Cytoskeletal Rearrangements and Transcriptional Activation of c-fos Serum Response Element by Rho-kinase*

The small GTPase Rho is implicated in cytoskeletal rearrangements including stress fiber and focal adhesion formation and in the transcriptional activation of c-fosserum response element. In vitro, Rho-kinase, which is activated by Rho, phosphorylates not only myosin light chain (MLC) (thereby activating myosin ATPase) but also myosin phosphatase, thus inactivating myosin phosphatase. Rho-kinase is involved in the formation of stress fibers and focal adhesions in fibroblasts. Here we show that the expression of constitutively active Rho-kinase increased the level of MLC phosphorylation. The activity of Rho-kinase was necessary for maintaining the vinculin-containing focal adhesions, whereas organized actin stress fibers were not necessary for this. The microinjection of constitutively active Rho-kinase into fibroblasts induced the formation of focal adhesions to some extent under the conditions where organized actin stress fibers were disrupted. The expression of constitutively active Rho-kinase also stimulated the transcriptional activity of c-fos serum response element. These results suggest that Rho-kinase has distinct roles in divergent pathways downstream of Rho, which include MLC phosphorylation leading to stress fiber formation, focal adhesion formation, and gene expression.

The Rho family of small GTPases exhibits both GDP/GTP binding and GTPase activities (for a review, see Ref. 1). They have GDP-bound inactive and GTP-bound active forms, which are interconvertible by GDP/GTP exchange and GTPase reactions. Members of the Rho family including RhoA, B, and C, Rac1 and 2, and Cdc42 share more than 50% sequence identity with each other (1). Rho regulates the formation of stress fibers and focal adhesions in response to extracellular signals such as lysophosphatidic acid and certain growth factors (2,3). Rho is implicated in the regulation of cell morphology (4), cell aggregation (5), cell motility (6), cytokinesis (7,8), smooth muscle contraction (9,10), and endocytosis (11). Other studies have indicated that Rho is involved in the regulation of phosphatidylinositol 3-kinase (12)(13)(14) and phosphatidylinositol 4-phosphate 5-kinase (15). Rho is also required for the transcriptional activation of c-fos SRE 1 via SRF (16), as well as for the regulation of cytoskeletal organization.
Specific targets of Rho are thought to mediate signals from Rho and to exert their biological functions. We have identified three target proteins of Rho: protein kinase N (17,18), Rhokinase (19), which is also known as ROK (20), and the MBS of myosin phosphatase (21). p160 Rho-associated coiled-coil containing protein kinase is an isoform of Rho-kinase (22). Rhokinase phosphorylates MBS and consequently inactivates myosin phosphatase (21). Rho-kinase also phosphorylates MLC and thereby activates myosin ATPase (23). Constitutively active Rho-kinase can induce the formation of stress fibers and focal adhesions when injected into intact cells (24,25). Because the phosphorylation of MLC enhances the actin-myosin interaction (for reviews, see Refs. 26 -30), the increase of the level of MLC phosphorylation regulated by Rho-kinase may account for the mechanism of smooth muscle contraction and stress fiber formation by Rho. Other targets of Rho with unknown functions include rhophilin, rhotekin, and citron (18,31).
To explore further the functions of Rho-kinase among these targets of Rho, we examined the effects of constitutively active Rho-kinase on cytoskeletal rearrangements and gene expression. We found that the expression of constitutively active Rho-kinase induced MLC phosphorylation, regulated the organization of myosin, and stimulated the transcriptional activity of c-fos SRE.

EXPERIMENTAL PROCEDURES
Materials and Chemicals-GST-CAT and GST-CAT-KD were produced in Spodoptera frugiperda cells (Sf9 cells) with a baculovirus system and purified on a glutathione-Sepharose column as described (23,32). pGEX-C3 was kindly provided by Dr. Alan Hall (University College of London). GST-RhoA I41 and GST-C3 were produced and purified from Escherichia coli as described (17). For microinjection, GST-C3 and GST-RhoA I41 were cleaved with thrombin, purified to remove the GST, and concentrated (33,34). Anti-vinculin antibody, tetramethylrhodamine isothiocyanate-labeled phalloidin, and anti-MLC antibody were purchased from Sigma. Anti-MLC-pS19 monoclonal antibody was prepared, and the method will be described elsewhere. 2 Calyculin A and cytochalasin D were purchased from Wako Pure Chemical Industries (Osaka, Japan). [␥-32 P]ATP was purchased from Amersham Corp. All materials used in the nucleic acid study were purchased from Takara Shuzo Co. (Kyoto, Japan). Other materials and chemicals were obtained from commercial sources.
