipitation with anti-Grb2 antibody and immunoblotting with an anti-Shc antibody. The intensities of the bands around 52 kDa and 46 kDa corresponding to Shc were enhanced by ISO stimulation (Fig. 5B), suggesting that the 52-and 46-kDa Shc form a complex with Grb2 after ISO stimulation.
Ras and Raf-1 Kinase Activation Is Essential to ISO-induced ERK Activation-Association of Grb2 with Shc usually results in the recruitment of a Ras activator Sos to the membrane fraction, leading to activation of Ras (46 -48). The role of Ras was next analyzed in ISO-stimulated cardiomyocytes. ISOinduced activation of ERKs was suppressed by overexpression of D.N.Ras (Fig. 6A), suggesting that activation of Ras is re-quired for ISO-induced activation of ERKs in cardiac myocytes. Activated Ras usually induces activation of ERKs through Raf-1 kinase and mitogen-activated protein kinase/ERK kinase (49). Activation of the transfected ERK2 by ISO or PHE was abolished by the overexpression of D.N.Raf-1 kinase (Fig. 6B), indicating that Raf-1 kinase is crucial for activation of ERKs by ISO and PHE in cardiac myocytes.
Activation of ERKs by ISO Requires PKA-dependent Phosphorylation of ␤-AR-We next examined how both G s /PKA-and G i /Ras-dependent pathways are involved in ISO-induced activation of ERKs. Quite recently Daaka et al. (26) have reported that phosphorylation of ␤ 2 -AR by PKA changes coupling of the receptor from G s to G i protein. We therefore examined this possibility by introducing the gene of ␤ 2 -ARmut, which lacks phosphorylation sites for PKA, into cardiomyocytes. ISO-induced activation of ERK2 was completely suppressed by overexpression of ␤ 2 -ARmut (Fig. 7). In contrast, overexpression of ␤ 2 -ARwt enhanced the activation of ERK2 by ISO. These results suggest that PKA-dependent phosphorylation of ␤-AR is necessary to evoke signals from ␤-AR to ERKs in cardiac myocytes.

ISO-induced Protein Synthesis Also Occurs through Both G sand G i -dependent Signal Transduction Pathways-ISO acti-
vates ERKs through both G s /PKA-and G i /Ras-dependent pathways. To determine whether cardiomyocyte hypertrophy is also induced through the same signal transduction pathways, we examined the effects of various inhibitory agents for signaling molecules on the ISO-induced increase in protein synthesis. We pretreated cardiac myocytes with PTX, genistein, manumycin (a Ras farnesyltransferase inhibitor), H89, or RpcAMP and stimulated the cells with ISO for 24 h. Although the pretreatment with these inhibitors alone did not affect the basal level of protein incorporation significantly (Fig. 8A), an ISO-induced increase in protein synthesis was suppressed significantly by all of these inhibitors (Fig. 8B), indicating that ISO enhanced protein synthesis through cAMP/PKA, G i , tyrosine kinases, and Ras. These results suggest that the signal transduction pathway leading to activation of ERKs is also important for ISO-induced protein synthesis in cardiac myocytes. DISCUSSION Stimulation of ␤-AR usually activates an effector enzyme adenylyl cyclase through G s . The activation of adenylyl cyclase induces an increase in cAMP levels, which in turn activates PKA (14,19,50,51). In addition, a previous report indicated that down-regulation of G s␣ abolishes ISO-induced ERK activation in human endothelial cells (25), suggesting that G s␣ is required for activation of ERKs by ␤-AR. Our previous (7) and present studies also showed that ISO activates ERKs through cAMP/PKA in cardiac myocytes, supporting that G s protein plays an essential role in the activation of ERKs. Many laboratories have reported that activation of cAMP-dependent PKA inhibits activation of the Raf-1 kinase/ERK cascade in various cell types such as Rat-1 cells, smooth muscle cells, Chinese hamster ovary cells, COS-7 cells, and adipocytes (16 -19). In other cell types such PC12 cells, S49 mouse lymphoma cells, and Swiss-3T3 cells, however, cAMP activates ERKs and potentiates the effects of growth factors on differentiation and gene expression (20 -24). Taken together, cAMP-dependent PKA activation may have different effects on ERKs among cell types, and PKA may activate ERKs in cultured cardiac myocytes.
