Genetically identical parthenogenetic mouse embryos produced by inhibition of the first meiotic cleavage with cytochalasin D

Cytochalasins are potent inhibitors of actin microfilaments and have been used to increase the ploidy of various cells. The presence of cytochalasin during cell division does not disturb the anaphase movement of chromosomes (karyokinesis), but inhibits the separation of daughter cells (cytokinesis), both sets of chromosomes remaining in a single cell, doubling its ploidy. Cytochalasin B has been used for the diploidisation of mouse oocytes activated parthenogenetically in vitro by blocking second polar body extrusion (Balakier and Tarkowski, 1976). It can also be used to produce tetraploid embryos (Snow, 1973), or even higher ploidy depending of number of cytokineses blocked (Petzoldt et al. 1983). During all these cell divisions, as for normal somatic cell mitosis, the cells undergo transition to interphase and form two interphase nuclei, which merge during the subsequent metaphase. Hashimoto and Kishimoto (1988) have applied cytochalasin D (CCD) during the first meiotic division of the mouse oocyte (i.e. at the transition between the first and second meiotic metaphase, MI and Mil respectively, which proceeds normally without formation of any interphase nucleus (Donahue, 1968)) and suggested that this drug blocks not only the first polar body extrusion, but also stabilizes maturation promoting factor (MPF) activity at high, metaphase levels. They also demonstrated that during the normal MI-MII transition the activity of MPF drops to the low level characteristic for interphase. Instead of 20 bivalents characteristic for MI, they found a single metaphase spindle with 40 monovalent chromosomes in cytochalasin D-treated oocytes after the time of the transition, and concluded that cytochalasin D allows splitting of the bivalents, but prevents anaphase I movement of monovalents during the first meiotic division. Since this cell division would be the only example of cytochalasin D-arrested karyokinesis, we have reexamined this phenomenon using immunofluorescence microscopy of tubulin and staining of chromatin. In this paper, we show that after cytochalasin D treatment oocytes are able to resume the normal course of meiosis and have the same developmental potential as classical diploid parthenogenotes.

Mouse oocytes were recovered from the ovaries of F₁(C57Bl/ DBA2), Fj(C57Bl/CBA) or Swiss females 7–8 h after hCG injection (i.e. 2-3 h before the first meiotic division). The experimental protocol is presented in Fig. 1. The experimental group of oocytes (CCD +−) was cultured in M2 medium containing 1 μg ml⁻¹ cytochalasin D (CCD), while the control group (CCD −+) was cultured in M2 medium alone. After 3–4 h, i.e. when the oocytes in the control group (CCD−+) had completed first polar body extrusion, the drug was washed out from the medium in which the experimental eggs were being incubated and the oocytes were further cultured in M2. Both groups were activated by treatment with 8% ethanol in M2 for 6 min between 16 and 20 h post-hCG (Cuthberson, 1983; Kaufman, 1983). The experimental group (CCD+−) was allowed to undergo normal extrusion of a polar body in M2, while the control group (CCD−+) was cultured for 3–4 h in the presence of 1 μg ml⁻¹ cytochalasin D in M2 to prevent extrusion of the second polar body. Activated oocytes were transplanted into the oviducts of randomly chosen females for 3 days or cultured overnight in vitro and transplanted after the first cleavage division into the oviducts of pseudopregnant females during the first day post coitum. The foster mothers were killed on the 9th or 10th day of gestation, and their uteri were examined for presence and morphology of fetuses.

The zonae pellucidae were removed by brief treatment with acid Tyrode’s solution, pH2.5 (Nicolson et al. 1975). Oocytes were placed in specially designed stainless steel chambers as described by Maro et al. (1984) except that the chambers were coated with 0.1 mg ml⁻¹ concanavalin A (Sigma). The samples were centrifuged at 450g for 8–10 min at 37°C, and after centrifugation fixed with 0.1% glutaraldehyde (Sigma) in PBS supplemented with 1 % Triton X-100 (Boehringer Mannheim GmbH) as described by de Pennart et al. (1988). After a 5 min wash in PBS, they were extracted with 2% Triton X-100 for 30 min, incubated in 10 mg ml⁻¹ NaBH₄ in PBS (three incubations of 10 min each), washed twice in PBS and processed for immunofluorescence as described by Maro et al. (1984).

