Regulated actin cytoskeleton assembly at filopodium tips controls their extension and retraction

Abstract

The extension and retraction of filopodia in response to extracellular cues is thought to be an important initial step that determines the direction of growth cone advance. We sought to understand how the dynamic behavior of the actin cytoskeleton is regulated to produce extension or retraction. By observing the movement of fiduciary marks on actin filaments in growth cones of a neuroblastoma cell line, we found that filopodium extension and retraction are governed by a balance between the rate of actin cytoskeleton assembly at the tip and retrograde flow. Both assembly and flow rate can vary with time in a single filopodium and between filopodia in a single growth cone. Regulation of assembly rate is the dominant factor in controlling filopodia behavior in our system.

Figure 1

Imaging actin dynamics in filopodia. A and B show a caged Q-rhodamine actin photoactivation experiment. C and D show a GFP-actin photobleaching experiment. (A) Three lower magnification views of a section of a growth cone are presented. From left to right these are the phase-contrast image, an image of the fluorescent mark, and a combined image in which the fluorescent mark appears as a red overlay on the phase image. B shows a detailed view of a photoactivation mark made on a filopodium (box in A) at successive time points. (C) Image of a whole growth cone containing GFP-actin is presented as it appears by phase-contrast (left) and epifluorescence (right). D shows a detailed view of a photobleach mark made on one filopodium (gray box in C). Arrows indicate the location of a photobleach mark near the filopodium tip. Bars: (A and C), 10 μm; (B and D) 5 μm.

Figure 2

Example of regulation by cytoskeleton assembly rate. Q rhodamine photoactivation experiment. (Left) Phase-contrast; (middle) epi-fluorescence; and (right) combined image. The black line indicates the mark and the purple line the filopodium tip. Arrows denote a filopodium that retracts while the one above it extends. B, the mark, tip and mark to tip distances for the longest filopodium in A plotted as a function of time. Note that switching of this filopodium from extending to retracting is coincident with a change in the cytoskeleton assembly rate from ∼1.8 to ∼0 μm/min with flow approximate constant. Bar, 5 μm.

Figure 2

Example of regulation by cytoskeleton assembly rate. Q rhodamine photoactivation experiment. (Left) Phase-contrast; (middle) epi-fluorescence; and (right) combined image. The black line indicates the mark and the purple line the filopodium tip. Arrows denote a filopodium that retracts while the one above it extends. B, the mark, tip and mark to tip distances for the longest filopodium in A plotted as a function of time. Note that switching of this filopodium from extending to retracting is coincident with a change in the cytoskeleton assembly rate from ∼1.8 to ∼0 μm/min with flow approximate constant. Bar, 5 μm.

Figure 3

Example of regulation by retrograde flow. This figure follows the layout of Fig. 2. Note that the change from slow retraction to fast retraction at 1.5 min in this example is coincident with an increase in flow rate while assembly remains approximately constant.

Figure 3

Example of regulation by retrograde flow. This figure follows the layout of Fig. 2. Note that the change from slow retraction to fast retraction at 1.5 min in this example is coincident with an increase in flow rate while assembly remains approximately constant.

Figure 4

Example of simultaneous regulation of assembly and flow rates. This figure follows the layout of Fig. 2. Note that the filopodium tip remains relatively stationary while both retrograde flow and actin assembly rate increase at 1.5 min.

Figure 4

Example of simultaneous regulation of assembly and flow rates. This figure follows the layout of Fig. 2. Note that the filopodium tip remains relatively stationary while both retrograde flow and actin assembly rate increase at 1.5 min.

Figure 5

Average assembly and flow rates grouped by rates of filopodium movement. Tip movement, assembly, and flow rates were determined for individual bouts of filopodia movement for all observations, and arbitrarily grouped on the basis of tip movement as follows. Rapidly retracting: retracting at greater than −1 μm/min. Slowly retracting: between −1 and −0.3 μm/min. Stationary: between −0.3 and +0.3 μm/min. Slowly extending: between +0.3 and +1 μm/min. Rapidly extending: extending at any rate above +1 μm/min. Error bars are SD of the mean values in each group. For analysis, the rates of tip movement, assembly, and flow were measured between each time point (every 30 s) for each filopodium and grouped. Different periods of movement of a single filopodium might contribute to several different categories. The plot pools 587 individual time point measurements taken from 75 separate filopodia from 18 different sequences. 14 were Q-rhodamine photoactivation experiments and 4 were GFP-photobleaching experiments. We were unable to detect differences between these two groups of data examined separately.

