Shaping the tracks: Regulation of microtubule dynamics by kinesins KIF21A and KIF21B

Abstract

Control of microtubule dynamics is important for cell morphogenesis. Kinesins, motor proteins known to function in cargo transport, were recently also implicated in altering the microtubule network. Several kinesins are described to cause microtubule network reorganization or stabilization, either via inducing microtubule depolymerization or inhibiting polymerization. With our studies we add the kinesins KIF21A and KIF21B, members of the kinesin-4 family, to the list of microtubule growth inhibitors. Using cellular assays and in vitro reconstitutions, we show that the motor domains of these kinesins have an intrinsic ability to inhibit microtubule polymerization. KIF21B was shown to be a strong inhibitor as few dimers could lead to a pause in microtubule growth. Furthermore, by dissecting the proteins we found different domains to be involved in regulation of protein activity. We show kinesin stalk and tail domains to be involved in recruiting the molecules to specific sites in the cell via interactions with binding partners, leading to site-specific stabilization of the microtubule network. In addition to the motor domain, microtubule binding sites were found in different parts of the kinesins, possibly all contributing to the ability of the kinesins to inhibit microtubule dynamics. To prevent futile use of energy and aberrant transport of cargo, kinesins can be inactivated via autoinhibition. We observed KIF21A and KIF21B to also inactivate through autoinhibition, which depends on a region of their coiled coil stalk interacting to the motor domain. We found that this inhibitory segment in the coiled coil folds into an intramolecular antiparallel coiled coil, a structure not yet described in other kinesins. Furthermore, regulation of kinesin activity was shown to be important, as KIF21A mutations found in the developmental syndrome Congenital Fibrosis of the Extraocular Muscles type 1 (CFEOM1) caused kinesin hyperactivation, leading to its increased accumulation in neuronal growth cones and defects in neuronal pathfinding. CFEOM1 mutations in KIF21A either alter the antiparallel coiled coil structure, or interfere with the binding interface between the inhibitory region and the motor domain, thus preventing the autoinhibitory self-folding of the kinesin. Relief of kinesin autoinhibition can be regulated by multiple factors. For KIF21B, it was shown to occur through microtubule binding by the WD40 domain in the tail of the molecule. In addition, we identified binding partners of both KIF21A and KIF21B which might regulate their activity, including several microtubule-associated proteins that bound to KIF21A and recruited KIF21A to microtubules. Also, we found that KIF21B binds to Ezrin, which is known to be involved in microtubule organization at immunological synapses. A possible role for KIF21B in this process was shown through knockout of KIF21B, which led to decreased spreading of T-cells on an activating substrate, possibly reflecting defects in synapse formation. With our work, we show KIF21A and KIF21B to be important regulators of microtubule dynamics, and demonstrate that microtubule growth inhibition is a general function of kinesin-4 family members. Microtubule growth is important in different cellular processes, and perturbations of its regulation might lead to pathogenesis, as observed for KIF21A and suggested for KIF21B.