Content area
Abstract
The rise and fall of cyclin-dependent kinase (CDK) activity coordinates progression through the eukaryotic cell division cycle. Human cells have multiple CDKs that bind preferred cyclin partners in a defined order, suggesting that individual CDKs have distinct roles in vivo. This model was challenged by the discovery that Cdk2, 4, and 6 are dispensable for proliferation and early development. Cdk1 replaces the missing interphase CDKs by binding non-canonical cyclin partners. Thus, a competing model was proposed in which passage through the G1/S and G2/M transitions depends exclusively on Cdk1.
During my PhD I sought to further our understanding of the cell cycle's organizing principles by testing these models. The latter predicts that Cdk1/cyclin complexes form early in the cell cycle, but I found cyclin binds Cdk2 over Cdk1 until mid-S phase. This suggests a "scaffold" function for Cdk2 that prevents premature Cdk1 activation. I hypothesized that when CDKs become active depends on how they form complexes with cyclins. CDK activity requires both cyclin binding and phosphorylation by a CDK-activating kinase (CAK). The only known CAK in metazoans is Cdk7, but it was unclear if it is required for both early and late cell cycle events. Using a chemical genetic approach I showed that Cdk7 activates both Cdk1 and Cdk2, but Cdk2 follows a kinetically distinct and favored activation pathway in vivo. Allowing Cdk1 to follow the Cdk2 pathway switches the binding preference of cyclin A, suggesting that Cdk2's scaffold function is derived from its unique activation pathway.
These studies suggest that Cdk2 activity should be required, but cells can proliferate without Cdk2 protein (i.e. when both catalytic and scaffold functions are lost). To test if Cdk2 activity is required when cyclin pairing is normal, I used cell lines in which Cdk2 was replaced with a drug-sensitive allele. By manipulating Cdk2 activity with small molecules, I showed that: 1) Cdk2 activity is required for proliferation; and 2) Cdk2 is a non-redundant regulator of S phase entry. Therefore, cyclin-pairing rules confer essential functions on Cdk2 and prevent premature Cdk1 activation, arguing against a role for Cdk1 in early cell cycle decisions.
