Vinculin associates with endothelial VE-cadherin junctions to control force-dependent remodeling

Abstract

To remodel endothelial cell-cell adhesion, inflammatory cytokine- and angiogenic growth factor-induced signals impinge on the vascular endothelial cadherin (VE-cadherin) complex, the central component of endothelial adherens junctions. This study demonstrates that junction remodeling takes place at a molecularly and phenotypically distinct subset of VE-cadherin adhesions, defined here as focal adherens junctions (FAJs). FAJs are attached to radial F-actin bundles and marked by the mechanosensory protein Vinculin. We show that endothelial hormones vascular endothelial growth factor, tumor necrosis factor α, and most prominently thrombin induced the transformation of stable junctions into FAJs. The actin cytoskeleton generated pulling forces specifically on FAJs, and inhibition of Rho-Rock-actomyosin contractility prevented the formation of FAJs and junction remodeling. FAJs formed normally in cells expressing a Vinculin binding-deficient mutant of α-catenin, showing that Vinculin recruitment is not required for adherens junction formation. Comparing Vinculin-devoid FAJs to wild-type FAJs revealed that Vinculin protects VE-cadherin junctions from opening during their force-dependent remodeling. These findings implicate Vinculin-dependent cadherin mechanosensing in endothelial processes such as leukocyte extravasation and angiogenesis.

Figure 1.

Vinculin marks distinct, remodeling cell–cell junctions attached to radial actin bundles. (a) Still images and enlarged views from time-lapse recordings (Video 2) showing perpendicularly oriented remodeling cell–cell junctions and linear stable/mature cell–cell junctions in a monolayer of HUVECs expressing VE-cadherin–GFP (green) and the F-actin probe Lifeact-mCherry (red). (b) IF images of HUVECs stained for Vinculin (green), VE-cadherin (red), and F-actin (blue) showing specific colocalization of Vinculin with perpendicular remodeling junctions, the FAJs (middle), and the absence of Vinculin from stable/mature linear junctions (bottom). (c, top) Merged IF images of HUVECs stained for Vinculin, phospho-Y118-Paxillin, phospho-Y397-FAK, or Talin (green) together with VE-cadherin (red). (c, bottom) Accompanying fluorescence intensities along the depicted lines showing that Vinculin, but not Paxillin, FAK, or Talin (green lines) colocalize with VE-cadherin (red lines) at FAJs. See also Fig. S1 for details. Bars: (a and b) 20 µm; (c) 5 µm.

Figure 2.

VEGF, TNF, and Thrombin induce the formation of FAJs. (a) IF images of Vinculin (green), VE-cadherin (red), and F-actin (blue) in HUVEC monolayers that were left untreated (control) or stimulated with VEGF for 4 h or TNF for 24 h. Arrows point to FAJs that are characteristic for VEGF and TNF treatments. Bar, 20 µm. (b) IF images of HUVECs that were untreated or stimulated with thrombin for 10 min, and stained as in a. Please note the strong induction of FAJs (arrows) by thrombin. Bar, 20 µm. (c) Graph shows a quantification of the fraction of Vinculin-positive junction fragments (detected by automated image segmentation based on VE-cadherin signal, see Materials and methods) in control (n = 15 images) and in thrombin (n = 17 images)-treated HUVECs of two independent experiments. Values are averages ± SEM (error bars). P-value was calculated with a two-tailed, homoscedastic Student’s t test.

Figure 3.

Thrombin induces FAJs formation by transformation of stable AJs. (a) Time-lapse images of HUVECs expressing p120-catenin–mCherry (red) and Vinculin-GFP (green) after thrombin stimulation. Merged images on the right highlight the rapid recruitment of Vinculin during thrombin-induced junction remodeling at the region of interest. See corresponding Video 4 for the ∼1-h time-lapse recording. (b) Time-lapse images of HUVECs expressing α-catenin–Dendra2 before and after photoswitching a fraction of a stable junction using a 405-nm confocal laser followed by thrombin-induced FAJ formation. See corresponding Video 5 for an 8-min time-lapse recording. Bars: (a, left) 10 µm; (a, right) 5 µm; (b) 10 µm

Figure 4.

FAJs require actomyosin contraction for their formation. (a) IF images of HUVECs stained for Vinculin (green), VE-cadherin (red), and F-actin (blue) that were treated with membrane-permeable C3 transferase for 4 h to inhibit Rho, Y-27632 for 10 min to inhibit Rock, blebbistatin for 30 min to inhibit myosin activities, or VE-cadherin blocking antibody for 2 h. Line scans on the right show intensities of Vinculin, VE-cadherin, and F-actin signal across indicated junctions. See also Fig. S3 for details. Bar, 20 µm. (b) Quantification (as in Fig. 2 c) of the fraction of Vinculin-positive junction fragments in HUVECs treated with C3 transferase (n = 10 images), Y-27632 (n = 9 images), or blebbistatin (n = 10 images) compared with control (n = 14 images) of two independent experiments. Values are averages ± SEM (error bars). P-values were calculated with a two-tailed, homoscedastic Student’s t test.

