Molecular Requirements for the Transition from Lateral to End-on Microtubule Binding and Dynamic Coupling

Abstract

Accurate chromosome segregation relies on the ill-understood ability of kinetochores to convert their lateral microtubule attachment into the microtubule plus-end association, capable of the processive motion with tubulin dynamics. Here we report that this transition can be recapitulated in vitro using only two components: the plus end-directed kinesin CENP-E and the microtubule wall-binding Ndc80 protein. CENP-E's primary role is to establish the end configurations for Ndc80, while Ndc80 mediates the maintenance of end attachment. To gain insights into the molecular requirements for end-conversion we paired CENP-E with other microtubule-binding proteins. Ska1, CENP-E Tail, EB1 and CLASP2 differed in their ability to retain the microtubule ends, and none performed as robustly as Ndc80. Likewise, a non-mitotic transporter Kinesin-1 failed to support the Ndc80-mediated end-conversion, implying that the pair of CENP-E kinesin and Ndc80 is optimally suited for this activity. To investigate the underlying mechanistic differences between these motors and MAPs, we developed a Brownian dynamics model for the molecular ensembles of proteins engaged stochastically in walking and diffusion on the microtubule wall. Modeling demonstrates that the observed differences in the end-retention by different MAPs can be largely attributed to their different residency times and rates of diffusion on the microtubule wall. The model also recapitulates the strikingly different behavior of Kinesin-1 and CENP-E, suggesting that it is rooted in their distinct force-detachment characteristics. Following a model prediction, we were able to achieve robust end-conversion with Kinesin-1 by amending its dynamic response via the reduced ATP concentration. Together, our results argue strongly that microtubule end-conversion is an emergent property of the ensemble of transporting motors and diffusing MAPs, limited by the microtubule end boundary. We propose that similar mechanism ensures microtubule end conversion at mitotic kinetochores.