|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
The Fels Institute for Cancer Research and Molecular Biology4
Department of Biochemistry,5 Temple University School of Medicine, Philadelphia, Pennsylvania 19140
| ABSTRACT |
|---|
|
|
|---|
-amanitin treatment. The effect of the maternal component of the embryonic genome on cytofragmentation constitutes the earliest known effect of the embryonic genome on mammalian embryo phenotype. The results also indicate that clinical procedures seeking to define or manipulate oocyte quality in humans should take into account early effects of the embryonic genome, particularly the maternal genome.
apoptosis, cytofragmentation, embryo, gene regulation, maternal effect, oocyte
| INTRODUCTION |
|---|
|
|
|---|
Under the control of ooplasmic factors, the sperm and egg genomes undergo an extensive series of modifications that collectively transform them into a newly formed embryonic genome that is able to initiate its programmed journey along the path leading to the formation of a new individual. This remarkable ability of the oocyte to direct this process likely constitutes the basis for successful cloning by nuclear transfer. Creating the embryonic genome requires the remodeling of chromatin to establish the capacity for transcriptional control [4, 5]. Chromatin remodeling is especially profound for the paternal genome, as the ooplasm mediates removal of protamines and their replacement with histones. Chromatin remodeling is followed by the acquisition of the ability of the ooplasm to support gene transcription, and ultimately, a major transcriptional activation event [4, 5]. During the critical time period preceding the major genome activation event, the metabolic and homeostatic functions of the early embryo are likewise largely directed by maternally inherited ooplasmic proteins, mRNAs, and other macromolecules. So pervasive is the ooplasm in controlling early embryogenesis that gene transcription before the major genome activation event is often perceived to be of little significance in mammalian embryos.
An increasing body of studies, however, indicates that a significant amount of gene transcription occurs well before the major genome activation event in most species [5]. Curiously, no functional role has yet been established for this early transcription. It has also become clear that the two parental genomes display differences in transcriptional activity before the major genome activation event in mice [6, 7]. The two genomes also display distinct molecular properties, including genomic imprinting and other epigenetic modifications [810]. Early gene transcription, coupled with early differences in pronuclear function, thus establish a possible basis for distinct but complementary effects of the two parental genomes on early embryo phenotype.
One critical, early function for the parental genomes may be to inhibit cell death or maintain blastomere integrity in the embryo. As either gamete in the absence of fertilization would be doomed to degenerate, an obligate consequence of fertilization is the suppression of pathways leading to cell death or degeneration. Under some circumstances, this suppression appears incomplete, leading to apoptotic-like events in the early embryo. These events include cytoplasmic fragmentation or blebbing, DNA fragmentation, and other changes often associated with cellular apoptosis [11, 12]. Cytoplasmic fragmentation can reduce the number of oocytes and embryos available for assisted reproduction protocols in humans. It may also reduce blastomere viability by selectively removing membrane-associated or cortical factors, such as growth factor receptors [12]. Higher pregnancy rates have been reported for human embryos with no fragmentation, or embryos judged to be of higher quality based in part on the degree of cytofragmentation [13, 14]. Other studies, however, found no such correlation [15], or found that certain spatial patterns of cytofragmentation correlated with reduced embryo viability, whereas other patterns of fragmentation did not [12, 16, 17]. Factors that increase the incidence of cytofragmentation can also decrease pregnancy rates [18], while culture conditions that reduce fragmentation are associated with improved blastocyst development [19]. Cytofragmentation in human embryos can be associated with expression of proapoptotic genes [20]. In mouse embryos, apoptosis and cytofragmentation are most commonly observed at the blastocyst stage [21, 22] and are associated with a variety of factors, including oxidative stress, diabetes, and specific gene mutations [2226]. Cytofragmentation also occurs at the two-cell stage in the mouse [27, 28]. This can be associated with expression of molecular markers of apoptosis [27], but is not associated with increased labeling to detect DNA fragmentation (TUNEL labeling) as a marker of apoptosis [28].
Although some studies suggest that cytofragmentation may reflect intrinsically poor oocyte quality [29] or effects of exogenous factors such as maternal age, repeated gonadal hyperstimulation, exogenous gonadotropin dosage, and embryo culture conditions [19, 3036], other studies indicate that the embryonic genome regulates cytofragmentation [27, 37, 38]. In humans, cytofragmentation occurs primarily from the six- to eight-cell stage onward, immediately before or at the time of genome activation [39, 40]. Likewise, cytofragmentation as early as the mid two-cell stage in mice also corresponds to the time just before the major genome activation event [27, 28]. In humans, embryo quality is determined during the zygote stage [40], and the paternal genotype can affect early embryo quality and cytofragmentation [41]. The incidence of cytofragmentation in two-cell-stage mouse embryos is influenced by the combination of maternal and paternal genotypes [27, 28]. Thus, early gene activity may affect the incidence of cytofragmentation.
To better understand how genetic factors may interact with epigenetic or cytoplasmic factors, and to what degree parental genomes may cooperatively regulate early processes such as cytofragmentation, we have undertaken studies in a murine genetic system. In this system, both maternal and paternal genotypes affect the incidence of blastomere fragmentation at the two-cell stage. In this study, we have examined the relative contributions of the maternal genome versus the ooplasm to the maternal genotype effect. We show here that the maternal pronucleus (PN) plays a predominant role in controlling this process, that this effect of the maternal PN is transcription-dependent, and that the ooplasm plays a lesser role in the overall effect of maternal genotype.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Embryos were obtained from adult (8- to 15-wk-old) female mice superovulated by the injection of eCG (Calbiochem, San Diego, CA) and hCG (Sigma-Aldrich, St. Louis, MO) 48 h later. Females received 5 IU of each hormone. All embryos were isolated at the one-cell stage at approximately 19 h post-hCG (hphCG). Adherent cumulus cells were removed by digestion with hyaluronidase (ICN Pharmaceuticals, Costa Mesa, CA, 500 U/mg) at a final concentration of approximately 120 U/ml in M2 medium at room temperature. Embryos were cultured in CZB medium [42] as in our previous study [28]. Embryos were examined at the late one-cell stage to remove any unfertilized eggs, morphologically abnormal embryos, or embryos with any cytoplasmic fragments. This permitted an unambiguous assessment of the frequency of cytofragmentation occurring in healthy, intact embryos during development to the two-cell stage. Cytofragmentation was assessed at 48 hphCG.
Mice of the C57BL/6 genotype were obtained from Harlan Sprague-Dawley (Indianapolis, IN), and those of the C3H/HeJ genotype were from The Jackson Laboratory (Bar Harbor, ME). All studies adhered to procedures consistent with the National Research Council's Guide for the Care and Use of Laboratory Animals.
Pronuclear Transfer
Pronuclear transfers were performed as described [43, 44]. Briefly, embryos were isolated from the oviducts of successfully mated, superovulated females, and then treated with 10 µg/ml cytochalasin B (Sigma-Aldrich) and 0.4 µg/ml demecolcine (Sigma-Aldrich) at least 20 min before manipulation. Pronuclei were then exchanged using sharpened, beveled pipettes (approximately 25 µm inner diameter). Low-energy pulses were delivered to the pipette tip using a piezo pipette driver to facilitate zona penetration. Maternal pronuclei were removed and replaced with maternal pronuclei from other embryos. During maternal PN removal, the second polar body and degenerated first polar bodies were removed. After transfer of a PN-containing karyoplast to the perivitelline space, embryos were returned to culture for approximately 30 min. Electrofusion (900 V/cm, 10 µsec) was performed in fusion medium (275 mM mannitol, 0.05 mM CaCl2, 0.10 mM MgSO4, 0.3% BSA). Only constructs that fused after a single pulse were included in the analysis. After fusion, embryos were examined for the presence of any small cytoplasmic blebs, and such embryos were discarded, leaving just single cell, intact embryos for the analysis (Fig. 1). Embryos were examined again at the late one-cell stage to ensure that no fragmentation had occurred. Embryos were then cultured overnight and examined at approximately 48 hphCG for signs of cyto-fragmentation.
|
Analysis of Cytofragmentation
Embryos were examined by 320x magnification phase contrast microscopy, and scored as either unfragmented, or with fragmentation graded into three categories (Fig. 1). Grade 1 fragmentation was evidenced by one or more small cytofragments smaller than the size of a polar body, and often clustered at one or both poles or in the crevice between blastomeres. Grade 2 fragmentation was evidenced by the presence of many more fragments, with total volume of fragments comprising an equivalent of less than one-half the volume of a blastomere. Grade 3 fragmentation was evidenced by larger fragments, with the total volume of fragments being approximately one-half of one blastomere or greater. The proportion of embryos falling into each category was recorded, and the significance of differences in the incidence of fragmentation between experimental groups was evaluated using a chi-square test of independence.
| RESULTS |
|---|
|
|
|---|
In previous studies, embryos from C57BL/6 females displayed a significantly greater incidence of fragmentation when sired by C3H/HeJ fathers compared with that from C57BL/6 fathers [27, 28]. This result was confirmed here (Table 1). Examples of fragmented and nonfragmented fertilized embryos are shown in Figure 1 (CE). C3H/HeJ homozygotes previously displayed a greater incidence of cytofragmentation than C57BL/6 homozygous embryos, but the difference was not found to be statistically significant [28]. A much larger number of C3H/HeJ homozygous embryos was assayed here. The embryos were scored for fragmentation at a slightly later time during the two-cell stage (48 hphCG) than in the previous study, as a precaution to ensure that cytofragmentation had reached its fullest extent and that cellular fragments were not incorrectly attributed to transient blebs that might arise at cleavage. Also, embryo examinations were conducted at a higher magnification than in the previous study for greater accuracy in grading. With this larger sample size and the other changes in the assay, the comparison here revealed that the incidence of cytofragmentation was significantly greater among the C3H/HeJ homozygous embryos than C57BL/6 homozygotes. Thus, although this difference did not reach the level of statistical significance previously, these data indicate that indeed, embryos from C3H/HeJ females are more prone to cytofragmentation than those from C57BL/6 females. No statistically significant effect of paternal genotype was observed for fertilized embryos from C3H/HeJ females. These data, based on larger sample sizes than our previous study [28], indicate a clear effect of maternal genotype on the incidence of cytofragmentation, and further indicate that the paternal genotype effect may be manifested only where an otherwise low incidence of cytofragmentation occurs, as for example, when using fertilized C57BL/ 6 oocytes.
|
The Cellular Basis for the Maternal Genotype Effect
To examine the basis for the maternal genotype effect, we employed maternal pronuclear transfer to create embryos in which the ooplasm and the maternal genome were of different origins (Fig. 1, FH and Fig. 2). A total of 1822 nuclear transfer embryos were produced for these studies. We observed a consistent and strong effect of maternal PN strain of origin on the rate of cytofragmentation (Fig. 1, I K and Table 2). Specifically, a C3H/HeJ maternal PN consistently produced a high rate of fragmentation and a C57BL/6 maternal PN consistently produced a lower rate of fragmentation.
|
|
The maternal PN studies accomplish the equivalent of nearly total ooplasm exchange, as the maternal PN is separated from its ooplasm during the transfer and introduced into a new ooplasm. Thus, the data in Table 2 also permit the effects of ooplasm to be evaluated. There was no significant difference between constructs made with C57BL/6 and C3H/HeJ recipient ooplasm, even at the level of severity of cytofragmentation. In particular, for constructs of equal resulting genotype, there was no difference in cyto-fragmentation attributable to the ooplasm alone (e.g., compare BCB with CCB constructs, and compare CCC with BCC constructs). A C57BL/6 maternal PN was even able to confer a lower rate of fragmentation on constructs prepared with C3H/HeJ ooplasm (compare CCC with CBC, and CCB with CBB). The absence of any significant effect of recipient ooplasm also excludes the possibility that the small amount of ooplasm transferred with the maternal PN has any discernible effect. These data indicate that the maternal PN is primarily responsible for the maternal genotype effect.
The paternal genotype effect was not evident in these maternal PN transfer studies, even for constructs prepared with C57BL/6 oocytes (i.e., compare BBB and BBC). There was a tendency to observe lower rates of fragmentation when the paternal genotype and the ooplasm genotype were the same, but this difference was not statistically significant in any combination. As stated above, the paternal genotype effect may be most apparent on a background of low overall incidence of cytofragmentation. The overall incidence of fragmentation in maternal PN transfer embryos was increased compared with that of unmanipulated embryos (e.g., 6.91% versus 3.45% for C57BL/6 homozygous constructs and embryos, respectively; Tables 1 and 2). This increased incidence of fragmentation in maternal PN transfer embryos may have obscured a possible effect of the paternal genotype (note: all data in Table 2 are from maternal PN transfer embryos and can be compared). These data thus indicate that, while a paternal genotype effect exists (Table 1), this effect can be obscured by procedures such as pronuclear transfer.
Effect of
-Amanitin on Fragmentation
The effect of the maternal PN on fragmentation indicated that early gene expression from the maternal PN could be required to suppress fragmentation. To test this possibility, we treated an additional 624 maternal pronuclear constructs with
-amanitin to inhibit transcription from 26 hphCG injection onward (Table 2, annotated with b). Because the paternal PN did not alter the outcome for maternal pronuclear transfer embryos, and because C57BL/6 males mated efficiently, C57BL/6 sires were employed for these studies. Treatment of constructs prepared with C57BL/6 oocytes yielded levels of fragmentation intermediate between those observed for untreated BCB and BBB constructs. A significant effect was also observed for
-amanitin treatment of CBB embryos (P < 0.05), which displayed a significantly increased rate of cytofragmentation, indicating that transcription was required for the C57BL/6 maternal PN to suppress fragmentation in a C3H/HeJ ooplasm.
| DISCUSSION |
|---|
|
|
|---|
-amanitin treatment, indicating a role for early gene transcription.
The data exclude a predominant role for the ooplasm in controlling cytofragmentation. If, for example, C57BL/6 ooplasm were to inhibit cytofragmentation, then it would not be expected that the presence of the C3H/HeJ maternal PN would increase cytofragmentation when transferred to a C57BL/6 ooplasm. Conversely, if C3H/HeJ ooplasm were to promote cytofragmentation, then a C57BL/6 maternal PN would not be expected to overcome the effects of the C3H/HeJ recipient ooplasm. The reciprocal independence of cytofragmentation on ooplasm excludes the possibility of a nonspecific effect of culture medium, as effects of culture medium would most likely be exerted via the ooplasm to affect cytofragmentation at this early stage. Because ooplasm of either type could display either high or low fragmentation, depending on the maternal PN present, such a culture effect does not appear evident. Thus, when the results of all combinations shown in Table 2 are considered together with the
-amanitin effect, it is clear that the maternal PN, not the ooplasm or other extranuclear oocyte components, plays the predominant role in controlling the incidence of cytofragmentation.
Although the maternal PN exerts the predominant role in controlling cytofragmentation, a lesser effect of the ooplasm likely exists. One indication of this is the intermediate level of fragmentation observed following
-amanitin treatment of maternal PN transfer constructs made with C57BL/6 oocytes. If a C57BL/6 maternal PN expresses a gene or genes required to suppress cytofragmentation, then the effect of
-amanitin treatment should be similar between constructs prepared with C3H/HeJ and C57BL/6 ooplasm (i.e.,
-amanitin treatment should increase fragmentation in BBB construct to the highest level). Although
-amanitin significantly reduced the ability of C57BL/6 maternal PN to suppress cytofragmentation in C3H/HeJ ooplasm and led to a high level of cytofragmentation, the same treatment led only to an intermediate level of cyto-fragmentation in constructs prepared with C57BL/6 ooplasm. Thus, the requirement for suppression of cytofragmentation may vary with ooplasm strain of origin, with the C57BL/6 ooplasm being endowed with an increased supply of anticytofragmentation gene products relative to C3H/HeJ ooplasm. The expression of anticytofragmentation gene products from the C57BL/6 maternal PN after fertilization may augment this maternal (ooplasmic) supply produced from the C57BL/6 maternal genome during oogenesis. If such an ooplasmic endowment exists, then the high level of cytofragmentation observed in untreated BCB constructs could reflect a combination of lack of expression of anticytofragmentation genes and overexpression of cytofragmentation-promoting genes from the C3H/HeJ maternal PN. The intermediate level of cytofragmentation observed in
-amanitin-treated BCB constructs would thus be expected to render equivalent levels of cytofragmentation in
-amanitin-treated BCB and BBB constructs, which indeed was observed. Thus, the greater predisposition of C3H/HeJ homozygotes compared with C57BL/6 homozygotes to display cytofragmentation (Table 1) likely reflects the combined effects of ooplasm and maternal PN.
To our knowledge, the effect of the maternal PN on cytofragmentation is the earliest documented effect of the embryonic genome on phenotype during development, and the first apparent effect of the gene transcription that occurs before the major genome activation event. Although several previous studies have revealed very early gene transcription activity (reviewed in [5]), none have associated this activity with an effect on immediate embryo phenotype.
Although it has been appreciated for some time that embryonic gene transcription occurs during the one-cell stage in several different mammalian species [5], the function of this early transcriptional activity has not been elucidated. Our data indicate that one function of early zygotic gene transcription may be to suppress the pathway leading to cytofragmentation and cell death, and thus maintain blastomere integrity and promote embryo survival. Consistent with this,
-amanitin treatment of maternal pronuclear transfer embryos modulated the effect of the maternal PN on the incidence of cytofragmentation, as is seen particularly for the CBB constructs. Thus, transcription appears to play a significant role in this early effect of the maternal PN. Also consistent with such an early role for gene transcription, homozygosity for a null mutation in the gene encoding the mouse regulator of G protein signaling (RGS14), which is transcribed in the one-cell embryo, leads to fragmentation at the two-cell stage, a phenotype related to cytokinesis (L.M. McCaffrey, D. Siderovski, and S. D'Souza, University of Western Ontario, personal communication). These observations for RGS14 also support an early role for early embryonic gene transcription in early processes related to cytofragmentation.
While the paternal PN exerts an effect in fertilized embryos obtained from C57BL/6 mothers, the paternal PN does not exert a strong effect in the context of maternal PN transfer. That paternal PN thus behaved differently from maternal PN of the same strain indicates a functional difference between the maternal and paternal genomes. This difference could be imposed on the two genomes either during gametogenesis (i.e., genomic imprinting) or after fertilization [4555]. In either case, this difference between maternal and paternal pronuclear function likely contributes to a need for cooperative interactions between the two genomes during early development. A requirement for correct interactions between complementary, differentially imprinted parental genomes during the first one to two cell cycles could serve as an early selection mechanism to promote survival of high-quality embryos [27]. Genetic diversity may lead to some maternal-paternal genotype combinations that are less compatible than others, a situation that could contribute to speciation during evolution. Inheritance of a wild-type allele of RGS14 from either parent prevents cytofragmentation in heterozygous RGS14 mutant embryos, indicating a lack of imprinting for this gene. Thus, genes other than RGS14 are likely expressed to account for the specific effect of the maternal PN. Cytofragmentation is thus likely to be a consequence of disruption in more than one pathway or early process in the early embryo. Further studies directed toward identifying the relevant genes controlling cytofragmentation will be invaluable, both for understanding the mechanism of cytofragmentation and for understanding the basis for the genetic and epigenetic effects seen here.
The finding that the maternal genome can exert a predominant effect on the incidence of cytofragmentation, independently of ooplasm origin, may be relevant to the clinical evaluation of human oocyte and embryo quality. Blastomere fragmentation is widely attributed to "oocyte quality." Novel approaches for improving oocyte quality include germinal vesicle transfer and ooplasm transfer [56, 57]. Such approaches are founded on the assumption that it is the ooplasm specifically that is responsible for compromised fertility. The results presented here indicate, however, that cytoplasmic factors, including mitochondrial properties, need not always be the only cause of poor embryo quality. Rather, genetic or developmental variation in the programming of the oocyte genome for early functions can also lead to defects such as cytofragmentation. In such cases, ooplasm manipulation procedures could be of limited value in overcoming compromised fertility. Additionally, recent studies of the human have indicated important roles for paternally derived factors in fertility [41, 58]. Further elucidation of the nature of these early genome functions and their consequences should thus be of value to both basic and clinical reproductive biology.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
2 Correspondence: Keith Latham, The Fels Institute for Cancer Research and Molecular Biology, Temple University School of Medicine, 3307 N. Broad Street, Philadelphia, PA 19140. FAX: 215 707 1454; klatham{at}temple.edu ![]()
3 Current address: Advanced Cell Technology, One Innovation Drive, Biotech Three, Worcester, MA 01605 ![]()
Received: 22 August 2004.
First decision: 17 September 2004.
Accepted: 29 October 2004.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
Z. Han, N. R. Mtango, B. G. Patel, C. Sapienza, and K. E. Latham Hybrid Vigor and Transgenerational Epigenetic Effects on Early Mouse Embryo Phenotype Biol Reprod, October 1, 2008; 79(4): 638 - 648. [Abstract] [Full Text] [PDF] |
||||
![]() |
L.-B. Wan, H. Pan, S. Hannenhalli, Y. Cheng, J. Ma, A. Fedoriw, V. Lobanenkov, K. E. Latham, R. M. Schultz, and M. S. Bartolomei Maternal depletion of CTCF reveals multiple functions during oocyte and preimplantation embryo development Development, August 15, 2008; 135(16): 2729 - 2738. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |