PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of jbcThe Journal of Biological Chemistry
 
J Biol Chem. 2012 November 30; 287(49): 41342–41351.
Published online 2012 October 12. doi:  10.1074/jbc.M112.421040
PMCID: PMC3510832

Direct Binding and Regulation of RhoA Protein by Cyclic GMP-dependent Protein Kinase Iα*

Abstract

Vascular smooth muscle cell (VSMC) tone is regulated by the state of myosin light chain (MLC) phosphorylation, which is in turn regulated by the balance between MLC kinase and MLC phosphatase (MLCP) activities. RhoA activates Rho kinase, which phosphorylates the regulatory subunit of MLC phosphatase, thereby inhibiting MLC phosphatase activity and increasing contraction and vascular tone. Nitric oxide is an important mediator of VSMC relaxation and vasodilation, which acts by increasing cyclic GMP (cGMP) levels in VSMC, thereby activating cGMP-dependent protein kinase Iα (PKGIα). PKGI is known to phosphorylate Rho kinase, preventing Rho-mediated inhibition of MLC phosphatase, promoting vasorelaxation, although the molecular mechanisms that mediate this are unclear. Here we identify RhoA as a target of activated PKGIα and show further that PKGIα binds directly to RhoA, inhibiting its activation and translocation. In protein pulldown and immunoprecipitation experiments, binding of RhoA and PKGIα was demonstrated via a direct interaction between the amino terminus of RhoA (residues 1–44), containing the switch I domain of RhoA, and the amino terminus of PKGIα (residues 1–59), which includes a leucine zipper heptad repeat motif. Affinity assays using cGMP-immobilized agarose showed that only activated PKGIα binds RhoA, and a leucine zipper mutant PKGIα was unable to bind RhoA even if activated. Furthermore, a catalytically inactive mutant of PKGIα bound RhoA but did not prevent RhoA activation and translocation. Collectively, these results support that RhoA is a PKGIα target and that direct binding of activated PKGIα to RhoA is central to cGMP-mediated inhibition of the VSMC Rho kinase contractile pathway.

Keywords: Cell Signaling, Cyclic GMP (cGMP), Protein Kinase G (PKG), RhoA, Vascular Biology

Introduction

Vascular contraction and relaxation are highly regulated in vertebrates. Control of smooth muscle cell contractile state occurs primarily through regulation of the phosphorylation state of myosin light chains (MLCs)3 (1). In the vascular smooth muscle cell (VSMC), contraction is initiated by both calcium-dependent and -insensitive mechanisms. Increases in intracellular calcium activate the calcium/calmodulin-dependent myosin light chain kinase, leading to MLC phosphorylation and VSMC contraction (2). The calcium-insensitive pathway is mediated by G protein-coupled receptor activation of the small GTPase RhoA, which activates Rho kinase. Rho kinase then phosphorylates the regulatory subunit of myosin light chain phosphatase and inhibits its activity (36), leading to VSMC contraction and increased vascular tone (2, 6).

Conversely, relaxation of vascular smooth muscle results from a decrease in cytosolic Ca2+ concentration and/or reduced Ca2+ sensitivity of the contractile apparatus. Nitric oxide (NO), the most important endogenous vasodilator, regulates VSMC relaxation by increasing intracellular cGMP and activating cGMP-dependent kinase I (PKGI) (7, 8). The two isoforms of PKGI, PKGIα and PKGIβ, are identical except for their amino termini, which contain different leucine zipper (LZ) heptad repeat motifs that are critical for protein targeting (9). PKGIα and PKGIβ both regulate VSMC tone through interactions with key VSMC contractile proteins mediated by their respective LZ domains (8, 1012). PKGI regulates both cytosolic Ca2+ concentration and contractile sensitivity to Ca2+. Although PKGIα decreases intracellular Ca2+ concentration by activating the regulator of G protein signaling 2 (RGS2) (13), PKGI also directly reduces the sensitivity of the contractile apparatus to Ca2+ through phosphorylation and activation of MLC phosphatase (10), as well as through inhibition of the Rho/Rho kinase contractile pathway (1417).

PKGIα phosphorylation of RhoA on serine 118 inhibits RhoA activation and membrane translocation, which then inhibits Rho/Rho kinase-induced VSMC stress fiber formation and contraction (16, 17). We have shown previously that mice with selective mutations in the amino-terminal leucine zipper domain of PKGIα (leucine zipper mutant (LZM) mice) display both decreased Ser-118 phosphorylation and increased activation of RhoA in their VSMCs (18). This dysregulation of RhoA contributes to VSMC contractile abnormalities, abnormal relaxation of large and resistance blood vessels, and increased systemic blood pressure (18), supporting the importance of PKGIα regulation of RhoA in maintaining normal VSMC function in vivo.

