2. MECHANICS OF UNCONVENTIONAL MYOSINS67to actin). Kinetic bottlenecks for myosin V and myosin VI are shifted compared to muscle myosin II, so that they occur before actin release rather than before actin binding. The duty ratio was estimated to be $ 0.7 for myosin V (49) and $ 0.9 for myosin VI (67). In the case of myosin VI, the low ATP a僴ity means that a signiant fraction of the motor populates the rigor state. Using the information gathered from these kinetic measurements, it is possible to create simple models of processivity to test aspects of kinetic tuning (35). In the scheme outlined in Eq. (1), the duty ratio is given by r k1/(k1 k1). The probability of reattachment of the free head from the DA state is given by PDA k1/(k1 k2). If the two heads were operating independently, the detachment rates for the st and second heads would be the same, so k1 k2. Under these circumstances, the probability of taking an additional step from the DA state, rather than detaching, is given by the duty ratio, r. The probability that a motor takes exactly n steps under these circumstances is therefore Pn rn(1r), and the processivity index is n0.5 ln 2/(jln r j) [see also (69)]. Such an independent head model predicts an average run length of $ 2 steps for a myosin V with a duty ratio of 0.7. In comparison, average run lengths of 66 steps have been observed in myosin V motility observed by total internal rection (31). Such runs would require a duty ratio of at least 0.99 by this model, which suggests that the heads in fact communicate their biochemical state to one another, to avoid simultaneous detachment. Other information suggests that communication between the two heads must occur. At least in the case of tightly coupled, hand-over-hand mechanisms, which seem to apply to myosin V (30, 70), it can be shown that the two heads cannot be truly independent. For myosin V, the rate of ADP release is at least 10-fold slower than the aggregate of all other rates in the ATPase cycle at saturating ATP (30, 49). If the heads were independent, they would each cycle at the same rate. Cycling of each of the individual heads under this circumstance would be well described by a single Poisson process with a high randomness (69). In this situation, the historical pattern of cycling is unimportant. Once one head completes a cycle, it has a 50% probability of completing another cycle before its partner. However, each sequential cycle in the leading head is coupled to a mechanical advance. This requires that one head may step multiple times along actin while its partner is left behind, which is an unlikely situation. Alternatively, the lagging head may prevent the leading head from advancing, so that futile hydrolyses occur, but this is inconsistent with tight coupling. Therefore, either cycling in the trailing head is accelerated, or cycling in the lead head is retarded, which requires that the two heads communicate their identities (�leading� vs. �trailing�) to each other. In fact, in optical trapping assays using the