We used rat YL1/2 antibody specific for tyrosinated alphatubulin (Kilmartin et al. 1982). Fluorescein-labelled anti-rat antibody (Miles) was used as a second layer. For visualizing chromatin, DAPI (5 μg ml⁻¹ in PBS) was included in the last wash of the specimens.

Histone H1 kinase activity was determined as described by Felix et al. (1989) in HK buffer (80 mM β-glycerophosphate, 20mM EGTApH7.3, 15 mM MgCh, 1 mM DTT, 1 mM PMSF, 10 μg ml⁻¹ leupeptin, 10 μg ml⁻¹ pepstatin, 10 μg ml⁻¹ aprotinin) using exogeneous histone H1 (HIII-S from calf thymus, Sigma) as substrate. Samples each containing 50 oocytes in 5 of water were lysed by freezing and thawing three times, diluted twice in two times concentrated HK buffer (2×HK) and incubated for 15 min at 20°C in the presence of 3.3mgml^-1 histone Hl, ImM ATP and 0.25 mCi ml⁻¹ [³²P]ATP. The reaction was stopped by the addition of a similar volume of two times concentrated sample buffer (Laemmli, 1970) and incubation for 2 min at 90°C. The samples were then electrophoresed on a 15 % SDS-polyacrylamide gel (Laemmli, 1970). In order to test the specificity of the reaction, the p34^cdc2 kinase (histone Hl kinase) was removed by centrifugation from the control sample using pl3^sucl-coated sepharose beads (gift of M. Dorée).

The assay was performed according to the method described by McLaren and Buehr (1981).

In the presence of cytochalasin D, the first meiotic division takes place, but cleavage is impaired and the first polar body is not extruded (Fig. 2A,B). Two separate spindles start to form around each group of chromosomes (Fig. 2C,D). The midbody microtubules, which separate the two groups of chromosomes, disassemble soon after anaphase (Fig. 2C), and both sets of chromosomes surrounded by microtubules merge to form a single metaphase plate (Fig. 2D,E). During the merging of the two spindles, the chromosomes are dispersed within the spindle volume indicating that spindle formation is not yet completed (Fig. 2D). Later on, the chromosomes form a well-organized metaphase plate within a compact, well-formed spindle between 1–2 h after anaphase I (Fig. 2E,F). This spindle is equivalent to the second meiotic spindle (Mil); however, it contains a tetraploid number of chromosomes instead of a diploid one.

The histone Hl kinase activity reflects the level of the maturation promoting factor (MPF) activity and is due to the p34^cdc2 protein, homolog of the product of the gene cdc2 in Schizosaccharomyces pombe (Draetta and Beach, 1988; Dunphy et al. 1988; Gautier et al. 1988; Labbe et al. 1988, 1989). We measured the activity of histone Hl kinase in the lysate of F₁(CBA/CS7Bl) oocytes before (MI), during (anaphase I), and after (Mil) the transition in the presence of cytochalasin D. Since, in the presence of the drug, the stage of karyokinesis is not visible in living oocytes due to the absence of polar body extrusion, we used 500 ng ml⁻¹ Hoechst 33342 to visualize chromosomes under a fluorescent microscope. Using this technique, we selected oocytes undergoing anaphase I and either lysed them immediately or allowed them to complete karyokinesis, culturing for a further hour. The MI oocytes were arrested in the M-phase by 5 μM nocod-azole added to the cytochalasin D-containing medium. The in vitro measurement of histone kinase activity in the lysates of these three groups of oocytes revealed a significant drop of this activity during the transitional period despite the presence of cytochalasin D (Fig. 3). The specificity of the reaction was tested in a control experiment where a lysate of Mil oocytes was incubated with sepharose beads coated with pl3^sucl, a protein that binds specifically to p34^cdc2 (Draetta et al. 1987). This allowed us to remove p34^cdc2 from the lysate. After centrifugation the supernatant was used in the assay and served as a reference point to estimate the histone Hl activity in the experimental groups (Fig. 3). During the normal transition between MI and Mil, using polar body extrusion as a reference point, we observed a more dramatic drop in histone kinase activity, characteristic for interphase (Kubiak, Weber and Maro, data not shown). The difference observed between the levels of Hl kinase activity during the MI–MII transition in the absence and the presence of cytochalasin D is caused by a poorer synchrony of the cytochalasin D-treated oocytes, since the time interval necessary for handling oocytes under a fluorescent microscope is much longer and in consequence some of the oocytes in the sample of anaphase oocytes could already have entered early MH.