The mechanism by which PKGIα inhibits RhoA activation is not understood. Therefore, in the present study, we tested whether PKGIα regulates RhoA activation through a direct PKGIα-RhoA interaction, using lysophosphatidic acid (LPA)-induced RhoA membrane translocation as an assay of RhoA activation. We further explored the specific PKGIα and RhoA domains required for co-interaction in vitro and in cellular lysates; and, using specific mutants of PKGIα, we determined the requirements of the PKGIα LZ domain for mediating PKGIα interaction with RhoA, and of PKGIα kinase activity for mediating inhibition of RhoA activation. Taken together, our findings define the mechanism by which PKGIα directly inhibits RhoA activity.

EXPERIMENTAL PROCEDURES

Cell Culture

The immortalized human aortic smooth muscle cell line Ao184 was established by infecting VSMCs, isolated from an explanted human aorta, with retroviral constructs containing the E6 and E7 human papillomavirus proteins as reported previously (13). COS-1 and Swiss 3T3 fibroblasts cells were obtained from American Type Culture Collection. Cells were cultured and passaged in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen) with 10% fetal bovine serum, penicillin (100 units/ml) and streptomycin (100 μg/ml). Cells were grown at 37 °C in a 5% CO2 humidified incubator. Smooth muscle cells used in this study were passages 10–18.

Subcellular Fractionation

VSMCs were grown on 100-mm dishes to ~80% confluence. The medium was then replaced with a low serum medium (DMEM containing 1% fetal bovine serum, 100 units/ml penicillin, and 100 mg/ml streptomycin) for 16 h to allow the cells to become quiescent. The cells were then treated with serum-free medium (DMEM containing antibiotics) for 4 h prior to agonist stimulation. 24 h after transfection, Swiss 3T3 cells were split into 6-well dishes, and 8 h later the medium was replaced with a low serum medium overnight, followed by treatment with serum-free medium for 4 h. The cells were stimulated with 50 μm LPA (Sigma) for different durations as noted in the results. Cells were washed with ice-cold PBS two times and scraped in 0.5 ml or 0.3 ml of lysis buffer (50 mm HEPES, pH 7.5, 50 mm NaCl, 1 mm MgCl2, 2 mm EDTA supplemented with a proteinase inhibitor mixture (Calbiochem)) in 100-mm or 6-well dishes, respectively. Cells were lysed by two sequential freeze-thaw cycles. The lysate was first centrifuged at 500 × g for 5 min to pellet the nuclear fraction and then centrifuged again at 120,000 × g for 45 min to pellet the membrane fraction. The pellet was dissolved with solubilization buffer (1% Triton X-100, 3% glycerol in lysis buffer). The pellet and the supernatant were dissolved separately in 2× sample buffer (100 mm Tris-HCl, 4% SDS, 20% glycerol, 5% 2-mercaptoethanol, pH 6.8) and boiled for 5 min (19).

Antibodies

Antibodies were raised against GST-peptides corresponding to PKGIα or LZM-PKGIα amino-terminal 59 amino acids as described (10, 13). The rabbit polyclonal anti-PKGI common (PKGIcom) antibody was from Stressgen. Anti-RhoA-interacting protein RhoA GDP dissociation inhibitor (RhoGDI) antibody was from Cell Signaling. The mouse monoclonal anti-RhoA antibody, the goat polyclonal anti-PKGIβ antibody, and the anti-goat peroxidase-conjugated secondary antibody were from Santa Cruz Biotechnology. Anti-mouse and anti-rabbit peroxidase-conjugated secondary antibodies were from Amersham Biosciences. The mouse monoclonal anti-human smooth muscle actin (1A4) antibody was from DAKO.

Immunoblotting

Protein concentrations were determined using the DC protein assay kit (Bio-Rad). Samples (20–40 μg of protein) were separated by SDS-PAGE (10 or 12.5% running, 4% stacking). The separated proteins were transferred by electrophoresis to a polyvinylidene difluoride membrane (Millipore) using a Mini Trans-Blot cell (Bio-Rad). The blots were blocked for 1 h at room temperature in PBS-T buffer (0.02% Tween 20) containing 3% nonfat milk before incubation with the primary antibody overnight at 4 °C. After washing the membranes three times in PBS-T buffer, a peroxidase-conjugated secondary antibody was added for 1 h. Immunodetection was performed with an ECL Western blotting detection kit (Amersham Biosciences). For densitometric analysis, immunoblotted bands were quantitated using an Alpha Innotech image analyzer.

