《遗传学》课程教学资源（学科前沿）遗传与表观遗传 Epigenetic events in mammalian germ-cell development

REVIEWS Epigenetic events in mammalian germ-cell development reprogramming and beyond Hiroyuki Sasaki* and Yasuhisa Matsui* Abstract The epigenetic profile of germ cells, which is defined by modifications of DNA and chromatin, changes dynamically during their development. Many of the changes are associated with the acquisition of the capacity to support post-fertilization development. Our knowledge of this aspect has greatly increased-for example insights into how the re-establishment of parental imprints is regulated. In addition, an emerging theme from recent studies is that epigenetic modifiers have key roles in germ-cell development itself-for example, epigenetics contributes to the gene- and genomic integrity. Understanding epigenetic regulation in germ cells har of meiosis expression programme that is required for germ-cell development, regulation implications for reproductive engineering technologies and human health Epigenetics refers to a collection of mechanisms and The role of epigenetics in germ cells can be viewed phenomena that define the phenotype of a cell without differently from that in somatic cells During somatic cell affecting the genotype. In molecular terms, it repre- differentiation, cells start in a pluripotent state and make a nts a range of chromatin modifications including series of decisions about their fates, thereby giving rise to DNA methylation, histone modifications, remodelling of a range of cell types. Their gene-expression programme nucleosomes and higher order chromatin reorganiza- become more restricted and potentially locked in by tion. These epigenetic modifications constitute a changes in epigenetic modifications. However, germ cells unique profile in each cell and define cellular identity are different in that, once their fate has been determined by regulating gene expression. Epigenetic profiles are during early development, there is no need for develop- Division of Human Genetics. modifiable during cellular differentiation, but herit- mental decisions to be made. Instead, germ cells have a Department of Integrated ability is an important aspect of epigenetics: it ensures specific fate and go through a series of epigenetic events Genetics, National Institute that daughter cells have the same phenotype as the that are unique to this cell type. The aspects of germ-cell Organization of Information parental cell development that are relevant to these epigenetic events The process of germ-cell development is regulated are the need for a unique gene-expression programme of Genetics. School of Life by both genetic and epigenetic mechanisms2-. Among that is different from somatic cells, the fact that germ Science The Graduate the various cell types that constitute an animal body, cells undergo meiosis and the particular importance of niversity for Advanced udies, 1111 yata germ cells are unique in that they can give rise to a new maintaining genomic integrity in these cells organism. On fertilization, the products of germ-cell In this Review, we discuss dynamic epigenetic Cell Resource Center for development, the oocyte and sperm cell, fuse to form changes that occur during mammalian germ-cell devel a zygote, which is totipotent- it can develop a whole opment. Recent studies have identified a number of epi stitute of Development, new organism. For the zygote to acquire this totipo- genetic modifiers, including DNA methyltransferase ging and Cancer, Tohoku niversity, Seiryo-machi4-I, tency, germ cells and the zygote undergo extensive ep histone-modification enzymes and their regulatory genetic reprogramming. In mammalian germ cells, factors, that have crucial influences on germ-cell devel Correspondence to HSor Y.M. reprogramming also strips existing parental imprints opment. There is also an increasing understanding of epigenetic marks that ensure parental-origin- the mechanisms of the epigenetic reprogramming that hisasakiglab nig acip: pecific monoallelic expression of about a hundred takes place during germ-cell development-for exam do:10.1038/nrg229 mammalian imprinted genes in the next generation ple, how imprints are re-established in the male and Published online and establishes new ones that are different in male female germ cells. Our discussion follows the temporal and female gametes progression of events during germ-cell development, NATURE REVIEWS GENETICS VOLUME 9 FEBRUARY 2008 129 @2008 Nature Publishing Group

Epigenetics refers to a collection of mechanisms and phenomena that define the phenotype of a cell without affecting the genotype1 . In molecular terms, it repre￾sents a range of chromatin modifications including DNA methylation, histone modifications, remodelling of nucleosomes and higher order chromatin reorganiza￾tion. These epigenetic modifications constitute a unique profile in each cell and define cellular identity by regulating gene expression. Epigenetic profiles are modifiable during cellular differentiation, but herit￾ability is an important aspect of epigenetics: it ensures that daughter cells have the same phenotype as the parental cell. The process of germ-cell development is regulated by both genetic and epigenetic mechanisms2–5. Among the various cell types that constitute an animal body, germ cells are unique in that they can give rise to a new organism. On fertilization, the products of germ-cell development, the oocyte and sperm cell, fuse to form a zygote, which is totipotent — it can develop a whole new organism2 . For the zygote to acquire this totipo￾tency, germ cells and the zygote undergo extensive epi￾genetic reprogramming2,3. In mammalian germ cells, reprogramming also strips existing parental imprints — epigenetic marks that ensure parental-origin￾specific monoallelic expression of about a hundred mammalian imprinted genes in the next generation — and establishes new ones that are different in male and female gametes. The role of epigenetics in germ cells can be viewed differently from that in somatic cells. During somatic cell differentiation, cells start in a pluripotent state and make a series of decisions about their fates, thereby giving rise to a range of cell types6 . Their gene-expression programmes become more restricted and potentially locked in by changes in epigenetic modifications. However, germ cells are different in that, once their fate has been determined during early development, there is no need for develop￾mental decisions to be made. Instead, germ cells have a specific fate and go through a series of epigenetic events that are unique to this cell type. The aspects of germ-cell development that are relevant to these epigenetic events are the need for a unique gene-expression programme that is different from somatic cells, the fact that germ cells undergo meiosis and the particular importance of maintaining genomic integrity in these cells. In this Review, we discuss dynamic epigenetic changes that occur during mammalian germ-cell devel￾opment. Recent studies have identified a number of epi￾genetic modifiers, including DNA methyltransferases, histone-modification enzymes and their regulatory factors, that have crucial influences on germ-cell devel￾opment. There is also an increasing understanding of the mechanisms of the epigenetic reprogramming that takes place during germ-cell development — for exam￾ple, how imprints are re-established in the male and female germ cells. Our discussion follows the temporal progression of events during germ-cell development, *Division of Human Genetics, Department of Integrated Genetics, National Institute of Genetics, Research Organization of Information and Systems & Department of Genetics, School of Life Science, The Graduate University for Advanced Studies, 1111 Yata, Mishima 411‑8540, Japan. ‡Cell Resource Center for Biomedical Research, Institute of Development, Aging and Cancer, Tohoku University, Seiryo-machi 4‑1, Sendai 980‑8575, Japan. Correspondence to H.S. or Y.M. e-mails: hisasaki@lab.nig.ac.jp; ymatsui@idac.tohoku.ac.jp doi:10.1038/nrg2295 Published online 16 January 2008 Epigenetic events in mammalian germ-cell development: reprogramming and beyond Hiroyuki Sasaki* and Yasuhisa Matsui‡ Abstract | The epigenetic profile of germ cells, which is defined by modifications of DNA and chromatin, changes dynamically during their development. Many of the changes are associated with the acquisition of the capacity to support post-fertilization development. Our knowledge of this aspect has greatly increased— for example, insights into how the re-establishment of parental imprints is regulated. In addition, an emerging theme from recent studies is that epigenetic modifiers have key roles in germ-cell development itself — for example, epigenetics contributes to the gene￾expression programme that is required for germ-cell development, regulation of meiosis and genomic integrity. Understanding epigenetic regulation in germ cells has implications for reproductive engineering technologies and human health. R E V I E W S nature reviews | genetics volume 9 | february 2008 | 129 © 2008 Nature Publishing Group

REVIEWS Paternal trol of meiosis Imprinting Germ-cell specification Birth/prepuberty/adult ression and activation Suppression of somatic genes Transposon repression PGC founder population PGCs zygote MIll oocyte/egg PGCs settled in gonad Onset of Sex differentiation ully grown oocyte E6.0 Ovulation reprogramming Imprint erasure E17.5 Birth/ prepuberty/adult X-chromosome reactivation Figure 1 Germ cell development and associated epigenetic events in mice. Chronology of mouse germ cell development and the main epigenetic events that occur. PGCs(primordial germ cells) first emerge at embryonic day 7. 25(E7. 25)as a cluster of about 20 cells. Subsequently, they rapidly proliferate with an average doubling time of approximately 16 hours Before they stop dividing at E13.5, their number reaches up to about 26,000 MSCl, meiotic sex-chromosome inactivation DNA methylation A covalent modification that epigenetic changes ar might be important suppression. PGC-like cel dinucelotides in the vertebrate contributions to germ-cell-specific functions at each in Blimpl-null embryos have aberrant expression of the nome.It is catalysed by stage. Understanding the epigenetic changes that take Hox genes6, which are normally repressed in PGCs. This DNA methyltransferases place during germ-cell development has important suggests that BLIMPI is crucial for suppression of the transcription directly by resses implications for animal cloning, assisted reproductive somatic programme, which might ensure that the PGC echnologies and human health precursors and nascent PGCs are restricted to the rganisms such as Caenorhabditis ele specific transcription factors, Germ-cell specification and differentiation and Drosophila melanogaster, this repression involves nd indirectly by recruiting Determination and maintenance of the germ-cell fate. global inhibition of RNA polymerase II (RNAPII) In post-implantation mammalian embryos, a popula- dependent transcription 7-19. In D. melanogaster, the and their associated repressive tion of pluripotent cells in the epiblast gives rise to pri- pole cells that develop into PGCs also have reduced chromatin-remodelling ctivities mondial germ cells(PGCs), the fate of which is specified levels of histone H3 lysine 4 methylation(H3K4me), by tissue interaction during gastrulation. In mice, PGCs a mark that is associated with the permissive(active) Histone modifications first emerge inside the extra-embryonic mesoderm at state, and are enriched in H3K9me, a mark that is translational modifications the posterior end of the primitive streak as a cluster of associated with repression, suggesting a role for epi that alter their interactions cells at embryonic day 7.25(E7. 25)(REFS 7-9)(FIG. 1). genetic modifications in suppressing the somatic pro Before the final specification of PGCs, their precursors gramme. Here, maternally inherited molecules such roteins. In particular, the tails are induced within the proximal epiblast cell popula- as the products of gcl (germ-cell-less)212, pgc(polar of histones H3 and H4 can b tion by signals from the adjacent extra-embryonic ecto- granule component)2.2and nanos u are involved in esidues. Modifications of dermo-ls. A transcriptional regulator, B-lymphocyte the transcriptional quiescence. Therefore, suppres- the tail include methylation, maturation-induced protein 1(BLIMPI, also known as sion of somatic differentiation through transcriptional etylation, phosphorylation PR-domain-containing 1), is expressed specifically in regulation might be an evolutionarily conserved theme the precursor cells as early as E6. 25(REF. 16), and this for germ-cell specification. However, as RNAPII is processes, including gene molecule is essential for PGC specification learly active in nascent PGCs in mice, the molecules The PGCprecursors need to suppress the somaticgene- and mechanisms that regulate the process might differ pression programme, and epigenetic modifications between species www.nature.com/reviews/genetics @2008 Nature Publishing Group

