M-RIP Targets Myosin Phosphatase to Stress Fibers to Regulate Myosin Light Chain Phosphorylation in Vascular Smooth Muscle Cells*

Vascular smooth muscle cell contraction and relaxation are directly related to the phosphorylation state of the regulatory myosin light chain. Myosin light chains are dephosphorylated by myosin phosphatase, leading to vascular smooth muscle relaxation. Myosin phosphatase is localized not only at actin-myosin stress fibers where it dephosphorylates myosin light chains, but also in the cytoplasm and at the cell membrane. The mechanisms by which myosin phosphatase is targeted to these loci are incompletely understood. We recently identified myosin phosphatase-Rho interacting protein as a member of the myosin phosphatase complex that directly binds both the myosin binding subunit of myosin phosphatase and RhoA and is localized to actin-myosin stress fibers. We hypothesized that myosin phosphatase-Rho interacting protein targets myosin phosphatase to the contractile apparatus to dephosphorylate myosin light chains. We used RNA interference to silence the expression of myosin phosphatase-Rho interacting protein in human vascular smooth muscle cells. Myosin phosphatase-Rho interacting protein silencing reduced the localization of the myosin binding subunit to stress fibers. This reduction in stress fiber myosin phosphatase-Rho interacting protein and myosin binding subunit increased basal and lysophosphatidic acid-stimulated myosin light chain phosphorylation. Neither cellular myosin phosphatase, myosin light chain kinase, nor RhoA activities were changed by myosin phosphatase-Rho interacting protein silencing. Furthermore, myosin phosphatase-Rho interacting protein silencing resulted in marked phenotypic changes in vascular smooth muscle cells, including increased numbers of stress fibers, increased cell area, and reduced stress fiber inhibition in response to a Rho-kinase inhibitor. These data support the importance of myosin phosphatase-Rho interacting protein-dependent targeting of myosin phosphatase to stress fibers for regulating myosin light chain phosphorylation state and morphology in human vascular smooth muscle cells.

Vascular smooth muscle cell contraction and relaxation are directly related to the phosphorylation state of the regulatory myosin light chain. Myosin light chains are dephosphorylated by myosin phosphatase, leading to vascular smooth muscle relaxation. Myosin phosphatase is localized not only at actin-myosin stress fibers where it dephosphorylates myosin light chains, but also in the cytoplasm and at the cell membrane. The mechanisms by which myosin phosphatase is targeted to these loci are incompletely understood. We recently identified myosin phosphatase-Rho interacting protein as a member of the myosin phosphatase complex that directly binds both the myosin binding subunit of myosin phosphatase and RhoA and is localized to actin-myosin stress fibers. We hypothesized that myosin phosphatase-Rho interacting protein targets myosin phosphatase to the contractile apparatus to dephosphorylate myosin light chains. We used RNA interference to silence the expression of myosin phosphatase-Rho interacting protein in human vascular smooth muscle cells. Myosin phosphatase-Rho interacting protein silencing reduced the localization of the myosin binding subunit to stress fibers. This reduction in stress fiber myosin phosphatase-Rho interacting protein and myosin binding subunit increased basal and lysophosphatidic acid-stimulated myosin light chain phosphorylation. Neither cellular myosin phosphatase, myosin light chain kinase, nor RhoA activities were changed by myosin phosphatase-Rho interacting protein silencing. Furthermore, myosin phosphatase-Rho interacting protein silencing resulted in marked phenotypic changes in vascular smooth muscle cells, including increased numbers of stress fibers, increased cell area, and reduced stress fiber inhibition in response to a Rho-kinase inhibitor. These data support the importance of myosin phosphatase-Rho interacting protein-dependent targeting of myosin phosphatase to stress fibers for regulating myosin light chain phosphorylation state and morphology in human vascular smooth muscle cells.
Blood vessel tone is important in the regulation of blood pressure and tissue perfusion. Disorders of blood vessel tone play a prominent role in the pathogenesis of hypertension, vascular spasm, and acute coronary syndromes (1)(2)(3)(4)(5). Blood vessel tone is itself controlled by the contraction and relaxation of vascular smooth muscle cells (VSMCs) 2 in the media of the blood vessel wall. VSMC contraction is determined by the phosphorylation state of the regulatory myosin light chain (MLC) (6). Phosphorylation of MLC at serine 19 leads to actin-activated myosin ATPase activity, cross-bridge cycling, and contraction (7,8). MLC is phosphorylated by the calcium/ calmodulin-regulated myosin light chain kinase (MLCK), which phosphorylates MLC leading to VSMC contraction (9,10). Myosin light chain phosphatase (MLCP) dephosphorylates MLC causing VSMC relaxation (11). The counter regulatory activities of MLCK and MLCP control MLC phosphorylation state in response to contractile agonists and vasodilators.
