| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Gamete Biology; |
Laboratory of Cellular Biochemistry, Animal Resource Sciences/Veterinary Medical Sciences, The University of Tokyo, Bunkyo-ku, Tokyo 113-8657, Japan
ABSTRACT
The oocyte-specific subtype of the linker histone H1 is H1FOO, which constitutes a major part of oocyte chromatin. H1foo is expressed in growing oocytes, through fertilization, up until the two–cell embryo stage, when it is subsequently replaced by somatic H1 subtypes. To elucidate whether an epigenetic mechanism is involved in the limited expression of H1foo, we analyzed the dynamics of the DNA methylation status of the H1foo locus in germ and somatic cells. We identified a tissue-dependent and differentially methylated region (T-DMR) upstream of the H1foo gene, which was hypermethylated in sperm, somatic cells, and stem cell lines. This region was specifically unmethylated in the ovulated oocyte, where H1foo is expressed. 5-Aza-2'-deoxycytidine treatments and luciferase assays provided in vitro evidence that DNA methylation plays a role in repressing H1foo in nonexpressing cells. DNA methylation analyses of fetal germ cells revealed the T-DMR to be hypomethylated in female and male germ cells at Embryonic Day 9.5 (E9.5), whereas it was highly methylated in somatic cells at this stage. Intriguingly, the unmethylated status was continuously observed throughout oogenesis at E9.5, E12.5, E15.5, E18.5, in mature oocytes, and after fertilization, in E3.5 blastocysts. In comparison, male germ cells acquired methylation beyond E18.5. These data demonstrate a continuously unmethylated circuit at the H1foo locus in the female germline.
DNA methylation, epigenetics, gametogenesis, gene regulation, histone H1, oocyte, oocyte development
Histones are architectural proteins involved in the packaging of DNA. They can be classified into two groups: core histones and linker histones. The linker histone H1 displays diversity between and within species when compared to core histones, but is conserved from protists to mammals [1]. A switch from one H1 subtype to another during development appears to be an evolutionarily conserved phenomenon. Humans and mice, as well as frogs, zebrafish, and sea urchins, all utilize different classes of H1 at different developmental stages.
In mammals, H1foo (oocyte-specific), Hist1h1t, and H1fnt (testis-specific) are restricted to the germline among the 10 known variants, including seven that are somatic variants. Transitions in H1 variants occur during critical periods in development, such as cell differentiation or embryonic gene activation [2, 3]. H1FOO is the predominant linker histone found in mouse oocytes, expressed beyond the germinal vesicle stage, in polar bodies as well as in the early zygote [4]. At the two-cell stage, when H1FOO expression is mostly extinguished, somatic subtypes take its place in assembling the embryonic chromatin. In fertilization, H1FOO replaces sperm nuclear proteins in the process of active decondensation, suggesting a role for this protein in the remodeling of paternal chromatin [4]. In addition, subtypes of histone H1 contribute to gene-specific regulation and the DNA methylation of selective gene loci [5–7].
DNA methylation is a principal epigenetic mechanism underlying gene regulation and normal development, regulating chromatin structure coordinately with the modification of histones. Genes of developmental importance, including Pou5f1 and Sry, are regulated through DNA methylation-mediated mechanisms [8, 9]. The mammalian genome contains a large number of tissue-dependent and differentially methylated regions (T-DMRs) that make up, as a whole, the DNA methylation "profile" corresponding to individual cell or tissue types [10–12]. Methylation analysis performed on sperm/testis DNA has revealed numerous differentially methylated loci in the male germline, revealing that germ cells exhibit a methylation profile distinct from somatic cells [13, 14]. In general, the hypomethylated state of DNA is permissive of gene expression, and demethylation occurs in advance of gene transcription. Since the DNA methylation profile is specific to the cell type, the formation of a certain methylation pattern during development may contain information that predicts a particular cell lineage.
