|Home | About | Journals | Submit | Contact Us | Français|
Vasodilator-stimulated phosphoprotein (VASP) can catalyze actin polymerization by elongating actin filaments. The elongation mechanism involves VASP oligomerization and its binding to profilin, a G-actin chaperone. Actin polymerization is required for tension generation during the contraction of airway smooth muscle (ASM); however, the role of VASP in regulating actin dynamics in ASM is not known. We stimulated ASM cells and tissues with the contractile agonist acetylcholine (ACh) or the adenylyl cyclase activator, forskolin (FSK), a dilatory agent. ACh and FSK stimulated VASP Ser157 phosphorylation by different kinases. Inhibition of VASP Ser157 phosphorylation by expression of the mutant VASP S157A in ASM tissues suppressed VASP phosphorylation and membrane localization in response to ACh, and also inhibited contraction and actin polymerization. ACh but not FSK triggered the formation of VASP-VASP complexes as well as VASP-vinculin and VASP-profilin complexes at membrane sites. VASP-VASP complex formation and the interaction of VASP with vinculin and profilin were inhibited by expression of the inactive vinculin mutant, vinculin Y1065F, but VASP phosphorylation and membrane localization were unaffected. We conclude that VASP phosphorylation at Ser157 mediates its localization at the membrane, but that VASP Ser157 phosphorylation and membrane localization are not sufficient to activate its actin catalytic activity. The interaction of VASP with activated vinculin at membrane adhesion sites is a necessary prerequisite for VASP-mediated molecular processes necessary for actin polymerization. Our results show that VASP is a critical regulator of actin dynamics and tension generation during the contractile activation of ASM.
Dynamic remodelling of the actin cytoskeleton is recognized as an important step in the agonist-induced activation of contraction and tension development in airway smooth muscle (ASM)2 and in other smooth muscle tissues (1). The regulation of actin filament remodelling is a complex process involving the coordinated activity of multiple proteins; but the molecular mechanisms for actin remodelling and the processes by which actin regulatory proteins are activated in response to contractile and dilatory agents in smooth muscle tissues are poorly understood. In ASM, neuronal Wiskott-Aldrich syndrome protein (N-WASp) plays a critical role in the initiation of actin polymerization during contractile stimulation by activating the Arp2/3 complex (2). The Arp2/3 complex is believed to nucleate the formation of new actin filaments that branch off the sides of existing filaments (3). However, the importance of other catalysts for actin polymerization in regulating the contraction of ASM has not been determined.
Members of the Ena/VASP family regulate cell motility and shape in a wide variety of cell types (4, 5). Although a number of molecular functions that affect actin dynamics have been attributed to Ena/VASP proteins, there is extensive evidence for their role as actin filament elongation factors, i.e. they bind to the barbed (fast growing) ends of existing actin filaments and promote filament lengthening (4, 6, 7). The mechanism for the elongation of actin filaments by VASP is proposed to require the assembly of VASP into tetrameric oligomers and the membrane recruitment and anchoring of VASP to the scaffolding proteins vinculin and zyxin at sites of actin filament assembly. Filament elongation can then occur via the recruitment of profilin-G actin complexes to bind to VASP tetramers, followed by the transfer and assembly of G-actin monomers into the barbed ends of the actin filaments that are also bound to VASP (8,–10). We evaluated ASM for evidence of a VASP-mediated process of actin elongation during contractile and dilatory stimulation.
Ena/VASP proteins consist of 3 domains, N- and C-terminal Ena/VASP homology 1 and 2 (EVH) domains and a central proline-rich region (4, 7). The EVH1 domain contains binding sites for several focal adhesion scaffolding proteins including vinculin; the proline-rich region contains binding sites for profilin-actin, a primary source of actin monomers for actin filament polymerization; and the EVH2 domain contains binding sites for filamentous (F)- and globular (G)-actin. The C-terminal coiled-coil region within the EVH2 domain of VASP mediates the assembly of VASP monomers into stable tetramers, believed to be an essential step for VASP to function as an elongation factor (4, 8, 11,–15). Ena/VASP proteins are also known substrates for both serine/threonine and tyrosine kinases (16,–18). The phosphorylation of VASP Ser157 has been implicated in the cellular localization of VASP (17, 19). VASP plays a role in the regulation of actin polymerization and contraction in aortic smooth muscle (20). VASP is expressed in ASM tissues and undergoes phosphorylation at Ser157 during β adrenergic stimulation (21, 22); but the function of VASP during the contraction and relaxation of ASM is unknown.
