Effects of electric field on fusion rate and survival of 2-cell rabbit embryos

Current fusion procedures applied to mammalian gametes and early embryos have certain inherent limitations and most often they are not accurately controlled (Zimmermann & Vienken, 1982). In contrast, the new electrofusion technique developed in the last few years appears to bypass some of the disadvantages by using the parameters of the vectorial system exactly defined and hence constantly repeated.

The electrofusion technique has been applied to red blood cells and lymphocytes (Zimmermann, 1982), sea-urchin eggs (Richter, Scheurich & Zimmermann, 1981) and plant protoplasts (Zimmermann, 1982), and very recently Kubiak & Tarkowksi (1985) reported successful fusion of 2-cell mouse embryos with various fusion parameters and their development beyond implantation. However, still little is known about the optimal parameters required for electrofusion of mammalian embryonic cells and about the behaviour and tolerance of such cells exposed to an electric field. Thus, it is necessary to determine the factors that provide the most advantageous conditions for the right balance between fusion rate, viability and developmental capacity of ‘fused’ cells.

The purpose of this study was to test the effects of physical parameters such as electric field strength and duration as well as the number of pulses applied at a wide range of values and in several combinations on the fusion, survival and developmental rate of 2-cell rabbit embryos.

Like other fusion procedures, the electrofusion technique can be used in the manipulation of ploidy and hence in the production of tetraploid embryos used in developmental and genetic research like other mammalian embryos of abnormal ploidy. Several attempts were made to induce tetraploidy in mice and a number of papers reported the effect of tetraploidy on the development of mouse eggs (Graham, 1971; Snow, 1973; Tarkowski, Witkowska & Opas, 1977; Eglitis, 1980; Spindle, 1981; Kubiak & Tarkowski, 1985). It was clearly demonstrated that mouse tetraploids can develop to blastocysts and a small proportion of them underwent postimplantation development (Snow, 1973, 1975, 1976; Tarkowski et al. 1977) but very seldom resulted in live-born young (Snow, 1973,1975).

In rabbits, no attempts were made to produce tetraploid embryos, but according to some experiments (Bomsel-Helmreich, 1965, 1971) triploid embryos can develop to day 16 of pregnancy. On the other hand, Harper & Chang (1971) showed that some presumably polyploid embryos (the ploidy of embryos was not determined) obtained after cytochalasin treatment were able to develop to blastocysts and even beyond postimplantation, but failed to survive after day 17. Thus, our report concerns the production of tetraploid embryos and testing of their postimplantation development in a species other than mouse.

Sexually mature mixed-bred female rabbits were superovulated with follicle-stimulating hormone (FSH) followed by pituitary luteinizing hormone (LH) according to the technique described by Kennelly & Foote (1965) and modified by Thibault (personal communication). They were given 2mg FSH in five injections at 12h intervals: 0·250, 0·250, 0·650, 0·650 and 0·250 mg, respectively. 12 h later, just before mating, they received one injection of 0·33 mg LH.

16 to 23 h after mating, fertilized ova or 2-cell embryos were recovered from the oviducts of the does by flushing the Fallopian tubes with phosphate-buffered saline (PBS). Embryos were washed in B2 medium (Ménézo, 1976) and placed before treatment in drops of this medium under liquid paraffin. Cultures were gassed with 5 % CO₂ in air and kept at 37°C.

The experimental procedure for electrically induced fusion was similar to that described by Zimmermann & Vienken (1982). Two cylindrical platinum wires, 400 μm in diameter, were mounted in a fusing chamber parallel and 200 µm apart (Fig. 1). The wires used as electrodes were connected to a stimulator (Anapulse Stimulator 301T WPI) equipped with a stimulus isolation unit (model 305R WPI). An oscilloscope was connected parallel to the electrodes for adjusting electric field pulse duration and strength. The experiment was carried out under a stereomicroscope at room temperature in a chamber filled with low conductive (15 to 32 kΩ) solution of 0·3M-glucose in distilled water.

