Capping protein increases the rate of actin-based motility by promoting filament nucleation by the Arp2/3 complex

Abstract

Capping protein (CP) is an integral component of Arp2/3-nucleated actin networks that drive amoeboid motility. Increasing the concentration of capping protein, which caps barbed ends of actin filaments and prevents elongation, increases the rate of actin-based motility in vivo and in vitro. We studied the synergy between CP and Arp2/3 using an in vitro actin-based motility system reconstituted from purified proteins. We find that capping protein increases the rate of motility by promoting more frequent filament nucleation by the Arp2/3 complex and not by increasing the rate of filament elongation as previously suggested. One consequence of this coupling between capping and nucleation is that, while the rate of motility depends strongly on the concentration of CP and Arp2/3, the net rate of actin assembly is insensitive to changes in either factor. By reorganizing their architecture, dendritic actin networks harness the same assembly kinetics to drive different rates of motility.

Figure 1. Effects of Arp2/3 and CP on steady-state motility

A. Panels: Four stages of motility—shell building, symmetry breaking, transient burst, and steady-state (clockwise, starting from top-left.) Arrowheads track individual beads through the panels. Graph: Total distance traveled by beads in the panels. Curves and speed figures are color-matched to the arrowheads. Steady-state speeds are the slopes of data between the dashed lines (30 to 180 minutes). Conditions: 5 µm ActA-coated beads mixed with 7.5 µM actin (5% TMR-labeled), 5 µM cofilin, 3 µM profilin, 120 nM Arp2/3, 168 nM CP. B–F. Representative actin comet tails assembled on 220 nm ActA-coated beads in the presence of various concentrations of Arp2/3 and CP as indicated. Images were acquired approximately 30 minutes after mixing. Brightfield bead images and fluorescent comet tails were digitally merged for presentation. Brightness and contrast of the fluorescent images were set to the same values. Conditions: 7.5 µM actin (5% TMR-labeled), 4 µM cofilin, and 3 µM profilin. C. Rate of motility decreases with [Arp2/3] and increases with [CP]. Each data point is the slope of the line fit to a time series of length measurements. Error bars are slopes of fits to the standard deviation of each time series (average n=140). D. Actin density in the comet tails increases with [Arp2/3] and decreases with [CP]. Fluorescence intensities along single-pixel spines of actin comet tails were measured. Each data point is the intensity averaged over the first micron length of the tails. Error bars are standard deviations for 20 comet tails. E. Rate of actin network assembly does not change with [Arp2/3] or [CP]. Rate of fluorescent actin incorporation is the product of tail growth rate (C) and fluorescence intensity (D). Error bars mark the product of the upper and lower error limits of the two terms. F. Soluble actin concentration at steady-state does not change with [Arp2/3] or [CP]. Error bars mark the standard error for 3 independent measurements. G. Rate of motility in brain cytosolic extract decreases with [Arp2/3] and increases with [CP]. Rates of tail growth on 220 nm ActA-coated beads were measured as in (C); see text for details. Each data point is the slope of the line fit to a time series of length measurements. Error bars are slopes of fits to the standard deviation of each time series (average n=77). Conditions: Brain cytosolic extract supplemented with 2 µM actin (10% TMR-labeled) and Arp2/3 and CP as indicated.

Figure 2. CP increases the rate of Arp2/3 nucleation during steady-state motility

A. Representative actin comet tails assembled on 220 nm ActA-coated beads in the presence of various concentrations of Arp2/3 (10% Alexa488-labeled) and CP as indicated. Images were acquired approximately 30 minutes after mixing. Tails are framed with the beads (not shown) to the left of each panel, except the lower left panel (75 nM Arp2/3, 42 nM CP) where the bead would be top-center. This condition, as well as lower [CP] than that shown for the higher two Arp2/3 concentrations, produces twin tails on every bead. The red (i.e. actin) channel of each image was set to 90% saturation; this value was scaled by the same factor to determine the green (i.e. Arp2/3) channel display limits for all images. Conditions: 7.5 µM actin (5% TMR-labeled), 4 µM cofilin, and 3 µM profilin. B. Fluorescence intensities along single-pixel spines of actin comet tails were measured. Each data point is the ratio of Alexa488-Arp2/3 to TMR-actin intensity averaged over the first micron length of the tails. Error bars are standard deviations (average n=27).