Signaling events that regulate actin polymerization during contractile stimulation of ASM are mediated by adhesome complexes at integrin-ECM adhesion junctions (23). Vinculin, a VASP ligand, plays an important structural role in these junctions by binding to the integrin-binding proteins talin and α-actinin as well as to actin filaments (24). Vinculin can assume a closed conformation in which it does not bind to actin or talin, and an open conformation in which its actin and talin binding sites are exposed (25, 26). The contractile stimulation of ASM tissues with ACh induces the recruitment of vinculin to membrane adhesion complexes and its activation to an open ligand-binding conformation (27, 28). Vinculin phosphorylation on Tyr1065 is necessary for vinculin to sustain an activated conformation in which it can bind to talin and actin filaments (27). VASP has been shown to bind to the proline-rich hinge region of vinculin at cell junctions (29,–31); thus we hypothesized that vinculin might play a role in the regulation of VASP-mediated actin dynamics in ASM. To test this hypothesis, we evaluated the molecular mechanisms by which contractile and dilatory stimuli regulate the activity of VASP and its interaction with vinculin in ASM.
Our results suggest that VASP functions as an actin elongation factor at the ASM plasma membrane, and that the interaction of VASP with activated vinculin is prerequisite to this function. We conclude that VASP is an important catalyst for actin polymerization during the contraction of ASM tissues.
Mongrel dogs were euthanized in accordance with the guidelines of the Institutional Animal Care and Use Committee (IACUC), Indiana University School of Medicine. The trachea was immediately removed and immersed in physiological saline solution at 22 °C. The solution was aerated with 95% O2, 5% CO2 to maintain a pH of 7.4. Rectangular strips of tracheal muscle 1 mm in diameter and 10 mm in length were dissected after removal of the epithelium and connective tissue layer. Each muscle strip was placed in physiological saline solution at 37 °C in a 25-ml organ bath and attached to a Grass force transducer for the measurement of force. At the beginning of each experiment, muscle length was progressively increased until the force of active contraction in response to a contractile stimulus reached a maximum (optimal length).
Antibodies used in these experiments were: mouse monoclonal human VASP (BD Biosciences Pharmingen), rabbit polyclonal human VASP (Cell Signaling), mouse monoclonal human VASP phospho-Ser157 (Abcam), mouse monoclonal human VASP phospho-Ser239 (Millipore); rabbit polyclonal human VASP phospho-Ser239 (Cell Signaling); rabbit polyclonal GFP (MBL); rabbit His tag (Cell Signaling), rabbit polyclonal bovine myosin light chain, custom made by BABCO (Richmond, CA); polyclonal vinculin (against canine cardiac vinculin); mouse monoclonal human profilin1 (Abcam); rabbit polyclonal human profilin (Cytoskeleton); mouse monoclonal human actin (Sigma); Alexa Fluor 488 and 546 (Invitrogen); and IRDye 800CW goat anti-mouse IgG and IRDye 680LT goat anti-rabbit IgG (Li-Cor).
Reagents included the Duolink in situ proximity ligation kit (PLA) and anti-mouse plus and anti-rabbit minus probes (Olink Bioscience, Uppsala, Sweden), lysis and F-actin stabilization buffers (Cytoskeleton), and protein kinase C inhibitor bisindolylmaleimide I (BIM) (Sigma). Plasmid vectors used included pcDNA3 EGFP full-length human VASP and VASP S157A (alanine substituted at Ser157, His6 tagged) (17, 18), pEGFP-vinculin full-length chicken vinculin (residues 1–1066) (32), and non-phosphorylatable vinculin Y1065F (27).
Plasmids encoding full-length EGFP VASP and EGFP VASP S157A were introduced into the smooth muscle strips by the method of reversible permeabilization as previously described (2, 33,–36). Muscle tissues were attached to a metal hooks and incubated successively in each of the following solutions: Solution 1, which contained 10 mm EGTA, 5 mm Na2ATP, 120 mm KCl, 2 mm MgCl2, and 20 mm TES (at 4 °C, pH 7.1, 100% O2 for 120 min); solution 2, which contained 0.1 mm EGTA, 5 mm Na2ATP, 120 mm KCl, 2 mm MgCl2, and 20 mm TES and 10 μg/ml of plasmids (at 4 °C, pH 7.1, overnight); solution 3, which contained 0.1 mm EGTA, 5 mm Na2ATP, 120 mm KCl, 10 mm MgCl2, and 20 mm TES (at 4 °C, pH 7.1, 100% O2 for 30 min); and solution 4, which contained 110 mm NaCl, 3.4 mm KCl, 0.8 mm MgSO4, 25.8 mm NaHCO3, 1.2 mm KH2PO4 (at 22 °C, pH 7.4, for 60 min and aerated with 95% O2 and 5% CO2). After 30 min in solution 4, CaCl2 was added gradually to reach a final concentration of 2.4 mm. The plasmid-treated tissue strips were then incubated at 37 °C for 2 days in DMEM containing 5 mm Na2ATP, 100 units/ml of penicillin, 100 μg/ml of streptomycin, and 10 μg/ml of plasmids to allow for expression of the recombinant proteins. Sham-treated tissues were subjected to identical procedures except that no plasmids were included in Solution 2. Tissues were then quickly frozen using liquid N2-cooled tongs for biochemical analysis or dissociated for cellular imaging studies.
