Cytogenetic study of silver-staining NOR in 8-cell-stage mouse blastomeres fused to 1-cell-stage embryos

There are many approaches that have been used to study nuclear–cytoplasmic interactions during early mammalian embryogenesis. One successful approach has been the analysis of the behaviour of blastomeres or somatic nuclei that have been transferred into the cytoplasm of activated and nonactivated mouse oocytes (Tarkowski & Balakier, 1980; Czolowska et al. 1984; and for review see Tarkowski, 1982), and into oocytes from other species (sheep: Willadsen, 1986; cow: Prather et al. 1987). These experiments have clearly demonstrated that the cytoplasm of 1-cell-stage embryos contains certain ‘factors’ that are capable of regulating the behaviour and activity of the introduced nuclei.

An additional approach, which is now widely used, involves the transplantation of male or female pronuclei or of embryonic nuclei isolated from more-advanced-stage cleaving mouse embryos into enucleated 1-cell-stage mouse embryos (McGrath & Solter, 1983, 1984a, 1984b, 1986; Surani et al. 1984, 1986; Robl et al. 1986; Howlett et al. 1987; Tsunoda et al. 1987). In these studies, the authors demonstrated that the 8-cell embryonic nuclei that had been transplanted into the enucleated 1-cell activated embryos were not capable of supporting normal preimplantation embryonic development (McGrath & Solter, 1984b, 1986; Howlett et al. 1987). However, transcriptionally active 8-cell embryonic nuclei, when transferred to the cytoplasm of 1-cell embryos, were still capable of expressing heat-shock proteins (hsp 68/70 × 10³ M_r) that are characteristic of the first embryonic gene activity, and which normally only occur after the first cleavage division (Howlett et al. 1987). This observation strongly suggests that the cytoplasm of the 1-cell embryo is capable of modifying gene expression in the nuclei of 8-cell-stage embryos. Because of the complex nature of nuclear–cytoplasmic reactions in such reconstructed embryos, further studies of this model system are clearly necessary, especially with respect to the genetic activity of the introduced nuclei from more-advanced-stage embryos. In our study, we have used another criterion to assess the genetic activity of nuclei isolated from 8-cell- and 16-cell-stage embryos, which had been transplanted into 1-cell-stage eggs/embryos. For this purpose, a silver-staining technique was used as a quick method of visualizing transcriptionally active rRNA genes (NOR) (for reviews, see Howell, 1982; Hubbell, 1985). It was shown previously that during mouse embryonic development silver-staining NOR can be detected from the 2-cell stage onwards but was not observed in the chromosomes of 1-cell-stage embryos (Engel et al. 1977; Hansmann et al. 1978; Patkin & Sorokin, 1983). Our goal was to investigate, using this approach, whether 1-cell-stage cytoplasm could influence the positive NOR silver staining in the chromosomes from 8- to 16-cell-stage embryos. To achieve this aim, 8- to 16-cell blastomeres were fused to fertilized, unfertilized or parthenogenetically activated 1-cell-stage mouse eggs/embryos.

8- to 12-week-old (C57BL×CBA)F¹hybrid female mice were injected with 5i.u. of pregnant mares’ serum gonadotrophin (PMSG) followed 48 h later by 5i.u. of human chorionic gonadotrophin (HCG) to induce superovulation. Shortly after the HCG injection, the females were caged with (C57BL×CBA) F¹ hybrid males. The presence of a vaginal plug early the next morning was taken as evidence of mating, and this was designated the first day of pregnancy. The female mice that had successfully mated were killed by cervical dislocation shortly before midday (at about 20 h after the injection of HCG), and their eggs were isolated into M16 tissue-culture medium (Whittingham, 1971) containing 1 mg ml^-1 hyaluronidase (Bovine testes, Sigma) to remove adherent cumulus cells and facilitate the visualization of pronuclei. The eggs were then transferred into Petri dishes containing microdrops of M16 culture medium and retained in this medium at 37 °C in an atmosphere of 5 % CO² in air until they were required for micromanipulation and fusion (see E below).

