The process by which cells divide consists of a cyclical series of phases, and the transition from one to the next is generally controlled by the expression, modification and destruction of various proteins. The switch from G1 phase to S phase in yeast, for instance, requires the breakdown of the Sic1 protein. Writing in the 29 November issue of Nature, Mike Tyers, Tony Pawson and colleagues show that this process has a built-in timer mechanism, which involves the accumulation of at least six phosphate groups on Sic1.

Yeast cells are driven into S phase by a complex that consists of a cyclin-dependent kinase (known as Cdc28) and its cyclin partners (Clb5/Clb6). During G1 phase, this complex is inhibited by Sic1, which ensures that cells do not enter S phase prematurely. At the G1/S transition, Sic1 is destroyed and Cdc28–Clb is no longer blocked. A prerequisite for the destruction of Sic1 is that it is phosphorylated (by Cdc28 in complex with another set of cyclins, Cln1/Cln2/Cln3); this enables Sic1 to be recognized by the cellular protein-degrading machinery through a protein called Cdc4. It was already known that Sic1 is phosphorylated many times before it can be detected and destroyed. But what wasn't clear was why this was the case.

To find out, Tyers and colleagues mutated the nine sites in Sic1 that can be phosphorylated by Cln–Cdc28. They then added the sites back, one at a time. They showed that the protein needs at least six of these sites before it can bind to Cdc4 or be degraded in vivo. Next, they identified a consensus amino-acid sequence surrounding a phosphorylation site that can bind to Cdc4 with high affinity. Cyclin E has this sequence, but none of the nine phosphorylation sites in Sic1 fits the consensus perfectly. Yet there is strength in numbers. Individually, these sites in Sic1 bind only weakly to Cdc4, but together they make a high-affinity Cdc4-binding site.

Moreover, it is biologically important that Sic1 has so many phosphorylation sites, as Tyers and colleagues discovered when they created a Sic1 protein that contained the single, high-affinity Cdc4-binding site from cyclin E. The hybrid protein was destroyed too efficiently, leading to premature S-phase entry and consequent genome instability.

All of which leads to the idea that cells time their entry into S phase to perfection by using a phosphate-dependent 'fuse'. If a single phosphorylation site were sufficient for Sic1 to be recognized and destroyed, then even a small amount of Cln–Cdc28 activity would drive cells into S phase, perhaps prematurely. But the fact that at least six sites need to be phosphorylated before Sic1 is destroyed ensures that a threshold level of Cln–Cdc28 activity must be reached. The use of this timer mechanism also means that the transition into S phase is inherently switch-like, in that it depends on the sixth order of Cln–Cdc28 kinase concentration in a highly nonlinear response.