Figure 2. Quantification of actin reorganization by live-cell confocal polarization microscopy.

(A) Fluorescence polarization microscopy of stress fibers in live interphase hTERT-RPE-1 cells labeled with SiR-actin using the LC-PolScope. Color saturation indicates degree of fluorophore alignment (anisotropy), hue indicates mean orientation of fluorescence dipoles as shown in upper left corner. (B) Fluorescence polarization microscopy of live hTERT-RPE-1 cells stably expressing H2B-mRFP (magenta), stained with SiR-actin (overlay of red and green) at representative stages during cytokinesis. Images were acquired with horizontal (red) and vertical (green) linear polarizers in the emission beam path. Untreated cells (upper panel) and cells treated with 50 µm para-nitroblebbistatin (lower panel) are shown. (C) Schematic drawing of the circular fit procedure during cell division. Circles (magenta) were fitted to the center of the equator (blue). (D) Curvature of the equatorial cell cortex in a representative cell. Dots indicate individual time points. Time = 0 s at anaphase onset. (E) Quantification of equatorial cortex curvature in 18 dividing cells. Line indicates mean, gray area s.d.. The first time point with significantly changed curvature was detected at 90 s after anaphase onset by two-tailed t-test; ***p<0.001. (F) Quantification of SiR-actin normalized emission ratio in the 18 dividing cells used for the analysis in (E) (line indicates mean, blue and red areas s.d. **p<0.01 by two-tailed t-test (G) Quantification of SiR-actin normalized emission ratio in dividing cells treated with 50 µm para-nitroblebbistatin. Time = 0 s at anaphase onset (lines indicate median, shaded areas indicate s.d.; equatorial measurement: n = 15 cells, polar: n = 11 cells). (D-G) The green area indicates time points where the absolute value of cortical curvature was equal or below that of the metaphase cell. Only these time points were used for further interpretation of fluorescence anisotropy measurements, to avoid potential artifacts by geometry effects. Scale bars = 10 µm.

Figure 2—figure supplement 1. Polarization microscopy with linear polarizers in the emission beam path.

(A) Confocal fluorescence polarization microscope setup used for live-cell imaging. A linearly polarized excitation laser that was aligned with the X-axis of the optical table, and two perpendicularly oriented polarization filters, horizontal (H) and vertical (V) relative to the optical table, respectively, were used. (B) Cleavage furrow ingression measured in hTERT-RPE-1 cells expressing MyrPalm-mEGFP to visualize the plasma membrane (in absence of SiR-actin), line indicates mean, shaded area indicates s.d. of 8 cells. Anaphase onset = 0 s. (C) Cleavage furrow ingression measured in hTERT-RPE-1 cells stained with SiR-actin, line indicates mean, shaded area indicates s.d. of 18 cells. Anaphase onset = 0 s. (D–F) Analysis of geometry effect caused by linear polarized excitation light (see Materials and methods) on the ratio image. (D) Confocal image of a live metaphase hTERT-RPE-1 cell stained with SiR-actin, stably expressing H2B-mRFP. Yellow dashed line indicates line scan measurement region and arrow indicates measurement direction. (E) Quantification of (D), ratio between horizontal (H) and vertical (V) linear emission polarization filter was calculated for each pixel and plotted against the respective tangential angle. Because fluorescent dipoles are excited with linear polarized laser light, the ratio between horizontal and vertical emission channels of a randomly oriented network (as a metaphase cell cortex) depends on the angle relative to the excitation light. This effect can be described by a squared sine regression model (indicated by red curve). (F) Each pixel of the ratio image of the metaphase cortex was divided by the respective value derived from evaluating the regression model developed in (E) using the tangential angle for a given pixel as input parameter. (G) Schematic showing cortex thickness measurement assay. Membrane marker (green), actin marker (magenta), and measurement region (grey line). (H) Representative images of live hTERT-RPE-1 metaphase cell stably expressing MyrPalm-EGFP, actin was stained with SiR-actin. Solid yellow line indicates measurement area. (I) Quantification of (H) by fitting Gaussian functions to the line scans of each color channel as shown in (H), drawn in the direction from the cell interior to the extracellular space. The displacement between the Gaussian maximum positions was used to calculate the half-width of cortical thickness. (J) Quantification of cortex thickness based on the displacement of MyrPalm-EGFP and SiR-actin as shown in (H, I), for metaphase, anaphase and interphase cells. Each dot represents a single cell (metaphase: n = 48, anaphase: n = 50, interphase: n = 71), line indicates median. (**p<0.026 by Mann-Whitney test). Scale bar for (D) is 10 µm and for (H) is 2 µm.

Figure 2—figure supplement 2. Analysis pipeline of live-cell polarization microscopy.

(A) Analysis pipeline for a representative live hTERT-RPE-1 cell stained with SiR-actin recorded with the confocal fluorescence polarization microscope setup. Cells were recorded with horizontal and vertical polarization emission filters. The cell cortex was segmented and furrow and pole midpoints were determined to generate separate measurement regions of each point. (B) To correct for the geometry effects caused by polarized excitation light, the pixel orientation of every pixel along the segmented contour was measured. Pixel orientation along the cortex was inferred by fitting a B-spline to the segmented contour. We then averaged a seven-pixel region perpendicular to the central pixel to capture both channel with a single mask. Yellow rods mark the averaged area, a seven-pixel region perpendicular to the tangential vector of the central pixel of the spline using the segmented cortex as the central pixel (red dot). The angle for each pixel was used to evaluate the squared sine regression model as described in Figure 2—figure supplement 1D–F. The regression model was developed for each cell independently and based on multiple pre-anaphase frames. Evaluation of the regression model using actual pixel orientation as input parameters resulted in the expected fluorescence ratio and is shown together with the actual fluorescence ratio measured for each pixel. The observed ratio for every pixel for every furrow/pole measurement region was divided by the expected ratio resulting in the corrected ratio. Dashed line indicates ratio of 1 indicating an isotropic dipole orientation and a random oriented network. Scale bars = 10 µm (A) and 2 µm (B).

Figure 2—figure supplement 3. Cortex organization in para-nitroblebbistatin-treated anaphase cells and lateral distribution of actin and myosin in untreated cells.

(A) Diameter at equatorial position measured in hTERT-RPE-1 cells expressing H2B-mRFP 

and stained with SiR-actin, in presence of 50 µM para-nitroblebbistatin, an inhibitor of myosin II. Line indicates mean, shaded area indicates s.d. of 10 cells. Anaphase onset = 0 s. (B, C) Quantification of SiR-actin fluorescence at cell poles and equator as illustrated for untreated and para-nitroblebbistatin treated cell in Figure 2B. Anaphase onset = 0 s (D) Confocal image of live hTERT-RPE-1 cell expressing non-muscle myosin IIc-EGFP (MLC-12B-EGFP), stained with SiR-actin. Dashed line indicates measurement region for line profiles, arrowhead indicates cell equator position. (E) Quantification of lateral distribution of MLC-12B-EGFP and actin filaments at late furrow ingression stage as in (D); lines represent mean fluorescence in cortical line profiles centred to the cell equator (position = 0 μm), shaded areas represent s.d.; n = 37 cells. Scale bars = 10 µm.