Suppression of chromosome condensation during meiotic maturation induces parthenogenetic development of mouse oocytes

Oocytes that have completed meiotic maturation become developmentally arrested until they are fertilized. This developmental arrest occurs at different stages in different animal species; in most vertebrates, it occurs at metaphase II (review: Masui, 1985). Metaphase Il-arrested oocytes can begin embryonic development not only after fertilization but also after parthenogenetic activation. In the mouse, a brief inhibition of protein synthesis can activate oocytes (Siracusa et al. 1978; Clarke & Masui, 1983). Second polar body emission and pronuclear formation occur during protein synthesis inhibition, and DNA replication and cleavage after its resumption. Thus, the developmental arrest of mouse oocytes appears to depend on the synthesis of short-lived proteins.

Proteins synthesized during maturation also regulate the behaviour of the oocyte chromosomes. When maturing oocytes at metaphase I are treated with puromycin, the cytoplasm becomes depleted of maturation-promoting factor (MPF) (Hashimoto & Kishimoto, 1988), and the chromosomes become decondensed within an interphase nucleus (Clarke & Masui, 1983, 1985). When the oocytes are allowed to resume protein synthesis, MPF reappears and the chromosomes become recondensed within 8 –10 h to a state resembling metaphase II. Thus, short-lived proteins synthesized during maturation are required not only to maintain developmental arrest at metaphase II, but also to maintain condensed chromosomes during maturation.

The results described above show further that oocytes at metaphase I, unlike those at metaphase II, are not parthenogenetically activated by puromycin treatment. This is indicated by the fact that after protein synthesis resumes they return to metaphase. whereas the metaphase II-treated oocytes remain in interphase of the first embryonic cell cycle. However, the oocytes treated at metaphase I with puromycin oocytes may also be maintained in interphase, even after protein synthesis resumes, by addition of dibutyryl cyclic AMP (dbcAMP) to the culture medium (Clarke & Masui, 1985). DbcAMP apparently acts post-translationally to prevent the production of the protein factors responsible for chromosome condensation, thus allowing the oocytes to remain in interphase. In the experiments reported here, we investigated whether these unactivated oocytes maintained in interphase under the influence of dbcAMP would begin embryonic development. To this end, we examined the ability of these oocytes to undergo DNA replication and cleavage.

3-to 10-week-old virgin female mice of the HS (McClearn et al. 1970) or CD-I (Charles River Canada) strain were given an injection of 5 i.u. pregnant mares’ serum gonadotropin (Sigma) and killed 44 –48 h later. Their ovaries were removed, placed in Hepes-buffered minimum essential medium (MEM) modified as described (Schroeder & Eppig, 1984) and supplemented with 50/xgmr¹ dbcAMP (Sigma), and oocytes were isolated by puncturing the ovarian follicles with forceps. The addition of dbcAMP prevented spontaneous germinal vesicle breakdown (GVBD) among the oocytes that were released from the punctured follicles (Cho et al. 1974). 100 –200 fully grown immature oocytes containing an intact germinal vesicle (GV) were collected from four mice in about 2h.

Metaphase I oocytes were obtained by incubating immature oocytes for 8 –9 h (Donahue, 1968) in dbcAMP-free MEM or Brinster’s medium (Brinster, 1965) in 5 % CO₂ in air at 37 °C, and selecting those that had undergone GVBD but had not emitted a polar body. Metaphase II oocytes were obtained by incubating the immature oocytes for 20 h (Donahue, 1968) and selecting those that had both undergone GVBD and emitted a polar body.

The metaphase I or metaphase II oocytes were transferred to modified Whitten’s medium (Whitten, 1971, NaCl reduced from 88 mm to 71 mm) and incubated in the presence of 10 μg ml ⁻¹ puromycin (Sigma) for 9h, or 50 μg m ⁻1 for 4 ·5 h, in 5 % CO₂ in air at 37 °C.

