Cell Boundary Elongation by Non-autonomous Contractility in Cell Oscillation  Yusuke Hara, Murat Shagirov, Yusuke Toyama  Current Biology  Volume 26, Issue 17, Pages 2388-2396 (September 2016) DOI: 10.1016/j.cub.2016.07.003 Copyright © 2016 Elsevier Ltd Terms and Conditions

Current Biology 2016 26, 2388-2396DOI: (10.1016/j.cub.2016.07.003) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 1 Relative Spatial Distribution of Medial MyoII Correlates with Cell Boundary Oscillation (A) Dynamics of amnioserosa cells during dorsal closure. Cell boundaries were visualized by Dα-cateninRFP (A′; Movie S1A). Forced-colored images show the rate of area change (A″) and the deformation rate of cell boundaries (A″′). Red and blue color indicate contracting area/boundaries and expanding area/boundaries in (A″) and (A″′), respectively. Asterisks indicate typical cells that show unsynchronized cell boundary behaviors. Scale bar, 20 μm. (B) Top: a junction and regions where medial MyoII is measured. The myosin ratio is a ratio between the average sqhGFP intensity in the red regions and the intensity in the blue regions. Bottom: a ratio that is larger or smaller than 1 represents the case where medial myosin accumulates more in “a” or “b” cells, respectively. (C) Dynamics of medial myosin accumulation during cell boundary oscillation. Cell boundary and MyoII are visualized by Dα-cateninRFP (magenta) and sqhGFP (green) (C′). Forced-colored images show the deformation rate (C″) and myosin ratio (C″′). The color code of (C″) is the same as in (A″′). Red and blue color in (C″′) indicate a high and low ratio, respectively. Yellow arrowheads indicate the boundary we quantify in (D). Scale bar, 10 μm. See also Movies S1D and S1E. (D) Evolution of the myosin ratio and deformation rate of an oscillating boundary highlighted in (C″). The red solid line, dashed red line, and green line represent the moving average of the deformation rate, deformation rate (raw data), and myosin ratio, respectively. (E) Cross-correlation analysis between the deformation rate and the rate of the myosin ratio. The black line shows the average correlation coefficients (n = 13 boundaries were randomly picked from an embryo). The faint colored lines indicate individual data. (F) Temporal relation between the deformation rate, myosin ratio, and rate of the myosin ratio. See also Figure S1B. (G) Cartoon showing a junction (black) and nearest (red) and next-nearest (green) junctions. (G′) Cross-correlation of changes in boundary length between adjacent cell boundaries (black and red junctions; black graph) and between next-nearest cell boundaries (black and green junction; gray graph) (n = 13 boundaries were randomly picked from an embryo). Each analysis was done between a black boundary and either 4 red or 8 green boundaries. Colored circles on the black line indicate the results of statistical analysis against the gray line. Statistical significance was calculated by Wilcoxon rank-sum test. Data represent mean ± SD. (H) Laser ablation targeting medial MyoII in a neighboring cell. Cell boundaries and MyoII are visualized by Dα-cateninRFP (white) and sqhGFP (green). Yellow dots indicate nodes of a boundary of interest. The white dotted lines denote the initial point of the vertex (i). The white arrow and red arrowhead indicate MyoII accumulation and the point of laser ablation, respectively. Scale bar, 5 μm. See Movie S2A. (H′ and H″) Kymographs of sqhGFP (H′) and Dα-cateninRFP (H″) along a line defined by two vertices (i and ii). White dotted lines denote the trajectories of the vertices. Red arrowheads and black dotted lines indicate the timing of laser ablation. The white arrowhead in (H′) highlights MyoII disruption. The white arrow in (H″) highlights the retraction of the vertex after ablation. Numbers at the bottom of (H′) and (H″) represent the timing (s) of still images in (H). Vertical and horizontal scale bars, 5 μm and 10 s, respectively. (I) Four different medial MyoII ablation experiments (i–iv). Green color represents the accumulation of MyoII. D indicates the length of a boundary analyzed. (I′) Boxplots showing the relative boundary length at 25 s after ablation (Dt25) with respect to the length at the time of ablation (Dt0), Dt25/Dt0. A value smaller or larger than 1 represents the shrinking or elongating boundary, respectively. Sample number is indicated at the upper right of each boxplot. See (I) for different conditions of the laser ablations (i–iv). Statistical tests were done by Steel-Dwass test. ∗∗p < 0.01, ∗∗∗p < 0.005. See also Figure S2. Current Biology 2016 26, 2388-2396DOI: (10.1016/j.cub.2016.07.003) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 2 Junctional Tension Correlates with Cell Junctional Dynamics (A) Left: three classes of cell junctional dynamics. Right: boundaries that connect to delaminating cells are categorized as “rosette.” Yellow arrowheads indicate rosette-type boundaries. Scale bar, 20 μm. (B) Left: definition of cell boundary waviness (Wn; Supplemental