Biochemical mechanisms cyclin-dependent kinases use to control cell division

Division Of Basic Sciences - Nci

Investigators

Abstract

Cancer is often described as a disease of uncontrolled cell division. Central to this process are cyclin-dependent kinases (CDKs) proteins that function like gas pedals, propelling the cell through its division cycle by phosphorylating key substrates. In contrast, CDK inhibitor proteins act as brakes, ensuring proper control of cell cycle progression. In cancer, this regulatory balance breaks down and the gas pedal gets stuck, the brakes fail, and cells divide uncontrollably. Our research seeks to understand how this control system operates in normal cells, how it becomes dysregulated in cancer, and how we can restore proper regulation to halt disease progression. This report summarizes our progress over the past year, organized around three research aims. Aim 1: Identify G1 cyclin-CDK substrates and define docking mechanisms in G1/S control Background: Cyclin D-CDK4/6 complexes are essential drivers of the G1/S transition, the point at which cells commit to division, but research has largely focused on a narrow set of targets, particularly the retinoblastoma (Rb) protein. Our goal is to map and characterize a broader set of cyclin D-CDK4/6 substrate interactions. Moreover, while current CDK4/6 inhibitors (e.g., palbociclib, ribociclib, abemaciclib) target kinase activity, they provide only modest benefit and are often met with resistance in patients. Therefore, a deeper, mechanistic understanding of how cyclin D-CDK4/6 engages its substrates is urgently needed to uncover new vulnerabilities and develop the next generation of targeted therapies. Hypothesis: Cyclin D-CDK4/6 phosphorylates a broader set of substrates beyond the Rb pathway through defined docking mechanisms, some of which may represent cancer-specific vulnerabilities. Experimental Systems: We use a chemical-genetic strategy with analog-sensitive CDK4/6 variants that accept bulky ATP analogs for selective substrate labeling. Quantitative mass spectrometry approaches are used to identify direct substrates. Complementary methods including crosslinking, proximity labeling, and AI-based predictions help validate and expand this substrate set. Progress: We developed analog-sensitive CDK4/6 variants, enabling direct substrate labeling in diverse set of cell lines. In non-transformed cells, cyclin D-CDK4/6 was found to phosphorylate hundreds of targets beyond Rb, including regulators of transcription, DNA repair, and RNA metabolism. Additionally, we have combined this approach with docking-deficient cyclin D mutants to distinguish canonical from non-canonical substrate engagement. Target validation is ongoing using biochemical and computational approaches focused on kinase specificity, docking mechanisms, and functional impact in G1. To explore CDK4/6 substrate functions beyond our expertise, we've partnered with other NCI labs including the Singer (transcription), Volkov (RNA metabolism), and Larson (RNA splicing) lab. Milestone: Our first independent publication revealed key differences between CDK4 and CDK6 in cell cycle regulation and drug sensitivity. Combining quantitative biochemistry and molecular dynamics simulations (in collaboration with the Nussinov lab), we showed that CDK6 has stronger coupling between its beta3-alfaC loop and G-loop, leading to higher activity and distinct inhibitor sensitivity. We also identified a new regulatory role for the unstructured CDK6 C-terminus. These insights refine our understanding of CDK4/6 regulation and provide a framework for designing more specific CDK inhibitors to improve targeted cancer therapy. Impact: This work provides a comprehensive map of cyclin D-CDK4/6 substrates, elucidates the structural basis of substrate recognition, and lays the foundation for next-generation therapeutics targeting CDK4/6 dependencies. Aim 2: Investigate promoter-specific regulation by cyclin-CDK complexes Background: CDKs and their cyclin partners are key regulators of the cell cycle, phosphorylating targets in a phase-specific manner. My recent work in yeast system found that G1 cyclin-CDK complexes directly phosphorylate RNA polymerase II at specific promoters, triggering transcriptional changes required for the G1/S transition. However, the broader role of these complexes in regulating transcription through promoter-associated proteins remains poorly understood. Hypothesis: We hypothesize that cyclin-CDK complexes regulate transcription in a promoter-specific manner by phosphorylating RNA polymerase II or other promoter-bound proteins, and that this is mediated by recruitment through sequence-specific transcription factors via docking motifs. Experimental Systems: We integrate biochemical assays, protein-DNA and protein-protein interaction mapping, transcriptomics, and AI-based screening methods (initial screen in collaboration with O'Reilly lab at NCI) to identify points of regulation across multiple cell types and model systems. Progress: We generated mammalian cell lines expressing affinity-tagged wild-type cyclins suitable for ChIP-seq. Initial experiments confirmed that G1-phase cyclin D binds to distinct chromatin regions, supporting the hypothesis that mammalian cyclins interact directly with DNA. Using AlphaFold2, we screened for co-structures between cyclin D and the entire human proteome, identifying new classes of interactors and candidate interface residues. These are being tested via targeted mutagenesis to generate separation-of-function cyclin D mutants that cannot bind DNA, enabling us to dissect recruitment mechanisms and their role in promoter-specific gene activation. Impact: As both cell cycle and transcriptional cyclin-CDK complexes are therapeutic targets, understanding how they cooperate to drive gene expression may reveal new strategies for combinatorial therapies. Cancer cells are often dependent on both uncontrolled proliferation and dysregulated transcription, making them especially vulnerable to dual-targeting approaches. Aim 3: Develop reversible tool compounds to disrupt cancer-specific cyclin-substrate interactions Background: Many cancers exhibit elevated cyclin-CDK activity due to the dysregulation of core cell cycle components such as activators, inhibitors, and key substrates. While most small molecule inhibitors developed so far target the CDK active site, less attention has been given to disrupting substrate recognition mediated by the cyclin subunit. Hypothesis: We hypothesize that cyclin-CDK complexes can be selectively inhibited by targeting specific cyclin-substrate docking interactions. Experimental System: We use experimental data as an input for AI-based models (AlphaFold2/3) to predict co-structures of cyclin-CDK complexes with their targets, identifying interaction interfaces and 3D binding pockets. These are validated through biochemical and cellular assays and used for virtual screening of small-molecule libraries. Progress: As a proof of principle, we validated our AI-based workflow in a yeast model system by identifying novel interaction pockets on the G1 cyclin Cln3, a functional homolog of mammalian cyclin D. We then extended this strategy to mammalian cyclin D, using data from CDK4/6 target identification for co-structure predictions. This approach revealed both known and novel docking sites on cyclin D that could potentially be targeted by small molecule compounds or peptidomimetics. In collaboration with the Tarasova lab at NCI, we launched virtual compound screens to identify small molecules that could target these identified pockets and disrupt cyclin D-target interactions. Promising candidates are currently being tested in biochemical and cellular assays. Impact: This work could lead to a new class of therapeutics that selectively block substrate-specific cyclin docking interactions, providing a novel strategy to target cancer-specific vulnerabilities.