Gα12/13-mediated Production of Reactive Oxygen Species Is Critical for Angiotensin Receptor-induced NFAT Activation in Cardiac Fibroblasts*

Angiotensin II (Ang II) activates multiple signaling pathways leading to hyperplasia of cardiac fibroblasts. Reactive oxygen species (ROS) produced by Ang II stimulation are assumed to play pivotal roles in this process. Here, we show that ROS mediate Ang II-induced activation of nuclear factor of activated T cells (NFAT) in rat cardiac fibroblasts. Ang II-induced NFAT activation was suppressed by diphenyleneiodonium (an NADPH oxidase inhibitor), dominant negative (DN)-Rac, DN-p47phox, and an inhibitor of Gα12/13 (Gα12/13-specific regulator of G protein signaling domain of p115RhoGEF, p115-regulator of G protein signaling (RGS)). Stimulation of Ang II receptor increased the intracellular ROS level in a Rac- and p47phox-dependent manner. Because p115-RGS suppressed Ang II-induced Rac activation, Ang II receptor-coupled Gα12/13 mediated NFAT activation through ROS production by Rac activation. Ang II-induced nuclear translocation of the green fluorescent protein (GFP)-tagged amino-terminal region of NFAT4 (GFP-NFAT4) was suppressed by p115-RGS or BAPTA but not by diphenyleneiodonium. The expression of constitutively active (CA)-Gα12/13, CA-G translocation α13, or CA-Rac increased the nuclear of GFP-NFAT4. These results suggest that NFAT activity is regulated by both Ca2+-dependent and ROS-dependent pathways. Furthermore, activation of c-Jun NH2-terminal kinase (JNK) induced by Ang II stimulation is required for NFAT activation because Ang II-induced NFAT activation was inhibited by SP600125, a selective JNK inhibitor. These results indicate that Ang II stimulates the nuclear translocation and activation of NFAT by integrated pathways including the activation of Gα12/13, Rac, NADPH oxidase, and JNK and that Gα12/13-mediated ROS production is essential for NFAT transcriptional activation.

In the normal heart, two-thirds of the cell population is composed of non-muscle cells, most of which are fibroblasts (1). The abnormal proliferation of cardiac fibroblasts with excessive accumulation of extracellular matrix proteins is one of the features of left ventricular remodeling, which leads ultimately to cardiac dysfunction (2,3). Ang II 1 is one of humoral factors that affect phenotype and function of cardiac fibroblasts (4,5). Ang II has been demonstrated to stimulate proliferation and induce the expression of genes of collagens, fibronectins, and integrins (6 -8). Receptor stimulation by Ang II is assumed to activate G q and G i and turns on various signaling cascades dependent on cell types. Ang II exerts cardiac hypertrophy and hyperplasia of cardiac fibroblasts by activating a number of intracellular signal transduction pathways through AT1R.
ROS such as hydrogen peroxide and oxygen radicals play various roles in living cells as a second messenger to elicit physiological responses or a toxic intermediate leading to cellular damage (9). Recent studies suggest that ROS work as a regulator of signal transduction (10,11). We have reported heterotrimeric G proteins G i and G o as putative target molecules of ROS (12,13). Although Ang II produces ROS in vascular smooth muscle cells, cardiac myocytes, and cardiac fibroblasts (14 -16), the molecular mechanism of ROS production and the target molecules of ROS are largely unknown.
