Abstract. Introduction. RBMOnline - Vol 7. No Reproductive BioMedicine Online; on web 12 August 2003

Transcription

1 RBMOnline - Vol 7. No Reproductive BioMedicine Online; on web 12 August 2003 Article Epigenetic reprogramming in early embryonic development: effects of in-vitro production and somatic nuclear transfer Christine Wrenzycki studied veterinary medicine at the Veterinary University of Hannover in Germany from 1987 to In 1995, she finished her doctoral studies and was promoted to Dr. med. vet. at the Veterinary University of Hannover. In 1997, she obtained the certification as specialist in Reproductive Medicine and in 2001 as specialist in Molecular Genetics and Gene Technology. Dr Wrenzycki s research area is animal biotechnology, especially the regulation of gene expression patterns in preimplantation bovine embryos of different origins. Dr Christine Wrenzycki Christine Wrenzycki 1, Heiner Niemann Department of Biotechnology, Institute for Animal Science (FAL), Neustadt, Germany 1 Correspondence: Wrenzycki@tzv.fal.de Abstract A considerable proportion of offspring, in particular in ruminants and mice, born from nuclear transfer (NT)-derived and invitro-produced (IVP) embryos is affected by multiple abnormalities of which a high birthweight and an extended gestation length are the predominant features; a phenomenon that has been called large offspring syndrome (LOS). The underlying mechanisms are largely unknown at present, but alterations of epigenetic modifications of embryonic and fetal gene expression patterns, primarily caused by alterations in DNA methylation are thought to be involved in this syndrome. In mammals, DNA methylation is essential for the regulation of transcription during development and differentiation. This review summarizes results from studies in which mrna expression patterns from IVP and NT-derived embryos were compared with those of their in-vivo counterparts. Numerous aberrations have been found ranging from suppression of expression to de-novo overexpression or more frequently to a significant up- or down-regulation of a specific gene. These observations emphasize the need for further epigenetic studies during preimplantation embryo development to gain insight into the molecular regulation correlated with an undisturbed embryonic and fetal development. Understanding molecular mechanisms will aid improvements in biotechnologies applied to early embryos in all species, including humans. Keywords: DNA methylation, embryo, epigenetics, in-vitro production, nuclear transfer Introduction Unusually large offspring have been born in ruminants following transfer of embryos that have either been subjected to somatic nuclear transfer (NT) or exposed to an unusual in-vivo or in-vitro environment. This phenomenon, called large offspring syndrome (LOS), is characterized by a great variety of abnormal phenotypes, including significant increases in birthweight, polyhydramnios, hydrops fetalis, altered organ growth, various placental and skeletal defects, immunological defects and increased perinatal death. Unusual phenotypes (for reviews, see Walker et al., 1996; Kruip and Den Daas, 1997; Young et al., 1998; Renard et al., 1999; Niemann and Wrenzycki, 2000; Sinclair et al., 2000) have also been observed in other species, including humans (Ménézo et al., 2000; Olivennes et al., 2001). Epigenetics includes gene-regulating activity that is not directly related to the DNA sequence itself and that can persist through one or more generations. It is primarily based on the methylation of the DNA sequence. DNA methylation is crucially involved in imprinting, X chromosome inactivation, genome stability, silencing of retrotransposons, tissue specific gene expression, and inactivation of genes in cancer (Bird and Wolffe, 1999). In mammalian cells, DNA methylation occurs predominantly at CpG dinucleotides and is catalysed by two important classes of DNA methyltransferases. The DNA methyltransferase 1 (Dnmt1) is a maintenance enzyme that methylates hemimethylated CpG dinucleotides after DNA replication (Bestor, 1992). Dnmt3a and Dnmt3b are required for de-novo methylation in vivo and for establishing new DNA-methylation patterns during development (Okano et al., 1999; Hsieh, 1999). Likewise, the oocyte-specific isoform Dnmt1 (Dnmt1o; Howell et al., 2001) is responsible for maintaining but not establishing maternal imprints. Dnmt3L, a protein that by itself has no 649

2 650 DNMT activity, co-localizes with Dnmt3a and Dnmt3b and is thought to be essential for establishing methylation imprints (see below) in the female germ line (Bourchis et al., 2001; Hata et al., 2002). The most dramatic changes in the methylation level occur during gametogenesis and early embryonic development (Monk et al., 1987; Kafri et al., 1993). Epigenetic reprogramming in gametes and epigenetic regulation of embryogenesis In differentiated somatic cells, genomic methylation patterns are stable and heritable due to the activity of Dnmt1 (Li et al., 1992; Bestor, 2000). During murine development, there are at least two developmental periods in which methylation patterns of the entire genome are reprogrammed (Reik et al., 2001). The first period occurs in primordial germ cells of both sexes and leads to rapid demethylation of imprinted genes (see below) and single copy sequences, followed by de-novo methylation in male and female germ cells several days later (Surani, 