MLCP is a heterotrimer consisting of the 37-kDa PP1 catalytic subunit, a 130-kDa myosin binding subunit (MBS), and a 20-kDa subunit of uncertain function (11)(12)(13)(14). The MBS has been shown to confer the specificity of PP1 for MLC (11) and contains protein interaction domains, including amino-terminal ankyrin repeats, and a carboxylterminal leucine zipper domain.
Recent data support an important role for MLCP in the regulation of blood vessel function in vivo. The use of specific Rho-kinase inhibitors has implicated inhibition of MLCP in the pathogenesis of hypertension (23), vascular spasm (24,25), response to injury (26,27), and atherosclerosis (25,28). Use of Rho-kinase inhibitors in small studies in humans has shown a potential role in the treatment of angina (29) and microvascular spasm (30).
MLCP has been localized in VSMCs and fibroblasts to actin-myosin stress fibers (20,31) and is also present in the cytosol and at the cell membrane (20,32). Recent data suggest that phosphorylation of MBS may play a role in the regulation of MLCP localization within the cell (33,34). Studies have shown that MBS can bind to actin-myosin stress fibers, supporting a potential role for MBS in the cellular localization of the heterotrimeric phosphatase (20,31,35). The regulation and intracellular targeting of MLCP is, however, complex and incompletely understood.
We recently identified a new member of the MLCP complex, myosin phosphatase-rho interacting protein (M-RIP) (36), the human homolog of murine p116RIP3 (37). We found that M-RIP can directly bind both RhoA and MLCP at adjacent sites and can form a complex of all three proteins (36). In VSMCs, M-RIP colocalized with actin filaments. We hypothesized that M-RIP, by virtue of these protein-protein interactions, targets MLCP to actin-myosin contractile filaments. Here we report that M-RIP silencing using RNA interference (RNAi) prevents MBS localization to stress fibers. This leads to increased MLC phosphorylation without a change in cellular MLCP, MLCK, or RhoA activity. M-RIP silencing also caused marked VSMC phenotypic changes, including increases in stress fiber number and cell area, and reduced stress fiber inhibition from a Rho-kinase inhibitor.

EXPERIMENTAL PROCEDURES
Materials-1-Oleoyl-2-hydroxy-sn-3-glycerol-3-phosphate sodium (LPA) was from Sigma, Y27632 was from Tocris, and bovine brain calmodulin was from Biomol. Rhotekin-RBD GST beads were purchased from Cytoskeleton. Antibodies were as follows: anti-M-RIP and anti-p-MLC (Thr 18 /Ser 19 ) were from BD Transduction Laboratories, anti-MBS was from Covance, monoclonal anti-PP1 and polyclonal anti-RhoA were from Santa Cruz Biotechnology, and monoclonal anti-MLC and anti-MLCK were from Sigma.
Cell Culture-Human aortic smooth muscle cells (Ao184) were cultured by the explant method as described before (36). Ao184 cells were immortalized using adenovirus-expressing E6 and E7 and selection for G418 resistance and were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum.
Myosin Light Chain Phosphorylation Assay-Human aortic smooth muscle cells (Ao184) were grown to near confluence in 12-well dishes, then were serum-deprived overnight. After changing the media, cells were treated with 1 M LPA for 10 min with or without pretreatment with 10 M Y27632 for 30 min. Following stimulation, cold 10% trichloroacetic acid, 2 mM EDTA, and 10 mM dithiothreitol (DTT) were added to the cells on ice. The cell pellet was then microcentrifuged and washed three times with cold acetone containing 2 mM DTT. The washed pellet was solubilized in SDS sample buffer, separated by SDS-PAGE on 7.5% gels, and immunoblotted for M-RIP, P-MLC, MLC, and MBS.