DNA methylation A covalent modification that occurs predominantly at CpG dinucelotides in the vertebrate genome. It is catalysed by DNA methyltransferases and converts cytosines to 5‑methylcytosines. It represses transcription directly by inhibiting the binding of specific transcription factors, and indirectly by recruiting methyl-CpG-binding proteins and their associated repressive chromatin-remodelling activities. Histone modifications Histones undergo post￾translational modifications that alter their interactions with DNA and nuclear proteins. In particular, the tails of histones H3 and H4 can be covalently modified at several residues. Modifications of the tail include methylation, acetylation, phosphorylation and ubiquitylation, and influence several biological processes, including gene expression, DNA repair and chromosome condensation. and we describe the epigenetic changes and their contributions to germ-cell-specific functions at each stage. Understanding the epigenetic changes that take place during germ-cell development has important implications for animal cloning, assisted reproductive technologies and human health. Germ-cell specification and differentiation Determination and maintenance of the germ-cell fate. In post-implantation mammalian embryos, a popula￾tion of pluripotent cells in the epiblast gives rise to pri￾mordial germ cells (PGCs), the fate of which is specified by tissue interaction during gastrulation. In mice, PGCs first emerge inside the extra-embryonic mesoderm at the posterior end of the primitive streak as a cluster of cells at embryonic day 7.25 (E7.25) (REFS 7–9) (FIG. 1). Before the final specification of PGCs, their precursors are induced within the proximal epiblast cell popula￾tion by signals from the adjacent extra-embryonic ecto￾derm10–15. A transcriptional regulator, B‑lymphocyte maturation-induced protein 1 (BLIMP1, also known as PR-domain-containing 1), is expressed specifically in the precursor cells as early as E6.25 (REF. 16), and this molecule is essential for PGC specification. The PGC precursors need to suppress the somatic gene￾expression programme, and epigenetic modifications might be important for this suppression. PGC-like cells in Blimp1-null embryos have aberrant expression of the Hox genes16, which are normally repressed in PGCs. This suggests that BLIMP1 is crucial for suppression of the somatic programme, which might ensure that the PGC precursors and nascent PGCs are restricted to the germ￾cell fate. In organisms such as Caenorhabditis elegans and Drosophila melanogaster, this repression involves global inhibition of RNA polymerase II (RNAPII)- dependent transcription17–19. In D. melanogaster, the pole cells that develop into PGCs also have reduced levels of histone H3 lysine 4 methylation (H3K4me), a mark that is associated with the permissive (active) state, and are enriched in H3K9me, a mark that is associated with repression, suggesting a role for epi￾genetic modifications in suppressing the somatic pro￾gramme20. Here, maternally inherited molecules such as the products of gcl (germ-cell-less)21,22, pgc (polar granule component)23,24 and nanos20,25 are involved in the transcriptional quiescence. Therefore, suppres￾sion of somatic differentiation through transcriptional regulation might be an evolutionarily conserved theme for germ-cell specification. However, as RNAPII is clearly active in nascent PGCs in mice, the molecules and mechanisms that regulate the process might differ between species. Nature Reviews | Genetics Transposon repression Germ-cell specification Suppression of somatic genes E3.5 E6.0 E7.25 E10.5 E12.5 Meiosis Sperm cell Fully grown oocyte MII oocyte/egg Zygote PGC precursors PGC founder population Migration Repression and activation of germ-cell-specific genes Imprint erasure PGCs settled in gonad Sex differentiation X-chromosome reactivation E13.5 E17.5 Birth/prepuberty/adult Onset of meiosis Maternal imprinting Genome-wide deacetylation Spermatocyte E13.5 Birth/prepuberty/adult Paternal imprinting Control of meiosis MSCI Histone– protamine exchange Spermatogonium PGCs Oocyte growth Maturation Ovulation Genome-wide reprogramming Non-growing oocyte Prospermatogonium Figure 1 | Germ cell development and associated epigenetic events in mice. Chronology of mouse germ cell development and the main epigenetic events that occur. PGCs (primordial germ cells) first emerge at embryonic day 7.25 (E7.25) as a cluster of about 20 cells. Subsequently, they rapidly proliferate with an average doubling time of approximately 16 hours. Before they stop dividing at E13.5, their number reaches up to about 26,000. MSCI, meiotic sex-chromosome inactivation. R E V I E W S 130 | february 2008 | volume 9 www.nature.com/reviews/genetics © 2008 Nature Publishing Group

REVIEWS 1In11111111111111111111111 5mec The global changes in repressive marks in migrat aIlllIll11 H3K9me12 ing PGCs might reflect the reprogramming of the PGC genome, which is eventually necessary for the zygote to acquire totipotency. Recent studies of chromatin modi I Transcriptional quiescence fications in embryonic stem cells(ES cells)showed that their pluripotency might be ensured by bivalent chro- E]05 En.5 E125 EB3.5 matin-that is, coincidence of H3K27me and H3K4me Arrival in at genes that encode key developmental transcrip- Entry into meiosis(female) tion factors2.30.Such modifications might temporarily keep the developmental genes poised for activation Figure 2 Epigenetic reprogramming in primordial germ cells(PGCs). Changes in undifferentiated ES cells. The increased level of in epigenetic modifications that occur during the genome-wide reprogramming that H3K27me3 and loss of other repressive marks in PGCs akes place during mouse PGC development Dashed lines indicate that the level of seem to make the PCG genome partly resemble such a the epigenetic modification is lower during these periods than that during the periods chromatin state. Understanding the epigenetic profile shown by solid lines. Sme C, 5-methylcytosine(the product of DNA methylation). of ES cells and germ cells should facilitate research on e exciting possibility of deriving functional gametes from pluripotent cells in culture(BOX 1) How BLIMPI regulates germ-cell specification and suppresses the somatic genes is currently obscure. Regulation of post-migratory PGC-specific genes by Although BLIMPI has a histone-methyltransferase epigenetic mechanisms. Recent studies have shown Nucleosome motif, such an activity has not been detected for this that changes in epigenetic modifications also have he basic structural subunit protein. As discussed below, BLIMPI binds to a histone- important roles in the regulation of post-migratory part for the compactness of a arginine methyltransferase, PRMT5 (protein argini GC-specific genes. Genes such as Ddx4 (DEAD chromosome. Each nucleosome N-methyltransferase 5), to repress premature expression box polypeptide 4, also known as Mvh), Sycp3(syn e of DNA of some germ-cell-specific genes in more advanced aptonemal complex protein 3)and Dazl(deleted in rapped around a histone core, PGCs, and it is possible that epigenetic modification azoospermia-like)are induced after migrating PGCs containing two copies of each might also contribute to the somatic repression role of have entered the genital ridge between E10. 5 and of the core histones: H2A, H2B. BLIMPI Ell.5(FIG. 1). DNA-methylation analysis revealed that, despite the genome-wide decrease in DNA methyl Genome-wide epigenetic changes during early PGC dif- ation after E8.0, the flanking regions of these genes ferentiation. An important recent insight into germ-cell remain methylated at E10.5, but become hypomethyl development comes from findings that unique epige- ated by E13.5 when they are expressed. Furthermore, eages usually all netic and transcriptional changes are seen in further these genes are derepressed in E9.5 embryos that lack lineages and a subset of differentiating, migrating PGCs2 2 When the germ- the maintenance DNA methyltransferase, DNMTI ctraembryonic lineages cell fate is established at E7. 25, levels of genome-wide (REF. 31). The results suggest that DNA methylation Epiblast DNA methylation, H3K9 dimethylation(H3K9me2) regulates the timing of activation of these genes n embryonic lineage that is and H3K27 trimethylation(H3K27 me3)-all marks This demethylation might be part of the second wave derived from the inner cell mass that are associated with transcriptional repression of demethylation that occurs around El1.5(REF. 32) of the blastocyst, which gives are similar to those in surrounding somatic cells. (see below). of the fetus. Subsequently, H3K9me2 starts to be erased at E7.5 A recent study has shown that histone methylation Gastrulation and DNA methylation decreases after about E8. 0, with that is mediated by BLIMPI and its associated histone the level of the former being clearly lower than in the arginine methyltransferase PRMT5 also regulates PGC- movements whereby the cells neighbouring somatic cells by E8.75(REFS 27, 28(FIG. 2). specific genes in post-migratory PGCs. In migrating of the blastula are rearranged Following this initial decrease in these two repressive PGCs, a nuclear-localized BLIMPI-PRMT5 com plan, which consists of the outer marks, the level of H3K27me3, another repressive mark, plex mediates dimethylation of histone H2AR3 and toderm. inner ectoderm and starts to be upregulated after E8. 25 and most PGCs show H4R3. After PGCs have settled in the genital ridge, terstitial mesoderm significantly higher levels of this mark at E9.5(REF 28) the BLIMP1-PRMT5 complex translocates to the FIG. 2). It is likely that this upregulation of H3K27me3 cytoplasm and the levels of H2AR3me2 and H4R3me2 Primitive streak complements the erasure of H3K9me2 to maintain a are diminished. Subsequently, Dhx38(DEAH box structure, which is present as proper repressive chromatin state of the PGC genome. polypeptide 38), a putative target of BLIMPI-PRMT5 relatively free of repressive chromatin between E7.5 and prevents premature expression of this gene Dhx38 E8. 25, which could result in deregulated transcription. encodes a protein that contains a DEAD box, which is gastrulation embryonic cells However, global RNAPII-dependent transcription is an RNA-helicase motif, and its homologue in C. elegans progress through the streak transiently repressed during this period, as demonstrated is involved in germ-cell development. It is interest m ce by the absence of both 5-bromouridine 5'triphosphate ing that H4R3me is associated with activation ofother A type of pluripotent stem cell(BrUTP) incorporation and RNAPII C-terminal- genes 4.35 and H3R8me, which is also mediated by that is derived from the inner domain phosphorylation. As RNAPII is active in PRMT5, can repress transcription6.Together,these cell mass of the early embryo nascent PGCs, as mentioned above, this transcrip- findings suggest that H2AR3me2 and H4R3me2 that of generating virtually all cell tional quiescence seems indifferent to the suppression are mediated by BLIMPI-PRMT5 might have a novel types of the organism. of the somatic program repressive role in PGCs @2008 Nature Publishing Group

