Rho-kinase Contributes to Sustained RhoA Activation through Phosphorylation of p190A RhoGAP*

RhoA is transiently activated by specific extracellular signals such as endothelin-1 (ET-1) in vascular smooth muscle cells. RhoGAP negatively regulates RhoA activity: thus, RhoA becomes the GDP-bound inactive form afterward. Sustained activation of RhoA is induced with high doses of the extracellular signals and is implicated in certain diseases such as vasospasms. However, it remains largely unknown how prolonged activation of RhoA is induced. Here we show that Rho-kinase, an effector of RhoA, phosphorylated p190A RhoGAP at Ser1150 and attenuated p190A RhoGAP activity in COS7 cells. Binding of Rnd to p190A RhoGAP is thought to enhance its activation. Phosphorylation of p190A RhoGAP by Rho-kinase impaired Rnd binding. Stimulation of vascular smooth muscle cells with a high dose of ET-1 provoked sustained RhoA activation and p190A RhoGAP phosphorylation, both of which were prohibited by a Rho-kinase inhibitor. The phosphomimic mutation of p190A RhoGAP weakened Rnd binding and RhoGAP activities. Taken together, these results suggest that ET-1 induces Rho-kinase activation and subsequent phosphorylation of p190A RhoGAP, leading to prolonged RhoA activation.

RhoA small GTPase is the molecular switch for various extracellular signals and is implicated in a variety of biological functions, including cell contraction, cell migration, cell adhesion, cell cycle progression, and gene expression (1,2). RhoA regulates these functions through its specific effectors such as Rho-kinase/ROCK/ROK and mDia (2). We previously found that Rho-kinase phosphorylates myosin phosphatase target protein 1 (MYPT1) 3 of myosin phosphatase and thereby inac-tivates its phosphatase activity, resulting in an increase in the phosphorylation of myosin light chain followed by smooth muscle contraction (3)(4)(5). Rho-kinase increases the agonistinduced Ca 2ϩ sensitivity and contributes to sustained contraction of smooth muscle (6).
RhoA cycles between the GTP-bound active and GDPbound inactive conformations. This cycle is under the direct control of three groups of regulatory proteins: the guanine nucleotide exchange factors (GEFs), which catalyze the exchange of GDP for GTP to activate RhoA; GTPase-activating proteins (GAPs), which enhance the intrinsic GTPase activity of RhoA to promote hydrolysis of GTP to GDP to inactivate RhoA; and the guanine nucleotide dissociation inhibitors, which sequester the GDP-bound RhoA and may also regulate its intracellular localization (1,2).
The typical RhoGEFs contain a catalytic Dbl homology domain and an adjacent pleckstrin homology domain. This Dbl homology-associated pleckstrin homology domain interacts with phospholipids, which may localize GEFs to the plasma membrane and activate GEF activity (7,8). RhoA activation is often mediated by G protein-coupled receptors. Three Rho-GEFs, which contain the regulator of G protein-signaling domains, including leukemia-associated RhoGEF, PDZ-Rho-GEF, and p115RhoGEF, directly link between G␣ 12 /G␣ 13 and RhoA (9 -11). G␣ 12 and G␣ 13 specifically interact with the regulator of G protein-signaling domains of these RhoGEFs and positively regulate their GEF activity (12). The typical RhoGAPs have a catalytic domain and various domains for protein-protein interaction. Recent studies suggest that RhoGAPs are regulated by various mechanisms, including protein-protein interaction, phospholipid interaction, phosphorylation, subcellular translocation, and proteolytic degradation (13,14). However, the precise mechanisms that regulate RhoGAP activity remain elusive in many cases.
When the smooth muscle cells are stimulated with agonists such as ET-1, RhoA is transiently activated presumably through RhoA-specific GEFs such as leukemia-associated RhoGEF and inactivated later (11,15). The RhoA-specific GAP appears to be responsible for RhoA inactivation under physiological conditions (16). Sustained RhoA/Rho-kinase activation occurs with high doses of ET-1 (17). Prolonged activation of RhoA/Rhokinase is implicated in the pathogenesis of certain vascular diseases, including subarachnoid hemorrhage-induced cerebral vasospasm, coronary vasospasm, essential hypertension, and pulmonary hypertension (18,19). For example, RhoA activity is higher in aortic smooth muscle cells derived from the strokeprone spontaneously hypertensive rat than from the wild-type rat, although the expression levels of RhoA are not different between mutant and wild-type rats (20). Rho-kinase activity is up-regulated, and phosphorylation levels of MYPT1 are increased in the coronary spastic lesion in a porcine swine model (21). Subarachnoid hemorrhage induces sustained Rhokinase activation in the canine basilar artery and subsequent cerebral vasospasm (22). Chronic hypoxia-induced pulmonary hypertension in rats is associated with an increase of RhoA activity in pulmonary artery (23). However, it remains largely unknown how prolonged activation of RhoA is induced.