Frozen muscle tissues were pulverized, and the proteins were extracted for electrophoresis or immunoprecipitation as previously described (2, 28). For immunoprecipitation, the extracts were precleared at 4 °C with protein A/G UltraLink Resin and incubated with antibodies against the target protein. Western blotting of immunoprecipitates or muscle extracts was performed to quantitate proteins. The proteins were visualized by ECL and digitally quantified using a Bio-Rad ChemiDoc XRS detection system, VASP phosphorylation at Ser157 was quantitated by probing membranes simultaneously with mouse anti-phospho-Ser157 VASP Ab (Abcam) and polyclonal rabbit anti-VASP Ab (Cell Signaling), and then using fluorescent probes to detect both antibodies simultaneously using a Li-Cor Odyssey infrared imaging system.
Muscle strips were rapidly frozen and then immersed in acetone containing 10% (w/v) trichloroacetic acid and 10 mm DTT that was precooled with dry ice. Strips were thawed in acetone/trichloroacetic acid/DTT at room temperature and then washed 4 times with acetone/DTT. Proteins were extracted for 2–3 h in 8 m urea, 20 mm Tris base, 22 mm glycine, and 10 mm DTT. Myosin light chains (MLCs) were separated by glycerol-urea polyacrylamide gel electrophoresis, transferred to nitrocellulose, and incubated with polyclonal rabbit MLC 20 antibody (2, 37, 38). Unphosphorylated and phosphorylated bands of MLCs were visualized by ECL and quantified by densitometry. MLC phosphorylation was calculated as the ratio of phosphorylated MLCs to total MLCs for each sample.
Smooth muscle cells were enzymatically dissociated from tracheal muscle strips, plated onto glass slides, and allowed to adhere for 60 min as previously described (2, 35). After cells were stimulated with ACh, FSK, or left unstimulated, cells were then fixed and visualized for EGFP fluorescence or stained for immunofluorescence analysis. The effects of stimulation with ACh or FSK on the localization of proteins were evaluated in freshly dissociated smooth muscle cells using a Zeiss LSM 510 confocal microscope. Images of smooth muscle cells were analyzed for regional differences in fluorescence intensity by quantifying the pixel intensity with a series of cross-sectional line scans along the entire length of each cell, excluding the nucleus (2, 35). The ratio of pixel intensity between the cell periphery and the cell interior was computed for each line scan by calculating the ratio of the average maximum pixel intensity at the cell periphery to the average minimum pixel intensity in the cell interior. The ratios of pixel intensities between the cell periphery and the cell interior for all line scans performed on a given cell were averaged to obtain a single value for each cell.
In situ proximity ligation assays were performed to detect interactions between VASP and vinculin, VASP and profilin, VASP-VASP complexes, and phospho-Ser157 VASP in dissociated cells. PLA provides a method for the precise detection of protein-protein complexes or interactions and protein modifications. Two primary antibodies against the target proteins or protein epitopes are raised in different species, and a pair of oligonucleotide-labeled secondary antibodies (+ and − PLA probes) are targeted to each pair of primary antibodies. The probes form circular DNA strands only when they are bound in very close proximity (<40 nm). These DNA circles serve as templates for localized rolling circle amplification, generating a fluorescent signal (spot) that enables individual interacting pairs of the target protein molecules to be visualized. The PLA signal thus allows for the detection of a complex between two target proteins at a very high resolution (39, 40).
Smooth muscle cells were fixed, permeabilized, and incubated with a pair of primary antibodies of different species against two target proteins followed by a pair of oligonucleotide-labeled secondary antibodies (Duolink + and − PLA probes). Mouse anti-VASP and rabbit anti-vinculin antibodies were used to probe VASP-vinculin interactions, rabbit anti-VASP and mouse anti-profilin antibodies were used to probe VASP-profilin interactions, rabbit anti-VASP and mouse anti-VASP phospho-Ser157 antibodies were used to detect VASP phospho-Ser157, and mouse and rabbit phospho-Ser239 VASP antibodies were used to detect VASP-VASP interactions. PLA probe hybridization, ligation, amplification, and detection media were administered according to the manufacturer's instructions (Olink Bioscience). Cells from unstimulated and ACh- or FSK-stimulated groups were analyzed for interactions by counting PLA fluorescent spots using a Zeiss LSM510 confocal microscope. Duolink ImageTool software was used to quantitate PLA signals.
The concentration of F-actin and G-actin in smooth muscle tissues was measured using an assay kit from Cytoskeleton Inc. Each of the smooth muscle strips was homogenized in F-actin stabilization buffer. The supernatants of protein extracts (G-actin fraction) were collected after high speed centrifugation at 150,000 × g for 60 min at 37 °C. The pellets were resuspended in ice-cold distilled H2O plus 10 μm cytochalasin D and then incubated on ice for 1 h to dissociate F-actin. The supernatant of the resuspended pellets was collected after centrifugation at 4 °C. Equal volumes of the first supernatant (G-actin) or second supernatant (F-actin) were subjected to analysis by immunoblot using anti-actin antibody. The amount of F-actin and G-actin was determined by densitometry and the ratio of F-actin to G-actin was used for the analysis.