Nucleosome The basic structural subunit of chromatin, responsible in part for the compactness of a chromosome. Each nucleosome consists of a sequence of DNA wrapped around a histone core, which is a histone octamer containing two copies of each of the core histones: H2A, H2B, H3 and H4. Pluripotent Able to give rise to a wide range of, but not all, cell lineages (usually all fetal lineages and a subset of extraembryonic lineages). Epiblast An embryonic lineage that is derived from the inner cell mass of the blastocyst, which gives rise to the body of the fetus. Gastrulation A process of cell and tissue movements whereby the cells of the blastula are rearranged to form a three-layered body plan, which consists of the outer ectoderm, inner ectoderm and interstitial mesoderm. Primitive streak A transitory embryonic structure, which is present as a strip of cells, that pre-figures the anterior–posterior axis of the embryo. During gastrulation embryonic cells progress through the streak. Embryonic stem cell A type of pluripotent stem cell that is derived from the inner cell mass of the early embryo. Pluripotent cells are capable of generating virtually all cell types of the organism. How BLIMP1 regulates germ-cell specification and suppresses the somatic genes is currently obscure. Although BLIMP1 has a histone-methyltransferase motif, such an activity has not been detected for this protein. As discussed below, BLIMP1 binds to a histone￾arginine methyltransferase, PRMT5 (protein arginine N-methyltransferase 5), to repress premature expression of some germ-cell-specific genes in more advanced PGCs26, and it is possible that epigenetic modification might also contribute to the somatic repression role of BLIMP1. Genome-wide epigenetic changes during early PGC dif￾ferentiation. An important recent insight into germ-cell development comes from findings that unique epige￾netic and transcriptional changes are seen in further differentiating, migrating PGCs27,28. When the germ￾cell fate is established at E7.25, levels of genome-wide DNA methylation, H3K9 dimethylation (H3K9me2) and H3K27 trimethylation (H3K27me3) — all marks that are associated with transcriptional repression — are similar to those in surrounding somatic cells. Subsequently, H3K9me2 starts to be erased at E7.5 and DNA methylation decreases after about E8.0, with the level of the former being clearly lower than in the neighbouring somatic cells by E8.75 (REFs 27,28) (FIG. 2). Following this initial decrease in these two repressive marks, the level of H3K27me3, another repressive mark, starts to be upregulated after E8.25 and most PGCs show significantly higher levels of this mark at E9.5 (REF. 28) (FIG. 2). It is likely that this upregulation of H3K27me3 complements the erasure of H3K9me2 to maintain a proper repressive chromatin state of the PGC genome. The observation indicates that the PGC genome is relatively free of repressive chromatin between E7.5 and E8.25, which could result in deregulated transcription. However, global RNAPII-dependent transcription is transiently repressed during this period, as demonstrated by the absence of both 5-bromouridine 5′ triphosphate (BrUTP) incorporation and RNAPII C-terminal￾domain phosphorylation28. As RNAPII is active in nascent PGCs, as mentioned above, this transcrip￾tional quiescence seems indifferent to the suppression of the somatic programme. The global changes in repressive marks in migrat￾ing PGCs might reflect the reprogramming of the PGC genome, which is eventually necessary for the zygote to acquire totipotency. Recent studies of chromatin modi￾fications in embryonic stem cells (ES cells) showed that their pluripotency might be ensured by bivalent chro￾matin — that is, coincidence of H3K27me and H3K4me — at genes that encode key developmental transcrip￾tion factors29,30. Such modifications might temporarily keep the developmental genes poised for activation in undifferentiated ES cells. The increased level of H3K27me3 and loss of other repressive marks in PGCs seem to make the PCG genome partly resemble such a chromatin state. Understanding the epigenetic profiles of ES cells and germ cells should facilitate research on the exciting possibility of deriving functional gametes from pluripotent cells in culture (BOX 1). Regulation of post-migratory PGC-specific genes by epigenetic mechanisms. Recent studies have shown that changes in epigenetic modifications also have important roles in the regulation of post-migratory PGC-specific genes. Genes such as Ddx4 (DEAD box polypeptide 4, also known as Mvh), Sycp3 (syn￾aptonemal complex protein 3) and Dazl (deleted in azoospermia-like) are induced after migrating PGCs have entered the genital ridge between E10.5 and E11.5 (FIG. 1). DNA-methylation analysis revealed that, despite the genome-wide decrease in DNA methyl￾ation after E8.0, the flanking regions of these genes remain methylated at E10.5, but become hypomethyl￾ated by E13.5 when they are expressed31. Furthermore, these genes are derepressed in E9.5 embryos that lack the maintenance DNA methyltransferase, DNMT1 (REF. 31). The results suggest that DNA methylation regulates the timing of activation of these genes. This demethylation might be part of the second wave of demethylation that occurs around E11.5 (REF. 32) (see below). A recent study has shown that histone methylation that is mediated by BLIMP1 and its associated histone￾arginine methyltransferase PRMT5 also regulates PGC￾specific genes in post-migratory PGCs26. In migrating PGCs, a nuclear-localized BLIMP1–PRMT5 com￾plex mediates dimethylation of histone H2AR3 and H4R3. After PGCs have settled in the genital ridge, the BLIMP1–PRMT5 complex translocates to the cytoplasm and the levels of H2AR3me2 and H4R3me2 are diminished. Subsequently, Dhx38 (DEAH box polypeptide 38), a putative target of BLIMP1–PRMT5, is upregulated, suggesting that arginine methylation prevents premature expression of this gene26. Dhx38 encodes a protein that contains a DEAD box, which is an RNA-helicase motif, and its homologue in C. elegans is involved in germ-cell development33. It is interest￾ing that H4R3me is associated with activation of other genes34,35 and H3R8me, which is also mediated by PRMT5, can repress transcription36. Together, these findings suggest that H2AR3me2 and H4R3me2 that are mediated by BLIMP1–PRMT5 might have a novel repressive role in PGCs. Nature Reviews | Genetics Migration Arrival in genital ridge Mitotic arrest at G1 (male) Entry into meiosis (female) 5meC H3K9me1/2 H3K27me3 Transcriptional quiescence E7.5 E8.5 E9.5 E10.5 E11.5 E12.5 E13.5 Figure 2 | Epigenetic reprogramming in primordial germ cells (PGCs). Changes in epigenetic modifications that occur during the genome-wide reprogramming that takes place during mouse PGC development. Dashed lines indicate that the level of the epigenetic modification is lower during these periods than that during the periods shown by solid lines. 5meC, 5-methylcytosine (the product of DNA methylation). R E V I E W S nature reviews | genetics volume 9 | february 2008 | 131 © 2008 Nature Publishing Group

REVIEWS Erasure and establishment of parental imprints Imprinting in the male and female germline. Once the An early stage of mammalian Imprint erasure in PGCs. When they arrive at the parental imprints have been erased, new imprints must mbryonic development at genital ridge, which occurs by E11.5, PGCs undergo be re-established according to the gender of the animal. which the first cell lineage extensive epigenetic reprogramming, including the This re-establishment occurs only after sex determina- erasure of parental imprints(FIG. 1). The erasure of tion has been initiated, and male and female germ-cell imprints is reflected by demethylation of the imprinted development diverges to give rise to sperm or oocytes, Spherical structure formed loci, which occurs concomitantly with demethylation respectively. In mice, the gonads of males and females by diferentiating ES cells in of other regions. The erasure occurs at different become morphologically distinguishable by E12.5.In culture, which resembles the imprinted loci at different times between E10.5 and female embryos, germ cells arrest in meiotic prophase E12.5, as shown in a study of cloned embryos that were at around E13. 0, whereas male germ cells enter into produced from PGC nuclei". Since the imprint eras- Gl-phase mitotic arrest at a similar time(FIGS 1,2).A ure at each locus is a rapid process that is completed number of environmental cues, including retinoic acid within one day of development, this might be an active signals from the mesonephroi, determine the timing demethylation process of entry into meiosis by germ cells In somatic cells of female mammals, one of the two G1-arrested male germ cells are called prosper X chromosomes is inactivated so that the dosage of the matogonia or gonocytes(FIG. 1). Paternal methylation genes on this chromosome is equalized between males imprints, which have been identified at just three loci, and females. The inactive X chromosome is reactivated are progressively established in these cells between during female germ-cell development. It had been E14.5 and the newborn stage7-st A germline-specific thought that this reactivation occurs around the time gene-knockout study indicated that the de novo dNA of imprint erasure 39. However, more recent studies methyltransferase, DNMT3A, has a central role in the e novo methylation process of all known paternally at an even earlier stagei. So, X-chromosome reactiva- methylated loci, and another de novo methyltrans tion occurs progressively over a prolonged period and ferase, DNMT3B, is involved only at the Raserfl(ras is completed in post-migratory PGCs. The initiation protein-specific guanine nucleotide-releasing factor 1) of reactivation in migrating PGCs is reminiscent of locus 0. 2. The reason why RasgrfI requires an addi the X reactivation in inner-cell-mass cells of female tional enzyme is unknown, but this could be related to blastocysts, but the mechanisms of these processes are the presence of several retrotransposon sequences at yet to be clarified. this locus(see below ). The establishment of paternal Not all sequences show DNA demethylation in methylation imprints at all loci requires another mem post-migratory PGCs For example, DNA methylation ber of the Dnmt3 family, DNMT3L, which is highly of the intra-cisternal A particle(IAP)retrotransposon expressed in prospermatogonia052-4. This protein has family is only partially reprogrammed. Incomplete no DNA-methyltransferase activity but forms a complex removal of epigenetic marks in the germ line can lead with DNMT3A and/or DNMT3B and stimulates their to epigenetic inheritance from one generation to the activities. next, evidence of which is now accumulating in both The established methylation imprints are then main- factors, it has been suggested that this phenomenon established from neonatal testes, can be maintained could provide a basis for adaptive evolution and/or stably in culture and can give rise to sperm when heritable disease susceptibility without changes in transplanted into testes- possess paternal methyla DNA sequence-[BOX 2). tion imprints, whereas their multipotent derivatives, mGS cells, show partial demethylation at these sites, similar to ES cells. GS cells and mGS cells provide Box 1 Derivation of germ cells from embryonic stem cells invaluable tools for germ-cell study and reproductive Various types of somatic cell, including blood cells and neural cells, have been In the female germline, the initiation of dna obtained from embryonic stem(ES)cells in culture dishes. Recent studies have methylation imprinting occurs after birth, during the revealed that it is also possible to generate gametes from ES cells. Gametes or gamete-like cells were derived when mouse ES cells were cultured under various oocyte growth?. The growing oocytes are at the diplo tene stage of meiotic prophase I, and the de novo methyl differentiation conditions including simple monolayer culture(oocyte), embryoid- ation process is complete by the fully-grown oocyte body formation (sperm)o, embryoid-body formation followed by treatment with retinoic acid (sperm)and retinoic acid induction alone (sperm). In the most ge(FIG. 1). Both DNMT3A and DNMT3L also have successful case, ES-derived sperm cells were able to fertilize oocytes after essential roles in this process 2.9, but DNMT3B seems and survived only up to five months, indicating that reprogramming of the o. K intracytoplasmic injection and support embryonic development to term1.The resultant pups, however, had abnormalities in DNA methylation at imprinted loci Recent studies have started to provide some clues on the mechanism by which the DNMT3A-DNMT3L germ-cell genome was not properly accomplished. When we fully understand the complex recognizes the imprinted loci(and some ret- mechanisms of germ-cell reprogramming, we might be able to derive appropriately rotransposons, see below). A crystallographic analysis reprogrammed, functional gametes from cultured cells, which will allow new of the complexed C-terminal domains of DNMT3A approaches to reproductive engineering, although ethical and safety issues must be and DNMT3L revealed a tetrameric structure with two carefully considered active sites. This structure suggests that DNA regions @2008 Nature Publishing Group

Blastocyst An early stage of mammalian embryonic development at which the first cell lineages become established. Embryoid body Spherical structure formed by differentiating ES cells in culture, which resembles the early embryo. Erasure and establishment of parental imprints Imprint erasure in PGCs. When they arrive at the genital ridge, which occurs by E11.5, PGCs undergo extensive epigenetic reprogramming, including the erasure of parental imprints (FIG. 1). The erasure of imprints is reflected by demethylation of the imprinted loci, which occurs concomitantly with demethylation of other regions32. The erasure occurs at different imprinted loci at different times between E10.5 and E12.5, as shown in a study of cloned embryos that were produced from PGC nuclei37. Since the imprint eras￾ure at each locus is a rapid process that is completed within one day of development, this might be an active demethylation process32. In somatic cells of female mammals, one of the two X chromosomes is inactivated so that the dosage of the genes on this chromosome is equalized between males and females. The inactive X chromosome is reactivated during female germ-cell development. It had been thought that this reactivation occurs around the time of imprint erasure38,39. However, more recent studies showed that it is initiated in the migratory stage40 or at an even earlier stage41. So, X-chromosome reactiva￾tion occurs progressively over a prolonged period and is completed in post-migratory PGCs. The initiation of reactivation in migrating PGCs is reminiscent of the X reactivation in inner-cell-mass cells of female blastocysts, but the mechanisms of these processes are yet to be clarified. Not all sequences show DNA demethylation in post-migratory PGCs. For example, DNA methylation of the intra-cisternal A particle (IAP) retrotransposon family is only partially reprogrammed42. Incomplete removal of epigenetic marks in the germ line can lead to epigenetic inheritance from one generation to the next, evidence of which is now accumulating in both mice and humans43–45. Together with the fact that epigenetic marks can be influenced by environmental factors, it has been suggested that this phenomenon could provide a basis for adaptive evolution and/or heritable disease susceptibility without changes in DNA sequence43–45 (BOX 2). Imprinting in the male and female germline. Once the parental imprints have been erased, new imprints must be re-established according to the gender of the animal. This re-establishment occurs only after sex determina￾tion has been initiated, and male and female germ-cell development diverges to give rise to sperm or oocytes, respectively. In mice, the gonads of males and females become morphologically distinguishable by E12.5. In female embryos, germ cells arrest in meiotic prophase at around E13.0, whereas male germ cells enter into G1-phase mitotic arrest at a similar time (FIGS 1,2). A number of environmental cues, including retinoic acid signals from the mesonephroi46, determine the timing of entry into meiosis by germ cells. G1-arrested male germ cells are called prosper￾matogonia or gonocytes (FIG. 1). Paternal methylation imprints, which have been identified at just three loci, are progressively established in these cells between E14.5 and the newborn stage47–51. A germline-specific gene-knockout study indicated that the de novo DNA methyltransferase, DNMT3A, has a central role in the de novo methylation process of all known paternally methylated loci, and another de novo methyltrans￾ferase, DNMT3B, is involved only at the Rasgrf1 (RAS protein-specific guanine nucleotide-releasing factor 1) locus50,52. The reason why Rasgrf1 requires an addi￾tional enzyme is unknown, but this could be related to the presence of several retrotransposon sequences at this locus (see below). The establishment of paternal methylation imprints at all loci requires another mem￾ber of the Dnmt3 family, DNMT3L, which is highly expressed in prospermatogonia50,52–54. This protein has no DNA-methyltransferase activity but forms a complex with DNMT3A and/or DNMT3B and stimulates their activities. The established methylation imprints are then main￾tained throughout the rest of male germ-cell develop￾ment. Notably, germline stem (GS) cells — which are established from neonatal testes, can be maintained stably in culture and can give rise to sperm when transplanted into testes — possess paternal methyla￾tion imprints, whereas their multipotent derivatives, mGS cells, show partial demethylation at these sites, similar to ES cells55. GS cells and mGS cells provide invaluable tools for germ-cell study and reproductive engineering. In the female germline, the initiation of DNA￾methylation imprinting occurs after birth, during the oocyte growth56,57. The growing oocytes are at the diplo￾tene stage of meiotic prophase I, and the de novo methyl￾ation process is complete by the fully-grown oocyte stage (FIG. 1). Both DNMT3A and DNMT3L also have essential roles in this process52,58,59, but DNMT3B seems dispensable52. Recent studies have started to provide some clues on the mechanism by which the DNMT3A–DNMT3L complex recognizes the imprinted loci (and some ret￾rotransposons, see below). A crystallographic analysis of the complexed C-terminal domains of DNMT3A and DNMT3L revealed a tetrameric structure with two active sites60. This structure suggests that DNA regions Box 1 | Derivation of germ cells from embryonic stem cells Various types of somatic cell, including blood cells and neural cells, have been obtained from embryonic stem (ES) cells in culture dishes. Recent studies have revealed that it is also possible to generate gametes from ES cells108–111. Gametes or gamete-like cells were derived when mouse ES cells were cultured under various differentiation conditions including simple monolayer culture (oocyte)108, embryoid￾body formation (sperm)109, embryoid-body formation followed by treatment with retinoic acid (sperm)110 and retinoic acid induction alone (sperm)111. In the most successful case, ES-derived sperm cells were able to fertilize oocytes after intracytoplasmic injection and support embryonic development to term111. The resultant pups, however, had abnormalities in DNA methylation at imprinted loci and survived only up to five months, indicating that reprogramming of the germ-cell genome was not properly accomplished. When we fully understand the mechanisms of germ-cell reprogramming, we might be able to derive appropriately reprogrammed, functional gametes from cultured cells, which will allow new approaches to reproductive engineering, although ethical and safety issues must be carefully considered. R E V I E W S 132 | february 2008 | volume 9 www.nature.com/reviews/genetics © 2008 Nature Publishing Group

