Kinesins are highly regulated mechanochemical motors that drive cargoes along microtubules to maintain cellular homeostasis and conduct transformative processes, such as migration and mitosis. Spatiotemporal regulation of kinesin is essential for cell function. Kinesin binding protein (KIFBP) is a recently characterized multi-kinesin inhibitor that is essential for neurodevelopment. Despite its relevance, KIFBP and its role in kinesin inhibition are critically undercharacterized. KIFBP is a 71kDa alpha-solenoid protein with tetratricopeptide-repeats (TPR) that inhibits select members of the kinesin-2, 3, 8, and 12 families. Typically, TPRs mediate versatile protein-protein interactions by binding charged disordered sequence motifs. KIFBP does not follow this convention. Instead, KIFBP binds kinesin’s motor domain and remodels the force-communicating ɑ4-helix. By sequestering kinesin’s ɑ4-helix, KIFBP inhibits kinesin from landing on microtubules. In this work, we utilize single-particle cryo-EM to understand how KIFBP binds and remodels kinesin motors. To this end, we determined a 3.8Å reconstruction of KIFBP bound to kinesin-2/KIF3A. Our structure allowed us to model the precise sequence of KIF3A-ɑ4 that is sequestered by KIFBP, which is a non-canonical TPR substrate. Notably, our structural work shows that the kinesin-2/KIF3A-ɑ4 helix flips upon remodeling into KIFBP and is stabilized in the TPR groove by large hydrophobic interactions. During our analysis, we identified a new “primed state” of KIFBP:kinesin-2/KIF3A in which the kinesin-2 motor is bound but not remodeled. This structural state, combined with an apo-KIFBP reconstruction at 3.6Å, allows us to propose how conformational changes in KIFBP allow kinesin-2 motor recognition and remodeling. This work provides a detailed view of KIFBP bound to kinesin-2/KIF3A while also highlighting a new mode of TPR-mediated substrate binding. Finally, this work opens the door for investigation of KIFBP in kinesin-2-dependent ciliogenesis and axonal transport.
Missman et al. (2026) studied this question.