NFAT was originally described as a transcription factor expressed in activated but not resting T cells (17)(18)(19). Growing evidence indicates that NFAT is not only a T cell-specific transcriptional factor but also is expressed in a variety of cells, including cardiac myocytes and cardiac fibroblasts (18,19). The relevance of the NFAT signaling pathway to cardiac hypertrophy is underscored by the observation that cardiac-targeted transgenic animals expressing constitutively activated forms of either calcineurin or NFAT produced ventricular hypertrophy * This work was supported in part by a research grant (to M. N. and H. Kurose) from the Ministry of Education, Science, Sports, and Culture of Japan and in part by a grant (to M. N.) from the Kurozumi Medical Foundation, the Nakajima Memorial Foundation, and the Mochida Memorial Foundation for Medical and Pharmaceutical Research. 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. (20,21). NFAT family proteins have three functional domains; they are the Rel-similarity domain, which is responsible for DNA binding activity and interaction with AP-1, the NFAThomology region, which regulates intracellular localization, and the transcriptional activation domain (19,22). The activation of NFAT includes dephosphorylation, nuclear translocation, and an increase in affinity for DNA binding. Because stimuli that elicit Ca 2ϩ mobilization result in rapid dephosphorylation of NFAT proteins by calcineurin and their translocation to nucleus, the Ca 2ϩ -calcineurin signaling pathway primarily regulates NFAT activity. On the other hand, recent reports indicate that ROS mediates NFAT activation induced by vanadium or nickel compounds (23,24). Although AT1R stimulation is sufficient to induce NFAT activation (21), it has not been examined whether ROS works as a mediator of NFAT activation. G 12 family G proteins G 12 and G 13 couple with various G protein-coupled receptors and mediate physiological responses by interacting with different signaling proteins (25). The role of G 12/13 in the heart, however, has not been revealed because of the absence of a specific inhibitor. Recent studies demonstrated that p115RhoGEF has an RGS domain specific for ␣ subunits of G␣ 12/13 (26,27), and this RGS domain works as an inhibitor of G␣ 12/13 (28 -30). We examine whether G␣ 12/13 is involved in Ang II-mediated signal transduction pathway in cardiac fibroblasts by using RGS domain of p115RhoGEF. In the present study we demonstrate that Ang II-induced NFAT activation requires ROS and that Ang II-induced ROS production is mediated by activation of G 12/13 .

EXPERIMENTAL PROCEDURES
Materials and Plasmid Construction-AT1R blocker CV11974 was provided from Takeda Chemical Industries Ltd. (Osaka, Japan). BAPTA-AM, AG1478, and JNK inhibitor II (SP600125) were purchased from Calbiochem. Angiotensin II, DPI, N-acetyl-L-cysteine, catalase, PD123319, carbonyl cyanide m-chlorophenylhydrazone, rotenone, nigericin, sodium nitroprusside dihydrate, N-nitro-L-arginine methyl ester hydrochloride, and Tiron were purchased from Sigma. Fura2/AM was from Dojindo (Kumamoto, Japan). Collagenase and FuGENE 6 were from Roche Applied Science. Dual-luciferase reagents were from Promega. 2Ј,7Ј-Dichlorofluorescein diacetate and dihydroethidium were from Molecular Probes. Glutathione-Sepharose beads were from Amersham Biosciences. Rat p47 phox was cloned from the rat heart, and the sequence was confirmed (31). The plasmid coding Prx II was provided by Dr. Sue Goo Rhee (National Institutes of Health) (32). The entire coding region of Prx II and p47 phox was amplified, sequenced, and then used for adenovirus production. The CRIB domain of p21-activated kinase was cloned from mouse brain, sequenced, and ligated into pGEX-4T-1 to make glutathione S-transferase fusion protein construct. pN-FAT-Luc and pRL-SV40 were from Stratagene. The GFP-tagged aminoterminal region of NFAT4 (GFP-NFAT4) was constructed as described (33). The cDNAs coding wild type and CA-Ras, -Rho, and -Rac1 were provided by Dr. Kozo Kaibuchi (Nagoya University, Japan). Horseradish peroxidase-conjugated anti-rabbit IgG or anti-mouse IgG antibodies were from Santa Cruz Biotechnology. Anti-Rac1 antibody was from Transduction Laboratories.
Cell Culture-Cardiac fibroblasts were prepared from ventricles of 1ϳ2-day-old Sprague-Dawley rats (12). Briefly, after digestion of ventricles with 0.1% collagenase, isolated fibroblasts were plated on a non-coated dish in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum and 50 units/ml penicillin/streptomycin. Subconfluent cells were serum-starved for 48 h and used for the experiments.
Considering the possibility that cardiac fibroblasts may lose the original characteristics after prolonged culture, cells were used within two passages in all experiments.
Measurement of NFAT Activity-After 2 h of adenoviral infection (100 m.o.i.) in serum-free medium, fibroblasts (3 ϫ 10 5 cells) plated on 24-well dishes were transiently co-transfected with 0.45 g of pNFAT-Luc and 0.05 g of pRL-SV40 control plasmid using FuGENE 6 (36). The receptor stimulation was performed 48 h after transfection, and luciferase activity was measured 6 h after stimulation with dual luciferase reagents. For measuring the translocation of GFP-NFAT4 (NFAT4 is also referred as NFAT-c3), cells (1 ϫ 10 5 ) plated on glassbottom 35-mm dishes were transfected for 48 h with 0.5 g of GFP-NFAT4 fusion protein construct and 0.5 g of construct encoding CA-G␣ or CA-small GTPase (indicated in Figs. 6 and 7). After receptor stimulation for 15 min by Ang II (100 nM), cells were fixed by 10% formaldehyde neutral buffer solution. The localization of GFP-NFAT4 was measured at an excitation wavelength of 488 nm with a laser scanning confocal imaging system (Carl Zeiss LSM510). More than 60 cells were scanned in each experiment and quantified the subcellular localization of GFP-NFAT4 using Photoshop.