Data are expressed as mean ± S.E. Differences between treatment groups were determined using paired or unpaired two-tailed Student's t test or analysis of variance. Differences were considered statistically significant when p < 0.05.
VASP phosphorylation at Ser157 was measured in ASM tissues stimulated with 10−4 or 10−5 m ACh for time periods up to 30 min (Fig. 1A). The shift in apparent molecular mass of VASP from 46 to 50 kDa by SDS-PAGE was used to analyze stoichiometric changes in the phosphorylation of VASP at Ser157. We confirmed that the 50-kDa band was serine 157-phosphorylated VASP by simultaneously probing each membrane using a site-specific VASP phospho-Ser157 antibody (Fig. 1, A and B) (41). Tissues were also stimulated with FSK for up to 20 min to evaluate VASP phosphorylation at Ser157 (Fig. 1, A and B). VASP Ser157 phosphorylation increased significantly with both ACh and FSK stimulation, approaching a maximum by 10 min (Fig. 1C). FSK induced significantly higher levels of VASP Ser157 phosphorylation than ACh. Stimulation of muscles with the adrenergic hormone epinephrine affected VASP Ser157 phosphorylation similarly to FSK, but with a somewhat more rapid time course (Fig. 1, A and B).
The PKC inhibitor BIM (42) was used to determine whether PKC is involved in the regulation of VASP Ser157 phosphorylation in response to ACh or FSK. Tracheal smooth muscle tissues were treated with 10 μm BIM and then stimulated with ACh or FSK (Fig. 1, D and E). BIM treatment significantly suppressed ACh-induced VASP Ser157 phosphorylation but it did not affect VASP phosphorylation stimulated by FSK, indicating that PKC mediates VASP Ser157 phosphorylation in response to stimulation with ACh but not FSK.
The cellular localization of phospho-Ser157 VASP was analyzed by PLA using probes against antibodies for phospho-Ser157 VASP and VASP. Close proximity (<40 nm) of the target epitopes on VASP results in the generation of a fluorescent spot indicating VASP phosphorylation. Phospho-Ser157 VASP was observed on the membrane of unstimulated and ACh- or FSK-stimulated cells; however, many more spots were observed in cells stimulated with ACh or FSK than in unstimulated cells (Fig. 1F).
Immunofluorescence was used to evaluate the localization of phospho-Ser157 VASP and total VASP in freshly dissociated cells (Fig. 1G). In both unstimulated cells and cells stimulated with ACh or FSK, phospho-Ser157 VASP was only detected at the cell membrane, whereas VASP was detected in both the cytoplasm and at the cell membrane. The intensity of phospho-Ser157 VASP fluorescence was much higher in ACh- or FSK-stimulated cells than in unstimulated cells. Treatment with BIM markedly reduced VASP phospho-Ser157 fluorescence in ACh-stimulated cells and inhibited VASP localization to the membrane (Fig. 1G).
His-EGFP VASP S157A or His-EGFP VASP WT was expressed in ASM tissues and force in response to 10−5 m ACh measured after expression (Fig. 2, A–C). VASP S157A does not undergo phosphorylation at Ser157 due to the substitution of alanine for serine (Fig. 2C). The expression of VASP S157A in ASM tissues significantly depressed endogenous VASP Ser157 phosphorylation and inhibited contractile force in response to 10 min stimulation with ACh (Fig. 2, A–E). Neither sham treatment nor expression of VASP WT affected force or VASP Ser157 phosphorylation in response to ACh.
The proportions of F-actin to G-actin were analyzed in unstimulated and ACh-stimulated muscle tissues expressing VASP S157A, VASP WT, or sham-treated tissues (Fig. 2F). Expression of VASP S157A prevented the increase in the F- to G-actin ratio in response to stimulation with ACh; whereas expression of VASP WT or sham treatment had no significant effect on actin polymerization in response to ACh. ACh stimulation increased actin polymerization in tracheal muscle tissues, whereas stimulation with FSK did not (Fig. 2G).
MLC phosphorylation was analyzed in unstimulated and ACh-stimulated muscle tissues expressing VASP S157A, VASP WT, or sham-treated tissues (Fig. 2, H and I). MLC phosphorylation in response to ACh stimulation increased significantly regardless of treatment and was not significantly different among tissues expressing VASP WT, VASP S157A, or sham-treated tissues.
Co-immunoprecipitation analysis was used to assess the interaction of VASP with vinculin in tissues stimulated with ACh or FSK (Fig. 3A). Stimulation with ACh significantly increased the amount of vinculin that co-precipitated with VASP compared with unstimulated tissues (Fig. 3B). In contrast, stimulation with FSK did not significantly increase the co-precipitation of vinculin with VASP.