Additional F¹ hybrid mice to those indicated in A above were injected with 5i.u. PMSG followed 48 h later by 5i.u. HCG to induce superovulation. Unfertilized eggs were isolated shortly before midday (at about 20 h after the HCG injection), exposed briefly to medium containing hyaluronidase, and subsequently washed in enzyme-free medium, then transferred into M16 tissue culture medium (as described above) until they were required for micromanipulation and fusion. A proportion of these eggs activate parthenogenetically following exposure to medium containing hyaluronidase (see C below). Only the nonactivated eggs were selected for subsequent micromanipulation.

Unfertilized eggs were isolated from superovulated F¹ hybrid females at 20 h after the HCG injection. The eggs were activated in vitro following a 15 min exposure to M16 medium containing hyaluronidase (Kaufman, 1983). The eggs were then washed in hyaluronidase-free medium and retained in M16 medium for a further 4–6 h. Only activated eggs in which an obvious second polar body was seen were selected for subsequent micromanipulation. Eggs of this type invariably develop a single haploid pronucleus if retained in culture.

8-cell/morula embryos were isolated from F¹ hybrid females during the morning on the 3rd day of pregnancy. In this group, naturally cycling (i.e. spontaneously ovulating) and superovulated females were mated with F¹ hybrid males. The embryos were initially isolated in M16 culture medium, then exposed to M2 culture medium (Whittingham & Wales, 1969) containing 0·5 % pronase (Calbio-chem–Behring) to remove the zona pellucida. Following its removal, the zona-denuded embryos were initially thoroughly washed in M2 medium, then transferred into Ca²⁺-free M16 medium for 15 min. During the latter period, the blastomeres were mechanically dissociated through a glass pipette with an inner diameter of about 20 μm. Once dissociated, the individual blastomeres were each stored in a separate microdrop of M16 medium until they were required for micromanipulation and fusion (see E, below).

Recipient 1-cell (fertilized, unfertilized or parthenogenetically activated) eggs were incubated in M16 medium containing l μg ml^-1 cytochalasin D (Sigma) and 0·1μg ml^-1 nocodazole (Sigma) for 15 min (Howlett et al. 1987). Individual eggs were then transferred into a glassslide chamber containing the same concentrations of cytochalasin D and nocodazole made up in M2 medium. A single dissociated 8-cell/morula blastomere was then introduced into the perivitelline space of the 1-cell-stage fertilized, unfertilized or parthenogenetically activated recipient egg in the presence of a small volume of inactivated Sendai virus, using the micromanipulation technique described in detail by McGrath & Solter (1983). In the majority of cases, it was possible to observe within about 1 h that successful fusion had occurred.

The fused embryos were retained in culture (in most instances in the presence of Colcemid, 0·035 mg ml^-1) for a further period of up to 5–10h, or until the pronuclear/nuclear membranes were observed to have disappeared. Airdried chromosome spreads were then made according to the method described by Dyban (1983) with slight modifications (the hypotonic solution used consisted of a 1:3 v/v mixture of 0·93% sodium citrate: 0·56% KC1, duration of treatment in the hypotonic solution was for not longer than 5 min at room temperature, then the embryos were transferred into cold fixative at 4°C for a further 5–10 min and then chromosome preparations made). Selective silver staining for detecting active nucleolar-organizing regions (NOR) in metaphase chromosomes was then carried out according to the method of Howell & Black (1980) modified by Patkin & Sorokin (1983).

In the unfertilized eggs, the second meiotic metaphase chromosomes were always present. Because the chromosomal preparations were made from postovulatory aged ova (28–30h after HCG, i.e. about 16–18h after ovulation), the metaphase chromosomes were invariably contracted and formed into a tight group, so that the exact number of chromosomes present could not be ascertained. Nevertheless, it was still possible to observe that there was no evidence of positive NOR silver staining (see Table 1, group 4).

In all of the 1-cell-stage fertilized and parthenogenetically activated embryos that were examined at metaphase of the first cleavage mitosis, no evidence of silver-staining NOR regions was observed (see Table 1, groups 2 and 3). The centromeric regions, however, stained dark brown, whereas the rest of the chromosomes stained light brown in colour. The staining pattern therefore closely resembled standard C-banding (Fig. 1). No difference was observed in the silver nitrate staining of the fertilized (diploid) or parthenogenetically activated (haploid) chromosomes.