1998). The second wave of methylation/demethylation is observed after fertilization. The paternal genome is demethylated actively only hours after fertilization by an unknown mechanism, while the maternal genome, which contains most of the methylation marks associated with imprints, undergoes further de-novo methylation (Mayer et al., 2000; Oswald et al., 2000; Santos et al., 2002). During subsequent cleavage divisions, the maternal genome is demethylated passively because the Dnmt1 is excluded from the nucleus (Howell et al., 2001). This global demethylation event erases the majority of the DNA methylation marks, except those of imprinted loci. It is followed by a denovo methylation surge around the time of implantation, which re-establishes overall genomic methylation levels (Jaenisch, 1997; Bird, 2002). The reprogramming events in the zygote via active paternal demethylation, passive demethylation in early cleavage stages and de-novo methylation thereafter appear to be conserved in mammals, although their relative timing with respect to developmental stage can differ (Dean et al., 2001). In mouse embryos, de-novo methylation occurs in the inner cell mass (ICM) cells of the expanded blastocyst, whereas in bovine embryos de-novo methylation occurs already from the 8- to 16- cell stage onwards (Dean et al., 2001). Imprinting It is estimated that approximately 1% of all genes in mammals are repressed on one of the chromosomes, depending on the parental origin of the gene. Imprinted genes are critically involved in the regulation of fetal growth, development, function of the placenta, and postnatal behaviour. Imprinted genes have been identified in mice, rats, sheep, cattle, horses and humans. However, the number of identified imprinted genes differs substantially between species (Morrison et al., 2001). Aberrant expression patterns of imprinted genes have been found in mice and humans (Moore and Reik, 1996) and were related to specific phenotypes, e.g. Beckwith Wiedemann syndrome (BWS; a congenital overgrowth disorder) in humans. Usually imprinting is associated with allele-specific methylation of DNA sequences, which interacts with modifications in chromatin structure and acetylation of the chromatin-associated histone proteins. In most of the imprinted genes differentially methylated regions have been identified (DMRs; Reik and Walter, 2001). Normally, DNA methylation is associated with gene silencing (Bird and Wolffe, 1999). However, seven of 18 imprinted genes have DMRs that are methylated on the active allele (Reik and Walter, 2001). Methylation of imprinted genes can be either associated with induction or silencing of gene expression, depending on the gene (Barlow, 1997). Any perturbation of the methylation process could result in dysregulation of development at later stages. Disruption of Dnmt1 in mice has resulted in abnormal imprinting, embryonic lethality and greatly reduced levels of DNA methylation (Li et al., 1992; Li and Jaenisch, 1993). However, few imprinted genes lack differential methylation or are even unaffected by disruption of the Dnmt1 gene, and many allele-specific methylations are not conserved throughout development, suggesting that other kinds of epigenetic modifications may also contribute to this phenomenon (Tucker et al., 1996). Another epigenetic modification that modulates transcription is histone acetylation (Grunstein, 1997). A mechanistic link between DNA methylation, histone deacetylation and transcriptional silencing of a specific gene has been suggested (Wade et al., 1999). Therefore, rather than DNA methylation, other chromatin reprogramming processes need to be examined in embryos as well. X chromosome inactivation (X inactivation) In female embryos, one of the two X chromosomes is inactivated during early embryogenesis to make X-linked gene dosages equivalent to the ones in males that possess a single X chromosome (Lyon, 1961). In the mouse, X inactivation is characterized by random inactivation of either one X chromosome in the epiblast (primitive ectoderm) lineage, which gives rise to all fetal tissues of the fetus and the extraembryonic mesoderm. The paternal X chromosome is preferentially inactivated in extra-embryonic lineages (Takagi and Sasaki, 1975). Maintenance of X inactivation in extraembryonic tissues also involves hypoacetylation by histone deacetylases, which have been shown to play a universal role in gene silencing in organisms with or without methylation, i.e. imprinted X inactivation is maintained by a mouse polycomb group gene eed (embryonic ectoderm development; Wang et al., 2001). Aberrations of the imprinted expression pattern of the X-inactive specific transcript (Xist) gene may lead to a complete or partial loss of X inactivation, thereby contributing to some of the phenotypes associated with LOS although the phenomenon has been observed in both sexes (Wilson et al., 1995). Recently, it has been shown that IVP and NT protocols affect dosage compensation of X-linked genes in female and male bovine embryos (Wrenzycki et al., 2002). Epigenetic reprogramming and gene expression in cloned embryos Although successful nuclear transfer has been achieved in several mammalian species, the rate of success of obtaining live young is only 1 5%. Successful cloning requires the reprogramming of the genetic activity of the differentiated