Plasmid Constructs-The cDNA fragment of RhoA I41 was generated by the site-directed mutagenesis of Asn-41 to Ile-41 and subcloned into pGEX-2T to yield pGEX-RhoA I41 . pEF-BOS-HA-RhoA V14 was constructed as described (17,35). pEF-BOS-myc-CAT, pEF-BOS-myc-CAT-KD, and pEF-BOS-myc-COIL were constructed as described (25). The cDNA fragment of MLC was subcloned into the BamHI site of pEF-BOS-myc to yield pEF-BOS-myc-MLC. For gene expression analysis, two copies of SRE.L (5Ј-GGTACCTCGAGCATGTACTGTATGTCCAT-ATTAGGACATCTATGTACTGTATGTCCATATTAGGACATCTGGA-TCC-3Ј) were inserted into the KpnI and BamHI sites of pGVB-tkluciferase to yield SRE.L-luciferase. pGVB-tk-luciferase contains thymidine kinase promoter. The cDNA fragment encoding E. coli ␤-galactosidase was inserted into the NotI site of pME18S which contains SR␣ promoter to yield pME18S-lacZ. Transfection into COS7 Cells-COS7 cells were cultured in DMEM with 10% FBS. Transfection of plasmids into COS7 cells was carried out by the standard DEAE-dextran method (36). Cells were seeded on a 60-mm dish at a cell density of 5 ϫ 10 5 cells/dish in DMEM with 10% FBS and cultured overnight. The medium was replaced with fresh DMEM with 10% FBS 2 h before transfection. The DEAE-dextran/DNA mixture was prepared and added to the dish with gentle agitation. After 2 h, the cells were treated with 10% dimethyl sulfoxide and phosphatebuffered saline. The cells were grown in DMEM with 10% FBS for 1 day and then in DMEM for 1 day. In some experiments, cells were treated with calyculin A (0.1 M) for 10 min. The cells were treated with 10% (w/v) trichloroacetic acid. The resulting precipitates were subjected to immunoblot analysis.
Microinjection and Immunofluorescense Analysis-REF52 cells were cultured in DMEM with 10% FBS. Cells were seeded at a density of 1-3 ϫ 10 3 cells onto 12-mm glass coverslips and cultured for 1-2 days. Swiss 3T3 cells were cultured in DMEM with 10% FBS. Cells were seeded at a density of 8 -10 ϫ 10 3 cells onto 12-mm glass coverslips. After culturing for 4 days, the cells were serum starved for 24 h in DMEM. Recombinant proteins were microinjected along with a marker protein (rabbit IgG at 1 mg/ml) into the cytoplasm of cells. Cells were pretreated with 0.1 M cytochalasin D for 30 min before the microinjection. After injection, the cells were incubated at 37°C for 20 min. Actin, vinculin, MLC, and phosphorylated MLC were visualized by tetramethylrhodamine isothiocyanate-labeled phalloidin and antibodies to vinculin, to MLC, and to phosphorylated MLC, respectively, as described (2).
Transfection into NIH 3T3 Cells-NIH 3T3 cells were cultured in DMEM with 10% calf serum. Transfection of plasmids into NIH 3T3 cells was carried out by the standard calcium phosphate method (37). Cells were seeded on a 35-mm dish at a cell density of 6 ϫ 10 4 cells/dish in DMEM with 10% calf serum and cultured overnight. The medium was replaced by DMEM with 10% calf serum and 10 mM Hepes at pH 7.5, 2 h before transfection. A calcium phosphate/DNA mixture was prepared and added to the dish with gentle agitation. After 16 h, the cells were washed with phosphate-buffered saline for 10 min twice and grown in DMEM for 36 h.
Other Procedures-SDS-polyacrylamide gel electrophoresis was performed as described previously (38). The immunoblot analysis was carried out as described (39).

Expression of Constitutively Active and Kinase-negative
Forms of Rho-kinase-We showed previously that the catalytic domain of Rho-kinase (CAT) serves as the constitutively active form (25). To express the various forms of Rho-kinase in mam-malian cells, the cDNAs encoding myc-tagged full-length Rhokinase, CAT and catalytic domain mutated at the ATP binding site (CAT-KD) were cloned into the pEF-BOS-myc expression plasmid. Plasmids encoding the full-length Rho-kinase, CAT, and CAT-KD were transfected into COS7 cells. The various forms of Rho-kinase were immunoprecipitated with anti-myc antibody (9E10) from lysate of the cells transfected with these plasmids. The immunoprecipitates were subjected to immunoblot analysis. The various forms of Rho-kinase were all expressed in COS7 cells (data not shown).
