Rotational bias of cell nuclei reversed from CW to ACW bias during cell attachment.

(A) Phase contrast and H33342 staining of nuclei (blue) at 0 min and 240 min of the time-lapse imaging (Video 1). (B) Imaging processing of nuclei tracking. (Left) After binarization of nuclei, the ellipses of nuclei were shown for time point 0 min (white) and 240 min (blue). (Right) The boundary of cell nuclei at different time points, demonstrating the tracking of nuclei from the beginning to the end point, colored with a gradient from white to blue. (C) Schematic showing calculation of angular velocity using the angle difference of the nucleus at two consecutive images, 5 min per interval for 4 hour and 30 min. (D) The angular velocity of the nucleus measured for one cell. (E) Probability of ACW rotation of nuclei for each 20 min interval (n = 26). (F) Cell projected areas of each cell measured by phase contrast images after initial attachment of cells, with 20 min interval (mean shown in red). (G) Circularity calculated by the area and perimeter of cells. (H) Circularity of each cell calculated with 15 min interval (mean shown in red). (I and J) Probability of ACW/CW rotation of nuclei against cell projected area with circularity greater than 0.7 (I) or less than 0.7 (J). Two-sample equal variance two-tailed t-test. Scale bar: 100 μm.

Reversal from CW to ACW chirality with increasing micro-island area.

(A) Micropatterned islands of cell-adherent fibronectin and cell-repellent Pluronic F127, with areas from 500 μm2 to 2500 μm2, on which rotational bias was measured based on the angular velocity of the long axis of the binarized nucleus ellipse. After the cell was fully spread on the micropatterned island (2–3 hours), time-lapse images were taken every 10 min for 2–3 hours to measure the angular velocity. (B to D) Probability of ACW/CW rotation of HFF-1 human fibroblasts (B), C2C12 mouse myoblasts (C), and human mesenchymal stem cells (hMSCs) (D) on micro-islands with varied area within 3 hours. Mean ± SEM. Two-sample equal variance two-tailed t-test.

Actin at cell periphery promotes ACW chirality.

(A) Fluorescent staining of F-actin (white) by phalloidin, DAPI-stained nucleus (blue), and a stacked image showing heatmap distribution on a 2500 μm2 island. (B) Fluorescent staining of F-actin (white) by phalloidin, DAPI-stained nucleus (blue), and a stacked image showing heatmap distribution on a 750 μm2 island. (C and D) Probability of ACW/CW rotation on (C) 2500 μm2 of control (n = 57), treatment of A23 (n = 50) and LatA (n = 54), and (D) 750 μm2 for control (n = 34) and treatment of A23 (n = 70) and LatA (n = 43). Mean + SEM. Two-sample equal variance two-tailed t-test. (E) Fluorescent staining of F-actin (white) by phalloidin and DAPI-stained nucleus (blue) with treatment of A23 (left) and LatA (right) on a 2500 μm2 island. (F) Stacked image of (E) showing distribution of actin by heatmaps. (G) Fluorescent staining of F-actin (white) by phalloidin and DAPI-stained nucleus (blue) with treatment of A23 (left) and LatA (right) on a 750 μm2 island. (H) Stacked image of (G) showing distribution of actin by heatmaps. (I) Differential heatmap showing intensity differences of actin, comparing the treatment of A23 (left) and LatA (right) to the control on a 2500 μm2 island. (J) Curves of actin intensity from cell center to boundary (left) and differential actin intensity (right) comparing A23 and LatA treatment to the control on a 2500 μm2 pattern. (K) Differential heatmap showing intensity differences of actin, comparing the treatment of A23 (left) and LatA (right) to control on a 750 μm2 island. (L) Curves of actin intensity from cell center to boundary (left) and differential actin intensity (right) comparing A23 and LatA treatment to the control on a 750 μm2 pattern. Scale bar: 20 μm.

Polymerization of radial fibers drives ACW chirality.

