(A) Dimeric myosin VI (M6) with and without swivels in the MT domain moves processively on actin (Blue). The N-terminal catalytic domain, bound calmodulins, and proximal tail (PT) domain are shown in yellow. The PT is followed by the medial tail (MT, Black), a GCN4 coiled-coil domain (Black), and a C-terminal YFP that replaces the cargo binding domain (Green). Three different constructs, swivel 1, 2, and 3, contain the amino acid sequence GGSGGSGGSGG inserted after residues L914, Q931, and R941 (Small Arrows), interrupting the MT. (B) Both control M6 dimer (referred to throughout as M6dimer; Black) and swivel 1 (Blue, fit in Red) move processively on actin. Run lengths of 780 ± 80 nm (M6dimer; N = 113) and 350 ± 30 nm (swivel 1; N = 97) were measured using single-molecule TIRF microscopy. Run lengths for the other swivel constructs are in Table 1.

(A) GNT schematic. M6dimer walks along surface-immobilized actin. Light from a 532 nm laser is totally internally reflected at the glass-water interface, creating evanescent illumination of the sample. The return beam is deflected by a small wedge mirror. Light scattered by gold particles (Thick Beam) passes through a 50/50 beam splitter and is imaged using a fast camera. (B) Dimeric swivel 1 takes large steps. A forward step size of 30 ± 1 nm (SEM, N = 222) was measured using gold nanoparticle tracking (GNT). (C) Averaged steps for M6dimer (Blue; 2,000 frames s^-1; N = 93), swivel 1 (Green, 500 frames s^-1; N = 60), and swivel 2 (Red, 2,000 frames s^-1; N = 46). Displacements due to the lever arm swing and the diffusive search occur faster than the time resolution of our measurement. Only steps with amplitudes between 30 and 45 nm are included in the above averages. These steps are expected to have substantial diffusive components in the case of the swivel constructs, which have stroke sizes measured at ∼20 nm.

The dimeric swivel 1 step size decreases under load. (A) Optical trapping with force-feedback was used to apply a constant load to processively stepping dimeric swivel 1. Each color represents a different molecule. An open circle indicates the step size at zero applied load as measured using GNT. A gray line shows a linear fit to the optical trap data. M6dimer step size data are indicated with a black ×. M6dimer step size data from ref. 21 are shown as black closed circles. (B) (Top) Under very low applied load the swivel 1 step size approximates the actin pseudohelical repeat (Purple). (Bottom) Higher loads result in decreased step size.

Intramolecular strain blocks ADP release in the front head of dimeric myosin VI. (A) Mechanism with ADP gating. Myosin VI walks right to left. Both ADP release (D; First Line) and ATP binding (T; Second Line) in the rear head precede the step. ADP release is blocked in the new lead head (Third Line). ADP release and ATP binding at the rear head again precede the next step (Fourth Line). The resulting dwell time distribution fits a sequential exponential model, with two rates reflecting the rates of rear head ADP release and ATP binding. (B) Mechanism without gating. ATP binds to the empty rear head (ϕ) resulting in a step (First and Second Lines). ADP leaves the front head prior to the next step (Third and Fourth Lines). The resulting stepping kinetics are single-exponential, and reflect the ATP binding rate. (C and D) Sequential exponential model fits (Red) to dwell time distributions measured for M6dimer under 1.5 pN load in the optical trap (C; 100 μM ATP, N = 320, fit rates of 4(3.8) ± 1 s^-1 and 1.0(0.94) ± 0.1 s^-1) and under no load measured using GNT (D; 120 μM ATP, N = 464, fit rates of 7(5.8) ± 1 and 1.7(1.8) ± 0.1 s^-1). Fit rates are derived from bootstrap analysis and MLE (parentheses). GNT and optical trap dwell time distributions collected in the presence of saturating ATP resolve sequential processes with rates of 5 and ∼30 s^-1, corresponding to ADP release and the next slowest process in the catalytic cycle (Figs. S4 and S7). The observation of the much faster 30 s^-1 rate confirms that both assays can resolve the ∼5 s^-1 rate we attribute to ADP release in panels C and D.