Cloning and Expression Constructs

Full-length RhoA cDNA and cDNA fragments encoding RhoA amino acids 1–44 and 1–147 were amplified by PCR from human aortic cDNA library (Clontech). The 5′-primer for RhoA was 5′-GCG GGA TCC ATG GCT GCC ATC CGG AAG-3′. The 3′-primers were 5′-CGC GTC GAC TCA CAA GAC AAG GCA CCC AGA TTT-3′, 5′-CGC GTC GAC TCA GCC ACA TAG TTC TCA AAC AC-3′, and 5′-CGC GTC GAC TCA CAT ATC TCT GCC TTC TTC AGG-3′ for full-length, RhoA amino acids 1–44, and 1–147, respectively. The PCR products were cleaved by BamHI-SalI and cloned into pGEX-4T-3 (Amersham Biosciences). To generate truncated PKGIα DNA fragments (amino acids 237–671, 360–619, 237–359, and 620–671), cDNA fragments were amplified from the full-length pcDNA 3.1/PKGIα or pcDNA 3.1/PKGIβ DNA (20) by PCR. The following primers were designed for the investigation. The 5′-primer for PKGIα(237–671) and PKGIα(237–359) was 5′-GCC GAA TTC CTT GAA GAG ACC CAC TAT GAA AAT-3′, for PKGIα(360–619) was 5′-GCC GAA TTC TTC AAC ATC ATT GAT ACC CTT GG-3′, and for PKGIα(620–671) was 5′-GCC CTC GAG ATC AGA CAG CTT CAG GTT GGC-3′. The 3′-primer for PKGIα(237–671) and PKGIα(620–671) was 5′-GCC CTC GAG GAA GTC TAT ATC CCA TCC TGA G-3′, for PKGIα(360–619) was 5′-GCC CTC GAG AAA CCA TTT GTG CTT TTG AAT GTC-3′, and for PKGIα(237–359) was 5′-GCC GAA TTC GAG GGC TTT AAC TGG GAA GGC-3′. The PCR products were digested with EcoRI-XhoI and inserted into the plasmid DNA pGEX-4T-1. The inserts were confirmed by sequencing. The catalytically inactive mutant of PKGIα (21) was kindly provided by A. P. Broun (University of Calgary).

Transient Transfection

PKGIα constructs and FLAG-RhoA proteins were overexpressed in COS-1 cells by using Lipofectamine 2000 (Invitrogen). Briefly, transfection was performed with 12 μg of DNA plasmid and 45 μl of Lipofectamine 2000 in 3 ml of Opti-MEM reduced serum medium (Invitrogen) in 100-mm dish plates (4 × 105 cells/dish). The PKGIα constructs were overexpressed in Swiss 3T3 cells using Lipofectamine LTX (Invitrogen). Briefly, transfection was performed with 15 μg of DNA plasmid, 30 μl of PLUS Reagent, and 90 μl of Lipofectamine LTX in 3 ml of Opti-MEM reduced serum medium in 100-mm dish plates, which contained 4 × 105 cells/dish.

GST Fusion Protein Expression in Escherichia coli

GST fusion proteins of RhoA fragments and PKGIα fragments were expressed in E. coli BL-21DE3 (Stratagene) and transcription induced with 0.2 or 1 mm isopropyl-d-thiogalactopyranoside as described previously (13). The expressed GST fusion proteins were purified on glutathione-Sepharose beads (Novagen).

In Vitro Pulldown Assay

PKGIα protein purified from bovine lung (Promega) in binding buffer (50 mm Tris-HCl, 250 mm NaCl, 0.05% Nonidet P-40, 30 mm MgCl2, pH 7.4) with protease inhibitor mixture (4 μg/ml) was mixed with 20–30 μl of GST-RhoA fragments bound to beads. Samples were rocked for 2 h at 4 °C (22). In control experiments, GST-alone beads were also tested for PKGI binding. For cell lysate assays, COS-1 cells were harvested in TLB buffer (20 mm Tris-HCl, 137 mm NaCl, 2 mm EDTA, 10% glycerol, 1% Triton X-100, 25 mm β-glycerol phosphate, pH 7.4) 48 h after transfection. The cells were lysed by five passes through a 27-gauge needle on ice. The supernatant obtained by centrifugation (15,000 rpm, 30 min) was then mixed with 20–30 μl of GST fusion protein beads and rocked for 2 h at 4 °C. After a thorough washing with binding buffer, the bound proteins were solubilized in 2×SDS sample buffer and analyzed by 12.5% SDS-PAGE and immunoblotting. To indicate the amount of protein used in each pulldown assay, 5% of the input lysate was loaded in the input lane. Beads loaded with GST alone were included as negative controls in all pulldown experiments. Similar loading of GST fusion proteins was confirmed by Ponceau S staining.

GDPβS and GTPγS Binding Assay

GST-RhoA fusion beads were prepared as described above and incubated with 200 μm stable GTP/GDP analogue GTPγS or GDPβS for 15 min at room temperature. The reaction was stopped with MgCl2. After centrifugation at top speed in a microcentrifuge, beads were washed three times with binding buffer followed by the addition of purified PKGIα (20 μg/ml) for 1 h at 4 ºC. Samples were then centrifuged as above, washed three times, and eluted with SDS sample buffer followed by SDS-PAGE and Western blotting for PKGI with the PKGIcom antibody.

