The transition from mitosis (M phase) to G1 phase is controlled by complex interactions between two key components of the cell-cycle machinery, the anaphase-promoting complex or cyclosome (APC/C) and cyclin A. The APC/C is a multisubunit ubiquitin ligase that controls sister-chromatid separation and triggers mitotic exit through the ubiquitin-mediated degradation of proteins (including, in M phase, cyclin A). Cyclin A is an activator of the cell-cycle regulators CDK1 and CDK2 and, in late G1 phase, is also responsible for APC/C inactivation.

How cyclin A can re-accumulate in the presence of active APC/C during G1, and then inactivate the APC/C to allow the re-accumulation of other mitotic cyclins during the DNA synthesis (S) and G2 phases, has been a puzzle. In Nature, Michael Rape and Marc Kirschner now report that they have solved the mystery, by studying the effects that different components and regulators of the APC/C have on the degradation of cyclin A.

Initially, using extracts from synchronized HeLa cells, the authors showed that, whereas other APC/C substrates were consistently degraded throughout G1 phase, cyclin-A degradation gradually became blocked as cells progressed through G1. The degradation of cyclin A relied specifically on the presence of a threshold level of the ubiquitin-conjugating enzyme (E2) UBCH10, whereas other APC/C substrates were still degraded in the presence of another E2, UBCH5. In keeping with this, the authors showed that the level of UBCH10 (but not UBCH5) fluctuated during the cell cycle, remaining high during cyclin-A degradation and staying low when cyclin A was stable.

Intriguingly, it was then found that UBCH10 was autoubiquitylated before also being degraded by APC/CCDH1 towards the end of G1. However, this process was hampered both by the slow rate of UBCH10 autoubiquitylation and by the presence of other APC/C substrates. In fact, the authors showed that, as long as other APC/C substrates are present in G1, UBCH10 preferentially ubiquitylates them, rather than itself, which effectively allows UBCH10 to orchestrate the timing of cyclin-A accumulation through its own, carefully timed, self destruction.

So, Rape and Kirschner have identified what they suggest is the 'self-perpetuating oscillator' that pushes the cell cycle through its phases (see figure). Accumulation of cyclin-A–CDK1 and cyclin-B–CDK1 promotes entry into mitosis and activation of APC/CCDC20, which, in turn, induces degradation of the mitotic cyclins and mitotic exit (a). APC/C acquires the CDH1 adaptor in G1 and degrades the remaining APC/CCDH1 substrates (b), which facilitates UBCH10 autoubiquitylation and degradation (c). Cyclin A then re-accumulates (d) and completely inactivates APC/CCDH1 (e), thereby initiating entry into S phase and allowing further mitotic-cyclin accumulation (f). So, as the saying goes: what goes around comes around...