In the present study, we attempted to elucidate the involvement of DNA methylation in the oocyte-specific expression of histone H1foo. We analyzed germ cells and somatic cells at different stages in development and found that the H1foo locus displays a unique methylation pattern specifically in the female germline.
All experiments were carried out according to the institutional guidelines for the care and use of laboratory animals (Graduate School of Agriculture and Life Sciences, The University of Tokyo).
Female mice (C57BL/6NCrj) were purchased from Charles River Laboratories, Inc. (Yokohama, Japan) and mated with GOF18 (Oct-4) delta PE/GFP transgenic males [15, 16]. Noontime of the day a vaginal plug was detected was designated as Embryonic Day 0.5 (E0.5). Dorsal mesenteries from E9.5 and gonads from E12.5, E15.5, and E18.5 fetuses were isolated in Dulbecco modified Eagle medium (DMEM; Gibco BRL, Rockville, MD) containing 15% fetal bovine serum (FBS) and incubated in PBS(-) containing 0.02% EDTA and 0.02% glucose (pH 7.2) at room temperature. Gonadal cells were disaggregated, and germ cells, selected for GFP expression, were collected using a micromanipulator (MMO-202N, Narishige, Tokyo, Japan) under an inverted fluorescence microscope (IX70/IX-FLA, Olympus, Tokyo, Japan). Each embryo was sexed by PCR at E9.5 and by the morphology of the gonads at E12.5–E18.5. For E12.5–E18.5 samples, 200–650 germ cells were collected from each individual embryo. For E9.5 samples, 540 and 510 germ cells were pooled from seven embryos each for male and female analysis, respectively. Fetal heart and liver were dissected and collected as controls. Whole gonads isolated from E12.5, E15.5, and E18.5 GFP-negative fetuses were used for RT-PCR analysis. All germ cells and fetal tissues were frozen in liquid nitrogen and stored at –80°C until use.
Collection of Mature Oocytes and Blastocysts
Female mice were injected with 7.5 IU of eCG (Teikoku Hormone Mfg. Co., Ltd., Tokyo, Japan) and 7.5 IU of hCG (Teikoku Hormone Mfg. Co., Ltd.) with a 48-h interval. Ovulated metaphase II oocytes were collected 24 h later by puncturing the swollen ampullae. Oocytes were washed and denuded of cumulus cells by treatment with 1% hyaluronidase (Sigma-Aldrich, Tokyo, Japan). Removal of the zona pellucida and first polar body was conducted using acidic Tyrode solution (pH 2.5) and gentle pipetting. Oocytes (70–90) were collected for DNA analysis. To obtain blastocysts, superovulated females were each caged with a wild-type stud male overnight and examined for a vaginal plug the following morning. At E3.5, 13–14 blastocysts were collected by flushing the uterus. Oocytes and blastocysts were collected twice from two separate rounds of superovulation. Samples were rinsed, frozen in liquid nitrogen, and stored at –80°C until use.
Wild-type (J1), DNA methyltransferase (Dnmt) 1–/–, Dnmt 3a–/–, Dnmt 3b–/–, Dnmt 3a–/–3b–/– embryonic stem (ES) cells of 129S4/SvJae background were cultured as previously described [17, 18].
Embryonic Germ Cell Culture and 5-Aza-2'-deoxycytidine Treatment
Female embryonic germ (EG) cells (12.5 cells) were derived from E12.5 oogonia of C57BL/6NCrj background as previously described [19] and maintained on EMFI feeder cells. The culture medium was DMEM (pH7.4) supplemented with 15% FBS, 100 µM/ml 2-mercaptoethanol, 100 µM/ml MEM nonessential amino acids, 1 mM/ml sodium pyruvate, 2 mM/ml L-glutamine, 100 U/ml penicillin, and 100 µg/ml streptomycin. To maintain an undifferentiated state, 1000 U/ml LIF was added to the culture medium, and the medium was changed at 1-day intervals.