REVIEWS Box 2 I Transgenerational influence of epigenetic alterations in germ cells in mammals because expression levels of the imprinted genes, which include ny important developmental Recent studies have suggested that exposure to chemicals and malnutrition conditions can affect not only the children of the affected individuals, but also genes, are unbalanced in such embryos. When the their grandchildren. This might be attributable to epigenetic alterations that occur imprinted genes are appropriately modified by genetic engineering and developmental manipulation, however, disruptors, the number of spermatogenic cells decreased in the F1 generation. This it was possible to derive adult femalemice with two mater effect was transmittable through the male germ line to subsequent generations, nal genomes and no paternal complement 66(BOX 3) and this was correlated with altered DNA-methylation patterns. In another As the method involves genetically engineered animals example, exposure to methyl-donor supplementation during midgestation and highly complex nuclear-transfer technologies, affected the epigenetic s of fetal germ cells. The mouse A y gene, which its direct application to livestock seems difficult. influences coat colour, is regulated by the DNA methylation status of an intra- cisternal A particle(lAP)retrotransposon inserted at pseudoexon 1A of the gene The methyl-donor supplementation shifted the coat colour of the F2 generation to - Pigenetic silencing of retrotransposons darker one. This suggests that the methyl donor directly or indirectly affected the Only germ cells can transmit genetic information to epigenetic status of Ay in fetal germ cells. Finally, epidemiological studies have the next generation. Therefore, transposons, which indicated that grandchildren of malnourished women show low birth weight4, mobilize in the genome and might cause insertional that grandchildren of men who were well-fed before adolescence have a greater nutations, have to be strictly controlled in these cells. Approximately 40-50% of the mammalian genome is have low cardiovascular mortality In these cases, it is possible that the nutrition occupied by retrotransposons, which mobilize through status caused epigenetic alterations in germ cells, but further studies are needed an RNA intermediate, although many of them are to confirm this possibility. truncated or have accumulated mutations Mammalian retrotransposons include short interspersed nuclear elements(SINEs), long interspersed nuclear elements with a 10-nucleotide CpG interval are a preferred sub- (LINEs) and endogenous retroviruses(long terminal strate,and these are found in many imprinted loci. repeat(LTR)-type retrotransposons However, there are many other regions in the genome One way to control transposable elements is through with the same CpG spacing. Another study showed that epigenetic mechanisms?. In the male germline, al DNMT3L interacts with unmodified H3K4 (REF. 61), retrotransposon sequences undergo de which might restrict targets to regions without H3K4me. methylation during the fetal prospermatogonium Together, both nucleosome modification and CpG spac- stage, concomitant with the de novo methylation ing might provide the basis for the recognition of the of the imprinted loci(FIG. 1). Gene-knockout studies imprinted loci by DNMT3A-DNMT3L(REF. 60)(FIG. 3). in mice showed both common and differential target The differential methylation of the imprinted loci in the specificities of DNMT3A and DNMT3B with respect male and female germlines might require additional to these sequences: SINEBl is mainly methylated factors. In the case of paternally imprinted H19, pro- by DNMT3A, whereas LINEl and IAP are methyl X-chromosome inactivation tection of this locus from de novo DNA methylation in ated by both DNMT3A and DNMT3B (. 50). By The process that occurs in oocytes requires CCCTC-binding factor(CTCE), which contrast, DNMT3L is required for methylation of is known to bind to an unmethylated H19(H19 fetal all these sequences, indicating the crucial function liver mRNA)regulatory region and broad specificity of this factor in de novo dNA the pair of x chromosomes is X-chromosome inactivation in female mice is imprinted methylation((FIG 3) ownregulated to match the in pre-implantation embryos and the extra-embryonic The functional importance of DNA methylation in tissues of post-implantation embryos, and in both cases retrotransposon silencing and germ-cell development that is present in males. The the paternal X chromosome is preferentially inactivated. was first seen in Dnmt3L knockout mice. The LINE and inactivation process involves This imprinted X inactivation depends on both an acti- AP retrotransposons, of which de novo methylation a range of epigenetic vating imprint on the maternal X chromosome and an was prevented by Dnmt3L mutations, were highly tran mechanisms on the inactivated inactivating imprint on the paternal X chromosome. scribed in spermatogonia and spermatocytes' hanges in dNa methylation As mentioned above, the inactive X chromosome is mutations also caused meiotic failure with widespread histone modifications. reactivated in female PGCs, but maintenance of the non-homologous chromosome synapsis and progressive active state of the maternal X chromosome beyond fer- loss of germ cells by the mid-pachytene stage[TABLE 1) Chromosome synapsis tilization requires an imprint. Nuclear transplantation This resulted in complete azoospermia in older ani wo pairs of sister chromatids experiments showed that this maternal imprint is set on mals. The non-homologous synapsis could arise from the X chromosomes during the growth of the oocyte, illegitimate interactions between dispersed retrotrans- chromosomes)that occurs at as with the imprints at autosomal loci. As the maternal poson sequences that were unmasked by demethyla X chromosome from Dnmt3a/Dnmt3b double-mutant tion or from single-or double-strand breaks that were oocytes seems to have normal imprints, the mecha- produced during replicative retrotransposition Argonaute proteins are the nism of this imprinting might be different from that of Recently, a link between a small-RNA pathway and central components of RNA. autosomal imprinting DNA methylation of retrotransposons was discovered silencing mechanisms. They k Parthenogenesis, which is a successful development of (FIG 3).MILL, a member of the Piwi subfamily of Argonaute roteins, Is and, if it were possible in mammals, would provide a way early as E12.5(REF. 70) and interacts with a class of small and the catalytic activity for to produce clones of livestock animals. However, imprint- RNAs called piwi-interacting RNAs(piRNAs) In ing is a major barrier to parthenogenetic development Mili-mutant testis, LINEl and IAP retrotransposons were NATURE REVIEWS GENETICS @2008 Nature Publishing Group

X-chromosome inactivation The process that occurs in female mammals by which gene expression from one of the pair of X chromosomes is downregulated to match the levels of gene expression from the single X chromosome that is present in males. The inactivation process involves a range of epigenetic mechanisms on the inactivated chromosome, including changes in DNA methylation and histone modifications. Chromosome synapsis The association or pairing of the two pairs of sister chromatids (representing homologous chromosomes) that occurs at the start of meiosis. Argonaute proteins Argonaute proteins are the central components of RNA￾silencing mechanisms. They provide the platform for target-mRNA recognition by short guide RNA strands and the catalytic activity for mRNA cleavage. with a 10-nucleotide CpG interval are a preferred sub￾strate, and these are found in many imprinted loci60. However, there are many other regions in the genome with the same CpG spacing. Another study showed that DNMT3L interacts with unmodified H3K4 (Ref. 61), which might restrict targets to regions without H3K4me. Together, both nucleosome modification and CpG spac￾ing might provide the basis for the recognition of the imprinted loci by DNMT3A–DNMT3L (Ref. 60) (FIG. 3). The differential methylation of the imprinted loci in the male and female germlines might require additional factors. In the case of paternally imprinted H19, pro￾tection of this locus from de novo DNA methylation in oocytes requires CCCTC-binding factor (CTCF), which is known to bind to an unmethylated H19 (H19 fetal liver mRNA) regulatory region62. X-chromosome inactivation in female mice is imprinted in pre-implantation embryos and the extra-embryonic tissues of post-implantation embryos, and in both cases the paternal X chromosome is preferentially inactivated. This imprinted X inactivation depends on both an acti￾vating imprint on the maternal X chromosome and an inactivating imprint on the paternal X chromosome. As mentioned above, the inactive X chromosome is reactivated in female PGCs, but maintenance of the active state of the maternal X chromosome beyond fer￾tilization requires an imprint. Nuclear transplantation experiments showed that this maternal imprint is set on the X chromosomes during the growth of the oocyte63, as with the imprints at autosomal loci. As the maternal X chromosome from Dnmt3a/Dnmt3b double-mutant oocytes seems to have normal imprints64, the mecha￾nism of this imprinting might be different from that of autosomal imprinting. Parthenogenesis, which is a successful development of unfertilized eggs, is observed in many vertebrate species and, if it were possible in mammals, would provide a way to produce clones of livestock animals. However, imprint￾ing is a major barrier to parthenogenetic development in mammals because expression levels of the imprinted genes, which include many important developmental genes, are unbalanced in such embryos. When the imprinted genes are appropriately modified by genetic engineering and developmental manipulation, however, it was possible to derive adult female mice with two mater￾nal genomes and no paternal complement65,66 (BOX 3). As the method involves genetically engineered animals and highly complex nuclear-transfer technologies, its direct application to livestock seems difficult. Epigenetic silencing of retrotransposons Only germ cells can transmit genetic information to the next generation. Therefore, transposons, which mobilize in the genome and might cause insertional mutations, have to be strictly controlled in these cells. Approximately 40–50% of the mammalian genome is occupied by retrotransposons, which mobilize through an RNA intermediate, although many of them are truncated or have accumulated mutations. Mammalian retrotransposons include short interspersed nuclear elements (SINEs), long interspersed nuclear elements (LINEs) and endogenous retroviruses (long terminal repeat (LTR)-type retrotransposons). One way to control transposable elements is through epigenetic mechanisms67. In the male germline, all retrotransposon sequences undergo de novo DNA methylation during the fetal prospermatogonium stage50, concomitant with the de novo methylation of the imprinted loci (FIG. 1). Gene-knockout studies in mice showed both common and differential target specificities of DNMT3A and DNMT3B with respect to these sequences: SINEB1 is mainly methylated by DNMT3A, whereas LINE1 and IAP are methyl￾ated by both DNMT3A and DNMT3B (REF. 50). By contrast, DNMT3L is required for methylation of all these sequences50, indicating the crucial function and broad specificity of this factor in de novo DNA methylation (FIG. 3). The functional importance of DNA methylation in retrotransposon silencing and germ-cell development was first seen in Dnmt3L knockout mice. The LINE and IAP retrotransposons, of which de novo methylation was prevented by Dnmt3L mutations, were highly tran￾scribed in spermatogonia and spermatocytes53,54,68. The mutations also caused meiotic failure with widespread non-homologous chromosome synapsis and progressive loss of germ cells by the mid-pachytene stage (TABLE 1). This resulted in complete azoospermia in older ani￾mals. The non-homologous synapsis could arise from illegitimate interactions between dispersed retrotrans￾poson sequences that were unmasked by demethyla￾tion or from single- or double-strand breaks that were produced during replicative retrotransposition53. Recently, a link between a small-RNA pathway and DNA methylation of retrotransposons was discovered69 (FIG. 3). MILI, a member of the Piwi subfamily of Argonaute proteins, is expressed in the male and female gonads as early as E12.5 (REF. 70) and interacts with a class of small RNAs called piwi-interacting RNAs (piRNAs)69,71. In Mili-mutant testis, LINE1 and IAP retrotransposons were Box 2 | Transgenerational influence of epigenetic alterations in germ cells Recent studies have suggested that exposure to chemicals and malnutrition conditions can affect not only the children of the affected individuals, but also their grandchildren. This might be attributable to epigenetic alterations that occur in fetal germ cells. When gestating female rats were exposed to some endocrine disruptors, the number of spermatogenic cells decreased in the F1 generation. This effect was transmittable through the male germ line to subsequent generations, and this was correlated with altered DNA-methylation patterns112. In another example, exposure to methyl-donor supplementation during midgestation affected the epigenetic status of fetal germ cells113. The mouse Avy gene, which influences coat colour, is regulated by the DNA methylation status of an intra￾cisternal A particle (IAP) retrotransposon inserted at pseudoexon 1A of the gene. The methyl-donor supplementation shifted the coat colour of the F2 generation to a darker one. This suggests that the methyl donor directly or indirectly affected the epigenetic status of Avy in fetal germ cells. Finally, epidemiological studies have indicated that grandchildren of malnourished women show low birth weight114, that grandchildren of men who were well-fed before adolescence have a greater risk of mortality from diabetes, and that descendants of men who suffered famine have low cardiovascular mortality115. In these cases, it is possible that the nutrition status caused epigenetic alterations in germ cells, but further studies are needed to confirm this possibility. R E V I E W S nature reviews | genetics volume 9 | february 2008 | 133 © 2008 Nature Publishing Group