3 somatic donor nucleus to an undifferentiated state to permit the temporal and spatial orchestration of the gene expression pattern correlated with regular embryonic and fetal development. In an effort to unravel effects of somatic nuclear transfer on transcriptional activities in cloned embryos, the relative abundance of eight developmentally important gene transcripts has recently been determined in reconstructed blastocysts derived from various modifications of the nuclear transfer protocol and compared the findings with those for invitro- and in-vivo-derived embryos (Wrenzycki et al., 2001a). The genes analysed were Dc II, E-cadherin, Glut-1, Hsp70.1, interferon τ, Mash2, IGF-2R, and Dnmt1 employing either fusion before activation (FBA) or activation and fusion simultaneously (AFS). Hsp mrnas could not be found in NT embryos from either protocol, whereas Hsp transcripts were detected from in-vitro- produced embryos that were analysed in parallel. The relative abundance of interferon τ transcripts was significantly increased in embryos derived from AFS and IVP compared with the FBA treatment. The use of either G 0 or G 1 donor cells significantly reduced the relative amounts of Dnmt transcripts and significantly increased the relative abundance of Mash2 compared with IVP embryos. In addition, the interferon τ transcript levels were significantly elevated in NT-derived blastocysts employing G 1 donor cells for NT compared with IVP embryos and those generated from G 0 cells. Similar results were obtained when donor cells from passages 5/6 versus 8 were compared in nuclear transfer. These data show that modifications of the nuclear transfer protocol are associated with distinct alterations in the expression patterns of the resulting embryos. In one study, an aberrant expression pattern in NT-derived embryos was found with respect to stress susceptibility, trophoblastic function and DNA methylation during preimplantation development (Wrenzycki et al., 2001b). Nuclear transfer also affected expression of X-chromosome linked genes. While the relative abundance of the gene for glucose-6-phosphate dehydrogenase (G6PD) and phosphoglycerate kinase (PGK) were similar in in-vitro- and in-vivo-produced and NT-derived embryos, the level for the X- inactive specific transcript (Xist) was significantly elevated in those embryos derived from adult donor cells versus those derived from fetal donor cells (Wrenzycki et al., 2002). Epigenetic changes such as loss of imprinting and/or alterations of the DNA methylation status may affect the ability of nuclei to initiate and sustain normal development. It has been postulated that these epigenetic changes of mrna expression in preimplantation embryos affect gene expression during fetal development and thus the phenotype of fetuses and/or neonates. In cloned bovine embryos, disturbed epigenetic reprogramming has been demonstrated. Highly aberrant methylation patterns were determined in various genomic regions of cloned bovine embryos (Dean et al., 2001). Frequently, cloned blastocysts resembled donor cells in their overall genomic methylation status (Kang et al., 2001a,b). Additionally, it has been shown that various epigenetic modifications occur among different genomic loci during ageing of bovine fibroblasts, often used for NT (Kang et al., 2001c). This was particularly obvious in trophectodermal cells of the blastocysts, which are undermethylated compared with the ICM in several mammalian species. The abnormally high levels of DNA methylation in the trophectodermal lineage, which constitutes an integral part of the placenta, could be responsible for the observed placental abnormalities in cloned pregnancies. This hypothesis is supported by the finding that the mrna expression of genes that are exclusively expressed in the trophectoderm was heavily affected in the early embryo already by the cloning procedure (Wrenzycki et al., 2001a). In contrast, typical demethylation processes were identified in repetitive sequences of the cloned donor genome during cleavage in porcine embryos, the patterns of which were similar to the ones detected in their fertilized counterparts indicating that species-specific differences may exist in modifying the epigenetic status of cloned donor genomes (pig: Kang et al., 2001d; mouse: Boiani et al., 2002). A recent comparative analysis of Dnmt1, Dnmt3a, Dnmt3b, and Xist mrna expression patterns in donor cells, in-vitromatured oocytes, 1-cell and 8-cell embryos (Figure 1) derived from different nuclear transfer protocols revealed that the Dnmt mrnas could not be detected in the donor cells, whereas expression of these transcripts was determined in matured oocytes and 1-cell embryos (Figure 1). At the 8-cell stage, a significant decrease in Dnmt1 transcripts was found compared with the 1-cell counterparts. Xist mrna levels were detected in the donor cells, were not detectable in matured oocytes, and again not detected in 1-cell embryos. At the 8-cell stage, a significant increase was seen in the experimental groups FBA (fusion before activation) and AFS (activation and fusion simultaneously), with the exception of 8-cell embryos, which have been generated employing pre-activated cytoplasts (PA) as recipient cells. At the blastocyt stage, a significant increase of Xist (females only) and Dnmt1 transcripts was determined in in-vitro-produced and NT-derived embryos compared with their in-vivo-generated counterparts. However, no differences were observed