| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 280, Issue 38, 33083-33095, September 23, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Pathology, St. Louis University School of Medicine, St. Louis, Missouri 63104
Received for publication, March 22, 2005 , and in revised form, July 5, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
cAMP is a primary mediator of the biological effects of 
-adrenergic hormones, including hormone-induced relaxation of smooth muscle tissue (6-9). In 1984 de Lanerolle et al. (10) proposed, as a mechanism for smooth muscle relaxation, that increases in cAMP lead to activation of PKA, which phosphorylates MLCK inhibiting its catalytic activity and causing a decrease in myosin II activity. Since this hypothesis was proposed, two general classes of MLCKs, nonmuscle and smooth muscle MLCK, have been identified. The nonmuscle or larger form ranges in size between 206 and 220 kDa (11-14), and the smooth muscle or smaller form ranges in size between 125 and 155 kDa (15-18). Both classes of MLCKs have identical biochemical properties in vitro and exhibit overlapping tissue distribution patterns (11, 19, 20).
Numerous studies have shown that smooth muscle MLCK125-155 has multiple sites that can be phosphorylated by several kinases such as PKA, protein kinase C, and Ca2+/CaM-dependent kinase II in vitro at two residues, site A (Ser992) and site B (Ser1005), in the C-terminal region near the calmodulin binding domain (8, 21, 22). Phosphorylation at these sites results in desensitization of MLCK to activation by Ca2+/CaM. In addition, inhibition of MLCK activity by p21-activated kinase has been reported (23, 24). However, little information is available documenting the effects of phosphorylation on the nonmuscle form of MLCK206-220.
In nonmuscle cells, elevated cAMP has been shown to cause the following: 1) alterations in cell morphology; 2) inhibition of cell motility; 3) disruption of actin filaments; 4) pericyte relaxation in a silicone rubber wrinkling assay; and 5) fibroblast relaxation in a collagen gel assay (1, 25-29). In endothelial cells, preliminary studies have shown that elevations in cAMP correlate with an increase in MLCK phosphorylation (13, 30) and a decrease in RLC phosphorylation. Nonmuscle MLCK has been shown to contain putative PKA consensus phosphorylation sites, and PKA has been reported to inhibit MLCK activity in vitro. Based on these initial observations, it has been inferred that the enhanced barrier function induced by cAMP results from PKA phosphorylation of MLCK and inhibition of MLCK activity leading to decreased RLC phosphorylation and endothelial cell relaxation. For the last 15 years there has been considerable controversy regarding the physiological importance of MLCK phosphorylation by PKA. It is clear that cAMP modulates endothelial barrier function and prevents increases in vascular permeability to inflammatory mediators. That barrier protection and inhibition of endothelial cell contraction are a result of PKA phosphorylation of MLCK remains unsubstantiated.
The Rho family of GTPases, which includes Cdc42, Rac, and Rho, has been shown to play a pivotal role in dynamic rearrangements of the cytoskeleton. Rho cycles between the GDP-bound inactive state and GTP-bound active state and has been implicated in formation of stress fibers, cell motility, cell adhesion, smooth and nonmuscle cell contraction, neurite retraction, and cell cycle progression (reviewed in Refs. 31-33). Upon binding to GTP, Rho interacts with and activates multiple downstream effectors such as mDia, Rho kinase, and phosphatidylinositol 4,5-kinase, which are involved in actin polymerization, formation of focal adhesion/stress fibers, and cell contraction. RhoA activity is tightly controlled by activating factors such as the guanine nucleotide exchange factors that catalyze the exchange of GDP for GTP and by the following inactivating factors: GTPase-activating proteins that enhance hydrolysis of GTP to GDP. In resting cells RhoA is also regulated by RhoGDI (guanosine nucleotide dissociation inhibitor) that binds GDP-Rho and sequesters the complex within the cytoplasm and prevents its activation. In addition, recent evidence has emerged suggesting the PKA/c-GMP-dependent protein kinase pathways act as negative regulators of RhoA function (34-36).
Clostridium botulinum exoenzyme C3 catalyzes ADP-ribosylation of RhoA at Asn41 (37) and exerts its effects on cell morphology and F-actin stress fibers. Inhibition of Rho by C3 leads to changes in cell morphology and loss of stress fibers similar to those observed in cells treated with either dibutyryl cAMP or forskolin. These observations led to the hypothesis that cAMP and C3 may share a common pathway. Lang et al. (36) studying the effects of cAMP and C3 on NK cells documented that both agents produced similar effects on cell morphology and motility. This observation led to the demonstration that PKA catalyzes phosphorylation of RhoA on Ser188, and this phosphorylation inhibits RhoA activity establishing an alternative pathway for terminating RhoA signaling. Since these initial reports, studies have confirmed that the RhoA signaling cascade is regulated by PKA phosphorylation establishing an alternative pathway for regulating Rho GTPases (35, 36, 38). Although evidence has accumulated establishing PKA as a modulator of RhoA activity, the downstream effects of PKA inhibition of RhoA remain unidentified.
