Extended Kinesins Take Highly Variable Steps
Kinesin moves along the microtubule with highly regular 8 nm steps, in contrast with such other motors as myosin VI (Rock et al., 2001) and dynein (Reck-Peterson et al., 2006), which take highly variable steps. Kinesin's center-of-mass step size may be restricted to 8 nm as a result of geometrical constrains. The dimensions of the two fully extended neck linkers in the dimer (∼10 nm span) would allow the rear head to step to the next available tubulin dimer along the protofilament and not extend further forward or sideways. However, this hypothesis has not been tested directly. By extending the length of the neck linkers, we tested whether the motor can now take different sized steps. One head of the kinesin dimers was labeled with a single quantum dot, which provides a bright signal for high precision (1 nm) fluorescence tracking with better time resolution (70 ms) than in prior studies with fluorescent dyes (Yildiz et al., 2003).
The Q-dot-labeled WT kinesin head took regular 16 nm steps (Figures 3A and 3B), consistent with prior results and indicative of hand-over-hand movement of the heads (Yildiz et al., 2004). However, the steps taken by extended kinesins were often larger and highly variable (Figures 3A, 3B, and S3). For example, 26P step size histogram showed a major peak at 24 nm as well as peaks at 8, 16, and 32 nm. For the very shortest insertions (0P, 2P, and 4P), the head step size was less variable and similar in size to that of WT kinesin (Figure S4). These results indicate that extending the reach of the kinesin dimer by elongating the neck linker with a rigid proline helix of sufficient length enables the rear head to move beyond the first available tubulin-binding site after it passes its partner head. The step sizes, however, were shorter than expected if the neck linker was predicted to be fully extended. For example, 13P (4 nm helix) approximately doubles the length of the native neck linker, but relatively few (∼2%) steps were 32 nm in size and some were even shorter than 16 nm. This could be due to the flexible hinges on either side of the proline helix, which would prevent the entire linker region from fully extending in one dimension (Figure S5). The structure of the neck linker extension also affected motor stepping, since 14GS showed a distinct step size distribution, compared with 13P. In fact, the average step size of 14GS is lower than that of WT kinesin because of the appearance of ∼8 nm steps (see discussion in Figure S5). A strong forward bias was still preserved, although the frequency of backward stepping was higher for extended kinesins (5%–7%) than for WT (