for Dnmt3a and Dnmt3b transcripts in embryos of in-vivo or in-vitro origin (Figure 1; Wrenzycki et al., 2001a), while the increase in the relative abundance of Dnmt3a transcripts in NT-derived embryos compared with in-vivo generated, parthenogenetic and IVP blastocysts reached statistical significance. It has been found that de-novo methylation occurs in bovine embryos at the 8- to 16-cell stage (Dean et al., 2001), consistent with the major activation of the bovine embryonic genome (Telford et al., 1990) suggesting that Dnmt3a or Dnmt3b are already activated at that developmental stage. This hypothesis is not supported by findings showing an increase in the relative abundance of Dnmt3a and Dnmt3b transcripts only from the 8-cell stage to the blastocyst. The 8- to 16-cell embryo in murine development is the stage at which the oocyte form of Dnmt1 (Dnmt1o) enters the nucleus for one cell cycle to get involved in the maintenance of imprinted methylation (Howell et al., 2001). Dnmt1 transcript levels were down-regulated from the 1-cell to the 8-cell stage. Speculatively, there is an analogy with the mouse, and the oocyte form of Dnmt1 is excluded from the nucleus, whereas the somatic form of the enzyme is introduced with the somatic donor nucleus and may remain associated with it. Significant differences in the relative abundance of gene transcripts between in-vivo- and in-vitroor NT-derived blastocysts were seen for Dnmt1, and between NT-derived blastocysts and the other experimental groups for Dnmt3a (Figure 1), suggesting that at the blastocyst stage, maintenance and de-novo DNA methylation are affected by 651

4 652 Figure 1. Effects of in-vitro production (IVP; light grey bars), parthenogenetic activation (Parth; vertical lined bars), different activation protocols for nuclear transfer protocols (NT; various grey bars; PA = pre-activated cytoplasts, FBA = fusion before activation, AFS = activation and fusion simultaneously) and in-vivo collection (horizontal lined bars) on the relative abundance of Xist (a), Dnmt1 (b), Dnmt3a (c), and Dnmt3b (d) transcripts during early bovine embryo development [Mat. ooc. (matured oocyte) = open bars; 1-cell and 8-cell embryos; and blastocysts] including donor cells used for NT. Bars with different superscripts within each developmental stage differ significantly (a,b P 0.05). Note that in (a) the higher relative expression of Xist in blastocysts made it necessary to use a different scale (shown on the right) from that used for earlier cleavage stage embryos (scale on the left). The scale break is indicated between the 8-cell stage and the blastocyst stage.

5 the in-vitro culture and/or the cloning procedure. An abnormal regulation of the expression of the oocyte and somatic DNA methyltransferase 1 has recently also been reported in cloned mouse embryos (Chung et al., 2003). In conclusion, these findings suggest that type and treatment of the donor cells affect the expression pattern of the reconstituted embryos and that aberrations of the subtle regulation of the methylation pattern may be an early causative mechanism for the occurrence of LOS. Epigenetic reprogramming and gene expression in IVP embryos It has been postulated that in-vitro culture of preimplantation embryos is associated with epigenetic modifications in the genome. Extended culture of murine embryos in a deficient medium (Whitten s medium) led to bi-allelic expression of the H19 gene, whereas in an optimised medium (KSOM), the regular pattern, a silent paternal allele, was maintained (Doherty et al., 2000). Analysis of multiple growth-related and imprinted genes in murine embryos cultured in a chemically defined medium (M16) with or without fetal calf serum revealed that culture in the presence of serum affected the regulation of imprinted genes (Khosla et al., 2001). Aberrant methylation patterns have also been observed for a number of growth-related imprinted genes in murine ES cells that were cultured in medium supplemented with serum. These epigenetic alterations were not corrected during postimplantation development and were associated with aberrant imprinted fetal gene expression and phenotypic abnormalities (Dean et al., 1998). In sheep, methylation and expression of the IGF-2R gene, subject to imprinting in the mouse, were reduced in oversized fetuses derived from IVP (in-vitro culture for 5 days in the presence of granulosa cells and/or serum) compared with the control group, thereby altering the subtle regulation of expression of the IGF family (Young et al., 2001). Extensive studies have shown that culture conditions affect the well-orchestrated expression pattern of developmentally important genes in bovine embryos (Wrenzycki et al., 1996, 1999, 2001a,b, 2002). In these studies, substantial evidence was accumulated that bovine embryos produced in vitro differ from their in-vivo-derived counterparts with regard to mrna expression. Transcripts for the gap junction protein connexin 43 (Cx43), which is necessary for the maintenance of compaction, were detected in bovine morulae and blastocysts grown in vivo. In contrast, blastocysts and hatched blastocysts produced in vitro did not show detectable amounts of this gene, independently of whether the culture system was supplemented with serum or BSA (Wrenzycki et al., 1996, 1998). In other studies, the expression pattern of the LIF LIFreceptor system (leukaemia inhibitory factor) was