Editing G2 phase (section)

==End of G<sub>2</sub>/entry into mitosis==
{{see also|Maturation promoting factor|Biochemical switches in the cell cycle}}
{{More citations needed|section|date=December 2023}}
Mitotic entry is determined by a threshold level of active cyclin-B1/CDK1 complex, also known as cyclin-B1/Cdc2 or the [[maturation promoting factor]] (MPF). Active cyclin-B1/CDK1 triggers irreversible actions in early mitosis, including [[centrosome]] separation, [[nuclear envelope]] breakdown, and [[Spindle apparatus|spindle]] assembly. In vertebrates, there are five cyclin B [[isoform]]s ([[Cyclin B1|B1]], [[Cyclin B2|B2]], [[Cyclin B3|B3]], [[Cyclin B4|B4]], and [[Cyclin B5|B5]]), but the specific role of each of these isoforms in regulating mitotic entry is still unclear. It is known that cyclin B1 can compensate for loss of both cyclin B2 (and vice versa in ''[[Drosophila]]'').<ref>{{cite journal | vauthors = Porter LA, Donoghue DJ | title = Cyclin B1 and CDK1: nuclear localization and upstream regulators | journal = Progress in Cell Cycle Research | volume = 5 | pages = 335–47 | date = 2003 | pmid = 14593728 }}</ref> ''Saccharomyces cerevisiae'' contains six B-type cyclins (Clb1-6), with Clb2 being the most essential for function. In both vertebrates and S. cerevisiae, it is speculated that the presence of multiple B-type cyclins allows different cyclins to regulate different portions of the G2/M transition while also making the transition [[Robustness|robust]] to perturbations.<ref name=":0">{{Cite book|title=The cell cycle : principles of control|last=Morgan, David Owen, 1958-|date=2007|publisher=New Science Press|isbn=978-0-19-920610-0|oclc=70173205}}</ref>

Subsequent discussions will focus on the spatial and temporal activation of cyclin B1/CDK in mammalian cells, but similar pathways are applicable in both other metazoans and in S. cerevisiae.

=== Cyclin B1 synthesis and degradation ===
Cyclin B1 levels are suppressed throughout G1 and S phases by the [[anaphase-promoting complex]] (APC), an E3 ubiquitin ligase which targets cyclin B1 for proteolysis. Transcription begins at the end of S phase after DNA replication, in response to phosphorylation of transcription factors such as [[NFYA|NF-Y]], [[FOXM1|FoxM1]] and [[MYB (gene)|B-Myb]] by upstream G1 and G1/S cyclin-CDK complexes.<ref>{{cite journal | vauthors = Katula KS, Wright KL, Paul H, Surman DR, Nuckolls FJ, Smith JW, Ting JP, Yates J, Cogswell JP | display-authors = 6 | title = Cyclin-dependent kinase activation and S-phase induction of the cyclin B1 gene are linked through the CCAAT elements | journal = Cell Growth & Differentiation | volume = 8 | issue = 7 | pages = 811–20 | date = July 1997 | pmid = 9218875 }}</ref>

=== Regulation of cyclin-B1/CDK1 activity ===
Increased levels of cyclin B1 cause rising levels of cyclin B1-CDK1 complexes throughout G2, but the complex remains inactive prior to the G2/M transition due to inhibitory phosphorylation by the Wee1 and Myt1 kinases. Wee1 is localized primarily to the nucleus and acts on the Tyr15 site, while Myt1 is localized to the outer surface of the ER and acts predominantly on the Thr14 site.

The effects of Wee1 and Myt1 are counteracted by phosphatases in the cdc25 family, which remove the inhibitory phosphates on CDK1 and thus convert the cyclin B1-CDK1 complex to its fully activated form, MPF.
[[File:G2-M feedback loops.png|thumb|This diagram illustrates the feedback loops underlying the G2/M transition. Cyclin-B1/CDK1 activates Plk and inactivates Wee1 and Myt1. Activated Plk activates cdc25. Activation of Cdc25 and inactivation of Wee1/Myt1 lead to further activation of Cyclin-B1/CDK1. Also shown is the putative role of cyclin-A/CDK2 and Cdc25A as initial activators of the feedback loop, discussed in a later section.]]
Active cyclinB1-CDK1 phosphorylates and modulates the activity of Wee1 and the Cdc25 isoforms A and C. Specifically, CDK1 phosphorylation inhibits Wee1 kinase activity, activates Cdc25C [[phosphatase]] activity via activating the intermediate kinase [[PLK1]], and stabilizes Cdc25A. Thus, CDK1 forms a [[positive feedback]] loop with Cdc25 and a double [[negative feedback]] loop with Wee1 (essentially a net positive feedback loop).

=== Positive feedback and switch-like activation ===
[[File:G2-M Bistability.png|thumb|This graph illustrates the stable equilibria for cyclin-B1/CDK1 activity at varying cyclin B1 concentrations, with the threshold of cyclin B concentration for entering mitosis higher than the threshold for exiting mitosis.]]

These positive feedback loops encode a [[Hysteresis|hysteretic]] [[Bistability|bistable]] switch in CDK1 activity relative to cyclin B1 levels (see figure). This switch is characterized by two distinct stable equilibria over a bistable region of cyclin B1 concentrations. One equilibrium corresponds to interphase and is characterized by inactivity of Cyclin-B1/CDK1 and Cdc25, and a high level of Wee1 and Myt1 activity. The other equilibrium corresponds to M-phase and is characterized by high activity of Cyclin-B1/CDK1 and Cdc25, and low Wee1 and Myt1 activity. Within the range of bistability, a cell’s state depends upon whether it was previously in interphase or M-phase: the threshold concentration for entering M-phase is higher than the minimum concentration that will sustain M-phase activity once a cell has already exited interphase.

Scientists have both theoretically and empirically validated the bistable nature of the G2/M transition. The [[Novak-Tyson model]] shows that the differential equations modelling the cyclin-B/CDK1-cdc25-Wee1-Myt1 feedback loop admit two stable equilibria over a range of cyclin-B concentrations.<ref>{{cite journal | vauthors = Novak B, Tyson JJ | title = Numerical analysis of a comprehensive model of M-phase control in Xenopus oocyte extracts and intact embryos | journal = Journal of Cell Science | volume = 106 | pages = 1153–68 | date = December 1993 | pmid = 8126097 | issue = 4 | doi = 10.1242/jcs.106.4.1153 }}</ref> Experimentally, bistability has been validated by blocking endogenous cyclin B1 synthesis and titrating interphase and M-phase cells with varying concentrations of non-degradable cyclin B1. These experiments show that the threshold concentration for entering M-phase is higher than the threshold for exiting M-phase: nuclear envelope break-down occurs between 32-40&nbsp;nm cyclin-B1 for cells exiting interphase, while the nucleus remains disintegrated at concentrations above 16-24&nbsp;nm in cells already in M-phase.<ref>{{cite journal | vauthors = Sha W, Moore J, Chen K, Lassaletta AD, Yi CS, Tyson JJ, Sible JC | title = Hysteresis drives cell-cycle transitions in Xenopus laevis egg extracts | journal = Proceedings of the National Academy of Sciences of the United States of America | volume = 100 | issue = 3 | pages = 975–80 | date = February 2003 | pmid = 12509509 | pmc = 298711 | doi = 10.1073/pnas.0235349100 | bibcode = 2003PNAS..100..975S | doi-access = free }}</ref>

This bistable, hysteretic switch is physiologically necessary for at least three reasons.<ref>{{cite journal | vauthors = Pomerening JR, Sontag ED, Ferrell JE | title = Building a cell cycle oscillator: hysteresis and bistability in the activation of Cdc2 | journal = Nature Cell Biology | volume = 5 | issue = 4 | pages = 346–51 | date = April 2003 | pmid = 12629549 | doi = 10.1038/ncb954 | s2cid = 11047458 }}</ref> First, the G2/M transition signals the initiation of several events, such as chromosome condensation and nuclear envelope breakdown, that markedly change the morphology of the cell and are only viable in dividing cells. It is therefore essential that cyclin-B1/CDK1 activation occurs in a switch-like manner; that is, cells should rapidly settle into a discrete M-phase state after the transition, and should not persist in a continuum of intermediate states (e.g., with a partially decomposed nuclear envelope). This requirement is satisfied by the sharp [[Discontinuity (mathematics)|discontinuity]] separating the interphase and M-phase equilibrium levels of CDK1 activity; as the cyclin-B concentration increases beyond the activation threshold, the cell rapidly switches to the M-phase equilibrium.

Secondly, it is also vital that the G2/M transition occur unidirectionally, or only once per cell cycle  Biological systems are inherently [[Noise (signal processing)|noisy]], and small fluctuations in cyclin B1 concentrations near the threshold for the G2/M transition should not cause the cell to switch back and forth between interphase and M-phase states. This is ensured by the bistable nature of the switch: after the cell transitions to the M-phase state, small decreases in the concentration of cyclin B do not cause the cell to switch back to interphase.

Finally, the continuation of the cell cycle requires persisting oscillations in cyclin-B/CDK1 activity as the cell and its descendants transition in and out of M-phase. Negative feedback provides one essential element of this long-term oscillation: cyclin-B/CDK activates APC/C, which causes degradation of cyclin-B from metaphase onwards, restoring CDK1 to its inactive state. However, simple negative feedback loops lead to [[Damping ratio|damped oscillations]] that eventually settle on a steady state. Kinetic models show that negative feedback loops coupled with bistable positive feedback motifs can lead to persistent, non-damped oscillations (see [[relaxation oscillator]]) of the kind required for long-term cell cycling.  

=== Positive feedback ===
The positive feedback loop mentioned above, in which cyclin-B1/CDK1 promotes its own activation by inhibiting Wee1 and Myst1 and activating cdc25, does not inherently include a “trigger” mechanism to initiate the feedback loop. Recently, evidence has emerged suggesting a more important role for [[cyclin A2]]/CDK complexes in regulating the initiation of this switch. Cyclin A2/[[CDK2 (gene)|CDK2]] activity begins in early S phase and increases during G<sub>2</sub>. Cdc25B has been shown to dephosphorylate Tyr15 on CDK2 in early-to-mid G<sub>2</sub> in a manner similar to the aforementioned CDK1 mechanism. Downregulation of cyclin A2 in U2OS cells delays cyclin-B1/CDK1 activation by increasing Wee1 activity and lowering Plk1 and Cdc25C activity. However, cyclin A2/CDK complexes do not function strictly as activators of cyclin B1/CDK1 in G<sub>2</sub>, as CDK2 has been shown to be required for activation of the p53-independent G<sub>2</sub> checkpoint activity, perhaps through a stabilizing phosphorylation on [[Cdc6]]. CDK2-/- cells also have aberrantly high levels of Cdc25A. Cyclin A2/CDK1 has also been shown to mediate proteasomal destruction of Cdc25B. These pathways are often deregulated in cancer.<ref name=":0" />

=== Spatial regulation ===
In addition to the bistable and hysteretic aspects of cyclin B1-CDK1 activation, regulation of subcellular protein localization also contributes to the G2/M transition. Inactive cyclin B1-CDK1 accumulates in the cytoplasm, begins to be activated by cytoplasmic cdc25, and then is rapidly sequestered into the nucleus during prophase (as it is further activated). In mammals, cyclin B1/CDK1 translocation to the [[Nucleus (cell)|nucleus]] is activated by phosphorylation of five [[serine]] sites on cyclin B1's cytoplasmic retention site (CRS): S116, S26, S128, S133, and S147. In ''[[Xenopus laevis]]'', cyclin B1 contains four analogous CRS serine phosphorylation sites (S94, S96, S101, and S113) indicating that this mechanism is highly conserved. Nuclear export is also inactivated by phosphorylation of cyclin B1's [[nuclear export signal]] (NES). The regulators of these phosphorylation sites are still largely unknown but several factors have been identified, including [[extracellular signal-regulated kinases]] (ERKs), [[PLK1]], and CDK1 itself. Upon reaching some threshold level of phosphorylation, translocation of cyclin B1/CDK1 to the nucleus is extremely rapid. Once in the nucleus, cyclin B1/CDK1 phosphorylates many targets in preparation for mitosis, including [[histone H1]], [[nuclear lamins]], [[Centrosome|centrosomal proteins]], and [[Microtubule-associated protein|microtubule associated proteins (MAPs)]].

The subcellular localization of cdc25 also shifts from the cytosol to the nucleus during prophase. This is accomplished via removal of nuclear localization sequence (NLS)-obscuring phosphates and phosphorylation of the nuclear export signal. It is thought that the simultaneous transport of cdc25 and cyclin-B1/CDK1 into the nucleus amplify the switch-like nature of the transition by increasing the effective concentrations of the proteins.<ref name=":0" />