Studies presented in this report characterize the effect of elevated intracellular cAMP on basal and agonist-stimulated isometric tension in confluent endothelial cell monolayers. We provide evidence that in endothelial cells the decrease in RLC phosphorylation and the drop in isometric tension upon elevation of cAMP are not mediated via PKA phosphorylation of MLCK210 but through RhoA inactivation, which alters the activation of effector proteins within the Rho/Rho kinase signaling cascade. Furthermore, evidence is presented showing that both myosin phosphatase and cofilin activity increase in the presence of RhoA inactivation leading to the decrease in RLC phosphorylation, disruption of F-actin, and the drop in endothelial cell isometric tension observed upon elevation of cAMP.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cell Labeling—BPAE monolayers were washed in phosphate-free MEM containing 0.25% bovine serum albumin and labeled with [32P]orthophosphoric acid as described previously (40). Monolayers were incubated in low phosphate media containing 10 µM Fsk, 40 µM IBMX (Fsk/IBMX) and 75 µCi/ml [32P]orthophosphoric acid for the duration of the experiment.
Protein Purification—Recombinant rabbit smooth muscle MLCK (MLCK155) was expressed in SF-9 cells (23, 41); recombinant nonmuscle myosin II RLC was expressed in BL21 (DE3) bacteria (41) and purified as described previously (42).
Antibodies—Polyclonal antibodies specific to MLCK155 were generated against purified full-length protein. Antibodies specific to nonmuscle myosin II RLC were generated to recombinant light chain (41) coupled to tuberculin PPD (Statens Seriuminstitut, Denmark) with glutaraldehyde. Polyclonal antibodies were raised to purified MLCK155 in New Zealand White rabbits, and the IgG fractions were purified from pooled rabbit serum. MLCK155 antibodies were affinity-purified as described by Goeckeler and Wysolmerski (40). A monoclonal antibody raised to chicken gizzard MLCK was purchased from Sigma (M7905; clone K36). Affinity-purified rabbit polyclonal cofilin (catalog number 3312) and phosphocofilin (Ser3; catalog number 3311) peptide antibodies were purchased from Cell Signaling Technology. Rabbit polyclonal antibody to RhoA (catalog number SC-178) was from Santa Cruz Biotechnology. A phospho-RhoA-Ser188 peptide antibody (catalog number ST1035) was from Calbiochem. Rabbit polyclonal antibody to LIM kinase (catalog number L2290) was from Sigma, and a phospho-LIM kinase 1/2 peptide antibody (catalog number NSB1076) was from Novus Biologicals, Inc.
Measurement of Intracellular cAMP and Protein Kinase A Activity—BPAE monolayers grown on 35 x 10-mm dishes were incubated with 0.5, 1, 5, 10, 20, 30, or 50µM Fsk alone and then with 10µM Fsk and 20, 40, 80, 100, and 500µM or 1 mM IBMX for 10 min. To determine cAMP levels, the cultures were washed with 1 ml of MEM, 0.25% bovine serum albumin containing 40 µM IBMX and flooded with 0.5 ml of working lysis reagent. Samples were processed, and cAMP was measured by using the cAMP Biotrack EIA system (Amersham Biosciences catalog number RPN 225).
To measure PKA activity, monolayers were flooded with 100 µl of ice-cold PKA homogenization buffer (50 mM Tris, 25 mM KH2PO4, 25 mM K2HPO4, 5 mM EDTA, 5 mM EGTA, 1 mM IBMX, 0.2 mM phenylmethylsulfonyl fluoride, 10 µg/ml L-1-tosylamido-2-phenylethyl chloromethyl ketone, 1-chloro-3-tosylamido-7-amino-2-heptanone, aprotinin, leupeptin, and pepstatin, pH 7.2). PKA activity was measured in supernatants immediately after cell lysis using a protein kinase A assay (Transbio Corp., Baltimore, MD) following the manufacturer's instructions. cAMP concentration and PKA activity were normalized to total protein.
Isometric Tension Measurements—Isometric tension measurements were performed as described in detail previously (40, 43). BPAE cells were seeded onto precast collagen gels and maintained at 37 °C in a humidified 5% CO2 atmosphere. Monolayers were confluent within 2-3 days after seeding and used for experiments after basal isometric tension stabilized (5-7 days after confluence).
