Introduction
Nuclear transfer (NT) experiments using adult mammalian somatic cells began 10 years ago (Wilmut et al., Reference Wilmut, Schnieke, McWhir, Kind and Campbell1997). However, the biology underlying this process remains unclear in many respects and cloning by somatic cell nuclear transfer (SCNT) still has a very low success rate. We have developed a microinjection-based nuclear transfer method to generate cloned mice from adult or foetal somatic cells (Wakayama et al., Reference Wakayama, Perry, Zuccotti, Johnson and Yanagimachi1998; Wakayama & Yanagimachi, Reference Wakayama and Yanagimachi1999, Reference Wakayama and Yanagimachi2001b), somatic stem cells (Mizutani et al., Reference Mizutani, Ohta, Kishigami, Van Thuan, Hikichi, Wakayama, Kosaka, Sato and Wakayama2006), or embryonic stem (ES) cells (Wakayama et al., Reference Wakayama, Tabar, Rodriguez, Perry, Studer and Mombaerts2001). Recently, we improved the success rate of cloning in mice up to six-fold using dimethyl sulphoxide (DMSO) (Wakayama & Yanagimachi, Reference Wakayama and Yanagimachi2001a), trichostatin A (Kishigami et al., Reference Kishigami, Mizutani, Ohta, Hikichi, Thuan, Wakayama, Bui and Wakayama2006a, Reference Kishigami, Bui, Wakayama, Tokunaga, Van Thuan, Hikichi, Mizutani, Ohta, Suetsugu, Sata and Wakayama2007) or serial nuclear transfer (Wakayama et al., Reference Wakayama, Mizutani, Kishigami, Thuan, Ohta, Hikichi, Bui, Miyake and Wakayama2005). However, the overall efficiency of mouse cloning is less than 6% and the offspring that survive to term have displayed many abnormalities (Wakayama, Reference Wakayama2007), which are probably caused by incomplete genomic reprogramming after nuclear transfer into the oocyte cytoplasm. The effect of the donor cell type on the cloning success rate has been reported previously and all studies used hybrid mouse oocytes because they have shown better results in in vitro culture. However, the quality, volume and reprogramming potential of the oocyte cytoplasm have not yet been analysed.
In general, the cytoplasm of a single oocyte has surplus potential for fertilizability, including its capacity to support sperm decondensation and male pronucleus formation. For example, a normal oocyte can accept and decondense up to four spermatozoa (Clarke & Masui, Reference Clarke and Masui1987), whereas a quarter-size oocyte can accept only one spermatozoon (Tarkowski & Balakier, Reference Tarkowski and Balakier1980; Wakayama & Yanagimachi, Reference Wakayama and Yanagimachi1998). Although quarter-size oocytes fail to develop after fertilization, half-size oocytes can develop to full term without a reduction in the birth rate (Wakayama & Yanagimachi, Reference Wakayama and Yanagimachi1998). However, it is not yet clear whether the cytoplasm of an oocyte is sufficient to reprogramme one somatic cell nucleus in SCNT experiments, because in nature, the reprogramming factor(s) of the oocyte cytoplasm is prepared for the receipt of a gamete nucleus, not a somatic cell nucleus. Unmatched reprogramming factor(s) may reprogramme the somatic cell nucleus inefficiently, or simply because of their inadequate reprogramming potential, the somatic cell nucleus may require more reprogramming factors to erase somatic memory and to achieve totipotency. In the latter case, cloning success rates might be improved by the addition of supplementary oocyte cytoplasm before the SCNT procedure.