By examining the signal transduction pathway of ISO-induced activation of ERKs, we found that the G i protein/Src/Ras pathway is also required for ISO-induced ERK activation. How do two different pathways, cAMP/PKA pathway and G i /Src/Ras pathway, converge at the ERK cascade? G s -coupled ␤ 2 -AR was reported to activate simultaneously the pathway that leads to Relative kinase activities of 42-kDa ERK were determined by scanning each band with a densitometer. Activities were expressed relative to that obtained from unstimulated cardiomyocytes. Data are presented as the mean Ϯ S.E. from three independent experiments. *p Ͻ 0.01 versus control.
functional inhibition of cAMP/PKA pathway via G i protein in cardiac cells (36). Moreover, it has been reported recently that PKA-induced phosphorylation of ␤-AR changes the coupling of the receptor from G s to G i and activates ERKs through the Src/Ras pathway in HEK293 cells (26). We therefore examined the role of G i protein in ISO-evoked ERK activation and found that G i protein is also essential to ISO-induced activation of ERKs in cardiac myocytes. Although it has been reported that inhibition of G i protein may lead to sustained stimulation of adenylyl cyclase and down-regulation of ␤-ARs in some types of cell (35), Western blot analysis revealed that blocking of G i by PTX did not affect the abundance of ␤-ARs in the membrane fraction of cardiac myocytes, suggesting that in cardiac myocytes, the inhibitory effect of PTX is attributable to suppression of signaling from ␤-ARs to downstream but is not caused by down-regulation of ␤-ARs.
It is necessary to define how G proteins are involved in the ISO-induced ERK activation in cardiac myocytes. It has been reported in COS-7 cells that activation of ␤-AR activates ERKs through G s␤␥ but not G s␣ . G s␤␥ activates ERKs through a Rasdependent pathway, whereas G s␣ inhibits activation of ERKs via cAMP/PKA. The balance between these two opposite mechanisms of regulation may control the ERK response to ␤-AR agonists (19). However, this model may not be applied to cardiac myocytes because ISO-stimulated ERK activation is dependent on cAMP/PKA, suggesting the importance of G s␣ protein. It is also possible that ISO activates ERKs through both G s and G i proteins independently. In the present study, however, because inhibition of cAMP/PKA or of G i completely suppressed ISO-induced ERK activation, it may be unlikely that G s and G i proteins are involved independently. We introduced ␤ 2 -ARmut lacking phosphorylation sites for PKA with HA-ERK2 into cardiac myocytes and showed that inhibition of ␤ 2 -AR phosphorylation inhibits ISO-induced activation of ERKs. These results are consistent with the hypothesis that activation of cAMP-dependent PKA phosphorylates the activated receptor, leading to the change in coupling of the receptor from G s to G i protein.
G ␤␥ subunit derived from PTX-sensitive G i protein regulates many effectors within the cell (22,41,52,53). Stimulation of various receptors such as ␣ 2A -adrenergic, M 2 muscarinic acetylcholine, D 2 dopamine, A 1 adenosine, or angiotensin II type 1 receptors induces Ras-dependent ERK activation via the G ␤␥ subunit in COS-7 cells (53). By introducing the ␤ARK1 495-689 polypeptide minigene, we showed in the present study that the G ␤␥ subunit is required for ISO-induced ERK activation in cardiac myocytes. There is a report showing that the ␤␥ subunit of G s protein activates ERKs in COS-7 cells (19). Because activation of adenylyl cyclase is usually mediated by the ␣ subunit rather than the ␤␥ subunit of G s (14,24,25,50,51) and ISO activated ERKs through a cAMP/PKA-dependent pathway in this study, ISO-induced activation of ERKs might be dependent on the ␣ subunit of G s and on the ␤␥ subunit of G i in cardiac myocytes.
