Abstract
A guiding hypothesis for cell cycle regulation asserts that regulated proteolysis constrains the directionality of certain cell cycle transitions1, 2. We have tested this hypothesis for mitotic exit, which is regulated by degradation of the Cdk1 activator, cyclin B3-5. Application of chemical Cdk1 inhibitors to cells in mitosis induces cytokinesis and other normal aspects of mitotic exit, including cyclin B degradation. However, chromatid segregation fails, resulting in entrapment of chromatin in the midbody. If cyclin B degradation is blocked with proteasome inhibitor or by expression of non-degradable cyclin B, Cdk inhibitors will nonetheless induce mitotic exit and cytokinesis. If after mitotic exit, the Cdk1 inhibitor is washed free from cells where cyclin B degradation is blocked, cells can reverse back to M phase. This reversal is characterized by chromosome recondensation, nuclear envelope breakdown, assembly of microtubules into a mitotic spindle, and in most cases, dissolution of the midbody, reopening of the cleavage furrow, and realignment of chromosomes at the metaphase plate. These findings demonstrate that proteasome-dependent protein degradation provides directionality for the M phase to G1 transition.
Cdk1, the major regulator of mitotic progression is activated through binding of cyclin A or B. Cyclin A is degraded during prometaphase when chromosomes move to align at the metaphase plate6, 7. Cyclin B degradation begins at metaphase and continues during chromatid segregation in anaphase and exit from M phase5. Cytokinesis is initiated shortly after anaphase onset. Cdk1 inactivation and dephosphorylation of Cdk1 substrates during mitotic exit likely serve as timing mechanisms to ensure that cytokinesis occurs after chromatid separation8-12. For example prior to anaphase, high Cdk1 activity blocks the accumulation of the cytokinetic regulators Aurora B and MKLP1 at the cleavage furrow and on the microtubules of the spindle midzone13-15.
Flavopiridol is a potent inhibitor of Cdk116. We found that treatment of vertebrate cells in mitosis with Flavopiridol resulted in premature mitotic exit accompanied by cytokinesis (Fig 1a and Supplementary Video 1). Similar results were recently found for the Cdk inhibitor, BMI-102617. Flavopiridol induced the microtubule network to undergo changes characteristic of anaphase and mitotic exit. The spindle poles moved apart, and microtubule bundles formed in the spindle midzone and at the equatorial cortex. Even though chromatid separation did not occur, cytokinetic furrows formed and ingressed to completion. The cleavage furrow trapped chromosomes in the midbody resulting in a “cut” phenotype. Nevertheless, the chromosomes decondensed and nuclear envelopes reformed. Eventually cytoplasmic contractile activity diminished as cells flattened fully onto the substratum, and the microtubule array established an interphase pattern.
During normal mitotic exit, Cdk1 activity is reduced by ubiquitylation and proteasome-mediated degradation of cyclin B3, 5. Proteasome inhibitors such as MG132 induce mitotic cells to arrest at metaphase. We found that Flavopiridol treatment overrode the metaphase arrest induced with MG132, causing mitotic exit and cytokinesis that was accompanied by chromosome decondensation and reformation of the nuclear envelope (Fig 1b and Supplementary Video 2). The proteolysis of cyclin B at mitotic exit is thought to ensure the uni-directionality of the M phase to G1 transition2. In cells where the proteasome was inhibited, we found that Flavopiridol-induced mitotic exit was reversible. Upon its removal, cells that had exited mitosis could return to metaphase (Fig. 1c and Supplementary Videos 3 and 4). The microtubules, having assumed an interphase configuration after Flavopiridol-induced mitotic exit, reassembled a mitotic spindle when Flavopiridol was removed. The midbody disappeared and the cytokinetic furrow retracted. The newly formed nuclear envelope dissolved. The chromosomes recondensed, attached to spindle microtubules, and realigned at the metaphase plate. This main finding is summarized in Supplementary Fig. 1. We found that cells induced to reverse back to metaphase could subsequently undergo a second, normal mitotic exit including chromatid separation and movement, as well as a second cytokinesis if the proteasome inhibitor was subsequently washed away (Supplementary Video 5). We used Flavopiridol for the majority of our experiments because of it high potency in inhibiting Cdk1 (Supplementary Fig 2). However, certain other chemical Cdk inhibitors also produced similar phenotypes (see Supplementary Information).