Occasionally, female mammals ovulate large diploid ‘giant oocytes’. The frequency varies from 0.1 to 0.5% in many mammalian species, including humans (Austin & Braden, Reference Austin and Braden1954; Austin, Reference Austin1960; Funaki & Mikamo, Reference Funaki and Mikamo1980; Balakier et al., Reference Balakier, Bouman, Sojecki, Librach and Squire2002). To overcome this low spontaneous ovulation rate, giant oocytes have been constructed artificially. However, there is a consensus that naturally derived mature giant oocytes have abnormal diploid chromosome content and contain either one mixed or two metaphase plates, instead of the normal haploid chromosomes of the mature oocyte. Therefore, to distinguish them, we here refer to them as ‘artificial giant (AG) oocytes’. To the best of our knowledge, the production of AG oocytes and embryos was first reported by Soupart (Reference Soupart1980), using Sendai virus fusion. However, this approach was replaced with less toxic and more reproducible treatments using polyethylene glycol (PEG) (Eglitis, Reference Eglitis1980; Gulyas et al., Reference Gulyas, Wood and Whittingham1984) and more recently, by electrofusion (Gordo et al., Reference Gordo, Rodrigues, Kurokawa, Jellerette, Exley, Warner and Fissore2002). Several reports of the use of these techniques have been published. Tesarik et al. (Reference Tesarik, Nagy, Mendoza and Greco2000) used mature mouse oocytes fused with germinal vesicle-stage oocytes and activated oocytes, zygotes, or aged oocytes to determine the effects of different stages of the cell cycle on development. Others have applied these techniques to study the mechanisms by which oocytes resume meiosis and regulate their surface membrane functions (Fulka et al., Reference Fulka, Flechon and Flechon1989, Reference Fulka, Notarianni, Passoni and Moor1993, Reference Fulka, Kalab, First and Moor1997); to determine the distribution of mitochondria (Fulka, Reference Fulka2004); to evaluate the nature of the block to polyspermy at the oolemma (Krukowska et al., Reference Krukowska, Wielkopolska, Czolowska, Maleszewski and Tarkowski1998); and to investigate cytoplasmic deficiencies in aged oocytes (Gordo et al., Reference Gordo, Rodrigues, Kurokawa, Jellerette, Exley, Warner and Fissore2002). In vitro developmental potential has also been examined using the fusion of two mature intact oocytes. When those diploid AG oocytes are activated parthenogenetically, they usually extrude two second polar bodies and form two female pronuclei (Gulyas et al., Reference Gulyas, Wood and Whittingham1984; Vassetzky & Sekirina, Reference Vassetzky and Sekirina1985; Karnikova et al., Reference Karnikova, Jacquet and Fulka2000), as do naturally ovulated and fertilized giant oocytes (Rosenbusch and Schneider, Reference Rosenbusch and Schneider1998; Karnikova et al., Reference Karnikova, Jacquet and Fulka2000; Balakier et al., Reference Balakier, Bouman, Sojecki, Librach and Squire2002). Such activated AG oocytes can develop to blastocysts in vitro, but their potential to develop to full term has never been demonstrated, because parthenogenetic embryos usually die (Gulyas et al., Reference Gulyas, Wood and Whittingham1984; McGrath & Solter, Reference McGrath and Solter1984; Surani et al., Reference Surani, Barton and Norris1984).
In this study, we first defined the optimal methods of constructing AG oocytes and examined the in vitro developmental potential of these oocytes using parthenogenetic activation. Secondly, we fused intact oocytes with enucleated oocytes and examined their in vivo developmental capacity after intracytoplasmic sperm injection (ICSI) with isolated sperm heads. Finally, we examined the nuclear reprogramming potential of the enucleated AG oocytes after SCNT.
Materials and Methods
Animals
Oocytes, spermatozoa and cumulus cells were harvested from B6D2F1 (C57BL/6 × DBA/2) strain mice. ICR strain pseudopregnant females mated with vasectomized males of the same strain were used as surrogate mothers for the transferred embryos.
All animals were maintained in accordance with the guidelines of the Laboratory Animal Service at RIKEN Center for Developmental Biology. The protocols for animal handling and treatment were reviewed and approved by the Animal Care and Use Committee at the same institution.
Oocyte collection
Female mice were induced to superovulate with consecutive injections of equine chorionic gonadotropin (eCG; 5 IU) and human CG (hCG; 5 IU) 48 h apart. Fourteen hours after the human CG injection, cumulus–oocyte complexes (COC) were collected from the oviducts. The oocytes were freed from the cumulus cells by the addition of 0.1% bovine testicular hyaluronidase (ICN Biochemicals) to the COC-containing medium. After the cumulus cells had been dispersed from the oocytes, the oocytes were rinsed twice with CZB medium (Chatot et al., Reference Chatot, Lewis, Torres and Ziomek1990) and stored under 5% CO2 at 37.5 °C for 30 min until use.
