Cyclin-CDK complexes: regulation of each cell cycle transition and checkpoint mechanisms

Intuition Beginner

The cell cycle has four phases — G1, S, G2, and M — and the cell needs a molecular engine to drive it from one phase to the next. That engine is built from two parts: cyclins, proteins whose concentrations rise and fall during the cycle, and CDKs (cyclin-dependent kinases), enzymes that are inactive on their own but spring to life when a cyclin binds them. Each cyclin-CDK pair acts as a timed switch, turning on at the right moment and being destroyed when its job is done.

Different cyclins appear at different times. Cyclin D rises first in G1, pairing with CDK4 and CDK6 to push the cell toward DNA replication. Cyclin E appears at the G1/S boundary, partnering with CDK2 to start S phase. Cyclin A takes over during S phase and G2, and cyclin B pairs with CDK1 to drive the cell into mitosis. As each cyclin accumulates, its partner CDK phosphorylates target proteins that execute the next phase of the cycle.

Checkpoints act as quality-control gates. The G1/S checkpoint asks: is the cell big enough, are nutrients available, and is the DNA undamaged? The G2/M checkpoint asks: has every base pair been copied? The spindle checkpoint during mitosis asks: is every chromosome properly attached to the pulling cables? If a checkpoint finds a problem, the cell activates CDK inhibitor proteins that shut down specific cyclin-CDK complexes, halting the cycle until the issue is resolved.

Visual Beginner

The diagram shows a circular cell cycle with cyclin-CDK pairs positioned at each transition. Cyclin D-CDK4/6 sits in G1. Cyclin E-CDK2 guards the G1/S gate. Cyclin A-CDK2 operates during S phase. Cyclin A-CDK1 bridges G2. Cyclin B-CDK1 drives mitosis. At each checkpoint, inhibitor proteins (p21, p27, Wee1) and checkpoint kinases (ATM, ATR, Chk1, Chk2) stand ready to block the next cyclin-CDK complex if conditions are not met.

Worked example Beginner

A cell in G1 receives a growth signal. Cyclin D protein is produced and binds CDK4. The cyclin D-CDK4 complex begins phosphorylating a protein called Rb, which normally holds back a transcription factor called E2F. As Rb gets more and more phosphate tags, it releases E2F. Free E2F turns on genes for cyclin E and DNA replication enzymes.

Now cyclin E accumulates and binds CDK2. The cyclin E-CDK2 complex pushes the cell through the G1/S checkpoint and into S phase. After S phase begins, cyclin E is rapidly degraded — its job is done. Cyclin A takes over, binding CDK2 to keep DNA replication going.

This handoff pattern repeats at every transition: one cyclin-CDK pair turns on, does its job, and is destroyed, making way for the next pair. The sequential rise and fall of cyclins is the clock that keeps the cell cycle running in the correct order.

Check your understanding Beginner

Formal definition Intermediate+

The cyclin-CDK pairs

The eukaryotic cell cycle is driven by a set of cyclin-dependent kinase complexes, each operating at a specific phase transition:

Transition	Cyclin	CDK partner	Primary targets
G1 progression	Cyclin D (D1, D2, D3)	CDK4, CDK6	Rb (progressive phosphorylation)
G1/S commitment	Cyclin E (E1, E2)	CDK2	Rb (hyperphosphorylation), replication origin licensing
S-phase progression	Cyclin A	CDK2	Replication forks, origin firing control
G2 preparation	Cyclin A	CDK1	Early mitotic targets
M-phase entry and execution	Cyclin B (B1, B2)	CDK1	Nuclear envelope breakdown, chromosome condensation, spindle assembly

CDK regulation at four levels

1. Cyclin binding. A CDK without a cyclin partner adopts a closed conformation that blocks the substrate-binding cleft. Cyclin binding reorients the PSTAIRE helix (in CDK1/2) or the equivalent helix, opening the active site and creating the substrate-recognition surface. Without cyclin, the catalytic site is misaligned and the activation segment (T-loop) blocks access.

2. Activating phosphorylation (T-loop). Full activity requires phosphorylation of a conserved threonine in the T-loop (Thr160 in CDK2, Thr161 in CDK1) by CDK-activating kinase (CAK), itself a complex of CDK7-cyclin H-MAT1. T-loop phosphorylation stabilizes the active conformation and increases substrate affinity by roughly 100-fold.

