Ubiquitin-Mediated Degradation of Aurora Kinases

Abstract

The Aurora kinases are essential regulators of mitosis in eukaryotes. In somatic cell divisions of higher eukaryotes, the paralogs Aurora kinase A (AurA) and Aurora kinase B (AurB) play non-overlapping roles that depend on their distinct spatiotemporal activities. These mitotic roles of Aurora kinases depend on their interactions with different partners that direct them to different mitotic destinations and different substrates: AurB is a component of the chromosome passenger complex that orchestrates the tasks of chromosome segregation and cytokinesis, while AurA has many known binding partners and mitotic roles, including a well-characterized interaction with TPX2 that mediates its role in mitotic spindle assembly. Beyond the spatial control conferred by different binding partners, Aurora kinases are subject to temporal control of their activation and inactivation. Ubiquitin-mediated proteolysis is a critical route to irreversible inactivation of these kinases, which must occur for ordered transition from mitosis back to interphase. Both AurA and AurB undergo targeted proteolysis after anaphase onset as substrates of the anaphase-promoting complex/cyclosome (APC/C) ubiquitin ligase, even while they continue to regulate steps during mitotic exit. Temporal control of Aurora kinase destruction ensures that AurB remains active at the midbody during cytokinesis long after AurA activity has been largely eliminated from the cell. Differential destruction of Aurora kinases is achieved despite the fact that they are targeted at the same time and by the same ubiquitin ligase, making these substrates an interesting case study for investigating molecular determinants of ubiquitin-mediated proteolysis in higher eukaryotes. The prevalence of Aurora overexpression in cancers and their potential as therapeutic targets add importance to the task of understanding the molecular determinants of Aurora kinase stability. Here, we review what is known about ubiquitin-mediated targeting of these critical mitotic regulators and discuss the different factors that contribute to proteolytic control of Aurora kinase activity in the cell.

Keywords: APC/C; AURKA; AURKB; Aurora kinase; mitosis; ubiquitin-mediated proteolysis.

Figure 1

Events during mitotic exit that are influenced by Aurora kinases, illustrated against their degradation profiles. Blue, AurA; red, AurB.

Figure 2

Kinetics of Aurora kinase destruction during mitotic exit. (A) Total levels of Venus-tagged AurA and AurB in cells passing through mitosis. Data taken from Min et al. (47). Fluorescence measurements from single cells were used to generate averaged progress curves for the degradation of each substrate (n ≥ 50). (B) Plots of the changing rate of degradation over time for the averaged progress curves show that the maximum rate of AurA degradation is fivefold higher than that of AurB. (C) Simulation of first-order (Michaelis–Menten) kinetics predicts a theoretical degradation curve showing an exponential decrease in substrate levels over time that resembles the degradation curves that we have previously described for other substrates of the APC/C, such as Plk1, RacGAP1, and KIFC1 (16, 48), and is consistent with the idea that for these substrates, ubiquitination is the single rate-limiting step for proteolysis (since the rate of proteolysis depends on the amount of substrate present). (D) Modeling of distributive ubiquitination of a substrate, where a threshold number of stepwise ubiquitin modifications is required to generate the product that can be processed for proteasomal degradation, compared to a processive ubiquitination process. The simulated reaction exhibits the sigmoidal/switch-like response that characterizes degradation of AurA. (E) Schematic to explain kinetics of degradation of different substrates. Processive ubiquitination of substrates, such as Plk1, is achieved by a single binding event to the APC/C, and substrates rate limited by single-step ubiquitination are degraded with first-order kinetics. By contrast, Aurora kinases bind to the APC/C multiple times to acquire polyubiquitin chains. AurA, rate-limited by this multistep ubiquitination, exhibits switch-like degradation kinetics. Degradation of AurB is likely governed by a post-ubiquitination step.

Figure 3

Conserved degrons in Aurora kinases. (A) Aurora kinase degrons are the most conserved motifs outside the kinase domain. The sequence alignment of AurA and AurB was converted to a vector of values corresponding to conservation at each position (from 4 for fully conserved position to 0 for no conservation). Rolling averages of a five-residue window across the whole alignment is presented as a heat map. Therefore, the shade of red indicates residue conservation between the two paralogs. (B) Degradation plots for A-box- (including S51-) and D-box-mutated versions of AurA-GFP, as described in Ref. (81). Fluorescence levels measured over time in single cells exiting mitosis are normalized to anaphase onset. ΔA-box = Δ31–66; D-box mutant = R371A, L374A. (C) Mitotic localization of mutants analyzed in (B), showing loss of functional localization of the D-box mutant.

Figure 4

Schematic illustrating the complexity of APC/C substrate specificity after anaphase onset. Most substrates can be targeted by either APC/C–Cdc20 or APC/C–Cdh1, working with both Ube2C and Ube2S (blue stream). Cdh1-only substrates, although they can still bind Cdc20, show restricted targeting by the APC/C–Ube2S that depends on Cdh1 (red stream). APC/C–Ube2S is required to generate the K11 linkages for rapid degradation at the proteasome.