The 2-cell embryos were washed in 0·3 M-glucose solution before they were subjected to the electric field. Each embryo was placed separately between the electrodes with the cleavage plane perpendicular to the direction of the electric field vector (Ë). Two series of experiments were carried out and the effect of the electric field, as well as of the number of pulses, was tested (groups I and II).

(A) 2-cell embryos were submitted to a single pulse and 35 ps of the following strengths: lkVcm⁻¹, l-5kVcm⁻¹, 2kVcm⁻¹, 2-5kVcm⁻¹ and 3kVcm⁻¹.

(B) Embryos were submitted to a train of four pulses of 35 ps each at 2 ms intervals, of the same strengths as in (A).

Embryos were exposed to one pulse of 2·5kVcm⁻¹ (subgroup A) or to 3kVcm⁻¹ (sub-group B) for 100, 500 or 1000 μs, respectively. Treated embryos were washed in B2 medium and then cultured. During the first 2–3 h embryos were observed under an inverted microscope (Diaphot, Nikon) fitted with a Nomarski interference contrast. The ‘nonfused’ embryos were separated from the ‘fused’ ones. Most eggs for all groups were cultured for at least 48 h. To test the postimplantation viability of ‘fused’ embryos, some of them from each group were transferred immediately after fusion to the oviduct of pseudopregnant does using the technique described by Staples (1971). Some ‘fused’ embryos belonging to group IB (four pulses of 3 kV cm⁻¹ for 35 μs each) and to group IIB (one pulse of 3 kVcm⁻¹ for 100 μs) were examined carefully after fusion till the next cleavage (see Results). Only embryos subjected to equal cleavages were transferred to pseudopregnant recipients. Autopsies of the recipients were performed between day 11 and day 20 of pregnancy and the number of implantation sites and live foetuses was determined.

Three control experiments were carried out (groups I, II and III).

Freshly collected 1- or 2-cell embryos were cultured in vitro for 5 h and then transferred to the pseudopregnant does. Efficiency of transfer and also the influence of short culture were determined by the implantation rate recorded on day 10 of pregnancy and by the number of live-born young.

• A certain number of ‘nonfused’ embryos treated with one or four pulses of 2kVcm⁻¹ were used as controls and transferred to recipients.

• 2-cell embryos placed between the electrodes with the cleavage plane parallel to Ë to prevent fusion were submitted either to one pulse of 3kVcm⁻¹ for 100 μs or to four pulses of 3kVcm⁻¹ for 35 μs and transferred to the oviduct of pseudopregnant does. In both subgroups the rate was determined by the number of live-born young.

Freshly collected 1- or 2-cell embryos were cultured for 96 h and the rate of development determined by the number of expanded blastocysts.

Ploidy of cultured ‘fused’ preimplantation embryos was determined by chromosome preparations according to the method described by King, Linares, Gustavsson & Bane (1979). Pieces of limbs, kidney, heart and liver were treated with trypsin (0·5 mg ml⁻¹ in Hank’s Ca²⁺- and Mg²⁺-free medium). Cells were washed in MEM and then centrifuged at 3000 rev min⁻¹ for 5 min. After 48 h of culture cells were treated with colcemide (0·01 mg ml⁻¹) at 37°C for 3h whereafter the standard karyotyping methods were used. Some embryos from which only pieces of limbs or limb buds were removed to assess their ploidy were used for histological analysis. Embryos were fixed in Bouin-Holland’s fluid and stained with trichrome Masson in one step. These observations will be described in detail elsewhere.

Fusion probably began at the lowest tested electric field strength (just below IkVcm⁻¹), since 27·3% of treated embryos fused at that level. Efficiency of fusion increased progressively reaching a maximum at 2·5 and 3kVcm⁻¹. Most treated embryos showed distinct signs of fusion within the first 15 min, though some fusions were only observed after 45 min. In this group no embryo lysis was seen and all fused embryos reached the early morula stage (16–32 cells) after 48 h of culture (Fig. 3F). Those in which ploidy could be assessed were diploid, tetraploid or diploid/tetraploid mosaics. Recipients with transferred ‘fused’ embryos were checked for implantation between day 11 and 14. Implantations were found in all recipients and their frequency increased with increasing electric field intensity. Among 15 implantations, 6 live foetuses with a normal appearance were found while 9 were retarded in their development, already dead or in various stages or resorption.

