Characterization of the protein-protein interaction network within the central domain of the S. cerevisiae kinetochore

Abstract

The kinetochore is a specialized structure composed of centromeric DNA and a large number of proteins. The primary function of the kinetochore is to connect chromosomes with the mitotic spindle throughout the cell cycle and to monitor the fidelity of these attachments in order to ensure proper chromosome segregation. Despite the fact that chromosome segregation is directed by the kinetochore, the architecture and assembly of such an intricate structure remains elusive. We use budding yeast as a model system to characterize direct interactions among the central domain of the kinetochore proteins, specifically the COMA-complex. The COMA-complex consists of four proteins; two nonessential proteins Ctf19 and Mcm21 and two essential proteins Ame1 and Okp1. Although the chromosome segregation is a highly conserved process, the human orthologues of the two essential Ame1 and Okp1 proteins have not been identified. The tetrameric complex is the core of the COMA-network which is composed of seven additional nonessential proteins: Ctf3, Mcm16, Mcm22, Chl4, Iml3, Nkp1 and Nkp2, more loosely associated. According to the central localization within the kinetochore, the COMA-complex represents one of the linker complexes (together with the Mtw1-, Ndc80- and Spc105-complexes) bridging the centromere-associate inner proteins with microtubule-bounded outer kinetochore proteins. Here we present a biochemical approach to reconstitute and to characterize the budding yeast COMA-network. Our first aim was to reconstitute the COMA-complex in vitro and in vivo. A stabile heterodimer consisting of Ctf19 and Mcm21 proteins could be reconstituted as a tetrameric complex in solution. The Ame1 and Okp1 heterodimer showed noticeable instability and we were not able to reconstitute it. Surprisingly, the trimeric Ctf19, Mcm21 and Okp1-complex could be assembled in vivo independently of the Ame1 essential protein. Moreover, we demonstrated that the Okp1 coiled-coil region per se is sufficient to form a complex with Ctf19 and Mcm21 proteins in vitro. The tetrameric COMA-complex could have also been reconstituted in vitro, but the amount and the stoichiometry of the components were not satisfactory. In solution, the COMA-complex showed oligomerization behavior. Through the protein purification experiments, we also found that the Ctf19, Mcm21 and Okp1-complex as well as the Ame1 protein separately may bind to unspecific, E.coli RNA. To determine other kinetochore subunits that can directly or indirectly associate with the COMA-components, we performed co-immunoprecipitation from yeast cells with either Okp1-TAP or Ame1-TAP tagged proteins. Among many known interacting partners, the Dsn1 component of the Mtw1-central kinetochore complex was identified. To support this finding and to test if the binding between the Dsn1 and the COMA-proteins is direct or indirect, we performed in vitro reticulocyte lysate binding assay. The interaction between the Mtw1- and the COMA-complexes via the Dsn1 protein was confirmed. Additional information has been gained from the co-immunoprecipitation experiments using budding yeast cells. We identified two proteins from the COMA-network, Nkp1 and Nkp2 proteins, as highly enriched. Since this may reflect the close proximity of these two proteins to the core of the COMA-network, we purified separately Nkp1 and Nkp2 dimer (which revealed the stabile heterodimer formation between these two Nkp proteins), combined it with the recombinant COMA-complex and reconstituted the hexameric protein complex at a 1:1:1:1:1:1 stoichiometry. Taken together, this study led to proposal of a new model for the spatial organization of the COMA-network. In summary, we used affinity based protein isolation to identify new direct binding partners within the central domain of the budding yeast kinetochore. Our findings improve the current understanding of the overall kinetochore architecture. The complete characterization of the kinetochore structure and organization has to be fully known, ultimately leading to three-dimensional vision and biochemical features of the kinetochore complexes, in order to unravel the mechanisms of chromosome segregation and maintenance of genome stability. Our work is therefore one step further in answering relevant biological and medical questions concerning faithful chromosome segregation during mitosis.