Determination of JNK Activation-Activity of JNK was determined as described (28,29). Briefly, cells on 60-mm dishes (3 ϫ 10 6 cells) were lysed in radioimmune precipitation assay buffer. Supernatant was incubated with anti-JNK antibody and protein A-Sepharose beads for 120 min at 4°C. The beads were washed and finally suspended in SDS sample buffer. The proteins were resolved by SDS-polyacrylamide gel electrophoresis, and radioactive bands of glutathione S-transferase-c-Jun were quantified using film-less autoradiographic analysis (BAS-2000 system, FUJIFILM).
Quantification of Intracellular ROS by Ang II Receptor Stimulation-ROS production was measured with a fluorescent dye, 2Ј,7Ј-dichlorofluorescein diacetate, as described (32). Cells were plated on glass-bottom 35-mm dishes and loaded with 2Ј,7Ј-dichlorofluorescein diacetate (10 M) in the cultured medium at 37°C for 10 min. The DCF fluorescence at an emission wavelength of 510 nm was observed at room temperature by exciting DCF at 488 nm using a video image analysis system (Aquacosmos, Hamamatsu Photonics).
Pull-down Assay-Activation of Rac was measured by the method of Ren and Schwartz (37) with a slight modification. Cells were stimulated by Ang II and lysed in buffer containing 50 mM Tris (pH 7.2), 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, 500 mM NaCl, 10 mM MgCl 2 , 10 g/ml leupeptin, 10 g/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride, 4 g glutathione S-transferase-PAK-CRIB. The supernatant was incubated with glutathione-Sepharose beads for 120 min at 4°C. The beads were washed and finally suspended in SDS sample buffer. Pull downed Rac was detected with anti-Rac1 antibody.

ROS Mediate Ang II-induced NFAT Activation in Cardiac
Fibroblasts-To define the signal transduction cascade leading to NFAT activation by angiotensin receptor stimulation, concentration dependence and time course of NFAT activation were first determined by NFAT-dependent luciferase assay. Stimulation of cardiac fibroblasts with Ang II resulted in a concentration-dependent activation of NFAT, which was peaked at 6 h (Figs. 1, A and B). This NFAT activation was completely suppressed by an intracellular Ca 2ϩ chelator BAPTA-AM, indicating that a Ca 2ϩ -activated signaling is re-quired for NFAT activation. The Ang II-induced NFAT activation was also completely suppressed by two chemical antioxidants or anti-oxidizing enzyme; that is, N-acetyl-L-cysteine (a radical scavenger), DPI (an NADPH oxidase inhibitor), or catalase (an H 2 O 2 -degrading enzyme) (Fig. 1C). The Ang II-induced NFAT activation was also suppressed by the expression of Prx II, an H 2 O 2 -scavenging enzyme (Fig. 1D). The involvement of ROS in Ang II-induced NFAT activation is further supported by the finding that the exposure of H 2 O 2 to cardiac fibroblasts resulted in a significant increase in NFAT activity (Fig. 1E) (Supplemental Fig. 1). These results suggest that ROS mediate Ang II-induced NFAT activation without the increase in [Ca 2ϩ ] i . G␣ 12/13 Mediate Ang II-induced NFAT Activation-We determined which subtype of angiotensin receptor is responsible for NFAT activation. AT1R blocker CV11974 but not AT2R blocker PD123319 inhibited Ang II-induced NFAT activation, indicating that AT1R mediates NFAT activation ( Fig. 2A). To examine which G proteins are involved in Ang II-induced NFAT activation, an RGS domain specific for G␣ q (GRK2-RGS) and G␣ 12/13 (p115-RGS) was used. Ang II-induced Ca 2ϩ release was significantly suppressed by the expression of GRK2-RGS but not by p115-RGS (Fig. 2B). Ang II-induced NFAT activation was completely inhibited by the expression of p115-RGS but not by the expression of GRK2-RGS (Fig. 2C). These results suggest that G␣ q mediates Ang II-induced Ca 2ϩ mobilization, but G␣ q -mediated increase in [Ca 2ϩ ] i is not enough for NFAT activation. In addition, the expression of CA-G␣ q did not increase NFAT activity, but the expression of CA-G␣ 12 or CA-G␣ 13 signifi-cantly increased NFAT activity (Fig. 2D). These results suggest that stimulation of AT1R activates G␣ 12/13 , leading to NFAT activation.
