Extended Kinesins Undergo Futile Cycles of ATP Hydrolysis
The ATPase and velocity data described above suggested that extended kinesins undergo “futile” ATPase cycles, hydrolysis events that do not lead to forward movement. We searched for additional evidence of such uncoupling by performing a fluctuation analysis of dwell times in between steps, which yields a randomness parameter, “r” (Schnitzer and Block, 1997) (Supplemental Text). An r value of 0 reflects a constant dwell time between steps, whereas r values of 1 and 0.5 are expected for motors with one and two rate-limiting transitions per step, respectively. Our dwell time analysis of WT kinesin (one head labeled) at low ATP revealed an r of 0.57 (Figure 3A, inset), reflecting the expected two successive ATP-binding events that occur between 16 nm steps (the head takes a 16 nm step during one ATP binding or turnover, but remains stationary during the next cycle while its partner head is moving). In contrast, the extended kinesins showed larger fluctuations of their dwell times (r = 0.97–1.01), which is most likely caused by a lower stepping probability per ATP-binding event (Schnitzer and Block, 1995). These data indicate that extended kinesin motors often fail to step forward after binding or hydrolyzing ATP, in contrast to WT kinesin. In addition, a histogram of WT kinesin dwell periods can be fit well to a convolution of two exponentials, as expected from a tight coupling of ATP binding to hand-over-hand movement of the heads (Yildiz et al., 2004). However, extended kinesins display an additional population of very long dwells, which likely reflects impaired coupling of ATP hydrolysis to a mechanical step (Figure S7).