Cyclin-dependent kinases (CDKs) are enzymes that act as master switches for cell division and gene activity. They work by attaching phosphate groups to other proteins, changing those proteins’ behavior, and in doing so they control when a cell copies its DNA, when it divides, and even how genes get read. The human genome contains 21 different CDK genes, and while many of them drive the cell cycle, others regulate processes that have nothing to do with division at all.
How CDKs Get Switched On
A CDK on its own is essentially inactive. It needs a partner protein called a cyclin to turn on, which is where the name “cyclin-dependent” comes from. When a cyclin binds to a CDK, it physically reshapes the enzyme’s structure. Part of the CDK called the T-loop normally blocks the active site like a door wedged shut. Cyclin binding pushes the T-loop out of the way and shifts another structural element closer to where the enzyme does its work, partially opening that door.
Partial activation isn’t enough for most CDKs, though. Full power requires a second step: another kinase (called CAK, for CDK-activating kinase) adds a phosphate group to a specific spot on the T-loop. In CDK2, for example, phosphorylation of a single amino acid (threonine 160) boosts catalytic activity roughly 300-fold. That phosphate acts as an organizing center, pulling together surrounding parts of the protein and reshaping the T-loop into a platform where target proteins can dock and get phosphorylated.
Not every CDK follows this two-step recipe. CDK8, which works in gene regulation rather than cell division, lacks the conserved threonine that other CDKs rely on. Instead, a subunit of a larger protein complex called Med12 physically stabilizes CDK8’s T-loop, standing in for the missing phosphorylation. CDK7 can also reach full activity without T-loop phosphorylation when it binds both cyclin H and a third protein called MAT1. These exceptions show that the family has evolved multiple ways to flip the same basic switch.
Driving Each Stage of Cell Division
The cell cycle has four phases: G1 (the cell grows and prepares), S (DNA is copied), G2 (the cell prepares to divide), and M (mitosis, the actual split). Different CDK-cyclin pairs take the lead at each transition, rising and falling in sequence like a relay team passing a baton.
- G1 phase: CDK4 and CDK6 pair with D-type cyclins (D1, D2, or D3). Their main job is to start disabling the retinoblastoma protein (Rb), a powerful brake on cell division. As CDK4/6 adds phosphate groups to Rb, it loosens Rb’s grip on transcription factors called E2Fs, allowing genes needed for DNA replication to begin turning on.
- G1-to-S transition: The newly freed E2Fs activate genes for cyclins E1 and E2, which bind CDK2. The cyclin E/CDK2 complex finishes phosphorylating Rb and triggers the cell’s commitment to DNA replication.
- S phase: Cyclin A replaces cyclin E and partners with CDK2. This complex helps coordinate DNA copying and, once replication is complete, shuts down S-phase signals by phosphorylating replication factors so they can’t fire a second time.
- G2 and mitosis: Cyclin A also partners with CDK1 to push the cell from G2 into mitosis. Then cyclin B takes over as CDK1’s partner, and the cyclin B/CDK1 complex orchestrates the dramatic physical events of cell division.
What CDK1 Does During Mitosis
The cyclin B/CDK1 complex is sometimes called the “master mitotic kinase” because it phosphorylates a wide range of proteins to trigger the visible events of cell division. Among its targets are proteins that make up the nuclear envelope, the double membrane surrounding the cell’s DNA. Phosphorylation of nuclear envelope and nuclear pore proteins causes the envelope to break apart, which is necessary for the chromosomes to be captured by the spindle fibers that will pull them to opposite ends of the cell. CDK1 also phosphorylates proteins involved in chromosome condensation, spindle assembly, and reorganization of the cell’s internal skeleton. Proteomic studies have identified CDK1 substrates at nearly every subcellular structure that changes shape during mitosis.
CDKs That Regulate Gene Transcription
Several CDKs have nothing to do with cell division. Instead, they control how RNA polymerase II (the enzyme that reads genes into messenger RNA) does its job. At least five CDKs participate at different stages of this process.
CDK7, paired with cyclin H, sits inside a complex called TFIIH that helps RNA polymerase II get started at the beginning of a gene. CDK7 phosphorylates a long, repetitive tail on the polymerase (the CTD, or C-terminal domain) at a specific position called serine 5. This early phosphorylation helps the cell add a protective cap to the beginning of the new RNA molecule. CDK7 also helps establish a pause point shortly after the polymerase starts moving, about 50 to 70 base pairs downstream of where transcription begins.
CDK9, paired with cyclin T in a factor called P-TEFb, releases the polymerase from that pause. It phosphorylates the CTD at serine 2, a mark associated with active elongation. CDK9 also influences a chemical tag on histone proteins (a modification called H2Bub1) that helps the polymerase move through tightly packed DNA. When CDK9 is depleted, cells lose this histone mark and the polymerase can overshoot the ends of genes.
CDK8 (and its close relative CDK19) work within the Mediator complex, a massive molecular bridge between gene-activating proteins and the transcription machinery. CDK12 and CDK13, both partnered with cyclin K, phosphorylate the polymerase’s tail during later stages of elongation and help the cell properly cut and finish RNA molecules at the ends of genes. Losing CDK12 leads to defective RNA processing because cleavage factors can no longer be recruited efficiently.
How Cells Put the Brakes on CDKs
Because uncontrolled CDK activity would mean uncontrolled cell division, cells have built-in braking systems. Two families of small inhibitor proteins keep CDKs in check.
The INK4 family (the best-known member is p16) specifically targets CDK4 and CDK6. These inhibitors bind directly to the CDK subunit and distort its shape so it can no longer pair with a cyclin or phosphorylate anything. Importantly, when p16 knocks cyclin D off CDK4/6, it frees up another inhibitor protein called p27, which then migrates to CDK2 complexes and shuts those down too. This creates a cascading brake that can halt the cell cycle at multiple points simultaneously. p16 ramps up in response to DNA damage, oxidative stress, or signals from overactive growth genes.
The CIP/KIP family (p21 and p27 are the main members) is more versatile. p21 can bind and inhibit most major CDK-cyclin combinations, including CDK4/6, CDK2, and CDK1 complexes. It accumulates in cells that have exited the cell cycle and sits on CDK complexes to prevent re-entry. When a cell detects DNA damage, p21 also binds CDK1 complexes to halt division at the G2 checkpoint, buying time for repair. Beyond direct binding, p21 can block CDK activation indirectly by interfering with the CDK-activating kinase (CAK) that provides the essential T-loop phosphorylation.
p27 is the cell’s primary enforcer of quiescence, the resting state most cells in your body spend their time in. In its active form, p27 binds and silences CDK2 and CDK1 complexes. When growth signals arrive, p27 gets phosphorylated on specific tyrosine residues, which loosens its grip and eventually targets it for destruction. The freed CDKs then push the cell back into division.
How Cyclin Destruction Resets the System
CDKs can only stay active as long as their cyclin partner is present, and the cell uses targeted protein destruction to bring cyclin levels crashing down at the right moment. The key player is the anaphase-promoting complex (APC/C), a large molecular machine that tags proteins with a small molecule called ubiquitin, marking them for shredding by the cell’s protein recycling machinery.
In mid-mitosis, the APC/C partners with an activator called Cdc20 and destroys cyclin B. This deactivates CDK1 and allows the cell to finish dividing and exit mitosis. After division, the APC/C switches to a different activator called Cdh1, which keeps cyclin levels low throughout G1 so the cell doesn’t accidentally re-enter division. Dozens of cell cycle proteins beyond the cyclins are also destroyed by the APC/C, making it a central reset button for the entire division machinery.
CDK Inhibitors in Cancer Treatment
Because cancer cells rely on hyperactive CDK signaling to keep dividing, blocking specific CDKs has become a major strategy in oncology. Three drugs that specifically inhibit CDK4 and CDK6 are now widely used.
Palbociclib (brand name Ibrance) was the first to win FDA approval, in February 2015, for hormone receptor-positive, HER2-negative metastatic breast cancer. It blocks CDK4 and CDK6 at very low concentrations while leaving other CDKs essentially untouched. Ribociclib (Kisqali) followed in March 2017, and abemaciclib (Verzenio) in September 2017, both for similar breast cancer subtypes. Abemaciclib is the most potent of the three against CDK4, effective at a concentration of just 2 nanomolar.
All three drugs work by the same principle: they prevent CDK4/6 from phosphorylating Rb, keeping Rb active and the brakes on cell division engaged. In practice, they are used alongside hormone therapies. Abemaciclib has also been approved for early-stage breast cancer in patients at high risk of recurrence, not just metastatic disease, making it the first CDK4/6 inhibitor used in the adjuvant (post-surgery) setting.