| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Unité de Biologie du Développement et Biotechnologie, Institut National de la Recherche Agronomique (INRA), 78352 Jouy en Josas, France
b Institute of Developmental Biology, Chinese Academy of Sciences (CAS), Beijing 100080, China
ABSTRACT
Mice have recently been successfully cloned from embryonic stem (ES) cells. However, these fast dividing cells provide a heterogenous population of donor nuclei, in terms of cell cycle stage. Here we used metaphases as a source of donor nuclei because they offer the advantage of being both unambiguously recognizable and synchronous with the recipient metaphase II oocyte. We showed that metaphases from ES cells can provide a significantly higher development rate to the morula or blastocyst stage (56–70%) than interphasic nuclei (up to 28%) following injection into a recipient oocyte. Selective detachment of mitotic cells after a demecolcin treatment greatly facilitates and accelerates the reconstruction of embryos by providing a nearly pure population of cells in metaphase and did not markedly affect the developmental rate. Most of the blastocysts obtained by this procedure were normal in terms of both morphology and ratio of inner cell mass and total cell number. After transfer into pseudopregnant recipients at the one- or two-cell stage, the ability of metaphase to be fully reprogrammed was demonstrated by the birth of two pups (1.5% of activated oocytes). Although the implantation rate was quite high (up to 32.9% of activated oocytes), the postimplantation development was characterized by a high and rapid mortality. Our data provide a clear situation to explore the long-lasting effects that can be induced by early reprogramming events.
developmental biology, early development, implantation
INTRODUCTION
Enucleated oocytes arrested at the metaphase stage of the second meiotic division (MII) have been shown to enable the full reprogramming of diploid nuclei obtained from cultured primary [1–6] or established cell lines [7,8] in different mammalian species. This ability is related to the high activity of the maturation-promoting factor (MPF) that catalyzes nuclear envelope breakdown and chromatin condensation in chromosomes in all mitotic and meiotic cells. In nuclear transfer (NT) embryos, this activity induces the premature chromosome condensation (PCC) of the transferred nucleus, leading to the direct exposure of chromatin to the cytoplasm of the oocyte. Although this is not essential to full term development [3, 9], this exposure may facilitate the access to key remodeling factors. PCC however induces chromatin remodeling that precisely depends on the donor cell cycle stage [10–12]. While replicating S-phase DNA is rapidly disorganized into small-dispersed fragments [13] G1/G0 (diploid) or G2 (tetraploid) nuclei form single and double chromatids, respectively, that can be fully reprogrammed into live offspring [7].
Different strategies have now been developed to maintain the correct ploidy of the reconstructed embryo (2n) after fusion with [14] or injection into [5, 7] an enucleated MII cytoplasm. Recent results in the mouse have shown that nuclei from cultured embryonic stem cell lines (ES cells) presumed to be either in the G1 phase or in the G2/M phase can support full term development after injection into enucleated MII oocytes [7]. In this work, neither the rates of in vitro development to the morula/blastocyst stages, nor that of in vivo development into fetuses and/or live offspring significantly differed between the two types of nuclei [7]. According to these authors this reflected the fact that nuclear totipotency defined by themselves as "the ability of a nucleus to direct full embryonic development" is not linked to the cell cycle stage of the donor cell [7]. Consequently, an initial asynchrony between the cell cycle stages of the cytoplasmic and the nuclear compartments of the reconstructed egg, respectively, can be considered as not developmentally essential as soon as normal ploidy (2n) is maintained. This would mean that the changes in chromatin that immediately affect a G1/G0 or a G2/M foreign nucleus upon its exposure to an MII cytoplasmic environment do not interfere with reprogramming. However, little is known about the developmental effects of initial perturbations affecting the NT embryos. To better understand the consequences of these early reprogramming events, it appears necessary to design a protocol in which the initial changes affecting chromatin condensation/decondensation could be minimized.
In the present study, we compared the developmental potential of mouse embryos reconstructed with metaphase nuclei to interphasic nuclei obtained from ES cells. Metaphase nuclei were considered as synchronous with the cytoplasm of recipient MII oocytes because they were introduced as a condensed chromatin already organized as that just removed from the cytoplasm of recipient MII oocytes. By contrast, interphasic (G1, S, or G2) nuclei were not synchronous with the recipient cytoplast because they were already in a decondensed state. Our results show that a large number of NT embryos reconstructed with M nuclei developed in vitro into morphologically normal blastocysts, whereas this number was much lower when embryos were reconstructed with interphasic nuclei. However, postimplantation development of metaphase-reconstructed embryos was rapidly compromised, even though we observed the birth of normal healthy young.
