Checkpoint control of the G2/M phase transition during the first mitotic cycle in mammalian eggs

J. Fulka, Jr1,2,4, N.L. First3, J. Fulka11 and R.M. Moor2

1 The Institute of Animal Production, CS-104 01 Prague 10, Czech Republic, 2 The Babraham Institute, Development and Genetics Programme, Babraham, Cambridge CB2 4AT, UK, and 3 3University of Wisconsin, Department of Meat and Animal Science, 1675 Observatory Drive, Madison, WI 53706, USA


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The high incidence of chromosomally abnormal human embryos is frequently assumed to be due to a lack of checkpoint controls operating during early embryogenesis. In our study we have analysed when these mechanisms first become functional. Mouse oocytes treated in late metaphase I with either of two different cyclin-dependent kinase inhibitors [butyrolactone 1 (BL1) or 6-dimethylaminopurine (6-DMAP)] form nuclei in the cytoplasm. BL1-treated eggs enter S-phase at 16–18 h post-treatment and, after completion of DNA synthesis, cleave to 2-cell stage embryos. 6-DMAP treatment results in the rapid initiation of DNA synthesis, its completion by 12 h and then arrest in the G2 phase. Thus, two different cell cycle stages can be obtained at the same time point after the initiation of treatment: G1- after BL1 and G2-staged nuclei after 6-DMAP treatment. That this approach greatly facilitates cell cycle studies has been shown by analysing checkpoint function during the first division. Whilst G2-staged eggs enter M phase within 2–3 h when 6-DMAP is washed out, the onset of M phase is delayed after their fusion to G1 (BL1) cells. Here M phase occurs only after the less advanced nucleus completes DNA replication. Our results indicate that checkpoints in mammalian eggs are functional during the first mitotic cycle.

Key words: checkpoints/cyclin-dependent kinase inhibitors/DNA replication/mitosis/oocytes


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Cell cycle checkpoint controls ensure the fidelity of cell division and can basically be divided into two groups: firstly, interphase checkpoints which monitor the completion of the preceding phase before the entry into the next phase, e.g. DNA replication must be completed before the onset of G2 phase and mitosis; and secondly, M-phase checkpoints controlling the precise separation of chromosomes during meiosis and mitosis (Elledge, 1996Go; Page and Orr-Weaver, 1997Go; Hardwick, 1998Go). The absence of these controls results in genetic instability associated with cancer, cell death or birth defects such as Down's syndrome. Whilst these controls are well documented in somatic cells and function during spermatogenesis (Odorisio et al., 1998Go), we have surprisingly little knowledge about their function during oogenesis and early embryogenesis in mammals (Fulka et al., 1998Go). The remarkably high incidence of chromosomal abnormalities in preimplantation human embryos has led to the postulation that cell cycle checkpoint controls are absent or restricted during the early stages of development (Delhanty and Handyside, 1995Go; Handyside and Delhanty, 1997Go). This has, however, not been proven mainly because appropriate methodological approaches are yet not available, embryonic cell cycle staging is rather complicated and embryos are sensitive when manipulated. Nevertheless, it has already been clearly demonstrated that these controls are absent in fully grown mammalian oocytes both in the immature (GV) oocyte and also during meiotic maturation (Fulka et al., 1995Go, 1997Go; LeMaire-Adkins et al., 1997Go). A novel technical advance we have made recently whilst studying the mechanisms involved in the exit from metaphase to anaphase has enabled us to extend our observations on checkpoint controls to the embryonic first mitotic cycle. The results in this paper show that checkpoint controls are functional during the first mitotic cycle in mammalian eggs.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Mouse oocytes were isolated from large antral follicles of pregnant mare's serum gonadotrophin-stimulated females (C57BL/6JxCBA/Ca) injected 44–48 h previously. The oocytes were manipulated in M2 medium supplemented with bovine serum albumin (BSA) (4 mg/ml). Their cumulus cells were removed by pipetting and only those oocytes with germinal vesicles (GV) were used and cultured in M199 medium (M 4530) containing sodium-pyruvate (0.2 mmol/l), gentamicin (25 µg/ml) and BSA (4 mg/ml) at 37°C in 5% CO2 in air. The oocytes were inspected after 90 min and those still containing visible GVs were discarded. The remaining oocytes were cultured for another 6.5 h and then transferred into medium supplemented with either (i) butyrolactone 1 (BL1) at a final concentration of 75 µmol/l, or (ii) 6-dimethylaminopurine (6-DMAP) at a concentration of 2.5 mmol/l and cultured thereafter as indicated in the Results section. Butyrolactone 1 is a selective inhibitor of the cyclin-dependent kinase family, whilst 6-DMAP inhibits a broad spectrum of protein kinases. The optimal concentration of inhibitors was assessed in preliminary experiments. BL1 [{alpha}-oxo-ß(p-hydroxyphenyl)-{gamma}-(para-hydroxy-m-3,3-dimethylallyl-benzyl)-{gamma}-methoxycarbonyl-{gamma}-butyrolactone] was purchased from Funakoshi (Tokyo, Japan) and dissolved in dimethylsulphoxide. DNA replication was analysed after bromodeoxyuridine (BrDU) labelling exactly as described in our previous paper (Ouhibi et al., 1994Go). Spindles, nuclei and chromosomes were viewed with a confocal microscope (MRC-600, BioRad, Cambridge, UK) after appropriate labelling with anti-tubulin antibody (McAb, Sera-Lab, Loughborough, UK) and staining with propidium iodide (Lin et al., 1996Go). Fusion between cells (see Results) was induced by polyethyleneglycol (PEG) exactly as described previously (Fulka et al., 1997Go). When GV-stage oocytes were used for fusion spontaneous germinal vesicle breakdown (GVBD) was prevented by dibutyrylcyclic AMP (dbcAMP-150 µg/ml). For optical microscopy oocytes were fixed in aceto-alcohol, stained in orcein and examined under phase-contrast optics. Each experiment was repeated at least three times.