Analysis of Myosin Light Chain Phosphorylation—Myosin II RLC phosphorylation states were separated by urea/glycerol gel electrophoresis (44). Control and treated monolayers were flooded with 1 ml of ice-cold 10% trichloroacetic acid containing 10 mM DTT and processed as described previously (23, 45). The pellets were dissolved in 80 µlof9.5 M urea, 10 mM DTT, 20 mM Tris, 23 mM glycine, and 0.04% bromphenol blue, pH 8.8.
Urea/glycerol PAGE and Western blotting were carried out as described previously (41). Myosin II RLCs were detected by ECL methodology (Amersham Biosciences). X-ray films were developed at various time intervals to obtain an exposure within the linear range of the films. Myosin RLC quantitation was performed as described previously (40).
MLCK Immunoprecipitation and Activity Measurements—MLCK210 was immunoprecipitated from either [32P]orthophosphoric labeled or unlabeled monolayers with a rabbit polyclonal MLCK antibody (227423) as described previously (23). Immunoprecipitates were electrophoresed on 7.5% SDS gels, transferred to PVDF membrane, probed with a mouse monoclonal to gizzard smooth muscle MLCK, and detected by ECL. PVDF membranes were air-dried and exposed to PhosphorImager plates for detection of 32PO4 incorporation in MLCK. To account for differences in recovery of immunopurified MLCK, auto-radiographs and corresponding Western blot ECL films were each quantitated by two-dimensional laser densitometry and MLCK phosphorylation expressed as a ratio of 32PO4 densitometric units to Western blot densitometric units. All values were normalized to untreated controls.
To measure catalytic activity, MLCK210 was immunoprecipitated from unlabeled BPAE monolayers treated with 10 µM Fsk, 40 µM IBMX for 5 and 30 min and used in an in vitro phosphorylation assay. Protein G beads containing MLCK210 were resuspended in 50 µl of phosphorylation buffer (25 mM Tris-HCl, pH 7.5, 1 mM MgCl2, 150 mM KCl, 1 µM CaM, 5 µM CaCl2, 1 µM RLC) containing 125 µM [
-32P]ATP and incubated at 30 °C for 30 min. Assays were terminated by transferring 10 µl of reaction mixture to P81 papers that were immediately immersed in 10% trichloroacetic acid, 2% Na4P2O7 for 10 min. The papers were washed in 1% trichloroacetic acid, 2% Na4P2O7, and the phosphorylated product was quantitated by liquid scintillation counting. The amount of immunopurified kinase in each reaction mixture was quantitated by Western blot using an affinity-purified rabbit polyclonal MLCK155 antibody. Known amounts of recombinant MLCK155 were electrophoresed in 7.5% SDS gels, transferred to PVDF membrane, and probed with affinity-purified MLCK155 antibody to generate a standard curve, which was used to compute the amount of immunopurified kinase in each reaction mixture. MLCK activity is expressed as picomoles of 32PO4 incorporated per pmol of RLC/ng of MLCK210.
RhoA Phosphorylation and Activity Assay—To determine whether RhoA was phosphorylated in response to elevated cAMP, treated and control monolayers were flooded with 100 µl of 100 °C Laemmli sample buffer, scraped up, sonicated to shear DNA, heated for 10 min at 100 °C, and centrifuged at 22,000 x g for 10 min to remove cell debris. Equivalent samples were electrophoresed on 12% SDS gels, transferred to PVDF membranes, and probed with a 1:1000 dilution of affinity-purified rabbit polyclonal antibody to RhoA or phospho-Ser188-RhoA. Western blots of phospho-RhoA and total RhoA were detected by ECL methodology. X-ray films were quantitated by laser densitometry, and the phospho-RhoA densitometric units were divided by the RhoA densitometric units for each time point. For each experiment, all phospho-RhoA values were then normalized to the control value and expressed as a ratio of control.
BPAE cells were plated at a density of 1 x 106 cells per 100-mm dish and used 6 days post-confluence. The amount of activated GTP-bound RhoA was determined following the methods described by Ren et al. (46, 47). Samples were electrophoresed on 12% gels, transferred to PVDF membranes, and probed with a 1:1000 dilution of a rabbit polyclonal antibody to RhoA. Western blots of GTP-bound RhoA and total cell RhoA were visualized with ECL detection reagents. X-ray films were developed at various time intervals to obtain an exposure within the linear range of the film. Films were scanned in a Personal Densitometer (Amersham Biosciences) and data analyzed as outlined by Ren et al. (46, 47).