The effects of ACh and FSK on the interaction of VASP and vinculin were also evaluated using PLA in cells freshly dissociated from ASM tissues (Fig. 3, C and D). Significantly more PLA spots indicating VASP-vinculin complexes were observed in ACh-stimulated cells than in FSK-stimulated or unstimulated cells.
The colocalization of VASP and vinculin was evaluated by double immunofluorescence staining. Smooth muscle cells were freshly dissociated from tracheal smooth muscle tissues and stimulated for 10 min with ACh or FSK or left unstimulated. In unstimulated cells, both VASP and vinculin were distributed throughout the cell cytoplasm and at the membrane. In cells stimulated with ACh, VASP and vinculin were co-localized at the cell membrane, and little fluorescence was observed in the cytoplasm for either protein. In cells stimulated with FSK, VASP was localized almost entirely at the membrane, whereas vinculin was distributed throughout the cytoplasm (Fig. 3E). The effect of FSK and ACh on the distribution of VASP and vinculin to the membrane versus the cytoplasm was analyzed in 65 cells from 3 separate experiments (Fig. 3F). ACh stimulation significantly increased the membrane localization of both VASP and vinculin. In contrast, FSK stimulated the membrane localization of VASP but not vinculin.
The role of VASP Ser157 phosphorylation in the cellular localization of VASP was evaluated by analyzing the localization of EGFP VASP S157A and EGFP VASP WT in dissociated cells stimulated with either ACh or FSK (Fig. 4). Stimulation with either ACh or FSK caused the recruitment of EGFP VASP WT to the membrane, but neither ACh nor FSK stimulated the recruitment of EGFP VASP S157A to the membrane (Fig. 4, A and B). The amount of EGFP VASP WT at the membrane was significantly higher in cells stimulated with ACh or FSK than in unstimulated cells (Fig. 4B).
PLA was used to evaluate the effect of VASP Ser157 phosphorylation on the interaction of vinculin and VASP (Fig. 4, C and D). In unstimulated freshly dissociated smooth muscle cells expressing VASP WT or VASP S157A, very few PLA spots were observed in the cell, indicating few protein complexes containing both VASP and vinculin. In cells expressing VASP WT, stimulation with ACh caused a significant increase in the number of PLA spots at the cell membrane, indicating that ACh stimulated the formation of VASP-vinculin protein complexes. The expression of VASP S157A significantly inhibited the formation of VASP-vinculin protein complexes in response to ACh.
Co-immunoprecipitation analysis was also used to evaluate the interaction of vinculin and VASP in tissues expressing VASP WT or VASP S157A. The amount of vinculin that co-precipitated with VASP increased significantly in response to ACh stimulation in sham-treated tissues and tissues expressing VASP WT, but not in tissues expressing VASP S157A (Fig. 4, E and F). Thus, the expression of VASP S157A significantly inhibited the formation of VASP-vinculin protein complexes in response to ACh in both tissues and isolated dissociated cells, demonstrating that the Ser157 phosphorylation of VASP and its membrane localization are necessary for the interaction of VASP with vinculin.
We previously demonstrated that vinculin phosphorylation at Tyr1065 is required for vinculin to maintain an open activated conformation (27). Smooth muscle tissues were stimulated with FSK for 10 min or ACh for 5 min, then vinculin Tyr1065 phosphorylation and VASP phosphorylation were analyzed by immunoblot (Fig. 5). Stimulation with ACh, but not with FSK, induced vinculin Tyr1065 phosphorylation, indicating that vinculin undergoes activation in response to ACh but not FSK (Fig. 5, A and B). In contrast VASP undergoes phosphorylation at Ser157 in response to both ACh and FSK (Figs. 5A and and1,1, A–C).
Wild type vinculin or the phosphorylation deficient vinculin mutant Y1065F were expressed in ASM tissues to determine whether vinculin activation is required for the formation of VASP-vinculin complexes (Fig. 5, C–G). The interaction between VASP and vinculin was evaluated in tissue extracts by immunoprecipitation (Fig. 5, C and D). The amount of endogenous vinculin that co-precipitated with VASP increased significantly in response to ACh stimulation in sham-treated muscles and tissues expressing WT vinculin, but not in tissues expressing vinculin Y1065F (Fig. 5, C and D). The expression of vinculin Y1065F did not affect VASP Ser157 phosphorylation in response to ACh. Immunofluorescence analysis confirmed that the expression of EGFP vinculin Y1065F also had no effect on the localization of VASP phospho-Ser157 to the membrane (Fig. 5E).
PLA assays were used to evaluate the interaction of vinculin and VASP in freshly dissociated smooth muscle cells from tissues expressing vinculin Y1065F or WT vinculin (Fig. 5, F and G). Very few spots were detected in unstimulated cells. In cells expressing WT vinculin, stimulation with ACh resulted in a dramatic increase in the number of PLA spots at the cell membrane, indicating that ACh stimulates formation of protein complexes containing both VASP and vinculin. In the cells expressing vinculin Y1065F, there were markedly fewer spots indicating VASP-vinculin complexes, demonstrating that vinculin activation is required for the interaction of VASP with vinculin in membrane complexes (Fig. 5, F and G).