Figure 5.

FAJs are tensile junctions. (a) Time-lapse images of HUVECs expressing p120-catenin–mCherry (red) and Vinculin-GFP (green) during treatment with a low dose of Cytochalasin D. Note the specific and rapid displacement of the Vinculin-containing junction, whereas the Vinculin-negative junction does not move in this time period. See corresponding Video 7 for 16-min time-lapse recordings. (b) Quantification of the mean distance of translocation ± SEM of the p120-catenin signal within 60 s after Cytochalasin D treatment of Vinculin-positive junctions (n = 14) and Vinculin-negative junctions (n = 14) from five different experiments as determined by manual measurements in ImageJ. P-value was calculated with a two-tailed, homoscedastic Student’s t test. (c) Time-lapse images of FAJs in HUVECs expressing VE-cadherin–GFP (green) and Lifeact-mCherry (red), before and 15 s after laser ablation at the indicated region. See corresponding Video 8 for an ∼1-min time-lapse recording, which is representative of multiple experiments. The image on the right is a kymograph showing the intensity of VE-cadherin–GFP in time along the dotted line (shown is the maximum intensity pixel of a 10-pixel-wide line). (d and e) IF images of control and thrombin-treated HUVECs stained with the conformation-sensitive α18 rat monoclonal antibody for α-catenin (green), phalloidin for F-actin (blue), and antibodies for VE-cadherin (d) or Vinculin (e; red). Bars: (a) 5 µm; (c) 5 µm; (d and e) 20 µm.

Figure 6.

Junctional Vinculin restrains force-dependent junction disruption by thrombin. (a) Schematic representation of Vinculin and α-catenin to illustrate which homolog’s domain (dark green) was swapped to generate a hybrid α-catenin–ΔVBS that is unable to associate with Vinculin. (b) Representative Western blot analysis of α-catenin and actin in lysates of HUVECs transduced with α-catenin shRNAs and rescued by α-catenin–GFP or α-catenin–ΔVBS–GFP. (c) IF images of control HUVECs and HUVECs transduced with lentiviral shRNA against human α-catenin stained for VE-cadherin (green) and F-actin (red). Bar, 20 µm. (d) IF images of α-catenin shRNA-transduced HUVECs rescued with wild-type α-catenin–GFP (top) or α-catenin–ΔVBS–GFP (bottom) that were stimulated with thrombin for 10 min, and stained for Vinculin (red) and F-actin (blue). Colocalization of Vinculin with α-catenin–GFP or α-catenin–ΔVBS–GFP was analyzed by line scans displaying signal intensity across the FAJs as indicated. Bar, 10 µm. (e) Representative Western blot analysis of GFP and Vinculin in total lysates and in GFP immunoprecipitations from thrombin-stimulated HUVECs expressing indicated GFP constructs. (f) Quantification of the average junction width ± SEM after 10 min of thrombin treatment as measured using ImageJ in IF stainings of two independent experiments of control HUVECs (13 images, n = 382 junction width measurements), α-catenin shRNA–transduced HUVECs rescued with α-catenin–GFP (10 images, n = 987), or α-catenin–ΔVBS–GFP (10 images, n = 1,201). P-value was calculated with a two-tailed, homoscedastic Student’s t test. (g) Time-lapse images of α-catenin shRNA transduced HUVECs rescued with α-catenin–GFP (top) or α-catenin–ΔVBS–GFP (bottom) that were stimulated with thrombin, showing that thrombin induces wider remodeling junctions that persist longer in α-catenin–ΔVBS–GFP cells than in α-catenin–GFP cells. See corresponding Video 10 for ∼3-h time-lapse recordings. Bar, 10 µm. (h) Quantification of the mean junction width ± SEM after thrombin of α-catenin–GFP (nine time-lapse recordings) and α-catenin–ΔVBS–GFP (seven time-lapse recordings) junctions of two independent experiments as measured in time-lapse recordings using ImageJ. The number of junction width measurements (n value) of α-catenin–GFP and α-catenin–ΔVBS–GFP at time points 0, 10, 15, 20, 30, and 60 min after thrombin were 208, 409; 212, 375; 143, 311; 67, 180; 69, 126; and 149, 230, respectively.