In light of these observations, we hypothesized that highly activated RhoA/Rho-kinase can inhibit Rho-specific GAP and lead to sustained RhoA activation. Here we show that Rhokinase phosphorylated p190A RhoGAP, the best characterized RhoA-specific GAP, at Ser 1150 in vitro and in vivo. Phosphorylation of p190A RhoGAP by Rho-kinase appeared to attenuate its GAP activity.
GAP Assay-The RhoA GAP assay was performed as previously described (26). Briefly, recombinant GST-RhoA was preloaded with 1 M [␥-32 P]GTP (222 TBq/mmol, PerkinElmer Life Sciences) in 25 l of buffer containing 50 mM HEPES, pH 7.4, 50 mM NaCl, 0.1 mM dithiothreitol, 0.1 mM EGTA, 5 mM EDTA, and 1 mg/ml bovine serum albumin for 10 min at 30°C before the addition of MgCl 2 to a final concentration of 10 mM. An aliquot of [␥-32 P]GTPloaded GST-RhoA was mixed with the GAP assay buffer, which contained 25 mM HEPES, pH 7.5, 50 mM NaCl, 1 mM MgCl 2 , 0.1 mM dithiothreitol, 0.1 mM GTP, and 1 mg/ml bovine serum albumin in the presence of nonphosphorylated GST-p190A RhoGAP-4 ϩ 5 or phosphorylated GST-p190A RhoGAP-4 ϩ 5. The reaction was performed for 5 min at 30°C and terminated by rapid addition of 5 ml of ice-cold buffer containing 50 mM HEPES, pH 7.5, 50 mM NaCl, and 1 mM MgCl 2 . The samples were then immediately deposited onto nitrocellulose filters. The radioactivity retained on the filter was then subjected to quantitative analysis by scintillation counting. RhoGAP activity was detected as the remainder of [␥-32 P]GTP bound to GST-RhoA.
GTP-Rho Pulldown Assay-The activity of RhoA was determined by pulldown assay with the GST-Rho-binding domain of Rhotekin (GST-Rhotekin-RBD) as previously described (27). Briefly, the cells were washed with ice-cold phosphatebuffered saline and lysed in 500 l of lysis buffer (50 mM Tris/HCl, pH 7.5, 1 mM EDTA, 10 mM MgCl 2 , 500 mM NaCl, 0.5% Nonidet P-40, 0.1 mM (p-amidinophenyl)methanesulfonyl fluoride, 2.5 g/ml aprotinin, 2.5 g/ml leupeptin) containing 20 g of GST-Rhotekin-RBD. The lysates were centrifuged at 20,000 ϫ g for 3 min at 4°C, and the supernatants were incubated with glutathione-Sepharose 4B beads for 30 min at 4°C. The beads were washed with an excess of lysis buffer and then eluted with SDS-sample buffer. The eluates were subjected to SDS-PAGE, followed by immunoblot analysis with anti-GFP antibody or anti-RhoA antibody.
Cell Culture and Agonist Stimulation-Human aortic smooth muscle cells were obtained from Takara Bio Inc. (Shiga, Japan) and cultured in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin. Human aortic smooth muscle cells at the sixth passage were transfected with GFP-p190A RhoGAP-FL by using the Nucleofector system (Amaxa, Cologne, Germany). 24 h after transfection, the cells were starved for serum of 8 h and then stimulated with 1 M ET-1.
For the experiments with inhibitors, the cells were pretreated with inhibitors 30 min before ET-1 stimulation.
Measurement of Cell Size-To measure the cell size, human aortic smooth muscle cells were transfected with plasmids by using the Nucleofector system and then seeded on glass coverslips coated with fibronectin (BD Biosciences Pharmingen). 24 h after transfection, the cells were fixed with 3.0% formaldehyde in phosphate-buffered saline for 10 min and then treated with phosphate-buffered saline containing 0.1% Triton X-100 for 10 min. The cell area was measured with a laser scanning confocal microscope (model LSM510, Carl Zeiss, Oberkochene, Germany).

RESULTS
Counteraction of p190A RhoGAP Function by Rho-kinase-p190A and p190B RhoGAPs are ubiquitously expressed in various tissues and display GAP activity exclusively toward RhoA in vivo (28,29). p190RhoGAP activity accounts for ϳ60% of the total RhoGAP activity detected in whole cell extracts in fibroblasts (30). Inhibition of p190RhoGAP activity is sufficient to promote RhoA activation in fibroblasts (30). Knockdown of p190A RhoGAP activity using siRNA increases RhoA activity in spreading microvascular endothelial cells (31). Thus, p190RhoGAP appears to account for the majority of RhoGAP activity.
Hence, we first examined whether Rho-kinase affects the p190A RhoGAP function in intact cells. GFP-RhoA was transfected into COS7 cells, and the amount of the GTP-bound form of GFP-RhoA was monitored by pulldown assay (Fig. 1). The expression of GFP-p190A RhoGAP-FL decreased the amount of GTP-bound GFP-RhoA in COS7 cells, suggesting that GFP-p190A RhoGAP-FL acts as RhoAGAP. The amount of GTPbound GFP-RhoA was greater in the cells expressing GFP-p190A RhoGAP-FL and GFP-Rho-kinase-CAT than that in the cells expressing GFP-p190A RhoGAP-FL alone. This result suggests that Rho-kinase counteracts the GAP activity of p190A RhoGAP in COS7 cells.
In Vitro Phosphorylation of p190A RhoGAP by Rho-kinase-We then examined whether p190A RhoGAP is phosphorylated by Rho-kinase in vitro. To make a full length of p190A RhoGAP, we transiently transfected COS7 cells with GFP-p190A RhoGAP-FL, and immunoprecipitated GFP-p190A RhoGAP-FL from cell lysates with a polyclonal anti-GFP antibody. The immunoprecipitated GFP-p190A RhoGAP-FL was effectively phosphorylated by GST-Rho-kinase-CAT in vitro (Fig. 2B). We also found that Rho-kinase phosphorylated p190B RhoGAP (data not shown).
In Vivo Phosphorylation of p190A RhoGAP by Rho-kinase-To examine the phosphorylation state of p190A RhoGAP by Rho-kinase in vivo, we prepared rabbit polyclonal antibodies that specifically recognize p190A RhoGAP phosphorylated at respective phosphorylation sites. Among them, the anti-p190A RhoGAP-pS1150 antibody specifically recognized GST-p190A  RhoGAP-4 phosphorylated by GST-Rho-kinase-CAT in a dosedependent manner, but not phosphorylated GST-p190A RhoGAP-4-S1150A (Fig. 3A), indicating that the antibody specifically recognized GST-p190A RhoGAP-4 that was phosphorylated at Ser 1150 . The anti-p190A RhoGAP-pT1173 and pS1174 antibodies only slightly recognized phosphorylated p190A RhoGAP, suggesting that Thr 1173 and Ser 1174 are not major phosphorylation sites. Alternatively, these antibodies did not work well on phosphorylated p190A RhoGAP, although they recognized the antigen phosphopeptides (data not shown). The anti-p190A RhoGAP-pT1226 and -pS1236 antibodies could recognize the phosphorylated p190A RhoGAP in vitro in a manner similar to that of anti-p190A RhoGAP-pS1150 antibody (data not shown).

Phosphorylation of p190A RhoGAP by Rho-kinase
To examine whether Rho-kinase phosphorylates p190A RhoGAP in intact cells, GFP-p190A RhoGAP-FL was exogenously co-transfected with GFP-Rho-kinase-CAT into COS7 cells. Co-transfection of Rho-kinase-CAT resulted in an increase of phosphorylated GFP-p190A RhoGAP-FL at Ser 1150 (Fig. 3B). Treatment of the cells with Y-27632 inhibited phosphorylation of GFP-p190A RhoGAP-FL by GFP-Rho-kinase-CAT. GFP-Rho-kinase-CAT failed to phosphorylate GFP-p190A RhoGAP-FL-S1150A. Under the same conditions, phosphorylation of endogenous p190A RhoGAP at Ser 1150 was not detected, presumably because the expression level of p190A RhoGAP was low in COS7 cells. Taken together, these results indicate that Rho-kinase can phosphorylate p190A RhoGAP at Ser 1150 in COS7 cells. Similarly, the immunoblot analysis, through the use of the anti-p190A RhoGAP-pT1226 and -pS1236 antibodies, revealed that Rho-kinase can phosphorylate p190A RhoGAP at Thr 1226 and Ser 1236 in COS7 cells (data not shown).
Effects of Phosphorylation of p190A RhoGAP by Rho-kinase on Its GAP Activity-Does phosphorylation of p190A RhoGAP by Rho-kinase affect p190A RhoGAP functions? To examine the effects of phosphorylation on the GAP activity of p190A RhoGAP, we tried to produce and purify the full length of p190A RhoGAP from E. coli and insect cells, but the procedure failed. We then prepared nonphosphorylated and phosphorylated GST-p190A RhoGAP-4ϩ5, which includes the identified five phosphorylation sites and the RhoGAP catalytic domain, and performed an in vitro GAP assay. Hydrolysis of GTP-bound GST-RhoA was accelerated by using purified GST-p190A RhoGAP-4ϩ5 in a dose-dependent and time-dependent manner (supplemental Fig. S2). The GAP activity of GFP-p190A RhoGAP-4ϩ5 was not dramatically affected by phosphorylation (Fig. 4, A and B).
We here found that Rho-kinase appeared to inhibit the p190A RhoGAP activity in COS7 cells (Fig. 1). How does Rhokinase regulate the GAP activity of p190A RhoGAP in intact cells? Small GTPase Rnd is a member of the distinct subgroup of the Rho family GTPases and regulates the organization of actin cytoskeleton (32). Expression of Rnd inhibits the formation of the stress fibers in response to lysophosphatidic acid stimulation in fibroblasts (33), suggesting that Rnd antagonizes the action of RhoA. Consistently, Rnd binds to p190A RhoGAP and increases its GAP activity toward GTP-bound RhoA, resulting in RhoA inactivation (34,35). This observation prompted us to examine whether phosphorylation of p190A RhoGAP affects its interaction with Rnd1. HA-Rnd1 efficiently interacted with GST-p190A RhoGAP-4 in the pulldown assay, and phosphorylation of GST-p190A RhoGAP-4 by Rho-kinase attenuated its interaction with Rnd1 (Fig. 5). Thus, it is conceivable that Rho-kinase suppresses the GAP activity of p190A RhoGAP by inhibiting the interaction with Rnd1 through phosphorylation.
Phosphorylation of p190A RhoGAP in Cultured Vascular Smooth Muscle Cells-To understand the physiological functions of p190A RhoGAP, we confirmed the expression profile of p190A RhoGAP in various rat tissues and found that p190A RhoGAP was highly expressed in brain, lung, and aorta (supplemental Fig. S3A). We also found that p190A RhoGAP was highly expressed in primary human aortic smooth muscle and endothelial cells (supplemental Fig. S3B).
Then, we monitored phosphorylation of endogenous p190A RhoGAP in human aortic smooth muscle cells. The basal phosphorylation level of p190A RhoGAP at Ser 1150 was not detected (Fig. 6A). ET-1 is known to activate the Rho/Rho-kinase pathway (17). Stimulation of smooth muscle cells by ET-1 induced phosphorylation of endogenous p190A RhoGAP at Ser 1150 (Fig.  6A). Treatment of the cells with Y-27632 inhibited ET-1-induced phosphorylation of p190A RhoGAP. As a positive control, the MYPT1 phosphorylation level was monitored. Phosphorylation of MYPT1 at Thr 853 decreases the activity of myosin phosphatase in vascular smooth muscle cells: this phosphorylation is used as an indicator of the activity of Rho-kinase (17). ET-1 induced phosphorylation of MYPT1 at Thr 853 , whereas Y-27632 completely inhibited this phosphorylation. Taken together, these results suggest that ET-1 provoked phosphorylation of endogenous p190A RhoGAP at Ser 1150 in a Rhokinase dependent fashion in cultured vascular smooth muscle cells.
Of note, the immunoblot analysis using the anti-p190A RhoGAP-pT1226 and -pS1236 antibodies revealed that p190A RhoGAP was phosphorylated upon treatment with ET-1, but these phosphorylations were not dramatically inhibited by Y-27632, suggesting that these sites are not major phosphorylation sites by Rho-kinase in aortic smooth muscle cells (data not shown).
A high concentration of ET-1 has been shown to induce activation of Rho/Rho-kinase and subsequent MYPT1 phosphorylation, thereby resulting in the sustained contraction of vascular smooth muscle (17). We found that a high concentration of ET-1 induced sustained RhoA activation (supplemental Fig.  S4), whereas Y-27632 suppressed ET-1-induced sustained RhoA activation (Fig. 6B and supplemental Fig. S4), suggesting that Rho-kinase is involved in prolonged RhoA activation.
