Contrôle spatio-temporel de la ségrégation des chromosomes – CHROMALIGN

Most solid tumors possess an abnormal number of chromosomes (aneuploidy), which is caused by defects during chromosome segregation. The research in our laboratory is focussed on understanding of the mechanisms that allow successfull chromosome alignment, and therefore accurate chromosome segregation. Once the bipolar microtubule spindle is assembled, most of the regulatory processes responsible for chromosome alignment operate at the kinetochore, the proteic interface between the primary constriction of the chromosomes (centromere), and the microtubule bundles emanating from opposite spindle poles. The kinetochore is also at the center stage of the mitotic checkpoint, which ensures that all chromosomes are correctly bi-oriented onto the mitotic spindle, before allowing the cells to proceed through anaphase. Although it is now known that about ten protein kinases are involved in chromosome alignment and/or segregation (for review, see Musacchio and Salmon, 2007), the signalling cascade taking place at the kinetochore remains elusive. Electron tomography provided a view of the kinetochore as a mesh (Dong et al., 2007) constituted of several low affinity microtubule-binding sites (Cheeseman et al., 2006), and the interplay between the different players is barely starting to emerge. We recently contributed to this issue by studying the regulation of the Xenopus kinetochore motor Cenp-E, an essential player of chromosome alignment (Wood et al., 1997; Schaar et al., 1997) and checkpoint signalling (Abrieu et al., 2000; Mao et al., 2003; Weaver et al., 2003). We showed that Cenp-E autoinhibition is relieved by MPS1 and/or CDK1-cyclin B dependent phosphorylation in vitro (Espeut et al., 2008). The next obvious steps are to understand whether this mechanism is conserved in vivo, and also applies to human Cenp-E. In order to fulfil these tasks we have started to identify the MPS1 and/or CDK1-cyclin B phosphorylation sites onto Cenp-E. We will generate Cenp-E phospho-mutants and we will analyze their motility in vitro. We will also use these Cenp-E mutants to reconstitute Xenopus egg extracts depleted from endogenous Cenp-E. In parallel, we will develop the tools to study whether this mechanism is conserved for human Cenp-E, both in vitro and in vivo. This required the characterisation of the substrate specificity of MPS1, in order to predict the MPS1-dependent phosphorylation sites onto human Cenp-E. For this purpose, we identified MPS1 phosphorylation consensus sequence and we found that it closely resembles to the one for PLK1. Yet, we showed that while Cenp-E can be phosphorylated onto the common sites by MPS1 and PLK1, some sites were found MPS1 specific. Importantly, the later sites were shown to regulate Cenp-E motility in vitro (Gaussen et al., submitted). MPS1 is a kinetochore-associated kinase essential for both chromosome alignment (Jelluma et al., 2008) and checkpoint signalling (Abrieu et al., 2001; Stucke et al., 2002). It is known that the kinase is hyperactived and hyperphosphorylated in mitosis (Stucke et al., 2002), but the regulatory mechanisms underlying these observations are unknown. We thus started an in-depth study of MPS1 regulation by phosphorylation. We identified and mutagenized eleven phosphorylation sites onto Xenopus MPS1, and we showed that the phosphorylation of one new site is essential for its activity (Morin, Abrieu, unpublished). Strikingly, we can restore MPS1 activity by introducing mimicking-phosphorylation acidic amino acid in that position. We will now dissect how phosphorylation of these sites affect 1) MPS1 kinetochore localization, 2) the mechanisms leading to the recruitment of other kinetochore components, 3) chromosome alignment and/or 4) checkpoint signalling. Altogether, our studies will uncover new regulatory mechanisms governing chromosome alignment, segregation, and therefore will contribute to our understanding of aneuploidy. Indeed, the proteins that we are studying are putative therapeutic targets. Inhibitors of MPS1 are starting to be developed (Dorer et al., 2005; Schmidt et al., 2005), and our detailed analysis of MPS1 activating mechanisms could contribute to rationally design more specific MPS1 inhibitors. For instance, thanks to our molecular modelling of MPS1 kinase domain (collaboration Andrey Kajava, CRBM), we might have identified the region that accounts for the similarity between MPS1 and PLK1 phosphorylation consensus. This valuable information could be used to target both MPS1 and PLK1. Since several PLK1 and Cenp-E inhibitors are being tested in phase I clinical assays (Steegmaier et al., 2007; Schmidt et al., 2007; Chua et al., 2007; Sutton et al., 2007), we think that our studies could contribute to their improvement. For this goal, we wish to determine whether the Cenp-E autoinhibitory mechanism (via Cenp-E motor-tail binding) that we uncovered could be linked to the effect of this allosteric Cenp-E inhibitor.