VASP can form tetrametric homo-oligomers through VASP-VASP binding interactions at its C terminus (11, 12). Although we could not probe specifically for VASP tetrameric homo-oligomers within the smooth muscle cells and tissues, we used PLA to assess for protein complexes containing multiple VASP molecules (VASP-VASP complexes) (Fig. 6). The effect of stimulation with ACh and FSK on VASP-VASP complex formation was investigated by using VASP phospho-Ser239 as a target for both +PLA and −PLA probes. VASP phospho-Ser239 was chosen as a target epitope for the PLA probes because stimulation with ACh does not cause a significant increase in phosphorylation at this site for the first 10 min (Fig. 6, A and B). Thus, an increase in the number of PLA spots detected after 5–10 min stimulation with ACh should reflect more VASP-VASP interactions rather than an increase in VASP Ser239 phosphorylation.
In cells stimulated for 10 min with ACh, VASP-VASP PLA spots were clearly observed on the membrane, but spots were barely evident in unstimulated cells (Fig. 6, C and D). In contrast, when PLA for VASP-VASP interactions was performed on cells stimulated with FSK for 10 min, very few PLA spots were observed (Fig. 6, C and D), even though FSK causes VASP Ser239 phosphorylation to increase by more than 5-fold within 10 min (Fig. 6, A and B). Thus, stimulation with ACh induced the formation of VASP-VASP complexes by membrane-localized VASP proteins, whereas stimulation with FSK did not. The fact that FSK induces a much higher level of VASP Ser239 phosphorylation than ACh confirms that an increase in the level of VASP Ser239 phosphorylation by itself does not result in an increase in the number of PLA spots.
The effect of VASP Ser157 phosphorylation on VASP-VASP complex formation was evaluated by PLA in cells expressing VASP S157A and VASP WT (Fig. 6, E and F). ACh-stimulated cells expressing VASP S157A had significantly fewer VASP-VASP PLA spots than cells expressing VASP WT, indicating that the expression of VASP S157A inhibits VASP-VASP complex formation in response to ACh. Thus, without Ser157 phosphorylation, VASP does not form VASP-VASP complexes, probably because it does not localize to the membrane.
The effect of vinculin activation on VASP-VASP complex formation after stimulation with ACh was also evaluated by PLA using cells expressing either WT vinculin or vinculin Y1065F. There were significantly more PLA spots on the cell membrane cells expressing WT vinculin than in cells expressing vinculin Y1065F (Fig. 6, G and H). The results demonstrate that vinculin activation is required for the formation of VASP-VASP complexes at the cell membrane in response to stimulation with ACh. Although FSK stimulates VASP Ser157 phosphorylation, it does not stimulate vinculin activation, and thus FSK does not induce VASP-VASP complex formation.
Profilin binds to VASP in its central proline-rich domain and is necessary for the transfer of G-actin monomers to the barbed ends of actin filaments by VASP (9). The effect of ACh and FSK on complex formation between VASP and profilin was evaluated using co-immunoprecipitation from tissue extracts and PLA in dissociated cells (Fig. 7). Stimulation with ACh but not FSK significantly increased the amount of profilin that co-immunoprecipitated with VASP compared with unstimulated tissues (Fig. 7, A and B). Significantly more PLA spots indicating VASP-profilin interactions were detected in ACh-stimulated cells than in FSK-stimulated or unstimulated cells (Fig. 7, C and D). These results indicate that ACh but not FSK induces the binding of profilin to VASP.
WT vinculin and the phosphorylation deficient mutant vinculin Y1065F were expressed in ASM tissues to determine whether vinculin activation is required for the binding of profilin to VASP. The formation of complexes between VASP and profilin was evaluated by co-immunoprecipitation and PLA (Fig. 8). The amount of profilin that co-immunoprecipitated with VASP increased significantly in response to ACh stimulation in sham-treated muscles and tissues expressing WT vinculin, but not in tissues expressing vinculin Y1065F (Fig. 8, A and B). In cells expressing WT vinculin, stimulation with ACh resulted in a significant increase in the number of PLA spots along the cell membrane, indicating more interactions between VASP and profilin. In contrast, in the cells expressing vinculin Y1065F, there were significantly fewer spots at the membrane indicating VASP-profilin interactions. The results demonstrate that vinculin activation is prerequisite to the interaction of profilin with VASP at the membrane.
Our studies demonstrate that VASP is a critical regulator of actin dynamics and tension generation during the contractile stimulation of ASM (Fig. 2). Our results also indicate that the interaction of VASP with activated vinculin at membrane adhesion sites is a necessary prerequisite for VASP-mediated molecular processes that are required for actin polymerization (Figs. 5, ,6,6, and and8).8). In ASM tissues, contractile stimulation triggered the formation of VASP-VASP complexes at the membrane and also stimulated the interaction of profilin with VASP (Figs. 66–8). VASP oligomerization and its binding to profilin have been shown to be necessary steps in the VASP-mediated actin filament elongation process (7, 9); thus, our data are consistent with a function for VASP as an actin elongation factor during contractile stimulation.
