Molecular insights into kinetochore-microtubule attachments in mitosis

Abstract

During mitosis, cells must ensure that chromosomes are correctly segregated to both daughter cells. Chromosomes must attach microtubules from opposite spindle poles and congress to the metaphase plate. Once all the chromosomes are bioriented, the cell proceeds to anaphase and chromosomes migrate to opposite poles. Microtubules are dynamic filaments that grow and shrink. It is essential that chromosomes maintain their attachment to microtubules during mitosis to allow for correct chromosome segregation. Chromosomes are attached to microtubules of the mitotic spindle via their kinetochore, a multiprotein complex scaffold that assembles onto centromeric DNA. First, I studied the molecular basis of outer kinetochore-microtubule attachments in the budding yeast Saccharomyces cerevisiae. There, the Dam1 complex is an outer kinetochore-localised heterodecamer that maintains correct kinetochore attachment to a single incoming microtubule in vivo. The organisation of the ten different subunits within a single complex and how multiple complexes oligomerise is unclear. Furthermore, which domains of which subunits are responsible for microtubule binding remains under debate. Using cross-linking coupled with mass spectrometry, I shed light on the subunit organisation of Dam1 heterodecamers and obtained information on domains mediating oligomerisation. I demonstrated that the flexible C termini of both Duo1 and Dam1 subunits were essential for microtubule binding by designing mutants and assessing their microtubule-binding abilities. This work shows Duo1 and Dam1 subunits are placed in close proximity to the microtubule lattice and furthers our understanding of the overall structure of the Dam1 complex. I then focussed on kinetochore-microtubule attachments in humans. In higher eukaryotes, chromosomes align along the metaphase plate and biorient via different processes involving motors and microtubules. Some are immediately captured by microtubules while others move to the spindle poles before sliding from the poles to the metaphase plate, following the lattice of a kinetochore fibre. CENP-E is a large kinesin motor essential for chromosome congression. However, how CENP-E is recruited to unattached kinetochores remains poorly defined. Using biochemistry and cell biology, I showed that a minimal kinetochore-targeting domain of CENP-E interacts with the spindle checkpoint protein BubR1 directly in vitro and that the C-terminal helix of the pseudokinase domain of BubR1 is essential but not sufficient for CENP-E binding. Conversely, CENP-E necessitates a conserved acidic patch close to its C terminus for BubR1 binding. This precise mapping of the interaction allowed us to interrogate the role of BubR1 in recruiting CENP-E to kinetochores. We revealed that this interaction is essential for CENP-E recruitment in early mitosis and in metaphase. These experiments also highlight that CENP-E is recruited to kinetochores via another, uncharacterised, pathway. This work led to the third aim of my thesis: identify an alternative recruitment pathway of CENP-E to kinetochores. CENP-E localises to the outer corona of kinetochores. Interestingly, CENP-E detaches from kinetochores after a short treatment with a CDK1 inhibitor. CENP-E is then found in a protein complex that contains other detached corona proteins such as the RZZ complex, the Dynein adaptor Spindly and Dynactin. I identified a minimal domain of CENP-E that localises to these detachable structures and identified key residues essential for this localisation. By designing mutants, I showed that the kinetochore-targeting domain of CENP-E contains two kinetochore binding sites suggesting that two different pools of CENP-E are recruited to kinetochores. Finally, using siRNA, I showed that Dynein forms a second recruitment pathway of CENP-E to kinetochores. Overall, this work broadens the knowledge of kinetochore-microtubule attachment during mitosis in both yeast and humans. It provides key information that will allow future structural work on the Dam1 complex as well as future studies on the binding partners of CENP-E at kinetochores.