Résumé
Entry and progression through mitosis has traditionally been linked directly to the activity of cyclin-dependent kinase 1 (Cdk1). In this study we utilized low doses of the Cdk1-specific inhibitor, RO3306 from early G(2) phase onwards. Addition of low doses of RO3306 in G(2) phase induced minor chromosome congression and segregation defects. In contrast, mild doses of RO3306 during G(2) phase resulted in cells entering an aberrant mitosis, with cells fragmenting centrosomes and failing to fully disassemble the nuclear envelope. Cells often underwent cytokinesis and metaphase simultaneously, despite the presence of an active spindle assembly checkpoint, which prevented degradation of cyclin B1 and securin, resulting in the random partitioning of whole chromosomes. This highly aberrant mitosis produced a significant increase in the proportion of viable polyploid cells present up to 3 days post-treatment. Furthermore, cells treated with medium doses of RO3306 were only able to reach the threshold of Cdk1 substrate phosphorylation required to initiate nuclear envelope breakdown, but failed to reach the levels of phosphorylation required to correctly complete pro-metaphase. Treatment with low doses of Okadaic acid, which primarily inhibits PP 2A, rescued the mitotic defects and increased the number of cells that completed a normal mitosis. This supports the current model that PP 2A is the primary phosphatase that counterbalances the activity of Cdk1 during mitosis. Taken together these results show that continuous and subtle disruption of Cdk1 activity from G(2) phase onwards has deleterious consequences on mitotic progression by disrupting the balance between Cdk1 and PP 2A.