After treatment with four pulses, the fusion rate slightly increased, in comparison with the same electric field values as in group IA, reaching a maximum level of 88·4 % at 3 kV cm⁻¹. As in group I A, all embryos cultured for 48 h reached the early morula stage. Those embryos in which ploidy could be assessed were diploid or tetraploid or diploid/tetraploid mosaics. As in group I A, a certain number of ‘fused’ embryos at each electric field value was transferred to recipients and implantations were observed in all groups of ‘fused’ embryos. Among 54 implantations recorded between day 11 and 18, 9 normal live embryos were found while the remaining embryos were either abnormal and obviously retarded in their development or already resorbed. We succeeded in confirming the ploidy of three live embryos. One was diploid and two were diploid/tetraploid mosaics. In abnormal dead embryos ploidy could not be assessed.

To test the effect of single pulse length, the values of electric field intensity chosen were those which in group I gave the maximum rate of fusions and also the relatively highest number of implantations (i.e. 2·5kVcm⁻¹ and 3kVcm⁻¹). At 2·5kVcm⁻¹ (subgroup IIA) the fusion rate decreased from 89-6% at 35 μs to 86·6 % at 100 μs and from 78·5 % at 500 μs to 72·3 % at 1000 μs. Parallel to that the lysis rate increased from 0 to 15·4%. Among the ‘fused’ embryos treated for 100 μs, over 83·3 % reached the expanded blastocyst stage after 96 h of culture. As the pulse duration increased, the number of embryos reaching the blastocyst stage (Fig. 3G) slightly decreased, but at least two-thirds of them were still able to complete preimplantation development in vitro. In all groups, implantation of ‘fused’ embryos and of live foetuses was recorded. For a single pulse of 3 kV cm⁻¹ (subgroup IIB) the fusion rate was 85% for 35 μs, but fell to 42·5 % for 1000 μs with a rapid increase (up to 50 %) in blastomere lysis. However, when applying the latter pulse length, 12 ‘fused’ embryos out of 16 cultured for 96h reached the blastocyst stage.

In several previous experiments cultured as well as transferred embryos were unexpectedly not uniformly tetraploids. In order to examine how a reversion of ‘fused’ (i.e. presumably tetraploid) embryos to a diploid state is possible, 103 embryos belonging to Group IB (treated with four pulses of 3 kVcm⁻¹ for 35 μs) and 79 embryos belonging to Group IIB (treated with one pulse of 3 kV cm⁻¹ and 100μs) were carefully examined after fusion during the whole subsequent cell cycle. The typical sequence of events is shown in Fig. 2. The majority of ‘fused’ embryos, i.e. 61·2 and 73·4% according to groups, followed this pattern. After fusion two diploid nuclei moved to the centre of the ‘fused’ cell (Fig. 2A-C) and established close contact (Fig. 2D). After that the embryos cleaved into two equal blastomeres (Fig. 2F-H). In some embryos, however, (6·8 % in embryos of group IB and 14 % in those of group IIB) small vacuoles were visible in the fusion plane (Fig. 3A) The vacuoles swelled and migrated to the centre of the fused cell between the two nuclei (Fig. 3B). In these embryos the nuclei were not able to come into close contact and remained separated from each other and finally the cell cleaved into three or four parts (Fig. 3C-E). However, this event cannot be responsible for all atypical cleavages because in those groups the direct cleavage into three or four blastomeres (18·4% and 10·1%, respectively) was observed under a light microscope without any visible vacuoles on the plane of fusion.

