Figure 3 - Actin-dependent movement of VE-cadherin.

(a) Cytochalasin D (1 g ml^-1) blocks cadherin flow. Kymography with sampling at the magenta line shown on the left. After the addition of cytochalasin D at the time indicated by the arrow, cadherin flow stops (see Supplementary Information, Movies 6, 8). The black arrowheads indicate the apical-most region of the junction. (b) Immunostaining for E-catenin (magenta) and actin filaments (green) in a VE-cadherin transfected A431D cell. The white box is enlarged on the right. E-catenin is localized along actin filaments (arrows). (c) Simultaneous time-lapse imaging of actin (magenta) and VE-cadherin (green) in Alexa 568 phalloidin-injected A431D cells. Upper right: cadherin clusters move on actin filaments — an example is traced with white arrows. Lower: cadherin clusters jump between actin filaments. A single cluster, indicated by the white arrow, splits into three pieces, which fuse again to form two pieces, during the jumping process (see Supplementary Information, Movie 7). The dotted lines indicate the apical-most portion of the junction, which is not in focus in each panel. (d) ML-7 (10 M) blocks cadherin flow. Kymography with sampling at the magenta line shown on the left. After the addition of ML-7 at the time indicated by the arrow, cadherin flow stops (see Supplementary Information, Movies 9). (e) Y-27632 (10 M) deforms apical cell junction, but does not block cadherin flow. After the addition of Y-27632, the outlines of apical cell junctions are vigorously distorted, including local collapse of the junctions (magenta arrow heads; compare the images taken 5 min before and 50 min after the addition of Y-27632. Cadherin flow still occurs at the sites where cell–cell contacts are maintained (green arrows; see Supplementary Information, Movies 10, as accurate kymograms could not be obtained, because of the dynamic relocation of cell junctions during the treatment). The scale bars represent 15 m in a, d and e, 30 m in b and 5 m in c.