Enucleation
For ICSI and NT experiments, the second meiotic metaphase (MII) spindle of the oocytes had to be removed before the AG oocytes were created. A group of oocytes was transferred to a droplet (about 10 μl) of HEPES–CZB medium containing 5 μg/ml cytochalasin B under mineral oil on a micromanipulation microscope stage at 25 °C. An oocyte was selected and immobilized at the tip of a holding pipette by gentle suction. The zona pellucida of the oocyte was then penetrated by a piezo-actuated (Prime Tech) enucleation pipette (Wakayama et al., Reference Wakayama, Perry, Zuccotti, Johnson and Yanagimachi1998). The MII chromosome–spindle complex, distinguished as a translucent region in the ooplasm, was drawn into the pipette with minimal accompanying ooplasm. As the pipette tip was withdrawn through the zona pellucida entry site, its contents were pinched off, leaving behind an enucleated oocyte (ooplast). One batch of 20 oocytes typically took 6 min to enucleate. Immediately following enucleation, the ooplasts were placed under oil in droplets of cytochalasin-B-free CZB medium under 5% CO2 at 37.5 °C for 1–3 h until use. The detailed protocols have been reported elsewhere.
Removal of the zona pellucida
Intact or enucleated oocytes were treated with acid Tyrode's solution (pH 2.5) for 30 s to dissolve the zona pellucida. Zona-free oocytes were washed with HEPES–CZB medium at least twice and then cultured in CZB medium, as above, for at least 30 min before use.
Oocyte fusion
In this study, we used two fusion methods, mechanical and electrofusion. For mechanical fusion (Fig. 1a), zona-free oocytes were transferred to HEPES–CZB medium in the micromanipulation chamber. One oocyte was held by the holding pipette and another was placed next to it. Both oocytes had to be exactly in line between the holding and fusing pipettes. The fusing pipette (ID: 3 μm) was then inserted into the cytoplasm of the second oocyte, almost to the membrane of the first oocyte, as shown in Fig. 1a. Finally, both oocyte cell membranes were perforated using a piezo-pulse and the fusing pipette was pulled out gently. The paired oocytes were kept on the micromanipulation stage for about 20 min until fusion began and were then transferred and stored in CZB medium for more than 1 h, as above, to recover from fusion damage.
Methods of mechanical and electrofusion. (a). Mechanical fusion. (b) Electrofusion. Nine oocytes were clustered using phytohaemagglutinin and then given a fusion pulse. About 30 min later, one group of six and one group of two oocytes had fused, but one oocyte remained apart.
For electrofusion (Fig. 1b), pairs of zona-free oocytes were placed together in HEPES–CZB medium containing 0.05 mg/ml phytohaemagglutinin (PHA) for 1 min. They were then moved to new HEPES–CZB medium and pushed together using a micropipette. Once the oocytes had become firmly attached, they were placed into Ca2+-free fusion medium (0.3 M mannitol, 0.1 mM MgSO4 and 0.1% polyvinyl pyrrolidone [PVP]) until they sank to the bottom of the droplet. Within a few minutes, the paired oocytes were picked up and transferred to a fusion chamber with electrodes 0.5 mm apart, covered with fusion medium. A current was then applied (20 V alternate current [AC] at 1 MHz) for 1–2 s to line up the paired oocytes. Immediately thereafter, 70 V, 100 V, or 150 V direct current (DC) pulses were applied for 20 μs to cause fusion. The oocyte pairs were kept in the chamber for 5 min, then were carefully removed, transferred to CZB medium and incubated for at least 1 h to recover from fusion damage. In some experiments, ten oocytes or more were lined up for electrofusion in an attempt to generate even larger AG oocytes, to evaluate the appropriate volume of cytoplasm for the cloning procedure.
After fusion, we examined whether the oocytes were spontaneously activated or not.
Parthenogenetic activation of AG oocytes
In these experiments, AG oocytes were generated from nonenucleated oocytes (Fig. 2a). The fused oocytes were transferred to activation medium (10 mM SrCl2 in Ca2+-free CZB medium; (Bos-Mikich et al., Reference Bos-Mikich, Whittingham and Jones1997). Six hours after activation, those oocytes had extruded their second polar bodies and contained female pronuclei. For the control experiments, single oocytes were activated with cytochalasin B, which prevents the extrusion of the second polar body, resulting in diploid embryos (Bos-Mikich et al., Reference Bos-Mikich, Whittingham and Jones1997). The parthenogenetically activated zygotes were washed in CZB medium at least twice and cultured in CZB medium, as above, for 96 h to examine the in vitro developmental potential of the embryos.