investigated in bovine in-vitro-produced embryos and was compared with that for in-vivo-derived embryos (Eckert and Niemann, 1998). The LIF LIF receptor system is thought to be critically involved in early differentiation and implantation of the mammalian embryo. No LIF ligand transcripts were found in in-vitro-produced morulae and blastocysts, while transcripts for this gene were detected in their in-vitro-derived counterparts. In contrast to in-vivo embryos the LIF-receptor β (LR-β) was not detected in in-vivo-derived morula to blastocyst stages, while the gp130 transcript was found from the morula to the hatched blastocyst stage in both in-vitro and in-vivo-produced embryos. For subsequent experiments, the qualitative RT-PCR assay was improved to a semi-quantitative level allowing determination of the relative abundance of specific genes in single embryos. Employing this sensitive assay, the relative abundance of a set of developmentally important marker genes in embryos cultured in media supplemented with either serum or PVA (polyvinyl alcohol, e.g. chemically defined medium) was investigated (Wrenzycki et al., 1999). The gene transcripts were involved in compaction and cavitation [(Cx43), desmosomal protein plakophilin (Plako), desmosomal glycoproteins, desmocollin II and III (Dc II and III)], energy metabolism [glucose transporter-1, (Glut-1)], RNA processing [polya polymerase, (PolyA)], stress-response [heat shock protein 70.1 (Hsp70.1)] and maternal recognition of pregnancy [interferon τ]. Transcription of the majority of the genes was increased in PVA supplemented/serum-free medium derived embryos compared with their serum-generated counterparts mainly beginning at the 8- to 16-cell stage of development, e.g. the maternal-embryonic transition of genomic activity. The exception from that observation was Hsp70.1, which was significantly increased in embryos derived from serum enriched culture conditions. These findings were interpreted as stress response of the early embryo to suboptimal culture conditions. In a recent study, the effects of two basic culture systems (TCM with 5% CO 2 in air versus SOF with 7% O 2, 88% N 2, 5% CO 2 ) each supplemented with various proteins (serum, BSA or PVA) on the expression of the above mentioned panel of marker genes was investigated. The results revealed that the basic culture system including medium composition and oxygen tension had profound effects on the level of specific transcripts whereas the protein source was less important (Wrenzycki et al., 2001b). More differences with regard to gene expression were found between TCM-generated embryos and their in-vivo counterparts rather than between SOF and in-vivo-derived embryos. Interestingly, no differences with respect to transcriptional activities of the above genes in embryos generated under chemically defined conditions in two different laboratories were detected (Wrenzycki et al., 2001b). These findings provide a major step forward towards the standardization of in-vitro production systems for the generation of bovine embryos of consistent high quality. The effects of two diverging culture systems (TCM supplemented with serum versus SOF supplemented with PVA) on the expression of genes from the insulin-like growth factor family (IGF-1 and -2, IGF-1R and IGF-2R) were also investigated (Yaseen et al., 2001). These genes are known to play an important role in the regulation of embryonic and fetal development. IGF-1 expression was not detected at any developmental stage. Significant differences were found in the timing and magnitude of gene expression levels between the two culture systems (Yaseen et al., 2001). Recently, it was possible for the first time to link mrna expression patterns with in-vivo development of embryos derived from IVP (Lazzari et al., 2002). In-vitro culture of bovine embryos in the presence of high concentrations of serum or BSA significantly increased the relative abundance of the transcripts for Hsp, Cu/Zn-SOD, Glut-3, Glut-4, bfgf, and IGFI-R when 653

6 654 compared with embryos from in-vivo production. Birth weights of calves derived from embryos out of the IVP system were significantly higher than those of calves derived from superovulation or artificial insemination. The results support the hypothesis that the persistence of early deviations in development with respect to gene expression patterns is causally involved in the incidence of LOS, in particular in increased birth weights (Lazzari et al., 2002). In-vitro culture also affects dosage compensation of X-linked genes, i.e. PGK and G6PD. Comparing male and female embryos of in-vivo and in-vitro origin, it has been shown that dosage compensation of PGK was delayed and no dosage compensation has occurred for G6PD in IVP embryos, although Xist transcripts were highly abundant in embryos of in-vitro origin. The absence of X-chromosome inactivation effect of the Xist transcripts in spite of its abundance in bovine IVP embryos suggests that timing and magnitude of Xist expression alteration do not necessarily correlate with a tightly controlled inactivation process (Wrenzycki et al., 2002). Recent studies have revealed that the abnormalities observed in gene expression of preimplantation embryos of different developmental stages are not related to the maturation process of the oocyte. Embryos derived from in-vitro- or in-vivomaturated oocytes did not display significant differences