Unless otherwise stated all chemicals were purchased from Sigma (Prague, Czech Republic).


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Phosphorylation inhibitors provide a reliable method of producing G1 and G2 stage zygotes
Under our culture conditions GVBD typically occurred within 1.5 h of explantation and when oocytes were cultured for 8 h meiosis had progressed to the metaphase I (MI) stage. Thereafter the transition from MI to anaphase–telophase I (A-TI) could be detected. When oocytes were transferred at 8 h into medium supplemented with BL1 a first polar body (1PB) was extruded but meiosis did not proceed to metaphase II (MII). Instead, an interphase nucleus (Figure 1Go) was formed in the cytoplasm in most cases (410/475; 86%). A similar situation was observed when 6-DMAP was added to the medium, but when using this inhibitor the results were less consistent (see Table IGo). Confocal microscopy showed that 6 h after addition of either BL1 or 6-DMAP fully formed nuclei were observed (Figure 2Go) but their size gradually increased (Figure 3Go) with further cultivation (10 oocytes examined in each group).



View larger version (116K):
[in this window]
[in a new window]
 
Figure 1. Mouse metaphase I oocyte treated with butyrolactone 1 (BL1) and examined after 16 h of culture. Nucleus (N) and the first polar body (PB) are clearly visible.x600.

 

View this table:
[in this window]
[in a new window]
 
Table I. The effect of butyrolactone 1 (BL1) or 6-dimethylaminopurine (6-DMAP) on metaphase I mouse oocytes
 


View larger version (83K):
[in this window]
[in a new window]
 
Figure 2. After 6 h post-BL1 or 6-dimethylaminopurine (6-DMAP) treatment of metaphase I oocytes, small nuclei with clear nucleoli are visible. Propidium iodide (PI) staining, confocal microscopy x1000.

 


View larger version (107K):
[in this window]
[in a new window]
 
Figure 3. Nuclei (BL1, 6-DMAP) increase their size with extended treatment (16 h). PI staining, confocal microscopy x1000.