REVIEWS H3K4 unmodified/ CpG spacing DNMT3A DNMIBL D|[ piRNA Figure 3 Cross-talk between DNA methylation, histone marks and the pirNa pathway in male germ cells In prospermatogonia of the fetal testis, de novo dNa methylation of paternally imprinted loci and retrotransposons occurs. The DNMT3A (DNA methyltransferase 3A)-DNMT3L complex recognizes its targets, such as imprinted loci, by sensing unmodified histone 3 lysine 4(H3K4 )and CpG spacing at 10-nucleotide intervals. Both DNMT3A- DNMT3L and DNMT3B-DNMT3L complexes methylate retrotransposons but how these sequences are recognized is unknown. MILL, a member of the Argonaute family of proteins, is involved in the piwi-interacting RNA(piRNA) pathway, but recent studies have shown that it also has a role in DNA methylation of retrotransposons. Whether MILI functions in de novo methylation or maintenance methylation is currently unknown. Sme C, 5-methylcytosine (the product of DNA methylation). derepressed, which is consistent with the fact that a large Epigenetic regulation of meiosis proportion of piRNAs from pre-pachytene spermato- Germ cells undergo several types of change in their cytes and prospermatogonia(Kuramochi-Miyagawa, S, epigenetic profile during the various stages of meiosis Watanabe, T, H.S. and Nakano, T, unpublished data) For example, in premeiotic PGCs and spermatogonia, were retrotransposon-derived. As RIWI, a member of unique patterns of histone modifications such as low the rat Piwi family, co-purifies with an RNA-cleavage H3K9me2 levels are observed,, but these patterns ctivity", it is likely that the suppression involves RNA- are dynamically changed upon the initiation of meiosis, guided cleavage of target RNAs. Interestingly, however, especially in male germ cells"(FIG 4) the Mili mutants also showed decreased DNA methyla- tion at LINEl retrotransposons and meiotic defects", The role of histone variants. Male germ cells also express similar to the phenotype of Dnmt3L mutants. At present, an unusually high number of histone variants, including how the MILl-PiRNA complex leads to DNA methyla- TH2A, TH2B, TH3, H3. 3A, H33B and HTl, which tion is an open question. Recent studies in Drosophila are incorporated in the nucleosomes of spermatogonia suggest a link between H3K9me3 and H3K9me2, het- and/or spermatocytes In oocytes, somatic histone Hl erochromatin protein 1(HPI)and Piwi proteinss, and is replaced by an oocyte-specific variant HIFoO dur- it is interesting to ask whether such a link exists in the ing meiotic prophase I (REF 5). The changes in histone mammalian germline. composition and modification might contribute to a Dnmt3L-mutant females can produce fully mature chromatin state that is required for meiosis to take place oocytes(although embryos derived from them die correctly, and for the further maturation of the gametes in utero because of the imprinting defects). However, (see REF. 5 for a detailed review of the role of histone a recent study showed that loss of LSH (lymphoid- variants during meiosis atin specific helicase), a member of the SNF2-helicase fam This is the highly compacted ily of chromatin remodelling proteins, leads to DNA Crucial functions of histone methyltransferases in uxtaposed to centromeres and demethylation and derepression of IAP retrotrans- prophase. The functional significance of histone modifi- contains large blocks of tandem posons in pachytene oocytes". This suggests a role cations during the process of meiosis is particularly clear repeats. It is irreversib for DNA methylation in retrotransposon suppression from gene-knockout studies(TABLE 1). Double mutations silenced and remains so throughout the cell cycle. in oocytes as well. Furthermore, the mutant oocytes of the H3K9 trimethyltransferase genes Suv39h1 and showed demethylation of repeats at pericentric hetero- Suy39h2 cause abnormal meiotic prophase in the testis chromatin as well as incomplete chromosome synapsis, The mutant spermatocytes lack pericentric H3K9me3 ructurally distinct, non- leading to a severe loss of oocytes at an early postnatal just before and during early meiotic prophase, and mei pical versions of the histone stage. In addition to DNA methylation, oocytes also otic chromosomes in these mutant cells undergo non- proteins. They are encoded use small-RNA pathways to silence retrotransposon omologous interactions, predominantly at centromeres (REF. 77 and Watanabe, T and H.S., unpublished data). and delayed synapsis. Meiosis is arrested at the pachytene that is distinct from that of So, both male and female germ cells use multiple stage in these cells, which triggers pronounced apopto the canonical histones. mechanisms for defence against retrotransposons sis of spermatocytes. The results suggest that specific @2008 Nature Publishing Group

Pericentric heterochromatin This is the highly compacted chromatin region that is juxtaposed to centromeres and contains large blocks of tandem repeats. It is irreversibly silenced and remains so throughout the cell cycle. Histone variants Structurally distinct, non￾typical versions of the histone proteins. They are encoded by independent genes and often subject to regulation that is distinct from that of the canonical histones. derepressed69, which is consistent with the fact that a large proportion of piRNAs from pre-pachytene spermato￾cytes69 and prospermatogonia (Kuramochi-Miyagawa, S., Watanabe, T., H.S. and Nakano, T., unpublished data) were retrotransposon-derived. As RIWI, a member of the rat Piwi family, co-purifies with an RNA-cleavage activity72, it is likely that the suppression involves RNA￾guided cleavage of target RNAs. Interestingly, however, the Mili mutants also showed decreased DNA methyla￾tion at LINE1 retrotransposons69 and meiotic defects73, similar to the phenotype of Dnmt3L mutants. At present, how the MILI–piRNA complex leads to DNA methyla￾tion is an open question. Recent studies in Drosophila suggest a link between H3K9me3 and H3K9me2, het￾erochromatin protein 1 (HP1) and Piwi proteins74,75, and it is interesting to ask whether such a link exists in the mammalian germline. Dnmt3L-mutant females can produce fully mature oocytes (although embryos derived from them die in utero because of the imprinting defects)58,59. However, a recent study showed that loss of LSH (lymphoid￾specific helicase), a member of the SNF2-helicase fam￾ily of chromatin remodelling proteins, leads to DNA demethylation and derepression of IAP retrotrans￾posons in pachytene oocytes76. This suggests a role for DNA methylation in retrotransposon suppression in oocytes as well. Furthermore, the mutant oocytes showed demethylation of repeats at pericentric hetero‑ chromatin as well as incomplete chromosome synapsis, leading to a severe loss of oocytes at an early postnatal stage. In addition to DNA methylation, oocytes also use small-RNA pathways to silence retrotransposons (REF. 77 and Watanabe, T. and H.S., unpublished data). So, both male and female germ cells use multiple mechanisms for defence against retrotransposons. Epigenetic regulation of meiosis Germ cells undergo several types of change in their epigenetic profile during the various stages of meiosis. For example, in premeiotic PGCs and spermatogonia, unique patterns of histone modifications such as low H3K9me2 levels are observed27,78, but these patterns are dynamically changed upon the initiation of meiosis, especially in male germ cells79 (FIG. 4). The role of histone variants. Male germ cells also express an unusually high number of histone variants, including TH2A, TH2B, TH3, H3.3A, H3.3B and HT1, which are incorporated in the nucleosomes of spermatogonia and/or spermatocytes5 . In oocytes, somatic histone H1 is replaced by an oocyte-specific variant H1FOO dur￾ing meiotic prophase I (Ref. 5). The changes in histone composition and modification might contribute to a chromatin state that is required for meiosis to take place correctly, and for the further maturation of the gametes (see REF. 5 for a detailed review of the role of histone variants during meiosis). Crucial functions of histone methyltransferases in meiotic prophase. The functional significance of histone modifi￾cations during the process of meiosis is particularly clear from gene-knockout studies (table 1).Double mutations of the H3K9 trimethyltransferase genes Suv39h1 and Suv39h2 cause abnormal meiotic prophase in the testis79. The mutant spermatocytes lack pericentric H3K9me3 just before and during early meiotic prophase, and mei￾otic chromosomes in these mutant cells undergo non￾homologous interactions, predominantly at centromeres, and delayed synapsis. Meiosis is arrested at the pachytene stage in these cells, which triggers pronounced apopto￾sis of spermatocytes. The results suggest that specific Nature Reviews | Genetics Imprinted locus Retrotransposon H3K4 unmodified/ CpG spacing Target-RNA degradation DNMT3A DNMT3L DNMT3L De novo DNA methylation DNMT3A/B Effect on DNA methylation MILI piRNA piRNA biogenesis 5meC 5meC Figure 3 | Cross-talk between DNA methylation, histone marks and the piRNA pathway in male germ cells. In prospermatogonia of the fetal testis, de novo DNA methylation of paternally imprinted loci and retrotransposons occurs. The DNMT3A (DNA methyltransferase 3A)–DNMT3L complex recognizes its targets, such as imprinted loci, by sensing unmodified histone 3 lysine 4 (H3K4) and CpG spacing at 10-nucleotide intervals. Both DNMT3A– DNMT3L and DNMT3B–DNMT3L complexes methylate retrotransposons but how these sequences are recognized is unknown. MILI, a member of the Argonaute family of proteins, is involved in the piwi-interacting RNA (piRNA) pathway, but recent studies have shown that it also has a role in DNA methylation of retrotransposons. Whether MILI functions in de novo methylation or maintenance methylation is currently unknown. 5meC, 5-methylcytosine (the product of DNA methylation). R E V I E W S 134 | february 2008 | volume 9 www.nature.com/reviews/genetics © 2008 Nature Publishing Group