(A) Lifeact-GFP transfected cell showing actin organization in the control on 2500 μm2 circular islands. With the increase in length, the radial fibers start to tilt rightward. Following the established direction, the transverse arcs (green arrows) flow toward the cell center and cause an overall ACW rotation of the entire cytoskeleton. (B and C) Photobleaching in a ring shape parallels the circumference of actin-GFP transfected cells of the untreated control (B) or LatA-treated samples (C). A kymograph (yellow box) of the photobleached untreated control was used to measure the centripetal velocity of transverse arcs and elongation rate of radial fibers. The green dashed line indicates the photobleached location of actin-GFP, while red and green arrows show the elongation of radial fiber and translocation of transverse arc to the cell center, respectively. In each cell sample, radial fiber elongation velocity was compared against the transverse arc translocation velocity. The velocity was taken as the average of 2–3 measurements from different locations of the same cell. (D) The radial fiber elongation velocity (left) and the transverse arc translocation velocity (right) in untreated control and cells with LatA treatment. For radial fiber elongation velocity measurement, n = 11 for the control and n =14 for LatA treatment, based on 4–5 cells. For the transverse arc translocation velocity measurement, n = 16 for the control and n =17 for LatA treatment based on 4–5 cells. Mean ± SEM. Two-sample equal variance two-tailed t-test. (E) Lifeact-mRuby transfected cell showing actin organization of 2500 μm2 circular pattern with LatA treatment. Motion vectors (green) demonstrate the accumulated motion of the actin structure. Polar histogram shows the distribution angle of motion vectors, indicating the transverse arcs’ CW swirling. Scale Bar: 20 μm.

Inhibition of myosin II enhances ACW chirality.

(A) Fluorescence staining of F-actin (red) by phalloidin, immunofluorescence-stained myosin IIa (yellow), and DAPI-stained nucleus (blue) of the control and cells with myosin II inhibitor Y27632 a 2500 μm2 pattern. (B) Stacked image of (A) showing heatmap distribution of actin (left) and myosin IIa (right). (C) The median intensity of myosin IIa in control (n = 52) and Y27632-treated (n = 41) cells. Mean± SEM. Two-sample equal variance two-tailed t-test. (D) Colocalization analysis by Pearson’s Rvalue of myosin IIa on actin filaments in control (n = 52) and Y27632-treated (n = 41) cells. Mean ± SEM. Two-sample equal variance two-tailed t-test. (E) Probability of ACW/CW rotation of control (reused data from Figure 3D, n = 57) and Y27632-treated (n = 53) cells on 2500 μm2 pattern. (F) Probability of ACW/CW rotation of control (reused data from Figure 3E, n = 34) and Y27632-treated (n = 53) cells on 750 μm2 pattern. Mean + SEM. Two-sample equal variance two-tailed t-test. Scale Bar: 20 μm.

Myosin IIa binding on actin fibers drives CW chirality.

(A) Fluorescent staining of control on a 2500 μm2 micropattern with myosin lla (left) and phalloidin-stained F-actin (right) with DAPI (blue), with the corresponding heatmap. (B) Western blotting validating the silencing effect of mDia2 (top) and Tpm4 (bottom) siRNA transfection compared to the control. (C) Fluorescent staining of mDia2-silenced HFF-1 cell on a 2500 μm2 micropattern with myosin lla (yellow) and phalloidin-stained (red) F-actin with the corresponding distribution heatmap. (D) The myosin IIa intensity analysis by the median intensity (left) and colocalization analysis by Pearson’s R-value (right), comparing the control (n = 15) and mDia2-silenced (n = 13) samples regarding the intensity of myosin IIa images and the colocalization of myosin IIa with actin fluorescence images. Mean ± SEM. Two-sample equal variance two-tailed t-test. (E) Probability of ACW/CW rotation on 2500 μm2 pattern with untreated control (reused data from Figure 3D, n = 57), mDia2 silencing (n = 48), and overexpression (n = 25). (F) Probability of ACW/CW rotation on 750 μm2 pattern with untreated control (reused data from Figure 3E, n = 34), mDia2 silencing (n = 53), and overexpressing (n = 22). Mean + SEM. Two-sample equal variance two-tailed t-test. (G) Fluorescent staining of Tpm4-silenced HFF-1 cell on a 2500 μm2 pattern with myosin IIa (yellow) and phalloidin-stained (red) F-actin with DAPI (blue), and the corresponding distribution heatmap. Differential heatmap shows the intensity difference between the actin of the Tpm4-silenced cells and the control on a 2500 μm2 pattern (right). (H) The myosin IIa intensity analysis by the median intensity (left) and colocalization analysis by Pearson’s R-value (right), comparing the control (reused data from (C), n = 15) and Tpm4 silenced (n = 10) samples in the intensity of myosin IIa images and the colocalization of myosin IIa with actin fluorescence. Mean ± SEM. Two-sample equal variance two-tailed t-test. (I) Probability of ACW/CW rotation on 2500 μm2 pattern with untreated control (reused data from Figure 3D, n = 57) and Tpm4- silenced cells (n = 44). (J) Probability of ACW/CW rotation on 750 μm2 pattern with untreated control (reused data from Figure 3E, n = 34) and Tpm4-silenced cells (n = 48). Mean + SEM. Two-sample equal variance two-tailed t-test. Scale Bar: 20 μm. GAPDH was used as a loading control in (B).

