The effect of the nucleocytoplasmic ratio on protein synthesis and expression of a stage-specific antigen in early cleaving mouse embryos

In many non-mammalian species, cleavage divisions are very fast, consisting mainly of the S- and M-phase only. During this period the egg’s cytoplasm is distributed to the resulting blastomeres until the nucleocytoplasmic ratio of somatic cells is reached. However, in early mouse development, the cell cycle time following the first cleavage division is about 10h (Bowman & McLaren, 1970), and G₁- and G₂- phases of varying lengths are observed (see Smith & Johnson, 1986 for review). Although transcription and translation start early in mouse development and increase considerably during cleavage (see Magnuson & Epstein, 1981 for review), there is no cell growth (Abramczuk & Sawicki, 1974; Lehtonen, 1980) and the protein content of embryos does not increase before the late blastocyst stage (Brinster, 1967; Schiffner & Spielmann, 1976). Therefore, the nucleo-cytoplasmic ratio gradually increases with every DNA replication step, reaching the somatic value at approximately the 120-cell stage (Surani, cited from Smith & Johnson, 1986).

In Xenopus laevis embryos it was shown that the achievement of a critical nucleocytoplasmic ratio or DNA/cytoplasmic ratio was required for the event of midblastula transition and the activation of the embryonic genome (Newport & Kirschner, 1982a,b). The direct influence of the nucleocytoplasmic ratio on stage-specific gene expression during early mouse development has not been clearly analysed due to several experimental factors. The classical constriction experiments (Spemann, 1928) applied by New-port & Kirschner (1982a,b) to the Xenopus egg, cannot be utilized due to the small size of the mouse egg. Gross alterations in the DNA/cytoplasmic ratio by the injection of DNA would require the injection of a large volume of liquid and might result in early developmental arrest (Wirak, Chalifour, Wassarman, Muller, Hassell & de Pamphilis, 1985).

Therefore, we decided to alter the nucleocytoplasmic ratio by two methods. In one experiment we removed cytoplasm by micropipette from one egg and injected it to another, an operation which had been successful in overcoming the ‘in vitro two-cell block’ in eggs of certain mouse strains (Muggleton-Harris, Whittingham & Wilson, 1982). The advantage of this technique is that only the amount of cytoplasm is altered without quantitatively changing the membrane. The second method we used was bisection of fertilized eggs (Tarkowski, 1977; Barton & Surani, 1983) leaving both pronuclei in one half of the egg or forming haploid eggs. This technique requires removal of cytoplasm as well as membrane components to achieve the required nucleocytoplasmic ratio.

The results presented in this paper compare protein synthesis and stage-specific surface antigen expression in experimental eggs that have had their nucleocytoplasmic ratio manipulated with control eggs at similar stages of development.

For the protein synthesis experiments, fertilized eggs were obtained from F₁ hybrid females (C57BL/6xCBA/H). They were superovulated with 7·5 i.u. of pregnant mare serum gonadotropin (PMSG: Organon) and 5 i.u. human chorionic gonadotropin (HCG: Organon) given 45 to 48 h apart, injecting the HCG usually at 14.00 h. The females were mated with Fj males (plug date = day 1 of gestation). For the immunofluorescence experiments C57BL/6-JHAN females were naturally mated with males of the same strain.

For nucleocytoplasmic ratio alteration by bisection fertilized eggs were collected around noon of day 1, freed from the cumulus cells by treatment with hyaluronidase (Sigma, 150i.u.ml^-1) and washed several times through Hepes-buffered mouse embryo culture medium M2 (Fulton & Whittingham, 1978). Zonae pellucidae were removed by treatment with pronase (Calbiochem, 5 mg ml^-1) in M2 with polyvinylpyrrolidone but without bovine serum albumin. The eggs were washed again several times in M2 and transferred into M16 (Whittingham, 1971) for recovery from the pronase treatment for at least 2h. Culture was performed in plastic dishes (Falcon or Sterilin) in droplets of medium 16 under paraffin oil (BDH or Fisher Scientific) at 37 °C in an atmosphere of 5% CO₂ in air or an atmosphere of 5 % CO₂, 5 % O₂ and 90 % N₂.