3. Inhibitory phosphorylation. Wee1 and Myt1 kinases phosphorylate CDK1 at Tyr15 and Thr14, preventing ATP binding even when cyclin is bound. This mechanism specifically holds cyclin B-CDK1 inactive during G2 until the cell is ready for mitosis. Cdc25 phosphatases (Cdc25A, B, C) remove these inhibitory phosphates, activating CDK1.

4. CDK inhibitors (CKIs). Two structurally unrelated families:

• INK4 family (p16/INK4a, p15/INK4b, p18/INK4c, p19/INK4d): bind CDK4 and CDK6 directly, preventing cyclin D association. Specific to the G1 cyclin-CDK complexes.
• CIP/KIP family (p21/CIP1, p27/KIP1, p57/KIP2): bind cyclin-CDK complexes (primarily cyclin E-CDK2 and cyclin A-CDK2) and inhibit activity. p21 is the primary transcriptional target of p53 and mediates DNA-damage-induced G1 arrest.

Cyclin destruction: the APC/C and SCF

Cyclin degradation is the primary mechanism for dropping CDK activity between phases. Two E3 ubiquitin ligases control cyclin destruction:

SCF (Skp1-Cul1-F-box complex) operates during G1/S and S phase. It recognizes phosphorylated substrates via F-box proteins (Skp2 targets p27 for degradation; beta-TrCP targets Emi1). SCF activity is constitutive — substrate availability, not SCF activity, is regulated.

APC/C (anaphase-promoting complex/cyclosome) operates from mid-M through G1. It is a multi-subunit RING-finger E3 ligase activated by one of two co-activators:

• Cdc20: activates APC/C from metaphase through anaphase, targeting securin and cyclin B for degradation.
• Cdh1: activates APC/C from late mitosis through G1, targeting cyclin B, cyclin A, Plk1, and other mitotic kinases. Cdh1-APC/C maintains low CDK activity during G1, allowing origin relicensing.

APC/C recognizes substrates through short degron motifs: the D-box (RxxLxxxxN) and the KEN-box. Cyclin B has a D-box; securin has both a D-box and a KEN-box.

Checkpoint mechanisms

The G1/S checkpoint (restriction point). Growth-factor signaling through RTK-MAPK 17.07.02 and PI3K-AKT 17.07.03 pending induces cyclin D transcription. Cyclin D-CDK4/6 phosphorylates Rb at a subset of sites. Partially phosphorylated Rb releases some E2F, allowing cyclin E transcription. Cyclin E-CDK2 completes Rb hyperphosphorylation, releasing all E2F and driving S-phase gene expression. The mutual activation of E2F and cyclin E-CDK2 creates a bistable switch: once E2F activity crosses a threshold, the system is self-sustaining even without growth-factor input.

DNA damage during G1 activates ATM/ATR kinases, which phosphorylate and stabilize p53. p53 transactivates p21, which binds and inhibits cyclin E-CDK2 and cyclin A-CDK2, blocking the G1/S transition until damage is repaired.

The G2/M checkpoint. Cyclin B-CDK1 is held inactive by Wee1/Myt1-mediated phosphorylation at Tyr15/Thr14. DNA damage or incomplete replication activates ATR/Chk1 or ATM/Chk2 pathways, which phosphorylate and inactivate Cdc25C (the phosphatase that removes the inhibitory phosphates). Without active Cdc25C, CDK1 remains inhibited and the cell cannot enter mitosis.

CDK1-Cdc25 forms a second bistable switch: active CDK1 phosphorylates and activates Cdc25 (positive feedback) while phosphorylating and inhibiting Wee1 (double-negative feedback). Once CDK1 activity crosses the activation threshold, the feedback drives full mitotic entry.

The spindle assembly checkpoint (SAC). Unattached kinetochores recruit Mad2, which undergoes a conformational change (open Mad2 to closed Mad2) and binds Cdc20, inhibiting APC/C-Cdc20. The mitotic checkpoint complex (MCC = Mad2-Cdc20-BubR1-Bub3) sequesters Cdc20. Without APC/C-Cdc20 activity, securin and cyclin B are not degraded. Securin holds separase inactive; without separase activity, cohesin rings remain intact and sister chromatids cannot separate.

Once all kinetochores achieve proper amphitelic attachment, MCC production ceases, existing MCC is disassembled by p31(comet), and APC/C-Cdc20 is released to initiate anaphase.

Key mechanism Intermediate+

Mechanism: The CDK1-Cdc25-Wee1 bistable switch at G2/M.

The G2/M transition is governed by a bistable switch built from two interlocking feedback loops involving CDK1, Cdc25, and Wee1. The switch ensures that mitotic entry is sharp, irreversible, and occurs only when preconditions are met.

The three components form the following circuit:

1. CDK1 activates Cdc25. Active cyclin B-CDK1 phosphorylates Cdc25 on multiple serine residues, increasing its phosphatase activity toward CDK1.

2. Cdc25 activates CDK1. Cdc25 removes the inhibitory phosphates (Tyr15, Thr14) from CDK1, converting it from inactive to active.

3. CDK1 inhibits Wee1. Active CDK1 phosphorylates Wee1, reducing its kinase activity toward CDK1.

4. Wee1 inhibits CDK1. Wee1 phosphorylates CDK1 at Tyr15 (and Myt1 at Thr14), keeping it inactive even when cyclin B is bound.

Loops 1+2 form a positive feedback (CDK1 activates Cdc25 activates CDK1). Loops 3+4 form a double-negative feedback (CDK1 inhibits Wee1 inhibits CDK1), which is functionally equivalent to positive feedback. Two positive feedback loops acting on the same variable (CDK1 activity) create a strong bistable switch.

Quantitative model. Let $x$ denote the fraction of CDK1 in the active (dephosphorylated) state. The rate equation is:

$$\frac{dx}{dt}=V_{\text{Cdc25}}(x)\cdot (1-x)-V_{\text{Wee1}}(x)\cdot x$$

where both $V_{\text{Cdc25}}$ and $V_{\text{Wee1}}$ depend on $x$ because CDK1 phosphorylates and activates Cdc25 while phosphorylating and inhibiting Wee1:

$$V_{\text{Cdc25}}(x)=V_{\text{Cdc25}}^{\min }+(V_{\text{Cdc25}}^{\max }-V_{\text{Cdc25}}^{\min })\cdot \frac{x^n}{K_{\text{Cdc25}}^n+x^n}$$ $$V_{\text{Wee1}}(x)=V_{\text{Wee1}}^{\max }-(V_{\text{Wee1}}^{\max }-V_{\text{Wee1}}^{\min })\cdot \frac{x^n}{K_{\text{Wee1}}^n+x^n}$$

The steady state $dx/dt=0$ gives:

$$V_{\text{Cdc25}}(x)\cdot (1-x)=V_{\text{Wee1}}(x)\cdot x$$

For Hill coefficient $n\ge 2$ and appropriate parameter choices, this equation has three solutions: a low-$x$ stable fixed point (interphase, low CDK1 activity), an unstable saddle point, and a high-$x$ stable fixed point (mitosis, high CDK1 activity). The switch between them occurs via saddle-node bifurcations as the cyclin B concentration $[\text{CycB}]$ increases — effectively tilting the balance between $V_{\text{Cdc25}}$ and $V_{\text{Wee1}}$.

The checkpoint modulates the switch by acting on the parameters. DNA damage activates Chk1, which phosphorylates Cdc25 and targets it for cytoplasmic sequestration (reducing effective $V_{\text{Cdc25}}^{\max }$) and also stabilizes Wee1 (increasing $V_{\text{Wee1}}^{\max }$). These parameter shifts raise the cyclin B threshold needed for switching, holding the cell in the low-CDK1 state until damage is repaired.

Irreversibility. Once CDK1 is fully activated and the cell enters mitosis, the switch is irreversible with respect to the cyclin B input. The positive feedback maintains high CDK1 activity even if cyclin B levels dip slightly. Exit from mitosis requires an entirely separate mechanism — APC/C-mediated cyclin B degradation — which drops $[\text{CycB}]$ below the lower saddle-node bifurcation point, snapping the system back to the low-CDK1 state. The separation between the entry and exit thresholds is hysteresis, the hallmark of bistability.

Exercises Intermediate+

The cyclin-CDK oscillator and APC/C dynamics Master

The Tyson-Novak oscillator

The cyclin-CDK system is one of the best-characterized biological oscillators. Novak and Tyson (1993, 2003) developed a series of increasingly detailed ODE models that reproduce the major features of the eukaryotic cell cycle: sequential phase transitions, checkpoint-dependent arrest, and limit-cycle oscillation.

The minimal model contains two coupled modules: a bistable CDK1-Cdc25-Wee1 switch (producing the G2/M transition) and an APC/C-Cdc20-Cdh1 negative feedback loop (resetting the switch). The core variables are:

• $[\text{CycB}]$: cyclin B concentration (synthesized at constant rate $k_1$, degraded by APC/C)
• $x$: fraction of CDK1 in the active state (governed by the bistable switch equations)
• $[\text{Cdc20}]$: APC/C co-activator concentration (synthesized in response to CDK1 activity, creating the delayed negative feedback)
• $[\text{Cdh1}]$: fraction of APC/C bound to Cdh1 (inhibited by CDK phosphorylation, activated by phosphatases at low CDK)

The limit cycle emerges because CDK1-CycB activates its own destroyer (Cdc20) with a time delay. As CDK1 activity rises, Cdc20 accumulates, APC/C-Cdc20 degrades cyclin B, CDK1 activity drops, Cdc20 production ceases, cyclin B re-accumulates, and the cycle repeats. The trajectory spirals outward from an unstable fixed point to a stable limit cycle in the $([\text{CycB}],x,[\text{Cdc20}])$ phase space.

The mathematical structure is that of a relaxation oscillator: the system spends most of its time in quasi-steady states (interphase and mitosis), with rapid transitions between them driven by the bistable switches. This matches the biological observation that G1, S, and G2 are long phases (hours) while the G2/M and M/G1 transitions take minutes.

APC/C substrate specificity: Cdc20 versus Cdh1

The APC/C uses two co-activators that define two temporally distinct substrate-degradation programs:

APC/C-Cdc20 (metaphase to anaphase). Cdc20 is transcriptionally induced by E2F and stabilized by CDK phosphorylation. Its primary substrates are:

1. Securin: Degradation releases separase, which cleaves cohesin to initiate anaphase.

2. Cyclin B: Degradation initiates CDK1 inactivation and mitotic exit.

3. Nek2A: A kinetochore-associated kinase degraded in early mitosis.

Cdc20 is itself regulated by the SAC: unattached kinetochores generate MCC (Mad2-Cdc20-BubR1-Bub3), which binds and inhibits APC/C-Cdc20. The SAC thus controls the timing of securin and cyclin B degradation relative to kinetochore attachment.

APC/C-Cdh1 (late mitosis through G1). Cdh1 is constitutively expressed but kept inactive during S, G2, and early M by CDK phosphorylation. As CDK activity drops (from APC/C-Cdc20-mediated cyclin B degradation), Cdh1 is dephosphorylated by Cdc14 phosphatase in yeast or by PP2A in mammals, and Cdh1-APC/C becomes active.

Cdh1-APC/C has broader substrate specificity than Cdc20-APC/C:

1. Cyclin B and cyclin A: Continued degradation maintains the low-CDK state.

2. Cdc20 itself: Cdh1 degrades Cdc20, ensuring the transition from Cdc20- to Cdh1-driven APC/C activity is unidirectional.

3. Plk1 and Aurora kinases: Removing these mitotic kinases is required for mitotic exit.

4. Geminin: Degradation releases Cdt1, enabling origin relicensing for the next S phase.

5. Skp2: Degradation stabilizes p27, reinforcing G1 arrest.

The Cdc20-to-Cdh1 transition is a ratchet: Cdh1 degrades Cdc20 (preventing backsliding), and CDK inactivation activates Cdh1 (positive feedback for mitotic exit). The ratchet is reset only when cyclin D-CDK4/6 phosphorylates and inactivates Cdh1 at the next G1/S transition.

Separase, cohesin, and the anaphase trigger

The anaphase trigger is the APC/C-Cdc20-mediated degradation of securin, which releases the protease separase. Active separase cleaves the cohesin subunit Rad21/Scc1 at two sites, opening the cohesin ring and dissolving the physical linkage between sister chromatids.

Separase is regulated at three levels:

1. Securin binding: Securin directly inhibits separase catalytic activity. Securin degradation is both necessary and sufficient for separase activation under normal conditions.

2. CDK1 phosphorylation: CDK1 phosphorylates separase at multiple sites, creating a binding site for cyclin B-CDK1. The bound cyclin B-CDK1 sterically inhibits separase even in the absence of securin. This creates a dual-lock system: both securin degradation and partial cyclin B degradation are needed to release separase.

3. Auto-cleavage: Once activated, separase cleaves itself at a specific site. This auto-cleavage is not required for activity but may serve as a negative-feedback mechanism.

The biological significance of the dual-lock is that it prevents premature cohesin cleavage. If securin were degraded by a proteasome glitch (without APC/C involvement), CDK1 phosphorylation of separase would still block cohesin cleavage. Only the ordered sequence — APC/C activation, securin degradation, partial cyclin B degradation, CDK1 inactivation — releases separase. This ordering is a nontrivial design feature of the anaphase control system.

p53 dynamics: oscillations versus sustained response

The p53 response to DNA damage exhibits two distinct dynamical modes depending on damage severity:

Pulsatile mode. Low to moderate DNA damage produces repeated pulses of p53, with each pulse having a fixed amplitude (roughly 2-3 fold above baseline) and a period of approximately 5-7 hours. The number of pulses increases with damage severity. This was first observed by Lahav et al. (2004) using fluorescent p53 reporters in single MCF7 cells. The pulsatile dynamics arise from the p53-MDM2 negative feedback loop: p53 induces MDM2 transcription, MDM2 protein accumulates after a transcriptional-translational delay, MDM2 ubiquitinates p53 and targets it for degradation, p53 drops, MDM2 drops, and the cycle repeats.

Sustained mode. Severe or irreparable damage switches p53 to a sustained high level. This transition is mediated by additional inputs: ATM/ATR phosphorylation of MDM2 (inhibiting its E3 ligase activity), p53 acetylation by CBP/p300 (blocking MDM2 binding), and stress-induced degradation of MDM2 itself. Sustained p53 drives apoptosis or senescence rather than transient arrest.

The switch from pulsatile to sustained p53 is itself a bifurcation in the p53-MDM2 ODE system. Theoretical work by Ma et al. (2005) and Puszynski et al. (2009) showed that the system transitions from a damped oscillation (transient pulses decaying to a steady state) through a Hopf bifurcation to a limit cycle (sustained oscillations) as damage increases, and then through a second bifurcation to a high-p53 fixed point (sustained response) when the damage signal overwhelms the MDM2 feedback.

Senescence versus quiescence

Two distinct non-dividing states exist in post-mitotic cells:

Quiescence (G0). A reversible exit from the cell cycle in G1. Cells in G0 have low CDK activity, hypophosphorylated Rb, and low E2F activity, but retain the capacity to re-enter the cycle when stimulated by growth factors. Quiescent cells have low metabolic activity but maintain viability. Liver hepatocytes, lymphocytes, and fibroblasts can reversibly enter and exit G0.

Senescence. An irreversible cell cycle arrest triggered by telomere shortening (replicative senescence), oncogene activation (oncogene-induced senescence, OIS), or severe DNA damage (therapy-induced senescence). Senescent cells have:

• High p16 and/or p21 expression maintaining CDK inhibition
• Rb in a hypophosphorylated state (even if cyclin D is present)
• Persistent DNA damage foci (senescence-associated heterochromatic foci, SAHF)
• A secretory phenotype (SASP) producing inflammatory cytokines, growth factors, and proteases

The distinction matters for cancer therapy: senescent cells are metabolically active and secrete pro-inflammatory factors that can promote tumor growth in neighboring cells (the SASP). Therapeutic strategies targeting senescent cells (senolytics) aim to clear these cells after chemotherapy or radiation.

Cell cycle dysregulation in cancer

The cyclin-CDK system is disrupted in virtually every human cancer. The most common alterations include:

RB1 pathway disruption. Rb is the gatekeeper of the restriction point. RB1 loss (retinoblastoma, osteosarcoma, small-cell lung cancer) or functional inactivation through cyclin D amplification, CDK4/6 amplification, or p16 loss bypasses the G1/S checkpoint entirely. The cell commits to S phase regardless of growth-factor availability or DNA integrity.

TP53 mutation. p53 loss eliminates the DNA damage checkpoint at G1/S and weakens the G2/M arrest response. The most frequent TP53 mutations are missense mutations in the DNA-binding domain (R175H, R248Q, R273H), which both abolish p53's transcriptional activity and exert dominant-negative effects over wild-type p53 through tetramerization.

Cyclin D1 amplification. CCND1 amplification or translocation occurs in mantle-cell lymphoma (t(11;14)), breast cancer (amplification in roughly 15% of cases), and head-and-neck cancer. Elevated cyclin D1 drives constitutive Rb phosphorylation and G1/S entry.

CDK4/6 inhibitors. The clinical success of CDK4/6 inhibitors validates the cyclin-CDK system as a therapeutic target. Three agents are FDA-approved:

1. Palbociclib (Pfizer, 2015): selective for CDK4/6, given orally, 3-weeks-on/1-week-off schedule to manage neutropenia.

2. Ribociclib (Novartis, 2017): similar selectivity, approved in combination with endocrine therapy for HR+/HER2- breast cancer.

3. Abemaciclib (Eli Lilly, 2017): broader kinase inhibition profile including CDK4/6 with some CDK2/9 activity, continuous dosing, and single-agent activity in some settings.

Resistance mechanisms include RB1 loss (bypassing the drug's target), cyclin E amplification (CDK2-driven Rb phosphorylation independent of CDK4/6), FAT1 loss (Hippo pathway dysregulation), and FGFR pathway upregulation.

Connections Master

1. DNA replication 17.05.01. The once-and-only-once replication mechanism depends on CDK activity being low during G1 (for origin licensing) and high during S phase (for origin firing, while simultaneously preventing relicensing). The cyclin-CDK oscillator described here provides the CDK activity wave that gates origin licensing and firing.

2. RTK-MAPK signaling 17.07.02. Growth-factor signaling through the MAPK cascade drives cyclin D transcription, providing the initial input to the cyclin-CDK engine. ERK directly phosphorylates transcription factors that activate the CCND1 promoter. Without this upstream signal, cyclin D is not produced and the restriction point is never crossed.

3. PI3K-AKT-mTOR signaling 17.07.03 pending. The PI3K-AKT pathway promotes G1 progression through multiple mechanisms: AKT phosphorylates and inhibits GSK3-beta (stabilizing cyclin D), activates mTORC1 (increasing protein synthesis for cell growth), and phosphorylates p21 and p27 (promoting their cytoplasmic sequestration and degradation).

4. DNA repair pathways 17.06.02 pending. The DNA damage checkpoints at G1/S and G2/M link directly to the repair machinery. ATM and ATR kinases, activated by different DNA lesions, phosphorylate p53 and Chk1/Chk2 to arrest the cell cycle while repair enzymes process the damage. The checkpoint-kinase system described here is the signaling layer that coordinates repair with cell cycle progression.

5. Cell cycle and mitosis overview 17.08.01. This unit deepens the cyclin-CDK content introduced in 17.08.01, focusing on the regulatory mechanisms at each transition rather than the structural events of mitosis. The checkpoint and CDK-inhibitor systems described here are the molecular implementation of the quality-control gates described in the overview unit.

6. Meiosis 17.08.03 pending. Meiosis shares the cyclin-CDK regulatory logic but modifies it for two consecutive divisions (meiosis I and II) without an intervening S phase. Cohesin cleavage is regulated differently: separase cleaves arm cohesin at anaphase I but protects centromeric cohesin (via Shugoshin) until anaphase II. The APC/C is regulated by the spindle checkpoint in both divisions.

7. Dynamical systems 02.12.01. The cell cycle is a limit-cycle oscillator, the bistable switches at G1/S and G2/M are saddle-node bifurcations, and the checkpoint modulation of switch thresholds is an example of bifurcation control. The Novak-Tyson models are dynamical systems whose analysis uses the same ODE and bifurcation tools covered in the mathematics curriculum.

Historical notes Master

The discovery of cyclins is one of the classic stories in cell biology. Tim Hunt, then at the Marine Biological Laboratory in Woods Hole, was studying protein synthesis in sea urchin (Arbacia punctulata) eggs in the summer of 1982. His student Joan Ruderman had developed an in vitro translation system, and Hunt noticed that one particular protein band on his gels accumulated steadily during interphase and then vanished abruptly at each cleavage division. Because the protein's concentration cycled, he named it cyclin. The discovery was published in Cell in 1983 (Evans, Rosenthal, Youngblom, Distel, and Hunt). Hunt shared the 2001 Nobel Prize with Leland Hartwell and Paul Nurse.

The CDK story began independently. Paul Nurse identified the cdc2 gene in fission yeast as a master regulator of the G2/M transition. In 1987, Nurse's lab showed that human CDK1 (then called CDC2) could complement a yeast cdc2 mutant, establishing that the cell cycle engine is conserved across all eukaryotes. The link between cyclins and CDKs was made in 1991 when several groups showed that cyclin binding is required for CDK activity, and that different cyclins direct CDKs to different cell cycle stages.

The concept of checkpoints was proposed by Hartwell and Weinert in 1989. They introduced the idea that the cell cycle contains surveillance mechanisms that monitor the completion of upstream events before allowing downstream events to proceed. This was a departure from the earlier view that the cell cycle is a simple sequence of dependent steps. The checkpoint concept has since become central to understanding cancer: most tumors have defects in one or more checkpoints.

The APC/C was discovered independently by two groups in 1995 — King, Peters, and Kirschner (who named it the cyclosome) and Sudakin, Ganoth, and Hershko (who named it the anaphase-promoting complex). The convergence on the same multisubunit E3 ligase from complementary approaches cemented its role as the master regulator of mitotic exit. The spindle checkpoint's molecular mechanism — Mad2 conformational change catalyzed by unattached kinetochores — was elucidated by Musacchio, Salmon, and others in the 2000s.

The mathematical modeling tradition began with Novak and Tyson in 1993 and has been extended by Pomerening, Sauria, and others. The Tyson-Novak models demonstrated that the cell cycle can be understood as a dynamical system with limit-cycle behavior, and that the bistability at each checkpoint can be analyzed using bifurcation theory. These models made quantitative predictions about checkpoint strength and transition timing that have been confirmed experimentally.

Bibliography Master

@article{Evans1983,
  author = {Evans, T. and Rosenthal, E. T. and Youngblom, J. and Distel, D. and Hunt, T.},
  title = {Cyclin: a protein specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage division},
  journal = {Cell},
  volume = {33},
  year = {1983},
  pages = {389-396},
}

@article{Nurse2002,
  author = {Nurse, P.},
  title = {Cyclin dependent kinases and cell cycle control},
  journal = {Angew. Chem. Int. Ed.},
  volume = {41},
  year = {2002},
  pages = {4924-4932},
}

@article{NovakTyson1993,
  author = {Novak, B. and Tyson, J. J.},
  title = {Numerical analysis of a comprehensive model of {M}-phase control in {Xenopus} oocyte extracts and intact embryos},
  journal = {J. Cell Sci.},
  volume = {106},
  year = {1993},
  pages = {1153-1168},
}

@article{WeinertHartwell1989,
  author = {Weinert, T. and Hartwell, L. H.},
  title = {Control of {G2} delay by the rad9 gene of {Saccharomyces cerevisiae}},
  journal = {J. Cell Sci.},
  volume = {12},
  year = {1989},
  pages = {165-168},
}

@article{King1995,
  author = {King, R. W. and Peters, J. M. and Tugendreich, S. and Rolfe, M. and Hieter, P. and Kirschner, M. W.},
  title = {A 20S complex containing {CDC27} and {CDC16} catalyzes the mitosis-specific conjugation of ubiquitin to cyclin {B}},
  journal = {Cell},
  volume = {81},
  year = {1995},
  pages = {279-288},
}

@article{Lahav2004,
  author = {Lahav, G. and Rosenfeld, N. and Sigal, A. and Geva-Zatorsky, N. and Levine, A. J. and Elowitz, M. B. and Alon, U.},
  title = {Dynamics of the {p53-MDM2} feedback loop in individual cells},
  journal = {Nat. Genet.},
  volume = {36},
  year = {2004},
  pages = {147-150},
}

@article{Pardee1974,
  author = {Pardee, A. B.},
  title = {A restriction point for control of normal animal cell proliferation},
  journal = {Proc. Natl. Acad. Sci. USA},
  volume = {71},
  year = {1974},
  pages = {1286-1290},
}

@book{Morgan2007,
  author = {Morgan, D. O.},
  title = {The Cell Cycle: Principles of Control},
  publisher = {New Science Press},
  year = {2007},
}

@book{Alberts2022,
  author = {Alberts, B. and Johnson, A. and Lewis, J. and Morgan, D. and Raff, M. and Roberts, K. and Walter, P.},
  title = {Molecular Biology of the Cell},
  publisher = {Garland Science},
  year = {2022},
  edition = {7th},
}

@book{Lodish2021,
  author = {Lodish, H. and Berk, A. and Kaiser, C. A. and Krieger, M. and Bretscher, A. and Ploegh, H. and Amon, A. and Martin, K. C.},
  title = {Molecular Cell Biology},
  publisher = {W. H. Freeman},
  year = {2021},
  edition = {9th},
}