Protein Phosphatase Activity Assay—BPAE monolayers were either untreated or incubated with 10 µM Fsk, 40 µM IBMX for 1, 2.5, 5, and 15 min, and the media were aspirated and monolayers snap-frozen. Dishes were placed on a 4 °C copper block to thaw, flooded with 400 µl of homogenization buffer (50 mM Tris, 0.1 mM EDTA, 5 mM -mercaptoethanol containing protease inhibitors), scraped up, and homogenized in a glass/glass tissue homogenizer, and spun at 8,000 x g for 10 min. 50-µl aliquots of the cell extracts were snap-frozen in liquid N2 for storage at -70 °C. Cell pellets were dissolved in 25 µl of 0.1% SDS, and protein determinations were performed using a Pierce BCA protein assay after chloroform/methanol extraction as outlined by Wessel and Flugge (48). For determination of phosphatase activity, untreated (control) and experimental cell extracts were diluted 20-fold in phosphatase assay buffer (20 mM Tris, 0.2 mM EDTA, 0.5 mM -mercaptoethanol, 1 mg/ml bovine serum albumin, pH 7.0), and equivalent amounts of protein were added to a 50-µl reaction mixture containing 0.5 µM 32PO4-labeled nonmuscle myosin II RLC (41) and incubated at 30 °C for 15 min. Reactions were terminated by adding equal volumes of ice-cold 20% trichloroacetic acid. Proteins were precipitated on ice for 1 h, and the samples washed three times with 100% acetone, and the pellets were dissolved in SDS sample buffer. 32PO4-RLCs were analyzed on 12% SDS gels, stained with Coomassie Blue, dried, and exposed to Phosphor-Imager plates. Data are expressed as the ratio of phosphorylated RLC in Fsk/IBMX-treated samples/phosphorylated RLC in untreated cell extracts. Dephosphorylation of 32PO4-labeled RLC by untreated cell extracts is therefore 1.0, and the dephosphorylation of 32PO4-labeled RLC by Fsk/IBMX-treated cell extracts is used as a measure of phosphatase activity.
Phospho-LIM Kinase and Phosphocofilin Determinations—Treated and control monolayers were processed and quantitated as outlined above under RhoA phosphorylation and probed with a 1:1000 dilution of affinity-purified rabbit polyclonal antibody to phospho-LIM kinase, LIM kinase, Ser3-phosphocofilin, or cofilin. For each experiment, all phosphoprotein densitometric units were divided by total protein densitometric units for each time point and then normalized to the control value and expressed as a ratio of control.
Actin Filament Depolymerization Assay—Depolymerization of actin filaments by cell extracts from Fsk/IBMX-treated monolayers was quantitated using an F-actin disassembly assay (49). Actin was polymerized in MKEI buffer (20 mM imidazole-HCl, pH 7.0, 100 mM KCl, 2 mM MgCl2, 1 mM EGTA) for 30 min at room temperature. F-actin at a concentration of 5 µM (containing 10% pyrene-labeled actin) was diluted 25-fold into G buffer (2 mM Tris-HCl, pH 8.0, 0.2 mM ATP, 0.1 mM DTT, 0.2 mM CaCl2, 0.005% NaN3) containing a 1:20 dilution of cell extract at 25 °C. The decrease in pyrene fluorescence accompanying actin depolymerization was monitored in an Aminco Bowman luminescence spectrometer as outlined by Bryan (49). Phalloidin saturated pyrene-actin filaments were prepared by addition of 20 µM phalloidin.
BPAE cell lysates were prepared by lysing monolayers in 300 µl of lysis buffer (20 mM Tris-HCl, pH 7.5, 5.5 mM EGTA, 0.5 mM MgCl2, 0.5 mM ATP, 1 mM DTT containing protease inhibitors) at the desired time intervals after treatment with Fsk/IBMX. Extracts were spun at 22,000 x g for 10 min, and supernatants were snap-frozen in liquid N2. For depolymerization assays, extracts were diluted 1:20 in G buffer in a quartz cuvette prior to addition of pyrene-labeled F-actin.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
PKA activity was determined by a [32P]ATP-based assay using a synthetic Leu-Arg-Arg-Ala-Ser-Leu-Gly (Kemptide) as the substrate. BPAE cellular PKA activity increased 3-fold within 1 min after Fsk/IBMX treatment and remained elevated over 30 min. The PKA substrate, VASP, is phosphorylated (supplemental Fig. S1) in response to elevated cAMP, indicating that the increase in PKA activity is not a result of post-homogenization activation.
Isometric Tension and Myosin Light Chain Phosphorylation—The effect of cAMP on basal isometric tension was determined in BPAE monolayers grown on pre-cast collagen gels as described previously (40). A stable basal tension of 58-60 dynes (Fig. 1A) develops within 4-5 days after seeding, and monolayers are used within 24 h of reaching a constant basal tension. Fig. 1A is a representative isometric tension tracing of a BPAE monolayer treated with Fsk/IBMX. Isometric tension declines immediately and rapidly within the first 3-4 min, decreases steadily over the next 8 min, and establishes a new base-line tension of 41 dynes (a 33% reduction) within 10 min. Isometric tension reproducibly declined to 32-35% of the pretreatment base line and remained at these levels for the duration of the experiments.