RNA Interference-M-RIP-specific oligonucleotides were predicted using oligoengine and Dharmacon programs. A 21-bp oligonucleotide corresponding to bp 3259 -3278 of the M-RIP cDNA was synthesized by Dharmacon. Nonspecific control X (Dharmacon) was used as the control scrambled dsRNA and contained similar GC content to the M-RIP dsRNA. Cells were transfected in 6-or 12-well dishes at 30 -60% confluence with 100 nM dsRNA using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. The cells were studied 96 h after transfection. For determination of silencing efficiency in Fig.  1, cell lysates were prepared as described in Ref. 36, and protein assays were performed using the MicroBCA kit (Pierce). Twenty micrograms of each lysate was loaded on 7.5 and 10% protein gels, subjected to SDS-PAGE, and transferred to nitrocellulose. Immunoblots were performed with the indicated antibodies.
Preparation of Regulatory Myosin Light Chains-The 32 P-labeled myosin light chain (MLC) was used as the substrate for MLCP assays, and unlabeled MLC was used as the substrate in MLCK assays. Chicken gizzard MLC cDNA was a gift from Dr. Kathy Trybus (University of Vermont). MLC cDNA was cloned into vector pQE (Qiagen) and verified by sequencing. HIS 6 MLC was expressed and purified from bacteria using nickel-nitrilotriacetic acid-agarose beads (Qiagen) as described (36). For MLCK assays, HIS 6 MLC was eluted from beads with 250 mM imidazole. For MLCP assays, HIS 6 MLC was phosphorylated with GST-ZIPkinase. GST-ZIPkinase was expressed and purified from bacteria as described before (38 Myosin Phosphatase Activity Assay-Subconfluent Ao184 cells grown in 100-mm dishes were serum-deprived overnight. Cells were treated with vehicle or LPA for 10 min, with and without pretreatment with Y27632 for 30 min. The cells were then rinsed with cold PBS on ice and lysed in 20 mM Tris, pH 7.3, 137 mM NaCl, 1% Triton X-100, 10% glycerol, 25 mM ␤-glycerol phosphate, 2 mM phenylmethylsulfonyl fluoride, 0.01 g/ml each of aprotinin, leupeptin, and pepstatin A. The lysates were homogenized by passage through a 26-gauge needle five times and were microcentrifuged for 20 min. Each lysate was then split in two for immunoprecipitation with nonimmune IgG, and anti-MBS. The samples were incubated at 4°C, rocking for 3 h, then protein A beads were added for an additional 2 h. The beads were washed twice in phosphatase assay buffer (50 mM Tris, pH 7.5, 4 mM EDTA, 2 mM EGTA, 2 mM DTT), then incubated for 20 min at 30°C with 10 M [ 32 P]MLC. The assay was terminated, and the protein precipitated by the addition of bovine serum albumin and trichloroacetic acid to 0.1 and 10% final concentration, respectively, incubation on ice for 10 min, and centrifugation for 15 min. The liberated free 32 P in the supernatant was measured by Cerenkov counting. Similar amounts of MBS were recovered in the immunopellets of scrambled and M-RIP RNAi Ao184 cells (data not shown). Phosphatase activity was normalized to the amount of MBS immunoprecipitated.
MLCK Activity Assay-The assay was based on the method published by Poperechnaya et al. (39). Subconfluent Ao184 cells with and without transfection with dsRNA were grown in 100-mm dishes and deprived of serum overnight. In control experiments, cells were treated with or without 10% serum. The cells were rinsed with cold PBS, then lysed in 0.35 ml of 10 mM MOPS, pH 7.0, 1% Nonidet P-40, 5 mM EDTA, 0.5 mM EGTA, 1 mM DTT, 50 mM MgCl 2 , 300 mM NaCl, 2 mM phenylmethylsulfonyl fluoride, 0.01 g/ml each of aprotinin, leupeptin, and pepstatin A. The cell lysates were collected and microfuged for 15 min at 4°C. The supernatant was diluted 1:4 with 20 mM MOPS, pH 7.0, 10 mM magnesium acetate, and protease inhibitors as above and added to anti-MLCK precoupled to protein G beads. Following incubation for 2 h at 4°C, the immunopellets were washed three times with cold 20 mM MOPS, pH 7.0, 10 mM magnesium acetate, 100 mM NaCl, and 0.5 mM DTT. The immunopellets were resuspended in 40 l of reaction buffer consisting of 10 mM MOPS, pH 7.0, 1 mM DTT, 4 mM MgCl 2 , 0.1 mM CaCl 2 , 1 M calmodulin, and 0.1 mM ATP. The reaction was incubated at 25°C for 20 min (39) then stopped by the addition of protein sample buffer and boiling. Aliquots of each reaction were subjected to SDS-PAGE and immunoblotting with anti-p-MLC, as described above, and anti-MLCK. MLC phosphorylation was normalized to the amount of immunoprecipitated MLCK. Control samples were assigned a value of 1.
