Aurora B (Aurkb) is one of the major protein kinases that ensures the proper execution and fidelity of mitosis.1 A member of the chromosomal passenger complex, Aurkb has been implicated in various mitotic functions, including chromosome-microtubule interactions, sister chromatid cohesion, the spindle-assembly checkpoint and cytokinesis.2 As it is upregulated in several human cancers and correlated with poor prognosis, Aurkb is believed to be an important anti-cancer drug target. In this connection, a number of small-molecule inhibitors have been developed and are currently at various stages of clinical trials.3 Therefore the effects of Aurkb inactivation is of considerable interest. In this issue of “Cell Cycle,” Malumbres and colleagues examined the effects of Aurkb inactivation in mouse cells.4 As genetic ablation of Aurkb results in mitotic aberrations and lethality after implantation in mice,5 the authors made use of conditional knockout mouse embryonic fibroblasts and chemical inhibition to tackle the issue. These tantalizing results establish a linkage between Aurkb and another major cell cycle regulator, cyclin-dependent kinase, through the CDK inhibitor p21Cip1/Waf1. Trakala et al. discovered that although mitotic entry is unaffected in Aurkb-deficient MEFs, the majority of the cells are unable to form a metaphase plate and exit mitosis prematurely. Moreover, these Aurkb-deficient MEFs also exit Taxol-mediated mitotic block precociously. One of the characteristic features of premature mitotic exit is the formation of polyploid cells, which can lead to cell death or genome instability in the subsequent division cycles. This may in part contribute to the increase in tumor incidence in Aurkb heterozygous mice.5 Premature mitotic exit in mammalian cells typically involves mitotic slippage, which is caused by the gradual destruction of cyclin B1 during the mitotic arrest.6 Intriguingly, the premature mitotic exit in the absence of Aurkb is associated with high expression of cyclin B1, suggesting that the process may resemble adaptation instead of classic slippage. The observation that the CDK inhibitor p21Cip1/Waf1 is induced after Aurkb inactivation provides a possible mechanistic basis of the premature mitotic exit.4 Indeed, downregulation of p21Cip1/Waf1 reverses the unscheduled mitotic exit. Yet a conceptual obstacle for p21Cip1/Waf1 in causing premature mitotic exit is that the canonical p21Cip1/Waf1 pathway is well established to be involved in interphase arrest (such as after DNA damage). To extricate from this problem, Trakala et al. proposes that the p21Cip1/Waf1 induced after Aurkb inactivation is at a level that is insufficient to prevent mitotic entry but result in premature exit due to partial inhibition of Cdk1. A critical question is why inhibition of Aurkb leads to an accumulation of p21Cip1/Waf1. Aurkb has been shown to phosphorylate p53 and downregulate its transactivation activity and protein stability.7,8 Hence inhibition of Aurkb is expected to activates p53 and its downstream targets such as p21Cip1/Waf1. In addition, it is conceivable that the mitotic stress induced after Aurkb inhibition can also lead to p53 activation. A somewhat unanticipated result of Trakala et al. is the pronounced effect of Aurkb-deficiency on interphase. Entry into S-phase from quiescence is delayed in Aurkb-deficient MEFs or after treatment with the Aurora kinase inhibitor ZM447439. A reduction of Aurkb in heterozygous mice also delays cell cycle entry after partial hepatectomy or skin wound healing. Although not tested directly, the accumulation of p21Cip1/Waf1 after Aurkb inhibition may blunt the activity of Cdk2, thereby causing the observed delay in G1-S transition. However, as Aurkb is degraded at the end of mitosis by APC/C-mediated ubiquitination, one has to speculate that the Aurkb present during interphase is adequate to suppress the accumulation of p53 and p21Cip1/Waf1 during normal cell cycle. A recurring theme in the regulation of mitosis is the inextricable links between the key players. Although the biological significance of the regulation of p53-p21Cip1/Waf1 axis by Aurkb remains to be defined, these observations suggest the possibility of a novel mechanism that regulates CDKs in both mitosis and interphase. In addition to the effects on the cell cycle, other consequences of p53 activation after the Aurkb inactivation will also be interesting for further investigation. Finally, whether a similar mechanism is present in cancer cells, and whether the loss of p53 confers different sensitivity to Aurkb inhibitors may have important implications in cancer therapies. Figure 1. A model of links between Aurkb and cyclin-dependent kinases.