The effect of Fsk/IBMX treatment on BPAE RLC phosphorylation was analyzed by glycerol/urea gel electrophoresis and Western blotting. Fig. 1B shows a representative blot from a time course of BPAE RLC phosphorylation after incubation with Fsk/IBMX. Untreated control monolayers have a basal level of RLC phosphorylation of 0.33 mol of PO4/mol of RLC. During the first 90 s of incubation with Fsk/IBMX, RLC phosphorylation decreases by 46% to 0.18 mol of PO4/mol of RLC (Fig. 1B, and TABLE ONE). Myosin RLC phosphorylation exhibits a similar reduction over the duration of the experiment decreasing 38% compared with control monolayers after 30 min. TABLE ONE shows the quantitative data from four time course experiments.
|
5 min but declined over the succeeding 20 min. Fig. 2B is a representative tension tracing showing the effect of thrombin on BPAE monolayers. Thrombin induced a rapid increase in isometric tension, which reached a maximum of 129 dynes in 3-4 min and was sustained at this level over 50 min.
|
|
MLCK210 Phosphorylation and Activity Measurements—We have shown recently (11) that cultured large and small vessel endothelial cells contain the 210-kDa nonmuscle MLCK but not the smaller smooth muscle form. BPAE cells used in this study express only MLCK210. To examine more directly the effects of cAMP on the state of phosphorylation and catalytic activity of MLCK210, two sets of experiments were performed. In the first set, cultures were labeled with [32P]orthophosphoric acid and incubated in Fsk/IBMX for 0, 5, 15, and 30 min. The cells were lysed, and MLCK210 was immunoprecipitated and quantitated by Western blot (supplemental Fig. S2A). The blot was subsequently exposed to a PhosphorImager screen to determine the incorporation of 32PO4 into MLCK210 (supplemental Fig. S2B). To account for differential recovery of MLCK210 during immunoprecipitation, phosphorimages and the corresponding Western blots were analyzed by laser densitometry, and data were expressed as a ratio of phosphorylated MLCK210/immunoblot densitometric units as outlined under "Materials and Methods." Immunoprecipitates from control monolayers show a basal level of MLCK210 phosphorylation (supplemental Fig. S2B, lane 1). After 5-, 15-, and 30-min incubations in Fsk/IBMX, no significant increase in 32PO4 incorporation into MLCK210 occurred (supplemental Fig. S2B, lanes 2-4).
To assess catalytic activity, MLCK210 was immunopurified from monolayers treated with Fsk/IBMX for 5 and 30 min, and enzymatic activity was determined as outlined under "Materials and Methods." Immunopurified MLCK210 from untreated monolayers incorporated 0.26 ± 0.02 pmol 32PO4/pmol of RLC/ng enzyme. The activity determinations obtained from MLCK210 immunoprecipitated from BPAE cells treated with Fsk/IBMX for 5 and 30 min were 0.27 ± 0.02 and 0.25 ± 0.02 pmol of 32PO4 incorporated per pmol of RLC/ng of enzyme, respectively. These data show that cAMP has no effect on MLCK210 catalytic activity.
Effects of the MLCK Inhibitor, KT5926, on Isometric Tension and Light Chain Phosphorylation—The data thus far established that cAMP cause a reduction in basal tension and myosin RLC phosphorylation in BPAE cells but did not support the involvement of MLCK210 in mediating these effects. In the next set of experiments, the effects of direct inhibition of MLCK210 on BPAE cell RLC phosphorylation and basal and agonist-stimulated isometric tension were determined. Fig. 3A shows a typical isometric tension tracing of BPAE monolayers incubated with 100 nM KT5926, a highly selective and potent inhibitor of MLCK (50). A 7-dyne decrease in tension occurred within the first 5 min. This modest decrease was maintained for an additional 10 min before tension slowly returned to base-line levels. Concentrations of up to 1 µM KT5926 produced the same minimal alteration in basal isometric tension (data not shown). Quantitative data of the effects of KT5926 on myosin RLC phosphorylation from four time course experiments are presented in TABLE THREE. Monolayers incubated with 100 nM KT5926 showed a rapid time-dependent decrease in the extent of RLC phosphorylation from 0.33 mol of PO4/mol of RLC to 0.03 mol of PO4/mol of RLC by 30 min, which represents a 90% decrease from control values.
|
|
RhoA Phosphorylation and RhoA Activation—The state of myosin II RLC phosphorylation is dependent on the rate of dephosphorylation by myosin light chain phosphatase, a protein phosphatase type 1 (PP1) phosphatase (32). Recent studies have shown that PKA phosphorylates RhoA at Ser188, and this phosphorylation results in decreased binding of RhoA to Rho kinase, an enzyme that has been shown to regulate the activity of myosin light chain phosphatase (38).