Detection of Phosphorylated MLC in COS7 Cells-We demonstrated previously that Rho-kinase phosphorylates the MBS of myosin phosphatase, thereby inhibiting myosin phosphatase activity, and that Rho-kinase also phosphorylates MLC at Ser-19 directly in vitro (21,23). These data suggest that Rhokinase elevates the level of MLC phosphorylation, resulting in an enhancement of the actin-myosin interaction for stress fiber formation in intact cells. The microinjection of CAT into serumstarved Swiss 3T3 cells induces stress fiber formation (25). To determine whether Rho-kinase could elevate the level of MLC phosphorylation in vivo, monoclonal antibody against MLC phosphorylated at Ser-19 (anti-MLC-pS19 antibody) was used. The specificity of this antibody was examined by immunoblot analysis. Equal amounts of MLC with various ratios between unphosphorylated and phosphorylated forms were loaded on the gel. MLC phosphorylated by Rho-kinase in vitro was spe- cifically detected by the anti-MLC-pS19 antibody in a dose-dependent manner (Fig. 1).
Because the amount of endogenous MLC expressed in COS7 cells was insufficient to analyze the effect of Rho-kinase on the level of MLC phosphorylation, pEF-BOS-myc-MLC was cotransfected with plasmids carrying the RhoA V14 or Rho-kinase cDNAs into COS7 cells. The level of MLC phosphorylation was low in serum-starved COS7 cells expressing myc-MLC alone. When the cells were treated with calyculin A, which is a phosphatase inhibitor, the level of MLC phosphorylation was increased (Fig. 2), indicating that the activity of phosphatases was involved in the regulation of the level of MLC phosphorylation in the resting COS7 cells. The expression of RhoA V14 or CAT, but not CAT-KD, increased the MLC phosphorylation. Full-length Rho-kinase had no effect on the MLC phosphorylation. These results indicate that constitutively active Rhokinase elevates the level of MLC phosphorylation in intact cells.
Organization of Cytoskeletal Proteins by Rho-kinase-Because Rho-kinase can regulate the level of MLC phosphorylation in vitro (21,23) and in vivo, we then examined the distribution of phosphorylated or unphosphorylated myosin by immunostaining in REF52 cells that were microinjected with C3, GTP␥S⅐RhoA I41 (containing the substitution of Asn for Ile), or CAT. C3 ADP-ribosylates Rho at Asn-41 and inactivates it, whereas Rho I41 fails to be ADP-ribosylated by C3 (4). REF52 cells grown in 10% FBS had many actin stress fibers and vinculin-containing focal adhesions and showed a filamentous periodical pattern of myosin on stress fibers (Fig. 3). Phosphorylated myosin showed a pattern similar to that of myosin, suggesting that most of the myosin was on stress fibers and phosphorylated under these conditions. The microinjection of C3 into REF52 cells disrupted the stress fibers and focal adhesions to some degree, perturbed the filamentous pattern of myosin, and decreased the phosphorylated myosin staining. Under these conditions, most of the myosin was scattered in the cytoplasm, whereas the phosphorylated myosin was clustered over the nucleus to some extent, although the meaning of this staining is not known. The coinjection of GTP␥S⅐RhoA I41 or CAT with C3 reversed the effects of C3. Therefore, Rho-kinase is thought to mediate signals from Rho and to maintain cytoskeletal organization. In the case of cells injected with CAT, the stress fibers sometimes attached to each other at a certain point in the presence or absence of C3.
Relationship between Actin Stress Fibers and Focal Adhesions-Nobes and Hall (40) proposed that Rho induces the formation of actin stress fibers and focal adhesions in independent pathways. Chrzanowska-Wodnicka and Burridge (41) proposed that the contractility of actin stress fibers triggered by myosin phosphorylation drives focal adhesion formation. We found that Rho-kinase elevates the level of MLC phosphorylation in vitro (21,23) and in vivo and that constitutively active Rho-kinase enhances the formation of actin stress fibers and focal adhesions in Swiss 3T3 cells (25). To investigate whether the formation of stress fibers and focal adhesions diverges downstream of Rho-kinase, we examined stress fiber and focal adhesion formation by CAT in the presence of cytochalasin D, which specifically inhibits actin polymerization. The treatment of REF52 cells with cytochalasin D disrupted stress fibers but not focal adhesions (Fig. 4). The microinjection of C3 into REF52 cells resulted in the loss of both stress fibers and focal adhesions (Figs. 3 and 4). These results suggest that actin stress fibers are not necessary for maintaining focal adhesions, whereas Rho is required for maintaining them. The microinjection of CAT or CAT with C3 in the presence of cytochalasin D conferred actin aggregates with a few short actin stress fibers and apparently normal focal adhesions (Fig. 4). This indicates that Rho-kinase prevents the effects of C3 and may be involved in maintaining focal adhesions. It may be noted that the microinjection of CAT induced the stellate actin clusters in the cytochalasin D-treated cells, whereas it induced the actin clusters connected to the striated actin fibers in the nontreated cells. The active Rho-kinase may induce the aggregation of short fragments of the actin fibers and subsequently the formation of the stellate actin clusters free from the striated actin fibers that are disrupted by cytochalasin D.