The method for cytoplasmic removal and injection followed that described in a previous publication (Muggleton-Harris et al. 1982). Batches of ten eggs were placed in M2 plus cytochalasin D at 0·5 μtgml^-1 for 15 min at 37 °C prior to the manipulation. The experimental eggs were manipulated approximately 23-26h post HCG. The experimental and control eggs were placed in small drops of M2 plus cytochalasin in a 60 mm plastic Petri dish with a glass coverslip insert. The drops were overlaid with paraffin oil. A Leitz Diavert inverted-phase microscope, and De Fonbrune micromanipulators with Agla syringes were used for all micromanipulations. Control eggs for the manipulative procedure were processed through M2 plus cytochalasin but were not manipulated. After manipulation the eggs were washed and transferred to M16 under oil for further development.

Attempts were made to remove and inject half of the egg cytoplasm but the survival rate of the operated eggs was extremely low. Therefore, approximately one quarter of the donor egg’s cytoplasm was removed and injected into the recipient egg. The amount of cytoplasm was not calculated in a precise fashion but the same pipette size allowed an approximation of the amount injected. Such eggs from which cytoplasm had been removed (increased nucleocytoplasmic ratio; Fig. 1, lane C) and eggs that had received cytoplasm (reduced nucleocytoplasmic ratio; Fig. 1, lane D) were capable of cleaving and developing to the blastocyst stage. However, there was an unexpected degree of developmental failure, especially in eggs from which cytoplasm had been removed. Thus the method of bisection was used for all subsequent experiments.

Bisection of fertilized eggs was performed in the afternoon of day 1. The methods of Tarkowski (1977) and Barton & Surani (1983) were modified in the following ways. Operations were performed in M2 over 1 % Agarose (dissolved in PBS-A and incubated with M2) at about 5 °C or at room temperature. Hand-pulled flame-polished micropipettes were used for deformation of the eggs into a cylindrical shape (Fig. 2A). Then the eggs were bisected by hand with a fine-drawn glass needle under a Wild dissecting microscope. The plane of dissection was chosen in such a way that either one pronucleus ended up in each half of the egg or one part of the egg contained both pronuclei and an enucleate cytoplasmic fragment was left (Fig. 1). Control eggs without a zona pellucida were kept with experimental eggs over agar for a period equivalent to the manipulation time. After the operation the bisected eggs were transferred to M16 and cultured individually in drops of M16 for up to 3 days. The eggs developed to the 2-cell stage by day 2, the 4-cell stage (or 4- to 8-cell stage) by day 3 and compacted at day 4 (Fig. 1). The majority of eggs containing one pronucleus retain a haploid genome during further development (Tarkowski & Rossant, 1976; Tarkowski, 1977). Though these haploid half embryos are not normal, their nucleocytoplasmic ratio is comparable to that of untreated cleaving controls.

For radioactive labelling, embryos were taken on the afternoon of day 2 and 3 and were transferred to M16 containing [³⁵S]methionine (New England Nuclear) at a final concentration of 0·5-l·0mCiml^-1 (specific activity ∼ 1000Cimmol^-1) for a period of ⁠. During the necessary handling procedure the zona-less embryos tended to aggregate and disaggregate. Some of the 4-cell embryos (controls as well as bisected embryos) started the next cleavage during the period of incorporation. Processing of embryos for biochemical analysis was done as described previously (Petzoldt & Hoppe, 1980).

Two-dimensional gel electrophoresis was performed according to O’Farrell (1975). For fluorography (Laskey & Mills, 1975) gels were treated with Amplify (Radiochemical Centre, Amersham) and the Kodak X-omat AR film was used for exposure. Two to fifteen embryos were applied per gel.

Half embryos and controls were analysed for the appearance of the stage-specific antigen SSEA-1 in the afternoon of day 3 and 4, using the monoclonal antibody α-SSEA-1 (Solter & Knowles, 1978; kindly provided by Dr D. Solter, Philadelphia). All procedures were performed as described elsewhere (Petzoldt, 1986) but using the M2 medium. The dilution of α-SSEA-1 was 1:100, that of the second antibody (rabbit anti-mouse IgM/FITC, Nordic) 1:50. Negative controls were carried out using the same protocol but replacing the first antibody by mouse ascites fluid containing IgM (diluted 1:100 in M2, kindly provided by Prof. Dr D. Haustein, Marburg and Dr E. Pfaff, Heidelberg). The embryos were examined for immunofluorescence with a Leitz Orthoplan immunofluorescence photomicroscope. Also during this procedure, denuded embryos tended to aggregate and disaggregate, and some 4-cell embryos started the next cleavage division.