Actin cytoskeleton and myosin IIa height profile.

(A) The maximum projection of fluorescence staining of F-actin with phalloidin (left) and immunofluorescence staining of myosin IIa (right), with the side view showing the height profile of the cell from the region of interest (top), as indicated by the dashed green line on 2500 μm2 pattern. Myosin images represent the bottom region and middle region (right). (B) The merged fluorescence images of F-actin (red) and myosin IIa (yellow). The height profile with zoom-in (blue box) at the region of interest (right) is indicated by the dashed green line (left). (C) The maximum projection of F-actin fluorescence staining by phalloidin with Tpm4-silenced HFF-1 cells. The side view (top) shows the height profile of the cell from the region of interest (green line). We zoomed in at the lower right region (bottom) to visualize the structure. The F-actin structure at the bottom focal plane, from 0 to 0.3 μm, is labeled in blue, while the two subsequent sections above are labeled purple and green. The actin image represents the bottom region and middle region (lower right), showing the radial fibers at the base and the transverse arcs at the upper layer, near the cell ceiling. Scale Bar: 20 μm.

Actomyosin dynamics of radial fibers and transverse arcs in contribution to cell chirality.

(A) Schematic showing the radial fiber-generating rightward force. As the barbed ends of the right-handed double helix of an actin filament are capped and tethered by formin, incorporation of new actin monomers to the right-handed double helix causes the actin right-screw motion of radial fibers during polymerization. (B) Schematic of the myosin IIa gliding motion on Tpm4-decorated actin filament. Myosin IIa walks leftward from pointed end to barbed end. The green arrow represents the left-handed direction of the myosin IIa head footprint. (C) Schematic showing a unit of sarcomere-like contraction of myosin IIa and actin filament. During contraction, the actin filaments contract to the pointed end through myosin IIa, with a gliding motion creating left-handed twirling of actin filaments to the pointed end. (D) Schematic of transverse arcs producing CW swirling. Transverse arcs are located immediately beneath the cell membrane, away from the substrate. With respect to the cell surface, the contraction of the sarcomere-like unit in the transverse arcs gives rise to tilting. The green arrow demonstrates the rotational force of the sarcomere-like unit in the transverse arc. (E) When radial fiber is dominant, radial fibers with right-screw motion tilt rightward, accompanying retrograde flow along the growing direction of radial fibers and eventually driving an overall ACW rotation of the entire cytoskeleton around the cell nucleus. (F) When transverse arcs are dominant, the tilting of transverse arc precursors causes a CW swirling of transverse arcs that approaches the nucleus during retrograde flow, eventually causing CW rotational wrapping around the nucleus.

Nucleus rotation on micro-islands with varied areas and shapes.

(A-B) Phase contrast images of cells and the probability of ACW/CW rotation with areas of 750 μm2 (A) and 2500 μm2 (B). The pattern shapes used are equilateral triangle (circularity = 0.605), square (circularity = 0.785), and 1:3 rectangle (circularity = 0.589). Data presents the mean + SEM. Two-sample equal variance two-tailed t-test was used. Scale bar, 20 μm.

Actin orientation analysis of HFF-1 cell.

Fluorescent staining of DAPI-stained nucleus (blue) and F-actin (white) by phalloidin on a 2500 μm2 micro-pattern. Orientation analysis of actin filament on 2500 μm2 island. Mean + SEM. Two-sample equal variance two-tailed t-test.