Among 182 embryos observed during the whole cell cycle after fusion, 121 (66·5 %) embryos cleaved into two regular and equal blastomeres. These embryos were selected and transferred to recipient does. Live foetuses were recorded until day 20 of pregnancy. On day 11, three live embryos were uniformly tetraploids and no visible external abnormalities were detected. The retarded and abnormal ones showed abnormalities mainly affecting the neural tube, particularly in the brain region. On day 14·5, four live foetuses were found (Fig. 3H). All looked normal although they were slighlty smaller than the diploid embryos at the same stage. All embryos were tetraploid. The remaining implantation sites contained abnormal dead embryos with several types of abnormalities or only remnants of foetal membranes and embryos. On day 18, only one live embryo out of six implants was found. This embryo was fully developed and corresponded exactly to day 18 of pregnancy. The karyological analysis revealed diploid/tetraploid mosaicism with the distinct majority of diploid cells. On day 20, one live embryo out of 27 implants was found. This embryo was tetraploid, fully developed and its size (24×10mm) corresponded to day 20 of pregnancy according to the table of external development of the rabbit (Edwards, 1968).

From 46 2-cell embryos transferred into recipient does, 31 normal embryos were recovered on day 10, while from 36 other transferred 2-cell embryos, 19 normal young were born (Table 5). Thus, in our experimental conditions the implantation rate after collection, culture for a few hours and surgical transfer was 67·4 % and the survival rate to term 52·7 %.

• (a) Cleavage plane of 2-cell embryos perpendicular to E. Fourteen embryos in which blastomeres did not fuse after treatment with one pulse of 2kVcm⁻¹ and 35 μs (group IA) were transferred to recipient does and gave birth to three normal live young. Ten ‘nonfused’ embryos treated with four pulses of 2kVcm⁻¹ and 35 ps (group IB) and transferred into pseudopregnant does gave birth to 10 normal live rabbits (Table 5).

• (b) Cleavage plane of 2-cell embryos parallel to E. Among 2-cell embryos treated with one pulse of 3 kV cm⁻¹ for 100 μs, fusion of blastomeres occurred in one case and among the remaining 24 transferred embryos, 13 young live rabbits were obtained (54·2 %) (Table 5). Among 32 embryos treated with four pulses of 3kVcm⁻¹ for 35μs, fusion of blastomeres occurred in 3 of them while 13 live young were obtained among the remaining 29 transferred embryos (44·8%) (Table 5).

Among 91 nontreated 2-cell embryos cultured for 96 h, 79 (86·8 %) reached the expanded blastocyst stage.

In spite of the double size of rabbit ova in comparison to mouse eggs, the minimal field strength causing fusion of blastomeres of 2-cell rabbit embryos was similar to that inducing the fusion of 2-cell mouse embryos (Kubiak & Tarkowski, 1985). It seems that the electrical breakdown of the membrane inducing the fusion of the embryonic cell is not as clearly related to cell size as in the case of other kinds of somatic cells (Zimmermann & Vienken, 1982). Electrofusion of oocytes and blastomeres taken from 2-, 4- and 8-cell embryos and small karyoplasts and cytoplasts also confirms this view (unpublished observations). The efficiency of fusion of rabbit blastomeres increases with the rise in the electric field pulse strength and duration, until these values reach a critical level. It was clearly demonstrated that the electrical breakdown of plasma membrane is dependent on pulse strength and duration (Zimmermann, 1982). However, at the lowest electric field intensity, a train of four pulses significantly increased the fusion efficiency (60 % instead of 27·2 %), most likely in relation to the development of a considerable number of pores hastening the process of fusion. At the highest field intensity (i.e. 2·5 and 3kVcm⁻¹) the duration and number of pulses did not improve the fusion rate, which remained almost constant between 85 and 90%. From 2·5kVcm⁻¹ the rate of lysis increased, showing at about 3kVcm⁻¹ the level of maximum tolerance for the plasma membrane. Above that critical value the effect of electric field pulse became harmful and irreversible (Benz & Zimmermann, 1980; Zimmermann, Vienken, Pilwat & Arnold, 1984). A small proportion of embryos never fused even after strong and multiple treatments, but we cannot explain this phenomenon. If the right position of embryos to be fused (i.e. cleavage plane perpendicular to E) appears to be important for fusion efficiency at the lower pulse values (see also Kubiak & Tarkowski, 1985) it should not make such a significant difference at higher field intensity. When the embryos were located between electrodes with the cleavage plane parallel to I? and treated with 2·5 or 3kVcm⁻¹, around 7% of them fused, showing that at certain field strength levels the electrical breakdown must occur on the whole surface of the cell. Similar observations were also reported by Richter et al. (1981). Mammalian embryos appeared to be more resistant to electric field treatment than small differentiated cells. Several types of somatic cells exhibited irreversible damage at electric field pulse values much below the critical values for embryonic cleavage stages (Zimmermann et al. 1984). Mouse thymocytes and follicle cells of much smaller volume (up to 40 times less) than the volume of oocyte were also more sensitive than mouse oocytes, blastomeres or even small karyoplasts (unpublished observations).

