This model does not preclude a “chemical gate,” which surely must also exist. Early kinetic experiments (Gilbert et al., 1998 and Hackney, 1994) showed that ADP release of one head was greatly accelerated by ATP binding to the partner head on the microtubule. However, the lack of ADP release in one head may not be simply explained by that head not being able to access the microtubule (Hackney, 2005). In addition, compelling experiments by Guydosh and Block (2006) showed that kinesin must first take a backward step to escape from a block created by the binding of a nonhydrolyzable nucleotide to one of the kinesin heads. The result was interpreted as evidence for strain-regulated nucleotide binding to the front head. However, the Guydosh and Block results can also be interpreted as nucleotide release from the rear head being inhibited, perhaps due to forward pointing position of the neck linker (Mori et al., 2007).
A chemical gate of kinesin could operate in concert with strain-dependent dissociation of the rear head from neck linker docking. However, a rear head detachment mechanism that is promoted by neck linking docking (depicted in Figure 7) cannot operate with a chemical gate in which the front head cannot bind ATP until the rear head detaches. Further experiments will be needed to determine definitively which mechanism is operational in a walking kinesin motor. Another point that remains to be addressed is whether strain-dependent release operates at very low ATP concentrations. Our experiments show that external strain can increase the velocities of extended and WT kinesin at subsaturating ATP concentrations. Recent work by Mori et al. (2007) demonstrates a “one-head-bound” state of kinesin at low ATP concentrations, with the weak binding head positioned behind the strongly bound head. The rear head may still bind weakly to microtubules and be subject to complete detachment by force. Thus, it is possible that the force-induced velocity increase at low ATP concentrations occurs by either promoting rear head release or preventing reattachment of the detached head to the rear tubulin-binding site. Further insight into chemical gating also might be gained by understanding how ATP turnover becomes uncoupled from mechanical stepping in the extended kinesins. Such uncoupling might arise from ADP/ATP exchange in the rear head without its associated forward movement due to its “relaxed” neck linker. However, other mechanisms also could account for uncoupling, and new assays will be needed to probe the chemical cycle of kinesin at a single-molecule level.
The Role of the Neck Linker in Kinesin Stepping
Rice et al. (1999) originally proposed that ATP-induced neck linker docking in the front kinesin head triggers the 16 nm displacement of the rear head. However, a criticism of the neck linker theory is whether the docking of this small peptide to the catalytic core provides sufficient energy to power kinesin movement under load (Rice et al., 2003 and Schief and Howard, 2001), although recent computational modeling suggests that additional energy may be derived through an interaction between the kinesin N-terminal peptide interacting with C-terminal neck linker in the ATP state (Hwang et al., 2008). Here, we show that kinesin without its neck linker will still walk processively if a “bias” is provided by an optical trap. This result reveals that, in the absence of a mechanical element, the motor domain responds to tension by releasing and rebinding to the microtubule. This result also implies that a key function of the neck linker and its ATP-driven conformational change is to provide a directional force that releases and biases the motor domain (which the trap can mimic in the 19P-NL experiment). As described above, a very small force may be needed to detach the rear head and bias its movement forward, whereas most of the energy needed to move kinesin forward against a 5–6 pN backward load would be derived from the subsequent reformation of tight kinesin-microtubule-binding interaction, which locks the step in place.