Myosin-5 varies its steps along the irregular F-actin track

Figures

• 

  Figure 1.. Trajectory along F-actin of a myosin-5a motor labelled with a gold bead. (A) Diagram of myosin-5a, truncated after the proximal coiled coil. (B) Schematic of two consecutive myosin steps, depicting first, the labelled head movement (blue arrow; B-state to A-state) and second, the AB transition of the bead (red arrows) accompanying the unlabeled head movement and allowing measurement of the labelled head stride (B-state to B-state). Also shown at right is actin’s short-pitch helix (top filament; bar indicates separation of successive subunits along that helix) and two long-pitch helices (middle filament; bar indicates spacing of successive crossovers). (C) The XY trace of a typical run along F-actin of a gold labelled myosin-5a molecule, highlighting labelled head strides (blue) and AB transitions (red). We report stride lengths as the Euclidean distance between the mean x and y values of consecutive B states. Localization precisions (σ), defined as the root sum of squares of x and y standard deviations of stationary states, marked above each B state. (D) The 2D kernel density estimates of bead positions highlighted in (C) with black points at each localization and white points at…

• 

  Figure 2.. Analysis of strides taken by myosin-5a on F-actin. (A) Histogram of measured myosin-5 strides (N = 96 molecules, 725 strides). A Gaussian kernel density estimate, with locally adaptive bandwidth, is shown (solid orange line), as well as seven individual Gaussians (dashed lines). Mean stride length (nm) of each Gaussian is labelled above the fits, and calculated number of actin subunits traversed marked. The single Gaussian that expresses the overall mean and SD of the dataset is shown (grey shaded), rescaled to show fit to relative frequencies of the strides. (B) A schematic depicting the set of 9 equations used to calculate the probability of a given step length (Px, where x = 9, 11, 13, 15, 17 actin subunits) from the areas of the Gaussian fits to the stride peaks listed on the left side. Center panel lists the calculated step length probabilities. On the right is the multiplication diamond using the calculated probabilities of each step length to determine the percentage frequency of a stride resulting from each combination of step lengths. Combinations of step lengths that result in the same stride length are of the same color. The total predicted frequencies of each stride…

• 

  Figure 3.. CryoEM and nsEM of myosin-5a walking on F-actin. (A) Gallery of single cryoEM images depicting myosin-5 molecules with step lengths of 11, 13 or 15 actin subunits. Contrast inverted (protein white). (B) Atomic models corresponding to (A), constructed according to Vale and Milligan42, assuming a 13/6 helix for F-actin. Actin subunits 11, 13 or 15 subunits away from the trailing motor domain are colored pink, red and yellow respectively. (C) Averaged nsEM images of myosin-5 molecules with step lengths of 11, 13 or 15 actin subunits. Note that for the step length of 11 subunits, the leading motor is in an unprimed conformation, unlike for 13 and 15 subunit steps. (D) Left panel: averaged cryoEM image of myosin-5 molecules with step length of 13 actin subunits. Right panels: result of classification into three subclasses based on features in the leading lever. Note that the right subclass has a convex leading lever shape, but the lever still emerges from the trailing side of the motor domain. (A-D) Actin barbed (+) end and myosin-5a leading head are on the right in all panels. All panels (A-D) are scaled to match the panels in (B), in which 13 actin subunits span a distance of 35.75…

• 

  Figure 4.. Implications of variable step length if F-actin has fixed or disordered helical parameters. (A) Schematic detailing how varying step and stride lengths would rotate myosin-5a trajectory around the actin filament given a fixed actin rotation per subunit, ϕ= −166.39°. Motor binding site on actin subunit depicted as white spot. A myosin-5a molecule is schematically shown at a stage where one motor (black) has detached from the black actin subunit and the attached motor (yellow) has undergone its power stroke. The detached motor is transiently dwelling off-axis, behind the plane of the figure and will bind to the right of the attached motor. Stride color scheme matches Figure 2. (B) F-actin with variable CAD. Filament 1 has fixed subunit rotation −166.39° and axial translation 2.75 nm. Filaments 2–6 were built incrementally from left to right by adding subunits with an addition to the rotation per subunit randomly drawn from a Gaussian distribution with mean 0° and standard deviation 5.28°. Actin subunits that lie closest to the same azimuth as the zeroth (black) subunit are highlighted with the step colors used in (A), to show expected myosin step lengths, indicated by brackets on…

• 

  Figure 5.. Comparison of F-actin cumulative angular disorder with myosin-5a step and stride length distributions. (A) Schematic numbering the actin subunits involved in steps and strides. Color scheme matches Figure 4. (B) and (E) Calculated probabilities of each step or stride length from iSCAT data, compared with probabilities from the angular disorder 0° azimuth fit, as shown in (C,D) and (F,G), respectively. (C) and (F) Radial plots, i.e., as if looking along the F-actin axis towards the barbed end, showing the azimuthal probability distributions of the nth subunit away from the initial (i.e., the black) subunit along F-actin with the subunit rotation and CAD values (ϕ= −166.39 ± 5.28°) obtained from least squares fitting to the iSCAT data. Probability densities have been normalized such that the sum of probabilities at μ = 0° is 1. (D) and (G) Expanded ± 10° region showing probabilities of the nth subunit lying at μ = 0°.

• 

  Figure 6.. Analysis of impact of stride length on dwell times of A and B states. (A) Distance vs. Time trace of a typical gold-labelled myosin-5a motor domain demonstrating variability in dwell times of labelled motor domain. Dwell times (ms) marked below A states (filled boxes) and above B states (empty boxes). (B) Measured dwell time of A states after stride and B states before stride as a function of stride length. Each observed dwell is shown as a horizontal line. Violin plot envelopes are kernel density estimates of the distributions with bandwidth of 150 ms. (C) Fitted variable rate constants, ks and kf, as a function of stride length. (D) Shared fit variable rate constant, k, as a function of stride length. Error bars in (C) and (D) denote the standard error of the fit parameters obtained from the observed Fisher information matrix.