In all groups the development of ‘fused’ and ‘nonfused’ and treated embryos was as good as the nontreated ones. Thus, it seems that in embryonic cells an electric field pulse even relatively strong (up to 2·5 or 3kV cm⁻¹) does not create adverse effects leading to irreversible damage of the plasma membrane and inside the cell. A high rate of development until the blastocyst stage was recorded even in groups of eggs that survived the treatment of 3 kV cm⁻¹ for 1000 ps representing as previously mentioned the critical cell resistance value. However, in these experiments, current intensity never exceeded 3·7 mA; when occasionally it reached 8–10 mA due to the presence of electrolytes brought along with embryos from the culture medium, fusion occurred and the embryos survived but they never developed any further (these cases were not included in our results). Zimmermann (1982) and more recently Kubiak & Tarkowski (1985) and Blangero & Teissie (1985)\showed that fusions can occur in the presence of electrolytes. Mouse embryos fused in PBS with an electric field pulse below 1 kV cm⁻¹ can develop in vitro and complete implantation in foster mothers (Kubiak & Tarkowski, 1985). Because the relationship between potential difference (U) and current intensity (I) created by an electric field between electrodes depends on the resistance (R) of the pulsating medium (Ohm’s law, U = RI), these observations indicate that fusion can be induced with a low potential difference, but a high current intensity. In PBS the resistance between electrodes is very low and for the same power applied between electrodes (Power: P = UI) the intensity will be greater than in non-conductive solutions. Inversely, the potential difference will be greater in a non conductive versus saline solution. It would be useful to consider the Uxl power dissipated between the electrodes instead of only U to define the best electrical parameter values for optimum fusion and survival. Blangero & Teissie (1985) also showed that the efficiency of electric-field-induced fusion of mammalian cells (CHO) depends on the ionic content of the pulsating medium. In our conditions, current and ionic contents of the pulsating medium were not controlled, but at least in the case of rabbit embryos fusion in a nonconductive solution with a strong electric field treatment ensured a normal and undisturbed development. Although the structure of in vitro developed blastocysts was not accurately examined, we observed that in blastocysts originated from ‘fused’ embryos, ICM size was often reduced in comparison with ICM of blastocysts developed from ‘nonfused’ or nontreated embryos. Similar observations were reported in mouse after tetraploidization with cytochalasin B (Snow, 1975; Tarkowski et al. 1977). The majority of blastocysts developed from treated embryos hatched from the zona pellucida and some of them reached in the following two days a very expanded form at about 400–500 μm in diameter.

The rate of implantation of embryos exposed to electric field pulse was similar to that of nontreated ones and confirms previous observations according to which developmental potency of ‘fused’ and ‘nonfused’ embryos is probably not altered by electric field treatment. It must be admitted that ‘fused’ embryos did not develop beyond two-thirds of pregnancy, but it is also difficult to relate the failure of their development to other causes than that of tetraploidy. All abnormal foetuses in which it was possible to assess ploidy were uniform tetraploids. On the other hand, several normal young rabbits were born from treated but ‘nonfused’ embryos. In this case the survival rate to term was similar to nontreated embryos (Table 5).