RhoA Activity Assay-The assay was based on the method of Ren et al. (40) and the manufacturer's instructions for Rhotekin-RBD GST beads. Subconfluent Ao184 cells with and without transfection with dsRNA were grown in 150-mm dishes and deprived of serum overnight. In control experiments, cells were treated with or without 10% serum or 1 M LPA for 10 min. The cells were rinsed with cold PBS, then lysed in 50 mM Tris, pH 7.5, 1% Igepal, 300 mM NaCl, 10 mM MgCl 2 , 2 mM phenylmethylsulfonyl fluoride, 0.01 g/ml each of aprotinin, leupeptin, and pepstatin A. The lysate was microcentrifuged for 5 min at 4°C, and the cleared supernatant was added to 120 g of Rhotekin-RBD GST beads. After incubation for 45 min at 4°C, the beads were washed three times in 25 mM Tris, pH 7.5, 40 mM NaCl, and 30 mM MgCl 2 and boiled in protein sample buffer. The RhoA bound to Rhotekin-RBD beads and the input RhoA were detected by SDS-PAGE and immunoblotting. Rhotekin-bound RhoA was normalized to the input RhoA.
Purification of Stress Fibers-Purified stress fibers were prepared from Ao184 cells by glycerol extraction as described (35,36). Briefly, cells were grown on 100-mm dishes to near confluence, washed with cold PBS, then extracted with 10 ml of triethanolamine extraction buffer (2.5 mM triethanolamine, 1 g/ml leupeptin and pepstatin A, 20 g/ml aprotinin) for 30 min, shaking, with replacement of extraction buffer every 2-3 min. Remaining cell components were then further extracted using 10 ml of glycerol buffer (50% glycerol, 1 g/ml each of leupeptin and pepstatin A, 20 g/ml aprotinin in PBS) for 5 min while shaking, with replacement of extraction buffer twice. The glycerol was then removed by washing with 10 ml of aprotinin-PBS (20 g/ml aprotinin, 1 g/ml of leupeptin and pepstatin A in PBS) for 8 min while shaking with one replacement of wash buffer. The remaining insoluble material was scraped in aprotinin-PBS and homogenized with a Z-shaped 21 gauge needle. The insoluble debris was pelleted at 1,000 ϫ g for 5 min, and stress fibers were isolated by centrifugation of the supernatant at 100,000 ϫ g for 1 h. The stress fiber pellet was boiled in protein sample buffer and subjected to SDS-PAGE and immunoblotting with the indicated antibodies. For immunostaining of purified stress fibers, cells were grown on coverslips and glycerol-extracted as described above. Stress fibers were then fixed and immunostained as described below.

Assessment of Stress Fiber Number and Cell Area and Stress Fiber
Response to Y27632-Ao184 cells were transfected with control or M-RIP oligonucleotides, incubated for 48 h, and then plated at 20% confluence on coverslips in 12-well dishes in Dulbecco's modified Eagle's medium with 10% fetal bovine serum. Seventy-two hours after transfection, the cells were fixed and permeabilized as described (36)  Data Analysis-All data were plotted and analyzed using SigmaPlot 5.0 software, and p value was determined using Student's t test.

Specific Inhibition of M-RIP Expression Using
RNAi-To test the function of M-RIP in VSMCs, M-RIP expression was inhibited using RNAi. Compared with control transfected VSMCs, M-RIP RNAi reduced M-RIP expression by a mean of 72% (data not shown) without affecting the expression level of MLC, the MLCP subunits MBS and PP1, RhoA, ROCK, or MLCK (Fig. 1).

