Mitotic spindle formation relies on the stochastic capture of microtubules at kinetochores. Kinetochore architecture affects the efficiency and fidelity of this process with large kinetochores expected to accelerate assembly at the expense of accuracy, and smaller kinetochores to suppress errors at the expense of efficiency. We demonstrate that upon mitotic entry, kinetochores in cultured human cells form large crescents that subsequently compact into discrete structures on opposite sides of the centromere. This compaction occurs only after the formation of end-on microtubule attachments. Live-cell microscopy reveals that centromere rotation mediated by lateral kinetochore-microtubule interactions precedes formation of end-on attachments and kinetochore compaction. Computational analyses of kinetochore expansion-compaction in the context of lateral interactions correctly predict experimentally-observed spindle assembly times with reasonable error rates. The computational model suggests that larger kinetochores reduce both errors and assembly times, which can explain the robustness of spindle assembly and the functional significance of enlarged kinetochores.

The primary non-motile cilium, a membrane-ensheathed, microtubule-bundled organelle, extends from virtually all cells and is important for development. Normal functioning of the cilium requires proper axoneme assembly, membrane biogenesis and ciliary protein localization, in tight coordination with the intraflagellar transport system and vesicular trafficking. Disruptions at any level can induce severe alterations in cell function, giving rise to a myriad of human genetic diseases known as ciliopathies. Here we show that the Abelson helper integration site 1 (Ahi1) gene, whose human ortholog is mutated in Joubert syndrome, regulates cilium formation via its interaction with Rab8a, a small GTPase critical for polarized membrane trafficking. We find that the Ahi1 protein localizes to a single centriole, the mother centriole, which becomes the basal body of the primary cilium. In order to determine whether Ahi1 functions in ciliogenesis, loss of function analysis of Ahi1 was performed in cell culture models of ciliogenesis. Knockdown of Ahi1 expression by shRNAi in cells or targeted deletion of Ahi1 (Ahi1 knockout mouse) leads to impairments in ciliogenesis. In Ahi1-knockdown cells, Rab8a is destabilized and does not properly localize to the basal body. Since Rab8a is implicated in vesicular trafficking, we next examined this process in Ahi1-knockdown cells. Defects in the trafficking of endocytic vesicles from the plasma membrane to the Golgi and back to the plasma membrane were observed in Ahi1-knockdown cells. Overall, our data indicate that the distribution and functioning of Rab8a is regulated by Ahi1, not only affecting cilium formation, but also vesicle transport.

In the absence of continuous K-fiber attachment between each kinetochore and the spindle pole, one or more additional mechanisms dependent on dynein-mediated kinetochore transport exist to ensure proper chromosome segregation during mitosis.