The puromycin-treated oocytes were cultured in puromycin-free modified Whitten’s medium that contained 5 μCi ml ⁻¹ methyl-[³H]thymidine (25Ci mmol ⁻¹, Amersham) in 5 % CO₂ in air at 37 °C. The labelling was carried out in the presence or absence of 100 μg ml ⁻¹ dbcAMP for periods varying between 2 and 30 h, as described in the results. Then the oocytes were fixed as described previously (Clarke & Masui, 1985). For DNA autoradiography, the fixed oocytes were mounted on glass slides, which were then washed twice in ice-chilled 5 % (w/v) trichloroacetic acid (British Drug Houses), rinsed thoroughly in cold distilled water and allowed to dry. The dry slides were coated with emulsion (Kodak NTB-2) previously warmed to 41 °C, stored for 2 days in light-proof boxes at –20 °C, and developed using D-19 (Kodak). Then the slides were stained using Giemsa as described previously (Clarke & Masui, 1985).

Metaphase I oocytes were exposed to puromycin (50 μg ml ⁻¹, 4 ·5 h), then incubated in puromycin-free modified Whitten’s medium in the presence of 100 μg ml ⁻¹ dbcAMP for 15 h. Those that retained a nucleus were transferred to the oviducts of day 0 ·5 (=day of plug) pseudopregnant foster mothers (Hogan et al. 1986). Control oocytes were treated in the same way, except that they were incubated in the absence of dbcAMP to allow them to return to metaphase, and then transferred to oviducts of recipients. Between 20 and 60 oocytes were transferred per oviduct. Two days later, the oviducts were flushed and the oocytes were examined. Those that had undergone cleavage were incubated in modified Whitten’s medium for 48 h. and then examined to determine the stage of development reached. Some were then incubated for 4h in the presence of 0 ·5 μg ml ⁻¹ colcemid, fixed and stained as described (Clarke & Masui, 1985), and examined to determine the ploidy.

The first experiments were designed to determine when DNA replication began in oocytes that had been matured to metaphase II in vitro and then exposed to 10 μg ml ⁻¹ puromycin for 9h (Siracusa et al. 1978; Clarke & Masui, 1983). In the first experiment, 47 of 79 oocytes (59 %) emitted the second polar body and developed a pronucleus containing decondensed chromosomes after the puromycin treatment, and 80 of 98 oocytes (82 %) did so when the experiment was repeated. The oocytes containing a pronucleus were transferred to puromycin-free medium that contained [³H]thymidine and incubated in the presence or absence of dbcAMP. Samples of oocytes were fixed at regular intervals following the withdrawal of puromycin and processed for autoradiography. DNA synthesis was considered to have occurred if the density of silver grains over the nucleus or chromosomes was obviously higher, as judged visually, than that over the cytoplasm (see Fig. 3A).

Fig. 1 shows that DNA synthesis began after the withdrawal of puromycin, regardless of the presence or absence of dbcAMP in the medium. About half of the activated oocytes began DNA synthesis by 6h after the withdrawal of puromycin. This 6h interval between the withdrawal of puromycin and the initiation of DNA synthesis is similar to the intervals between pronuclear formation and the initiation of DNA replication that have been reported previously for fertilized or parthenogenetically activated mouse eggs (Luthardt & Donahue, 1973; Graham & Deussen, 1974; Abramczuk & Sawicki, 1975; Krishna & Generoso, 1977; Howlett & Bolton, 1985). These results confirmed that oocytes matured in vitro and then treated with puromycin began DNA replication after puromycin withdrawal.

Next, we investigated whether oocytes that had been treated at metaphase 1 with puromycin would begin DNA synthesis after withdrawal of puromycin. Depending on the batch, 50 –90 % of the oocytes treated at metaphase I with puromycin (10 μg ml⁻¹, 9h) emitted a polar body and developed a nucleus (Fig. 2). Oocytes with a nucleus from each batch were divided into two groups, transferred to puromycin-free medium that contained [³H]thymidine and incubated either in the presence or in the absence of dbcAMP. The oocytes incubated in the absence of dbcAMP returned to metaphase by 8–10 h after puromycin withdrawal, as previously observed (Clarke & Masui, 1985), indicating that they had not been parthenogenetically activated by the puromycin treatment. In contrast, most of the oocytes incubated in the presence of dbcAMP remained in interphase for up to 30 h (Fig. 2). Some oocytes became fragmented during the prolonged incubation in the presence of dbcAMP. These results confirmed that incubation in the presence of dbcAMP maintained the oocytes in interphase after the resumption of protein synthesis.

