Kinetochore attachments require an interaction between unstructured tails on microtubules and Ndc80(Hec1).

Kinetochore attachments to microtubules are tight enough to move chromosomes, yet the microtubules' plus ends must remain dynamic and reposition within the attachment pocket during depolymerization-coupled movement. Kinetochores are unable to bind microtubules after any of the four subunits of the Ndc80 complex are knocked down [2, 4]; however, because the Ndc80 complex has important structural roles [1-3], it is unclear whether it directly mediates kinetochore-microtubule attachments. The Ndc80(Hec1) subunit (Hec1) has a microtubule-binding site composed of both an unstructured N-terminal tail and a calponin homology domain [5-7]. Here, we show that, surprisingly, the N-terminal tail is sufficient for microtubule-binding affinity in vitro. The interaction is salt sensitive, and the positively charged Hec1 tail cannot bind microtubules lacking negatively charged tails. We have replaced the endogenous Hec1 subunit with a mutant lacking the N-terminal tail. These cells assemble kinetochores properly but are unable to congress chromosomes, generate tension across sister kinetochores, or establish cold-stable kinetochore-microtubule attachments. Our data argue that the highest affinity interactions between kinetochores and microtubules are ionic attractions between two unstructured domains. We discuss the importance of this finding for models of repositioning of microtubules in the kinetochore during depolymerization.