The occurrence of an atypical cleavage of ‘fused’ embryos was unexpected. To our knowledge such an event has not been described yet. In fact, the hybrid cell resulting from fusion is confronted with at least two problems.

The first problem is the resorption or degradation of a certain area of the plasma membrane that might reach 20 % of the original surface since the fusion between cells is accompanied by a surface reduction of constant volume (Zimmermann et al. 1984). A particular defective membrane resorption in the contact zone during the fusion may result in formation of vesicles located between two nuclei (see Fig. 3A,B) leading to an atypical cleavage in the next mitosis (Fig. 3C,D,F).

The second problem is the organization of a specific micro tubular system, the role of which is to bring the two diploid nuclei in close contact after fusion and before the onset of mitosis, a system that appears to push the pronuclei towards the centre of the egg. Such a system has been well described by several authors in mouse eggs (Maro, Howlett & Webb, 1985; Schatten, Simerly & Schatten, 1985). In the case of fertilized rabbit eggs we observed that at least 5 % of 1-cell eggs cleave spontaneously into four parts (unpublished observation). This event resembles the one we described as ‘fused’ embryos and could be explained by an incorrect organization of microtubular systems causing the formation of two widely separated metaphase plates inside the cell. In such a cell, two contractile rings must operate simultaneously, cleaving the cell into four regular blastomeres (Fig. 3E). The development of diploid and diploid/tetraploid mosaic embryos could be the result of such a disfunction. In some ‘fused’ blastomeres we observed a transient ‘apparent’ fragmentation which lasted for about one hour. These cells relaxed, however, and always restored their spherical shape before the next mitosis. This event is not frequent (8·7%) (Table 3), but may suggest that application of a certain electric field pulse range may exert a reversible effect on the internal structure of the egg.

The obtained results clearly show that tetraploid rabbit embryos derived from ‘fused’ blastomeres are capable of a limited postimplantation development. Although 54·5 % of tetraploids succeed in their implantation, only a few reach midpregnancy. The majority of them die between day 10 and 14. Several embryos were obviously retarded in their development, showing severe malformations mainly in the brain region. In some of them the whole central nervous system was affected. However, some tetraploids and diploid/tetraploid mosaics appeared fully normal and their morphology was consistent with gestational age (Edwards, 1968). These results are different from those obtained in mouse by Snow (1975) and by Tarkowski et al. (1977) who did not obtain normal 4N embryos. However, it should be mentioned that Golbus, Buchman, Witlse & Hall (1976) reported the birth of a tetraploid child which survived one year and a 2-year-old 2N/4N mosaic infant reported by Kelly & Rary (1974). A more precise analysis of post-implantation development of tetraploid rabbits is required and will be further described. Collection of a 2N/4N mosaic embryo on day 18 was surprising because in this group embryos transferred at the 2-cell stage were those selected from the ‘fused’ ones in which regular mitosis with common metaphase plate occurred. This observation suggests a more complex background modifying the chromosome distribution in daughter cells.

In the light of our results on rabbit embryos and those previously obtained by Kubiak & Tarkowski (1985) in mouse embryos, we may conclude that the electric field treatment of embryonic cells does not jeopardize their development and that a wide range of pulse intensities can be used to obtain a variety of embryonic cell hybrids with a sufficient efficiency so that the technique offers new perspectives in experimental embryology (for a review see Markert, 1981; Gulyas, 1985).

We are grateful to Mrs N. Bordas for excellent technical assistance, particularly in surgical transfer of embryos, to Dr D. Szôllôsi and Prof. A. K. Tarkowski for their critical advice and their stimulating discussions; Mrs K. Rérat for revision of the English manuscript; Mr B. Moret and Mr M. Baratte for preparation and care of animals; Mr R. Scandolo and Mr C. Slagmulder for the photography; Mrs M. E. Marmillod for typing the manuscript.