Prior to treatment with 5-Aza-2'-deoxycytidine (5-aza-dC; Sigma), EG cells were precultured for 24 h and then cultured for 48 h in medium containing 0, 0.01, 0.05, 0.5, or 1.0 µM 5-aza-dC.
Analysis of H1foo Gene Expression by RT-PCR
Total RNA was isolated from tissues, gonads, and EG cells using TRIzol reagent (Invitrogen Corp., Carlsbad, CA) according to the manufacturer's instructions. Prior to first strand cDNA synthesis, total RNA was treated with RNase-free DNase I (Takara, Kyoto, Japan) to eliminate contamination of residual DNA. First strand cDNA synthesis was performed using the SuperScript III First-Strand Synthesis System for RT-PCR (Invitrogen). PCR was performed using the primer sets for H1foo, Actb, and Ifitm3 (see Supplementary Table 1 available online at www.biolreprod.org). Ifitm3 primers were previously described [20]. The annealing temperatures and PCR cycles used were H1foo: 55°C, 30 cycles; Actb: 55°C, 25 cycles; and Ifitm3: 58°C, 25 cycles. PCR products were subjected to agarose gel electrophoresis and stained using ethidium bromide.
A 278-bp cDNA fragment corresponding to the nucleotide positions 136–413 of H1foo cDNA was amplified by RT-PCR and cloned into a pGEM-T Easy vector (Promega, Madison, WI). The probe covers exon 2 and detects both known transcripts of the H1foo gene [21]. In situ hybridization was performed as previously described [22] by GenoStaff (Tokyo, Japan). Sense or anti-sense digoxigenin (DIG)-labeled RNA probes were generated using DIG RNA Labeling Mix (Roche Molecular Biochemicals, Tokyo, Japan). Paraffin-embedded sections of E18.5 and adult ovaries were subjected to hybridization with the antisense or sense probes at the concentration of 100 ng/ml in the Probe Diluent (GenoStaff) at 60°C overnight. Hybridized probes were detected by alkaline phosphatase-conjugated anti-DIG IgG and visualized with NBT/BCIP solution (Roche Molecular Biochemicals). The sections were counterstained with Kernechtrot stain solution (Muto Chemical, Tokyo, Japan), dehydrated, and mounted with Malinol (Muto Chemical).
The 5'-flanking region of the H1foo gene (–1629 bp 
+42 bp) was amplified by PCR from the genomic DNA of C57BL/6NCrj mice. The fragment was cloned into a pGL3-Basic vector and amplified using the dam– dcm– bacterial strain, SCS110 (Stratagene, La Jolla, CA). Reporter constructs were methylated in vitro using three units of SssI methylase (New England BioLabs, Inc., Beverly, MA) per microgram of DNA in the presence of 160 µM S-adenosylmethionine at 37°C for 1.5 h. Completion of methylation was confirmed by resistance to methylcytosine-sensitive restriction enzymes.
Female EG cells (12.5) were cultured on a 24-well plate (1.7 x 105 cells per well) for 24 h and then transfected with 1.2 µg of the reporter constructs using LipofectAMINE reagent (Invitrogen). To normalize firefly luciferase activity, an internal control plasmid (0.07 µg) expressing Renilla luciferase (pRL-TK vector; Promega) was cotransfected into the cells. The activities of both luciferases were determined 48 h after transfection using a Dual-Luciferase reporter assay system (Promega) according to the manufacturer's instructions. Assays were performed twice in triplicate.
Genomic DNA from tissues, cultured cells, germ cells, and embryos were extracted as previously described [9]. DNA was digested with EcoRI (Takara), and bisulfite treatment was carried out as described [8]. The ethanol-precipitated DNA was suspended in water and amplified by PCR using a combination of primers designed to amplify the 5'-flanking region of the H1foo gene (Supplementary Table 1). All data were collected on a single PCR basis, and no nested PCR was used. For E12.5–E18.5 germ cells, oocytes, and blastocysts, PCR was performed at least twice on two independently collected and bisulfite-treated samples to avoid inconsistencies between sample preparations. PCR products were cloned into pGEM-T Easy vector (Promega) and sequenced. A representative set of 10 clones is shown for each sample.
Developmental Stage and Cell Type-Specific Expression of H1foo mRNA
To elucidate the timing and specificity of H1foo transcription, expression analyses of adult and fetal tissues were performed. In sections of the adult ovary, H1foo was detected in oocytes encased in primordial follicles, with the signal intensifying with follicular growth (Fig. 1A, a–c). H1foo was not detected in surrounding follicle cells (Fig. 1A) nor in other somatic tissues (data not shown), thereby confirming oocyte-specific expression. Analysis of fetal gonads was performed to determine whether H1foo transcription is initiated before the primordial follicle stage. H1foo expression was barely detectable in E18.5 ovaries as well as in E15.5 and E18.5 testis using RT-PCR (Fig. 1A, bottom). In addition, in situ hybridization of E18.5 ovaries failed to reveal any specific signal for H1foo (Fig. 1A, d and e), supporting subdued expression in fetal gonads. Thus, H1foo was transcribed in a cell type-specific and stage-dependent manner, primarily in the female germ cell lineage.
|
Cell Type-Dependent DNA Methylation Status of the H1foo Locus
The limited expression of H1foo in oocytes led us to investigate the DNA methylation status of the 5'-flanking region in various tissues and cell types. The upstream region of the H1foo gene is relatively limited in the number of CpG sites, with 36 CpGs found within 3 kb of the transcription start site. We focused on the 1.6-kb region, which contains 16 CpG sites, including one CpG in the 5'-UTR (Fig. 1B top). Genomic DNA from germ cells, embryos, adult tissues, and stem cell lines were subjected to bisulfite conversion, and the methylation status of these cells was analyzed (Fig. 1B). In the sperm, where H1foo is not expressed, all 16 CpGs were heavily methylated. In contrast, these CpGs were clearly hypomethylated in the oocyte. Further analysis of the methylation status of adult tissues, including liver, brain, and heart, revealed the H1foo locus to be highly methylated in all tissues examined. Blastocysts collected at E3.5 were the only other cell type to exhibit an unmethylated pattern of the H1foo locus in vivo. DNA hypermethylation was observed in EG cells and ES cells. As compared to wild-type ES cells, Dnmt1–/– ES cells displayed a marked decrease in the overall methylation. ES cells deficient in either Dnmt3a or Dnmt3b maintained a near-normal methylation level comparable to the wild type. However, a lack of both Dnmt3a and Dnmt3b resulted in a dramatic loss of methylation throughout the examined region. Therefore, a combination of DNMT1 and at least one of the two de novo DNMTs, DNMT3A or DNMT3B, is essential for methylating the H1foo locus in ES cells. Thus, we defined a 1.6-kb region as the T-DMR of the H1foo gene.
Alteration of DNA Methylation Affects H1foo Gene Expression In Vitro
We next examined the role of DNA methylation in H1foo gene expression. H1foo expression was not detected in EG cells derived from E12.5 oogonia (Fig. 1C, top). Treatment of EG cells with 5-aza-dC (an inhibitor of DNA methylation) resulted in the upregulation of H1foo as analyzed by RT-PCR, suggesting the involvement of DNA methylation in H1foo gene regulation. To further address this issue, luciferase assays were conducted to analyze the effect of DNA methylation on H1foo promoter activity. A reporter plasmid was constructed by ligating the H1foo T-DMR (–1629 to +42 bp) to a pGL3-Basic vector. Promoter activity was analyzed using EG cells, since EG cells demonstrated a capacity for H1foo expression with 5-aza-dC treatment. The H1foo reporter construct exhibited 3- to 4-fold activity relative to that of the empty vector, indicating that the T-DMR has promoter activity in EG cells (Fig. 1C, bottom). In vitro methylation of this construct caused a severe suppression of luciferase luminescence, indicating a role for CpG methylation in H1foo gene silencing.