G␣ 12/13 Mediate Ang II-induced ROS Production-Because NFAT activity was increased by the exposure of H 2 O 2 ( Fig. 1), we investigated the involvement of G␣ 12/13 in Ang II-induced ROS production. The addition of Ang II gradually increased the DCF fluorescence intensity without morphological changes, indicating ROS production by Ang II stimulation (Fig. 3A). The maximal ROS concentration generated by Ang II was about 2 M, as compared with the fluorescence intensity generated by exogenously added H 2 O 2 . DCF fluorescence probe was increased by reacting with ROS but was also sensitive to reactive nitrogen species. The increase in DCF fluorescence by Ang II stimulation was not suppressed by N-nitro-L-arginine methyl ester hydrochloride, an inhibitor of NO synthase, but was suppressed by Tiron, a superoxide scavenger (Supplemental Fig.  2A). Because the exposure of sodium nitroprusside dihydrate, an NO donor, revealed a small increase in DCF fluorescence (Supplemental Fig. 2B), DCF was more sensitive to ROS than reactive nitrogen species. Furthermore, the Ang II-induced ROS production was also observed by using a superoxide-specific fluorogenic probe, dihydroethidium (Supplemental Fig.  2C). These results strongly suggest that the increases in DCF fluorescence induced by Ang II are mainly caused by ROS. The Ang II-induced ROS production was inhibited by p115-RGS but not by GRK2-RGS (Fig. 3, B and C). The increase in DCF fluorescence intensity by AT1R stimulation was not due to changes in intracellular pH or [Ca 2ϩ ] i , as the treatment with a H ϩ ionophore (nigericin or carbonyl cyanide m-chlorophenylhydrazone) or Ca 2ϩ chelator BAPTA-AM and the exclusion of extracellular Ca 2ϩ did not affect Ang II-induced ROS production (Fig. 3D). These results suggest that the origin of Ang II-induced ROS production is not mitochondria, and G␣ 12/13 mediate ROS production.
Rac and p47 phox Mediate ROS Production and NFAT Activation-Because DPI suppressed Ang II-induced NFAT activation (Fig. 1C), we suspected that NADPH oxidase plays a crit- ical role in Ang II-induced ROS production and NFAT activation. Rac is one of the small GTP-binding proteins that participate in ROS production by activating NADPH oxidase in neutrophils (39). Recent studies have suggested that Rac also mediates NADPH oxidase activation in vascular smooth muscle cells (14,15). To examine the involvement of NADPH oxidase in Ang II-induced ROS production in cardiac fibroblasts, DN-Rac and DN-p47 phox are used. Both GTP-bound Rac and -phosphorylated p47 phox are necessary for activation of NADPH oxidase. Because phosphorylation of Ser 304 and Thr 305 of rat p47 phox is required for activation of NADPH oxidase, alanine substitution of these residues results in a dominant negative mutant of p47 phox . Ang II-induced ROS production and NFAT activation were suppressed by DN-Rac or DN-p47 phox (Fig. 4, A  and B). The blocking effect of DN-Rac or DN-p47 phox on NFAT activation is consistent with the result that NFAT activation is dependent on ROS production. However, DN-Rac or DN-p47 phox did not affect Ang II-induced Ca 2ϩ release (Fig. 4C). Because DN-Rac suppressed NFAT activation that was mediated by G␣ 12/13 , we examined whether Rac activation was suppressed by p115-RGS or GRK2-RGS. Ang II stimulation increased Rac activity by 3-fold, as determined by pull-down assay, which was completely suppressed by p115-RGS but not by GRK2-RGS. However, Ang II-stimulated Rac activation was not suppressed by AG1478, an EGF receptor kinase inhibitor (Fig. 4D). It indicates that EGF receptor transactivation by Ang II stimulation is not involved in Rac activation. These results suggest that AT1R-coupled G␣ 12/13 stimulate ROS production through Rac-mediated activation of NADPH oxidase in cardiac fibroblasts.
Ang II Stimulation Induces Nuclear Localization of NFAT in a ROS-independent Pathway-It is believed that NFAT family members move into the nucleus by the mechanism of persistent elevation of [Ca 2ϩ ] i and continuous activation of calcineurin (18). Therefore, we examined the involvement of G␣ 12/13 or ROS in NFAT translocation. Because it has been reported that pressure overload-and Ang II-induced cardiac hypertrophy are attenuated in NFAT4-null mice (40), the NFAT4 construct was used in the present study. Transfection of GFP-NFAT4 into cardiac fibroblasts revealed that, under basal conditions, the fusion protein was distributed evenly in the cytosol (Fig. 5A). The stimulation with Ang II (100 nM) for 15 min resulted in a shift of GFP-NFAT4 localization from the cytosol to the nucleus. This GFP-NFAT4 translocation was not affected by DPI. The treatment with ionomycin, a Ca 2ϩ ionophore, dramatically increased the nuclear localization of GFP-NFAT4, which was not affected by DPI. Exposure of H 2 O 2 did not affect GFP-NFAT4 localization. These results suggest that nuclear translocation of NFAT is predominantly regulated by [Ca 2ϩ ] i and that ROS does not participate in NFAT translocation by Ang II stimulation. Ang II-induced but not ionomycin-induced translocation of GFP-NFAT4 was attenuated in the p115-RGS-expressing cells (Fig. 5B). Because p115-RGS did not affect an Ang II-induced increase in [Ca 2ϩ ] i (Fig. 2B), these results suggest that G␣ 12/13 -mediated NFAT translocation requires [Ca 2ϩ ] i at the resting level. Fig. 5B also shows that GRK2-RGS did not affect the translocation of GFP-NFAT4 to the nucleus (Fig. 5B). This result suggests that the transient increase in [Ca 2ϩ ] i (or Ca 2ϩ release) by Ang II stimulation is not enough for the translocation of GFP-NFAT. The translocation of GFP-NFAT4 fusion protein in nucleus was quantitatively determined by calculating the percentage of GFP-NFAT4 in the nucleus from the percentage of the fusion protein locating in the cytosol and the nucleus of individual cells (it is referred here as nuclear predominant fluorescence). Compared with the basal conditions (13 Ϯ 1%), Ang II significantly increased the nuclear predominant fluorescence of GFP-NFAT4 (33 Ϯ 2%). The expression of p115-RGS but not the treatment with DPI suppressed the nuclear translocation of GFP-NFAT4 (Fig. 5C). The action of p115-RGS is specific, as p115-RGS did not affect ionomycin-induced the nuclear translocation of GFP-NFAT4 (Fig. 5C). These results substantiate confocal images of GFP-NFAT4 translocation.
Activated G␣ 12/13 Proteins Induce Nuclear Localization of NFAT4 -The expression of CA-G␣ 12 and CA-G␣ 13 , but not CA-G␣ q , resulted in a 3-5-fold increase in nuclear predominant fluorescence of the GFP-NFAT4 fusion protein (Fig. 6, A and  B). This result is correlated with the increase in NFAT activity after expression of each CA-G␣ proteins (Fig. 2D). The shift in GFP-NFAT4 localization by the expression of CA-G␣ 13 was sensitive to [Ca 2ϩ ] i , as GFP-NFAT4 located predominantly in the cytosol by the treatment with BAPTA-AM (Fig. 6C). As expected, DPI did not change the distribution of GFP-NFAT4 in CA-G␣ 13 -expressing cells. These results suggest that G␣ 12/ 13-promoted nuclear translocation of NFAT4 requires a basal level of [Ca 2ϩ ] i . Activated Rac and Rho Induce Nuclear Localization of NFAT4 -We further examined which small GTPases are involved in nuclear localization of NFAT4. The expression of CA-Rac and CA-Rho, but not CA-Ras, resulted in a 3-fold increase in nuclear predominant fluorescence of GFP-NFAT4 (Fig. 7, A and  B). The expression of CA-Rac and CA-Rho, but not CA-Ras, significantly increased NFAT activity (Fig. 7C). These results suggest that Rac and Rho promote nuclear localization of NFAT4, leading to NFAT activation in cardiac fibroblasts.