We found that the stimulation of ASM with either ACh or FSK induces the phosphorylation of VASP on Ser157 (Fig. 1). The ACh-stimulated VASP Ser157 phosphorylation was mediated by PKC. We evaluated the role of VASP Ser157 phosphorylation on VASP activity by expressing the point mutant VASP S157A in ASM tissues. VASP S157A expression inhibited endogenous VASP phosphorylation on Ser157 and prevented the localization of VASP to the membrane in response to stimulation with either ACh or FSK (Figs. 2 and and4).4). When VASP localization to the membrane was prevented, actin polymerization and tension generation in response to contractile stimulation (ACh) were inhibited, but MLC phosphorylation was unaffected (Figs. 2 and and4).4). We have previously found that in ASM, the inhibition of actin polymerization does not affect the increase in MLC phosphorylation in response to ACh, but that tension generation is markedly reduced (2, 37). This suggests that MLC phosphorylation and actin polymerization are independently regulated and that both processes are necessary for tension development (1).
Although FSK induced VASP Ser157 phosphorylation and membrane localization, FSK did not stimulate actin polymerization (Fig. 2), the formation of VASP-VASP complexes (Fig. 6), or the interaction of profilin with VASP (Fig. 7). Thus, our results suggest that VASP Ser157 phosphorylation and membrane localization of VASP are necessary but not sufficient steps to initiate the activity of VASP in catalyzing actin polymerization.
We next evaluated the role of vinculin activation on the function of VASP in ASM tissues. We have previously shown that ACh stimulation induces the recruitment of vinculin to membrane adhesion complexes and vinculin activation (27, 28). In the present study, we found that FSK does not stimulate the recruitment of vinculin to the membrane (Fig. 3). Thus even though FSK stimulates the phosphorylation of VASP at Ser157 and the recruitment of VASP to the membrane, the phosphorylated VASP does not interact with vinculin.
We expressed the vinculin point mutant Y1065F in ASM tissues to determine the role of vinculin activation on the regulation of VASP activity. Using FRET technology, we previously demonstrated that the phosphorylation of vinculin on Tyr1065 is necessary for vinculin to sustain its activated ligand-binding conformation, and that the mutant vinculin Y1065F localizes to the membrane in a closed inactive conformation and inhibits the activation of endogenous vinculin (Fig. 5E) (27). In the present study, when vinculin activation was inhibited by the expression of vinculin Y1065F, contractile stimulation induced the phosphorylation of VASP on Ser157; but VASP did not interact with vinculin (Fig. 5). Our observations suggest that even when both vinculin and VASP are localized at the membrane, VASP can only bind to vinculin when vinculin is in its open activated conformation. The inhibition of vinculin activation by the expression of vinculin Y1065F also prevented the formation of VASP-VASP complexes and the interaction of VASP with profilin (Figs. 6 and and8).8). These results suggest that vinculin activation and the binding of VASP to vinculin are necessary for VASP-mediated actin polymerization. The activation of vinculin may enable it to spatially and temporally coordinate the activity of VASP with other actin filament assembly promoting proteins.
The phosphorylation of VASP by cyclic nucleotide-dependent PKA and PKG protein kinases is well documented in multiple cell types including platelets, endothelial cells, and vascular and ASM cells (17, 19,–21). PKC-dependent VASP Ser157 phosphorylation has been reported in vascular smooth muscle cells activated by serum stimulation (19, 43). Our findings are consistent with observations in migrating vascular endothelial cells that VASP phosphorylation on Ser157 provides a signal for VASP localization to the membrane or the leading edge of the cell (17, 19); however, these studies also reported that VASP Ser157 phosphorylation had a minor impact on actin polymerization. In contrast, we find that stimulus induced actin polymerization in ASM tissues requires the Ser157 phosphorylation of VASP (Fig. 2). However, our results suggest that this does not result from a direct effect of VASP phosphorylation on its actin catalytic activity; rather it results from the requirement that VASP localize to the membrane and bind to activated vinculin to initiate its actin catalytic function. We note that VASP Ser157 phosphorylation does not provide a reliable indicator of the VASP activation with respect to actin dynamics.
VASP has been shown to bind to the proline-rich motif in the hinge region of vinculin between its head and tail domains; whereas the binding sites for vinculin on VASP are located within its EVH1 domain (5, 29, 31, 44). Vinculin binding has also been proposed to mediate the recruitment of VASP to focal adhesion sites (29, 31). However, our studies demonstrate that in ASM, VASP recruitment to the membrane occurs in the absence of vinculin recruitment, and that the recruitment of VASP does not depend on vinculin activation or the interaction of VASP and vinculin. Our results are thus consistent with VASP Ser157 phosphorylation as a primary event in the regulation of VASP localization to adhesion junction complexes.