To determine whether DNA synthesis had occurred in these oocytes, samples were fixed at regular intervals following the withdrawal of puromycin and processed for autoradiography. None of the oocytes had begun DNA synthesis by 10 h after puromycin treatment. This is not likely to be due to a toxic effect of the drugs, because 75 % of the metaphase 11 oocytes treated with puromycin in the same way as the metaphase I oocytes had begun DNA replication by this time (Fig. 1). Rather, it appears that metaphase I oocytes treated with puromycin were unable to initiate DNA synthesis during the first 10 h after puromycin withdrawal.

When these oocytes were incubated for longer than 10 h in dbcAMP, they began DNA synthesis (Figs 1 and 3A). By 20h after puromycin withdrawal, 50% had begun DNA synthesis, and 90% eventually did so. In contrast, none of the oocytes incubated in the absence of dbcAMP, which returned to metaphase, began DNA synthesis. As well, oocytes that had not been treated with puromycin, but were incubated in the presence of dbcAMP, remained at metaphase and did not incorporate [³H]thymidine (not shown). These results indicated that, even though the meta phase-1 oocytes had not been activated by puromycin treatment, they underwent DNA synthesis when allowed to resume protein synthesis under conditions that maintained them in interphase.

The fact that the oocytes maintained in interphase underwent DNA synthesis suggested that they had begun parthenogenetic embryonic development. To test this idea, oocytes were treated at metaphase I with puromycin (50 μg ml⁻¹, 4–5h) and incubated in the presence of dbcAMP for 15 h, and those that remained in interphase were transferred to the oviducts of foster mothers. When the oviducts were flushed 2 days later, 22 % of the transferred oocytes had reached the 4-to 8-cell stage of development (Table 1). The remaining oocytes were arrested at the 2-cell stage or were fragmented.

In contrast to these results, puromycin-treated oocytes that were allowed to return to metaphase, by incubation in the absence of dbcAMP, did not undergo cleavage after transfer to the oviduct. This observation rules out the possibilities that transfer into the oviduct had parthenogenetically activated the oocytes or that the embryos we observed had developed from parthenogenetically activated eggs of the host mother. Rather, these embryos developed from the puromycin-treated oocytes that were maintained in interphase.

The 4-to 8-cell embryos were then cultured for 48 h. During this time, 40% of them developed into blastocysts (Table 1, Fig. 3B). Karyotypic analysis of colcemid-treated blastocysts indicated that they were diploid (Fig. 3C), which was expected because they had developed from oocytes that had not undergone the second meiotic division. These results showed that oocytes treated at metaphase I with puromycin and then maintained in interphase under the influence of dbcAMP began parthenogenetic embryonic development.

The results presented here show that oocytes treated with puromycin at metaphase 1, and maintained in interphase by incubation in the presence of dbcAMP, could initiate DNA synthesis. There are two reasons for believing that the DNA synthesis that was observed was replicative, rather than repair synthesis. First, DNA synthesis was never observed in the puromycin-treated oocytes transferred to dbcAMP-free medium, which returned to metaphase, although repair synthesis can occur in metaphase oocytes of the mouse (Masui & Pedersen, 1975; Brazill & Masui, 1978). Second, in the experiments described here, DNA synthesis did not begin until 20 h after puromycin withdrawal, although repair synthesis can begin within 2h of exposure to DNA-damaging agents (Brazill, 1977). Therefore, it seems reasonable to conclude that the DNA synthesis observed here represents S phase of the first embryonic cell cycle.