in the relative abundance of Glut-1, Dc II, Plako and E-cadherin transcripts (Knijn et al., 2001; Dieleman et al., 2002). These findings suggest that IVM is not the major step in the IVPprocess with regard to modulation of expression of these genes in the post-fertilization period. However, further studies involving more gene transcripts are needed to confirm this observation. Recently, it was found that gene expression is related to the two lineages in the bovine blastocyst, the inner cell mass and trophectoderm (ICM and TE; Wrenzycki et al., 2003). As demonstrated in cloned mice, a spatial perturbation of gene expression patterns could be another source of an aberrant development (Boiani et al., 2002). Concluding remarks Possible explanations for the abnormal phenotypes frequently observed in offspring born after transfer of IVP and NTderived embryos include reprogramming errors, epigenetic dysregulation incurred during in-vitro culture of embryos prior to transfer into the uterus, and in the case of NT-derived embryos, undefined parameters of the nuclear transfer procedure per se. The data support the hypothesis that the invitro environment and the nuclear transfer procedure itself can have profound effects on mrna expression patterns in preimplantation bovine embryos. The prominent aberrations in IVP or NT-derived embryos show that current in-vitro production systems as well as the cloning technology may lead to persistent alterations of gene expression patterns during embryonic and fetal development and may thus be a causative mechanism involved in the frequent incidence of LOS. These alterations are thought to be induced by epigenetic modulation of gene expression patterns, most likely by changes in the methylation pattern (Niemann and Wrenzycki, 2000). It has been shown that the deregulation of the expression of imprinted and non-imprinted genes lead to perturbed placental and fetal growth, attributed to the cumulative action of many abnormally expressed genes which may have opposing effects on fetal and neonatal growth (Bartolomei and Tilghman, 1997; Reik and Walter, 2001). Remarkably, effects of a single imprinted gene are obviously not enough to induce a significant overgrowth (Humpherys et al., 2001). However, the widespread dysregulation of imprinted and non-imprinted genes in NT- and IVP-derived embryos that survive the postnatal period suggests that mammalian development can tolerate a substantial degree of epigenetic abnormality, and it is likely that epigenetic deviations and associated phenotypes will occur in all species including humans (Jaenisch and Wilmut, 2001). Therefore, more basic research is needed on molecular mechanisms underlying the epigenetic reprogramming during early embryonic development as well as sophisticated studies on the components and component interactions of culture media. One also has to take into account that mrna expression is only the first step in genome activity and that the protein defines the phenotype. The experience gained from studies with farm and laboratory animals suggests that the adoption of new protocols (i.e. intracytoplasmic sperm injection, ICSI) in human assisted reproductive technologies (ART) with an extended in-vitro culture up to the blastocyst stage to allow transfer of a single embryo to avoid multiple pregnancies or the development of new technologies (i.e. oocyte-induced haploidization of somatic cells to overcome age-related oocyte aneuploidy, Tesarik et al., 2001; Lacham-Kaplan et al., 2001; Palermo et al., 2002; Tesarik and Mendozo, 2003), may introduce a new set of problems in human infertility treatment (De Rycke et al., 2002; Powell, 2003) as shown recently (Cox et al., 2002; DeBraun et al., 2003; Gicquel et al., 2003). The bovine embryo could serve as a useful model as there is growing evidence that the bovine is, compared with the mouse, a better model for human preimplantation development; especially with regard to microtubule patterns during fertilization, timing of genome activation, intermediate metabolism and interaction with the culture medium (Navara et al., 1995; Anderiesz et al., 2000; Ménézo et al., 2000; Neuber and Powers, 2000; Niemann and Wrenzycki, 2000). Acknowledgements The research underlying this review work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG Ni 256/12 2 and DFG Ni 256/18 1) and the European Union (EU-Fair CT ). The authors are grateful to Drs Dave Wells and Paul Booth for the supply of cloned embryos that have been analysed to characterize the mrna expression pattern of the Dnmt family and to Doris Herrmann for competent analysis throughout these studies. References Anderiesz C, Ferraretti A-P, Magli C et al Effect of recombinant human gonadotrophins on human, bovine, and murine oocyte meiosis, fertilization and embryonic development in vitro. Human Reproduction 15, Barlow DP 1997 Competition a common motif for the imprinting mechanism? EMBO Journal 16, Bartolomei MS, Tilghman SM 1997 Genomic imprinting in mammals. Annual Review in Genetics 31, Bestor TH 1992 Activation of mammalian DNA methyltransferase by cleavage of a Zn binding regulatory domain. EMBO Journal 11,