Experimental Procedures). A Wn of 1.0 denotes a straight boundary. Right: examples of different Wn. (C) Comparison of the recoil velocities of different junctional dynamics and waviness. Sample number is indicated at the upper right of each boxplot. Statistical tests were done by Wilcoxon rank-sum test. ∗p < 0.05, ∗∗p < 0.01. n.s., not significant. Statistical tests between different junctional dynamics among the same category of waviness are shown in Figures S3A–S3C. (D–F) Correlations between deformation rate and recoil velocity (D), deformation rate and medial myosin ratio (E), and medial myosin ratio and recoil velocity (F). Red, green, blue, and purple dots denote individual data of the contracting (low Wn, n = 31; high Wn, n = 30), stable (low Wn, n = 20; high Wn, n = 16), elongating (low Wn, n = 36; high Wn, n = 25), and rosette boundaries (n = 16), respectively. Solid black lines represent the linear fit of plots. Gray regions indicate the confidence interval of the fitting. The correlation coefficient (R) and its significance are shown in each graph. ∗p < 0.05, ∗∗∗p < 0.005. See also Figures S3A–S3C. Current Biology 2016 26, 2388-2396DOI: (10.1016/j.cub.2016.07.003) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 3 Non-invasive Estimation of Cell Junctional Tension Distribution across the Amnioserosa and Its Validation (A) Categorization of junctions with respect to their relative tension and behavior. The categories (I–V) are defined by the junctional tension measurements (Figure 2C). The tension values are normalized by the lowest forces, which are represented by the stable boundaries with high Wn and the elongating boundaries. Cont., contracting; Stb., stable; Elng., elongating; lowWn, low waviness; highWn, high waviness. (B) Still from a movie showing a tension map across the AS (Movie S3B). All boundaries except those at the edge of the tissue and around the anterior- and posterior-most regions were analyzed. Colors denote each category. Anterior of the embryo is to the left. (B′) The fraction of each category (n ∼ 340 boundaries over 60 frames from an embryo). (B″) Still from a movie shows hot/cold spot analyses across AS cells (Movie S3C). Colors denote the degree of hot/cold spots. (C) Measurement of the average tension. Junctional tension was divided into a medio-lateral component (TML) and an anterior-posterior component (TAP). (C′ and C″) Average of TML and TAP in each A-P section (<TML>, C′) and M-L section (<TAP>, C″), respectively. The number of boundaries analyzed (n) is shown in the graphs. Data are represented as mean ± SD. Statistical tests were done by Steel-Dwass test, and the results are shown at the bottom of the panels. (D) Illustration of the tissue-level laser ablations. Laser ablations along either the A-P axis (AP-cut) or M-L axis (ML-cut). Red and blue arrows represent the relaxation of the tissue upon laser ablation. See also Figures S4A and S4B. (D′ and D″) Average recoil speed upon AP-cut (D′) and ML-cut (D″) at different positions (Supplemental Experimental Procedures; Figures S4A and S4B). The number of laser ablations analyzed is shown at the top of the graphs. Data represent mean ± SD. Statistical tests were done by Kruskal-Wallis one-way ANOVA. No significant differences were found. Current Biology 2016 26, 2388-2396DOI: (10.1016/j.cub.2016.07.003) Copyright © 2016 Elsevier Ltd Terms and Conditions

Figure 4 Changes in Junctional Tension during Cell Boundary Oscillation Correlate with Vinculin Dynamics (A) Changes in the level of vinculinGFP, boundary length, and relative junctional tension during cell boundary oscillation. Open and filled triangles indicate elongating and contracting boundaries, respectively. Vinculin was visualized by expressing hs-vinculinGFP. Colors denote the level of vinculinGFP (A′; Movie S4B). Forced-colored images show the deformation rate of cell boundaries (A″) and the estimated relative junctional tension (A″′; Movie S4C). Scale bar, 5 μm. (B) Evolution of boundary length (top, red), vinculinGFP intensity (top, green), change in vinculin intensity (three-frame moving average; middle), and the estimated relative junctional tension of a boundary highlighted in (A). Faint red and white regions indicate times of higher and lower junctional tension, respectively. (C, C′, and D) Cross-correlation analysis between boundary length and vinculinGFP intensity (C), between boundary length and membraneGFP (PLCγ-PH-GFP; PH domain of phospholipase C-γ fused with GFP) intensity (C′), and between junctional tension and the rate of vinculinGFP intensity (D). Black lines show the average correlation coefficients (n = 10 boundaries were randomly picked from an embryo in C; n = 13 in C′; and n = 10 in D). The faint colored lines indicate individual data. See also Figures S4E′–S4E″′. (E) A model of AS cell boundary oscillation and vinculin dynamics during early dorsal closure. Current Biology 2016 26, 2388-2396DOI: (10.1016/j.cub.2016.07.003) Copyright © 2016 Elsevier Ltd Terms and Conditions