In Xenopus S3 cells arrested at metaphase with MG132, Aurora B kinase was concentrated at centromeres and MKLP1 was localized diffusely in the cytoplasm (Fig. 1d). Upon treatment of cells with Flavopiridol, Aurora B and MKLP1 rapidly accumulated at the equatorial cell cortex associating with the nascent cleavage furrows and at the midzone overlaying the microtubule bundles. Eventually the proteins became highly concentrated at the midbody. These changes were consistent with the typical relocation of Aurora B and MKLP1 in normal anaphase. We found that these protein translocation events were also reversible. In cells that were induced through mitotic exit then allowed to reverse back metaphase, Aurora B and MKLP1 returned to their typical metaphase distributions (Fig 1d, 60 min).
Metaphase cells treated with Flavopiridol in the presence of proteasome inhibitor advanced through mitotic exit and cytokinesis, reached the midbody stage with decondensed chromosomes and reassembled nuclear envelopes by 25 minutes post treatment (Fig 1b). Reversibility of mitotic exit was dependent on the duration of exposure to Flavopiridol. Treatment at 5 μM for 17 to 30 min resulted in the majority of cells (81 of 106) returning to metaphase upon Flavopiridol release (Table 1). With longer exposure to Flavopiridol the proportion of cells that underwent reversal declined. However, a few cells (3 of 24) still reversed when Flavopiridol was removed 80 min after its addition. We detected no reversal among cells treated for 90 min (n = 26). Reversibility was entirely dependent on the presence of the proteasome inhibitor. In medium without proteasome inhibitor, premature mitotic exit and cytokinesis induced with Flavopiridol did not reverse in any of 13 prometaphase cells when the Flavopiridol was removed at 17 – 20 min.
Table 1.
Duration of Flavopiridol treatment (min) | # cells imaged | # cells reversed through mitotic exit (% reversed) |
---|---|---|
17 |
29 |
28 (97%) |
20 |
27 |
25 (93%) |
25 |
27 |
17 (63%) |
30 |
23 |
11 (48%) |
33 - 37 |
17 |
8 (47%) |
40 |
22 |
6 (27%) |
50 |
19 |
4 (21%) |
60 |
17 |
1 (06%) |
70 |
24 |
1 (04%) |
80 |
24 |
3 (12%) |
90 | 26 | 0 (0%) |
Previous studies have shown that several early mitotic regulators (e.g. cyclin A) are normally degraded during prophase and prometaphase and, as well, during the prometaphase arrest induced with microtubule drugs6, 7. Xenopus S3 cells were arrested with the microtubule drug, nocodazole, to allow degradation of early mitotic regulators. Cells were then released from nocodazole to medium containing proteasome inhibitor. The cells assembled spindles and arrested at metaphase. These cells were fully capable of undergoing Flavopiridol-induced mitotic exit and reversal (Fig. 2a and Supplementary Video 6). This result was consistent with the interpretation that mitotic exit reversal in cells treated with proteasome inhibitor was due to the protection of cyclin B from degradation and not due to the preservation of early mitotic regulators.