M-RIP Targets MLCP to Stress Fibers in Vascular Smooth
Muscle Cells-M-RIP has been shown to bind MLCP, actin, and myosin (36,(41)(42)(43). To test the possibility that M-RIP targets MLCP to stress fibers in VSMCs where MLCP is known to regulate MLC phosphorylation, MBS binding to stress fibers was studied. Because of abundant cytoplasmic and membrane MBS, stress fiber binding of MBS is difficult to detect in whole cells (31). Therefore, stress fibers were purified from cultured Ao184 smooth muscle cells. Ao184 cells were serially extracted, and the retained stress fibers were immunostained and immunoblotted for actin, M-RIP, and MBS. Silencing of M-RIP resulted in less M-RIP ( Fig. 2A) and MBS (Fig. 2B) fluorescence on the purified stress fibers. When purified stress fiber preparations were immunoblotted, more actin was found in the M-RIP-silenced stress fiber preparations than in the control stress fiber preparations, although total cellular actin was similar in both conditions. A similar amount of MBS was found in the M-RIP-silenced and control cells (Fig. 2C). When normalized to the actin content, there was a 50% reduction in MBS on M-RIP-silenced stress fibers (p Ͻ 0.05, n ϭ 3, Fig. 2D), whereas the amount of MLCK normalized to actin content was not significantly changed (Fig. 2E).
M-RIP Knockdown Increases Myosin Light Chain Phosphorylation-We hypothesized that reduced MLCP targeting to stress fibers would increase MLC phosphorylation. To test this hypothesis, we next assayed MLC phosphorylation state in VSMCs using phospho-specific antibodies. In control experiments, LPA treatment resulted in a 2.1-fold increase in phospho-MLC (p ϭ 0.007, in triplicate, n ϭ 3) that was inhibited by the Rho-kinase inhibitor Y27632, consistent with LPA receptor activation leading to sequential activation of RhoA and Rho-kinase and inhibition of MLCP-dependent dephosphorylation of MLC (Fig. 3A) (44). Silencing of M-RIP lead to a 74% increase in basal and an 83% increase in LPA-stimulated MLC phosphorylation compared with control transfection (p Ͻ 0.05 for basal, p Ͻ 0.05 for LPA, in triplicate, n ϭ 3 independent experiments, Fig. 3, B and C).
M-RIP Silencing Does Not Change Cellular MLCP Activity-To determine whether the increase in phospho-MLC shown in Fig. 3 was associated with inhibition of MLCP activity, MBS was immunoprecipitated from vascular smooth muscle cells, and the activity of the MLCP complex was measured using 32 P-labeled myosin light chain as a substrate. Immunoprecipitated MBS recovered PP1 protein (Fig. 4A). Immunoprecipitated phosphatase activity against MLC was inhibited by 2 M but not 2 nM okadaic acid, consistent with PP1 phosphatase activity (45,46) (Fig. 4B). To test whether immunoprecipitated phosphatase activity reflects cellular MLCP activity, vascular smooth muscle cells were treated with LPA to inhibit MLCP via RhoA/Rho-kinase mediated MLCP phosphorylation, with and without the Rho-kinase inhibitor Y27632. MLCP activity from cells treated with LPA was inhibited by 39% versus control (p Ͻ 0.05, n ϭ 3, Fig. 4C), and the inhibition was prevented by preincubation with Y27632 (p Ͻ 0.05, n ϭ 3, Fig. 4C), reflecting RhoA/Rho-kinase mediated regulation of MLCP in the cell. Next, cellular MLCP activity was assayed with and without M-RIP silencing. M-RIP silencing to similar levels that increased MLC phosphorylation (data not shown) did not change cellular MLCP activity (p ϭ NS, n ϭ 3, Fig. 4D).
M-RIP Silencing Does Not Change Cellular MLCK Activity-Increased MLC phosphorylation may result from loss of MLCP targeting to stress fibers, or alternatively, from increased MLCK activity. To assess cellular MLCK activity, MLCK was immunoprecipitated from Ao184 cells and incubated with recombinant MLC in the presence of ATP, calmodulin, and calcium. Stimulation of the cells with serum caused a 3-fold increase in MLCK activity (p ϭ 0.04, n ϭ 3) indicating that this assay reflected changes in cellular MLCK activity (Fig. 5, A and  B). Silencing M-RIP expression in these cells, however, did not change MLCK activity (p ϭ NS, n ϭ 3), supporting that the increase in MLC phosphorylation is due to loss of MLCP targeting rather than activation of MLCK (Fig. 5, A and B).