Figure 6

Actin dynamics vary between filopodia in a single growth cone. By following photoactivation marks on multiple filopodia in single growth cones we measured spatial variation in dynamics parameters. (A) Phase-contrast (left), fluorescence (middle), and combined (right) images. The outline of the growth cone at t = 0 are drawn on the combined image at 8–45 min. Photoactivation marks were made at t = − 0.3 min and again at t = 15 min. Arrows in t = 45 min indicate the remaining zones of marked actin from the first (1) and second (2) marking. Note that marks made near the top of the growth cone flow backward faster than marks made near the bottom. Arrowheads at 10 min indicate a zone of slow relatively slow flow. A typical feature of long sequences, apparent here, is the coalescing of filopodium actin bundles as they flow toward the neck of the growth cone. At t = 45 min, actin filaments marked at the tip are still visible and have been transported to the neck of the growth cone where they undergo lateral compression. From this sequence, we tracked mark and tip movements for 13 individual filopodia (indicated with numbers at t = 0, right column) over the course of 10 min and determined average assembly and flow rates by linear regression. (B) Average tip movement (black bars), assembly (white bars), and flow (gray bars) rates for the individual filopodia numbered as in A. Error bars indicate the SD of the regression coefficients, a measure dominated by temporal variation in dynamics. Bar, 5 μM.

Figure 6

Actin dynamics vary between filopodia in a single growth cone. By following photoactivation marks on multiple filopodia in single growth cones we measured spatial variation in dynamics parameters. (A) Phase-contrast (left), fluorescence (middle), and combined (right) images. The outline of the growth cone at t = 0 are drawn on the combined image at 8–45 min. Photoactivation marks were made at t = − 0.3 min and again at t = 15 min. Arrows in t = 45 min indicate the remaining zones of marked actin from the first (1) and second (2) marking. Note that marks made near the top of the growth cone flow backward faster than marks made near the bottom. Arrowheads at 10 min indicate a zone of slow relatively slow flow. A typical feature of long sequences, apparent here, is the coalescing of filopodium actin bundles as they flow toward the neck of the growth cone. At t = 45 min, actin filaments marked at the tip are still visible and have been transported to the neck of the growth cone where they undergo lateral compression. From this sequence, we tracked mark and tip movements for 13 individual filopodia (indicated with numbers at t = 0, right column) over the course of 10 min and determined average assembly and flow rates by linear regression. (B) Average tip movement (black bars), assembly (white bars), and flow (gray bars) rates for the individual filopodia numbered as in A. Error bars indicate the SD of the regression coefficients, a measure dominated by temporal variation in dynamics. Bar, 5 μM.

Figure 7

Photoactivation of a lamellipodium. (A) Phase-contrast and (B) fluorescence. NG108 cells often formed lamellipodia, characterized by lack of long filopodia and ruffling behavior. Photoactivation marks on lamellipodia disappeared quickly, indicating fast actin filament turnover compared with filopodia.

Figure 8

Model for regulation of actin dynamics in filopodia. Chevrons indicate actin monomers, with filaments oriented barbed end distal. Assembly is driven by actin polymerization at the tip, presumably regulated by factors that influence polymerization rate. Flow is driven by motors pulling on the filaments, though the precise identity of the motors, and what they pull relative to, is unknown. Flow is presumably regulated by direct regulation of the motors and/or coupling of actin filaments to the substrate via adhesion systems. Assembly and flow can be regulated independently: temporally within a single filopodium and spatially between filopodia within a growth cone. In our data set, assembly rate varied frequently while the flow rate was more constant (Fig. 5). Thus, the direction and rate of filopodium tip movement was governed primarily by assembly rate.