The cell-division cycle is the means by which cells accurately duplicate themselves and then divide in two. In recent years a general model has emerged in which protein complexes consisting of cyclins and cyclin-dependent protein kinases (Cdks) regulate progression through this cycle1. In addition, ‘checkpoints’ exist to monitor the integrity and replication status of the genetic material before cells commit either to replicate their DNA (in S phase) or to segregate it (during mitosis)2. These checkpoints are signal-transduction pathways whose effectors interact with cyclin/Cdk complexes to block the cell cycle. A report by Chan et al.3 on page 616 of this issue helps us to understand how checkpoints do this.
Why the need to block the cell-division cycle? If DNA is damaged, a delay in the cell cycle facilitates repair, minimizing the replication and subsequent segregation of damaged DNA. Indeed, loss of checkpoint integrity is a hallmark of many human cancers4. Cells respond to DNA damage by stopping the cell cycle either at a stage known as G1, before DNA replication (via the G1 DNA-damage checkpoint), or just before mitosis by the G2 DNA-damage checkpoint.
The p53 tumour-suppressor protein is an essential component of the G1 DNA-damage checkpoint5. In response to damage, cells that lack p53 initiate a delay in the cell cycle at G2, but they neither delay entry into S phase nor maintain the arrest at G26,7. These properties have important clinical implications, because agents that disrupt the G2 DNA-damage checkpoint selectively potentiate the cytotoxic effects of DNA-damaging agents on cancer cells that lack p538. Although the function of p53 in controlling the G1 checkpoint is fairly well understood at the molecular level, the contribution made by p53 to control of the G2 DNA-damage checkpoint is less clear. Chan et al.3 have now begun to address this.
The p53 protein activates transcription of other proteins, among them 14-3-3σ (ref. 9). The seven human 14-3-3 proteins are highly conserved, phosphoserine-binding proteins involved in cellular proliferation, checkpoint control and apoptosis10. In epithelial cells, expression of 14-3-3σ is induced in a p53-dependent manner in response to DNA damage and, when overproduced, 14-3-3σ blocks the cell cycle at G29.
To test the contribution of 14-3-3σ to G2 checkpoint control, Chan et al. disrupted the gene encoding 14-3-3σ in a human colorectal cancer cell line containing functional p53. After DNA damage, these cells arrested in the G2 phase of the cell cycle, indicating that the G2 DNA-damage checkpoint was intact. But the cells could not maintain cell-cycle arrest, and they ultimately perished by non-apoptotic cell death. These findings indicate that expression of 14-3-3σ is not required for cells to initiate the DNA damage-induced G2 delay, but that it is critical for cells to sustain — and, perhaps, to recover from — the delay.
Why does loss of 14-3-3σ allow cells with DNA damage to escape arrest and enter mitosis prematurely? In human cells, entry into mitosis requires at least two events: activation of a mitosis-promoting protein called cdc2 by the cdc25C phosphatase; and accumulation of active cdc2 in the nucleus. In cells containing functional p53, the DNA-damage checkpoint seems to block both processes. First, cdc25C is prevented from activating cdc2. This is done by maintaining cdc25C in a phosphorylated form that binds various 14-3-3 proteins, preventing the cdc25C from accumulating in the nucleus11,12, and activating cdc2 there. But, because cdc25C does not bind 14-3-3σ (ref. 3), this pathway is intact in p53/14-3-3σ-deficient cells. Second, cdc2 is exported to the cytoplasm. Normally, cdc2, in a complex with cyclin B1, is constantly shuttled between the nucleus and the cytoplasm (Fig. 1). Cyclin B1 contains a nuclear-export sequence that facilitates the efficient export of this complex into the cytoplasm13. Loss of 14-3-3σ seems to interfere with this second regulatory pathway, allowing cdc2/cyclin B1 complexes to accumulate in the nucleus and, ultimately, bypass of the G2 checkpoint.
Clearly, the molecular underpinnings of the p53/14-3-3σ pathway deserve further attention. Although cdc2 co-immunoprecipitates with 14-3-3σ, we do not yet know whether 14-3-3σ binds directly to cdc2/cyclin B1 or to other proteins in the complex. We also need to know whether 14-3-3σ anchors cdc2/cyclin B1 complexes in the cytoplasm, as Chan and colleagues propose3, or whether it regulates shuttling of these complexes between the nucleus and the cytoplasm. Furthermore, although loss of 14-3-3σ results in the nuclear accumulation of cdc2, this alone is not expected to cause entry into mitosis — cdc2 must also be activated. But it is not clear how. The cdc25C regulatory pathway seems to be intact in cells that lack 14-3-3σ. Moreover, the wee1 tyrosine kinase, which negatively regulates cdc2, can, presumably, inactivate nuclear pools of cdc2.
Chan and colleagues also report that wee1 is present in 14-3-3σ immunoprecipitates. So one possibility is that 14-3-3σ facilitates interactions between cdc2 and wee1. In that case, loss of 14-3-3σ might promote the nuclear accumulation of cdc2 and also disrupt the interactions between cdc2 and wee1. Together, these two events would be expected to generate active cdc2 in the nucleus and, in turn, to promote premature mitosis. Finally, future studies should be aimed at understanding why 14-3-3σ-deficient cells choose mitotic catastrophe over adaptation as their response to DNA damage.
A model now emerges for epithelial cells in which distinct members of the 14-3-3 family target specific mitotic regulators to control various aspects of the G2 DNA-damage checkpoint (Fig. 1). Initiation of the checkpoint does not depend on p53, but it does require interactions between cdc25C and members of the 14-3-3 family. Maintenance of the checkpoint, however, depends on p53 and also involves interactions between cdc2, cyclin B1 and 14-3-3σ. The 14-3-3 proteins use a common mechanism for initiating and maintaining the checkpoint — namely, exclusion of mitotic regulators from the nucleus. Components of the p53/14-3-3σ and cdc25C/14-3-3 pathways could, moreover, be therapeutic targets for drug development. In combination with established therapies that use DNA-damaging agents to kill tumour cells, such agents should enhance the anti-tumour effectiveness. Furthermore these agents are expected to be selective for cancer cells — because nonselective fools rush in where angels fear to tread.
Morgan, D. O. Annu. Rev. Cell Dev. Biol. 13, 261–291 (1997).
Hartwell, L. H. & Weinert, T. A. Science 246, 629–634 (1989).
Chan, T. A., Hermeking, H., Lengauer, C., Kinzler, K. W. & Vogelstein, B. Nature 401, 616–620 (1999).
Hartwell, L. H. & Kastan, M. B. Science 266, 1821–1828 (1994).
Kastan, M. B., Onyekwere, B., Sidransky, D., Vogelstein, B. & Craig, R. W. Cancer Res. 51, 6304–6311 (1991).
Paules, R. S. et al. Cancer Res. 55, 1763–1773 (1995).
Bunz, F. et al. Science 282, 1497–1501 (1998).
Wang, O. et al. J. Natl Cancer Inst. 88, 956–965 (1996).
Hermeking, H. et al. Mol. Cell 1, 3–13 (1997).
Aitken, A. Trends Cell Biol. 6, 341–347 (1996).
Peng, C.-Y. et al. Science 277, 1501–1505 (1997).
Dalal, S. N. et al. Mol. Cell Biol. 19, 4465–4479 (1999).
Yang, J. & Kornbluth, S. Trends Cell Biol. 9, 201–210 (1999).
About this article
Aberrant methylation profile of 14-3-3 sigma and its reduced transcription/expression levels in Chinese sporadic female breast carcinogenesis
Medical Oncology (2010)
Detection of somaclonal variation during cocoa somatic embryogenesis characterised using cleaved amplified polymorphic sequence and the new freeware Artbio
Molecular Breeding (2010)
Benzylidene derivatives of andrographolide inhibit growth of breast and colon cancer cells in vitro by inducing G1 arrest and apoptosis
British Journal of Pharmacology (2009)
The clinicopathological and prognostic impact of 14-3-3 sigma expression on vulvar squamous cell carcinomas
BMC Cancer (2008)
Molecular and Cellular Biochemistry (2007)