In order to determine whether cAMP affected RhoA phosphorylation in BPAE cells, monolayers were incubated in Fsk/IBMX for 0, 5, 15, and 30 min, and RhoA phosphorylation was analyzed as outlined under "Materials and Methods." Control monolayers exhibit basal levels of RhoA phosphorylation on Ser188 (Fig. 5A). Incubation of monolayers in Fsk/IBMX for 5 min resulted in a 4.6-fold increase in RhoA phosphorylation at Ser188. Maximal phosphorylation, 10-fold above control, occurred by 15 min and was maintained for the 30-min duration of the experiment. These results demonstrate that RhoA is phosphorylated at Ser188 in response to cAMP in endothelial cells.
We next sought to determine whether RhoA phosphorylation at Ser188 altered the state of RhoA activity in BPAE monolayers. To assess RhoA activation, an RBD assay was performed on BPAE monolayers by utilizing the RBD fragment from the RhoA effector Rhotekin as outlined by Ren et al. (46). Fig. 5B depicts representative Western blots as well as quantitative data from four time course experiments demonstrating the effects of cAMP on basal and thrombin-stimulated RhoA activity. BPAE monolayers were pretreated with Fsk/IBMX for 30 min, and Rho activity was assessed after addition of 2 units/ml thrombin. Unstimulated monolayers incubated in media containing vehicle only (0.01% Me2SO) for 30 min indicate that control monolayers exhibit basal RhoA activity (Fig. 5B, inset, lane 1), whereas monolayers pretreated with Fsk/IBMX show a 65% reduction in GTP-bound RhoA (lane 2) compared with controls (lane 1). Sixty seconds after adding thrombin to Fsk/IBMX-pretreated cultures, GTP-bound RhoA increased by 65% (Fig. 5B, inset, lane 3). In contrast, thrombin stimulation of BPAE monolayers preincubated in media alone for 60 s induced a 400% increase in GTP-bound RhoA. Taken together, these data indicate that increased levels of cAMP induce RhoA phosphorylation, inhibiting RhoA activity which prevents activation of the Rho/Rho kinase pathway in BPAE cells.
Numerous agonists have been shown to affect cellular functions through Rho and its downstream effectors. The Rho effector, Rho kinase, regulates myosin II and cell contraction by catalyzing phosphorylation of the regulatory subunit of myosin phosphatase by inhibiting its catalytic activity, which results in an indirect increase in RLC phosphorylation. Antagonism of the Rho/Rho kinase pathway by inhibition of either RhoA or Rho kinase uncouples a key regulatory cascade that maintains the balance between phosphorylation/dephosphorylation events. Our data have established that treatment with Fsk/IBMX induced phosphorylation of RhoA and inhibition of RhoA activation. Because RhoA activation has been linked to myosin phosphatase regulation, we sought to determine whether cAMP affected myosin light chain phosphatase activity. For quantitation of myosin phosphatase activity, 32PO4-labeled nonmuscle myosin II RLCs were used as substrate. Quantitative data from three time course experiments are presented in Fig. 6. Treatment with Fsk/IBMX for 60 s caused a rapid increase in phosphatase activity as shown by a 50% increase in release of 32PO4 from labeled RLC substrate (Fig. 6, inset, lane 2). An additional slight increase in phosphatase activity occurred over the ensuing 14 min (Fig. 6, inset, lanes 3-5).
|
To determine whether tautomycin also prevented the decrease in RLC phosphorylation elicited by cAMP, BPAE monolayers were pretreated with 10 nM tautomycin and myosin RLC phosphorylation analyzed after addition of Fsk/IBMX in the presence of inhibitor. Untreated monolayers have a base-line phosphorylation of 0.31 ± 0.02 mol of PO4/mol of RLC (Fig. 7B, lane 1). In monolayers treated with Fsk/IBMX alone, phosphorylation decreases to 0.17 ± 0.03 mol of PO4/mol of RLC (Fig. 7B, lane 2), whereas tautomycin alone results in a greater than 2-fold increase in phosphorylation to 0.71 ± 0.05 mol of PO4/mol of RLC (Fig. 7B, lane 3). Monolayers pretreated with 10 nM tautomycin for 15 min followed by addition of Fsk/IBMX for an additional 15 min in the presence of tautomycin showed levels of phosphorylation (0.76 ± 0.03 mol of PO4/mol of RLC; Fig. 7B, lane 4) comparable with tautomycin treatment alone. These results demonstrate that increased phosphatase activity is necessary for the decrease in isometric tension and RLC phosphorylation and that inhibition of myosin phosphatase blocks the effects induced by cAMP.