7 Bestor TH 2000 The DNA methyltransferases of mammals. Human Molecular Genetics 9, Bird A 2002 DNA methylation pattern and epigenetic memory. Genes and Development 16, Bird AP, Wolffe AP 1999 Methylation-induced repression-belts, braces and chromatin. Cell 24, Boiani M, Eckardt S, Schöler HR et al Oct4 distribution and level in mouse clones: consequences for pluripotency. Genes and Development 16, Bourchis D, Xu GL, Lin CS et al Dnmt3L and the establishment of maternal genomic imprints. Science 294, Cox GF, Burger J, Lip V et al Intracytoplasmic sperm injection may increase the risk of imprinting defects. American Journal of Human Genetics 71, Chung YG, Ratnam S, Chaillet JR et al Abnormal regulation of DNA methyltransferase expression in cloned mouse embryos. Biology of Reproduction, in press. Dean W, Bowden L, Aitchison A et al Altered imprinted gene methylation and expression in completely ES cell-derived mouse fetuses: association with aberrant phenotypes. Development 125, Dean W, Santos F, Stojkovic M et al Conservation of methylation reprogramming in mammalian development: aberrant reprogramming in cloned embryos. Proceedings of the National Academy of Science of the USA 98, DeBraun MR, Niemitz EL, Feinberg AP 2003 Association of in vitro fertilization with Beckwith Wiedemann syndrome and epigenetic alterations of LIT1 and H19. American Journal of Human Genetics 72, De Rycke M, Liebaers I, Van Steirteghem A 2002 Epigenetic risks related to assisted reproductive technologies. Risk analysis and epigenetic inheritance. Human Reproduction 17, Dieleman SJ, Hendriksen PJM, Viuff D et al Effects of in vivo prematuration and in vivo final maturation on developmental capacity and quality of pre-implantation embryos. Theriogenology 57, Doherty AS, Mann MRW, Tremblay KD et al Differential effects of culture on imprinted H19 expression in the preimplantation mouse embryo. Biology of Reproduction 62, Eckert J, Niemann H 1998 mrna expression of leukemia inhibitory factor (LIF) and its receptor subunits glycoprotein 130 (gp 130) and LIF receptor-β (LR-β) in bovine embryos derived in vitro or in vivo. Molecular Human Reproduction 4, Gicquel C, Gaston V, Mandelbaum J et al In vitro fertilization may increase the risk of Beckwith Wiedemann syndrome related to the abnormal imprinting of the KCN1OT gene. American Journal of Human Genetics 72, Grunstein M 1997 Histone acetylation in chromatin structure and transcription. Nature 389, Hata K, Okano M, Lei H, Li E 2002 Dnmt3L co-operates with the Dnmt3 family of de novo DNA methyltransferases to establish maternal imprints in mice. Development 129, Howell CY, Bestor TH, Ding F et al Genomic imprinting disrupted by a maternal effect mutation of the Dnmt1 gene. Cell 104, Hsieh CL 1999 In vivo activity of murine de novo methyltransferases Dnmt3a and Dnmt3b. Molecular and Cellular Biology 19, Humpherys D, Eggan K, Akutsu H et al Epigenetic instability in ES cells and cloned mice. Science 293, Jaenisch R 1997 DNA methylation and imprinting: why bother? Trends in Genetics 13, Jaenisch R, Wilmut I 2001 Developmental biology. Don t clone humans! Science 291, Kafri T, Gao X, Razin A 1993 Mechanistic aspects of genome-wide demethylation in the preimplantation mouse embryo. Proceedings of the National Academy of Sciences of the USA, 90, Kang Y-K, Koo D-B, Park J-S et al. 2001a Aberrant methylation of donor genome in cloned bovine embryos. Nature Genetics 28, Kang Y-K, Koo D-B, Park J-S et al. 2001b Influence of oocyte nuclei on demethylation of donor genome in cloned bovine embryos. FEBS Letters 499, Kang Y-K, Koo D-B, Park J-S et al. 2001c Differential inheritance modes of DNA methylation between euchromatic and heterochromatic DNA sequences in ageing fetal bovine fibroblasts. FEBS Letters 498, 1 5. Kang Y-K, Koo D-B, Park J-S et al. 2001d Typical demethylation events in cloned pig embryos. Clues on species-specific differences in epigenetic reprogramming of a cloned donor genome. Journal of Biological Chemistry 276, Khosla S, Dean W, Brown D et al Culture of preimplantation mouse embryos affects fetal development and the expression of imprinted genes. Biology of Reproduction 64, Knijn HM, Wrenzycki C, Hendriksen PJM et al Effects of oocytes maturation regimen on the relative abundance of gene transcripts in bovine blastocysts derived in vitro or in vivo. Reproduction 124, Kruip ThAM, Den Daas JHG 1997 In vitro produced and cloned embryos: effects on pregnancy, parturition and offspring. Theriogenology 47, Lacham-Kaplan O, Daniels R, Trounson A 2001 Fertilization of mouse oocytes using somatic cells as male germ cell. Reproductive BioMedicine Online 3, Lazzari G, Wrenzycki C, Duchi R et al Cellular and molecular deviations in bovine in vitro produced embryos are related to the large offspring syndrome. Biology of Reproduction 67, Li EBC, Jaenisch R 1993 Role of DNA methylation in genomic imprinting. Nature 366, Li E, Bestor TH, Jaenisch R 1992 Targeted mutation of the DNA methyltransferase gene results in embryonic lethality. Cell 69, Lyon M 1961 Gene action in the X-chromosome of the mouse (Mus musculus L.). Nature 190, Mayer W, Niveleau A, Walter J et al Demethylation of the zygote paternal genome. Nature 403, Ménézo YJR, Veiga A, Dale B 2000 Assisted reproductive technology (ART) in humans: facts and uncertainties. Theriogenology 53, Monk M, Boubelik M, Lehnert S 1987 Temporal and reginal changes in DNA methylation in embryonic, extraembryonic and germ cell lineages during mouse embryo development. Development 99, Moore T, Reik W 1996 Genetic conflict in early development: parental imprinting in normal and abnormal growth. Reviews in Reproduction 1, Morrison IM, Paton CJ, Cleverley