KT-MT interface · Molecular organization of the kinetochore-microtubule interface

Molecular organization of the kinetochore-microtubule interface

The faithful division of replicated DNA during mitosis is one of the most intriguing biological processes. The attachment of chromosomes to microtubules is a heavily regulated step of this process. The main challenge for the cell is to ‘sense’ the incomplete and incorrect attachments, and subsequently correct them if necessary. Only after all chromosomes are attached correctly, the cell cycle, which is halted in mitosis through the activity of the spindle assembly checkpoint (SAC), is allowed to resume. The link between the microtubules and chromosomes is not direct, but is mediated by a large multi-subunit and multi-copy protein complex named the kinetochore. By providing both the microtubule binding capacity and signaling platform for the SAC, the kinetochore is at the apex of several essential mitotic processes. In this challenging project we aimed to get a deeper understanding of the kinetochore architecture and its functions. To this end, we have (in a lab-wide effort) reconstituted kinetochore particles from recombinantly expressed and purified proteins on the centromere-specific CenpA-nucleosome. So far, we have reconstituted kinetochore particles up to 26 components in solution and on solid surfaces (glass and beads) by using a biotinylated SNAP fusion with CENP-C as bait. These reconstitution efforts, combined with labeling specific subcomplexes with fluorescent moieties with complementary emission spectra, allowed us to map the architectural hierarchy within the kinetochore. These multi-colored kinetochore particles can also used to study the microtubule binding behavior of the full complex or smaller subcomplexes. In these assays, taxol-stabilized microtubules are immobilized to the streptavidin-coated bottom of a glass flow-cell. By monitoring the co-localization of microtubules and protein fluorescence, we could confirm the integrity and stability of our recombinantly reconstituted kinetochores. Moreover we confirmed a preference for CenpA-containing nucleosomes over the canonical H3-containing nucleosomes. This assay will now be used to study minimal requirements of microtubule binding and the cooperativity between different microtubule binding proteins. The second main function of (unattached) kinetochores is to serve as a binding platform for the SAC signaling. This SAC generates a ‘wait’ signal that halts mitosis until all chromosomes are bi-oriented. The biochemical manifestation of this process is the Mitotic Checkpoint Complex (MCC), which is a potent inhibitor of the APC/C. Inhibiting the E3 ubiquitin ligase APC/C prevents CyclinB and Securin degradation and subsequently inhibits progression into anaphase. However, although the MCC complex does form spontaneously, the formation kinetics (several hours) are too slow to explain to create the responsive inhibitory system seen in vivo (several minutes). The MCC consists of four proteins (Cdc20, Mad2, BubR1 and Bub3), and especially the Mad2-Cdc20 interaction is particularly slow due to an obligatory and energetically unfavorable conformational change in Mad2. To solve this paradox, it has been proposed that the kinetochore acts as a catalyst to the formation of MCC. Using this elegant solution, the unattached kinetochore can quickly create an abundance of MCC, and thus inhibiting the APC/C. This property is lost when chromosomes bi-orient, and due to the constant generation of new Cdc20, this shifts the equilibrium to free Cdc20, which will activate the APC/C and the cell will progress into anaphase. To identify the requirements and components of the catalytic function of the kinetochore, we aimed to reconstitute this process in vitro. We created FRET-based sensors that measure the kinetics of MCC formation and disassembly. We created two FRET sensors by fusing CFP to either Cdc20 or BubR1 and TAMRA to Mad2, which measure the binding of Mad2 to either BubR1 or Cdc20. Using equilibrium experiments, we could measure the binding constants and confirmed that the MCC is a cooperative assembly. Upon mixing the individual MCC proteins, we could also monitor the generation of the FRET signal over time. This was used to determine both assembly (kon) and disassembly kinetics (koff), confirming that the complex indeed forms and disassembles very slowly (typically about 10 hours to completion depending on conditions). After adding the purified recombinant SAC proteins MPS1, Bub1/Bub3 and Mad1/2, we could increase the assembly kinetics dramatically, where the MCC complex now forms in about 3 minutes. Similarly, we also observed catalysis of MCC disassembly. After preforming MCC overnight without catalyst, we added a dark FRET competitor (Mad2) in large excess. By monitoring the disappearing FRET signal, one can determine the disassembly rate constant. When monitoring MCC disassembly on its own, however, even after eight hours the observed complex turnover was minimal. Yet, when we added the AAA+-ATPase TRIP13 together with p31 (both proteins previously reported to be involved in MCC disassembly) greatly accelerated the dismantlement of the MCC in an ATP dependent manner. Overall, we have produced very exciting reagents that we will allow for very detailed and controlled studies on essential processes in mitosis. Firstly, we have produced >20 subunit kinetochore particles on centromere specific nucleosomes solely from recombinant and purified material. We have confirmed their integrity and ability to bind microtubules. Secondly, using a FRET-based approach we have now been able to reconstitute and directly study MCC formation kinetics, a process that lies at the heart of cell cycle control.

Data: CORDIS, © European Union