H1 and H2 form an inhibitory finger in the Unc104 monomer.

The familiar arm-over-arm kinesin swing was recently challenged by proposals that KIF1A acts as a processive, but monomeric, kinesin motor. The proposal involved a rotation of the motor that, along with a microtubule-binding patch, would bias further movement in one direction.

But now Al-Bassam et al. add to the growing evidence that the KIF1A orthologue in worms, Unc104, may function as a dimer, like conventional kinesin (page 743). The monomer, they propose, is instead a regulated form whose full activity is only restored when motors are crowded onto cargo vesicles.

The unusual prospect of a monomeric motor, and expression problems that had led to the use of a KIF1A/conventional kinesin hybrid in the earlier work, led the authors to study Unc104. Images obtained by cryo-EM revealed a motor domain with a protruding finger that the group identified as representing two neck helices, paired in parallel.

Under other conditions, the finger unfolds (i.e., disappears in the EM), and its helices pair intermolecularly with neck helices from another Unc104 motor. The dimeric motor that is now visible by EM should move vesicle traffic along axons processively.

Dimerization would be favored when the motor is at high concentrations on vesicle cargo surfaces. But more sparsely spaced Unc104 may stay monomeric because the two neck helices pair with each other. Al-Bassam et al. prevent this inhibitory pairing, and thus the formation of the finger, by deleting a hinge between the two helices. This deleted protein can move well in vitro, but is a poor replacement for wild-type gene function in worms. Perhaps this motor is activated too easily, and therefore whizzes out to the axon before it has cargo.

An important question remains: does any kinesin truly operate as a monomer? The sizeable rotation seen in the KIF1A studies (which was proposed to drive monomeric movement) is minimal in Unc104, based on the new structures. But a decisive conclusion may require further experiments. If dimerization is required for movement, then mutations in neck helices that disrupt dimerization should always destroy movement. Many such correlations may be necessary before the monomeric camp gives up its cause. ▪