Photographs of the bisected eggs were taken with a Leitz inverted microscope. From the plane of the pictures the volume of the fragments was determined using either the formula for spheres or for ellipsoids.

When the nucleocytoplasmic ratio of fertilized eggs was altered the amount of cytoplasm removed or injected was not quantified exactly. However, by experimentation it was found that approximately one quarter of the cytoplasmic content could be with-drawn from the donor and injected into the recipient with a reasonable survival and cleavage. At the 2-cell stage, both receiver and donor embryos expressed similar protein synthesis patterns to control embryos (gels not shown). Although the survival of the donors was quite low, the recipients were able to cleave further. They sometimes started compaction at day 3 and synthesized polypeptide sets equivalent to those of normal controls at that time and stage (Fig. 3). Although this method failed to provide sufficient experimental material of the required nucleocytoplasmic ratio, it did provide an additional control for the manual bisection method and supportive data that a change of the nucleocytoplasmic ratio, at the level achieved, does not influence stage-specific gene ex-pression during early cleavage.

In a major experiment, fertilized eggs were manually bisected leaving one pronucleus in each half of the cytoplasm or both in one half. Using this procedure eggs were not dissected into exact halves, especially when both pronuclei ended up in one cytoplasmic fragment. This fragment was significantly larger than the enucleated one (P<0·0001), when the volume of bisected eggs was measured using an equivalent sample of eggs from a series of experiments (Fig. 2C,D; Table 1). Although these calculations showed that half eggs with two pronuclei were slightly larger than those halves with one pro’nucleus (Fig. IE; Table 1), this difference was not found to be significant (P> 0·1). Half eggs with two pronuclei were regarded therefore as real half eggs. The nucleo-cytoplasmic ratio is double that of the control eggs and half eggs with one pronucleus only, but identical to the ratio in control 2-cell embryos. At the 2-cell stage these eggs have a nucleocytoplasmic ratio equivalent to a 4-cell embryo and at the 4-cell stage a ratio equivalent to that of an 8-cell embryo.

When bisected eggs were kept in culture the majority cleaved, reaching the 2-cell stage the next day and the 4-cell stage the following day, but they were smaller than control embryos (Fig. 2F,G). A considerable number compacted at day 4. There was no general delay in development observed following bisection (Tarkowski, 1977; Barton & Surani, 1983). In our experiments we found that operated eggs were delayed in comparison with the controls at the first cleavage division. This delay was also visible at the following cleavage stages and more apparent in haploid than in diploid bisected embryos.

2- and 4-cell embryos were incubated with [³⁵S]methionine and analysed for their protein syn- thesis by two-dimensional polyacrylamide gel electro- phoresis. The periods for radioactive labelling were chosen so that they included a common part of the cell cycle for controls and bisected embryos. Embryos of the 2-cell stage were at the G₂-phase at that time, the majority of 4-cell embryos passed from S- to G₂+M-phase (see Smith & Johnson, 1986 for the variations of cell cycles during first cleavage divisions in mouse embryos). Fig. 4 shows representative gels (each gel was selected from at least five gel runs for every stage analysed) for 2- and 4-cell stages. It is clearly visible that 2-cell control embryos and 2-cell bisected embryos with normal (haploid half eggs) as well as with double (diploid half eggs) nucleocyto-plasmic ratio express protein synthesis patterns that are closely related to each other. A set of proteins predominant in 2-cell stages is marked by arrows and a triangle. Equivalently, 4-cell controls and bisected embryos with normal and double nucleocytoplasmic ratio exhibit protein synthesis patterns that are similar to each other but definitely different from those of the 2-cell embryos. Here a variety of polypeptide spots more typical for 4-cell stages is indicated by arrowheads (Fig. 4). There are variations between different gels of the same stage but, evidently, 2-cell embryos with a double nucleocytoplasmic ratio, which is equivalent to that of a 4-cell control embryo, synthesize proteins typical for the 2-cell and not the 4-cell stage.

In additional experiments embryos of day 3 at the 4-cell or 4-to 8-cell stage and compacted embryos of day 4 were analysed for the expression of the stage-specific antigen SSEA-1. In control embryos and bisected diploid embryos with double nucleocytoplasmic ratio of day 4,100 % reacted with the monoclonal antibody against SSEA-1, but only 71% of the bisected haploid embryos with normal nucleocytoplasmic ratio of day 4 were positive (Table 2; Fig. 5). Almost no reaction was found in day-3 embryos (Fig. 5), independent of whether they were at the 4-cell or 4- to 8-cell stage. Only one bisected embryo with normal nucleocytoplasmic ratio showed a slight reaction with the antibody (Table 2). These results also show that the expression of a stage-specific antigen, which normally appears at the 6- to 8-cell stage (Solter & Knowles, 1978), is not accelerated in 4-cell embryos with double nucleocytoplasmic ratio.

Early development is characterized by a variety of morphogenetic processes and the stage-specific activation of the embryonic genome. The time schedule of all these events is regulated by different programmes including cytoplasmic clock mechanisms, counting of DNA replication cycles and the quantitative nucleocytoplasmic (or DNA/cytoplasmic) ratio (see Satoh, 1982, for review). Such different programmes might cooperate within the same embryos each controlling a different developmental event as was shown for cleavage synchrony, midblastula transition and onset of gastrulation in Xenopus laevis (reviewed by Kirschner, Newport & Gerhart, 1985).

In recent years, the action of different timing mechanisms has been analysed also in mouse embryos. The existence of cytoplasmic clocks was verified as being responsible for certain morphogenetic events, e.g. in enucleated eggs cortical activity was observed at the time nucleate eggs cleave (Waksmundska, Krysiak, Karasiewicz, Czolowska & Tarkowski, 1984). For compaction and blastocyst formation conflicting data are reported. Experiments of Smith & McLaren (1977) led to the conclusion that cavitation of blastocysts is controlled by the counting of replication cycles. By using aphidicolin, an inhibitor for DNA polymerase-α, other investigators demonstrated that compaction as well as blastocyst formation occur at the correct time, even when one or two preceding DNA replication steps are blocked (Alexandre, 1982; Dean & Rossant, 1984). Embryos should have reached at least the 4-cell stage before exposure to the drug (Smith & Johnson, 1985). These results show that these morphogenetic changes are independent of a defined number of DNA replications but triggered by an unknown event occurring sometime during cleavage but not evident at the 1-cell stage. Recent results from Prather & First (1986) postulate that such an internal programme for the process of cavitation can be overcome by cell-cell interaction. Chimaeras formed from a single 8-cell blastomere and a 2-cell embryo cavitate at a time intermediate between both programmes. Such a programme can be manipulated even further. Treatment of 2-cell embryos with wheat germ agglutinin induced premature compaction- and cavitation-like events, thus inducating that proteins and glycoproteins necessary for these processes might be present already at that stage (Johnson, 1986). Actually, this idea is supported by experiments to inhibit protein synthesis at the 2- to 4-cell stage. In such embryos an accelerated expression of some features occurred, which are typical for compaction (Kidder & McLachlin, 1985; Levy, Johnson, Goodall & Maro, 1986). None of these results really clarify which mechanisms are responsible for the timing of compaction. Different signals might cooperate in unknown combinations (Levy et al. 1986).

Stage-specific gene activity in early embryos is completely independent of cleavage with regard to transcription, translation, surface antigen expression and the activation of the paternal genome (Surani, Barton & Burling, 1980; Pratt, Chakraborty & Surani, 1981; Petzoldt, Bürki, Illmensee & Illmensee, 1983; Petzoldt, 1986). Also post-translational modifications of maternal mRNA-derived proteins, which are characteristic for meiosis and mitosis, proceed in cleavage-arrested eggs after cytochalasin D treatment but cease when embryos are arrested in metaphase (Howlett, 1986a).

