Kinesin-1 motors can circumvent permanent roadblocks by side-shifting to neighboring protofilaments

Abstract

Obstacles on the surface of microtubules can lead to defective cargo transport, proposed to play a role in neurological diseases such as Alzheimer's. However, little is known about how motor proteins, which follow individual microtubule protofilaments (such as kinesin-1), deal with obstacles on the molecular level. Here, we used rigor-binding mutants of kinesin-1 as roadblocks to permanently obstruct individual microtubule binding sites and studied the movement of individual kinesin-1 motors by single-molecule fluorescence and dark-field scattering microscopy in vitro. In the presence of roadblocks, kinesin-1 often stopped for ∼ 0.4 s before either detaching or continuing to move, whereby the latter circumvention events occurred in >30% after a stopping event. Consequently, and in agreement with numerical simulations, the mean velocity, mean run length, and mean dwell time of the kinesin-1 motors decreased upon increasing the roadblock density. Tracking individual kinesin-1 motors labeled by 40 nm gold particles with 6 nm spatial and 1 ms temporal precision revealed that ∼ 70% of the circumvention events were associated with significant transverse shifts perpendicular to the axis of the microtubule. These side-shifts, which occurred with equal likelihood to the left and right, were accompanied by a range of longitudinal shifts suggesting that roadblock circumvention involves the unbinding and rebinding of the motors. Thus, processive motors, which commonly follow individual protofilaments in the absence of obstacles, appear to possess intrinsic circumvention mechanisms. These mechanisms were potentially optimized by evolution for the motor's specific intracellular tasks and environments.

Figure 1

Motility of GFP-labeled motors in the presence of unlabeled roadblocks studied by TIRF microscopy. (A) Schematic illustration of the stepping assay and the engineered protein constructs. (B) Definition of the motility parameters. The total distance and the duration of the trajectory were termed run length and dwell time, respectively. Switches from a moving to a waiting phase were termed stopping events. Switches from a waiting to a moving phase were termed starting events. The velocity of an individual motor was obtained by dividing the run length by the dwell time. (C) Mean waiting times of stopping and waiting motors (open circles, 0.37 ± 0.09 s, mean ± SD, N = 6 roadblock concentrations) were similar to the mean waiting times of pausing and starting motors (black spheres, 0.41 ± 0.09 s). (D) Kymographs of individual motors (at 1 mM ATP) walking along MTs decorated by the presence of 0, 0.75, 1.5, and 15 nM roadblock concentration. To see this figure in color, go online.

Figure 2

Experimental and simulated motility parameters in the presence of roadblocks. (A–C) Deterioration of the motility parameters mean velocity (A), mean run length (B), and mean dwell time (C), as a function of the roadblock concentration (black dots, mean ± SD, N = 3 movies with >250 motors total) and comparison with simulated data (red, orange, and green data points according to the scenarios in (D), mean ± SD, N = 3 simulations with 1000 motors each). The dotted line in (B) represents the experimentally determined mean roadblock spacing (see Supporting Material C and Fig. S5). (D) Simulated roadblock-encounter scenarios: (i) waiting and circumvention: motors wait for 0.4 s and subsequently continue in 30% and detach in 70% of the encounters (green box); (ii) waiting and no circumvention: motors detach after 0.4 s of waiting (orange box); (iii) no waiting and no circumvention: motors immediately detach upon roadblock encounter (red box).

Figure 3

Investigation of the mechanism by which motors circumvent roadblocks. (A) Individual AuNP-loaded motors (coupled via GFP-antibodies to GFP-labeled motor tails) were allowed to interact with surface-immobilized, roadblock-decorated MTs in the presence of 1 mM ATP. (Inset) Switching between MT protofilaments requires the motor head to move sideways by ∼6 nm, whereas the accompanied transverse shift of the AuNP is amplified by the extended motor tail and the AuNP diameter. (B) The transverse shift is determined by the difference in transverse displacement associated with stopping and starting events. Waiting phases (W, underlayed red) and moving phases (M, underlayed green) were deduced from the longitudinal displacement (see C, left). (C) Exemplary trajectory of a pausing motor: longitudinal (left) and transverse displacement (right) versus time. The sizes of the transverse shifts are given for each starting event. A negative shift depicts movements toward the right. (D) Histogram of transverse shifts (N = 106 starting events from pausing motors). Contributions by significant and insignificant shifts are shown in red and gray, respectively. (E) Histogram of longitudinal shifts (corresponding to the events in D). Contributions from events with significant and insignificant transverse shifts are shown in red and gray, respectively. (F) Heat map of longitudinal shifts sorted into shifts associated with significant transverse shifts to the left and right as well as insignificant transverse shifts (corresponding to D and E). The dashed white line denotes the longitudinal displacement before the circumvention event and the red cross denotes the presumable longitudinal position of the roadblock.

Figure 4

Behavior of AuNP-loaded kinesin-1 motors in the absence and presence of 2 nM added roadblocks. (A) In the absence of added roadblocks (at 0 nM roadblocks), motors detached with a probability of 0.73% per step. Upon encounter of an intrinsic obstacle (dotted box, here with a probability of 0.52% per step), motors switched into a waiting phase lasting on average 0.39 s. Subsequently, motors either detached in 46.7% of the cases or continued to walk with a chance of 53.3% of the cases. Circumvention events were accompanied by significant transverse shifts in 61% of the cases. (B) In the presence of 2 nM added roadblocks, motors detached with a 1.9-fold higher probability (1.36% per step). Roadblocks were encountered with a 2.4-fold higher probability (1.27% per step) but the duration of waiting phases remained unchanged (0.39 s). Waiting phases were exited by detachment in 48.1% and by continuation of walking in 51.9%. Circumvention events were accompanied by significant transverse shifts in 80% of the cases. See Table S4 for the full analysis including the condition of 4 nM added roadblocks. To see this figure in color, go online.