cAMP Effect on LIM Kinase and Cofilin—Our data indicate that a signaling cascade downstream of the Rho/Rho kinase is responsible for the decline in tension upon elevation of cAMP. Our next objective was to identify which downstream pathway(s) was activated by cAMP. One major pathway downstream from Rho/Rho kinase is the LIM kinase to ADF/cofilin pathway. ADF/cofilins (53-55) are a family of proteins that exert their effects through RhoGTPase-triggered actin reorganization. Rho/Rho kinase catalyzes phosphorylation and activation of LIM kinase, which in turn phosphorylates and inactivates cofilin. Active cofilin severs and depolymerizes F-actin, whereas inactive phosphocofilin most likely stabilizes F-actin by decreasing subunit turnover rates. Thus the Rho/Rho kinase signaling pathway affects actin dynamics both by inhibiting the severing activity of cofilin and by stabilizing F-actin. To determine whether cofilin activity was altered in response to cAMP, we assessed cofilin activity in treated and control monolayers.
Monolayers were incubated with Fsk/IBMX for the indicated time intervals, and changes in phospho-LIM kinase and phosphocofilin were analyzed and quantitated as described under "Materials and Methods." Untreated monolayers exhibit a constitutive level of LIM kinase phosphorylation (Fig. 8A, lane 1). A 46% decline in phosphorylation occurred after 1 min of incubation with Fsk/IBMX (Fig. 8A, lane 2), whereas by 5 min the level of phospho-LIM kinase was 30% of control (Fig. 8A, lane 3). Phospho-LIM kinase levels begin to recover by 15 min (Fig. 8A, lane 4), and within 30 min were similar to control. Concomitant with the decline in phospho-LIM kinase, levels of Ser3-phosphocofilin dropped by 40% within 1 min (Fig. 8B, lane 2; Fig. 8D) after stimulation with Fsk/IBMX and steadily declined to 20% of control values by 30 min (Fig. 8B, lane 5; Fig. 8D).
|
|
|
|
|
In an attempt to ascertain if cofilin was the severing protein responsible for cAMP-induced F-actin depolymerizing activity, we determined if extracts would disassemble F-actin stabilized with phalloidin (Fig. 9C). Phalloidin inhibits the severing activity of the cofilin family of proteins but does not prevent severing of gelsolin-like proteins (56, 57). As shown in Fig. 9B, recombinant human cofilin induces disassembly of pyrene-labeled F-actin but is unable to depolymerize phalloidin-stabilized filaments. Saturation of pyrene-labeled F-actin with phalloidin prevented actin depolymerization activity present in cell extracts from Fsk/IBMX-treated monolayers (Fig. 9C). Therefore, the presence of phalloidin-sensitive depolymerizing activity in cell extracts indicated that cofilin-mediated depolymerization is turned on in response to elevated cAMP.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Agents that elevate cAMP are known to relax vascular smooth muscle (6-8). The majority of studies in smooth muscle have focused on cAMP effects on lowering intracellular Ca2+ and/or its effect on altering MLCK catalytic activity (9). Investigators examining the role of cAMP in endothelial permeability found no alterations in Ca2+ homeostasis, suggesting that the mechanism of action of cAMP is downstream of Ca2+ entry (2). To identify the mechanism responsible for the cAMP action on endothelial cells, we monitored the effects of Fsk/IBMX treatment on basal and agonist-induced tension development and RLC phosphorylation. In agreement with previous work (13, 58), we found Fsk/IBMX treatment resulted in a rapid, sustained elevation of cAMP and maximal activation of PKA which correlates with a decline in tension and a decrease in RLC phosphorylation. Recent reports have documented a decline in RLC phosphorylation upon elevation of cAMP and attributed this decrease to PKA-catalyzed phosphorylation and inhibition of MLCK (1, 58-60). Identification of PKA consensus sequences in nonmuscle MLCK (19, 59) contributed additional support to this idea. Although this hypothesis, akin to that proposed for regulation of smooth muscle relaxation (6, 7, 10), is attractive, strong supportive in vivo evidence is lacking.
Our results demonstrate that unstimulated monolayers exhibit constitutive phosphorylation of MLCK210. Treatment with Fsk/IBMX causes no significant increase in MLCK210 phosphorylation nor a change in catalytic activity. In vitro, PKA phosphorylates MLCK and decreases its catalytic activity by reducing its affinity for Ca2+/CaM (7, 10). Initial studies examining an in vivo role for PKA-mediated phosphorylation and inhibition of MLCK in smooth muscle have yielded conflicting results (8, 10). However, more detailed examination of the Ca2+/CaM activation properties and MLCK catalytic activity in smooth muscle tissue has shown that cAMP/PKA activation has no effect on MLCK activation or catalytic activity (8, 61).