Experimental design. (a) Two intact oocytes were fused and activated parthenogenetically to examine their developmental potential in vitro. (b) Intact and enucleated oocytes were fused and then sperm heads were injected to examine the potential of the fertilized embryos to develop to term in vivo. (c) Two enucleated oocytes were fused and then somatic or ES cell nuclei were injected to examine the reprogramming potential of the AG oocyte cytoplasm.
Fertilization of AG oocytes by ICSI
In this experiment, AG oocytes were generated by the fusion of pairs of nonenucleated and enucleated oocytes (Fig. 2b). Thus, the AG oocytes possessed only one MII spindle. For sperm preparation, the cauda epididymis was removed from a mature male mouse (B6D2F1, 12–16 weeks old). Blood and adipose tissue were removed from the surface and the caudal (enlarged) part was excised with a pair of fine scissors. This tissue was compressed to release a dense mass of spermatozoa from which drops (about 2 μl) were taken. These spermatozoa were placed in a dish containing 0.4 ml of TYH (Toyoda et al., Reference Toyoda, Yokoyama and Hoshi1971) and cultured in a CO2 incubator at 37 °C for 20–30 min. The sperm concentration was approximately 3–5 × 106/ml at this stage.
ICSI was carried out using the technique described by Kimura & Yanagimachi (Kimura & Yanagimachi, Reference Kimura and Yanagimachi1995) with modifications. A droplet of sperm suspension (2–3 μl) was mixed thoroughly with 20 μl of HEPES–CZB medium containing 12% (w/v) PVP (Mr 360 kDa) in the micromanipulation chamber. The sperm head was separated from the midpiece and tail by the application of one or more piezo pulses and injected into an AG oocyte (Fig. 2b). Injected oocytes were transferred into CZB medium under mineral oil and incubated as described above. In the control experiments, ICSI was performed on single intact or zona-free oocytes and the zygotes were cultured in the same way.
Nuclear transfer into AG oocytes
In these experiments, AG oocytes were generated by the fusion of two enucleated oocytes, so there were no MII spindles (Fig. 2c). Nuclear transfer to cumulus and ES cells (line GRP14, which was established in our laboratory in the BD129F1 genetic background) was performed as described previously (Wakayama et al., Reference Wakayama, Perry, Zuccotti, Johnson and Yanagimachi1998, Reference Wakayama, Mizutani, Kishigami, Thuan, Ohta, Hikichi, Bui, Miyake and Wakayama2005; Kishigami et al., Reference Kishigami, Wakayama, Thuan, Ohta, Mizutani, Hikichi, Bui, Balbach, Ogura, Boiani and Wakayama2006b). Briefly, cumulus cells were collected at the same time as the oocytes were collected. After treatment with hyaluronidase, 2 μl aliquots of the condensed cumulus cell suspension were transferred to PVP medium (as with ICSI, described above). The ES cells were thawed two days before each experiment and collected by trypsinization. To remove the trypsin, the cells were washed at least twice in phosphate-buffered saline (PBS) and 2 μl aliquots of the condensed cell suspension were transferred to PVP medium.
For nuclear injections, AG oocytes were transferred to a micromanipulation chamber. The donor cell was then drawn in and out of the injection pipette until the cell membrane was broken. In some cases, a few piezo pulses were applied to break the plasma membrane. After each cell nucleus had been drawn deep into the pipette, another was drawn into the same pipette. Within a few minutes, several nuclei were lined up within a single pipette and were injected one by one into separate AG oocytes at room temperature (Kishigami et al., Reference Kishigami, Wakayama, Thuan, Ohta, Mizutani, Hikichi, Bui, Balbach, Ogura, Boiani and Wakayama2006b).
After NT, the reconstructed oocytes were activated with 10 mM SrCl2 in Ca2+-free CZB medium in the presence of 5 μg/ml cytochalasin B and 1% DMSO and cultured for four days in KSOM medium (Specialty Media).