The assembly of VASP into tetrameric homo-oligomers is critical for the VASP-mediated process of actin filament elongation in vitro (4, 8, 13,–15). Although VASP molecules self-assemble into stable tetramers in vitro; there is no evidence regarding the oligomerization state of VASP in living cells. We used a PLA to probe for multimeric VASP-VASP complexes in dissociated smooth muscle cells. Although PLA reports on complex formation between individual VASP proteins, it is not possible to confirm that the VASP protein complexes detected by PLA are tetramers. Other multimeric forms of VASP could be detected by PLA, as well as protein complexes that contain multiple VASP proteins or multimers bound in close proximity to an intermediary protein such as F-actin. In vitro studies have shown that VASP tetramers remain bound to the growing barbed ends of F-actin filaments during the elongation process (15). However, we observed few or no PLA spots in FSK-stimulated cells or in unstimulated cells, and there were also very few spots in ACh-stimulated cells treated with VASP S157A or vinculin Y1065F. These observations are consistent with the possibility that VASP exists in a monomeric form in unstimulated cells and assembles into tetramers at membrane sites after ACh stimulation, and that the assembly of VASP tetramers is prerequisite to the functional role of VASP in actin dynamics.
PLA provides a very high resolution method for the detection of protein-protein interactions and protein complexes in situ (39, 40). We targeted the phospho-Ser239 epitope on VASP proteins using two different species antibodies; thus the PLA signal should reflect the interaction between at least two VASP monomers. The VASP Ser239 epitope was selected as a target because VASP Ser239 phosphorylation does not increase above basal levels within the first 10 min of ACh stimulation (Fig. 6B); therefore an increase in the PLA signal in response to ACh stimulation cannot reflect an increase in VASP Ser239 phosphorylation. In addition, although stimulation of the cells with FSK causes a 5–6-fold increase in VASP Ser239 phosphorylation (Fig. 6, A and B); no PLA signal for VASP-VASP complexes was detected in FSK-stimulated cells, which further confirms that an increase in VASP Ser239 phosphorylation does not cause an increase in the PLA signal (Fig. 6C). It remains possible that the VASP-VASP PLA signal reflects the formation of VASP multimeric complexes with other proteins in response to stimulation with ACh. However, if VASP is present primarily in tetrametric form in unstimulated cells, FSK-treated cells, and in cells treated with mutant VASP or vinculin, it is unclear why these VASP tetramers do not generate a PLA signal.
Our previous studies demonstrated that the actin polymerization catalyst N-WASp is important for the activation of actin polymerization by contractile agonists. We proposed that N-WASp-mediated actin polymerization is likely to occur in submembranous regions of the smooth muscle cell, and that these actin filaments are therefore most likely distinct from the actin filaments that participate in actomyosin cross-bridge cycling (1). VASP is known to catalyze actin polymerization by mechanisms that are different from those of N-WASp (4, 7). It is therefore possible that VASP regulates actin polymerization in a subset of actin filaments that are distinct from those regulated by N-WASp, and that VASP-mediated actin polymerization constitutes a separate pathway for actin remodelling in ASM (2). However, our data show that VASP-protein interactions that are necessary for its actin catalytic activity occur at adhesome junctions at the plasma membrane and not in the cytoplasm; VASP-mediated actin polymerization is most likely also confined to submembranous pools of actin at the cell cortex that are distinct from actin filaments in the contractile apparatus in the interior of the cell. Although contractile filament actin may undergo remodelling during ASM contraction and relaxation (45), mechanisms that might regulate the remodelling of cytoplasmic contractile actin have not yet been described.
Actin polymerization is recognized as an essential component of the cellular response to agonist stimulation in many smooth muscle cell types; however, its role in the cellular processes that regulate the activation of signaling pathways and the functional responses of the smooth muscle cells remain unclear. Cortical actin may serve a structural role to fortify linkages between the membrane junctional adhesome sites and actin filaments within the contractile apparatus. In this capacity, newly polymerized actin would serve to strengthen linkages between membrane junctional proteins such as α-actinin, vinculin, and talin and the actin filaments that interact with myosin to mediate cell shortening and tension development. Regulation of the connections between the contractile apparatus and the cell membrane could also serve to modify the orientation of the contractile filaments to adapt the cell to shape changes imposed by forces within its external environment. This could be particularly important in ASM, which is continuously subjected to changes in stress and strain during breathing (46). However, the function of cortical actin dynamics may not be entirely structural: cortical actin filaments may also provide a lattice for the assembly of junctional complexes that transduce signals from environmental stimuli to the interior of the cell. Although the function and regulation of actin cytoskeletal structures within smooth muscle remain to be established, dynamic remodelling of the actin cytoskeletal lattice is likely to serve multiple functions within the ASM cell.
*This work was supported, in whole or in part, by National Institutes of Health Grants R01 HL29289, HL074099, and R01HL109629.
2The abbreviations used are: