Vimentin dephosphorylation at ser-56 is regulated by type 1 protein phosphatase in smooth muscle

All experimental protocols were approved by the Institutional Animal Care and Usage Committee (Animal Welfare Assurance Number A3099-01). C57BL/6 mice (25 ± 5 g, 8–12 weeks old) were originally purchased from Taconic Biosciences and bred in the specific pathogen free housing of Animal Research Facility, Albany Medical College. The animal housing was kept at 21–22 °C with 45–55 relative humidity. The light/dark cycle of the housing was 7 am- 7 pm for fluorescent/LED lights and 7 pm – 7 am for red lights. The numbers of cage companions were 3–8 each based on animal ages and gender. Animals were healthy and able to breed normally. They were fed with Purina Lab Diet 5P76 and had continuous access to food and water. Both male and female mice were randomized allocated to the experimental or control groups. The experimental unit represented each trachea from individual animals.

Animals were sacrificed with overdosed pentobarbital sodium (100 mg/kg). Tracheal rings (4–5 mm long) were immediately removed and placed at room temperature in physiological saline solution (PSS) containing 110 mM NaCl, 3.4 mM KCl, 2.4 mM CaCl₂, 0.8 mM MgSO₄, 25.8 mM NaHCO₃, 1.2 mM KH₂PO₄, and 5.6 mM glucose. The solution was aerated with 95 % O₂-5 % CO₂ to maintain a pH of 7.4. Tracheal rings were then placed in PSS at 37 °C in a 25-ml organ bath and attached to a Grass force transducer that had been connected to a Gould recorder or a computer with A/D converter (Grass). At the beginning of each experiment, 0.5 g passive tension was applied to tracheal rings. After 60 min equilibrium they were stimulated with 80 mM KCl repeatedly until contractile responses and passive tension reached a steady state.

For lentivirus-mediated shRNA in tissues, the thin epithelium layer of tracheal rings was removed by using forceps. They were then transduced with lentivirus encoding pan PP1 shRNA (sc-43533-V, Santa Cruz Biotechnology) or viruses for control shRNA (sc-108080, Santa Cruz Biotechnology) for 4 days. Contraction in response to agonist activation was compared before and after lentivirus transduction. For biochemical analysis, tissues were frozen using liquid nitrogen and pulverized as previously described [3, 4, 21].

Pulverized tissues were lysed in SDS sample buffer composed of 1.5 % dithiothreitol, 2 % SDS, 80 mM Tris–HCl (pH 6.8), 10 % glycerol, 0.01 % bromophenol blue, phosphatase inhibitors (2 mM sodium orthovanadate, 2 mM molybdate, and 2 mM sodium pyrophosphate) and protease inhibitors (2 mM benzamidine, 0.5 mM aprotinin and 1 mM phenylmethylsulfonyl fluoride). The lysates were boiled in the buffer for 5 min and separated by SDS-PAGE. Proteins were transferred to a nitrocellulose membrane. The membrane was treated with blockers for 1 h and probed with use of primary antibody followed by horseradish peroxidase-conjugated secondary antibody (Fisher Scientific). Proteins were visualized by enhanced chemiluminescence (Fisher Scientific) using the LAS 400 Fuji imaging system or GE Amersham Imager 600 system. Antibodies against PP1, PP2A, phospho-myosin light chain (Ser-19), myosin light chain, phospho-c-Abl (Tyr-412) and pan c-Abl were purchased from Santa Cruz Biotechnology. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) antibody was purchased from Fitzgerald (MA, USA). Antibodies against vimentin and p130CAS were obtained from BD Biosciences. Phospho-vimentin (Ser-56) antibody was custom made by our laboratory as previously described [4, 9, 16]. Antibody against α-actin was acquired from Sigma-Aldrich. The levels of proteins were quantified by scanning densitometry of immunoblots (Fuji Multigauge Software or GE IQTL software). The luminescent signals from all immunoblots were within the linear range [20, 21, 28, 40, 41].

Co-immunoprecipitation analysis was used to evaluate protein-protein interactions as previously described [9, 20, 22]. Briefly, tissue extracts were incubated overnight with corresponding antibodies and then incubated for 2–3 h with 125 μl of a 10 % suspension of protein A-Sepharose beads. Immunocomplexes were washed four times in buffer containing 50 mM Tris–HCl (pH 7.6), 150 mM NaCl and 0.1 % Triton X-100. The immunoprecipitates were separated by SDS-PAGE followed by transfer to nitrocellulose membranes. The membranes of immunoprecipitates were probed with corresponding antibodies.

Cytoskeletal vimentin was collected using the method as previously described [4, 9, 10, 16, 31]. Briefly, tissues were homogenized in buffer containing 1 % Nonidet P-40, 10 % glycerol, 20 mM Hepes, pH 7.6, 150 mM NaCl, 2 mM sodium orthovanadate, 2 mM molybdate, 2 mM sodium pyrophosphate and protease inhibitors (2 mM benzamidine, 0.5 mM aprotinin and 1 mM phenylmethylsulfonyl fluoride). The homogenates were immediately incubated in the same buffer at 37 °C for 30 min. The soluble (depolymerized) and insoluble (cytoskeletal) fractions were separated after centrifugation at 5,200 rpm, 4 °C for 30 min. The equal amount of cytoskeletal vimentin was separated by 10 % SDS-PAGE, and was transferred to nitrocellulose membranes. The membranes were cut into two parts; the upper part was probed with monoclonal p130CAS antibody. The lower part of the membrane was blotted with vimentin antibody. The ratio of p130CAS to vimentin was calculated after densitometrical analysis of immunoblots.

The content of F-actin and G-actin in smooth muscle was measured using a method as previously described [21, 22]. Briefly, tissues were treated with F-actin stabilization buffer (50 mM PIPES, pH 6.9, 50 mM NaCl, 5 mM MgCl₂, 5 mM EGTA, 5 % glycerol, 0.1 % Triton X-100, 0.1 % Nonidet P40, 0.1 % Tween 20, 0.1 % beta-mercapto-ethanol, 1 mM ATP, 1 μg/ml pepstatin, 1 μg/ml leupeptin, 10 μg/ml benzamidine). The supernatants of protein extracts were collected after centrifugation at 151,000 g for 60 min at 37 °C. The pellets were resuspended in ice-cold H₂O plus 1 μM cytochalasin D and then incubated on ice for 1 h to dissociate F-actin. The resuspended pellets were gently mixed every 15 min. The supernatant of the resuspended pellets was collected after centrifugation at 16,100 g for 2 min at 4 °C. Equal volume of the first supernatant (G-actin) or second supernatant (F-actin) was subjected to immunoblot analysis using α-smooth muscle actin antibody (Sigma). The amount of F-actin and G-actin was determined by scanning densitometry.

Although protein phosphatases have been implicated in vascular and uterine smooth muscle contraction [36, 37], the role of protein phosphatases in airway smooth muscle is not well elucidated. Mouse tracheal rings were treated with okadaic acid, a potent PP1 and 2A inhibitor [36, 37]. Contractile responses were then measured using a muscle research system [4, 5, 20, 28, 29, 31]. A representative force tracing after treatment with okadaic acid was shown in Fig. 1a. Treatment with okadaic acid induced mouse tracheal smooth muscle contraction, which was dose-dependent (Fig. 1b). However, treatment with okadaic acid did not affect the ACh-induced contraction (Fig. 1a). The results suggest that the inhibition of PP1 and/or PP2A induces airway smooth muscle contraction, but not sensitize the precontracted smooth muscle.

Previous studies suggest that vimentin phosphorylation at Ser-56 is critical for smooth muscle contraction [3‐5, 9, 12, 27]. Vimentin phosphorylation at this position regulates smooth muscle contraction by inducing the redistribution of signaling molecules and the spatial reorientation of the vimentin network [3‐5, 9]. Furthermore, vimentin phosphorylation at Ser-56 is catalyzed by PAK1 [4, 5, 9, 12, 16]. However, it is currently unknown which phosphatases mediate vimentin Ser-56 dephosphorylation in smooth muscle. Mouse tracheal smooth muscle tissues were treated with okadaic acid for 1–10 min. Vimentin phosphorylation at Ser-56 and contraction were then evaluated. Treatment with okadaic acid induced vimentin phosphorylation at Ser-56 (Fig. 1c), which was coordinately associated with an increase in smooth muscle contraction (Fig. 1d). These results suggest that PP1 and/or PP2A may mediate vimentin dephosphorylation at this position in airway smooth muscle.

To assess the role of PP1 and/or PP2A, we evaluated the interaction of the phosphatases with vimentin. Smooth muscle tissues were stimulated with acetylcholine (ACh), or left unstimulated. Extracts of these tissues were immunoprecipitated with vimentin antibody and immunoblotted with antibodies against vimentin, PP1 or PP2A.

In unstimulated tissues, PP1 was detected in vimentin immunoprecipitates. Upon stimulation with ACh, the ratios of PP1 over vimentin immunoprecipitates were reduced as compared to unstimulated tissues (Fig. 2a). However, PP2A was barely found in the precipitates. Moreover, the ratios of PP2A/vimentin were not significantly affected by contractile stimulation (Fig. 2b). The results suggest that vimentin predominantly interacts with PP1 in unstimulated smooth muscle. Contractile activation induces the dissociation of PP1 from vimentin in smooth muscle. We conclude that PP1 is likely a candidate for vimentin Ser-56 dephosphorylation in smooth muscle.

To determine the role of PP1, we utilized a lentivirus-mediated shRNA approach [19‐21, 28, 41] to inhibit the expression of PP1. Contractile response of mouse tracheal rings to ACh was evaluated. The rings were then transduced with lentiviruses encoding control shRNA or PP1 shRNA. These tissues were incubated in the medium for 4 days. Immunoblot analysis verified the lower expression of PP1 in tissues transduced with viruses encoding PP1 shRNA compared to tissues transduced with viruses for control shRNA (Fig. 3a).

We then evaluated vimentin phosphorylation at Ser-56 in these tissues. Compared to tissues expressing control shRNA, basal vimentin phosphorylation at Ser-56 was not significantly altered in tissues expressing PP1 shRNA. More importantly, ACh-induced vimentin phosphorylation at Ser-56 was higher in PP1-deficient tissues than in tissues expressing control shRNA (Fig. 3b). We also determined the effects of PP1 knockdown on smooth muscle contraction. In tissues treated with lentivirus expressing control shRNA, ACh stimulation was able to stimulate contractile response. Moreover, contraction in PP1-deficient tissues was higher as compared to tissues expressing control shRNA (Fig. 3c). Passive tension in PP1-deficient tissues was slightly increased. We conclude that PP1 mediates vimentin dephosphorylation at Ser-56 and inhibits smooth muscle contraction.

Our previous studies have shown that vimentin phosphorylation enhances the disengagement of p130CAS from vimentin, which may promote smooth muscle contraction [3‐5, 10, 11]. We determined the effects of okadaic acid on the association of p130CAS with vimentin. Cytoskeletal vimentin for smooth muscle tissues was collected using the intermediate filament fractionation assay [4, 5, 9, 31]. The equal amount of cytoskeletal vimentin was separated by SDS-PAGE and immunoblotted with antibodies against vimentin and p130CAS. In tissues not treated with okadaic acid, the amount of p130CAS was relatively higher. In contrast, the amount of p130CAS in cytoskeletal vimentin was reduced in rings treated with okadaic acid. The ratios of p130CAS/vimentin were reduced upon treatment with okadaic acid (Fig. 4a).

We also assessed the effects of PP1 knockdown on the interaction of p130CAS with vimentin. Compared to tissues expressing control shRNA, the amount of p130CAS in the vimentin fraction was lower in PP1-deficient tissues during ACh stimulation (Fig. 4b). Taken together, we conclude that PP1 is able to inhibit the disassociation of p130CAS from the vimentin cytoskeleton.

It is well recognized that smooth muscle contraction is regulated by actin filament polymerization [12, 24, 26, 27]. Because PP1 is able to regulate smooth muscle force development, this raises a possibility that PP1 may modulate actin dynamics. To test this, we assessed the effects of PP1 knockdown on actin polymerization by using the actin fractionation assay [20‐22, 28]. In tissues treated with control shRNA, ACh stimulation was able to increase F/G-actin ratios, an indication of actin polymerization [20, 24]. Furthermore, the ACh-induced increases in F/G-actin ratios were higher in PP1-deficient tissues than in tissues expressing control shRNA (Fig. 5a). The results suggest that PP1 inhibits the agonist-induced actin filament assembly in smooth muscle.

c-Abl is a nonreceptor protein tyrosine kinase that regulates actin dynamics and smooth muscle contraction [21, 22, 27, 29, 30]. Since PP1 silencing affects actin filament polymerization, we questioned whether PP1 has a role in regulating c-Abl tyrosine kinase. Tissues expressing control shRNA and PP1-deficient tissues were stimulated with ACh, or left unstimulated. c-Abl phosphorylation at Tyr-412, an indication of c-Abl activation [12, 22, 24, 27], in these tissues was evaluated by immunoblot analysis. The ACh-induced c-Abl tyrosine phosphorylation in PP1-deficient tissues was similar to that in tissues expressing control shRNA. In addition, basal c-Abl phosphorylation was not affected by PP1 knockdown (Fig. 5b). We conclude that PP1 is not involved in the phosphorylation of c-Abl tyrosine kinase.

Because myosin light chain phosphorylation at Ser-19 is one of the major cellular processes in smooth muscle in response to agonist stimulation [32‐35], we determined the effects of PP1 knockdown on myosin light chain phosphorylation. Compared to tissues expressing control shRNA, myosin light chain phosphorylation at Ser-19 upon ACh stimulation was higher in PP1- deficient tissues (Fig. 6). Knockdown of PP1 did not significantly affect myosin light chain phosphorylation without ACh stimulation (Fig. 6).

Since PP1 knockdown enhances the ACh-induced vimentin phosphorylation at Ser-56 (Fig. 3b), myosin light chain phosphorylation (Fig. 6) and contraction (Fig. 3c), it is likely that PP1 regulates smooth muscle contraction by controlling both the vimentin-associated pathway and myosin light chain-associated pathway.