In vitro and in vivo development
The oocytes were examined using an inverted microscope between 5 and 7 h after sperm injection or artificial activation. When the AG oocytes contained pronuclei or pseudopronuclei, they were considered to be activated. In the ICSI experiments, only those oocytes with two distinct pronuclei and a single second polar body were considered normally fertilized. These oocytes were cultured for up to 72 h to evaluate embryo development. Because they lacked the zona pellucida, these embryos often aggregated. To prevent this, each embryo was placed individually into a small hole made by pressing a fine needle onto the plastic dish under the medium.
When the embryos reached the morula–blastocyst stage, they were transferred to the uteri of pseudopregnant foster mothers (ICR strain, albino), which had been mated with vasectomized ICR males three days previously. All recipient females were euthanized at 19.5 days post coitus (dpc) and their uteri were examined for the presence of fetuses and implantation sites. Live fetuses were raised by lactating foster mothers (ICR strain).
Observation of oocyte spindle morphology by confocal laser microscopy
AG oocytes and control oocytes were fixed for 30 min in PBS–polyvinyl alcohol (0.1 mg/ml PVA) containing 3.5% (w/v) paraformaldehyde and 0.2% (v/v) Triton X-100. The fixed oocytes were washed twice in PBS–PVA and stored overnight at 4 °C in PBS supplemented with 3% (w/v) bovine serum albumin (PBS–BSA; Sigma–Aldrich) and 1% (v/v) Triton X-100 (Nacalai Tesque, Inc.).
To stain their α-tubulin, the oocytes were incubated with anti-α-tubulin antibody conjugated with fluorescein isothiocyanate (FITC; 1:100; Sigma–Aldrich) for 60 min at room temperature. After the oocytes had been washed three times in PBS–BSA for 10 min each, their DNA was stained with 400 μg/ml propidium iodide (Sigma–Aldrich). When they had been washed thoroughly, the oocytes were mounted on slides using Vectashield Mounting Medium (Vector Laboratories Inc.) and observed under a Bio-Rad Radiance 2100 confocal scanning laser microscope (Bio-Rad Laboratories).
Production of chimeric mice
Control embryos with increased size rather than volume were produced by the aggregation of two 8-cell embryos, as described previously (Nagy et al., Reference Nagy, Gertsenstein, Vintersten and Behringer2003). Briefly, BDF1 females were mated with BDF1 males and 8-cell embryos were collected at 3 dpc. The embryos were aggregated in a small hole in a culture dish. The next day, the aggregated blastocysts were transferred into 2.5 dpc pseudopregnant ICR females. At 19.5 dpc, the pregnant females were euthanized by cervical dislocation and the chimeric offspring were collected by Caesarean section.
Statistical analysis
Outcomes were evaluated using chi-squared tests and p < 0.01 was regarded as statistically significant.
Results
Fusion conditions
It has been reported that mouse AG oocytes produced using PEG fusion cannot be activated consistently. However, in mouse and human oocytes, the stimulation caused by pipette penetration does not, by itself, activate the oocytes (Tesarik & Testart, Reference Tesarik and Testart1994; Kimura & Yanagimachi, Reference Kimura and Yanagimachi1995; Ogura et al., Reference Ogura, Ogonuki, Takano and Inoue2001). To avoid any undesired oocyte activation caused by the fusion protocol, we developed a new oocyte fusion method by modifying the method of Tesarik et al. (Reference Tesarik, Nagy, Mendoza and Greco2000), in which paired zona-free oocytes were fused using a relatively large pipette. In this experiment, we did not use phytohaemagglutinin P because all the oocytes treated with this agent died just after fusion (data not shown). When two intact zona-free oocytes were arranged in a straight line and fused with a single piezo pulse, 53% of the pairs fused successfully and none of them was activated spontaneously (Table 1). When the AG oocytes were observed by confocal microscopy after immunostaining for α-tubulin, no abnormalities were apparent compared with the control single oocytes (Fig. 3f).
Fusion rate of paired mouse oocytes by electro- or mechanical method.
DC period was fixed at 20 μs.