We further examined whether Rho-kinase could induce the new synthesis of focal adhesions in a manner independent of actin stress fibers in Swiss 3T3 cells. Serum-starved Swiss 3T3 cells had very few stress fibers and focal adhesions, and the microinjection of CAT into the cells induced both (Fig. 5) as described previously (25). In the presence of cytochalasin D, the microinjection of CAT failed to induce stress fiber formation but induced focal adhesion formation, although the focal adhe-sions were smaller than those in the absence of cytochalasin D (Fig. 5).
Transcriptional Activation of c-fos SRE by Rho-kinase-Rho has been reported to induce the transcriptional activation of the c-fos SRE mediated by SRF (16). To determine whether Rho-kinase mediates the Rho-induced transcriptional activation, we examined the regulation of c-fos SRE by Rho-kinase. A fusion gene (SRE.L-luciferase) containing SRE.L (which is linked to the coding sequence of the firefly luciferase gene) was used as a reporter. SRE.L is a derivative of c-fos SRE and contains an intact binding site for SRF but lacks the ternary complex factor binding site (16). To standardize the transfection efficiency, pME18S-lacZ was used as an internal control. The cotransfection of SRE.L-luciferase and pME18S-lacZ with pEF-BOS-myc-CAT into NIH 3T3 cells increased the luciferase activity to an extent similar to that observed when the cells were cotransfected with pEF-BOS-HA-RhoA V14 (Fig. 6). pEF-BOS-myc-Rho-kinase had a weaker stimulatory effect. pEF-BOS-myc-CAT-KD was inactive in this capacity. Neither pEF-BOS-myc-CAT nor pEF-BOS-HA-RhoA V14 had a stimulatory effect when the cells were transfected with pGVB-tk-luciferase, which lacks the SRE.L elements. pEF-BOS-myc-CAT-KD but not pEF-BOS-myc-COIL (which encodes the coiled-coil domain of Rho-kinase lacking the Rho binding domain) inhibited the SRE.L-luciferase expression induced by RhoA V14 or CAT (Fig.  7). The inhibitory effect of CAT-KD was weaker on CAT than on RhoA V14 . These results suggest that Rho-kinase is implicated in the SRE activation regulated by Rho.

Regulation of the Level of MLC Phosphorylation-
Here, we demonstrated that constitutively active Rho-kinase (CAT) increases the level of MLC phosphorylation in COS7 cells to an extent similar to that obtained by activated Rho. Full-length Rho-kinase had no effect under the same conditions, indicating that full-length Rho-kinase is almost inactive in intact cells. The kinase-negative CAT (CAT-KD) had no effect. CAT-KD is thought to serve as the dominant negative form for the Rho- induced stress fiber formation in Swiss 3T3 cells (25). CAT-KD did not inhibit the Rho-induced MLC phosphorylation in COS7 cells (data not shown). CAT-KD is thought to titrate out substrates for Rho-kinase, thereby inhibiting Rho-kinase. CAT-KD may not efficiently tie up MLC because MLC was overexpressed in the COS7 cells. Taken together, these results suggest that Rho-kinase induces MLC phosphorylation in vivo downstream of Rho. We showed previously that Rho-kinase elevates the level of MLC phosphorylation in vitro via the direct phosphorylation of MLC and the inhibition of myosin phosphatase (21,23). It remains to be clarified which pathway mainly contributes to the elevation of the levels of MLC phos-phorylation in intact cells.