The H1foo T-DMR Is Unmethylated Throughout Female Germ Cell Development
To explore the correlation of DNA methylation with stage-specific expression of H1foo, the T-DMR of oogonia/oocytes from different stages in development, i.e., at E9.5, E12.5, E15.5, and E18.5, were analyzed. At E9.5, the entire T-DMR was hypomethylated in germ cells (Fig. 2A). This unmethylated state was also evident in E12.5 oogonia, implying that the H1foo locus remains hypomethylated in germ cells both before and after the colonization of the gonad. E15.5 and E18.5 oocytes were also unmethylated, suggesting continuous hypomethylation of the T-DMR during oogenesis. The unmethylated status of female germ cells raised the question of whether male germ cells and somatic cells follow the same dynamics during this fetal growth period. Male germ cells at E9.5, E12.5, and E15.5 exhibited low levels of methylation comparable to their female counterparts (Fig. 2B). However, in E18.5 prospermatogonia, the T-DMR became extensively methylated, and methylation appeared complete in the adult sperm. The T-DMR was also largely methylated in somatic cells of the E9.5 fetus, except for CpGs 1–4, which eventually become methylated in the adult (Figs. 2C and 1B). Hypermethylation was found in fetal heart of both sexes. CpGs 5–16 were methylated in the fetal liver as well, indicating early establishment of a hypermethylated state in somatic tissues.
|
We found that the H1foo locus contains a T-DMR that is responsible for gene repression. The T-DMR was consistently hypomethylated in the female germline at all stages examined. This unmethylated status is lineage-specific, as male germ cells and somatic cells became methylated during early development. Figure 3 summarizes the methylation dynamics of the H1foo T-DMR in female/male germ cells and somatic cells. Male and female germ cells at E9.5, E12.5, and E15.5 are indistinguishable in terms of the methylation status of the H1foo T-DMR. Somatic cells differ in methylation pattern from germ cells at E9.5, although neither germ cells nor somatic cells express H1foo at this stage.
|
The DNA methylation dynamics of the H1foo locus are unique compared to other genes studied in the context of germ cell development. Recent studies have documented the methylation patterns of several imprinted, nonimprinted, and germ cell-specific genes in primordial germ cells (PGCs) [23–27]. PGCs are the earliest recognizable precursors of gametes and are first detectable as alkaline phosphatase-positive cells around E7.0. The paternally imprinted genes Peg3 and Kcnq1ot1 and the nonimprinted gene Myl3 all follow the same dynamics of hypermethylation in E11.5 PGCs and hypomethylation at approximately E12.5 [23]. Similarly, at the maternally imprinted H19 locus, the paternal allele is highly methylated at E9.5 and E10.5, with a sharp decrease in methylation at E11.5 [23, 25]. Germ cell-specific genes, including Ddx4 and Dazl, are also extensively demethylated as the PGCs enter the genital ridges [26]. Another gene, Pgk2, exhibits hypermethylation until E15.5, and then is gradually demethylated before the onset of transcription in meiotic spermatocytes [27]. However, H1foo is unmethylated in both premigratory and postmigratory gonocytes as well as in E3.5 blastocysts, leading us to speculate continuous hypomethylation of this locus. The unmethylated state of E3.5 blastocysts suggests that the paternal allele is also demethylated after fertilization. It seemed that this demethylation occurred in both male and female embryos, for two sets of individually pooled blastocysts (a total of 27 blastocysts) yielded hardly any methylated clones. The T-DMR in the male germline acquires methylation at approximately E18.5, which likely accounts for the subsequent H1foo repression. H1foo-expressing oocytes, including transcriptionally quiescent mature oocytes, are the only cells found to maintain a hypomethylated state in the T-DMR in the adult.