AG oocytes, giant embryos and full-term development. (a) Naturally ovulated oocyte, 6 h after strontium activation with cytochalasin B. (b) Control intact oocyte and fused oocytes that were two to five times the normal size. (c) Activated oocytes extruded the expected number of second polar bodies, equivalent to the number of fused oocytes. (d, e) Sperm heads were microinjected into single and double-sized AG oocytes. These extruded single second polar bodies (d) and had developed to the morula stage (e) at 72 h. (f) The spindle apparatus of a double-sized oocyte, observed with confocal microscopy. Colours: α-tubulin, green; DNA, red. (g) Offspring (named Giko) derived from a mechanically fused AG oocyte microinjected with a sperm head.
This mechanical technique requires specialized skill and it is difficult to construct many AG oocytes, so we developed an electrofusion protocol. In preliminary experiments, we found that 2 MHz and 2 s AC pulse was sufficient for the line up of oocyte and caused minimal damage if the oocytes were pretreated with PHA. We then compared the effects of DC voltages ranging from 70 V to 150 V. As shown in table 1, maximal fusion rates (78.8%) were obtained when a 100 V electrical pulse was applied. We used that setting for all subsequent experiments.
Activation and in vitro development
Next, we examined the in vitro developmental potential of AG oocytes after parthenogenetic activation. In this experiment, we made AG oocytes that were two to nine times larger than normal oocytes by electrofusion and double-sized AG oocytes by mechanical fusion of nonenucleated oocytes. It was rare to get three or more oocytes to fuse electrically at the same time. Usually, only a few oocytes within any line fused. For example, when 10 oocytes were aligned together and an electrical pulse was applied, some oocytes died or did not fuse and only a few remaining oocytes fused with each other (Figs. 1b, 3b). The largest fusion product we achieved was nine times larger than a normal oocyte. None of these mouse oocytes was spontaneously activated by either electrofusion or mechanical fusion, in contrast to the human oocytes studied by Cohen et al. (Reference Cohen, Scott, Alikani, Schimmel, Munne, Levron, Wu, Brenner, Warner and Willadsen1998).
About 2 h after fusion, which was required for recovery, the oocytes was activated parthenogenetically using strontium treatment. Almost all the electrofused AG oocytes were activated, irrespective of size. However, about 10% of the mechanically fused AG oocytes failed to activate. All activated AG oocytes formed pronuclei and extruded the expected number of second polar bodies. For example, constructs produced by fusing seven oocytes extruded seven second polar bodies within 2 h and formed seven female pronuclei (Fig. 3c).
When we examined their in vitro development 96 h after activation, most AG oocytes (88%–100%) had reached the morula or blastocyst stage, except for the double-sized AG oocytes produced by electrofusion (77%; Table 2). Interestingly, two of the nine-times normal-sized AG oocytes had divided into more than five cells at 24 h after activation, but they continued to grow and formed normal-looking blastocysts, except that they were much larger than normal 96 h after activation.
Preimplantation development rates of mouse AG oocytes followed by parthenogenetic activation.
In vivo developmental potential after ICSI
We examined the full-term developmental potential of double-sized AG oocytes after ICSI (Table 3). These AG oocytes were formed by fusing non-enucleated and enucleated oocytes, so they had only one metaphase spindle. After mechanical fusion, 26 of the AG oocytes survived after ICSI; 23 (88%) of the embryos developed to the morula or blastocyst stage in vitro (Fig. 3d, e) and after embryo transfer, one of these (4%) reached full term (Fig. 3g). After electrofusion, although the rate of development to morula or blastocyst stage was significantly lower (71/148; 48%) than after mechanical fusion, 11 of these embryos produced live offspring (15%; not significantly different from that of embryos derived from mechanical fusion). All offspring were born by Caesarean section and six survived to become healthy adults (three females, three males; the remaining six were cannibalized by their foster mothers). One mouse derived from mechanical fusion and five from electrofused oocytes were mated and all demonstrated normal fertility.
Full term development of AG oocyte after ICSI and embryo transfer.
Full-term development of twice large oocyte after NT and embryo transfer.
aFusion rate of enucleated oocytes were decreased than intact oocytes.