Although data from several studies in endothelial cells have led investigators to evoke the PKA phosphorylation theory for the regulation of MLCK activity to explain the decrease in RLC phosphorylation and enhanced barrier function in response to elevated cAMP, neither MLCK phosphorylation nor changes in MLCK catalytic activity were measured in these studies. A more recent study measured MLCK phosphorylation and activity and reported that cholera toxin (13) induced a 2.5-fold increase in MLCK phosphorylation and a 4-fold decrease in MLCK activity. It is difficult to compare this study with ours in view of the different agents employed to elevate cAMP and the use of a synthetic peptide antibody to MLCK (D119) (12, 15), which reacts more avidly with myosin IIA (11) than with MLCK. Furthermore, normalization of experimental samples was not undertaken to account for differential recovery of MLCK during immunoprecipitation reactions. It is conceivable that the difference reported for MLCK activity could be accounted for by variable recovery of immunoprecipitated kinase.
To probe more directly the role of MLCK210 in isometric tension development and constitutive RLC phosphorylation, the enzyme was directly inhibited with KT5926, a highly selective and potent inhibitor of MLCK that interacts with the enzyme at its catalytic site (50). KT5926, in contrast to Fsk/IBMX, induced a precipitous drop in RLC phosphorylation, yet caused only a small (8-10 dyne) decrease in basal tension. It inhibited thrombin-induced tension development and RLC phosphorylation. These results indicate that the cAMP-induced decline in tension occurs by a mechanism independent of MLCK210 inactivation and further demonstrate that agonist-induced tension is dependent on MLCK210-catalyzed activation of myosin II, consistent with results observed in agonist-stimulated smooth muscle preparations (62, 63). Thus the results obtained from MLCK210 phosphorylation experiments, MLCK210 activity assays, and inhibitor studies do not support the PKA phosphorylation hypothesis (6, 7, 10) for regulation of BPAE MLCK210. In addition, cAMP elicited a rapid drop in tension and RLC phosphorylation in a cell line that possesses no MLCK isoforms (45).3 Therefore, we conclude that PKA-mediated inhibition of MLCK cannot be the mechanism mediating the decline in BPAE isometric tension and RLC phosphorylation in response to elevated cAMP.
A member of the Ras superfamily, RhoA, has been implicated as the molecular switch involved in the regulation of nonmuscle and smooth muscle contraction (32, 33). RhoA regulates cellular tension through concerted activities of its RhoA-dependent effectors Dia (64, 65) and Rho kinase (31). Its major effector, Rho kinase, changes cellular tension either indirectly by inhibiting myosin phosphatase activity or directly through phosphorylation of myosin II RLC (32, 33). Studies in vitro and in vivo (35, 36, 38, 67) have shown that PKA phosphorylates RhoA on Ser188, and this phosphorylation has been proposed as a molecular mechanism by which cAMP directly inactivates RhoA or RhoA effectors. Fsk/IBMX-treated endothelial cell monolayers showed a 10-fold increase in RhoA phosphorylation within 5 min. Concomitant with the increase in RhoA phosphorylation, cAMP caused a 65% inhibition of RhoA activity in control cultures and 200% inhibition in thrombin-stimulated monolayers. These results are consistent with previous reports (36, 38) showing that agents that elevate cAMP or cell-permeable analogs of cAMP induce RhoA phosphorylation and inhibit RhoA activity. Essler et al. (68) failed to demonstrate RhoA phosphorylation in human umbilical vein endothelial cells in response to cholera toxin and concluded that cAMP interferes with Rho signaling by a PKA-independent mechanism, but in a more recent study, Qiao et al. (69) showed that cAMP inhibited thrombin-induced RhoA activation in microvascular endothelial cells. Studies focusing on the ability of PKA/cGMP-dependent protein kinase to induce smooth muscle relaxation have established PKA-catalyzed phosphorylation of RhoA as a negative regulator of the GTPase signaling cascade demonstrating a critical role for RhoA in governing cellular tension (34, 70).
Of the many Rho effector proteins that are activated by GTP-RhoA, it is Rho kinase that most likely plays a pivotal role in governing cellular tension because of its ability to affect both the state of RLC phosphorylation and actin filament stability. Active Rho kinase catalyzes phosphorylation of the myosin targeting subunit (MYPT1) of myosin light chain phosphatase causing inhibition of its activity; active Rho kinase also directly phosphorylates myosin RLC. Phosphorylation of either substrate results in a net increase in RLC phosphorylation, myosin II activation, and contraction. Actin filament stability is achieved through Rho kinase-catalyzed phosphorylation and activation of LIM kinase which in turn phosphorylates and inhibits the actin severing and depolymerizing activity of cofilin. Thus antagonism of the Rho kinase pathway has two major consequences: 1) up-regulation of myosin phosphatase activity, and 2) activation of the severing and depolymerizing activity of cofilins.
The experimental data presented in this report showed that Fsk/IBMX up-regulated myosin phosphatase activity by 76% above unstimulated cont