Requirement of Rac-mediated JNK Activation for Ang IIinduced NFAT Activation-We demonstrated that Ang II stimulation of AT1R activated G␣ 12/13 , leading to Rac activation in cardiac fibroblasts. Activated Rac induces NFAT activation through a ROS-dependent pathway(s) that still requires the basal level of [Ca 2ϩ ] i . NFAT proteins interact synergistically with AP-1 (Fos/Jun proteins) on composite DNA elements, as the NFAT binding site is close to the AP-1 binding site. They form highly a stable ternary complex to regulate the expression of diverse inducible genes (19,22). Activation of JNK promotes the phosphorylation and activation of Fos and Jun transcriptional factors. Because Ang II-induced Rac activation also induces activation of mitogen-activated protein kinases (41), we examined whether Rac-mediated JNK activation regulates NFAT activity. Stimulation of AT1R increased JNK activity by 4-fold, which was significantly inhibited by p115-RGS (Fig. 8A). Ang II-induced JNK activation is ROS-dependent as DN-Rac or DN-p47 phox inhibited JNK activation. Expression of CA-Rac activated JNK by 2.5-fold, which was significantly suppressed by DPI (Fig. 8B). Ang II-induced NFAT activation was also significantly suppressed by SP600125, a selective JNK inhibitor (Fig. 8C). These results suggest that Rac-mediated JNK activation through ROS production also contributes to Ang II-induced NFAT activation. DISCUSSION The present study demonstrated that Ang II-induced NFAT transcriptional activation was mediated by a ROS-dependent signal transduction pathway(s). Stimulation of Ang II receptor activates G 12/13 , leading to ROS production through release of G␣ 12/13 and subsequent activation of Rac in cardiac fibroblasts (Fig. 9). Because not only CA-Rac but also CA-Rho resulted in a 4-fold increase in NFAT activity (Fig. 7), Rho may play a role in Ang II-induced NFAT activation. We have recently observed using rat neonatal cardiac myocytes that the stimulation of AT1R activates Rac through a Rho/Rho kinase pathway (30). Because Ang II stimulation activates G␣ 12/13 and G␣ 12/13 mediate Rho activation through RhoGEF, Rho may activate Rac in cardiac fibroblasts by a similar mechanism to Rho/Rho kinase pathway in cardiac myocytes. Previous reports indicated that Ang II stimulation produces ROS in vascular smooth muscle cells and suggested the role of ROS as a mediator of Ang II action (14 -16). The present study further revealed the upstream signaling pathways of Ang II-induced ROS production in cardiac fibroblasts using G␣-selective RGS domains, DN mutants of Rac and p47 phox , and chemical inhibitors.
The reagents that inhibit ROS production suppressed Ang II-induced NFAT activation. Activation of G␣ 12/13 induced by AT1R stimulation resulted in Rac activation (Fig. 4D), and Ang II-induced NFAT activation was suppressed by DN-Rac. Furthermore, the expression of CA-Rac induced NFAT activation (Fig. 7), accompanying JNK activation (Fig. 8). Because the JNK activation was sensitive to ROS and a JNK inhibitor SP600125 suppressed Ang II-induced NFAT activation, Rac may mediate NFAT activation mainly by ROS production. Requirement of ROS for NFAT activation is further supported by the following evidence. Prx II or catalase (H 2 O 2 -scavenging enzyme) suppressed Ang II-induced NFAT activation, and the exposure of H 2 O 2 to cells induced NFAT activation (Fig. 1). Therefore, we speculate that NADPH oxidase-generated superoxide anion is converted to H 2 O 2 , and H 2 O 2 forms more reactive species that participate in NFAT activation machinery.
Some reports have supported the idea that G 12/13 are involved in Ang II-mediated signaling (42,43). For example, Gohla et al. (42) has reported that G␣ 12 or G␣ 13 /EGF receptor pathway is involved in stress fiber assembly in fibroblasts. However, in the present study an EGF receptor kinase inhibitor AG1478 failed to inhibit Ang II-induced Rac activation (Fig.  4D). Murasawa et al. (41) has also reported that Ang II-induced JNK activation is dependent on Rac activation but is independent of EGF receptor transactivation in cardiac fibroblasts. Thus, EGF receptor transactivation may not be necessary for Ang II-induced ROS production and subsequent signaling in cardiac fibroblasts. Macrez et al. (43) reported that a ␤␥ dimmer derived from G 13 transduces AT1R signaling. However, we observed that the G␤␥-sequestering peptide GRK2-ct did not inhibit nuclear localization of GFP-NFAT4 (32 Ϯ 3%; compared with the data in Fig. 5C) and Rac activation (2.8 Ϯ 0.4-fold; compared with the data in Fig. 4D) induced by Ang II stimulation. These results indicate that G␣ 12/13 but not G␤␥ play a critical role in ROS production and NFAT activation induced by Ang II stimulation.
