Aurora borealis wraps Plk1 and CDK together

Abstract

Proper execution of mitosis requires the phosphorylation of hundreds of functional proteins by multiple protein kinases. Activation and inactivation of these kinases needs to be coordinated so that the correct order and timing of mitotic events allows faithful chromosome inheritance. The intrinsic oscillation of one of these kinases, the cyclin-dependent kinase (CDK), has been shown to be composed of a negative feedback loop and a time delay.1 CDK first activates the anaphase-promoting complex/cyclosome (APC/C), an E3 ubiquitin ligase for mitotic cyclins, which, in turn, inactivates the CDK, forming a negative feedback loop. It takes about 15 min for the CDK to activate the APC/C by an unknown mechanism (an intrinsic time delay). In addition APC/C activation is further delayed either by the spindle assembly checkpoint in somatic cells or by a cyto-static factor (CSF) that is present in unfertilized eggs. In embryonic cycles, CDK activity oscillates as the cells enter and exit mitosis, so how are other mitotic kinases controlled during this period, while CDK activity remains high? Polo-like kinase 1 (Plk1) is one such mitotic kinase, and its activity also oscillates, peaking in mitosis like CDK. The protein levels of Plk1 are relatively stable in eggs and in early development stages, so its oscillating activity must be initiated by some other mechanism. This activation is performed by Aurora A (Aur A), another mitotic kinase, which phosphorylates and activates Plk1 on entering mitosis. A recent study by Feine et al. now reveals how Plk1 activity is coordinated with CDK activity after entering mitosis via the regulated stability of a Plk1-interacting factor, Aurora borealis (Bora).2 Bora was originally identified as a Drosophila mutant, which shows a phenotype (defective in asymmetric cell division) similar to that of a mutant of Aur A.3 When Bora is repressed in higher vertebrates, cells form multipolar mitotic spindles, and they have aberrantly segregated chromosomes. It was also shown that Bora specifically enhances Aur A-mediated T-loop phosphorylation of Plk1 in vivo and in vitro, which is required for full activation of Plk1.4 Stability of the Bora protein is downregulated by both CDK and Plk1. Initially Bora is phosphorylated at Ser252 by CDK, which then permits Plk1 to phosphorylate Ser497 and Thr501, which, in turn, leads to phosphorylation-dependent protein degradation by the SCFβTrCP–proteasome pathway.5,6 Therefore, Plk1 and Bora constitute a negative feedback loop, and the question posed is how is Bora stabilized in mitosis while the activity of CDK and Plk1, which target Bora for degradation, remain high? Their study reported a key finding which shows how the Bora protein is stabilized in mitosis.2 They took advantage of the frog egg system and used extracts in which the cell cycle can be arrested by CSF at meiotic metaphase-II, where levels of both CDK and Plk1 activities are high. Fertilization-induced cytoplasmic calcium ion flush releases the CSF arrest (allowing progression into interphase) by activating the calmodulin-dependent kinase-II and the calmodulin-dependent phosphatase, calcineurin.7 Hinted by a calcineurin-dependent destabilization of Bora, which is stable in CSF extract, they found that Bora is phosphorylated at Thr52 (and 14 other sites) in CSF extract. When they added a Thr52-to-Ala substitution (non-phosphorylatable) mutant protein of Bora to CSF extract it became less stable than wild-type even in the absence of calcineurin activity, suggesting that phosphorylation at this site (probably by CDK) protects it from Ser252 phosphorylation-dependent degradation of Bora. This finding gives us a simple model in which any time delay for CDK is copied for Plk1. That is, Plk1 is kept active as long as Bora is phosphorylated on Thr52 by active CDK. On fertilization, calcineurin activates the APC/C and causes Bora to be degraded, by dephosphorylating Thr52, leading to simultaneous inactivations of CDK and Plk1.2,7 Stabilization of Bora in mitosis was also confirmed in somatic human cells, suggesting that this mechanism is conserved among most higher eukaryotes. Further questions arise from this report. Which phosphatase is responsible for dephosphorylating Thr52 in somatic cells, where calcineurin does not get activated? How does dephosphorylation of Thr52 precede that of S252 to upregulate Bora degradation, even though both residues are apparently phosphorylated by CDK? Because Bora is related to the localizaiton6 and activation3 of Aur A, does Bora coordinate more mitotic kinase activities, such as Aur A?