RhoA Activation State Is Unchanged by M-RIP Silencing-A recent report raised the possibility that M-RIP may have Rho-GAP activity (43). In the setting of M-RIP silencing, loss of RhoA-GAP activity could result in RhoA and Rho-kinase activation and increased myosin phosphorylation. To determine if loss of Rho-GAP activity caused increased MLC phosphorylation in the M-RIP-silenced Ao184 cells, RhoA activation assays were performed. RhoA activity was assessed by measuring binding of cellular RhoA to the Rhotekin-Rho binding domain, which preferentially interacts with Rho-GTP. In untransfected Ao184 cells, serum and LPA both increased RhoA binding to Rhotekin, consistent  with RhoA activation (4.75-fold increase, serum compared with control, p Ͻ 0.001, n ϭ 3) (Fig. 6, A and B). However, in Ao184 cells in which M-RIP expression was silenced, RhoA activation state was not significantly altered compared with scrambled control RNAi (Fig. 6, A and B), supporting that the M-RIP silencing did not activate cellular RhoA.

M-RIP Expression Regulates Vascular Smooth Muscle Cell
Morphology-To explore further the effect of M-RIP on VSMC phenotype, intact VSMCs subjected to control or M-RIP RNAi were examined by immunofluorescence microscopy. Control cells had the appearance of migrating smooth muscle cells, with lamellopodia at the leading edge and prominent stress fibers oriented along the long axis of the cell culminating at the retracting tail (Fig. 7A). When M-RIP was silenced by RNAi, the cells had fewer lamellopodia, more abundant stress fibers, and appeared larger (Fig. 7, A and B). Cells in which M-RIP levels were reduced had three times more stress fibers than control transfected cells (Fig. 7B, p Ͻ 0.001, n ϭ 3) and three times greater cell area compared with control transfected cells (Fig. 7C, p Ͻ 0.001, n ϭ 3). The increase in stress fibers noted here in intact cells after M-RIP silencing correlates with the increase in actin content in purified stress fiber preparations after M-RIP silencing in Fig. 2C.
M-RIP Expression Level Correlates with Y27632-mediated Stress Fiber Inhibition-Rho-kinase induces stress fiber formation when activated by RhoA (47). The Rho-kinase inhibitor Y27632 inhibits stress fibers at least in part by relieving RhoA/Rho-kinase-mediated MLCP inhibition (44). Activated MLCP dephosphorylates MLC leading to stress fiber destabilization (48,49). We therefore postulated that M-RIP silencing, by preventing MLCP targeting to stress fibers, would prevent Y27632-mediated stress fiber inhibition. Ao184 cells with control RNAi treated with Y27632 had very few remaining stress fibers (Fig. 8A). However, M-RIP RNAi Ao184 cells treated with Y27632 retained abundant stress fibers (Fig. 8A). When both stress fiber number and M-RIP expression were analyzed in M-RIP RNAi cells treated with Y27632, cells with higher levels of M-RIP expression had stress fiber inhibition, whereas cells with trace amounts of M-RIP had little effect from Y27632 (Fig. 8B).

DISCUSSION
We recently found that M-RIP is a member of the myosin phosphatase complex that directly binds both the MBS and RhoA (36). Furthermore, we have shown that M-RIP colocalizes with actin-myosin stress fibers (36), and others have found that M-RIP can bind both actin (36,41) and myosin (41,43), suggesting a role for M-RIP in targeting the MLCP complex. We tested this hypothesis using purified stress fibers from VSMCs and found that M-RIP silencing reduced the localization of MBS to these stress fibers. We further hypothesized that uncoupling of MLCP from actin-myosin stress fibers would prevent MLCP-mediated dephosphorylation of MLC, thus increasing MLC phosphorylation. Indeed, M-RIP silencing increased basal and LPA-stimulated MLC phosphorylation in VSMCs. M-RIP silencing caused changes in VSMC morphology, including increased numbers of stress fibers and increased cell area. M-RIP silencing also prevented Y27632-mediated stress fiber inhibition.
It has recently been suggested that p116RIP might target MLCP within neuronal cells (42). Silencing of p116RIP shifted MBS from a particulate to a soluble fraction in N1E-115 cells, consistent with a tar-  geting function (42). Our data confirm and extend those findings to show that M-RIP targets MBS in VSMCs where MLCP regulates contractile state, and we have localized the targeting of MBS by M-RIP to actin-myosin stress fibers.
