The ultimate goal of mitosis is to physically separate the entire set of duplicated chromosomes in order to propagate the genetic information during cell division. The perpetuation of a faithful maintenance of the genome integrity can only be guaranteed by the existence of multiple cell cycle checkpoints.1 In particular, mitosis is a highly regulated process where checkpoints assume a major importance.2 The latest stages of mitosis have been extensively studied; however, the mechanisms underlying the coordination between chromosome separation during anaphase and mitotic exit remain to be unveiled. It has long been assumed that anaphase and telophase are events that occur passively after satisfaction of the spindle-assembly checkpoint (SAC).2 However, it seems counterintuitive to assume that these crucial moments of mitosis are not being tightly monitored. A basic assumption for successful chromosome segregation during anaphase is that the 2 sets of chromosomes effectively separate before nuclear envelope reformation (NER). Recently, we showed that an inverse correlation exists between chromosome separation velocity and anaphase duration in Drosophila and human cultured cells, suggesting the presence of an additional surveillance mechanism that spatially regulates the anaphase-telophase transition.3 Importantly, spatial regulation implies the presence of a sensory ruler capable of “measuring” the chromosome position along the division axis. In fact, we showed that the previously reported Aurora B phosphorylation gradient4 creates an “area of exclusion” that inhibits DNA decondensation and NER (Fig. 1). When global Aurora B activity is impaired, or its localization at the spindle midzone affected, this spatial regulation is lost and NER occurs independently of the chromosome position. The precocious assembly of the nuclear envelope around unseparated chromosomes may lead to aneuploidy and/or polyploidy that will ultimately perturb cell or tissue homeostasis. Figure 1. After SAC satisfaction, mitotic cells proceed into anaphase. Aurora B is transferred to the spindle midzone and establishes a phosphorylation gradient where its substrates are phosphorylated and chromosomes remain condensed. This phosphorylation gradient ... Although Aurora B activity dictates a spatial control of the anaphase-telophase transition, the prolonged extension of anaphase after the expression of non-degradable Cyclin B1 mutants indicates that a final step of degradation is required to determine true mitotic exit. Additionally, inhibition of Cdk1 activity during anaphase shortens anaphase duration, suggesting that Cyclin B1 is necessary to provide time for proper chromosome separation. Therefore, both Aurora B and Cdk1 kinase activities must be coordinated to assure the spatiotemporal control of the anaphase-telophase transition. An additional level of regulation is related to the phosphatases, which must be in place to counteract the kinase activities. Inhibition of PP1/PP2A at the anaphase onset caused an “endless anaphase” phenotype, with persistent condensed chromosomes and no evident NER. Our data suggests that the kinase activity is predominant in the spindle midzone and, as chromosomes move to opposite poles, phosphatase activity will prevail and account for the dephosphorylation of substrates and consequent mitotic exit (Fig. 1). Some outstanding questions still remain and in the future, it would be important to understand how AuroraB, Cdk1 and PP1/PP2A activities are mechanistically coordinated: is there a feedback loop combining these 3 players? Or is there a hierarchical series of phosphorylation/dephosphorylation events that promote the balanced activity that ultimately leads to accurate chromosome segregation and NER? In the context of a potential chromosome separation checkpoint it will be important to understand which are the substrates that mediate the anaphase-telophase transition in coordination with chromosome separation. In this context, the localization of the Drosophila Condensin I subunit Barren, whose loading onto chromosomes is dependent on Aurora B,5 was found to be sensitive to the midzone Aurora B phosphorylation gradient (Fig. 1). Additionally, Barren depletion caused a similar phenotype to the one found with Aurora B inhibition, further supporting that chromosome condensation could be a readout for chromosome position during anaphase. Overall, sister-chromatid separation during anaphase is actively monitored by the presence of a phosphorylation gradient of Aurora B established at the spindle midzone. This hub of kinase activity constitutively phosphorylates its substrates as long as they are under the influence of the gradient. The inhibitory action of Aurora B activity must therefore feedback to Cdk1 in order to dampen the kinase activity and allow the phosphatases to prevail. The proposed feedback control of the anaphase-telophase transition is just the tip of the iceberg - how it is molecularly regulated and whether this mechanism is conserved between different cells in different tissues and organisms will certainly be a direction of further research.