MATERIALS AND METHODS
Experimental Design
Our first objective was to characterize the population of donor cells. We used cytometric analyses to examine the distribution of cells in the cell cycle and whether we could use the cell diameter to discriminate cells in G1 or G2 phases. We used two different synchronization protocols in order to increase the proportion of nuclei, respectively, in G1, M, and G2 stages. Before using the different nuclei in NT experiments, we also improved the efficiency of our NT technique. Finally, reconstructed embryos were transferred in pseudopregnant females to examine their postimplantation development.
Cells and Media
The R1 ES cell line was used as donor cells. These cells were a generous gift from Dr. A. Nagy (Mount Sinai Hospital, Toronto, Canada) and were derived from 129/SvJx129/Sv blastocysts [15]. They were routinely cultured onto mitomycin-inactivated embryonic fibroblasts in Dulbecco modified Eagle medium (DMEM) high glucose (Gibco [Life Technologies], Gaithersburg, MD), supplemented with 15% heat-inactivated fetal calf serum (FCS), 2 mM glutamine, and 0.1 mM ß-mercaptoethanol in the presence of penicillin and streptomycin.
When mentioned, cells were grown on gelatin-coated plates and the medium was then supplemented with 2000 U/ml ESGRO (Gibco).
Flow Cytometric Analysis of the DNA Content
Cells were collected either by trypsin treatment or by shake-off. The cell suspension was centrifuged and the cell pellet was resuspended in GM/0.5 mM EDTA solution (1.1 g/L glucose, 8 g/L NaCl, 4 g/L KCl, 0.39 g/L Na2HPO4·12H2O, 0.15 g/L KH2PO4, 0.5 mM EDTA) [16] and fixed by adding three volumes of cold 95% ethanol. After 30 min of fixation, the cells were centrifuged, washed once in PBS, and stained for 20 min with 40 µg/ml propidium iodide in PBS (106 cells/ml). For each sample, 104 events were stored for analysis using a FACScan cytometer (Becton-Dickinson, San Jose, CA) equipped with an argon laser (excitation 488 nm) and a 610-nm bypass filter. Data were analyzed with the Cell Fit software.
Bromodeoxyuridine Incorporation and Nuclear Staining
A BrdU incorporation assay was used to detect in situ DNA synthesis. Cultured cells were incubated for 30–60 min with 10 µM BrdU (Sigma, St. Louis, MO), before fixation with cold methanol. DNA was denatured by 15 min incubation with 2 N HCl, followed by neutralization with borate buffer [17]. The BrdU staining was revealed by incubation with a monoclonal anti-BrdU antibody (Roche Molecular Biochemicals, Indianapolis, IN) diluted 1:50 and fluorescein isothiocyanate-conjugated anti-mouse IgG (Sigma). Counterstaining of the nuclei was achieved by incubation with 1 µg/ml propidium iodide.
For cytometric studies, cells in suspension were plated on polysine-coated coverslips to allow rapid attachment of the cells. They were fixed with 2.5% paraformaldehyde, rinsed in PBS, and incubated for 10 min with 1 µg/ml propidium iodide.
Coverslips were mounted with the antifading agent Vectashield (Vector, Burlingame, CA) and analyzed with a confocal laser scanning microscope (see below).
Image Cytometric Studies
Fixed cells were observed under a confocal microscope (CLSM 310, Carl Zeiss, Oberkocher, Germany) with a 40x objective (numerical aperture = 1.3), and cytometric analyses from images were done with the ImageTool software (developed at the University of Texas Health Science Center at San Antonio, Texas and available from the Internet by anonymous FTP from maxrad6.uthscsa.edu), using script files developed in our laboratory. Briefly, cells and nuclei were selected by manual drawing and thresholding, and three parameters were measured for each cell: the nuclear intensity, the cell and nucleus diameters. At least 200 cells were analyzed. Linear regression curves and the R2 statistics were calculated, and t-tests were used to determine whether the correlation between each couple of parameters (cell and nucleus diameters, nucleus intensity and diameter) was different from 0, at level 1%. To analyze the distribution of the cells in S-phase according to the nucleus diameter, each nucleus was evaluated for the presence of BrdU staining.
Synchronization of the Donor Cells
Embryonic stem cells were synchronized in metaphase according to the protocol described [18] with slight modifications. Subconfluent cultures were passaged the day before and diluted twice on 0.2% gelatin-coated culture flasks. The flasks were then preshaked to remove the loosely adherent cells. The medium was replaced with culture medium containing 0.05 µg/ml demecolcin (Sigma). After 3 h of incubation, the mitotic cells appeared as loosely attached cells and were collected by again shaking the flasks. They were washed twice with warm complete medium and resuspended in medium before use.
For G1 synchronization, mitotic cells were collected after an overnight incubation with 0.02 µg/ml demecolcin, washed twice with culture medium, and incubated for 2–3 h to allow the resumption of the cell cycle.
The population of cells released from the demecolcin block rapidly desynchronized and thus could not be used to obtain cells in G2. Synchronization in G2 phase was thus carried out in two steps, according to Crissman and colleagues [16]. First, cells were presynchronized in S-phase by two successive thymidine blocks, and second they were arrested in G2 by incubation with the topoisomerase II inhibitor Hoechst 33342. Cells were preplated to get rid of most of the feeder cells, then plated onto gelatin-coated dishes. The following day, they were incubated for 7 h with medium containing 0.8 mg/ml thymidine (Sigma). They were washed twice with culture medium to remove thymidine and then replaced in the incubator for 15 h. They were incubated for 7 additional h with thymidine, washed twice, and incubated for 15 h with culture medium containing 5 µg/ml Hoechst 33342 (Sigma). The cells were harvested by trypsin treatment for subsequent use.
Nuclear Transfer
Eight-week old (C57Bl/6 x CBA/J) F1 females were superovulated with eCG (10 IU) and hCG (5 IU). Oocytes were collected from oviducts 13 h after hCG injection. Cumulus cells were removed with hyaluronidase (300 IU/ml), and oocytes were washed with Hepes-CZB several times. They were then cultured in CZB medium at 37.5°C, under 5 % CO2 in air.
Manipulations (Nikon-Narishige Micromanipulators MO-188; Nikon, Tokyo, Japan) were performed under differential interference contrast (DIC equipped Nikon Diaphot) at 20x (objective) magnification. Oocytes were first incubated for 5 min at 37.5°C in Hepes-CZB containing 5 µg/ml cytochalasin B (CB), then were placed in a chamber containing 200 µl of the same medium. Oocyte position was carefully adjusted using a holding pipette to visualize the MII chromatin. The translucent region of the chromatin spindle was aspirated into the pipette with a minimal volume of oocyte cytoplasm. The details have been described previously [19]. The enucleated oocytes were subsequently washed and cultured in CZB medium until use.
Donor nuclei or chromosomes were removed from ES cells by gently aspirating in and out of the injection pipette (outer diameter 10–15 µm and sometimes 20 µm). They were then injected into cytoplasts either using a previously described method [19] or a modified method that we named ‘hole removal’ (after transferring the ES cell nucleus into an oocyte, the ‘hole structure’ formed by the injection pipette was removed by gently aspirating it using the same pipette in the presence of 5 µg/ml CB). This modification was also suitable for the Piezo-assisted injection method. At least three replicates were done for each type of nucleus.
Oocyte Activation and Embryo Culture
Nuclear transfer embryos were activated as described by Wakayama et al. [20]. They were placed in Ca2+-free CZB containing 10 mM Sr2 for 6 h. Cytochalasin B (5 µg/ml) was also added for embryos reconstructed with small cell nuclei during the activation period. Embryos with visible nuclei were considered as activated. They were transferred into Sr2+-free, CB-free M16 medium, and incubation was continued at 37.5°C, under 5% CO2 in air.
Differential Staining of Inner Cell Mass and Trophectoderm
Blastocysts were incubated in Hepes-CZB medium for 45 min at 37°C in the presence of an anti-bovine trophoblast serum (generous gift from Dr. C. Galli), washed in PBS supplemented with 2% FCS, and incubated in PBS for 15 min at 37°C with the anti-guinea pig complement serum (Sigma). Embryos were washed in PBS-2% FCS at 37°C in the presence of 10 µg/ml propidium iodide (Sigma) to label nuclei in permeabilized cells. They were fixed in 2.5% paraformaldehyde in PBS for 20 min at room temperature, washed in PBS-2% FCS, and then stained for 30 min at 37°C with 2 µM SYTO11 (Molecular Probes, Eugene, OR), a permeant DNA marker that labeled nuclei in all cells. Blastocysts were mounted on slides with the antifading agent Vectashield (Vector) and observed under a confocal laser scanning microscope (LSM 310; Carl Zeiss). For each nuclear staining, optical sections were done in the blastocyst, and bicolor images were generated from two successive sections. This colorimetric method to identify nuclei in the depth of the embryo prevented that a nucleus visualized into two sections was counted two times. Nuclear counting was done with the ImageTool software (see above) using the "count and tag" function. Nuclear transfer embryos were incubated with the anti-trophoblast serum 120 h following oocyte activation. At that time, one blastocyst out of four was already hatched (n = 35). This criterion was used to determine the time for analyzing control embryos. Because (C57bl/6xCBA)F1 recipient oocytes were used for NT, (C57bl/6xCBA)F1 embryos were used as controls. They were collected at the one-cell stage and cultured until 115–116 h post-hCG.
We used a test proposed by Dufour and Torres [21], at the level of 1%, to determine whether there was a difference in total cell number between control and NT blastocysts. This test is derived from the Student test procedure.
Embryo Transfer
One- to two-cell stage embryos were transferred into the oviducts of pseudopregnant (C57bl/6xCBA)F1 females that had been mated 1 day before with vasectomized males of proven sterility. Animals were either killed at Embryonic Days 5–8 (E5–8) or at E19. Live pups were delivered by cesarian sections.
Polymerase Chain Reaction Analysis of Genomic DNA
Polymerase chain reaction (PCR) amplification of the microsatellite markers D4Mit204 and D7Mit22 was performed as already described [7]. Sequences for the primer pairs were found on the Mouse Genome Informatics web site (http://www.informatics.jax.org/searches/probe_form.shtml). DNA was extracted from tail tips or cell pellet with the QiaAmp DNA mini kit (Qiagen, Hilden, Germany). Reactions (20 µl) were subjected to 34 cycles of 30 sec 94°C, 1 min 60°C, 2 min 72°C, and products were separated on a 4% agarose gel and visualized after staining with ethidium bromide.
RESULTS
Relationship Between Cell Size and Cell Cycle Stages of the Donor ES Cells
We first analyzed the distribution of the ES donor cells among the cell cycle phases. Flow cytometric analysis (Fig. 1A) and BrdU incorporation (Fig. 1B) showed that about two thirds of ES cells (64%) were in S-phase, while the remaining ones were equally distributed in G2/M (18%) and G1 (18%). Propidium iodide staining after fixation on coverslips revealed that the mean proportion of nuclei in M phase was low (only 6%, data not shown).
|
Nuclear diameter and intensity of DNA labeling (that reflects the DNA content, Fig. 1C) and cell and nucleus diameters (Fig. 1D) were then plotted on a graph and regression curves were calculated. We found a significant correlation between both cell and nucleus diameters (P < 0.01) and nuclear diameter and labeling intensity (P < 0.01). Individual values were more dispersed (around the theoretical curve) in the former than in the latter distributions (respective values for R2, 0.69 and 0.94). Synthesizing nuclei (S phase, BrdU positive), although predominantly found in cells with a mean diameter of 8–10 µm, were also found in all the classes of cell sizes. The small cell population (with a diameter lower than 12 µm) was approximately equally distributed in G1 and S phase, whereas the large cell population (with a diameter above 15 µm) was equally distributed in S phase and G2/M (Fig. 1D).
In order to increase the proportion of donor cells at a given cell cycle stage at the time of NT, we proceeded to cell synchronization treatments using a demecolcin or a thymidine/Hoechst treatment (see Materials and Methods) before flow cytometric analysis or propidium iodide staining (Fig. 2). Upon removal of demecolcin, 91% of the cells were in metaphase (Fig. 2, A and B). When kept up to 3 h in culture after the removal of the drug about 40% of the cells were in G1 (Fig. 2C), the remaining being still in mitosis or degenerating. No cells in S phase could be detected (no BrdU staining, data not shown). A large proportion of the cells were present as doublets corresponding to two daughter cells just exiting M phase and thus in early G1. When kept in culture for a longer period after exposure to demecolcin (after 6 h), cells became rapidly desynchronized and their isolation as G2 cells (4n) was unsuccessful. A thymidine/Hoechst treatment allowed us, however, to enrich the population of cells with 4n (G2) instead of 2n (G1) nuclei (see Fig. 2D). Using this procedure up to 50% of the cells were considered to be in G2 (Fig. 2D), the remaining ones being predominantly in S (data not shown).
|
We concluded that metaphase cells could be easily recognized and isolated from synchronized populations but also from asynchronous ones. The size (diameter) of asynchronous ES cells is a poor predictor of their position in the cell cycle. After a synchronization treatment, however, the small cells isolated were almost exclusively at the G1 phase, whereas large cells were either in G2 or S.
Technical Improvement of Mouse Embryo Reconstruction by Nuclear Injection
The extreme sensitivity of mouse oocytes to invasive manipulative techniques has hampered for several years the use of this species as a model in cloning experiments. A few years ago, Yanagimachi and coworkers succeeded in producing mice through direct injection of somatic [5, 7, 20] and germinal [22] foreign nuclei into recipient oocytes. In a previous study, using the same method, we obtained a lower survival rate of reconstructed embryos, though we obtained fully competent mouse zygotes from adult somatic donor nuclei [18]. Thus, we first focused on the injection technique itself. We found that removing the piece of membrane located at the site of entry of the injection pipette increased the survival rate of the reconstructed oocytes (Table 1). This extra step consisted of aspirating a minimum volume of membrane and cytoplasm at this very site during withdrawal of the injection pipette. This procedure, that we named "the hole removal procedure," makes it possible to use a large injection pipette (up to 20 µm, outside diameter) that is needed to inject metaphase nuclei (see below). All these manipulations could be successfully performed without a Piezo device (our manual method [18]). The latter, however, proved to be helpful for faster manipulations.
|
In Vitro Development of Enucleated MII Oocytes Injected with Metaphase or Interphase Nuclei from ES Cells
Using the above-mentioned morphological criterion to select donor cells, we compared the in vitro development of embryos reconstructed with different types of synchronized cells: metaphase donor cell, small cells (early G1) and large cells (G2 or S). We also used asynchronous cells to estimate the effect of the synchronization treatment on nuclear viability. Results presented in Table 2 show that metaphase nuclei (both from synchronized or asynchronous cells) provided a high development rate of reconstructed embryos into morula (71.8–86.7%) or blastocysts (56.8–70%). Synchronization of donor cells in metaphase slightly reduced the development of the NT embryos to the blastocyst stage (56.8% for synchronized metaphase, versus 70% for asynchronous ones, P < 0.01). However, we found it much easier to pick up metaphases from the synchronized population than the rare metaphases present in an asynchronous culture. For this reason, only a few embryos can be reconstructed using metaphases naturally present when the culture is not treated. We found that most of metaphase-reconstructed embryos normally cleaved to the one- or two-cell stage (more than 95%).
|
This contrasted with the situation observed with interphasic nuclei obtained from cells with small (G1) or large (G2 and S) diameters. For both types of donor cells, up to 28.2% of morulae and up to 14.1% of blastocysts were obtained. Their development was often retarded, resulting in the formation of blastocysts of poor morphological quality. This was true both with the asynchronous and the synchronized cell populations.
Cell Counting in Blastocysts Developed from Embryos Reconstructed with Metaphase Nuclei
Embryos reconstructed with metaphase-synchronized nuclei were collected on Day 5 of culture, and their morphology was examined after differential nuclear staining. One quarter of them (13 of 48, 27%) contained nuclei of irregular size without clear distinction between inner cell mass (ICM) and trophectoderm cells (TE). These embryos obviously showed signs of degeneration and were not included in the cell counting. The others exhibited a visible blastocoelic cavity and ICM, and 26% of them were hatching. To analyze the population of NT blastocysts further, we classified them according to their morphology and sizes. Eighty percent (26 of 35) were fully expanded blastocysts, with no or few signs of nuclear fragmentation (Q1 blastocysts, Fig. 3A, a and a'). The remaining ones were of smaller size with some pycnotic nuclei (Q2 blastocysts, Fig. 3A, b and b'). Cultured blastocysts derived from in vivo-fertilized zygotes were used as controls (Fig. 3A, c and c'). Their mean number of cells was significantly higher (by a factor of 2.7) than that of Q2 blastocysts (Fig. 3B), although the ratio between ICM and total cell number was similar between the three categories (Q1, Q2, and control).
|
In Vivo Development of Embryos Reconstructed with Metaphase ES Nuclei
Embryos reconstructed with metaphase nuclei from ES cells were transferred at the one- or two-cell stage into oviducts of pseudopregnant mice. These foster mice were sacrificed at either E5–8, or at E19 (Table 3). When the population of donor ES cells was used between passages 15 and 19, we found that only one third (32.9%) of the NT embryos implanted. At E19, a still lower number of implantations was observed (20%), and 15% only consisted of a placenta. Finally, two live pups (1.5%) were delivered (Fig. 4A). Microsatellite analysis confirmed that the two cloned mice originated from R1 ES cell line (Fig. 4B). They have now proved to be fertile.
|
|
When synchronized ES cells isolated at passage 20–25 were used as a source of donor nuclei, the implantation rate dropped significantly (11.5% versus 32.9%), and no live pups could be delivered.
The effect of the passage number on implantation rate led us to examine whether this effect could be detected already during in vitro development of embryos reconstructed with metaphase nuclei. The results showed that the cell passage had no noticeable effect on preimplantation development (Table 4).
|
DISCUSSION
A Mitotic Rather Than an Interphasic Donor Nucleus Is Beneficial for the Preimplantation Development of Mouse NT Embryos
Proliferating mouse ES cells can be successfully used to generate live offspring by injecting them into enucleated recipient oocytes (see Wakayama et al. [7] and our present work). Embryonic stem cells are fast dividing cells, with short G1 and G2 phases [18]. Previous NT experiments with ES cells as donors used the criterion of cell size to select cells in G1 and G2/M [7], considering that small cells are predominant in G1 and large cells in G2 or M. To determine to which extent cell size and/or nuclear size of individual cells reflected their position in the cell cycle at the time of NT, we performed cytometric studies on R1 ES donor cells. We showed that the simple criterion of cell size was not accurate enough to determine the status of ES nuclei with respect to the replication phase. We found that DNA-synthesizing nuclei (S phase) were always present in addition to G1 nuclei or G2 nuclei in the population of small and large cells, respectively. This contamination with S-phase nuclei has been shown to be incompatible with development, because PCC causes pulverization of the replicating chromosomes [11, 14, 23]. By contrast, cells in metaphase can be clearly recognized in a population of ES cells by the presence of condensed chromosomes organized on the spindle.
The development rates we obtained here with interphasic nuclei isolated from small and large cells were similar to those reported by Wakayama and colleagues in terms of morula/blastocyst formation [7]. However, with metaphase nuclei the rate of development up to the blastocyst stage became at least twice higher. Moreover, the best development obtained with metaphase nuclei was similar to that of in vitro-fertilized embryos (unpublished observations). This shows that preventing PCC in MII oocytes is beneficial to early reprogramming of cultured nuclei. Cell cycle stage synchronization obtained by exposure to demecolcin and selective detachment of mitotic cells slightly reduced the developmental potential of nuclei. However, we found that this procedure greatly enhanced the performance of the embryo reconstruction by reducing the time required to isolate metaphases from a homogenous population of nuclei. Moreover, the additional step we introduced in our procedure, which we called "the hole removal procedure," proved to be highly efficient to reduce the risks of cell lysis after embryo reconstruction. The fact that metaphase donor nuclei can lead to high development rate up to the blastocyst stage (more than 80%) in the mouse species had been shown some years ago by Kwon and Kono [24]. In this work the authors had used metaphases obtained from still totipotent donor blastomere (isolated at the four-cell stage) and had adopted a two-step procedure in which the transferred nucleus was submitted to a second round of transfer into an enucleated one-cell-stage recipient egg. It was thus not possible to conclude from this work whether the high rate of blastocysts obtained resulted from the use of metaphase chromatin, to the prolonged exposure of the remodeled nuclei to the cytoplasm of a fertilized egg, or to both procedures. Our data suggest that the introduction of a metaphase chromatin can per se lead to high preimplantation development in the mouse. In the work of Kwon and Kono, however, a high proportion of embryos (more than 50%) developed to term, which was not the case in our work (only 3.1%). This could mean that a prolonged exposure of foreign chromatin to the cytoplasm of an enucleated fertilized egg would only be beneficial at postimplantation stages (see below). Alternatively, this high development rate could simply be due to the fact that donor nuclei were obtained from four-cell-stage blastomeres whose nuclei are still largely in an undifferentiated state. With metaphases from embryonic or differentiated donor cells we have the experimental support to address this issue.
Cell-cycle synchrony between the donor chromatin and the recipient oocyte has been shown in mammals to be critical for the functional reprogramming of a transplanted nucleus [1]. The use of metaphase nuclei transferred into MII-arrested, enucleated oocytes ensures a high degree of synchrony between the two components of a reconstructed embryo. Indeed, our data demonstrate that this synchrony results in a significantly higher development to the blastocyst stage than when a time gap is introduced as is the case with G1 or G2 donor nuclei. The reasons for this beneficial effect remain unclear. Transferring a nucleus as condensed chromosomes is a convenient means of ensuring that nuclear RNA synthesis is already repressed. This has been considered to be particularly important in the mouse where the onset of zygotic activities is required early after the first cleavage [25]. Many transcription factors are dissociated from chromatin at mitosis [26] that could be beneficial for the reprogramming of nuclear activities. However, mitotic repression of transcription and loss of transcription factors from mitotic DNA can occur independently of nucleosomal chromatin condensation [27]. Thus, this functional aspect of the reorganization of chromatin from metaphase donor nuclei should be considered cautiously. Structural changes affecting the organization of foreign chromatin directly exposed to the cytoplasm of MII oocytes are also potentially determinant. Its condensation after introduction into a recipient MII cytoplasm often leads to abnormal metaphase with chromosomes dispersed along the spindle [23]. In contrast, a chromatin already condensed and organized into chromosomes offers an increased mechanical resistance to the forces of the mitotic spindle [28] and should avoid entanglement of the fibers. This seems to be the case with mouse embryos because we observed that the metaphase chromosomes almost always remained organized or rapidly reacquired a metaphase-like organization after their injection into the recipient cytoplasm, this before the formation of the polar body (our unpublished data). A generalization to other species should however be considered with caution. In the bovine species, chromosomal abnormalities are frequently observed just after the transfer of metaphase donor cells, and polar body-like extrusion only occurs in about one half of the reconstructed embryos [29]. In the pig, metaphase karyoplasts can remain condensed for up to 10 h after their fusion with metaphase enucleated oocytes, but polar body-like extrusion does not apparently occur, resulting probably in the production of tetraploid embryos [30]. This difference between species that could stress differences in microtubule organization needs to be documented further. In cattle, for instance, as in many other mammals, the functional reproducing centrosome is normally paternally inherited, whereas in mice, multiple microtubule organizing centers with no centriole control microtubule assembly [31, 32]. In this respect, the fate of the duplicated centrosome following transfer of metaphase nuclei is currently under investigation in our laboratory.
Use of Metaphase Nuclei Provides a Situation where Only Postimplantation Defects Are Observed
By eliminating extensive structural modifications like nuclear envelope breakdown, PCC, and de novo formation of a spindle, we improved the rate of blastocyst formation. About three quarters of the blastocyst population was apparently normal, in terms of morphology and allocation of cells between ICM and TE. The mean number of cells at hatching was lower than in control in vivo-fertilized embryos. This could be a direct consequence of the manipulation required for injecting a foreign nucleus because it has been shown that zona drilling allowed hatching of blastocysts with a lower number of cells than in nonmanipulated controls [33]. However, we observed that the hatching of an NT blastocyst occurred about 1 day later than controls. Whether some restriction points exist during or after the cleavage period that might explain this delay is under investigation.
Despite the high implantation rate obtained with metaphase-reconstructed embryos, their fetal development is rapidly compromised, and our preliminary observations reveal that most of the fetuses died at the egg cylinder stage. Intriguingly enough, this high mortality precisely occurs at the stage at which de novo methylation takes place and is essential for gene expression [34]. This would indicate that global alterations that occur during the remodeling of the exogenous chromatin might persist along the preimplantation stages affecting the establishment of methylation-dependent patterns of gene expression. In our work all the embryos were transferred at the one- or two-cell stage to minimize the superimposition of in vitro-induced epigenetic effects on development [35, 36]. Deregulation in the expression of some specific genes acting as a key modulator of lineage commitment such as the transcription factor oct-4 [37] should also be considered.
Cell Passage Effect on Implantation Rate
In our work, a significant drop in implantation was observed when donor cells were used beyond passage 19, this despite the fact that the rate of blastocyst formation was not altered. Such an effect was apparently not observed by Wakayama and colleagues [7], who used an ES line of the same genetic background (R1) up to passage 32. However, in their work the number of passages after subcloning was apparently lower than 19. It is known in ES technology that a high number of passages decrease the germline transmission in ES cell chimeras. This phenomenon has been partly related to a progressive loss of euploidy, resulting in less than 50% of cells with normal karyotype after passage 20 [38]. This explains why ES cells are generally used at a lower number of passages. However, karyotype analysis of our donor R1 ES cells at passages 20 and 28 revealed that about 70% of nuclei were still euploid and did not provide evidence of an increase in the rate of aneuploidy between these two passages (data not shown). Our observation recalls previous studies on completely ES cell-derived mice generated by aggregation with tetraploid embryos [15]. According to these authors, epigenetic changes accumulate during the culture of ES cells and become critical after a given number of passages (around 14 in their study). If our observation was to be confirmed this would mean that obtaining mutant mice in a single generation through NT would be submitted to the same constraints in terms of culture conditions of donor cells as those of the germline ES chimera technology.