In our experiments, blastomeres were isolated from 8-cell/morulae obtained following superovulation as well as after natural matings. These two sources provided a heterogeneous population of embryos, some of which were at the early morula stage and had only just entered the fourth cleavage division (with 8 blastomeres present, some of which were in mitosis). Yet other morulae, with 16- or rarely more cells present, some of which were in mitosis, had already entered the fifth cleavage division.

In all 37 instances, when control embryos were stained with silver nitrate, the characteristic NOR-positive regions (a pair of black dots close to the centromere region) were observed in the metaphase spreads. The number of NOR per metaphase plate varied. There were never less than four NOR present (Fig. 2), and never more than eight present (Fig. 3) (see Table 1, group 1). No metaphase preparations were observed, even those that were particularly contracted following prolonged culture (for more than 4h) in medium containing Colcemid, in which NOR silver-staining regions were not clearly observed (Fig. 4).

The morula nuclei that were in interphase or very early prophase also had a characteristic appearance following silver nitrate staining. The nucleoli characteristically contained fine reticular material which stained black, and contrasted with the rest of the nucleolar material.

In all instances, when attempts were made to fuse isolated blastomeres with unfertilized eggs, fusion was successfully achieved (Table 2, line 3). However, when the fusion products were cultured overnight in medium containing Colcemid, none of the introduced blastomere nuclei entered mitosis, and most were found to be pyknotic. In a few cases, the nuclear contents were dispersed into clumps of chromatin-like material, while in other instances the remnants of nucleoli were seen. By contrast, in all cases, the oocyte’s chromosomes remained at metaphase of the second meiotic division and only very minor abnormalities of chromosome morphology were observed (severe contraction of chromosomes, clumping into compact groups, etc). In no instances did the introduction of a blastomere nucleus, or the fusion process per se, induce the parthenogenetic activation of the egg. The cytoplasm of the unfertilized egg did not appear to be capable of stimulating the blastomere nucleus to enter into mitosis.

In this series of experiments, all the fertilized eggs had obvious pronuclei present and were obtained at about 20 h after the HCG injection to induce superovulation. In all cases, fusion was successfully achieved, but not all of the hybrid embryos entered into mitosis. Indeed, 16 out of a total of 48 embryos in this group were still in interphase following an overnight period in culture (Table 2, line 1). These eggs were examined at not less than 40 h after the HCG injection, at possibly 8–10h after normal fertilized embryos might be expected to enter into the first cleavage mitosis. This finding, however, may have been a direct consequence of the experimental procedure employed.

In all of the remaining (32) eggs, metaphase chromosomes from both pronuclei as well as from the blastomere nucleus were invariably seen as three distinct metaphase plates – a diploid plate derived from the blastomere nucleus and two haploid plates of male and female pronuclear origin (Fig. 5A–D).

When individual blastomeres from 8- to 16-cell embryos were fused with pronuclear-stage haploid parthenogenones, successful fusion was always achieved (Table 2, line 2). However, not all of the hybrid embryos entered into the first cleavage mitosis and, in all of these cases, neither the single haploid pronucleus nor the blastomere nucleus entered into mitosis.

In 12 of the hybrid embryos, while the single haploid pronucleus entered into mitosis, the blastomere nucleus remained in interphase, despite the fact that it was located in close proximity to the haploid metaphase plate (Fig. 6). These blastomere nuclei contained silver-nitrate-positive-staining reticular material in their nucleoli (Fig. 6). In other cases, the blastomere nuclei were obviously swollen and their nucleoli failed to show this characteristic reticular material following staining with silver nitrate. In those instances where the pronuclei were swollen but failed to enter mitosis, the majority of the blastomere nuclei and nucleoli showed a similar reticular-free silver-staining pattern (Fig. 7). In all, a total of 19 hybrid embryos out of 44 displayed a haploid and diploid metaphase plate, of parthenogenetic and blastomere origin, respectively (Fig. 8).

No difference was observed in the silver nitrate staining of haploid metaphase chromosomes in the hybrid eggs following fertilization or parthenogenetic activiation. The introduced blastomere nucleus was therefore unable to influence them to display NOR silver staining (a) when the blastomere nucleus remained in interphase and (b) when the latter was in mitosis.

In 4 out of 29 blastomere interphase nuclei studied, no change was observed in the characteristic NOR-positive staining reaction (which was typical of control 8-cell/blastomere nuclei) in these hybrid embryos. More commonly, the interphase or early prophase blastomere nuclei were swollen in appearance and clearly resembled pronuclei, and their nucleoli lacked the silver-nitrate-positive NOR reticular staining component. In some instances, remnants of the silver staining material were present in the nucleoli (Fig. 9). Similar changes were observed in blastomere interphase nuclei which were located in close proximity to haploid metaphase plates of parthenogenetic origin (Fig. 10).

In nearly all instances, when blastomere nuclei entered into mitosis, their chromosomes did not display silver-nitrate-staining NOR (Fig. 11). Rarely (2 out of 49 hybrids studied), the blastomere-derived chromosomes in these hybrids displayed silver-nitrate-staining NOR.

All of the haploid metaphase plates of pronuclear origin, derived from both fertilized and parthenogenetic embryos, had a normal chromosome constitution, with no obvious structural abnormalities present. By contrast, the blastomere chromosomes in four hybrids displayed morphological abnormalities. In one hybrid embryo, premature separation of the chromatids had clearly occurred (Fig. 12), while in two others most of the chromosomes present displayed chromatid breakages and evidence of exchanges between chromatids. In another case, all of the blastomere chromosomes appeared to be ‘pulverized’ (Fig. 13).

Our findings have demonstrated that a difference exists in the ability of activated compared to nonactivated mouse eggs to influence the 8- to 16-cell morula nucleus introduced into it by fusion. When blastomeres were fused with unfertilized eggs, their nuclei were unable to enter into mitosis. Our observations did not confirm the suggestion of Hansmann et al. (1978) that after the ageing of unfertilized mouse eggs the positive-silver-nitrate-staining regions can be seen in meiotic metaphase chromosome preparations. Moreover, the oocyte’s cytoplasm appeared to be hostile to the blastomere nuclei because in most cases, clear evidence of nuclear degeneration was observed. This finding confirms the earlier conclusion of Tarkowski & Balakier (1980; see also Czolowska et al. 1984) that only following activation is the cytoplasm of a mouse egg capable of transforming somatic nuclei into pronuclei-like structures. In the sheep (Willadsen, 1986) and cow (Prather et al. 1987), on the other hand, successful development to term has been achieved following the fusion of blastomere nuclei to ‘enucleated’ oocytes, suggesting that there may be important species differences possibly related to the time of onset of rRNA synthesis. It is possible that, in these species, activation of the cytoplasm is induced during the fusion procedure or shortly afterwards.

When blastomere nuclei were fused to fertilized mouse eggs all three nuclei increased in volume and entered into mitosis. Similarly, in experiments in which blastomere nuclei were fused with 1-cell single-pronuclear haploid parthenogenones, one discrete diploid metaphase plate derived from the blastomere nucleus was observed, in association with a single haploid metaphase plate of parthenogenetic origin. In the vast majority of cases (49/51 instances), no evidence of silver-positive NOR regions was observed in any of the metaphase chromosomes.

In most instances in which neither pronuclei nor blastomere nuclei entered mitosis, the nucleoli of the blastomere interphase nuclei no longer displayed their silver-positive reticular component. Thus the 1cell cytoplasm appeared to be capable of reprogramming nuclei from 8- to 16-cell embryos and, as a result, the silver-positive NOR material was not seen in the metaphase chromosomes and interphase nucleoli. Rarely, the 8- to 16-cell interphase nuclei still demonstrated some NOR-positive reticular material and, in two cases, the blastomere-derived metaphase chromosomes displayed some NOR-positive silver-staining regions.

The nature of the silver-staining material responsible for the positive-NOR visualization has been extensively studied by different authors (for review, see Howell, 1982; Goessens, 1984; Hubbell, 1985). This method detects acidic nonhistone proteins, rather than ribosomal cistrons. At least three proteins (C23, B23 and one designated as ‘AgNOR protein’) were stained by silver nitrate. One of these proteins probably corresponds to a large subunit of RNA polymerase I (195000M_r doublet) (Williams et al. 1982; Goessens, 1984). This enzyme plays a key role in the synthesis of rRNA, and silver staining, which is widely used to detect NOR in the karyotypes of several organisms including mice, is widely accepted as a method of visualizing the chromosome regions that were actively engaged in the synthesis of rRNA during the preceding cell cycle (Miller et al. 1976; Goessens, 1984).