The influence of a cytoplasmic clock on stage-specific gene activity cannot be excluded. Bolton, Oades & Johnson (1984) showed by short-term treatment of late 1-cell eggs or early 2-cell embryos with aphidicolin, that α-amanitin-sensitive late 2-cell proteins were synthesized without an S-phase occurring. When treating early fertilized eggs with aphidicolin, only a few 2-cell-specific proteins showed up the next day (Petzoldt, 1984a; Howlett, 1986b). When such eggs were incubated with the drug even for 2 or 3 days they did not progress further in stage-specific protein synthesis or in surface antigen expression (Petzoldt, 1984a,b). A toxic side effect of aphidicolin cannot be excluded, however, when used over such a long period. Therefore, the action of a cytoplasmic clock to regulate stage-specific gene expression during cleavage cannot be ruled out completely.

We have tried to differentiate between the timing mechanisms by either counting the cycles of DNA replication or measuring the nucleocytoplasmic ratio in cells during cleavage. For this purpose, fertilized eggs were either bisected or the amount of cytoplasm was manipulated using micropipettes. Both techniques disturb the cellular architecture of the egg and its quantitative (and qualitative?) amounts of organelles and cytoplasmic substances. In the case of removal and addition of cytoplasm the disturbance is restricted to cytoplasmic organelles and/or sub-stances. However, the bisection method also removes amounts of the cell membrane. All these components are of major importance for pronuclear movements and regular development after fertilization (see e.g. Schatten, Simerly & Schatten, 1985; van Blerkom & Bell, 1986). Nevertheless, orderly cleavage is neither grossly influenced after bisection (see also Tarkowski & Rossant, 1976; Tarkowski, 1977; Barton & Surani, 1983), nor is it after cytoplasmic transfer experiments, which served as additional controls.

During successive cleavage, such embryos were analysed for stage-specific protein synthesis and surface antigen expression. Protein synthesis patterns changed in bisected eggs in a manner equivalent to controls. These changes include shifts of proteins translated from maternal mRNA as well as new embryonic gene products, as shown in previous publications (Petzoldt, Illmensee, Biirki, Hoppe & Illmensee, 1981; Petzoldt et al. 1983). Even though <x- amanitin sensitivity of newly appearing polypeptides was not tested here, it is obvious by comparison of our protein synthesis patterns with those in the literature that these newly appearing spots are α- amanitin sensitive (see Flach, Johnson, Braude, Taylor & Bolton, 1982). The expression of the stage-specific embryonic antigen SSEA-1 was not accelerated and occurred at day 4, when bisected and control embryos compacted.

It was quite clear from our results that doubling of the nucleocytoplasmic ratio does not induce the embryo to express proteins and surface antigens one cleavage step earlier than embryos with normal nucleocytoplasmic ratio. Furthermore, decreasing the ratio by injecting cytoplasm into fertilized eggs did not visibly slow down stage-specific protein synthesis. This is in agreement with previous results, where the nucleocytoplasmic ratio was approximately halved by removing one pronucleus from fertilized eggs and restoring the normal ratio by diploidizing the egg with cytochalasin B (Petzoldt et al. 1981). Protein synthesis followed the number of DNA cycles of replication and not the nucleocytoplasmic ratio. On the basis of the results presented here, we favour the idea that the consecutive events of stage-specific gene activity during early mouse development are closely related to the progress of DNA replication as experiments with cytochalasins have indicated previously (Surani et al. 1980; Pratt et al. 1981; Petzoldt et al. 1983; Petzoldt, 1986). The molecular basis for this mechanism is not known. Satoh (1982) has proposed that fertilized eggs contain certain fully or uniformly methylated DNA sequences. During cleavage defined consecutive demethylation takes place at specific sites of the DNA, allowing the embryo the capability to count the replication steps and activate the genome in a stage-specific manner. However, Sanford, Rossant & Chapman (1985) have shown that the DNA in mouse oocytes and preimplantation embryos is undermethylated and de novo methylation is not found until implantation. Future work in this area will help elucidate the role of methylation in early mammalian development.

We acknowledge the help and discussions of Prof. I. Hansmann, Inst, für Humangenetik, Gottingen and Dr A. H. Handyside, MRC, Carshalton. We thank Dr D. G. Whittingham for providing facilities and encouragement during a working visit to the MRC EETUnit Carshalton by U.P., Ms J. GieB and Mrs S. Petzoldt for excellent technical assistance. The work of U.P. was further supported by the Deutsche Forschungsgemeinschaft with a Heisenberg-Sti-pendium (Pe 166/8-1,2) and the grants (Pe 166/9-1,2,3).