 
Two entirely different responses were observed when oocytes were cultured overnight (12–14 h) in 6-DMAP or BL1 followed by the removal of the inhibitor and subsequent culture in inhibitor-free medium. Oocytes treated with 6-DMAP underwent nuclear envelope breakdown within 2–3 h and metaphase II-like chromosomes (90/96) were detected after staining. In contrast, after inhibitor removal nuclear envelope breakdown and chromosome condensation in BL1-treated cells occurred much later (8–10 h) than after 6-DMAP treatment, and oocytes progressed directly to 2-cell stage embryos (60/82). When metaphases (37) were analysed in BL1-treated embryos during cleavage, the chromosomes showed the typical morphological characteristics observed in normal embryos during cleavage from the 1- to 2-cell stage (Figure 4Go). Embryos at the 2-cell stage derived from BL1-treated oocytes were morphologically indistinquishable from normal embryos, their nuclei synthesized DNA and they developed readily up to the blastocyst stage (data not shown).



View larger version (78K):
[in this window]
[in a new window]
 
Figure 4. A composed confocal microscopy picture showing the first mitotic metaphase in BL1-treated oocytes. M, metaphase; S, spindle. x500.

 
In addition to the differences outlined above, BL1- and 6-DMAP-treated oocytes differed in another important respect. For example, nuclear envelope breakdown in 6-DMAP treated oocytes was prevented when these cells were cultured in dbcAMP-supplemented medium (54/57). This was not the case in BL1-treated cells. In these eggs cleavage occurred even in the presence of dbcAMP (27/38). In order to explain this paradoxical situation, oocytes of both groups were incubated in medium with BrDU and thereafter processed for fluorescence microscopy. DNA fluorescence could be detected in 6-DMAP-treated oocytes (75/97; 77%) when fixed at 14–16 h after the beginning of 6-DMAP treatment (Figure 5Go). When oocytes were incubated with BrDU after this interval no labelled nuclei were detected (0/48). The opposite situation was observed in BL1-treated oocytes, where after 14–16 h of incubation in BL1 with BrDU only sporadically labelled nuclei were detected (5/72;7%). This suggested that DNA replication might occur in these cells with some delay. Indeed, by extending the labelling period beyond 16 h or by starting BrDU labelling later it was clearly demonstrated that DNA replication began in these eggs at ~18 h after BL1 treatment (70/101; 69%). Thus at the same time point two different cell cycle stages could be obtained from the same original population of oocytes by differential inhibition. At 14–16 h after the beginning of 6-DMAP treatment the oocytes had already undergone DNA replication and were arrested in G2 stage. On the other hand oocytes treated with BL1 were in late G1 stage (just before the onset of DNA replication) at this time.



View larger version (90K):
[in this window]
[in a new window]
 
Figure 5. DNA replication detected in nuclei of BL1- or 6-DMAP-treated metaphase I oocytes. Confocal microscopy x1200.

 
Functional interphase checkpoints are established in the first mitotic cycle
Advantage was taken of the ability to produce eggs at precisely defined cell cycle stages in fusion experiments designed to identify potential checkpoints in the first mitotic cycle. The results showed that when two 6-DMAP-treated oocytes were fused together and thereafter cultured in normal medium, both nuclear membrane disassembly and chromatin condensation were detected in all cases 2–3 h later; this time interval was the same as that in both unfused and control oocytes. In contrast, when two BL1-treated oocytes were fused and cultured in normal medium, nuclear envelope breakdown occurred (as in controls) within 8–10 h. Surprisingly, after fusion of one 6-DMAP oocyte to one BL1-treated oocyte, both nuclei were still intact after 2–3 h of culture in normal medium, and chromosome condensation was not detected until 7–9 h after fusion (Figure 6Go). This is in contrast to the situation in unfused cells, where one oocyte underwent nuclear envelope breakdown within 2–3 h (6-DMAP) whilst the other (BL1) contained an intact nucleus at this time. In giant cells produced by fusion of G1 (BL1) with G2 (6-DMAP) stage eggs, the onset of nuclear envelope breakdown and chromosome condensation were accelerated by approximately 1–2 h as compared with fusions of G1 with G1 stage eggs (BL1xBL1). This indicated that the less advanced nucleus (BL1) must have replicated its DNA, and only after this process was completed could nuclear membrane breakdown and chromosomes condensation occur in the more advanced fusion partner (6-DMAP): this constraint is comparable to that in somatic cells. The results are summarized in Table IIGo. Giant BL1x6-DMAP cells were labelled with BrDU shortly after fusion and fixed before the expected time of nuclear envelope breakdown to verify that cell cycle progression in the G2-fusion partner was delayed until S-phase was completed. As expected, only one nucleus (BL1) was labelled whilst the other showed no labelling (Figure 7Go). In total we evaluated 57 giant cells produced by fusion. In 48 of these one nucleus was labelled whilst the other showed no labelling. In nine remaining cells no labelling could be detected in either nucleus.



View larger version (139K):
[in this window]
[in a new window]
 
Figure 6. When BL1- and 6-DMAP-treated mouse metaphase I oocytes with nuclei are fused and cultured in normal medium, both nuclei remain intact after 2–3 h of culture. Phase contrast x800.

 

View this table:
[in this window]
[in a new window]
 
Table II. Fusion of butyrolactone (BL1)-treated oocytes (G1) with 6-dimethylaminopurine (6-DMAP)-treated (G2) or with GV-stage oocytes
 


View larger version (98K):
[in this window]
[in a new window]
 
Figure 7. A composed picture showing bromodeoxyuridine (BrDU) labelling of fused cells (BL1 and 6-DMAP). Left, nuclei are stained red with propidium iodide (PI). Right, only one nucleus exhibits DNA replication (stained green, BL1). Original magnification x400.

 
The conclusion that the less advanced nucleus must replicate DNA before the M-phase onset was supported also by the morphology of chromosomes (Figure 8Go) which had a normal metaphase structure (52). In only a few cases (n = 9) was a normal metaphase detected together with prematurely condensed S-phase chromosomes (PCC) (Figure 9Go). The giant cells were mostly arrested at the metaphase stage and only rarely cleaved to a 2-cell-like stage. These results clearly demonstrated the reappearance of checkpoint controls which were able to detect the unreplicated DNA and prevent M-phase onset. The same results were obtained when different cell cycle stage combinations (only after BL1 treatment) were tested: G1xS and G2xS (not shown). In order to analyse further these mechanisms we fused BL1-treated oocytes at 18 h (G1) to immature GV-stage oocytes (G2). Fused cells were cultured for 1 h after fusion in dbcAMP-supplemented medium and thereafter in inhibitor-free medium. This fusion combination, however, resulted in a completely different response from that outlined above. In all cases germinal vesicle breakdown occurred together with nuclear envelope breakdown and premature chromosome condensation of BL1 nucleus (35 giant cells evaluated; Figure 10Go; Table IIGo). The findings showed that the G2-stage nucleus in these oocytes was not responsive to the inhibitory influence of a G1 partner. These observations indicated that checkpoint controls were only established in early development after the nucleus had become competent to replicate DNA. Taken together we believe that our results clearly demonstrated that in mammalian embryos the basic checkpoint controls become established during the first mitotic cell cycle but only after the nuclei have become competent to replicate DNA.



View larger version (120K):
[in this window]
[in a new window]
 
Figure 8. After DNA replication is completed in BL1 nucleus, nuclear envelope breakdown (NEBD) of both nuclei occurred and typical M-phase condensed chromosomes are observed. Phase contrast. x800

 


View larger version (121K):
[in this window]
[in a new window]
 
Figure 9. Exceptionally NEBD occurred even in the presence of unreplicated chromatin in one nucleus. (A) replicated chromosomes originated from BL1 nucleus, (B) fragmented S-phase-like prematurely condensed chromosomes from 6-DMAP nucleus. Phase contrast x900.

 


View larger version (121K):
[in this window]
[in a new window]
 
Figure 10. Fusion of GV-stage oocytes (G2) to BL1 (G1)-treated oocytes always resulted in germinal vesicle breakdown and premature NEBD of BL1 nucleus. (A) germinal vesicle chromosomes, (B) prematurely condensed chromosomes of BL1 nucleus. Phase contrast x800.

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Our results indicate that at least certain checkpoint controls operate in mammalian eggs during the first mitotic cell cycle stage. However, as pointed out (Delhanty and Handyside; 1995) some postzygotic abnormalities cannot arise during gametogenesis or at fertilization but must arise during early cleavages. To account for this apparent contradiction, one explanation is that the first cell cycle in mammalian embryos may be rather specific and that checkpoint controls may thereafter temporarily disappear during subsequent cleavages. The second explanation is that the early mitotic checkpoints are more error-prone than those in somatic cells. Thirdly, as we outlined above, there are basically two types of checkpoints: M phase and interphase type. Our experiments used the scheme originally developed in 1971 (Johnson and Rao, 1971Go). These workers found in somatic cells that after fusion of two different cell cycle stage cells, for example G1xG2, the onset of M phase is delayed until the less advanced cell completes DNA replication and thereafter both nuclei enter into M phase. These observations are the basis for our current checkpoint control studies and our results clearly show that these mechanisms operate in the same way in mammalian eggs. On the other hand in amphibian embryos these controls are established much later (mid-blastula stage) and the cleavage occurs even in the presence of DNA synthesis inhibitors (Kimelman et al, 1987Go). Our results thus indicate that the defect could be specifically in the M phase checkpoint control where the mechanisms are probably more fragile than those operating in interphase. As we demonstrated recently in fused maturing mouse oocytes, one normal metaphase is sufficient to drive meiotic maturation irrespective of the presence of a second damaged metaphase (Fulka et al., 1997Go) and the same situation exists also in somatic cells (Rieder et al., 1997Go). Clearly some additional studies with early mammalian embryos will be necessary to demonstrate why post-zygotic abnormalities occur. In the present experiment the fusion of a BL1 oocyte to a GV-stage oocyte does not prevent germinal vesicle breakdown. Thus it is possible that both nuclei must be competent to replicate DNA for the full establishment of checkpoint controls. Interestingly even in some fused (BL1xBL1) cells we observed chromosomes with a normal morphology together with non-replicated prematurely condensed chromosomes. Also in some intact BL1-treated cells DNA replication did not occur. Thus it is possible that some giant cells contained one replicating nucleus and one nucleus which does not replicate DNA for some unknown reason(s). The defective nucleus behaves as an inert cell component and does not influence the onset of M phase. An analogous situation could exist theoretically in human zygotes when one pronucleus will replicate DNA whilst the other will not. The developmental fate of these zygotes is unknown. Equally it is not known whether DNA, unreplicated during the first round, can undergo replication at a later stage. An interesting phenomenon we observed, but only after BL1 treatment, was the direct transition from MI through interphase into mitosis with subsequent cleavage to the 2-cell embryonic stage. This has already been described when mouse MI oocytes are treated with cycloheximide. When the inhibitor is washed out and oocytes with nuclei are incubated in medium with dbcAMP DNA replication is initiated and embryos develop up to the blastocyst stage (Clarke et al., 1988Go). When BL1 is used DNA replication is initiated irrespective of its continued presence or absence in the medium. The intention of our study was not the examination of development or the evaluation of ploidy in embryos produced. These embryos, however, seem to be diploid, their nuclei synthesise DNA and they can develop at least to the blastocyst stage. The molecular mechanisms involved in the regulation of direct transition from MI into interphase, DNA replication, embryo ploidy and development certainly deserve some additional studies which are now in progress.

In conclusion, we believe that our study clearly shows that at least in the mouse the basic checkpoint controls are already established during the first mitotic cell cycle after the nuclei become competent to replicate DNA. Whether the establishment of checkpoint controls is associated with the activation of the embryonic genome is still unclear and additional experiments will solve this question. The high incidence of chromosomal embryonic abnormalities in humans may also indicate that these controls are established in several steps. Further studies are clearly necessary to assess when all the checkpoint controls are fully functional in mammalian embryos. It would be also interesting to know which mechanisms, from the cell cycle control perspective, regulate the onset of M phase in binucleate cells with one nucleus still replicating DNA. Recent results in yeasts and human somatic cells identified cdc25C as a target for checkpoint control (Weinert, 1997Go). We believe that the simple and effective approach described in the present paper will be useful and accelerate further studies in mammalian embryos (cell cycle experiments, DNA replication regulation, M phase exit, etc.) because it eliminates the problems associated with embryo sensitivity to different manipulations which frequently influence the results of experiments in a negative manner.