Despite these encouraging results, the overall success rate for full-term development in embryos produced from AG oocytes was much lower than for usual ICSI (Yanagimachi, Reference Yanagimachi1998). As controls, we also examined the effects of zona pellucida removal, the ICSI and in vitro fertilization (IVF) procedures and the size of the embryos on full-term development (Table 3). When we injected sperm heads into zona-free single oocytes or removed the zona after ICSI, most of the embryos developed to blastocysts in vitro, but only two (6%) and five (10%) offspring, respectively, were born at full term (Control A and B). In contrast, most zona-intact oocytes (78%) developed into full-term pups after ICSI. We also found that ICSI and IVF themselves had detrimental effects on zona-free embryo development, because the zona-free in-vivo fertilized zygotes showed little better development (37%) than zona-free ICSI or IVF embryos (Control D and E). Conversely, the rate of full-term development of aggregation chimeras (double-sized embryos) was almost the same as that of intact embryos (83%), which suggests that the increased mass of an embryo per se does not impair development (Control G).
Developmental potential of cloned embryos
Finally, we examined the reprogramming potential of AG oocytes by injecting either somatic (cumulus cell) or ES cell nuclei into AG oocytes produced by electrically fusing two enucleated oocytes. The fusion rate for such pairs of enucleated oocytes was lower than that observed in the other experiments; most nonfused oocytes lysed. The oocytes may have been damaged by enucleation, for example during cytochalasin B treatment.
Using cumulus cell NT, 66 oocytes survived injection, but only 16 AG oocytes (24%) contained pseudopronuclei after activation. From those, 11 embryos started first cell division and three cloned embryos (19%) developed to the morula or blastocyst stage, but none of them developed to full term after embryo transfer. In the ES cell experiments, 28 AG oocytes survived after nuclear injection and 14 (50%) contained pseudopronuclei. However, only three embryos started first cell division and none of them developed to the morula or blastocyst stage. In the control experiments, mice cloned from both somatic and ES cells were obtained when donor nuclei were injected into zona-intact enucleated oocytes.
Discussion
Giant oocytes provide interesting material for basic biological research. In our laboratory, the frequency of spontaneously ovulated giant oocytes has not been recorded, but at most, only one is seen every year (Fig. 3a), about 0.01% to 0.001% of all the oocytes we study. Therefore, in this study, we generated giant oocytes artificially using both mechanical and electrofusion methods. We succeeded in obtaining normal offspring from those AG oocytes after ICSI and first report that doubling the volume of the cytoplasm in the mouse oocyte allows normal developmental potential to full term after fertilization.
About 9% of the AG oocytes generated mechanically by piezo-activated fusion failed to be activated. This finding suggests that the mechanical approach produced considerable damage in the oocytes, from which some did not recover by the time of activation. However, once the AG oocytes were activated, all of them extruded second polar bodies and formed female pronuclei in numbers that corresponded to the number of oocytes used for fusion. The MII spindles probably did not mix after oocyte fusion. Most of those parthenogenetically activated oocytes developed to the morula or blastocyst stage, irrespective of their size and chromosome numbers. This situation clearly demonstrates that the amount of cytoplasm of an oocyte per se does not impair in vitro preimplantation development. This finding is consistent with previous reports of double-sized AG oocytes produced with Sendai virus or PEG-mediated fusion (Soupart, Reference Soupart1980; Gulyas et al., Reference Gulyas, Wood and Whittingham1984; Karnikova et al., Reference Karnikova, Jacquet and Fulka2000) (Table 2).
