Bending flexibility of actin filaments during motor-induced sliding

Abstract

Muscle contraction and other forms of cell motility occur as a result of cyclic interactions between myosin molecules and actin filaments. Force generation is generally attributed to ATP-driven structural changes in myosin, whereas a passive role is ascribed to actin. However, some results challenge this view, predicting structural changes in actin during motor activity, e.g., when the actin filaments slide on a myosin-coated surface in vitro. Here, we analyzed statistical properties of the sliding filament paths, allowing us to detect changes of this type. It is interesting to note that evidence for substantial structural changes that led to increased bending flexibility of the filaments was found in phalloidin-stabilized, but not in phalloidin-free, actin filaments. The results are in accordance with the idea that a high-flexibility structural state of actin is a prerequisite for force production, but not the idea that a low-to-high flexibility transition of the actin filament should be an important component of the force-generating step per se. Finally, our data challenge the general view that phalloidin-stabilized filaments behave as native actin filaments in their interaction with myosin. This has important implications, since phalloidin stabilization is a routine procedure in most studies of actomyosin function.

FIGURE 1

Sodium dodecyl sulfate polyacrylamide gel electrophoresis of actin preparation used in the experiments described in this study. Both lanes represent preparation after freezing and thawing: lane 1, actin without NHS-rhodamine; lane 2, actin labeled with NHS-rhodamine. Note the high degree of purity and, in particular, the complete absence of any bands corresponding to tropomyosin and troponin components. Numbers to the left of the gel refer to the molecular mass (in kDa).

FIGURE 2

Filament paths and free fluctuating filaments in A80 solution at 23°C. (A) Paths (light blue) of HMM-propelled actin filaments without phalloidin (Ph⁻) obtained by overlaying 50 subsequent epifluorescence microscopy images. Images acquired at a frame rate of 10 s⁻¹ over a total of 5 s. The filament position at the onset of recording and after 4.8 s is indicated by yellow markers. Red parts of paths correspond to overlap of the leading end of the filament at the onset of recording with the trailing end of the filament after 4.8 s, which occurred for very long filaments. In addition, some positions where filament paths cross are color-coded in red. (Inset) Example of a grayscale image of Ph⁻ filaments (labeled with NHS-rhodamine) in solution, used for determination of (B) The same type of image as in A, but for Ph⁺ filaments (labeled with Alexa-488 phalloidin). Scale bar, 10 μm.

FIGURE 3

Persistence lengths () of actin filaments in solution. (A) Actin filaments with (shaded squares) or without (solid circles) phalloidin. Ph⁺ filaments were stabilized with APh and Ph⁻ filaments were fluorescence-labeled with NHS-rhodamine. The temperature was 23 ± 1.2°C, and there was no MgATP in the observation solution. The numerical value of was obtained by fitting Eq. 3 to experimental data (total number of data points in parentheses) from two to four flow cells and 82–238 filaments for each experimental condition. (B) Cosine correlation functions for actin filaments in solution (A80). Individual data points (from 57–179 filaments) are given with error bars (mean ± SE). Data for Ph⁻ filaments (solid circles) and for Ph⁺ filaments labeled with APh (shaded) or with rPh + NHS-rhodamine (open squares). Equation 3 was fitted to the data (solid lines) by nonlinear regression and shown with 95% confidence bands (dashed lines). (C) Numerical values of for Ph⁻ filaments (solid) in the absence (four flow cells) and presence (three flow cells) of MgATP. Data are also given for Ph⁺ filaments labeled with 1), rPh (open bars) in the absence (three flow cells) and presence (four flow cells) of MgATP; and 2), APh (shaded) in the absence and presence of MgATP (three flow cells for each case). The number of data points used for fitting of Eq. 3 is given in parentheses. Error bars in panels A and C represent 95% confidence intervals for obtained in the nonlinear regression analysis.

FIGURE 4

Sliding velocity of HMM-propelled actin filaments labeled with Alexa-488 phalloidin (shaded squares) or NHS-rhodamine (solid circles). Error bars represent 95% confidence intervals. (A) Experiments were performed in solutions of different ionic strength (40–130 mM) in the absence (A40, A80) and presence of methylcellulose (A80MC, A130MC), at 23 ± 1.2°C (solid symbols) or 29 ± 0.7°C (open symbols). Data for each condition were obtained from 17–138 filament paths using three to six flow cells, except in the cases of A40 data at high temperature (one flow cell) and MC data at low temperature (two to three flow cells). Error bars represent 95% confidence intervals. (B) Velocity data obtained in A80 solution (2–4 flow cells), comparing the effects of phalloidin stabilization with APh and rPh. Numbers in parentheses represent the number of filament paths analyzed.

FIGURE 5

Persistence lengths of actin filament paths. Ph⁺ and Ph⁻ filaments are represented by shaded squares and solid circles, respectively. (A) L_P data at 23 ± 1.2°C. (B) L_P data at 29 ± 0.7°C. Data in A and B were derived by fitting of Eq. 3 to the cosine correlation function. Error bars represent 95% confidence intervals. Numbers in parentheses indicate the total number of data points from 27–155 filaments (each tracked for one or two independent trajectories of 20 μm length) for each fit (condition). Experiments were the same as in Fig. 4. (C) Cosine correlation functions for HMM-propelled filament paths in A80 solution for Ph⁻ (solid circles) and Ph⁺ filaments (shaded squares, APh; open squares, rPh). Lines represent fits of Eq. 4 to the data by nonlinear regression. Plotted average data (mean ± SE) are based on a total of 1400–3240 individual data points. (D) L_P data for the Ph⁻ filaments in A and B (shaded symbols and shaded lines; A80MC data omitted) superimposed on data (dashed lines; from Fig. 3 A) for both Ph⁺ (shaded) and Ph⁻ (solid) filaments. (E) L_P data for the Ph⁺ filaments in A and B (shaded symbols and shaded lines) superimposed on data (dashed lines; from Fig. 3 A) for both Ph⁺ (shaded) and Ph⁻ (solid) filaments. (F) Kinetic scheme, developed from Kozuka et al. (14), representing different actin filament structural states in inactive (I and IPh) and active (A, APh, MA, and MAPh) conformations. The binding of phalloidin is assumed to be irreversible on the experimental timescale. Otherwise, equilibria are assumed to exhibit rapid kinetics (14) (equilibrium constants K_IA, and K_AM). For details, see text.