The ␤␥ subunit of G i protein activates ERKs through tyrosine kinases including Src family tyrosine kinases (43,54). Because ␤-AR itself does not possess tyrosine kinase activity, non-receptor-type tyrosine kinases may be responsible for ISOinduced ERK activation. We therefore examined the involvement of tyrosine kinases including Src family tyrosine kinases in ISO-induced ERK activation in cardiac myocytes. Pretreatment with a tyrosine kinase inhibitor or overexpression of Csk strongly inhibited ISO-induced activation of ERKs in cardiac myocytes. In contrast, PHE-induced activation of ERKs was not dependent on the tyrosine kinase pathway. We have reported that angiotensin II activates ERKs through PKC but not through tyrosine kinases in cardiac myocytes (27). These results collectively suggest that tyrosine kinases including Src family tyrosine kinases mediated G i -coupled receptor-induced ERK activation, whereas PKC but not tyrosine kinases play a critical role in activation of ERKs by G q -coupled receptors in cardiac myocytes.
Activation of tyrosine kinases leads to Ras activation through adaptor proteins such as Shc and Grb2 and the guanine exchange factor Sos (46 -48). It has been demonstrated that Shc is tyrosine phosphorylated in response to angiotensin II in cardiac myocytes (55) and that Shc serves as a converging target in the growth factor-and G protein-coupled receptorstimulated signal transduction events resulting in activation of Ras protein (56). The present results indicated that ISO rapidly induces phosphorylation of Shc and association of Shc with Grb2, which might result in the translocation of the Ras guanine nucleotide exchange factor Sos to the membrane fraction. It has been recognized that Ras plays a key role in a variety of cell functions through the sequential activation of Raf-1 kinase and ERKs (57,58). On the other hand, it has been reported that cAMP activates ERKs through a B-Raf-and a PKA-activated Rap1-dependent pathway, but not through Ras/Raf-1 kinase-dependent pathway in PC12 cells (24,59). It has also been shown that cAMP activates ERKs independently of Raf-1 kinase (60). By contrast, we observed in the present study that Ras and Raf-1 kinase are necessary for ISO-induced ERK activation. Taken collectively, these findings suggest that ISO may activate ERKs through the signaling pathway consisting of Src, Shc, Grb2, Sos, Ras, and Raf-1 kinase in cardiac myocytes.
As reported previously, ERK activation is required for cell growth including cardiac hypertrophy (6,7,34,61,62). We also showed in the present study that the signal transduction pathway leading to activation of ERKs is important for ISO-induced protein synthesis in cardiac myocytes. The effect of isoproterenol on the index of net protein synthesis is small compared with severalfold increases in the activity of each signaling steps. Although there are great differences between signaling steps and protein synthesis in many aspects such as the basal levels, the responsive ability to stimulation, and the methods of examination among them, it is also possible that the initiation of protein synthesis needs a much greater fold-increase in the activity of signaling molecules. Many previous studies also showed that signaling steps such as ERKs are activated more than severalfold by stimulation, whereas an increase in the protein synthesis rate is less than 50% (6,7,62,63). Although it is unknown at present why there are big differences in the time course for protein synthesis and activation of signaling molecules, there are several possibilities. Protein synthesis may need activation of many signaling steps, or the initiation step of protein synthesis needs some time period after activation of molecules responsible to protein synthesis. Further studies are necessary to determine these possibilities.
In summary, the ␤-AR-evoked signal transduction pathways to activation of ERKs are different among cell types. In cardiac myocytes, ISO activates ERKs through the signal transduction pathway consisting of ␤-AR phosphorylation by G s /cAMP-dependent PKA, G ␤␥ subunits derived from G i , Src family tyrosine kinases, the formation of the Shc-Grb2-Sos complex, Ras, and Raf-1 kinase. We showed in this study that the phosphorylation of ␤-AR by PKA and the change of coupling of the receptor from G s to G i play an important role in ISO-induced cardiomyocyte hypertrophy.