Oocytes that had the first meiotic cleavage suppressed by cytochalasin D were able to respond to an activating stimulus a few hours later (Fig. 4). In most cases, they extruded a polar body and formed a single pronucleus. Others underwent immediate cleavage, or abortive activation (extrusion of a polar body and reentry in a subsequent metaphase; (Kubiak, 1989)). The relative percentages of oocytes reacting in these ways depended on the age of the oocytes at the time of activation. A high proportion of oocytes activated soon after the MI–MII transition (17 h post-hCG) reacted by abortive activation, while later activation (21 h post-hCG) resulted in a normal response, i.e. pronuclear formation (either one or two pronuclei), as well as the appearance of some immediately cleaving oocytes (Table 1). These results show that the oocytes reach a stage physiologically equivalent to Mil, despite the presence of cytochalasin D during the first meiotic division.

Both the experimental (CCD + −) and control (CCD −+) groups of embryos obtained from Fi(C57Bl/DBA2) females heterozygous at the GPI locus were either cultured in vitro or transplanted into oviducts of random-bred females for three days to test their viability. They underwent similar development up to the blastocyst stage in both conditions. Since we were able to obtain a greater number of well-developed blastocysts using the in vivo technique, we used these embryos to test their GPI constitution. We made an electrophoretic analysis of single expanded blastocysts on their fifth day of development. All successfully tested blastocysts (n=32) from the experimental group (CCD + −) show three bands corresponding to the three isoforms of the enzyme (A/A, A/B and B/B) characteristic for heterozygotes. 83% of blastocysts in the control group (CCD−+) (n=35) were homozygous for this locus. The heterozygotes A/B (17%) represent blastocysts developed from oocytes in which the GPI locus underwent recombination (Table 2). Haploid blastocysts (n=20) of the same strain and age obtained by activation of the control oocytes without cytochalasin D show only one form (A or B) of the enzyme indicating disappearance of the maternal isozymes at the specified period of development.

Activated (CCD+−) oocytes which extruded the polar body and formed a single pronucleus (potentially diploid heterozygotes), as well as control oocytes (CCD−+), which underwent the second meiotic division in the presence of cytochalasin D, were transplanted to the oviducts of pseudopregnant females. The implantation rate was around 70% in both groups, a percentage similar to the one observed with normal control fertilized embryos. The foster mothers were killed on the 9th or 10th day of gestation. Most of the parthenogenetic fetuses in both groups had degenerated by day 9 (Table 3). The surviving embryos from both groups recovered at this age were poorly developed and resembled egg-cylinder-stage embryos of day 7 of normal development except for one well developed fetus with 12 pairs of somites and a beating heart in the experimental group (CCD+−) on day 10.