Cell cycle markers were measured during induced mitotic exit and reversal. Nocodazole-arrested Hela cells were treated with Flavopiridol according to the scheme depicted in Fig 2b. Mitotic indices were calculated from the percentage of cells with condensed chromosomes (Fig 2c, top), and samples were blotted for various cell cycle markers (Fig 2c, bottom). While the level of Cdk1 protein remained constant, Flavopiridol induced mitotic exit and degradation of cyclins B1 and B2 in the absence of the proteasome inhibitor, MG132 (Group 3, −MG132). In medium containing MG132, Flavopiridol induced mitotic exit, but cyclin B1 and B2 were preserved (Group 3, +MG132). Flavopiridol removal at 25 min caused reversal back to M phase in the presence of MG132 (Group 5, +MG132, asterisk), but not in its absence (Group 5, −MG132). As expected, Cyclin A, prominent when the initial population of mitotic cells was analyzed −45 min (Group 1) was substantially degraded during the subsequent 45 min incubation before Flavopiridol addition (Group 2). When the Flavopiridol was removed at 25 min for the mitotic exit reversal experiments, cyclin A was very low (Group 3) and thus was unlikely to contribute significantly to the reactivation of Cdk1. At longer times after treatment with Flavopiridol, some resynthesis of cyclin A protein appeared to occur in cells with MG132 (Group 4 and 5, +MG132). In summary, reversal of mitotic exit after removal of Flavopiridol correlates with the preservation of cyclin B1 and B2.
Flavopiridol was an effective inhibitor when purified Cdk1/CyclinB was used to phosphorylate histone H1 in vitro showing higher potency than certain commonly used Cdk1 inhibitors such as olomoucine and roscovitine (Supplementary Fig. 2). In living cells Flavopiridol resulted in the loss of known Cdk1 target phosphorylations. In both the presence and absence of MG132, Flavopiridol eliminated labeling with the TG-3 antibody that recognizes a Cdk1-catalyzed phosphoepitope on the protein Nucleolin18 (Group 3). Cdh1, an activator of the mitotic ubiquitin ligase APC/C, is inhibited by Cdk1 phosphorylation during M phase19. Flavopiridol treatment induced an increase the electrophoretic mobility of Cdh1 but, for yet unknown reasons, only in the absence of MG132. Cdh1 dephosphorylation may be involved the ubiquitylation and breakdown of cyclin B after Flavopiridol treatment in the absence MG132 (Group 3, −MG132).
Like Xenopus S3 cells, Hela cells cultured in MG132 exhibited a defined period after Flavopiridol-induced mitotic exit when removal of Flavopiridol led to reversal back to metaphase. When Flavopiridol was removed 25 min after its addition, 10 of 10 cells (100%) reversed back to metaphase. When removed at 60 min, only 2 of 12 cells (17%) underwent reversal. To test whether preservation of cyclin B alone was sufficient for reversal of mitotic exit and cytokinesis, a series of experiments were performed using non-degradable cyclin B1 fused to green fluorescent protein (R42A-CyclinB1-GFP) in place of the proteasome inhibitor. Most Hela cells expressing non-degradable cyclin B arrested at metaphase as was previously shown12. When these metaphase-arrested cells were treated with Flavopiridol, they completed cytokinesis and exited mitosis. In some cells, Flavopiridol treatment also induced chromatids to separate and to move to the poles before mitotic exit and cytokinesis. When the Flavopiridol was removed 25 min after its addition, 17 of 37 transfected cells underwent reversal back to M phase. In all 17 cells that underwent mitotic exit reversal, the chromosomes recondensed within 10 min after Flavopiridol removal. Cells that had not separated their chromatids opened their cleavage furrows rapidly (average time = 13 min after Flavopiridol removal) and reversed back to metaphase within 30 min (Fig. 3a and Supplementary Video 7). Flavopiridol addition induced chromatid separation in 7 of the 17 cells that subsequently underwent mitotic exit reversal. In these cells the chromatids recondensed rapidly (within 10 minutes after Flavopiridol removal) but opening of the cleavage furrow was somewhat delayed (average time = 45 min after Flavopiridol removal) (Figure 3b and Supplementary Video 8). The separated chromatids remained near the center of the cell but did not realign to form a metaphase plate. Two of the 7 cells with separated chromatids did not exhibit evidence of reopening their cleavage furrows at their last observation time of 90 min after Flavopiridol removal. Thus cells with separated chromatids can undergo reversal of mitotic exit and cytokinesis but, predictably, the chromatids cannot realign at the metaphase plate if sister chromatids have lost cohesion. Moreover while cytokinesis is reversible after chromatid segregation, reopening of the cleavage furrow generally took longer to initiate. Possibly the trapped chromatin physically facilitates reopening of the furrow, or cells that separate their chromatids may have biochemical differences that restrict reopening of the furrow. Mitotic cells expressing high levels of wild type cyclin B fused to GFP underwent mitotic exit and cytokinesis upon Flavopiridol treatment, but as expected, removing Flavopiridol at 25 min never resulted in reversal of mitotic exit (n = 6) (Fig. 3c and Supplementary Video 9).
Cells expressing high levels of non-degradable cyclin B, as measured by GFP fluorescence, were more likely to undergo reversal (Fig. 3d). Previous work suggests that cell cycle regulation in M phase exhibits the property of hysteresis because entry into M phase from G2 requires higher Cdk1 activity than does maintaining M phase once it is achieved20, 21. This hysteresis may explain why many cells expressing relatively low levels of non-degradable cyclin B arrested at metaphase before Flavopiridol treatment but were unable to undergo reversal when Flavopiridol was removed. In addition, in confirmation of our previous conclusion that the reversal of mitotic exit and the reestablishment of high Cdk1 activity levels in proteasome inhibitor are likely driven by cyclin B and not cyclin A, we found cells arrested at metaphase by expression of non-degradable cyclin B had nearly undetectable levels of cyclin A (Fig. 3e).
The studies above were completed using immortalized cell lines. To determine whether these findings were broadly applicable to non-immortalized cells, we also tested primary human keratinocytes. We found that these cells, cultured in medium containing MG132, underwent mitotic exit and cytokinesis in response to Flavopiridol treatment, and reversed back to metaphase when Flavopiridol was removed (Fig. 4 and Supplementary Video 10).
The importance of regulated proteolysis for guiding the directionality of cell cycle transitions is a central tenet of modern cell cycle theory extended from the founding work on the proteolysis of cyclins in early embryos1-4, 22. Here we have tested this idea, showing that if cyclin B is stabilized, the events of mitotic exit and cytokinesis are immediately reversible even in cells that have progressed through chromosome decondensation, reconstitution of the interphase nucleus, and mitotic spindle disassembly and cytokinesis to the stage of midbody formation. However, with increasing time after Flavopiridol-induced mitotic exit, reversibility wanes. Thus cyclin degradation does provide directionality during mitotic exit but this function is later supplemented by other events that inhibit backtracking. These events are likely to involve other biochemical changes to Cdk1 including phosphorylation changes or the binding of small molecule inhibitors. The ability to regulate mitotic exit in the absence of cyclin degradation in vertebrate cells will facilitate elucidating these downstream regulators of the M phase to G1 phase transition in vertebrate cells.
METHODS
Cell culture and live cell imaging
Xenopus S3 cells stably expressing GFP-α-tubulin were grown at 23°C in 70% L-15 medium supplemented with 15% FBS. For control images of these cells transiting M phase see Supplementary Video 12. HeLa cells were grown in DMEM with 10% FBS in 5% CO2 at 37°C. Normal human epidermal keratinocytes (Clonectics) were grown in Defined Keratinocyte SFM (InVitrogen). Time-lapse phase contrast and fluorescence images were collected from cells grown on glass coverslips using with a Zeiss Axiovert 200M microscope equipped with a Hamamatsu ORCA camera. When applicable, cells were incubated for 30 - 60 min with 25 μM MG132 (Calbiochem) or with 100ng/ml nocodazole (Sigma). Flavopiridol was provided by the National Cancer Institute (Drug Synthesis and Chemistry Branch, Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis), and used at 5 - 10μM. In cells expressing Cyclin B1 fused to GFP, levels of expression were quantified using Metamorph software (Molecular Devices).
Immunofluorescence and Immunoblotting
Cells on coverslips were fixed in 2 – 3% formaldehyde in 60mM PIPES, 25Mm HEPES (pH 6.8), 10mM EDTA, 4mM MgCl2 (PHEM) for 15 minutes, then permeabilized with 0.5% or 1% Triton X-100. Hela cells in suspension were fixed in PHEM buffer containing 1.5% formaldehyde and 0.5% Triton X-100. Antibodies to Xenopus MKLP1 and Aurora B were prepared in rabbits. Mouse antibody to human cyclin A was obtained from J. Gannon and T. Hunt or obtained commercially (Abcam). Fluorescence images were gathered with a Zeiss Axioplan II microscope with a Hamamatsu ORCA-II camera. Samples were treated with DAPI and mounted in Vectashield containing 10 mM MgSO4. Cells were lysed in NuPAGE (Invitrogen) western blotting sample buffer containing 50 mM DTT. Samples were electrophoresed on 4% - 12% gradient gels, transferred to PVDF membranes, and probed with primary antibodies overnight at 4°C. Blots were treated with appropriate HRP-conjugated secondary antibodies and were developed using SuperSignal West Pico Chemiluminescent Substrate (Pierce).
Supplementary Material
Acknowledgements
We thank Wendy Martin and staff of the OMRF Imaging Center for technical assistance. We thank Sanofi-Aventis Pharmaceuticals, Inc. and the National Cancer Institute for providing Flavopiridol. We thank Dr. Jonathan Pines for providing cyclin B plasmids and Dr. Peter Davies for TG-3 anti-Phospho-Nucleolin antibody. We thank Drs. Julian Gannon and Tim Hunt for antibodies to Cdk1 and to cyclins. This work was supported by grants from the National Institutes of Health (to GJG and PTS), from the American Cancer Society (to PTS), and from an NIH training grant (to DLS).
Footnotes
Supplementary Information is linked to the online version of the paper at www.nature.com/nature. A figure summarizing the primary finding of the paper is available as Supplementary Fig. 1.
References
- 1.Minshull J, et al. The role of cyclin synthesis, modification and destruction in the control of cell division. J Cell Sci. 1989;(Suppl 12):77–97. doi: 10.1242/jcs.1989.supplement_12.8. [DOI] [PubMed] [Google Scholar]
- 2.Reed SI. Ratchets and clocks: the cell cycle, ubiquitylation and protein turnover. Nat Rev Mol Cell Biol. 2003;4:855–64. doi: 10.1038/nrm1246. [DOI] [PubMed] [Google Scholar]
- 3.Glotzer M, Murray AW, Kirschner MW. Cyclin is degraded by the ubiquitin pathway. Nature. 1991;349:132–8. doi: 10.1038/349132a0. [DOI] [PubMed] [Google Scholar]
- 4.Murray AW, Kirschner MW. Cyclin synthesis drives the early embryonic cell cycle. Nature. 1989;339:275–80. doi: 10.1038/339275a0. [DOI] [PubMed] [Google Scholar]
- 5.Clute P, Pines J. Temporal and spatial control of cyclin B1 destruction in metaphase. Nat Cell Biol. 1999;1:82–7. doi: 10.1038/10049. [DOI] [PubMed] [Google Scholar]
- 6.Geley S, et al. Anaphase-promoting complex/cyclosome-dependent proteolysis of human cyclin A starts at the beginning of mitosis and is not subject to the spindle assembly checkpoint. J Cell Biol. 2001;153:137–48. doi: 10.1083/jcb.153.1.137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.den Elzen N, Pines J. Cyclin A is destroyed in prometaphase and can delay chromosome alignment and anaphase. J Cell Biol. 2001;153:121–36. doi: 10.1083/jcb.153.1.121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cross FR, Schroeder L, Kruse M, Chen KC. Quantitative characterization of a mitotic cyclin threshold regulating exit from mitosis. Mol Biol Cell. 2005;16:2129–38. doi: 10.1091/mbc.E04-10-0897. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Parry DH, O'Farrell PH. The schedule of destruction of three mitotic cyclins can dictate the timing of events during exit from mitosis. Curr Biol. 2001;11:671–83. doi: 10.1016/s0960-9822(01)00204-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sigrist S, Jacobs H, Stratmann R, Lehner CF. Exit from mitosis is regulated by Drosophila fizzy and the sequential destruction of cyclins A, B and B3. Embo J. 1995;14:4827–38. doi: 10.1002/j.1460-2075.1995.tb00164.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Visintin R, et al. The phosphatase Cdc14 triggers mitotic exit by reversal of Cdk-dependent phosphorylation. Mol Cell. 1998;2:709–18. doi: 10.1016/s1097-2765(00)80286-5. [DOI] [PubMed] [Google Scholar]
- 12.Chang DC, Xu N, Luo KQ. Degradation of cyclin B is required for the onset of anaphase in Mammalian cells. J Biol Chem. 2003;278:37865–73. doi: 10.1074/jbc.M306376200. [DOI] [PubMed] [Google Scholar]
- 13.Carmena M, Earnshaw WC. The cellular geography of aurora kinases. Nat Rev Mol Cell Biol. 2003;4:842–54. doi: 10.1038/nrm1245. [DOI] [PubMed] [Google Scholar]
- 14.Glotzer M. The molecular requirements for cytokinesis. Science. 2005;307:1735–9. doi: 10.1126/science.1096896. [DOI] [PubMed] [Google Scholar]
- 15.Mishima M, Pavicic V, Gruneberg U, Nigg EA, Glotzer M. Cell cycle regulation of central spindle assembly. Nature. 2004;430:908–13. doi: 10.1038/nature02767. [DOI] [PubMed] [Google Scholar]
- 16.Losiewicz MD, Carlson BA, Kaur G, Sausville EA, Worland PJ. Potent inhibition of CDC2 kinase activity by the flavonoid L86-8275. Biochem Biophys Res Commun. 1994;201:589–95. doi: 10.1006/bbrc.1994.1742. [DOI] [PubMed] [Google Scholar]
- 17.Niiya F, Xie X, Lee KS, Inoue H, Miki T. Inhibition of cyclin-dependent kinase 1 induces cytokinesis without chromosome segregation in an ECT2 and MgcRacGAP-dependent manner. J Biol Chem. 2005;280:36502–9. doi: 10.1074/jbc.M508007200. [DOI] [PubMed] [Google Scholar]
- 18.Dranovsky A, et al. Cdc2 phosphorylation of nucleolin demarcates mitotic stages and Alzheimer's disease pathology. Neurobiol Aging. 2001;22:517–28. doi: 10.1016/s0197-4580(00)00248-7. [DOI] [PubMed] [Google Scholar]
- 19.Kramer ER, Scheuringer N, Podtelejnikov AV, Mann M, Peters JM. Mitotic regulation of the APC activator proteins CDC20 and CDH1. Mol Biol Cell. 2000;11:1555–69. doi: 10.1091/mbc.11.5.1555. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pomerening JR, Sontag ED, Ferrell JE., Jr. Building a cell cycle oscillator: hysteresis and bistability in the activation of Cdc2. Nat Cell Biol. 2003;5:346–51. doi: 10.1038/ncb954. [DOI] [PubMed] [Google Scholar]
- 21.Sha W, et al. Hysteresis drives cell-cycle transitions in Xenopus laevis egg extracts. Proc Natl Acad Sci U S A. 2003;100:975–80. doi: 10.1073/pnas.0235349100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Evans T, Rosenthal ET, Youngblom J, Distel D, Hunt T. Cyclin: a protein specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage division. Cell. 1983;33:389–96. doi: 10.1016/0092-8674(83)90420-8. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.