SD 2001 The imprinted gene and parent-of-origin effect database. Nucleic Acids Research 29, Navara CS, Simerly C, Schatten G 1995 Imaging mortality during fertilization. Theriogenology 44, Neuber E, Powers RD 2000 Is the mouse a clinically relevant model for human fertilization failures? Human Reproduction 15, Niemann H, Wrenzycki C 2000 Alterations of expression of developmentally important genes in preimplantation bovine embryos by in vitro culture conditions: implications for subsequent development. Theriogenology 53, Okano M, Bell DW, Haber DA et al DNA methyltransferases Dnmt3a and Dnmt3b are essential for de novo methylation and mammalian development. Cell 99, Olivennes F, Mannaerts B, Struijs M et al Perinatal outcome of pregnancy after GnRH antagonist (ganirelix) treatment during ovarian stimulation for convential IVF or ICSI: a preliminary report. Human Reproduction 16, Oswald J, Engemann S, Lane N et al Active demethylation of the paternal genome in the mouse zygote. Current Biology 10, Palermo GD, Takeuchi T, Rosenwaks Z 2002 Oocyte-induced 655

8 656 haploidization. Reproductive BioMedicine Online 4, Powell K 2003 Seeds of doubt. Nature 422, Reik W, Walter J 2001 Genomic imprinting: parental influence on the genome. Nature Review Genetics 2, Reik W, Dean W, Walter J 2001 Epigenetic reprogramming in mammalian development. Science 293, Renard JP, Chastant S, Chesne P et al Lymphoid hypoplasia and somatic cloning. Lancet 353, Santos F, Hendrich B, Reik W et al Dynamic reprogramming of DNA methylation in the early mouse embryo. Developmental Biology 241, Sinclair KD, Young LE, Wilmut I et al In-utero overgrowth in ruminants following embryo culture: lessons from mice and a warning to men. Human Reproduction 15, Surani MA 1998 Imprinting and the initiation of gene silencing in the germ line. Cell 93, Takagi N, Sasaki M 1975 Preferential inactivation of the paternally derived X chromosome in the extraembryonic membranes of mouse. Nature 256, Telford NA, Watson AJ, Schultz GA 1990 Transition from maternal to embryonic control in early mammalian development: a comparison of several species. Molecular Reproduction and Development 26, Tesarik J, Mendoza C 2003 Somatic cell haploidization: an update. Reproductive BioMedicine Online 6, Tesarik J, Nagy ZP, Sousa M et al Fertilizable oocytes reconstructed from patient s somatic cell nuclei and donor ooplasts. Reproductive BioMedicine Online 2, Tucker KL, Beard C, Dausmann J et al Germ-line passage is required for establishment of methylation and expression patterns of imprinted but not of nonimprinted genes. Genes and Development 10, Wade PA, Gegonne A, Jones P et al Mi-2 complex couples DNA methylation to chromatin remodelling and histone deacetylation. Nature Genetics 23, Walker SK, Hartwich KM, Seamark RF 1996 The production of unusually large offspring following embryo manipulation: concepts and challenges. Theriogenology 45, Wang J, Mager J, Chen Y et al Imprinted X inactivation maintained by a mouse Polycomb group gene. Nature Genetics 28, Wilson JM, Williams JD, Bondioli KR et al Comparison of birth weight and growth characteristics of bovine calves produced by nuclear transfer (cloning), embryo transfer and natural mating. Animal Reproduction Science 38, Wrenzycki C, Herrmann D, Carnwath JW et al Expression of the gap junction gene Connexin 43 (Cx43) in preimplantation bovine embryos derived in vitro or in vivo. Journal of Reproduction and Fertility 108, Wrenzycki C, Herrmann D, Carnwath JW et al Expression of RNA from developmentally important genes in preimplantation bovine embryos produced in TCM supplemented with bovine serum albumin (BSA). Journal of Reproduction and Fertility 112, Wrenzycki C, Herrmann D, Carnwath JW et al Alterations in the relative abundance of gene transcripts in preimplantation bovine embryos cultured in medium supplemented with either serum or PVA. Molecular Reproduction and Development 53, Wrenzycki C, Wells D, Herrmann D et al. 2001a Nuclear transfer protocol affects messenger RNA expression patterns in cloned bovine blastocysts. Biology of Reproduction 65, Wrenzycki C, Herrmann D, Keskintepe L et al. 2001b Effects of culture system and protein supplementation on mrna expression in pre-implantation bovine embryos. Human Reproduction 16, Wrenzycki C, Lucas-Hahn A, Herrmann D et al In vitro production and nuclear transfer affect dosage compensation of the X-linked gene transcripts G6PD, PGK and Xist in preimplantation bovine embryos. Biology of Reproduction 66, Wrenzycki C, Hermann D, Wilkening S et al Timing of blastocyst expansion affects spatial expression patterns of genes in bovine blastocysts produced in vitro. Biology of Reproduction 68, in press. Yaseen MA, Wrenzycki C, Herrmann D et al Alterations in the relative abundance of mrna transcripts for insulin-like growth factor ligands (IGF-I and IGF-II) and their receptors (IGF- IR/IGF-IIR) in preimplantation bovine embryos derived from different in vitro systems. Reproduction 122, Young LE, Sinclair KD, Wilmut I 1998 Large offspring syndrome in cattle and sheep. Reviews in Reproduction 3, Young LE, Fernandes K, McEvoy TG et al Epigenetic change in IGF2R is associated with fetal overgrowth after sheep embryo culture. Nature Genetics 27, Paper based on contribution presented at the TWIN Meeting Alpha Andrology 2003 in Antwerp, Belgium, September Received 9 May 2003; refereed 10 June 2003; accepted 8 July 2003.