MLCP was originally purified as a myosin-associated phosphatase activity. Several studies have found that the MBS can bind to myosin. In these protein binding studies, myosin was purified from chicken gizzard smooth muscle and MBS was either purified from chicken gizzard or expressed in Sf9 cells (13,33,50,51). These previous studies raise the question as to the role of M-RIP in targeting MLCP to actin-myosin stress fibers if the MBS itself has myosin binding activity. One possibility is that both MBS and M-RIP directly bind actin-myosin stress fibers, and both are required for optimal targeting and binding of the phospha-tase to dephosphorylate its substrate phospho-MLC. This possibility would explain why there is residual MBS on stress fibers when M-RIP is silenced (although this could also be due to targeting by remaining   M-RIP). A second possibility is that myosin fractions used in previous studies showing MBS binding may have also contained M-RIP that was responsible for the MBS binding. In either case, the current experiments establish an important role for M-RIP in targeting MLCP to actin-myosin stress fibers.
There are conflicting published data regarding the ability of M-RIP to regulate MLC phosphorylation in non-muscle cells (42,43). Mulder et al. (42) recently found that, although p116RIP localized MBS to an insoluble fraction in N1E-115 cells, silencing of p116RIP did not change MLC phosphorylation in NIH3T3 cells. Our data, however, support that M-RIP does in fact regulate MLC phosphorylation in human VSMCs. This discrepancy raises the possibility that M-RIP may have cell-type-specific functions.
Increased MLC phosphorylation may result from reduced MLCP targeting or activity, or increased MLCK targeting or activity. Our data support that M-RIP silencing reduced MLCP targeting to stress fibers, but other potential mechanisms were also investigated. We examined the effect of M-RIP silencing on MLCP activity. The immunoprecipitation phosphatase assay used reflected intracellular regulation of MLCP in control experiments. However, we found that M-RIP silencing had no significant effect on cellular MLCP activity. We also investigated whether M-RIP silencing affected MLCK targeting or activity. Stress fiber preparations from control and M-RIP-silenced Ao184 cells did not show a difference in MLCK content. Furthermore, intracellular MLCK activity was not affected by M-RIP silencing. These data support that uncoupling of MLCP from stress fibers, rather than a change in MLCP activity or MLCK targeting or activity, is the primary mechanism of increased phospho-MLC after M-RIP silencing.
Koga and Ikebe (43) found that p116RIP overexpression in HeLa cells decreased RhoA-GTP, suggesting that p116RIP is a GTPase-activating protein (GAP) for RhoA in these cells. We tested RhoA activation in VSMCs and found that, whereas serum and LPA increased RhoA-GTP, M-RIP silencing did not. This indicates that RhoA activation is not a cause of MLC phosphorylation with M-RIP silencing in our experiments. However, there may be M-RIP-associated RhoA GAP activity under different conditions or in different cells.
Reducing the expression of M-RIP in VSMCs had a dramatic effect on cell morphology. Cells in which M-RIP was silenced had three times more stress fibers and three times greater cell area. Increased numbers of stress fibers can be attributed to the loss of MLCP targeting to stress fibers causing increased phospho-MLC, because phospho-MLC is known to promote and to stabilize stress fibers (52). In support of this, stress fibers have been shown to increase when MLCP is inhibited by hypoxia in pulmonary artery VSMCs (49), and when MBS is silenced in HeLa cells (48). It is also possible that the phenotypic changes noted after M-RIP silencing are due to changes in gene transcription.
Whereas control RNAi VSMCs had a dramatic reduction in stress fiber content after treatment with the Rho-kinase inhibitor Y27632, M-RIP RNAi VSMCs were resistant. In fact, the cells with the least M-RIP expression had the least response to Y27632, whereas cells with higher M-RIP expression appeared morphologically similar to control RNAi cells treated with Y27632. These data support the model shown in Fig. 9 where loss of M-RIP expression uncouples MLCP from stress fibers. Once uncoupled, the effect of Y27632 on MLCP activity no longer affects stress fiber assembly.
MLCP activity has emerged as a critical determinant of VSMC contractile state under the control of both vasoconstrictor and vasodilator compounds. Recent evidence in vivo demonstrates its role in the pathogenesis of vascular disease and its therapeutic potential (23-30). M-RIP, known to be a RhoA, MLCP, actin and myosin binding protein, regu-lates MLC phosphorylation state in VSMCs. Our data support that this effect is primarily due to M-RIP-dependent targeting of MLCP to stress fibers, resulting in increased numbers of stress fibers and cell shape changes. By targeting MLCP to actin-myosin stress fibers, M-RIP may therefore be a critical determinant of VSMC contractile state.