Abstract
Cyclin-dependent kinases (CDKs) play a key role in activating essential cell biology processes including cellular proliferation. Inappropriate regulation of CDKs has been implicated in driving several different forms of cancer. One of the regulatory factors is the need to bind to Cyclin-partners before they can be activated and advance the cell cycle. Cyclins are overexpressed in several different cancers, hence activating their relevant CDK. Hyperactive CDK2 in particular is implicated in many cancers, and while many drugs have shown preclinical promise, none have successfully passed through clinical development. Among the complications of targeting CDK2 is the fact that non-classical cyclin partners from the Speedy/RINGO family of proteins can alter the conformation of the kinase. Using computational approaches, we provide data supporting that the active site of CDK2 differs when bound to Spy1 as compared to classical cyclins. Furthermore, combining computational models with experimental techniques we provide data that many small molecule inhibitors have reduced activity against Spy1-bound CDKs. This work supports the need to develop new inhibitors capable of inhibiting the Spy1-CDK2 complex, and suggests that computational tools can be beneficial toward accomplishing this goal.