Ternary complex of Kif2A-bound tandem tubulin heterodimers represents a kinesin-13-mediated microtubule depolymerization reaction intermediate

Abstract

Kinesin-13 proteins are major microtubule (MT) regulatory factors that catalyze removal of tubulin subunits from MT ends. The class-specific “neck” and loop 2 regions of these motors are required for MT depolymerization, but their contributing roles are still unresolved because their interactions with MT ends have not been observed directly. Here we report the crystal structure of a catalytically active kinesin-13 monomer (Kif2A) in complex with two bent αβ-tubulin heterodimers in a head-to-tail array, providing a view of these interactions. The neck of Kif2A binds to one tubulin dimer and the motor core to the other, guiding insertion of the KVD motif of loop 2 in between them. AMPPNP-bound Kif2A can form stable complexes with tubulin in solution and trigger MT depolymerization. We also demonstrate the importance of the neck in modulating ATP turnover and catalytic depolymerization of MTs. These results provide mechanistic insights into the catalytic cycles of kinesin-13.

Evidence for conformational change-induced hydrolysis of β-tubulin-GTP

“…Our results showed that GTP hydrolysis required the associated tubulin dimers to be released from AMP-PNP-bound KIF2A-NM (i.e. bound only with DARPin, which disrupts the 1:2 KIF2A:tubulin dimer complex, as previously shown [49]). This suggests that the bending of tubulin dimers by KIF2A-NM per se does not trigger GTP hydrolysis; instead it is the return of free tubulin dimers to their native conformation that promotes the activity.…”

“…However, their depolymerizing function only happens at the ends of MTs, where they use the energy derived from ATP hydrolysis to dissociate tubulin dimers [47,48]. Our structural study shows each kinesin-13 monomer can bind to two tubulin dimers in tandem and bend both the intra-and inter-dimer curvature [49]. We postulate that this extreme bending of tubulin dimers by kinesin-13 ultimately leads to their dissociations from protofilament ends.…”

“…The presence of GDP in β-tubulin, in our previous structural study of the KIF2A-tubulin ternary complex [49], prompted us to determine if the hydrolysis of tubulin GTP occurred spontaneously during the sample preparation and crystallization processes, or if it was induced by forming the complex with KIF2A-NM (neck+motor; amino acids 153-553). To address this directly, we first used a fluorescence-based assay to detect the presence of GDP over time when tubulin dimers either were incubated alone or with KIF2A-NM in the presence of AMP-PNP and DARPin (Designed Ankyrin Repeat Protein, a polypeptide that binds to β-tubulin and prevents MT polymerization).…”

“…Note that tubulin dimers incubated with AMP-PNP-bound KIF2A alone or with DARPin alone did not promote GTP hydrolysis. From our previous study [49], we knew that DARPin could compete with the neck region of KIF2A-NM for binding to β-tubulin and thereby release KIF2a-NM from the complex, facilitating GTP turnover. Therefore, excess DARPin should disrupt the complex formation and suppress GTP hydrolysis.…”

“…From our previous study [49], we know that each kinesin-13-NM motor molecule can form a complex with two tubulin dimers in tandem and changes their curvatures, but we do not know whether this interaction triggers hydrolysis of both GTP molecules or just one. Our crystal structure suggests one of the two dimers is more curved than the other when they bind to KIF2A-NM with AMP-PNP.…”