Assembly principles and stoichiometry of a complete human kinetochore module

Abstract

Centromeres are epigenetically determined chromosomal loci that seed kinetochore assembly to promote chromosome segregation during cell division. CENP-A, a centromere-specific histone H3 variant, establishes the foundations for centromere epigenetic memory and kinetochore assembly. It recruits the constitutive centromere-associated network (CCAN), which in turn assembles the microtubule-binding interface. How the specific organization of centromeric chromatin relates to kinetochore assembly and to centromere identity through cell division remains conjectural. Here, we break new ground by reconstituting a functional full-length version of CENP-C, the largest human CCAN subunit and a blueprint of kinetochore assembly. We show that full-length CENP-C, a dimer, binds stably to two nucleosomes, and permits further assembly of all other kinetochore subunits in vitro with relative ratios that closely match those of endogenous human kinetochores. Our results imply that human kinetochores emerge from clustering multiple copies of a fundamental module, and may have important implications for trans-generational inheritance of centromeric chromatin.