Previous data published by Hashimoto and Kischimoto (1988) suggested that, in contrast to other cell divisions, the first meiotic division is arrested by cytochalasin D at both the karyokinesis and cytokinesis levels. As the first meiotic division proceeds, a particular modification of the cell cycle takes place, namely one metaphase (MI) is directly followed by another metaphase (MII). The period between these two metaphases represents, in fact, a very short interphase, short enough to prevent interphase nucleus formation. Since we know of no other data showing an interference with anaphase movements by cytochalasin D, we were intrigued by the reported exceptional action of this drug on this particular transition. Our results demonstrate that cytochalasin D acts similarly during the first meiotic division of mouse oocytes, during the second meiotic division and during mitosis. However, the result of cytochalasin D presence during the first meiotic division i.e. formation of a single metaphase plate, was to some extent unexpected, and was probably the reason for the previous misinterpretation of the action of this drug (Hashimoto and Kishimoto, 1988). The merging of both groups of chromosomes is possible due to the rapid disassembling of midbody microtubules, which takes place before definitive formation of the spindles. In the case of the second meiotic division, as well as during mitotic divisions, the transition to interphase takes place, and midbody microtubules persist for a prolonged time (Kidder et al. 1988; Kubiak, Weber and Maro, data not shown). A similar sequence of events occurs in cells treated with cytochalasin D. When the first meiotic division of mouse oocytes takes place in the presence of cytochalasin D and 20 μM puromycin (an inhibitor of protein synthesis causing transition of the oocytes to interphase), two nuclei separated by midbody microtubules formed in all oocytes treated with both drugs. This demonstrates that midbody microtubule behaviour does not depend on the presence or absence of cytochalasin D, but on cell-cycle-dependent modifications of cytoplasmic conditions.

The activity of histone Hl kinase drops during the MI–MII transition in the presence of cytochalasin D as it does in the normal MI–MII transition (Kubiak, Weber and Maro, data not shown). This represents a change in the activity of the p34^cdc2 protein kinase, which is the active part of MPF (Draetta and Beach, 1988; Dunphy et al. 1988; Gautier et al. 1988; Labbe et al. 1988, 1989). Thus, the drop in this activity reflects the inactivation of MPF during the transitional period in the oocytes despite the continuous presence of cytochalasin D. These results remain in conflict with those of Hashimoto and Kischimoto (1988) that suggested that cytochalasin D stabilizes the MPF activity in a similar manner to colcemid, but this could be explained by the difficulty of getting well-synchronized samples in the presence of cytochalasin D.

The normal course of meiosis with respect to the physiological properties of oocytes is also demonstrated by their ability to undergo activation even after the inhibition of first polar body extrusion by cytochalasin D. Oocytes in MI can not be activated (Clarke and Masui, 1983; Iwamatsu and Chang, 1972) except in the case of the LT/Sv strain (Kaufman and Howlett, 1986), and this ability develops at the beginning of the Mil arrest (Kubiak, 1989). Since we observed a similar sequence of events in cytochalasin D-treated oocytes after washing out the drug, it demonstrates that they are indeed at a stage physiologically equivalent to MII

The experimental protocol presented in this paper allowed us to obtain tetraploid oocytes, which could be activated to produce parthenogenetic embryos identical to each other and to their mothers – a true genetic clone. These embryos retain such properties since their genomes contain a single copy of each chromatid of every maternal monovalent chromosome. This means that, even in the case of recombined genes after crossing-over, the resulting embryonic genome represents a clone of the mother, only with the genes differently arranged on the sister chromatids. This effect could not be obtained using other methods of producing parthenogenetic embryos.

The cytochalasin D arrest of second polar body extrusion results in the formation of diploid embryos that have various gene compositions, since only a half of the original maternal chromosomes enter their genome. In addition, the majority of embryonic genes are in a homozygous state, with the exception of those that underwent crossing over. In contrast, the inhibition of first polar body extrusion by cytochalasin D provides an experimental method of creating diploid heterozygous, parthenogenetic embryos (with a genome similar to the genome of the mother).

We would like to thank Jacques Jami for helpful discussions and material support, Marcel Dorée for the gift of pl3^sucl-coated sepharose beads, John Kilmartin for the gift of the monoclonal anti-tubulin antibody, P. Debey for gift of Hoechst 33342 and discussions on its use, Kevin Sadler for critical reading of the manuscript and Richard Schwarztmann for his expert photographic work. This work was supported by grants from the Fondation pour la Recherche Médicale and from the Association pour la Recherche sur le Cancer to BM. JK is in receipt of a fellowship from the Fondation pour la Recherche Médicale and AP is in receipt of a fellowship from the Institut National pour la Santé et la Recherche Médicale.