Immunofluorescent image of actin distribution of untreated control and enucleated cell on an island with an area of 2500 μm2.

Scale Bar: 20 μm.

Actin structure pixel number and percentage of cell spreading on 2500 μm2 and 750 μm2 patterns.

(A) Number of pixels of radial fiber (RF) (left) or transverse arc (TA) (right) are compared between the cells on 2500 μm2 (n = 43) and 750 μm2 (n = 31) patterns. (B) Percentage of RF and TA in cells on 2500 μm2 (n = 43) and 750 μm2 (n = 31) patterns. Mean + SEM. Two-sample equal variance two-tailed t-test.

Rotational bias of cell nucleus and actin distribution with actin-related inhibitors on HFF-1.

(A-B) Probability of ACW/CW rotation on 2500 μm2 (A) and 750 μm2 (B) islands with cytoskeleton inhibitors NSC (left), stacked image of actin showing heatmap distribution (middle) and intensity difference (right) of actin comparing cells treated with NSC to untreated control. Data present the mean + SEM. Two-sample, equal variance two-tailed t-test was used. (C-D) Curves of actin intensity from cell center to boundary (left) and differential actin intensity (right) comparing cells treated with NSC to untreated control on (C) 2500 μm2 and (D) 750 μm2 island.

Actin percentage of cell with the control and LatA treatment on 2500 μm2 pattern.

Percentage of RF and TA in cells of untreated control (n = 43, reused data in Figure 3—figure supplement 3) and cell under LatA treatment (n = 42) on 2500 μm2 pattern. Mean + SEM. Two-sample equal variance two-tailed t-test.

Actin percentage of cell with the control and A23 treatment on 750 μm2 pattern.

Percentage of RF and TA in cells of untreated control (n = 31, reused data in Figure 3—figure supplement 3) and cell under A23 treatment (n = 29) on 750 μm2 pattern. Mean + SEM. Two-sample equal variance two-tailed t-test.

Transverse arcs crosslinking protein overexpression.

(A) Transfection of a-actinin-1-GFP showing a-actinin-1 distribution. Scale Bar: 20 μm. (B) Probability of ACW/CW rotation with increscent fluorescence intensity.

Myosin II distribution of Y27632 treatment and control from cell center to the pattern boundary on 2500 μm2 pattern.

Curves of myosin II intensity from cell center to boundary (left) and differential myosin lla intensity (right) comparing Y27632 treatment to the control on a 2500 μm2 pattern.

Immunofluorescence staining of actin and Tpm4 when spread on a 2500 μm2 pattern.

Scale bar: 20 μm.

Myosin IIa distribution of mDia2 silencing and Tpm4 silencing cells compared to the control from cell center to the pattern boundary on 2500 μm2 pattern.

Curves of myosin IIa intensity of cell from cell center to boundary (left) and differential myosin IIa intensity (right) of mdia2 silencing and Tpm4 silencing compared to the control on a 2500 μm2 pattern.

Nucleus rotation with different tropomyosin siRNA transfection.

(A-C) Probability of nucleus rotational bias on 2500 μm2 (upper) and 750 μm2 (lower) with gene silencing in tropomyosin isoform, Tpm1 (A), Tpm2 (B) and Tpm3 (C). Data present the mean + SEM. Two-sample equal variance two-tailed t-test was used.

Actin structure by z-stack of confocal microscopy.

(A) Maximum projection of DAPI-stained nucleus (blue) and phalloidin-stained F-actin (red). (B) Maximum projection of DAPI-stained nucleus (blue) and phalloidin image at 2.35 μm from the bottom of the z-stacked microscopy. The merged image of phalloidin and DAPI shows transverse arcs are present on top of the nucleus with conjunction to a perinuclear actin cap. Scale bar: 20 μm.

Natural spreading size of cell.

(A) Cell culture staining with cell tracker (green) and DAPI (blue) (upper) and the binarized area of cell (lower). Scale bar, 200 μm. (B) The projected area measurement of HFF-1 (2294.55 ± 93.1125 μm2, n = 110), C2C12 (860.8984 ± 26.7324 μm2, n = 175) and hMSCs (4026.123 ± 172.0287 μm2, n = 113). Data presents the mean + SEM.