In conclusion our results show that NT of metaphase donor nuclei into enucleated mouse oocytes leads to an experimental situation in which a high rate of preimplantation development is followed by a rapid and dramatic occurrence of fetal abnormalities. In other words, the in vivo developmental potential of cloned mouse blastocysts becomes rapidly compromised although their morphology is normal. Cloning from cultured somatic donor cells has been considered since the birth of Dolly as a promising approach for therapeutic applications in humans through the generation of pluripotent stem cells from cloned blastocysts [39]. This has been already demonstrated by two different groups [40, 41], although on a limited number of NT-derived ES lines. This nevertheless suggests that ICM of NT blastocysts are pluripotent, after their derivation in culture. In this respect, the high rate of fetal mortality we observed in vivo points to a deregulation in the establishment of extraembryonic lineages and/or in gradients of expression of some morphogenetic genes.
ACKNOWLEDGMENTS
We thank Dr. A. Nagy for the gift of R1 ES cells and Dr. C. Galli for the gift of anti-trophoblast antibody. We are grateful to Pr. F. Jouneau for his help in the statistical analyses.
FOOTNOTES
First decision: 5 January 2001.
1 This work was supported by the French Ministère de la Recherche, Direction de la Recherche, through an Action Concertée Incitative grant on Development and Integrative Physiology, and also by an Advanced Research Program between China and France (PRA BOO-004). Q.Z. is a recipient of a fellowship from INRA and is supported by the National Natural Science Foundation of China (ref. 9802801). 
2 Correspondence: FAX: 33 1 34 65 26 77; renard{at}jouy.inra.fr
Accepted: March 16, 2001.
Received: December 6, 2000.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  Y. Jincho, Y. Sotomaru, M. Kawahara, Y. Ono, H. Ogawa, Y. Obata, and T. Kono Identification of Genes Aberrantly Expressed in Mouse Embryonic Stem Cell-Cloned Blastocysts Biol Reprod, April 1, 2008; 78(4): 568 - 576. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Yu, C. Ding, E. Wang, X. Chen, X. Li, C. Zhao, Y. Fan, L. Wang, N. Beaujean, Q. Zhou, et al. Piezo-assisted nuclear transfer affects cloning efficiency and may cause apoptosis Reproduction, May 1, 2007; 133(5): 947 - 954. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ono and T. Kono Irreversible Barrier to the Reprogramming of Donor Cells in Cloning with Mouse Embryos and Embryonic Stem Cells Biol Reprod, August 1, 2006; 75(2): 210 - 216. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Jouneau, Q. Zhou, A. Camus, V. Brochard, L. Maulny, J. Collignon, and J.-P. Renard Developmental abnormalities of NT mouse embryos appear early after implantation. Development, April 1, 2006; 133(8): 1597 - 1607. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. J. Mullins, I. Wilmut, and J. J. Mullins Nuclear transfer in rodents J. Physiol., January 1, 2004; 554(1): 4 - 12. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Gasparrini, S. Gao, A. Ainslie, J. Fletcher, M. McGarry, W.A. Ritchie, A.J. Springbett, E.W. Overstrom, I. Wilmut, and P.A. De Sousa Cloned Mice Derived from Embryonic Stem Cell Karyoplasts and Activated Cytoplasts Prepared by Induced Enucleation Biol Reprod, April 1, 2003; 68(4): 1259 - 1266. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Gao, M. McGarry, T. Ferrier, B. Pallante, B. Gasparrini, J. Fletcher, L. Harkness, P. De Sousa, J. McWhir, and I. Wilmut Effect of Cell Confluence on Production of Cloned Mice Using an Inbred Embryonic Stem Cell Line Biol Reprod, February 1, 2003; 68(2): 595 - 603. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kurosaka, Y. Nagao, N. Minami, M. Yamada, and H. Imai Dependence of DNA Synthesis and In Vitro Development of Bovine Nuclear Transfer Embryos on the Stage of the Cell Cycle of Donor Cells and Recipient Cytoplasts Biol Reprod, August 1, 2002; 67(2): 643 - 647. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