REVIEWS Box 3 Production of bi-maternal mice genes was upregulated in EHMT2-deficient germ uggesting that silencing of these genes by EhMT2 ocyte-derived genomes and no sperm-derived mediated H3K9 mono-and dimethylation might be genome using tricks(see below)to adjust the expression levels of imprinted genes sential for proper synapsis. ne mice were ini eferred to as parthenogenetic but are now called bi-maternal A functional link between H3K4 methylation and mice. The success relied on the fact that the number of loci that are imprinted in the male germ cells(the paternal meiosis-specific gene expression has also been demon- ed loci) is smaller(only three are known)than hat of the maternally imprinted loci(more than ten). In their previous experiments, strated.PRDM9(PR domain-containing 9, also known eggs were reconstituted so that they have one genome from fully grown(imprinted) as Meisetz) is an H3K4 trimethyltransferase that causes ocytes and the other from newbom or rowing(non-imprinted or default transcriptional activation. It is specifically expressed in ocytes, early meiotic germ cells both in the testis and ovary, and (bottom, middle)developed better than the control parthenogenetic embryos(second analysis of Prdmg-null mice revealed that it is necessary from the right) for three embryonic days. The extended development was probably for synapsis and recombination of homologous chromo due to the fact that, in the reconstituted eggs, all loci except the three paternally somes during meiotic prophase. In PRDM9-deficient imprinted loci from the newborn oocytes had appropriate imprints. To adjust the spermatocytes, expression of a number of autosomal expression levels of the three loci, non-growing oocytes were obtained from mice in genes, including those that are specifically expressed which either one (gf2 (insulin-like growth factor 2 ))or two(af2 and Dik/GtlZ)of the in meiotic germ cells, was repressed. The results sug- ternally imprinted loci were genetically engineered so that the genes in these loci are gest that the genes that are activated by the function of propriately expressed. These engineered non-growing oocytes were then used to reconstitute eggs. This resulted in a low(0.5%, one locus) or surprisingly high (30%, PRDM9 might include those that are involved in the fter correction for the expected frequency of the desired genotype, both loci) synapsis of homologous chromosomes and/or recom- production rate of adult bi-maternal mice(second from the left). The results clearly bination. However, it is also possible that H3K4me3 indicate that imprinting is the main barrier to parthenogenesis in mammals and that that is introduced by PRDM9, well as H3K9me and sperm-derived RNAs or proteins are unnecessary for full development. H3K9me2 that are mediated by EHMT2, is important PGC and non-growing oocyte for specific chromosomal structures that are required for events such as the search for homologous chromo- Imprinting somes and synapsis. Further studies are needed to test Default Maternal imprints 10 loci) Genetic Mechanisms of meiotic sex-chromosome inactivation In male germ cells, X and Y chromosomes undergo syn apsis only within the pseudo-autosomal region during prophase I. Their chromatin is subsequently condensed Male to form a macrochromatin body that is called the XY or sex body and the chromosomes become transcription termal imprints (3 loci) ally silent. This form of X-chromosome inactivation is called meiotic sex-chromosome inactivation(MSCI (FIGS 1.4), and might be necessary for synapsis of the Parthenogenetic Androgenetic pseudo-autosomal regions 3. It is also suggested that the XY body might mask asynapsed axes of non-pseudo- autosomal regions to prevent sensing by the pachytene heckpoint machinery and subsequent meiotic arrest MSCI is mediated by some key molecules including the histone H2A variant, H2AX, and its regulatory pro Developmental potential teins. 8. During the leptotene phase of meiosis, H2AX PGC, primordial germ celL. that is phosphorylated at Ser139(YH2AX), which is known to recruit the DNA repair machinery to damaged chromatin is localized at sites of dna double-strand changes in modification of pericentric chromatin by breaks. yH2AX subsequently disappears from autosomes SUV39H might be necessary for the proper progression by pachytene when synapsis is completed. yH2AX accu- of meiotic prophase. Abnormalities of female germ cells mulates in the XY body at the zygotene-pachytene transi have also been observed in these double mutants, but tion, and analysis of H2AX-deficient mice demonstrated the molecular details of these defects are unknown. that this is essential for XY-body formation, MSCI and A recent study showed that H3K9 mono-and synapsis between the sex chromosomes. The functional dimethylation by EHMT2(euchromatic histone-lysine importance and mechanisms of H2AX phosphorylation N-methyltransferase 2, also known as G9a)is also are suggested by studies of additional molecules such essential for early meiotic progression". A germ-cell- as ATR (ataxia telangiectasia and RAD3-related)and specific homozygous mutation of Ehmt2 caused arrest the tumour suppressor BRCAl(breast cancer 1). ATR, of meiosis at the early pachytene stage, and synapsis a member of the Pl3-like kinase family, co-localizes between homologous chromosomes was not properly with yH2AX only on X and Y chromosomes at the formed in either testis or ovary. In the mutant sper- onset of their inactivation. Studies of BRCAl-deficient matocytes, most H3K9me and H3K9me2 signals were spermatocytes indicated that localization of ATR on lost, but H3K9me3 was unaffected. Expression of some XY chromatin and proper formation of the XY body NATURE REVIEWS GENETICS @2008 Nature Publishing Group

changes in modification of pericentric chromatin by SUV39H might be necessary for the proper progression of meiotic prophase. Abnormalities of female germ cells have also been observed in these double mutants79, but the molecular details of these defects are unknown. A recent study showed that H3K9 mono- and dimethylation by EHMT2 (euchromatic histone-lysine N-methyltransferase 2, also known as G9a) is also essential for early meiotic progression80. A germ-cell￾specific homozygous mutation of Ehmt2 caused arrest of meiosis at the early pachytene stage, and synapsis between homologous chromosomes was not properly formed in either testis or ovary. In the mutant sper￾matocytes, most H3K9me and H3K9me2 signals were lost, but H3K9me3 was unaffected. Expression of some genes was upregulated in EHMT2-deficient germ cells, suggesting that silencing of these genes by EHMT2- mediated H3K9 mono- and dimethylation might be essential for proper synapsis. A functional link between H3K4 methylation and meiosis-specific gene expression has also been demon￾strated81. PRDM9 (PR domain-containing 9, also known as Meisetz) is an H3K4 trimethyltransferase that causes transcriptional activation. It is specifically expressed in early meiotic germ cells both in the testis and ovary, and analysis of Prdm9-null mice revealed that it is necessary for synapsis and recombination of homologous chromo￾somes during meiotic prophase. In PRDM9-deficient spermatocytes, expression of a number of autosomal genes, including those that are specifically expressed in meiotic germ cells, was repressed. The results sug￾gest that the genes that are activated by the function of PRDM9 might include those that are involved in the synapsis of homologous chromosomes and/or recom￾bination. However, it is also possible that H3K4me3 that is introduced by PRDM9, as well as H3K9me and H3K9me2 that are mediated by EHMT2, is important for specific chromosomal structures that are required for events such as the search for homologous chromo￾somes and synapsis. Further studies are needed to test this possibility. Mechanisms of meiotic sex-chromosome inactivation. In male germ cells, X and Y chromosomes undergo syn￾apsis only within the pseudo-autosomal region during prophase I. Their chromatin is subsequently condensed to form a macrochromatin body that is called the XY or sex body, and the chromosomes become transcription￾ally silent82. This form of X-chromosome inactivation is called meiotic sex-chromosome inactivation (MSCI) (FIGS 1,4), and might be necessary for synapsis of the pseudo-autosomal regions83. It is also suggested that the XY body might mask asynapsed axes of non-pseudo￾autosomal regions to prevent sensing by the pachytene checkpoint machinery and subsequent meiotic arrest83. MSCI is mediated by some key molecules including the histone H2A variant, H2AX, and its regulatory pro￾teins83,84. During the leptotene phase of meiosis, H2AX that is phosphorylated at Ser139 (γH2AX), which is known to recruit the DNA repair machinery to damaged chromatin, is localized at sites of DNA double-strand breaks. γH2AX subsequently disappears from autosomes by pachytene when synapsis is completed. γH2AX accu￾mulates in the XY body at the zygotene–pachytene transi￾tion, and analysis of H2AX-deficient mice demonstrated that this is essential for XY‑body formation, MSCI and synapsis between the sex chromosomes83. The functional importance and mechanisms of H2AX phosphorylation are suggested by studies of additional molecules such as ATR (ataxia telangiectasia and RAD3-related) and the tumour suppressor BRCA1 (breast cancer 1). ATR, a member of the PI3-like kinase family, co-localizes with γH2AX only on X and Y chromosomes at the onset of their inactivation84. Studies of BRCA1-deficient spermatocytes indicated that localization of ATR on XY chromatin and proper formation of the XY body Nature Reviews | Genetics Female Male PGC and non-growing oocyte Imprinting Imprinting Maternal imprints (> 10 loci) Paternal imprints (3 loci) Genetic engineering Fertilized Bi-maternal Parthenogenetic Androgenetic Developmental potential > > > > Default Default PGC Box 3 | Production of bi-maternal mice Kono et al. produced mice with two oocyte-derived genomes and no sperm-derived genome using tricks (see below) to adjust the expression levels of imprinted genes65,66. The mice were initially referred to as parthenogenetic but are now called bi-maternal mice. The success relied on the fact that the number of loci that are imprinted in the male germ cells (the paternally imprinted loci) is smaller (only three are known) than that of the maternally imprinted loci (more than ten). In their previous experiments, eggs were reconstituted so that they have one genome from fully grown (imprinted) oocytes and the other from newborn or non-growing (non-imprinted or default) oocytes, using a serial nuclear-transfer technology that they devised. These eggs (bottom, middle) developed better than the control parthenogenetic embryos (second from the right) for three embryonic days116. The extended development was probably due to the fact that, in the reconstituted eggs, all loci except the three paternally imprinted loci from the newborn oocytes had appropriate imprints. To adjust the expression levels of the three loci, non-growing oocytes were obtained from mice in which either one (Igf2 (insulin-like growth factor 2)) or two (Igf2 and Dlk/Gtl2) of the paternally imprinted loci were genetically engineered so that the genes in these loci are appropriately expressed. These engineered non-growing oocytes were then used to reconstitute eggs. This resulted in a low (0.5%, one locus)65 or surprisingly high (30%, after correction for the expected frequency of the desired genotype, both loci)66 production rate of adult bi-maternal mice (second from the left). The results clearly indicate that imprinting is the main barrier to parthenogenesis in mammals and that sperm-derived RNAs or proteins are unnecessary for full development. PGC, primordial germ cell. R E V I E W S nature reviews | genetics volume 9 | february 2008 | 135 © 2008 Nature Publishing Group

REVIEWS Table 1 Impaired meiosis caused by defects in epigenetic modifiers Gene Function Mutant phenotype Suv39h1 and H3K9 trimethyltransfera Male) Arrest at mid to late hromosome synapsis: impaired modification of pericentric heterochromatin H3K9 m (Male and female) Arrest at early pachytene and apoptosis: impaired chromosome dimethyitransferase sis: deregulation of target genes H3K4 trimethyltransfera nd female)Arrest at early pachytene and apoptosis: impaired chromo psis and recombination; impaired activation of meiosis-specific genes Dnmt3a De novo DNA methyltransferase ( Male) Arrest at pachytene and apoptosis; imprinting failure 52 Dnmt3L Regulator of DNA methylation (Male) Arrest at pachytene and apoptosis: impaired chromosome synapsis; 53,54.68 imprinting failure; derepression of retrotransposons Chromatin remodelling protein (Female)Arrest and apoptosis at diplotene; impaired chromosome synapsis and of SNF2-helicase family Sami Polycomb group protein (Male)Apoptosis at late pachytene; abnormal chromatin modifications at XY body 86 Piwi family protein; small RNA (Male) Arrest at early pachytene and apoptosis; derepression of retrotransposon ulation Dnmt3a, DNA methyltransferase 3a: Ehmt2, euchromatic histone-lysine N-methyltransferase 2, also known as G9a: Lsh, lymphoid-specific helicase: Prdm9, PR domain-containing 9, also known as Meisetz: Scmhl, sex comb on midleg homologue 1. depends on the function of BRCAl, and that yH2AX that autosomal transgenes carrying an X-inactivation localizes on sex chromosomes in an ATR-dependent centre do not undergo MSCI but can induce imprinted inactivation of the inserted region". Furthermore, it wa The XY body shows characteristic changes in histone shown that the paternal X chromosome is transcribed modifications including deacetylation of histones H3 at zygotic gene activation in female embryos, arguing and H4 and dimethylation of H3K9 during pachytene, against the pre-inactivation modelo. Therefore, the rela consistent with the inactive state. Also, H3K4me2 tionship between MSCI and imprinted X inactivation is and H3K4me3 are over- and under-represented, yet to be clarified. respectively. Interestingly, loss of function of PRDM9 causes a failure of XY-body formation. This sug- Histone deacetylation in maturating oocytes and segre gests that PRDM9 might methylate autosomal loci gation of meiotic chromosomes. In mouse oocytes, his- hat are crucial for XY-body formation to keep them tones H3 and H4 are generally acetylated at prophase I transcriptionally active. of meiosis, but they rapidly become deacetylated at SCMHI(sex comb on midleg homologue 1), which is metaphase I (MI) by histone deacetylases(HDACs) component of Polycomb repressive complex 1(PRC1)(FIG 4). In vitro culture of germinal-vesicle-stage oocytes is involved in histone modifications of the XY body that have been arrested at prophase I in the presence and progression of meiotic prophase 6. In cells that lack of trichostatin A, an inhibitor of HDACs, showed functional SCMHl, meiosis arrests at late pachytene. that neither meiotic maturation, fertilization nor pre Ra normal spermatocytes, PRCl components as well implantation development is affected%2.However,a as H3K27me3 are excluded from the XY body at late closer examination revealed that chromosomes are not pachytene, whereas they are abnormally retained on the properly aligned at the metaphase plate in MIl oocytes, XY body in Scmhl-mutant spermatocytes A failure of and that aneuploidy is frequent in single-cell zygotes H3K9me and H3K9me2 accumulation in the XY body Consequently, about half of embryos derived from the and a failure of phosphorylated RNAPII exclusion were trichostatin-A-treated oocytes died in utero Importanth also observed in the mutants. Therefore, SCMHI is histones remained acetylated in the oocytes of older the xY body ho,. c changes in histone modification in (10-month old)female mice, suggesting that the high modifications, the inactivation of XY chromosomes is in older pregnancies might be due to inadequate histone not affected by the Scmhl mutation, and it is currently deacetylation. Histone deacetylation might be involved sted in specific chromosomal structure that is important fo It has been suggested that MSCI might have a role chromosome segregation. imprinted X-chromosome inactivation. A model has been proposed in which imprinted X inactivation results Epigenetic changes during gamete maturation from inheritance of an X chromosome that has been After meiosis, both male and female germ cells undergo Aneuploidy pre-inactivated by MSCI from the paternal germline?. final developmental changes, at the level of both their esence of an abnormal number of chromosome This model is supported by the observation that the morphology he epigenome, to allow them to carry X chromosome is persistently repressed in postmeiotic out their roles in fertilization and the initial stages of of trisomies, an extra copy of haploid cells and retains repressive modifications such as zygotic development In haploid round spermatids, glo- H3K9me2 (REFS 88, 89). However, other studies showed bal nuclear remodelling occurs, although some histone @2008 Nature Publishing Group

Aneuploidy Presence of an abnormal number of chromosomes. For example, in the case of trisomies, an extra copy of a chromosome is present. depends on the function of BRCA1, and that γH2AX localizes on sex chromosomes in an ATR-dependent manner84. The XY body shows characteristic changes in histone modifications including deacetylation of histones H3 and H4 and dimethylation of H3K9 during pachytene, consistent with the inactive state85. Also, H3K4me2 and H3K4me3 are over- and under-represented, respectively. Interestingly, loss of function of PRDM9 causes a failure of XY-body formation81. This sug￾gests that PRDM9 might methylate autosomal loci that are crucial for XY-body formation to keep them transcriptionally active. SCMH1 (sex comb on midleg homologue 1), which is a component of Polycomb repressive complex 1 (PRC1) is involved in histone modifications of the XY body and progression of meiotic prophase86. In cells that lack functional SCMH1, meiosis arrests at late pachytene86. In normal spermatocytes, PRC1 components as well as H3K27me3 are excluded from the XY body at late pachytene, whereas they are abnormally retained on the XY body in Scmh1-mutant spermatocytes. A failure of H3K9me and H3K9me2 accumulation in the XY body and a failure of phosphorylated RNAPII exclusion were also observed in the mutants. Therefore, SCMH1 is required for specific changes in histone modification in the XY body. However, in spite of the aberrant histone modifications, the inactivation of XY chromosomes is not affected by the Scmh1 mutation, and it is currently unclear why meiosis is arrested86. It has been suggested that MSCI might have a role in imprinted X-chromosome inactivation. A model has been proposed in which imprinted X inactivation results from inheritance of an X chromosome that has been pre-inactivated by MSCI from the paternal germline87. This model is supported by the observation that the X chromosome is persistently repressed in postmeiotic haploid cells and retains repressive modifications such as H3K9me2 (REFS 88,89). However, other studies showed that autosomal transgenes carrying an X-inactivation centre do not undergo MSCI but can induce imprinted inactivation of the inserted region90. Furthermore, it was shown that the paternal X chromosome is transcribed at zygotic gene activation in female embryos, arguing against the pre-inactivation model90. Therefore, the rela￾tionship between MSCI and imprinted X inactivation is yet to be clarified. Histone deacetylation in maturating oocytes and segre￾gation of meiotic chromosomes. In mouse oocytes, his￾tones H3 and H4 are generally acetylated at prophase I of meiosis, but they rapidly become deacetylated at metaphase I (MI) by histone deacetylases (HDACs)91 (FIG. 4). In vitro culture of germinal-vesicle-stage oocytes that have been arrested at prophase I in the presence of trichostatin A, an inhibitor of HDACs, showed that neither meiotic maturation, fertilization nor pre￾implantation development is affected92. However, a closer examination revealed that chromosomes are not properly aligned at the metaphase plate in MII oocytes, and that aneuploidy is frequent in single-cell zygotes. Consequently, about half of embryos derived from the trichostatin-A-treated oocytes died in utero. Importantly, histones remained acetylated in the oocytes of older (10-month old) female mice, suggesting that the high incidence of aneuploidy (such as trisomy 21 in humans) in older pregnancies might be due to inadequate histone deacetylation. Histone deacetylation might be involved in specific chromosomal structure that is important for chromosome segregation. Epigenetic changes during gamete maturation After meiosis, both male and female germ cells undergo final developmental changes, at the level of both their morphology and the epigenome, to allow them to carry out their roles in fertilization and the initial stages of zygotic development. In haploid round spermatids, glo￾bal nuclear remodelling occurs, although some histone Table 1 | Impaired meiosis caused by defects in epigenetic modifiers Gene Function of protein Mutant phenotype Refs Suv39h1 and Suv39h2 H3K9 trimethyltransferase (Male) Arrest at mid to late pachytene and apoptosis in double mutants; impaired chromosome synapsis; impaired modification of pericentric heterochromatin 79 Ehmt2 H3K9 mono- and dimethyltransferase (Male and female) Arrest at early pachytene and apoptosis; impaired chromosome synapsis; deregulation of target genes 80 Prdm9 H3K4 trimethyltransferase (Male and female) Arrest at early pachytene and apoptosis; impaired chromosome synapsis and recombination; impaired activation of meiosis-specific genes; impaired XY body formation 81 Dnmt3a De novo DNA methyltransferase (Male) Arrest at pachytene and apoptosis; imprinting failure 52 Dnmt3L Regulator of DNA methylation (Male) Arrest at pachytene and apoptosis; impaired chromosome synapsis; imprinting failure; derepression of retrotransposons 53, 54, 68 Lsh Chromatin remodelling protein of SNF2-helicase family (Female) Arrest and apoptosis at diplotene; impaired chromosome synapsis and recombination; demethylation of retrotransposons and tandem repeats 76 Scmh1 Polycomb group protein (Male) Apoptosis at late pachytene; abnormal chromatin modifications at XY body 86 Mili Piwi family protein; small RNA regulation (Male) Arrest at early pachytene and apoptosis; derepression of retrotransposons 70, 71 Dnmt3a, DNA methyltransferase 3a; Ehmt2, euchromatic histone-lysine N-methyltransferase 2, also known as G9a; Lsh, lymphoid-specific helicase; Prdm9, PR domain-containing 9, also known as Meisetz; Scmh1, sex comb on midleg homologue 1. R E V I E W S 136 | february 2008 | volume 9 www.nature.com/reviews/genetics © 2008 Nature Publishing Group

REVIEWS Primary spermatocyte Diakinesis I H3K9mel 2 H3K9me3 1111111111111111111111 H3K27me3 a!I1 H3ac/H4ac protamin XY body Primary oocyte econdary oocyte Leptotene Pachytene oocyte/egg Growth Extrusion of first Maternal imprinting Figure 4 Epigenetic changes that occur during meiosis in male and female gametogenesis. Changes in epigenetic modifications that occur at various stages of meiosis in male and female germ-cell development are shown. Dashed lines indicate that the level of epigenetic modification is lower during these periods than that during the periods shown by the solid lines. FG. fully grown; MSCI, meiotic sex-chromosome inactivation: ML, metaphase l MIl, metaphase lI; NG, non-growing: PGC, primordial germ celL. marks such as H3K9me2 on the inactive X chromosome chromatin retains up to 15% of the spermatid histones, are retained. A testis-specific linker histone variant and some regions- such as the human protamine gene HIT2 appears at this stage and has an important role cluster and imprinted IG F2 (insulin-like growth factor 2) in chromatin condensation during spermiogenesis 3. -have been reported to be histone-rich. The presence Later, another linker histone variant HILSI(histone-l- of somatic-like chromatin in the sperm nucleus could like protein in spermatids 1)appears in elongated sper- provide a means to transmit epigenetic information matids. In the process of histone-protamine exchange, to the offspring histones are first replaced by TNPI(transition protein 1) The genome-wide DNA methylation pattern changes and TNP2 and then by protamines, and phosphoryla- little during spermiogenesis, as its acquisition has been tion and dephosphorylation of these proteins regulate completed by the end of the pachytene spermatocyte the process. Very recently, the JmjC-domain-contain- stage. However, there is evidence that specific loci such ng histone demethylase 2A(IHDM2A, also known as as Pak2 (phosphoglycerate kinase 2)become de novo JMJD1A), which is an H3K9me1/2-specific demethylase, methylated as late as the sperm-maturation period in the was shown to be necessary for the specific activation of epididymis, the mechanism of which is unknown. As InpI and PrmI (protamine 1). The incorporation of Pgk2 is only expressed in spermatocytes and spermatids, protamines into sperm chromatin induces DNA compac- this methylation might preclude unnecessary activation tion, which is necessary for the formation of spermatozoa of this f and for providing a safe environment for the genome, A recent report showed that there are numerous resistant to physical damage and chemical agents. intra-and inter-individual differences in DNA methyla However, an interesting twist is that mammalian sperm tion in human sperm samples", which could contribute NATURE REVIEWS GENETICS @2008 Nature Publishing Group

marks such as H3K9me2 on the inactive X chromosome are retained88,89. A testis-specific linker histone variant H1T2 appears at this stage and has an important role in chromatin condensation during spermiogenesis93. Later, another linker histone variant HILS1 (histone-1- like protein in spermatids 1) appears in elongated sper￾matids. In the process of histone–protamine exchange, histones are first replaced by TNP1 (transition protein 1) and TNP2 and then by protamines5,94, and phosphoryla￾tion and dephosphorylation of these proteins regulate the process5 . Very recently, the JmjC-domain-contain￾ing histone demethylase 2A (JHDM2A, also known as JMJD1A), which is an H3K9me1/2-specific demethylase, was shown to be necessary for the specific activation of Tnp1 and Prm1 (protamine 1)95. The incorporation of protamines into sperm chromatin induces DNA compac￾tion, which is necessary for the formation of spermatozoa and for providing a safe environment for the genome, resistant to physical damage and chemical agents. However, an interesting twist is that mammalian sperm chromatin retains up to 15% of the spermatid histones, and some regions — such as the human protamine gene cluster and imprinted IGF2 (insulin-like growth factor 2) — have been reported to be histone-rich96. The presence of somatic-like chromatin in the sperm nucleus could provide a means to transmit epigenetic information to the offspring. The genome-wide DNA methylation pattern changes little during spermiogenesis, as its acquisition has been completed by the end of the pachytene spermatocyte stage97. However, there is evidence that specific loci such as Pgk2 (phosphoglycerate kinase 2) become de novo methylated as late as the sperm-maturation period in the epididymis98, the mechanism of which is unknown. As Pgk2 is only expressed in spermatocytes and spermatids, this methylation might preclude unnecessary activation of this gene during post-fertilization development. A recent report showed that there are numerous intra- and inter-individual differences in DNA methyla￾tion in human sperm samples99, which could contribute Secondary spermatocyte Spermatid Spermatozoan (sperm cell) Spermatogonium Leptotene Zygotene Pachytene Diplotene Synapsis Recombination Extrusion of first polar body Ovulation Maturation Resumption of meiosis MSCI XY body Histone– protamine exchange Diakinesis MI MII Leptotene Zygotene Pachytene Diplotene MI PGC Secondary oocyte (FG oocyte) (MII oocyte/egg) Synapsis Recombination Growth (NG oocyte) Maternal imprinting Nature Reviews | Genetics H3K9me1/2 H3ac/H4ac H3K9me3 H3K27me3 Primary spermatocyte Primary oocyte H3K9me1/2 H3K9me3 H3ac/H4ac Entry into meiosis Entry into meiosis Figure 4 | Epigenetic changes that occur during meiosis in male and female gametogenesis. Changes in epigenetic modifications that occur at various stages of meiosis in male and female germ-cell development are shown. Dashed lines indicate that the level of epigenetic modification is lower during these periods than that during the periods shown by the solid lines. FG, fully grown; MSCI, meiotic sex-chromosome inactivation; MI, metaphase I; MII, metaphase II; NG, non-growing; PGC, primordial germ cell. R E V I E W S nature reviews | genetics volume 9 | february 2008 | 137 © 2008 Nature Publishing Group