It is believed that the chromosomes of 1-cell-stage mouse embryos are genetically inactive and ribosomal RNA synthesis is not detectable until the 2-cell stage is achieved (Clegg & Piko, 1983; see for reviews Johnson, 1981; Johnson et al. 1984; Schultz, 1986; Dyban & Baranov, 1987). This is why the NOR region can only be visualized in mouse metaphase chromosomes from the 2-cell stage and onwards (Engel et al. 1977; Hansmann et al. 1978; Patkin & Sorokin, 1983).

Our experiments confirm that there were no silver-staining NOR regions in 1-cell-stage embryos (fertilized or parthenogenetically activated), but they were present in the chromosomes from control 8- to 16-cell mouse embryos. When, however, these nuclei were transferred into 1-cell-stage embryos, silver-positive NOR were as a rule not detectable.

There is a one-to-one correlation between the chromosomal sites showing hybridization of 28S and 18S rRNA, and those staining with silver nitrate (Goodpasture & Bloom, 1975). We therefore believe that the disappearance of the NOR silver-staining regions in the chromosomes from the 8- to 16-cell blastomeres can be interpreted as the switching off of transcription of ribosomal genes as a result of the inhibitory action of the 1-cell-stage cytoplasm. In our study, there were a few cases in which interphase nuclei and metaphase chromosomes from 8-cell-stage embryos still retained NOR-positive silver staining. In these exceptional embryos, ribosomal genes were probably active despite the presence of the cytoplasm from the 1-cell-stage embryo. Further studies will clearly be required to investigate these exceptional cases.

There was a difference in the behaviour of blastomere nuclei that had been introduced into fertilized eggs compared with those that had been transferred into parthenogenetic eggs. In all the former cases, the blastomere nuclei entered mitosis following their transfer to the fertilized egg. When blastomeres were fused with parthenogenetic haploid 1-cell embryos, however, in only a proportion of cases were metaphase plates observed that were derived from pronuclei and from blastomere nuclei. In others, only pronuclei entered into mitosis, while blastomere nuclei remained in interphase or progressed to very early prophase. This finding may have been due to the fact that these blastomere nuclei may have been asynchronous in their chromosomal cycle compared with the pronuclear chromosomal cycle. Alternatively, the increased postovulatory age of this population of eggs may have detrimentally influenced their developmental potential after micromanipulation.

In products of fusion in which there were metaphase plates from pronuclei as well as from blastomere nuclei the pronuclei-derived haploid metaphase plates were always normal. By contrast, the chromosomes within the metaphase plates derived from the blastomere nuclei were sometimes morphologically abnormal, with many structural aberrations present.

This morphologically abnormal appearance may have resulted from the fusion of a blastomere nucleus that was significantly out of synchrony with the 1-cell parthenogenetic haploid embryo. If a blastomere nucleus enters into mitosis before its DNA replication cycle has been completed, this may lead to structural chromosomal aberrations in these reconstructed embryos. Further cytogenetic investigations are clearly required to investigate the events that occur in reconstructed mouse embryos obtained by fusion of blastomeres with parthenogenetic, gynogenetic and androgenetic haploid 1-cell-stage embryos, and with enucleated fertilized ova. It is possible that these studies may shed light on the reason why such reconstructed embryos in this species (but possibly not in other species, such as in the sheep (Willadsen, 1986) and cow (Prather et al. 1987), are not capable of supporting normal preimplantation development when nuclei from advanced embryos are introduced into an enucleated activated 1-cell embryo.

The work described in this paper was carried out during the visit of APD to the Department of Anatomy, University of Edinburgh. APD thanks the British Council for financial support. The research was partly supported by grants (to MHK) from Action Research for the Crippled Child, the Cancer Research Campaign and the Melville Trust for the Care and Cure of Cancer.