Because the sensitivity of the anti-p190A RhoGAP-pS1150 antibody is relatively low, we transiently transfected cultured vascular smooth muscle cells with GFP-p190A RhoGAP-FL to examine whether ET-1 induces sustained phosphorylation of p190A RhoGAP in a Rho/Rho-kinase-dependent manner (Fig.  6C). ET-1 induced sustained phosphorylation of GFP-p190A RhoGAP-FL at Ser 1150 , and Y-27632 effectively inhibited ET-1induced phosphorylation of GFP-p190A RhoGAP-FL. Under the same conditions, ET-1 induced sustained phosphorylation of MYPT1 at Thr 853 , whereas Y-27632 completely inhibited this phosphorylation (Fig. 6C). Other vasoconstrictors such as angiotensin II, acetylcholine, and thrombin slightly induced phosphorylation of p190A RhoGAP (data not shown). These results suggest that ET-1 induces sustained phosphorylation of p190A RhoGAP at Ser 1150 in a Rho/ Rho-kinase-dependent manner in cultured vascular smooth muscle cells.
There are three different types of ET receptors, ET A , ET B , and ET C . ET B receptors are classified into two subtypes, ET B1 and ET B2 . ET A and ET B2 receptors are expressed in vascular smooth muscle, and both receptors mediate vascular smooth muscle contraction (36). To examine which types of ET receptors are involved in phosphorylation of p190A RhoGAP, vascular smooth muscle cells transfected with GFP-p190A RhoGAP-FL were treated with a selective ET A or ET B antagonist. BQ-123, a selective ET A antagonist, inhibited ET-1-induced RhoA activation and phosphorylation of p190A RhoGAP, whereas BQ-788, a selective ET B antagonist, minimally affected ET-1-induced RhoA activation and phosphorylation of p190A RhoGAP (Fig. 6, D and E). These results suggest that ET-1 induces p190A RhoGAP phosphorylation via RhoA activation through ET A receptor in cultured vascular smooth muscle cells.
Because the activity of p190A RhoGAP is thought to be regulated through the tyrosine phosphorylation by Src family kinases and Abl family kinases, we examined whether ET-1 affects the tyrosine phosphorylation state of p190A RhoGAP. Sodium orthovanadate, an inhibitor of phosphotyrosine phosphatase, increased the tyrosine phosphorylation level of GFP-190A RhoGAP-FL. Under the same condition, ET-1 did not induce the tyrosine phosphorylation of p190A RhoGAP (supplemental Fig. S5). These results suggest that ET-1 did not induce the tyrosine phosphorylation of p190A RhoGAP in cultured vascular smooth muscle cells.
Effect of Phosphomimic Mutant of p190A RhoGAP on Vascular Smooth Muscle Cell Size-To explore the effect of phosphorylation of p190A RhoGAP in vitro, we replaced three putative phosphorylation sites in p190A RhoGAP-4 with Glu to produce FIGURE 7. Effects of phosphomimic mutation of p190A RhoGAP. A, binding of phosphomimic mutant to Rnd1. COS7 cells were transiently transfected with pEF-BOS-HA-Rnd1. The cell lysates were incubated with glutathione-Sepharose 4B beads coated with 200 pmol of GST, GST-p190A RhoGAP-4-WT, and GST-p190A RhoGAP-4-EEE for 1 h at 4°C. The beads were washed, and the eluates were subjected to SDS-PAGE, followed by immunoblot analysis with anti-HA Ab. GST fusion proteins were visualized by silver staining. The result is representative of three independent experiments. B, effect of GFP-p190A RhoGAP-FL-5E on RhoA in COS7 cells. The activity of RhoA was monitored by pulldown assay with the GST-Rhotekin-RBD. The eluates were analyzed by immunoblotting with anti-GFP Ab (top). The ratio of GFP-RhoA-GTP to total GFP-RhoA is shown (bottom). Data represent the means Ϯ S.E. of four independent experiments. C, effect of phosphomimic mutant of p190A RhoGAP on vascular smooth muscle cell size. The area of cells transfected with the indicated constructs was measured. Data are indicated as mean Ϯ S.D. (n Ͼ 100 in each experiment), and these results are representative of three independent experiments. The asterisk indicates a significant difference (p Ͻ 0.01) from the value of GFP-p190A RhoGAP-FL-WT.