E12.5 oogonia have an unmethylated T-DMR, whereas the T-DMR in E12.5 female EG cells is highly methylated. EG cells can be derived from E8.0–E12.5 PGCs by culturing them on feeder layers with the addition of LIF, KITL, and FGF2 [19, 28]. The methylation status of a number of imprinted loci has been examined in EG cells. Labosky et al. [29] revealed that both the paternal and maternal alleles of Igf2r region 2 were unmethylated in 10 EG cell lines derived from E12.5 male PGCs, which is consistent with the disappearance of methylation in PGCs at this stage [24]. Comparable results have been obtained for imprinted genes including Snrpn and Peg3 [30]. For the H1foo T-DMR, a different situation is found, where EG cells acquire de novo methylation upon derivation from PGCs. A similar situation is seen with ES cells, where the T-DMR is highly methylated, in contrast to the hypomethylation seen in blastocysts. The H1foo T-DMR likely becomes methylated during the establishment of stem cell lines in culture, thus ensuring the inhibition of ectopic gene expression in cultured cells.
Improper expression of germline H1s in somatic lineages may affect selective gene expression and/or cell cycle progression, which has been reported for cells overexpressing subtypes of somatic H1 [31]. Histone H1s modulate chromatin structure and interact with epigenetic factors such as Cbx5 (formerly known as HP1
) [32–34]. They should normally function within a network of chromatin binding proteins that remodel chromatin at the levels of both the nucleosome and the higher order compaction of chromatin fibers [35]. It has been shown that the testis-specific histone Hist1h1t is hypermethylated in somatic cells, in contrast to the hypomethylated status in primary spermatocytes [36]. Taken together with our findings, histone H1 should be regarded as an epigenetic modifier that is, in itself, regulated by epigenetics.
The T-DMR of H1foo in the female germline was mostly hypomethylated regardless of its transcription status, whereas it was hypermethylated in somatic cells. What could be the function of this continuous hypomethylation in germ cells? The fact that the oocyte-specific histone subtype is conserved from the sea urchin to humans supports an evolutionarily preserved role for this constitutive protein. Phylogenetic analyses suggest that the oocyte-type H1 (B4/H1M) diverged from the rest of the H1 species long before the differentiation of other somatic subtypes [37]. Our results suggest that the unmethylated state of the T-DMR is a precondition for H1foo gene expression and might distinguish cells that have the potential to develop into female germ cells. We suggest that the H1foo T-DMR remains hypomethylated in the female germline throughout consecutive generations, with implications for germ cell-specific gene expression.
In conclusion, the present study proposes DNA methylation to be a key mechanism underlying the repression of H1foo in nonexpressing cells and suggests a continuously unmethylated status of this locus throughout the life span of the female germline.
ACKNOWLEDGMENTS
We thank Dr. En Li for providing us with the ES cells, Dr. Yasuhisa Matsui for GOF18 (Oct-4) delta PE/GFP transgenic mice, Dr. Yoshikazu Arai for technical assistance, and Dr. Jun Ohgane for helpful comments on the manuscript.
FOOTNOTES
1Supported by the Program for Promotion of Basic Research Activities for Innovative Biosciences and the Grant-in-aid for Scientific Research, Ministry of Education, Culture, Sports, Science and Technology, Japan (15080202) to K.S. 
Correspondence: 2Kunio Shiota, Laboratory of Cellular Biochemistry, Animal Resource Sciences/Veterinary Medical Sciences, The University of Tokyo, Bunkyo-ku, Tokyo 113-8657, Japan. FAX: 81 3 5841 8189; e-mail: ashiota{at}mail.ecc.u-tokyo.ac.jp
Received: 6 November 2007.
First decision: 30 November 2007.
Accepted: 27 December 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
