Fig. 4. A tug-of-war model predicts the observed parameters of bidirectional transport in vitro.

A tug-of-war model [21] was used to analyze the observed motility of neuronal transport vesicles moving bidirectionally along MTs in vitro. Model parameters were based on previous experimental observations [21, 29]. The two free parameters in the model are the number of actively engaged plus- and minus-end directed motors. (A) The relative fraction of time vesicles are moving toward either the plus or minus ends of the MT depends strongly on the ratio of the number of engaged plus- and minus-end directed motors. Predictions are shown over a range of dynein:kinesin-1 and dynein:kinesin-2 ratios. (B) Experimental data from vesicle motility along MTs is best described by a mole ratio of 7:1 dynein to kinesin-1 motors or 3:2 dynein to kinesin-2 motors. Data from vesicles treated with an Ab to DHC and DIC suggest a fraction of the dynein is inhibited under these conditions. (C) The Gillespie method [28] was used to generate simulated trajectories for control vesicles as well as vesicles incubated with pAb-DHC for motility driven by dynein and kinesin-1 (red), and dynein and kinesin-2 (black). Simulated trajectories are compared to kymographs of vesicle motility (excerpted from Fig. S4). (D) Regulation of bidirectional transport likely occurs at several levels. Recruitment, activation, or inhibition of motor proteins regulates the number of active motors associated with vesicular cargo. Regulation on a longer time scale (τ_long) likely involves motor effectors. At shorter time scales (τ_short), net directionality of movement results from a stochastic “tug-of-war” among opposing motor proteins bound to the same cargo and actively engaged with the MT.