G␣ q is generally thought to mediate almost all Ang II-induced responses. Because activation of G␣ q induces Ca 2ϩ mobilization through phospholipase C activation, the activated G␣ q was expected to increase NFAT transcriptional activity. However, the present study did not reveal any significant roles of G␣ q in Ang II-induced Rac activation, ROS production, and  9. Schema of AT1R-stimulated NFAT activation in cardiac fibroblasts. Stimulation of AT1R activates NFAT through G 12/13 signaling pathways. The activated G␣ 12/13 induces ROS production through Rac activation. Rho may mediate Ang II-induced Rac activation. Because p47 phox inhibition completely suppressed Ang II-induced ROS production, NADPH oxidase participated in Ang II-induced ROS production. The ROS-mediated activation of JNK may activate AP-1 proteins, which synergistically regulate NFAT activity. AT1R stimulation increases [Ca 2ϩ ] i by an unknown pathway (?), which is necessary for Ang II-induced NFAT activation.
NFAT activation, because GRK2-RGS did not affect these responses. We can exclude the possibility of insufficient expression of GRK2-RGS, because GRK2-RGS inhibited Ang II-induced Ca 2ϩ release from endoplasmic reticulum (Fig. 2B). These results suggest that Ca 2ϩ release induced by G␣ q activation is not enough for Ang II-induced NFAT activation. Because the NFAT transcriptional activity is maintained by persistent elevation of [Ca 2ϩ ] i and the continuous activation of calcineurin (18), Ang II stimulation should lead to a persistent increase in [Ca 2ϩ ] i for NFAT activation. Our previous report using myoblast cell line H9c2 indicated that expression of CA-Rac increases the basal level of [Ca 2ϩ ] i and activates JNK (38). Because G␣ 12/13 regulates Rac activity, one possible pathway is that G␣ 12/13 -activated Rac increases the basal level of [Ca 2ϩ ] i in cardiac fibroblasts. However, further study will be required for understanding the mechanism of sustained increase in the basal level of [Ca 2ϩ ] i induced by Ang II stimulation.
Previous study has demonstrated that Ras, but not Rac and Rho, regulates NFAT3 activity in cardiac myocytes (44). In contrast, expression of CA-Ras is not sufficient to activate NFAT in cardiac fibroblasts. These differences may relate to the differences in Ca 2ϩ handling between non-excitable fibroblast cells and excitable myocytes. In fact, CA-Ras could not activate NFAT in T cells, but Ras synergizes with Ca 2ϩ -activating signal to fully activate NFAT (45).
It has been reported that phosphorylation of NFAT4 by JNK causes exclusion from the nucleus (46). However, the present study demonstrated that a JNK inhibitor significantly suppressed Ang II-induced NFAT activation. This discrepancy can be explained by using the different methods for determining NFAT activation. Almost all data reported by Chow et al. (46) are shown by localization of NFAT4, but the data in the present study were determined by an NFAT-dependent luciferase assay. Thus, JNK activation increases AP-1 transcriptional activity that cooperates with NFAT family members such as NFAT3, NFATp, and NFATc and increases NFAT transcriptional activity.
ROS has been reported to be involved in Ang II-induced osteopontin gene expression (16) or endothelin-1-induced endothelin-1 gene expression in cardiac fibroblasts (47). In both cases ROS-dependent mitogen-activated protein kinase activation mediates receptor-stimulated gene expression (16,47). We demonstrated in the present study that Rac activation increased JNK activity in a ROS-dependent manner and that a JNK inhibitor suppressed Ang II-induced NFAT activation (Fig. 8). Thus, ROS-dependent JNK activation also contributes to NFAT activation induced by Ang II stimulation (Fig. 9).
In conclusion, we have demonstrated a signal transduction pathway of Ang II-induced ROS production and NFAT activation: AT1R 3 G 12/13 3 Rac1 3 ROS production 3 JNK 3 NFAT activation. The involvement of G␣ 12/13 in ROS production in cardiac fibroblasts will provide a new insight into the AT1R-mediated signaling pathway.