How mitotic spindle, kinetochores and chromosomes are coordinated in space and time to ensure a faithful and accurate cell division?
Our group investigates the fundamental mechanisms of cell division ensuring faithful chromosome segregation in human cells, how a deregulation of these mechanisms contribute to genetic instability in cancer cells, and how this erroneous process can reveal new targets for anti-cancer treatments or be exploited by novel anti-cancer drugs.
Originally, we studied chromosome segregation from the perspective of kinetochores, the microtubule-attachment sites on chromosomes (see e.g. Amaro et al, NCB, 2010, McClelland et al., EMBO J., 2007, Jaqaman et al., JCB, 2010, Mchedlishvili et al., JCS 2012 or Vladimirou et al, Dev Cell, 2013), in the last years we aimed to obtain a more integrated view that considers all the elements of the mitotic machinery, including centrosomes, the different microtubule populations of the mitotic spindle, kinetochores, chromosomes, and the cell cortex.
How centrosomes and spindle poles contribute to the overall regulation of the mitotic spindle?
In earlier studies we had demonstrated that centrosome age regulates the stability of kinetochore microtubules (Gasic et al., Elife 2015), and that the number of centrioles can control spindle size and the location of the cytokinetic furrow (Tan et al., Elife 2015). In a larger effort to understand the role of centrosomes in the spindle, we established an experimental system based on cells with only 1 centrosome, which form asymmetric bipolar spindles containing one centrosome in one pole and no centrosome in the other pole. By comparing both types of half-spindles in the same cell, we demonstrated that centrosomes are major regulators of spindle length and microtubule dynamics, and that these effects reach all the way to kinetochores, by regulating kinetochore-microtubule plus-end dynamics. Our work identified the spindle associated protein HURP as a major intermediary between centrosomes and kinetochore-microtubule plus ends, as its presence reinforced spindle asymmetry and was required that differential plus-end microtubule dynamics (Dudka et al, Curr. Biol. 2019). Pursuing this work, we recently found in collaboration with Andrew McAinsh (Warwick University) that endogenously tagged HURP has a unique binding pattern on kinetochore-microtubules: it is excluded from the growing microtubule binding parts (forming a HURP-gap) due to its preferentially binding to GDP-tubulin over GTP-tubulin. HURP’s binding preference was validated both in vitro with recombinant proteins, and in vivo with the exogenous expression of WT and non-hydrolyzable tubulin mutants (Castrogiovanni et al., Nat Commun, 2022). We are particularly excited as this study revealed that for the first time that mitotic spindles contain large micron-sized zone that are neither GDP-tubulin positive, nor GTP-tubulin positive (EB3-negative), suggesting the existence of large mixed-nucleotide zones. We speculate that the length of the HURP-gaps can serve as a readout for the GTP-tubulin hydrolysis on k-fibres. In another centrosomal project, we studied the function of the microcephaly-linked protein WDR62 and found that it acts as an anchor protein at spindle poles and microtubule minus-ends for the microtubule-severing enzyme katanin. This localization is required to drive efficient minus-end depolymerization and poleward microtubule flux to ensure synchronous anaphase movements (Guerreiro et al, J. Cell Biol. 2021). Indeed, both WDR62 and Katanin loss slow down microtubule minus-end depolymerization and result in lagging chromosomes in anaphase. Our results point to a general pattern, as WDR62 is, after ASPM, the second microcephaly linked gene shown to be required for katanin localization at spindle poles and for poleward microtubule flux function (Jiang et al., NCB 2017).
How kinetochore-microtubule regulation influences the mitotic spindle?
In terms of kinetochore-microtubule regulation, we used a novel microtubule-interfering drug, BAL27862, to investigate the role of microtubule occupancy at kinetochores. Kinetochores bind on average 20 microtubules, which is vastly more than what is needed to move chromosomes. BAL27862 treatment reduced this number by 1/3, allowing us to probe for the role of high microtubule occupancy. Surprisingly, this did not prevent the satisfaction of the spindle assembly checkpoint, which is engaged by unattached kinetochores, indicating that kinetochores can silence the checkpoint at partial occupancy. Instead, we found that full microtubule occupancy favors the resolution of merotelic chromosomes (one kinetochore bound by microtubules from both poles) in anaphase, by providing a sufficiently high force to prevent an extended tug-of-war situation (Dudka et al., Nat. Commun. 2018).
Which mechanisms drive the formation and the orientation of the mitotic spindle?
While in earlier work we studied the contribution of kinetochores to bipolar spindle assembly (Toso et al., JCB, 2009), more recently, we established in collaboration with Monica Gotta (Univ. of Geneva) for the first time a link between cell polarity and formation of the mitotic spindle, as we found that cell polarity controls centrosome separation speed in two-cell stage C. elegans embryos. This differential regulation depends on the asymmetric distribution of the mitotic kinase Plk1, which controls the abundance of the dynein-interactor NuMA at the cell cortex (Bondaz et al., J. Cell Biol. 2019). We also showed with the Gotta group that the AAA-ATPase p97 substrate adaptor p37 regulates spindle orientation in human cells by limiting the amount of NuMA at the cell cortex via the protein phosphatase PP1/Repoman (Lee et al., J. Cell Biol 2018), revealing a new pathway controlling spindle orientation.
What is the link between human cell division and cancer?
We recently investigated how mild replication stress, an early cancer hallmark defined by a reduced DNA replication speed, leads to chromosome segregation errors. An earlier study had shown that replication stress led to whole chromosome gain/loss in mitosis (Burell et al, Nature, 2013); the mechanism, however, was unknown. Our group demonstrated that mild replication stress in non-transformed cells induces premature dis-engagement of centrioles within centrosomes. Normally the two centrioles in a centrosome disengage at the end of mitosis to allow centrosomes to duplicate. In our study we found that this dis-engagement occurred already at mitotic entry, resulting in the formation of transient multi-polar spindles favouring chromosome mis-segregation (Wilhelm et al., Nat Commun. 2019). This deregulation sensitized cells to drugs that prevent the clustering of extra spindle poles, such as taxol, indicating that this phenotype could be exploited. In parallel, we initiated a major collaborative effort at better targeting cancer cells with the group of Patrycja Nowak-Sliwinska (Univ. of Geneva). In particular we characterized a new combinatorial anti-cancer treatment that selectively targets cancer cells with extra centrosomes and/or multipolar spindles (Weiss et al., Cancers, 2019 & European patent EP19199136), and we contributed to other studies of the Nowak-Slinwinska group, studying the link between cell division and anti-cancer drugs (Berndsen et al., Sci Rep. 2017; Berndsen et al., Br. J. Cancer 2019).