DNA synthesis did not begin in these oocytes, however, until 20 h after puromycin withdrawal. This interval is considerably longer than the 6h interval between puromycin withdrawal and DNA synthesis that was observed using metaphase II oocytes. Thus, entry into interphase alone was insufficient to initiate DNA synthesis. It appears that DNA replication also requires conditions that are absent in puromycin-treated metaphase I oocytes, but appear after the oocytes resume protein synthesis. A number of stage specific changes in the pattern of oocyte protein synthesis occur during mouse oocyte maturation (Schultz & Wassarman, 1977; van Blerkom, 1981; Cascio & Wassarman, 1982; Richter & McGaughey, 1982; Howlett & Bolton, 1985). Although the precise functions of these newly synthesized proteins are largely unknown, they may provide factors that control specific biochemical events required for the initiation of development. In the frog, the ability to induce DNA replication develops during oocyte maturation (Gurdon, 1967; Benbow, 1985). By analogy, this ability may develop in maturing mouse oocytes after metaphase 1.

The metaphase 1 oocytes that were treated with puromycin and maintained in interphase in the presence of dbcAMP not only began DNA synthesis, but also could subsequently undergo several cleavage divisions, and some developed into blastocysts. Thus, the experimental treatments had caused the oocytes to begin embryonic development. It is clear, however, that the metaphase 1 oocytes had not been parthenogenetically activated by the puromycin treatment, because, when they were allowed to resume protein synthesis in the absence of dbcAMP, these oocytes returned to metaphase and neither synthesized DNA nor cleaved. As well, oocytes treated only with dbcAMP remained developmentally arrested at metaphase II. Even metaphase I oocytes treated with puromycin and then exposed to dbcAMP for 6h returned to metaphase after transfer to dbcAMP-free medium (Clarke & Masui, 1985). These results demonstrate that neither puromycin nor dbcAMP treatment could activate metaphase I oocytes. Rather, it is apparent that, when chromosome condensation was suppressed during maturation, the maturing oocytes could begin embryonic development in the absence of an activating stimulus. This suggests that the mechanism responsible for developmental arrest of mature metaphase II oocytes depends on the cytoplasmic condition that causes chromosome condensation to a metaphase state.

In the frog, the unactivated oocyte also is arrested at metaphase II. In this case, arrest is caused by a cytoplasmic activity, termed cytostatic factor (CSF), which develops in oocytes during maturation and disappears after activation (Masui & Markert, 1971; Masui, 1974; Meyerhof & Masui, 1977; Newport & Kirschner, 1984). Cytoplasmic transfer experiments have shown that CSF arrests the cell cycle at metaphase (review: Masui & Shibuya, 1987). It has been speculated that a cytoplasmic factor resembling CSF plays a role in the developmental arrest of unactivated oocytes in other species, including molluscs (Néant & Guerrier, 1988) and the mouse (Balakier & Czolowska, 1977). These observations taken together suggest that developmental arrest of unactivated mature oocytes of animals in general may depend on the presence of a cytoplasmic factor which causes cell cycle arrest, and that its removal leads to initiation of development.

In the case of the mouse, the metaphase II arrest must depend on the continuous synthesis of shortlived proteins, because brief inhibition of protein synthesis will parthenogenetically activate metaphase II oocytes (Siracusa et al. 1978; Clarke & Masui, 1983), as has also been observed in oocytes of certain marine invertebrates (Fautrez & Fautrez-Firlefyn, 1961; Zampetti-Bosseler et al. 1973). As it is also known that certain proteins of the mouse oocyte become phosphorylated only during metaphase (Howlett, 1986), it may be speculated that the developmental arrest at metaphase II of mouse oocytes is regulated by short-lived, metaphase-specific phosphoproteins. When oocytes treated at metaphase I with puromycin are maintained in interphase, by the presence of dbcAMP, these proteins may not appear in the cytoplasm or may fail to exert their inhibitory effect. As a result, oocytes having completed meiotic maturation may directly begin embryonic development in the absence of an activating stimulus.

This work was supported by grants from the Natural Sciences and Engineering Research Council and the Medical Research Council of Canada