Co-immunoprecipitation (CoIP) Assay

Cells were harvested in ice-cold TLB buffer 48 h after transfection (22). The cells were lysed by five passes through a 27-gauge needle on ice, and the supernatant was obtained by centrifugation (15,000 rpm, 30 min). The supernatant (200 μg of total protein) was mixed with target antibodies (3–4 μg) or mouse immunoglobulin as a negative control for 12 h rocking at 4 °C. Fresh protein G conjugated to agarose was then added, followed by 2–3-h rocking at 4 °C. Immunoprecipitates were centrifuged at 2,500 rpm for 2 min at 4 °C. The supernatant was then discarded, and the pellet was washed four times with binding buffer and then resuspended with the same volume of 2×SDS sample buffer.

cGMP-Agarose Affinity Assay

For the assay with purified protein, purified PKGIα protein in TLB buffer with a proteinase inhibitor mixture at a concentration of 4 μg/ml was mixed with His-RhoA (4 μg/ml) and rocked overnight at 4 °C (22, 23). For the cell lysate assays, the supernatants were obtained by centrifugation (15,000 rpm, 30 min) and then rocked for 2 h at 4 °C. Fresh nonconjugated agarose, cGMP or Rp-8AET-cGMPS conjugated agarose (BIOLOG LIFE Science Institute) were then added with a further 2–3-h rocking. Complexes were centrifuged at 2,500 rpm for 2 min at 4 °C, the supernatant was discarded, and the pellet was washed four times with TLB buffer. The pellet was then resuspended in same volume of 2×SDS sample buffer, boiled for 5 min, and the supernatant was analyzed by 10% SDS-PAGE followed by immunoblotting. To indicate the amount of protein used in each pulldown assay, 5% of the input lysate was loaded in the input lane. Controls using agarose-only loaded beads were included as negative controls in all affinity assay experiments.

Statistics

All results are expressed as the mean ± S.E. Significance was tested by Student's t test. A p value of <0.05 was considered significant.

RESULTS

PKGIα Regulates RhoA Translocation in Human Aortic Smooth Muscle Cells

In response to a variety of stimuli, GTP binding activates RhoA and stimulates its translocation to the plasma membrane, making this process a reliable assay for RhoA activation. In cultured human Ao184 VSMCs, treatment with the RhoA activator LPA (50 μm) induced RhoA translocation from the cytosolic to the membrane fraction in a time-dependent manner (0, 5, 15, 30 min) (data not shown), reaching a plateau at 30 min with membrane-associated RhoA increasing ~2-fold compared with vehicle-treated cells (Fig. 1A). Pretreatment with the cGMP analogue and direct PKG activator 8-Br-cGMP (100 μm) completely inhibited LPA-stimulated RhoA membrane translocation (Fig. 1A). In each experiment, the quantities of RhoA in the cytosolic fractions and of PKGI in both the cytosolic and membrane fractions did not change (data not shown). Western blotting with PKGI isoform-specific antibodies confirmed that >90% of PKGI in Ao184 cells was PKGIα (data not shown). Furthermore, in Swiss 3T3 cells transfected with PKGIα, LPA-induced RhoA translocation was also inhibited by 8-Br-cGMP treatment (Fig. 1B). However, in Swiss 3T cells transfected with leucine zipper mutant PKGIα (PKGIα LZM), the inhibitory effect of 8-Br-cGMP was abolished (Fig. 1C). These data support a critical role of PKGIα, through its LZ domain, in regulating membrane translocation of RhoA.

FIGURE 1.
PKGI blocks LPA-induced membrane translocation of RhoA. A, representative Western blot and summary densitometry data of RhoA distribution in the membrane fraction prepared from human Ao184 vascular smooth muscle cells. Cells were treated for 30 min with ...

PKGIα Interacts Directly with RhoA

Next, we tested for a direct interaction of PKGIα with RhoA. First, we performed pulldown experiments incubating GST-RhoA fusion proteins with lysates from COS-1 cells overexpressing PKGIα and demonstrated that PKGIα complexes with GST-RhoA, but not with GST alone (Fig. 2A). In vitro mixing assays using purified PKGIα and GST-RhoA demonstrated that this complex formation results from a direct interaction between PKGIα and RhoA (Fig. 2B). In separate experiments, CoIP studies using lysates of COS-1 cells overexpressing both FLAG-tagged RhoA and full-length PKGIα also demonstrated complex formation between PKGIα and RhoA (Fig. 3). Identical results were observed with in vitro immunoprecipitations using purified His-RhoA and purified PKGIα (data not shown). Because PKGIβ is also expressed in the VSMC we examined whether RhoA binds PKGIβ in lysates from COS-1 cells transfected with the PKGIβ isoform. GST-RhoA pulled down PKGIβ, but to a significantly lesser extent than PKGIα (Fig. 2C). We next investigated the effect of RhoA nucleotide binding state on its interaction with PKGIα. Purified PKGIα complexed similarly with GDP-treated GST-RhoA, compared with GTP-treated GST-RhoA, supporting that the PKGIα-RhoA interaction is not regulated by RhoA nucleotide binding (Fig. 4A). Furthermore, RhoGDI did not precipitate with either full-length GST-RhoA (Fig. 4B) or with PKGI precipitated with cGMP-conjugated beads from Ao184 cells (Fig. 4C).

FIGURE 2.
GST-RhoA interacts directly with PKGIα. Immunoblots (IB) from cell lysates (A and C) or purified PKGIα (B) are shown. GST-RhoA fusion proteins were immobilized with GSH beads and incubated for 2 h at 4 ºC with lysates from COS-1 ...
FIGURE 3.
Co-immunoprecipitation of RhoA with PKGIα in cell lysates. Immunoprecipitation (IP) of RhoA from COS-1 cells transiently transfected for 48 h with FLAG-RhoA and/or PKGIα is shown. Immunocomplexes from overnight incubation with antibody ...
FIGURE 4.
Examination of dependence of RhoA-PKGIα interaction on RhoA nucleotide binding state. A, GST-RhoA fusion proteins were incubated with either GTPγS (denoted GTP) or GDPβS (denoted GDP), followed by incubation with purified PKGIα. ...

RhoA Switch I Domain Interacts Directly with PKGIα

To define the specific domains of RhoA required for binding to PKGIα, GST fusion proteins were generated containing RhoA Switch I domain (residues 1–44); RhoA Switch I and Switch II domains (residues 1–147); full-length RhoA (residues 1–192); or GST alone (Fig. 5A). All three GST-RhoA constructs, but not GST alone, precipitated PKGIα, supporting that residues 1–44, containing the Switch I domain, are sufficient for interaction with PKGIα (Fig. 5B).

FIGURE 5.
Identification of the PKGIα binding domains of RhoA. A, schematic of RhoA deletion mutants used in the experiments is shown. ED, effector domain; IR, insertion region (not in Ras); HV, hypervariable region (differs between Rho/Ras isoforms). ...

PKGI LZ Domain Interacts Directly with RhoA

We next investigated the domain(s) of PKGIα mediating the PKGIα-RhoA interaction. GST fusion proteins containing specific domains of PKGIα were generated as represented in Fig. 6A and incubated with purified His-RhoA, followed by immunoblotting with anti-RhoA antibody. RhoA was detected only in pulldowns containing the PKGIα LZ domain, using either the PKGIα LZ domain alone (residues 1–59) or the LZ-containing-N-terminal portion of PKGIα (residues 1–236) (Fig. 6B).

FIGURE 6.
Identification of the RhoA binding domains of PKGIα. A, schematic of PKGIα deletion mutants used in the experiments is shown. B, GST-truncated PKGIα fusion proteins were immobilized with GSH beads and incubated for 2 h at 4 ºC ...

cGMP-activated PKGIα, but Not Inactive PKGIα, Binds RhoA Directly

We next investigated whether RhoA-PKGIα binding requires cGMP binding of PKGIα. Lysates of COS-1 cells transfected with both PKGIα and FLAG-RhoA were incubated with agarose beads conjugated with either cGMP or with the cGMP antagonist Rp-8AET-cGMPS to isolate activated or inhibited PKGIα, respectively. Agarose beads were used as controls. cGMP-bound PKGIα formed a complex with RhoA, but Rp-8AET-cGMPS-bound (inactivated) PKGIα failed to complex with RhoA (Fig. 7, A and B). Similar results were observed using purified His-RhoA and purified PKGIα rather than cell lysates (Fig. 7, C and D). cGMP-conjugated agarose did not directly bind with RhoA (data not shown). To test whether binding of cGMP-bound PKGIα requires a functional LZ domain, COS-1 cells transfected with FLAG-RhoA and with either wild type PKGIα or with PKGIα-LZM were incubated with cGMP-conjugated agarose or with agarose alone. Whereas cGMP-bound PKGIα interacted with RhoA, cGMP-bound LZM-PKGIα precipitated substantially less RhoA (Fig. 8).

FIGURE 7.
PKGIα activation by cGMP mediates RhoA-PKGIα binding. Co-expression of PKGIα and FLAG-RhoA in COS-1 cells lysed 48 h after transient transfection (A and B) or mixture of purified PKGIα and His-RhoA (C and D), followed by ...
FIGURE 8.
Requirement of functional PKGIα leucine zipper domain for RhoA interaction with activated PKGIα. A, co-expression of FLAG-RhoA and PKGIα or LZM-PKGIα in COS-1cells, followed by isolation on cGMP agarose for 2 h at 4 ºC. ...

PKGIα Kinase Activity Is Required for RhoA Inhibition by PKGIα, but Not for PKGIα-RhoA Binding

We next examined whether PKGIα kinase activity is required for PKG-mediated inhibition of RhoA (as assayed by RhoA membrane translocation). Swiss 3T3 cells, which lack detectable endogenous PKG (17), were transfected with mutant PKGIα lacking kinase activity (kinase inactive; KI-PKGIα) (21) and were treated with LPA with or without 8-Br-cGMP. In the KI-PKGIα-transfected cells, LPA induced RhoA membrane translocation, but 8-Br-cGMP failed to inhibit LPA-induced RhoA membrane translocation (Fig. 9A), supporting that PKGIα requires intact kinase activity to inhibit RhoA. Next, we investigated whether PKGIα kinase activity is also required for PKGIα-RhoA interaction. cGMP-agarose or agarose-alone binding assays were performed on lysates from COS-1 cells transfected with FLAG-RhoA and with either KI-PKGIα or full-length PKGIα, followed by Western blotting for FLAG-RhoA. We observed no decrease in RhoA interaction with cGMP-bound KI-PKGIα compared with wild type PKGIα (Fig. 9C). These results indicate that PKGIα kinase activity is required for inhibition of RhoA activation, but not for cGMP-induced interaction with RhoA.

FIGURE 9.
PKGIα catalytic activity is required to inhibit RhoA membrane anchoring, but not for RhoA-PKGIα co-interaction. A and B, distribution of RhoA in cytosolic and membrane fractions prepared from Swiss 3T3 fibroblast cells 48 h after transient ...

DISCUSSION

In the present study, we define the mechanism by which PKGIα inhibits RhoA activation in vitro and in cultured cells. First, we have demonstrated that the Iα isoform of PKGI inhibits LPA-induced RhoA membrane translocation in both aortic smooth muscle cells and in 3T3 cells. We show further that PKGIα and RhoA form a complex using both GST pulldown and CoIP assays. In vitro mixing experiments support direct binding between these two proteins. Furthermore, this interaction requires both an intact PKGIα LZ domain and cGMP activation, but complex formation does not require intact PKGIα kinase activity. In contrast to this, intact kinase activity is required for PKGI-mediated inhibition of RhoA activation and membrane translocation. Taken together, these data support a mechanism by which PKGIα binds directly to RhoA through an interaction involving the PKGIα LZ domain and inhibits RhoA activation through an event requiring PKGIα kinase activity.

Our prior work has demonstrated that homozygous knock-in mice harboring discrete mutations in the PKGIα LZ domain (LZM mice) develop hypertension and derangements in vascular relaxation, accompanied by decreased RhoA phosphorylation in the VSMCs and increased RhoA activity (18). Therefore, the current study clarifies that this PKGIα regulation of RhoA occurs via direct binding and defines a mechanism of PKGIα-RhoA regulation of vascular relaxation.

Other groups have previously observed that PKGI can phosphorylate RhoA in vitro (16) and that cGMP regulates RhoA effects on VSMC structure and function (1618). However, our findings represent the first demonstration of the requirement for direct binding of PKGIα and RhoA for regulation of RhoA function. Moreover, we have observed co-precipitation of these proteins by GST pulldown assays, immunoprecipitations, and through cGMP-conjugated agarose affinity assays.

Our GST binding studies demonstrate that LZ-containing domains of PKGIα mediate PKGIα-RhoA binding. Our GST pulldown experiments support that the LZ domain of PKGIα alone is sufficient to bind RhoA. LZM PKGIα was unable to be purified, thus precluding comparison of the in vitro binding of WT and LZM PKGIα to RhoA (data not shown). However, our experiments in cultured cells with the full-length PKGIα-LZM protein demonstrate that the normal domain is necessary for full binding of RhoA-PKGIα. Additionally, we also detected increased RhoA pulldown from the PKGIα(1–236) domain compared with the PKGIα(1–59) domain alone. Although this analysis is admittedly qualitative and measuring true Kd of this interaction is beyond the scope of this study, these findings do suggest that additional domains of PKGIα between amino acids 59 and 236 may contribute to the PKGIα-RhoA interaction. In previous work, we have identified other LZ-mediated binding partners of PKGIα, including RGS2 (13), formin homology domain protein (FHOD1) (22), and MLC phosphatase (10). Our finding of an LZ-mediated interaction between PKGIα and RhoA therefore reveals RhoA as a new cGMP-regulated PKGIα target in the VSMCs of potential importance in vascular function and regulation.

Our GST pulldown studies also demonstrate that PKGIα binds primarily to the conserved Switch I domain of RhoA. Importantly, the Switch I domain is highly conserved throughout a number of small GTPase molecules, including RGS2, Ras, and CDC42 (24), suggesting a possible shared region of PKGIα binding. These data therefore suggest potential mechanisms of regulation of other Rho family GTPases by PKGIα.

We have also observed, from the cGMP-conjugated bead experiments, that PKGIα binding by cGMP increases PKGIα-RhoA interaction, but that PKGIα kinase activity is not required for PKGIα-RhoA binding. One potentially confusing finding is that Rp-8AET-cGMPS, which inhibits PKGI kinase activity by preventing disengagement of the kinase and autoinhibitory sites, also disrupts normal LZ-dependent PKGIα-RhoA binding. Others have shown, however, that liberation of the PKIα kinase domain from the autoinhibitory site leads to increased solubility of the LZ domain (25). Therefore, we interpret our findings to support that cGMP-mediated positioning of the PKGIα LZ domain, as well as its kinase activating effect, mediates PKGIα RhoA binding. These findings clarify the importance of cGMP in regulating RhoA activation; and our observations that kinase-inactive PKGIα binds RhoA identically to WT PKGIα, but fails to inhibit RhoA activation, represent the first demonstration of a direct requirement for PKGI kinase activity in the regulation of RhoA function. Together, these data define a mechanism by which cGMP-bound PKGIα binds RhoA in a LZ domain-dependent manner, presumably leading to RhoA phosphorlyation and inhibition of RhoA activity.

An understanding of the molecular mechanism of cGMP and PKGIα regulation of RhoA is of particular importance to the study of hypertension. Current dogma for human hypertension states that all causes of hypertension arise from renal-mediated derangements in salt handling. However, a number of mouse models, including the LZM model described above (18), support that primary vascular abnormalities are sufficient to induce hypertension. The current findings elucidating the mechanism by which PKGIα regulates RhoA activation in VSMCs therefore have potentially important implications for our understanding of the homeostatic mechanisms that maintain normal systemic blood pressure.

Acknowledgments

We thank Andrew P. Braun (University of Calgary, Calgary, Alberta, Canada) for kindly supplying the catalytically inactive mutant of PKGIα.

*This work was supported by the National Institutes of Health (HL55309, to M. E. M.), the American Heart Association (10SDG2630161, to R. M. B.), and by Daiichi Sankyo Co., Ltd., Tokyo, Japan.

3The abbreviations used are:

MLC
myosin light chain
8-Br-cGMP
8-bromo-cGMP
GDBβS
guanosine 5′-O-(β-thio)diphosphate
GTPγS
guanosine 5′-(γ-thio)triphosphate
KI
kinase-inactive
LPA
lysophosphatidic acid
LZ
leucine zipper
LZM
LZ mutant
PKGI
cGMP-dependent protein kinase I
RGS2
regulator of G protein signaling 2
RhoGDI
RhoA GDP dissociation inhibitor
VSMC
vascular smooth muscle cell.

REFERENCES

1. Somlyo A. P. (1997) Signal transduction: rhomantic interludes raise blood pressure. Nature 389, 908–911 [PubMed]
2. Somlyo A. P., Somlyo A. V. (1994) Signal transduction and regulation in smooth muscle. Nature 372, 231–236 [PubMed]
3. Gong M. C., Iizuka K., Nixon G., Browne J. P., Hall A., Eccleston J. F., Sugai M., Kobayashi S., Somlyo A. V., Somlyo A. P. (1996) Role of guanine nucleotide-binding proteins–ras-family or trimeric or both–in Ca2+ sensitization of smooth muscle. Proc. Natl. Acad. Sci. U.S.A. 93, 1340–1345 [PubMed]
4. Otto B., Steusloff A., Just I., Aktories K., Pfitzer G. (1996) Role of Rho proteins in carbachol-induced contractions in intact and permeabilized guinea-pig intestinal smooth muscle. J. Physiol. 496, 317–329 [PubMed]
5. Fujihara H., Walker L. A., Gong M. C., Lemichez E., Boquet P., Somlyo A. V., Somlyo A. P. (1997) Inhibition of RhoA translocation and calcium sensitization by in vivo ADP-ribosylation with the chimeric toxin DC3B. Mol. Biol. Cell 8, 2437–2447 [PMC free article] [PubMed]
6. Uehata M., Ishizaki T., Satoh H., Ono T., Kawahara T., Morishita T., Tamakawa H., Yamagami K., Inui J., Maekawa M., Narumiya S. (1997) Calcium sensitization of smooth muscle mediated by a Rho-associated protein kinase in hypertension. Nature 389, 990–994 [PubMed]
7. Murad F. (2006) Nitric oxide and cyclic GMP in cell signaling and drug development. N. Engl. J. Med. 355, 2003–2011 [PubMed]
8. Hofmann F. (2005) The biology of cyclic GMP-dependent protein kinases. J. Biol. Chem. 280, 1–4 [PubMed]
9. Orstavik S., Natarajan V., Taskén K., Jahnsen T., Sandberg M. (1997) Characterization of the human gene encoding the type Iα and type Iβ cGMP-dependent protein kinase (PRKG1). Genomics 42, 311–318 [PubMed]
10. Surks H. K., Mochizuki N., Kasai Y., Georgescu S. P., Tang K. M., Ito M., Lincoln T. M., Mendelsohn M. E. (1999) Regulation of myosin phosphatase by a specific interaction with cGMP-dependent protein kinase Iα. Science 286, 1583–1587 [PubMed]
11. Ignarro L. J., Kadowitz P. J. (1985) The pharmacological and physiological role of cyclic GMP in vascular smooth muscle relaxation. Annu. Rev. Pharmacol. Toxicol. 25, 171–191 [PubMed]
12. Geiselhöringer A., Werner M., Sigl K., Smital P., Wörner R., Acheo L., Stieber J., Weinmeister P., Feil R., Feil S., Wegener J., Hofmann F., Schlossmann J. (2004) IRAG is essential for relaxation of receptor-triggered smooth muscle contraction by cGMP kinase. EMBO J. 23, 4222–4231 [PubMed]
13. Tang K. M., Wang G. R., Lu P., Karas R. H., Aronovitz M., Heximer S. P., Kaltenbronn K. M., Blumer K. J., Siderovski D. P., Zhu Y., Mendelsohn M. E., Tang M., Wang G. (2003) Regulation of G-protein signaling 2 mediates vascular pressure. Nat. Med. 9, 1506–1512 [PubMed]
14. Somlyo A. P., Somlyo A. V. (2000) Signal transduction by G-proteins, Rho-kinase and protein phosphatase to smooth muscle and nonmuscle myosin II. J. Physiol. 522, 177–185 [PubMed]
15. Fukata Y., Amano M., Kaibuchi K. (2001) Rho-Rho-kinase pathway in smooth muscle contraction and cytoskeletal reorganization of nonmuscle cells. Trends Pharmacol. Sci. 22, 32–39 [PubMed]
16. Sauzeau V., Le Jeune H., Cario-Toumaniantz C., Smolenski A., Lohmann S. M., Bertoglio J., Chardin P., Pacaud P., Loirand G. (2000) Cyclic GMP-dependent protein kinase signaling pathway inhibits RhoA-induced Ca2+ sensitization of contraction in vascular smooth muscle. J. Biol. Chem. 275, 21722–21729 [PubMed]
17. Sauzeau V., Rolli-Derkinderen M., Marionneau C., Loirand G., Pacaud P. (2003) RhoA expression is controlled by nitric oxide through cGMP-dependent protein kinase activation. J. Biol. Chem. 278, 9472–9480 [PubMed]
18. Michael S. K., Surks H. K., Wang Y., Zhu Y., Blanton R., Jamnongjit M., Aronovitz M., Baur W., Ohtani K., Wilkerson M. K., Bonev A. D., Nelson M. T., Karas R. H., Mendelsohn M. E. (2008) High blood pressure arising from a defect in vascular function. Proc. Natl. Acad. Sci. U.S.A. 105, 6702–6707 [PubMed]
19. Fleming I. N., Elliott C. M., Exton J. H. (1996) Differential translocation of Rho family GTPases by lysophosphatidic acid, endothelin-1, and platelet-derived growth factor. J. Biol. Chem. 271, 33067–33073 [PubMed]
20. Christensen E. N., Mendelsohn M. E. (2006) Cyclic GMP-dependent protein kinase Iα inhibits thrombin receptor-mediated calcium mobilization in vascular smooth muscle cells. J. Biol. Chem. 281, 8409–8416 [PubMed]
21. Swayze R. D., Braun A. P. (2001) A catalytically inactive mutant of type I cGMP-dependent protein kinase prevents enhancement of large conductance, calcium-sensitive K+ channels by sodium nitroprusside and cGMP. J. Biol. Chem. 276, 19729–19737 [PubMed]
22. Wang Y., El-Zaru M. R., Surks H. K., Mendelsohn M. E. (2004) Formin homology domain protein (FHOD1) is a cyclic GMP-dependent protein kinase I-binding protein and substrate in vascular smooth muscle cells. J. Biol. Chem. 279, 24420–24426 [PubMed]
23. Burgoyne J. R., Madhani M., Cuello F., Charles R. L., Brennan J. P., Schröder E., Browning D. D., Eaton P. (2007) Cysteine redox sensor in PKGIα enables oxidant-induced activation. Science 317, 1393–1397 [PubMed]
24. Ridley A. J. (2001) Rho proteins: linking signaling with membrane trafficking. Traffic 2, 303–310 [PubMed]
25. Alverdi V., Mazon H., Versluis C., Hemrika W., Esposito G., van den Heuvel R., Scholten A., Heck A. J. (2008) cGMP-binding prepares PKG for substrate binding by disclosing the C-terminal domain. J. Mol. Biol. 375, 1380–1393 [PubMed]

Articles from The Journal of Biological Chemistry are provided here courtesy of American Society for Biochemistry and Molecular Biology