Myosin II Filament Dynamics in Actin Networks Revealed with Interferometric Scattering Microscopy

Abstract

The plasma membrane and the underlying cytoskeletal cortex constitute active platforms for a variety of cellular processes. Recent work has shown that the remodeling acto-myosin network modifies local membrane organization, but the molecular details are only partly understood because of difficulties with experimentally accessing the relevant time and length scales. Here, we use interferometric scattering microscopy to investigate a minimal acto-myosin network linked to a supported lipid bilayer membrane. Using the magnitude of the interferometric contrast, which is proportional to molecular mass, and fast acquisition rates, we detect and image individual membrane-attached actin filaments diffusing within the acto-myosin network and follow individual myosin II filament dynamics. We quantify myosin II filament dwell times and processivity as functions of ATP concentration, providing experimental evidence for the predicted ensemble behavior of myosin head domains. Our results show how decreasing ATP concentrations lead to both increasing dwell times of individual myosin II filaments and a global change from a remodeling to a contractile state of the acto-myosin network.

Figure 1

Experimental setup. (A) Shown is a diagram of the in vitro system consisting of a supported lipid bilayer (orange), actin-membrane linker protein decahistidine-KCK-ezrin actin binding domain (HKE; dark gray and red), actin filaments (gray), and muscle myosin II filaments (blue); arrows indicate principle of iSCAT microscopy. (B and C) Shown are example images of actin filaments (B) and myosin II filaments (C), both recorded with a 445-nm laser iSCAT system. Scale bars, 2 μm. (D) Shown is a histogram depicting the interferometric contrast distribution along actin filaments (gray, N_measure = 562 measurements along N_fil = 12 filaments) and myosin II filaments (blue, N_measure = 303 measurements along N_fil = 14 filaments); solid lines depict fits of a normal distribution with mean ± standard deviation noted above them. (E) Shown is an example of tracking a single actin filament (red dashed line, l = 0.4 μm) inside an actin network imaged over 137 s (from dark blue to cyan: 0–137 s; Video S3). Scale bars, 1 μm. (F) Shown is the corresponding MSD of the filament’s center of mass (black) and the components parallel (green) and perpendicular (red) to the actin filament orientation. (G) Shown is the MSD of the center of mass of a long actin filament (l = 4 μm, Video S4) (black) and its components parallel (green) and perpendicular (red) to the actin filament orientation, indicating confinement at timescales >10 s. (H) Shown is a box plot comparing the diffusive behavior (characterized by logarithmic slopes of MSD plots) of short (<1.5 μm, N = 25) and long (>1.5 μm, N = 14) actin filaments for Δt = 5 - 10 s indicating confined diffusion for filaments longer than 1.5 μm due to the surrounding actin meshwork, middle line: median, square: mean, whiskers: standard deviation. To see this figure in color, go online.

Figure 2

Myosin II filament dynamics in the acto-myosin network at different ATP concentrations. (A and B) Shown is a histogram of myosin II filament dwell times on actin at t = 1 min (100 μM ATP) (N = 6400) (A) and at t = 16 min (N = 8000) (B); each diamond represents a five frame = 1 s bin. (C–E) Shown are histograms of the myosin II filament velocities (C), run lengths (D), and frequency plot of run length versus dwell time (E) extracted from a subset of the myosin II filament kymographs at t = 1 min (N = 432). (F–H) Shown are histograms of the myosin II filament velocities (F), run lengths (G), and frequency plot of run length versus dwell time (H) extracted from a subset of the myosin II filament kymographs at t = 16 min (N = 1133). Best fits for the distributions were computed with the MEMLET fitting routine: (A and B) double exponential decay (solid green line, dashed lines indicate 95% confidence level), (C) double Gaussian (red line), (F) single Gaussian (red line), and (D and G) single exponential decay (black line). Data displayed in (E) and (H) are displayed in the range of 0–10 s to highlight this dwell time regime. To see this figure in color, go online.

Figure 3

Single-particle tracking of myosin II filaments provides further insights into the change of myosin dynamics with decreasing ATP concentration. (A) Shown is an example image sequence showing the tracking of a myosin II filament (purple circle locates the current detection, cyan line depicts the particle’s traces). Scale bars, 1 μm. (B) Corresponding track with segmentation into segments of directed (blue) and random (red) motion. (C and D) Images depicting all tracked myosin II filaments (in alternating colors to distinguish individual tracks) at (C) t = 1 min (100 μM ATP) (N = 14,016) and (D) t = 16 min (N = 15,647). Scale bars, 5 μm. (E) Shown is a corresponding histogram of myosin II filament dwell times on actin (t = 1 min: green; t = 16 min: purple); best fits for the distributions were computed with the MEMLET fitting routine: triple exponential decay with t_min = 0.3 s (solid line); boundaries depict 95% confidence level. (F) Shown is a box plot depicting the ratio of time of directed motion versus dwell time for myosin II filaments displaying directed motion (N_1 min = 221; N_16 min = 640). (G) Shown is a radial histogram depicting the angular difference α between myosin II filament orientation and its velocity vector for all detected myosin II filament steps. (H) Shown is a radial histogram plot depicting the ratio of steps in directed motion versus all detected steps for each angular difference α, indicating that there is a clear increase of filaments that are aligned along their axis of propagation in the contractile state. To see this figure in color, go online.

Figure 4

Dynamics within acto-myosin clusters. (A) Shown is an image sequence depicting the dynamics of myosin II filaments within an acto-myosin cluster; to make the dynamics in the cluster visible, the time median of the image sequence was subtracted from each frame; 635-nm iSCAT setup. Scale bars, 2 μm. (B) Shown is a graph depicting the change of interferometric contrast from the time median of the regions depicted in (A) inside (green) and outside (magenta) the cluster. (C) Shown is a schematic summarizing the different observed binding modes of myosin II filaments and their assumed effect on the acto-myosin network dynamics and organization. To see this figure in color, go online.