Organization of Cytoskeletal Proteins-REF52 cells grown in the presence of serum had many stress fibers and focal adhesions. The MLC staining in these cells is the discrete fibrous pattern as described (41), and phosphorylated MLC showed a similar pattern. Thus, some MLC is thought to be phosphorylated and to exist on stress fibers. The microinjection of C3 disrupted both stress fibers and focal adhesions to some degree. C3 also perturbed the fibrous pattern of MLC and reduced the phosphorylated MLC content. The coinjection of GTP␥S⅐ RhoA I41 or CAT with C3 reversed the inhibitory effect of C3, suggesting that Rho increases the level of MLC phosphorylation to incorporate myosin into stress fibers, most likely through Rho-kinase. This observation is consistent with the notion that the state of MLC phosphorylation regulates the organization of stress fibers (42). The microinjection of CAT appears to cause cross-linking between stress fibers. The reason why such a pattern is observed when the cells were injected with CAT is unclear. Other signals from Rho or the proper localization of Rho-kinase may be necessary to organize stress fibers normally as discussed (25).
Focal Adhesion Formation by Rho-kinase-The microinjection of Rho-kinase in Swiss 3T3 cells induces the formation of stress fibers and focal adhesions (25). The phosphorylation of MLC induced by Rho-kinase stimulates the actin-myosin interaction, which may be one of the final steps of stress fiber formation. We further investigated whether MLC phosphorylation is sufficient for focal adhesion formation. For this purpose, we examined the effects of Rho-kinase in the presence of cytochalasin D, which inhibits actin polymerization and thereby inhibits the contractility of actin stress fibers induced by actin-myosin interaction. In the REF52 cells, actin stress fibers became thinner or disappeared by the treatment with cytochalasin D, whereas there was no remarkable change in vinculin-containing focal adhesions under the same conditions. As shown above, C3 disrupted the focal adhesions, whereas the C3 effect was blocked by coinjection with CAT regardless of the presence of cytochalasin D. These results suggest that the maintenance of mature focal adhesion does not require highly organized actin stress fibers but does require the activity of Rho or Rho-kinase.
To examine whether organized actin stress fibers are necessary for triggering focal adhesion formation, we injected CAT into serum-starved Swiss 3T3 cells in the presence of cytochalasin D. Consequently, the focal adhesions were newly synthesized, but they were smaller than those induced by CAT in the absence of cytochalasin D. Thus, it seems likely that CAT can stimulate focal adhesion formation, although it may be insufficient to complete it. Focal adhesion formation appears to be partly linked to actin stress fiber formation. Organized actin stress fibers and their contractility induced by actin-myosin interaction may be necessary to form mature focal adhesions, as described (41). Nevertheless, it is likely that Rho-kinase phosphorylates certain proteins that trigger focal adhesion formation.
Gene Expression and Rho-kinase-It was shown that Rho regulates the transcriptional activation by SRF (16). PRK2 shows strong sequence homology with protein kinase N (43) and is identified as a Rho target (44). PRK2 weakly potentiates the transcriptional activation of SRE by activated Rho (44). Consistently, the constitutively active form of protein kinase N weakly activates SRE (up to 2-fold) (data not shown). We showed in this study that CAT strongly activates SRE and that CAT-KD inhibits the activation of SRE by CAT or RhoA V14 . The inhibitory effect of CAT-KD is weaker on the CAT-induced SRE activation than on that induced by RhoA V14 . CAT-KD may not efficiently inhibit the activity of CAT which is exogenously overexpressed. Taken together, these results indicate that Rhokinase can serve as a mediator of the signaling pathway for the transcriptional activation of SRE by RhoA V14 . It remains to be determined whether the formation of stress fibers and focal adhesions controls the SRE-linked transcriptional activation.
Pleiotropic Functions of Rho-kinase-We showed here that Rho-kinase regulates both cytoskeletal organization, including stress fiber and focal adhesion formation and enhancement of MLC phosphorylation, and gene expression downstream of Rho. Thus, Rho-kinase may mediate plural pathways from Rho and function in cooperation with other Rho targets. The identification of physiological substrates for Rho-kinase is necessary to understand the molecular mechanism underlying the phenomena induced by Rho-kinase. To date, the MBS of myosin phosphatase and MLC have been identified as physiological substrates for Rho-kinase. Rho-kinase phosphorylates several proteins such as GFAP stoichiometrically (45). In the case of GFAP, the phosphorylation of the head domain by Rho-kinase induces the disassembly of GFAP (45). In addition to MBS, MLC, and GFAP, we speculate that Rho-kinase phosphorylates at least two classes of proteins; 1) protein(s) that can stimulate focal adhesion formation, and 2) protein(s) implicated in gene expression via SRF. Further studies are necessary to identify these substrates for Rho-kinase.