ADF/cofilin regulates actomyosin assembly through competitive inhibition of myosin II binding to F-actin

Abstract

The contractile actin cortex is important for diverse fundamental cell processes, but little is known about how the assembly of F-actin and myosin II motors is regulated. We report that depletion of actin depolymerizing factor (ADF)/cofilin proteins in human cells causes increased contractile cortical actomyosin assembly. Remarkably, our data reveal that the major cellular defects resulting from ADF/cofilin depletion, including cortical F-actin accumulation, were largely due to excessive myosin II activity. We identify that ADF/cofilins from unicellular organisms to humans share a conserved activity to inhibit myosin II binding to F-actin, indicating a mechanistic rationale for our cellular results. Our study establishes an essential requirement for ADF/cofilin proteins in the control of normal cortical contractility and in processes such as mitotic karyokinesis. We propose that ADF/cofilin proteins are necessary for controlling actomyosin assembly and intracellular contractile force generation, a function of equal physiological importance to their established roles in mediating F-actin turnover.

Figure 1. ADF/cofilin silencing results in distinct cell morphological defects

HeLa cells treated with indicated siRNAs were examined at 72h post treatment. (A) DIC images of live cells. (B) Quantification of cells with plasma membrane blebs. Values are mean ± sd, n > 200 cells/treatment from at least 3 experiments. (C) Representative blots of total ADF/cofilin, cofilin (cofilin specific antibody), and GAPDH (loading control) following siRNA treatment. (D) Representative confocal sections (apical, mid) and maximum intensity projections of F-actin fluorescence staining. (E) Apical confocal sections depicting abnormal F-actin accumulation at the base of rosettes of small membrane blebs (left panels) and z-projections showing similar F-actin accumulation at the base of large membrane protrusions (right panels) in fixed ADF/cofilin depleted cells. Bars, 10 μm.

Figure 2. Formation of persistent blebs and aberrant contractile actomyosin structures in cells depleted of ADF and cofilin

(A) Representative time series showing recruitment of actin to the cortex of a dynamic bleb (arrowhead). (B) Persistent blebs fail to assemble a new actin cortex (arrowhead, GFP-actin upper panels; DIC overlay, lower panels). Arrow shows emergence of a new persistent bleb. Bars, 5 μm (A, B). (C) Kymograph of the bleb in (A, arrowhead) along line shown (A; t = 11s). (D) Kymograph of bleb activity for the bleb in (B, arrow) along line shown (B; t= 88s). (E) Time-lapse series of GFP-actin (upper panels; green, bottom panels) overlaid to DIC images (bottom panels) in HeLa cells depleted of both ADF and cofilin shows accumulation of actin at blebs which fail to retract. Actin accumulates at blebs after their initiation (arrow). Insets show magnified images of an actin aggregate associated with a bleb (arrowhead) which migrates along actin filaments towards the cell body (top of image). Bar 10 μm. (F) Time series of GFP-actin shows formation of a contractile actin ring structure (arrow) at the base of a bleb (see Movie S1). Insets show the same field for actin (pseudocolored red) overlain to a DIC image of the bleb (pseudocolored green). (G) Quantification of persistent bleb formation. Values are mean ± sd, n > 200 cells/treatment from at least 3 experiments. (H) Kymograph showing motility of the bleb actin aggregate depicted in (E, insets). Mean velocity ± sd, n=9. (I and J) Accumulation of GFP-MyoIIA (I, arrowheads) and GFP-MyoIIB (J, arrowheads) at contractile structures at the base of blebs in COF+ADF depleted cells. Bars, 5 μm.

Figure 3. Increased actomyosin assembly and cortical contractility in ADF/cofilin depleted cells

(A) Representative confocal immunofluorescence images show elevated cortical myosin II recruitment following cofilin siRNA treatment. Bars, 10 μm. (B and C) Intensity linescans (B, white lines in merged images) of myosin foci, detected as distinct peaks in plots (illustrated in C, upper panels) along actin filaments plotted in (C). Bars, 1 μm. (D) Quantification of p-MLC intensity at cortical cell regions. Values are mean ± sd, n ≥18 cells/treatment. (E) Illustrative time series of F-actin fluorescence intensity (RFP-Lifeact) along a contractile cortical F-actin bundle (line in image, t = 0s) graphed (right) over time following treatment of control cells with latrunculin B. Arrows in images and graph depict the position at which fiber fissure occurs (t = 390s), magnified in insets. Note the decrease in actin intensity coupled to increased intensity on both sides of the point of fissure, pre- and post-breakage. Bar, 10 μm. (F) Retraction kinetics of fiber segments to the left and right of the point of fissure for the fiber depicted in (E). (G) Retraction kinetics of cortical fibers with spontaneous breakage following latrunculin B treatment of control and cofilin depleted cells. Values are mean ± sd, n= 13 control; n = 12 COF siRNA. (H) Graph of fiber retraction kinetics following fissure in a control or COF siRNA treated cell. Solid lines represent a fit to the Kelvin-Voigt viscoelastic model, represented by a viscous dashpot in parallel with an elastic spring. D is the measured retraction distance at time, t, D₀ is the derived asymptotic retraction distance and the derived time constant, τ, is the ratio of viscosity to the elastic modulus.

Figure 4. Rescue of cytoskeletal and bleb defects in COF+ADF depleted cells by ectopic ADF/cofilin expression or myosin II silencing

(A) Confocal fluorescence images of cells depleted of both ADF and cofilin and ectopically expressing either wild type Xenopus ADF/cofilin (GFP-XAC wt) or a constitutively inactive cofilin (GFP-XAC E3). (B) Quantification of the rescue of bleb phenotypes by adenoviral mediated XAC or transfected hCOF siRNA resistant plasmid expression. Values are mean ± sd, n > 400 from at least 3 experiments. (C) Effects of co-depletion of myosin II isoforms and ADF/cofilin on cell morphology (live cell DIC images, upper panels) and F-actin organization (fluorescence images, lower panels). (D) Cells are more elongated following myoIIA silencing. Box lines are 25%, median and 75%, dots are mean values and whiskers show full data range. (E) Increased cell area following myoIIB silencing. Bars are mean values. (F) Quantification of bleb phenotypes following co-depletion of ADF/cofilin and myosin II. Values are mean ± sd, n ≥ 700 cells/treatment from at least 3 experiments. Bars, 10 μm (A, C).

Figure 5. Blebbing and abnormal cortical actin accumulation in ADF/cofilin depleted cells is dependent upon myosin II activity

(A and B) DIC time-lapse series showing inhibition of dynamic blebbing (A, arrowhead) in cofilin depleted cells and retraction of persistent blebs (B, arrowhead) in cells depleted of both ADF and cofilin, following blebbistatin treatment. Time is relative to blebbistatin addition. See also Movie S2. (C) Quantification of blebbing prior to and after blebbistatin treatment of cofilin and ADF+COF depleted cells. Values are mean ± sd, n = 123, from 3 experiments. (D) Confocal time-lapse series of mCherry-actin (upper panels; and pseudocolored green in DIC overlays, lower panels; see also Movie S3) in ADF/cofilin depleted cells pre and post drug treatment. Arrowhead depicts a persistent bleb which retracts following blebbistatin addition. (E) Kymograph of bleb depicted by arrowhead (D). (F) Graph of actin intensities along lines shown in (D, t = −14m, bottom panel) during retraction of the depicted bleb. Bars, 10 μm.

Figure 6. ADF/cofilin inhibits myosin S1 binding to F-actin

(A) Structure of the ADF-homology domain (ADF-H, blue) bound to G-actin (ribbon structure) (Paavilainen et al., 2008). ADF-H binds to a hydrophobic pocket between subdomains 1 and 3 of actin. Putative sites of interaction on actin for ADF/cofilin are colored green, those for myosin S1 are in red, and overlapping sites yellow. (B) Model structure of myosin S1 (heavy chain, violet; light chain, pink) interaction with F-actin ( 2 monomers, grey) (Lorenz and Holmes, 2010). Putative positions of interaction at the primary actin binding site (monomer A1) for ADF/cofilin are green and secondary sites of interaction (monomer A2), red. Overlapping sites for ADF/cofilin and myosin S1 are yellow (monomer A1) and cyan (monomer A2). (C) Coomassie-Blue stained gels of supernatants (s) and pellets (p) from a cosedimentation assay where myosin subfragment-1 (S1) was preincubated with F-actin in the presence of 120 μM ATP prior to addition of Xenopus cofilin (XAC), as indicated. S1-HC, myosin S1 heavy chain, lower band corresponds to a common ~ 76 kDa proteolytic product; S1-LC, myosin light chain A1. (D) Quantification of actin bound myosin S1 following simultaneous exposure of both myosin S1 and XAC to F-actin in the presence or absence of ATP. Results are mean of 3 experiments ± sd. (E and F) Representative gels (E) and quantification (F) of actin bound myosin S1 in the presence of ADF/cofilins from human, chick, xenopus, drosophila (twinstar), acanthamoeba (actophorin), starfish (depactin) and yeast. Graphs are mean of 3 experiments.

Figure 7. ADF/cofilin depletion results in mitotic karyokinesis defects

(A and B) Fluorescence images of cofilin depleted cells at telophase (A) and late cytokinesis (B), labeled for F-actin (red) and DNA (blue). Boxed region in (B) is expanded in inset and shows F-actin (red)/microtubule (green) overlay. (C) Quantification of karyokinesis defects. Values are mean ± sd, n > 170 from 3 experiments. (D and E) DIC images of blebbing at cytokinesis in control cells (D; see Movie S4) and of dynamic blebbing (E, arrowhead in time series) during metaphase in ADF depleted cells. (F) Quantification of cytokinetic bleb and spindle displacement phenotypes. Values are mean ± sd, n ≥ 24 from at least 3 experiments. (G and H) Time-lapse series showing weak spindle displacement behavior (G, DIC images of ADF depleted cells) and strong spindle oscillation behavior (H, GFP-GPI fluorescence images of cofilin depleted cells; see also Movie S5). Arrowhead in (G) shows retraction of a large bleb, white and green arrows show the initial and displaced chromosome positions respectively. Arrowheads in (H) mark chromosome positions. Scale bar for all images, 10μm. (I–L) Bleb height and diameter measurements during cytokinesis. (J) Values are mean ± sd (Control n=18 cells, 110 blebs; ADF n=22 cells, 241 blebs; COF n= 14 cells, 197 blebs). *, P < 0.001 relative to control, t-test. Box lines are 25%, median and 75%, whiskers show 10–90% range and circles are outliers (K, L). (M) Model depicting two physiologically significant functions for ADF/cofilin: 1) F-actin turnover through filament severing and/or depolymeriztion, 2) Actomyosin assembly through inhibition of myosin II binding to F-actin.