We examined the developmental potential of embryos produced from AG oocytes in vivo. When the oocytes were fertilized by ICSI and the giant morulae or blastocysts were transferred into recipient females, we obtained normal healthy and fertile offspring. To the best of our knowledge, this outcome is the first demonstration that such AG oocytes possess normal developmental potential to term after fertilization. The rate of development of giant embryos to the morula or blastocyst stage was similar to the rate observed for control single embryos. However, the full-term development of giant embryos was very low: only one pup (4% of embryos) resulted from mechanical fusion and 11 pups (15%) from electrofusion. This was significantly lower than the success rate following intact embryo transfer (78%). In the control experiments, normal-sized in vivo-fertilized zygotes cultured without the zona pellucida (Control E) developed to blastocysts at similar rates as intact oocytes (Control F), but their full-term developmental rate was significantly decreased, as in previous studies (Naito et al., Reference Naito, Toyoda and Yanagimachi1992; Suzuki et al., Reference Suzuki, Togashi, Adachi and Toyoda1995). Moreover, ICSI and IVF itself reduced the success rate of offspring production, whether the zona was removed before or after ICSI/IVF (Control A, B and D). Then, in these AG oocytes, both zona removal and ICSI had negative effects on full-term development. We also examined the developmental potential of SCNT-cloned embryos using AG oocytes. Before starting the experiments, we hypothesized that AG oocytes may possess extra nuclear reprogramming potential by dint of their increased cytoplasmic mass or increased reprogramming factor(s), so that the cloning success rate would be improved. However, none of the cloned embryos developed to full term. Recently, cloned cattle and mice have been born from zona-free oocytes fused with somatic or ES cells (Vajta et al., Reference Vajta, Lewis, Trounson, Purup, Maddox-Hyttel, Schmidt, Pedersen, Greve and Callesen2003; Tecirlioglu et al., Reference Tecirlioglu, French, Lewis, Vajta, Korfiatis, Hall, Ruddock, Cooney and Trounson2004; Ribas et al., Reference Ribas, Oback, Ritchie, Chebotareva, Ferrier, Clarke, Taylor, Gallagher, Mauricio, Sousa and Wilmut2005). Moreover, double-sized or larger chimeric cloned embryos produced by aggregation developed to term more efficiently than control embryos (Peura et al., Reference Peura, Lewis and Trounson1998; Boiani et al., Reference Boiani, Eckardt, Leu, Scholer and McLaughlin2003; Vajta et al., Reference Vajta, Lewis, Trounson, Purup, Maddox-Hyttel, Schmidt, Pedersen, Greve and Callesen2003). These reports suggest that neither zona-free nor large-sized embryos are disadvantaged. Why did the AG oocytes we studied here have a reduced success rate in terms of the survival of cloned embryos to term? Previously, we have reported that cloned mouse embryos often fail to generate normal spindles (Van Thuan et al., Reference Van Thuan, Wakayama, Kishigami and Wakayama2006) and form asymmetric 2-cell embryos (Wakayama et al., Reference Wakayama, Perry, Zuccotti, Johnson and Yanagimachi1998). One possibility is that the AG oocytes have twice as much cytoplasm as normal. Then, the diameter is theoretically 1.26 times greater than that of normal oocytes (volume of double size AG oocyte = 2(4/3πr3) = 4/3π(r + X)3; increased half diameter X = 0.26), which may impair the internal cell machinery of the first cell division of the giant cloned embryos. if so, even if the AG oocytes have twice the amount of nuclear reprogramming factor(s) and are able to reprogram the somatic cell nuclei more efficiently, the subsequent cloned embryos may stop at the 1-cell stage or become aneuploid. In fact, 31% of somatic cloned embryos derived from AG oocytes failed first cell division, whereas in control oocytes, none of cloned embryo failed at this stage. It is known that abnormal karyotypes, such as triploidy (Sherard et al., Reference Sherard, Bean, Bove, DelDuca, Esterly, Karcsh, Munshi, Reamer, Suazo and Wilmoth1986; Henery & Kaufman, Reference Henery and Kaufman1993), tetraploidy (Snow, Reference Snow1975), or damaged chromosomes (Kusakabe et al., Reference Kusakabe, Szczygiel, Whittingham and Yanagimachi2001), do not prevent preimplantation development, but most of these embryos die during foetal development. Therefore, even if giant cloned embryos could develop to morulae or blastocysts, they might have no hope of developing to full term.
In this experiment, although we did not obtain any cloned mice, 12 healthy offspring were obtained after ICSI, which is the first demonstration that AG oocytes can develop to term when one chromosome set is removed before fertilization. Although further analysis is required, AG oocytes should provide us with interesting tools for basic biological research, including providing new material for the study of nuclear reprogramming and genomic imprinting and may be important in widening the treatment options in human assisted reproductive technology.
Acknowledgements
We are grateful to the Laboratory for Animal Resources and Genetic Engineering for housing the mice. This work was supported by Scientific Research in Priority Areas (grant no. 15080211) and by a project for the realization of regenerative medicine (research toward technical developments in stem cell manipulation) to T. W. from the Ministry of Education, Culture, Sports, Science and Technology of Japan.