Figure 3. CP is necessary for shell growth and symmetry breaking

A. Actin assembly is restricted to the bead surface during shell growth. We mixed 5 µm ActA-coated beads with 2 µM actin (5% Alexa488-labeled), 50 nM Arp2/3, and 140 nM CP. After 30 seconds, the reaction was diluted ten-fold into a solution of identical protein composition containing 5% TMR-labeled actin. B. Actin assembly is not restricted to the bead surface in the absence of CP. Top: Selected frames from a time-lapse series of actin halo growth imaged using fluorescent speckle microscopy. Conditions: 5.8 µm ActA-coated beads mixed with 2 µM actin (1/16,000 TMR-labeled) and 80 nM Arp2/3. Bottom: Actin speckles are stationary and new speckles appear farther away from the bead surface. 20 frames of the series shown above, representing 6:20 minutes of elapsed time, were used to generate this maximum intensity projection. Images were filtered to highlight the speckles. Time is represented by changing hues, starting with yellow at early time points, progressing through orange to red. C. CP directs new actin assembly to the bead surface. Top: New actin assembly was followed using fluorescence pulse-chase as in (A). 5 µm ActA-coated beads were mixed with 2 µM actin (first label 5% Alexa488, second label 5% TMR) , 80 nM Arp2/3, and CP as indicated in top right corner of each panel. Images were acquired 10–15 minutes after mixing. Bottom: Intensity profiles of the green and red actin fluorescence along the 4 µm lines are plotted, starting from the endpoints closest to the beads. Each profile is normalized to its maximum value in the original image. D. The concentration of CP required for symmetry breaking varies directly with the Arp2/3 concentration. Panel shows representative actin structures assembled on 5 µm ActA-coated beads in the presence of 2 µM actin (5% TMR-labeled), and Arp2/3 and CP as indicated. Dashed line marks the symmetry breaking threshold. Images were acquired 10–15 minutes after mixing.

Figure 4. Effects of Arp2/3 and CP on shell growth

A–F. Reactions were Latrunculin B/phalloidin arrested 60 seconds after mixing, fractionated and analyzed as described. Conditions: 5.8 µm ActA-coated beads mixed with 3 µM actin (5% TMR-labeled), and Arp2/3 and CP as indicated. A. Representative shells after arrest, fractionation, and washing. B. Shell thickness decreases with [Arp2/3] and increases with [CP]. Error bars represent standard deviation (average n=17). C. Soluble actin concentration at the time of arrest decreases with increasing [CP]. D. Total amount of actin in the shells does not vary with [Arp2/3] or [CP]. E. Average actin density in the shells increases with [Arp2/3] and decreases with [CP]. Density was calculated by dividing the actin mass in (D) by the volume of the shells as derived from (B). F. Total number filaments in the shells is independent of [Arp2/3] but increases with [CP].

Figure 5. The kinetics of shell growth is independent of the kinetics of actin assembly

A–F. Aliquots from each reaction were Latrunculin B/phalloidin arrested 15, 30, 45, and 60 seconds after mixing, fractionated, and analyzed as described. Conditions: 5.8 µm ActA-coated beads mixed with 3 µM actin (5% TMR-labeled), and Arp2/3 and CP as indicated. A. Representative shells after arrest, fractionation, and washing. B. Rate of shell growth decreases with [Arp2/3] and increases with [CP]. Error bars represent standard deviation (average n=27). C. Rate of actin assembly is independent of [Arp2/3] and [CP]. D. Rate of filament nucleation is independent of [Arp2/3] but increases with [CP]. Nucleation rate per square micron of bead surface was derived from the slopes of the time courses. E. Two alternate mechanisms for increasing the density of a dendritic actin network imply different actin assembly kinetics. See text for details.

Figure 6. Arp2/3 slows down shell growth by tethering the actin network to the bead surface

A. Dephosphorylated Arp2/3 complex slows down shell growth as effectively as the nucleation-competent complex. Conditions: 5.8 µm ActA-coated beads mixed with 3 µM actin (5% TMR-labeled), 112 nM CP, and Arp2/3 as indicated. Actin assembly was arrested 60 seconds after mixing. B. Shell sizes for the conditions in (A). Error bars represent standard deviation (average n=20). C. Monomer Gating: NPF-bound actin monomers can either associate with nearby barbed ends (1) and contribute to filament elongation, or they can participate in dendritic nucleation (2). Elongation is kinetically favored over nucleation, which requires the stable association of four protein partners—the mother filament, Arp/3, NPF, and the actin monomer—through a slow activation step. CP, by eliminating the competition from barbed ends, gates the biochemical path of actin monomers toward nucleation.