REVIEWS to phenotypic differences in the next generation(see also adopted zao, because null mutants were lethal. Therefore, BOX 3). Furthermore, it has been reported that sperm the current collection of known epigenetic modifiers that amples from oligospermic patients often contain are essential for germ cells is probably just the tip of an DNA-methylation defects at the imprinted loci 0, 1o. iceberg, which suggests that we will need to study many The significance of these findings and also when and more genes. Although conditional knockout continues how these changes arise are important questions for to be a key technology, recently established GS cells, future studies which can be stably maintained in culture, are genetically In oocytes, except for the rapid deacetylation of modifiable and can differentiate to give rise to functional histones H3 and H4 at the germinal-vesicle break- sperm when transplanted into infertile testes, might offer down mentioned above, histone modifications seem an alternative experimental system 03. to remain unchanged throughout meiosis (FIG. 4]. After Recent advances in epigenomic analysis technolo the first meiotic division, ovulation occurs by releasing gies that allow high-resolution mapping of chromatin the secondary oocyte that is arrested at MIl from the modifications across the genome(for example, REF. 29) ovarian follicle. The ovulated MIl oocyte is now ready also show promise for future progress in this field. to fuse with a sperm cell and the second meiotic divi- These technologies will allow us to know the normal sion is completed after fertilization. Subsequently, the epigenomic landscape of germ cells and epigenomic paternal and maternal genomes of the zygote undergo changes that occur in mutants and disease conditions, The oocyte contributes not only the maternal genome modifiers. However, challenges exist in this respect and its associated epigenetic information, but also some such as the amount of germ-cell samples that are avail factors that are required for post-fertilization reprogram- able to researchers. Further improvement of small-scale ming, which are present in either the cytoplasm or the chromatin immunoprecipitation d unbiased dna nucleus. These factors are important for the develop- amplification methods os will make the technologies ment of both fertilized embryos and nuclear-transferred much more useful. Nuclear-transfer technologies, embryos. Identification of the reprogramming factors, which allow the expansion of cells possessing a specific which might include transcription factors and epigenetic epigenetic profile in cloned embryos " and single-cell modifiers, is an important subject for future study. analysis technologies including live cell imaging are also important. To know the whole picture of the epigenetic regulation of germ-cell development, further studies It has become clear that many epigenetic modifiers, as well as improvements of these key technologies are including DNA methyltransferases, histone-modifica- needed. tion enzymes and their regulatory proteins, have essen- Finally, some of the findings described here will have tial roles in germ-cell development. Mutations that affect an immediate impact on studies of human fertility: mice some of these factors cause early germ-cell loss, whereas harbouring mutations in epigenetic modifier genes can mutations that affect the others cause arrest at specific be models for human infertility conditions. However, meiotic stages and subsequent apoptosis. The functions increasing knowledge of epigenetic regulation of germ of these modifiers are either reprogramming of the germ- cell development will affect a wider biomedical area. cell genome, repression or activation of the downstream For example, one goal of current germ-cell research is target genes, or establishment of a chromatin state that is the efficient and stable derivation of functional gametes appropriate for germ-cell-specific events such as chro- from pluripotent stem cells in culture(BOX 1), success of mosome pairing and genetic crossing over. Furthermore, which depends crucially on knowledge of genetic and the functional links between DNA methylation, histone epigenetic mechanisms of germ cells. Together with modifications and even small-RNA metabolism in germ the use of recently described induced pluripotent stem cells are beginning to be understood, as shown by the cells%.10,, such a technology will allow a new approach recent studies on DNMT3L and MILI (REFS 61, 69). for saving endangered species and overcoming infertil- e Some epigenetic modifiers are specifically expressed ity. In addition, epigenetic changes that are induced by germ cells whereas others are more widely expressed; environmental stress and trans-generational epigenetic this has implications for future studies. The crucial inheritance, which could have relevance to many humai roles of germ-cell-specific genes such as Dnmt3L and diseases, will become an exciting area, for which research iess.5%; the importance of the widely expressed genes implications of these studies are improvements in an. Prdm9 were revealed by conventional knockout stud- on germ cells is particularly important. The far-reachin such as Dnmt3a and Ehmt2 became explicit only when mal cloning, livestock husbandry, assisted reproductive a germline-specific conditional knockout strategy was technologies and human health. 4. Allegrucci, C, Thurston, A, Lucas, E. a Young, L pigenetics: a landscape takes shape. Cell 128. and the germline Reproduction 129 germ cells in the mouse embryo during gastrulation. Surani, M. A, Hayashi, K. Hajkova, P Genetic and 5, Ki ato, M. et al. Identification of PGC7, a new gene s of pluripotency. Cell 128, remodels. a sassone-Corsi,PChromatin Nature434,583-58902005 germ cells. Mech. Dev. 113, 91-94(2002). organ, 6. Reik, w. Stability and flexibility of epigenetic gene 9. Saitou. M. Barton. s C. gulation in mammalian development. Nature 447. mice. Nature418,293-300[2002 38 FEBRU www.nature.com/reviews/genetics @2008 Nature Publishing Group

to phenotypic differences in the next generation (see also BOX 3). Furthermore, it has been reported that sperm samples from oligospermic patients often contain DNA-methylation defects at the imprinted loci100,101. The significance of these findings and also when and how these changes arise are important questions for future studies. In oocytes, except for the rapid deacetylation of histones H3 and H4 at the germinal-vesicle break￾down mentioned above, histone modifications seem to remain unchanged throughout meiosis (FIG. 4). After the first meiotic division, ovulation occurs by releasing the secondary oocyte that is arrested at MII from the ovarian follicle. The ovulated MII oocyte is now ready to fuse with a sperm cell and the second meiotic divi￾sion is completed after fertilization. Subsequently, the paternal and maternal genomes of the zygote undergo further reprogramming to acquire ultimate totipotency. The oocyte contributes not only the maternal genome and its associated epigenetic information, but also some factors that are required for post-fertilization reprogram￾ming, which are present in either the cytoplasm or the nucleus102. These factors are important for the develop￾ment of both fertilized embryos and nuclear-transferred embryos. Identification of the reprogramming factors, which might include transcription factors and epigenetic modifiers, is an important subject for future study. Conclusion It has become clear that many epigenetic modifiers, including DNA methyltransferases, histone-modifica￾tion enzymes and their regulatory proteins, have essen￾tial roles in germ-cell development. Mutations that affect some of these factors cause early germ-cell loss, whereas mutations that affect the others cause arrest at specific meiotic stages and subsequent apoptosis. The functions of these modifiers are either reprogramming of the germ￾cell genome, repression or activation of the downstream target genes, or establishment of a chromatin state that is appropriate for germ-cell-specific events such as chro￾mosome pairing and genetic crossing over. Furthermore, the functional links between DNA methylation, histone modifications and even small-RNA metabolism in germ cells are beginning to be understood, as shown by the recent studies on DNMT3L and MILI (REFS 61,69). Some epigenetic modifiers are specifically expressed in germ cells whereas others are more widely expressed; this has implications for future studies. The crucial roles of germ-cell-specific genes such as Dnmt3L and Prdm9 were revealed by conventional knockout stud￾ies58,59,81; the importance of the widely expressed genes such as Dnmt3a and Ehmt2 became explicit only when a germline-specific conditional knockout strategy was adopted52,80, because null mutants were lethal. Therefore, the current collection of known epigenetic modifiers that are essential for germ cells is probably just the tip of an iceberg, which suggests that we will need to study many more genes. Although conditional knockout continues to be a key technology, recently established GS cells, which can be stably maintained in culture, are genetically modifiable and can differentiate to give rise to functional sperm when transplanted into infertile testes, might offer an alternative experimental system103. Recent advances in epigenomic analysis technolo￾gies that allow high-resolution mapping of chromatin modifications across the genome (for example, REF. 29) also show promise for future progress in this field. These technologies will allow us to know the normal epigenomic landscape of germ cells and epigenomic changes that occur in mutants and disease conditions, and to identify the downstream targets of epigenetic modifiers. However, challenges exist in this respect, such as the amount of germ-cell samples that are avail￾able to researchers. Further improvement of small-scale chromatin immunoprecipitation104 and unbiased DNA￾amplification methods105 will make the technologies much more useful. Nuclear-transfer technologies, which allow the expansion of cells possessing a specific epigenetic profile in cloned embryos37, and single-cell analysis technologies including live cell imaging are also important. To know the whole picture of the epigenetic regulation of germ-cell development, further studies as well as improvements of these key technologies are needed. Finally, some of the findings described here will have an immediate impact on studies of human fertility: mice harbouring mutations in epigenetic modifier genes can be models for human infertility conditions. However, increasing knowledge of epigenetic regulation of germ￾cell development will affect a wider biomedical area. For example, one goal of current germ-cell research is the efficient and stable derivation of functional gametes from pluripotent stem cells in culture (BOX 1), success of which depends crucially on knowledge of genetic and epigenetic mechanisms of germ cells. Together with the use of recently described induced pluripotent stem cells106,107, such a technology will allow a new approach for saving endangered species and overcoming infertil￾ity. In addition, epigenetic changes that are induced by environmental stress and trans-generational epigenetic inheritance, which could have relevance to many human diseases, will become an exciting area, for which research on germ cells is particularly important. The far-reaching implications of these studies are improvements in ani￾mal cloning, livestock husbandry, assisted reproductive technologies and human health. 1. Goldberg, A. D., Allis, C. D. & Bernstein, E. Epigenetics: a landscape takes shape. Cell 128, 635–638 (2007). 2. Surani, M. A., Hayashi, K. & Hajkova, P. Genetic and epigenetic regulators of pluripotency. Cell 128, 747–762 (2007). 3. Morgan, H. D., Santos, F., Green, K., Dean, W. & Reik, W. Epigenetic reprogramming in mammals. Hum. Mol. Genet. 14, R47–R58 (2005). 4. Allegrucci, C., Thurston, A., Lucas, E. & Young, L. Epigenetics and the germline. Reproduction 129, 137–149 (2005). 5. Kimmins, S. & Sassone-Corsi, P. Chromatin remodelling and epigenetic features of germ cells. Nature 434, 583–589 (2005). 6. Reik, W. Stability and flexibility of epigenetic gene regulation in mammalian development. Nature 447, 425–432 (2007). 7. Ginsburg, M., Snow, M. H. & McLaren A. Primordial germ cells in the mouse embryo during gastrulation. Development 110, 521–528 (1990). 8. Sato, M. et al. Identification of PGC7, a new gene expressed specifically in preimplantation embryos and germ cells. Mech. Dev. 113, 91–94 (2002). 9. Saitou, M., Barton, S. C. & Surani, M. A. A molecular programme for the specification of germ cell fate in mice. Nature 418, 293–300 (2002). R E V I E W S 138 | february 2008 | volume 9 www.nature.com/reviews/genetics © 2008 Nature Publishing Group