VASP phosphorylation at serine239 regulates the effects of NO on smooth muscle cell invasion and contraction of collagen

Abstract

Nitric oxide triggers cGMP-dependent kinase-mediated phosphorylation of the actin regulator vasodilator-stimulated phosphoprotein (VASP) at residue serine239. The function of this phosphorylation for smooth muscle cell (SMC) adhesion, spreading, matrix contraction, and invasion is not well understood. We reconstituted VASP deficient SMC with wild-type VASP (wt-VASP) or VASP mutants that mimic "locked" serine239 phosphorylation (S239D-VASP) or "blocked" serine239 phosphorylation (S239A-VASP). Collagen gel contraction was reduced in S239D-VASP compared to S239A-VASP and wt-VASP expressing cells and nitric oxide (NO) stimulation decreased gel contraction of wt-VASP reconstituted SMC. Invasion of collagen was enhanced in S239D-VASP and NO-stimulated wild-type SMCs compared to S239A-VASP expressing cells. Expression of S239D-VASP impaired SMC attachment to collagen, reduced the number of membrane protrusions, and caused cell rounding compared to expression of S239A-VASP. Treatment of wt-VASP reconstituted SMCs with NO exerted similar effects as expression of S239D-VASP. As unstimulated cells were spreading on collagen S239A-VASP and wt-VASP localized to actin fibers whereas S239D-VASP was enriched in the cytosol. NO interferes with SMC invasion and contraction of collagen matrices. This requires phosphorylation of VASP on serine239, which reduces VASP binding to actin fibers. These findings support the conclusion that VASP phosphorylation at serine239 regulates cytoskeleton remodeling.

Figure 1

NO and expression of S239D-VASP in VASP^-/- SMCs inhibit SMC-mediated collagen gel contraction. (A) Western blots of cell lysates of growing aortic SMCs (C57BL/6 wild-type SMCs and the following all derived from VASP^-/- SMCs: LXSN SMCs, wt-VASP SMCs, S239A-VASP SMCs and S239D-VASP SMCs) were probed with antibodies against VASP or β-actin. (B) Collagen gel contraction by SMCs isolated from wild-type C57BL/6 mice or by wt-VASP SMCs was investigated in the absence or presence of a NO donor (100 μM deta NONOate). When indicated, SMCs were pretreated with deta NONOate for 30 minutes at room temperature. Data (mean±SEM; n=3-5 experiments) are expressed as percent of gel contraction. * With vs. without deta NONOate, P ≤ 0.05. (C) Collagen gel contraction by LXSN, wt-VASP, S239A-VASP or S239D-VASP SMCs. Data (mean±SEM; n=3-5 experiments) are expressed as the percent of gel contraction. * S239D-VASP SMCs vs. other cell types, P ≤ 0.05. (D) Collagen gel contraction by LXSN, wt-VASP, S239A-VASP or S239D-VASP SMCs in absence or presence of deta NONOate. Data (mean±SEM; n=3-5 experiments) are expressed as percent of gel contraction. * With vs. without deta NONOate, P ≤ 0.05.

Figure 1

NO and expression of S239D-VASP in VASP^-/- SMCs inhibit SMC-mediated collagen gel contraction. (A) Western blots of cell lysates of growing aortic SMCs (C57BL/6 wild-type SMCs and the following all derived from VASP^-/- SMCs: LXSN SMCs, wt-VASP SMCs, S239A-VASP SMCs and S239D-VASP SMCs) were probed with antibodies against VASP or β-actin. (B) Collagen gel contraction by SMCs isolated from wild-type C57BL/6 mice or by wt-VASP SMCs was investigated in the absence or presence of a NO donor (100 μM deta NONOate). When indicated, SMCs were pretreated with deta NONOate for 30 minutes at room temperature. Data (mean±SEM; n=3-5 experiments) are expressed as percent of gel contraction. * With vs. without deta NONOate, P ≤ 0.05. (C) Collagen gel contraction by LXSN, wt-VASP, S239A-VASP or S239D-VASP SMCs. Data (mean±SEM; n=3-5 experiments) are expressed as the percent of gel contraction. * S239D-VASP SMCs vs. other cell types, P ≤ 0.05. (D) Collagen gel contraction by LXSN, wt-VASP, S239A-VASP or S239D-VASP SMCs in absence or presence of deta NONOate. Data (mean±SEM; n=3-5 experiments) are expressed as percent of gel contraction. * With vs. without deta NONOate, P ≤ 0.05.

Figure 2

NO and expression of S239D-VASP enhance invasion through a collagen gel barrier. When indicated, SMCs were pretreated with detaNONOate (100 μM) for 30 minutes at room temperature. Invasion of SMCs through a thick collagen gel with PDGF-BB (10ng/mL) in the bottom chambers was investigated in the absence or presence of NO. Data (mean±SEM; n=3 experiments) are expressed as cells / mm². * P ≤ 0.05.

Figure 3

NO and expression of S239D-VASP inhibit adhesion. When indicated, SMCs were pretreated with 100 μM deta NONOate for 30 minutes at room temperature. Adhesion to polymeric collagen was investigated in the absence or presence of deta NONOate. Data (mean±SEM; n=3) are expressed as percent of adhesion of wt-VASP SMCS. *P ≤ 0.05.

Figure 4

NO and expression of S239D-VASP inhibit membrane protrusion formation and SMC spreading within collagen. (A) Cell morphology was analyzed by phase contrast microscopy 30 and 60 minutes after the cells were seeded within collagen. Scale bar represents 10μm. (B) wt-VASP, LXSN, S239A-VASP or S239D-VASP SMCs were pretreated with or without 100 μM deta NONOate for 30 minutes at room temperature and then seeded within collagen gel in presence or absence of deta NONOate. After 30 and 60 minutes, the cells were analyzed by fluorescence microscopy after staining with TRITC-conjugated phalloidin. Scale bar represents 5μm. The photomicrographs are representative of three independent experiments.

Figure 4

NO and expression of S239D-VASP inhibit membrane protrusion formation and SMC spreading within collagen. (A) Cell morphology was analyzed by phase contrast microscopy 30 and 60 minutes after the cells were seeded within collagen. Scale bar represents 10μm. (B) wt-VASP, LXSN, S239A-VASP or S239D-VASP SMCs were pretreated with or without 100 μM deta NONOate for 30 minutes at room temperature and then seeded within collagen gel in presence or absence of deta NONOate. After 30 and 60 minutes, the cells were analyzed by fluorescence microscopy after staining with TRITC-conjugated phalloidin. Scale bar represents 5μm. The photomicrographs are representative of three independent experiments.

Figure 5

Colocalization of VASP and F-actin is inhibited by phosphorylation of VASP serine239. wt-VASP, LXSN, S239A-VASP or S239D-VASP SMCs were plated on polymeric collagen-coated glass slides, and wt-VASP SMCs were pretreated with 100 μM deta NONOate for 30 minutes at room temperature and then plated on polymeric collagen-coated glass slides in the presence of 100 μM deta NONOate. 3 hours after plating localization of VASP and F-actin was analyzed by staining with rhodamine-phalloidin (red) and VASP (green). Boxed regions are enlarged in insets. Arrows indicated where VASP and F-actin are colocalized (yellow). Scale bar represents 5μm.