    Acknowledgments
 
This experiment was in part (J.F.Jr.) supported via a fellowship under the OECD Co-operative Research Programme: Biological Resource Management for Sustainable Agricultural Systems. The support of J.F.Jr's laboratory from The US–Czech Science and Technology Joint Fund in co-operation with USDA (95047), GACR 524/96/K162 and MZe NAZV EP0960006200 is also gratefully acknowledged.


    Notes
 
4 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Clarke, H.J., Rossant, J. and Masui, Y. (1988) Suppression of chromosome condensation during meiotic maturation induces parthenogenetic development of mouse oocytes. Development, 104, 97–103.[Abstract]

Delhanty, J.D.A., and Handyside, A.H. (1995) The origin of genetic defects in the human and their detection in the preimplantation embryo. Hum. Reprod. Update, 1, 201–215.[ISI][Medline]

Elledge, S.J. (1996) Cell cycle checkpoints: preventing an identity crisis. Science, 274, 1664–1672.[Abstract/Free Full Text]

Fulka, Jr, J., Moor, R.M. and Fulka, J. (1995) Replicating DNA does not block germinal vesicle breakdown in mouse oocytes. J. Exp. Zool., 272, 245–248.[ISI][Medline]

Fulka, Jr, J., Kalab, P., First, N.L. et al. (1997) Damaged chromatin does not prevent the exit from metaphase I in fused mouse oocytes. Hum. Reprod., 12, 2473–2476.[Abstract]

Fulka, Jr, J., First, N.L and Moor, R.M. (1998) Nuclear and cytoplasmic determinants involved in the regulation of mammalian oocyte maturation. Mol. Hum. Reprod., 4, 41–49.[Abstract]

Handyside, A.H. and Delhanty, J.D.A. (1997) Preimplantation genetic diagnosis: strategies and surprises. Trends Genet., 13, 270–275.[ISI][Medline]

Hardwick, K.G. (1998) The spindle checkpoint. Trends Genet., 14, 1–4.[ISI][Medline]

Johnson, R.T. and Rao, P.N. (1971) Nucleo-cytoplasmic interactions in the achievement of nuclear synchrony in DNA synthesis and mitosis in multinucleate cells. Biol. Rev., 46, 97–155.[ISI][Medline]

Kimelman, T., Kirschner, M. and Scherson, T. (1987) The events of the midblastula transition in Xenopus are regulated by changes in the cell cycle. Cell, 48, 399–407.[ISI][Medline]

LeMaire-Adkins, R., Radke, K. and Hunt, P.A. (1997) Lack of checkpoint control at the metaphase/anaphase transition: a mechanism of meiotic nondisjunction in mammalian females. J. Cell. Biol., 139, 1611–1619.[Abstract/Free Full Text]

Lin, L., Lee, C. and Moor, R.M. (1996) DNA synthesis, microtubule and nuclear dynamics in porcine parthenotes. Zygote, 4, 139–144.[ISI][Medline]

Odorisio, T., Rodriguez, T.A., Evans, E.P. et al. (1998) The meiotic checkpoint monitoring synapsis eliminates spermatocytes via p53-independent apoptosis. Nature Genet., 18, 257–261.[ISI][Medline]

Ouhibi, N., Fulka, Jr, J., Kanka, J. et al. (1994) A reversible block at the G1/S border during cell cycle progression of mouse embryos. Int. J. Dev. Biol., 38, 731–736.[ISI][Medline]

Page, A.W. and Orr-Weaver, T.L. (1997) Stopping and starting the meiotic cell cycle. Curr. Opin. Genet. Dev., 7, 23–31.[ISI][Medline]

Rieder, C.L., Khodjakov, A., Paliulis, L.V. et al. (1997) Mitosis in vertebrate somatic cells with two spindles: implication for the metaphase/anaphase transition checkpoint and cleavage. Proc. Nat. Acad. Sci. USA, 94, 5107–5112.[Abstract/Free Full Text]

Weinert, T. (1997) A DNA damage checkpoint meets the cell cycle engine. Science, 277, 1450–1451.[Free Full Text]

Submitted on December 7, 1998; accepted on March 3, 1999.