Alterations in epigenetic modifications during oocyte growth in mice

R
EPRODUCTIONRESEARCH
Alterations in epigenetic modifications during oocyte growthin mice

Shun-ichiro Kageyama1, Honglin Liu1,2, Naoto Kaneko1, Masatoshi Ooga1, Masao Nagata1

and Fugaku Aoki1

1Department of Integrated Biosciences, Graduate School of Frontier Sciences, University of Tokyo, Room #302,Seimei-Building, Kashiwa, Chiba 277-8571, Japan and 2College of Animal Science and Technology, NanjingAgricultural University, Nanjing 210095, Japan

Correspondence should be addressed to F Aoki; Email: [email protected]

Abstract

During oocyte growth, chromatin structure is altered globally and gene expression is silenced. To investigate the involvement of

epigenetic modifications in the regulation of these phenomena, changes in global DNA methylation and in various histone

modifications, i.e. acetylation ofH3K9,H3K18,H4K5, andH4K12, andmethylation ofH3K4 andH3K9,were examined during the

growth of mouse oocytes. Immunocytochemical analysis revealed that the signal intensities of all these modifications increased

during growth and that fully grown, germinal vesicle (GV)-stage oocytes showed themost modifications. Since acetylation of most

of the lysine residues on histones andmethylation of H3K4 are associated with active gene expression, the increased levels of these

modifications do not seem to be associated with gene silencing in GV-stage oocytes. Given that there are two types of GV-stage

oocytes with different chromatin configurations and transcriptional activities, the epigenetic modification statuses of these two

types were compared. The levels of all the epigenetic modifications examined were higher in the SN(surrounded nucleolus)-type

oocytes, inwhich highly condensed chromatin is concentrated in the area around the nucleolus and gene expression is silenced than

in theNSN(not surrounded nucleolus)-type oocytes, in which less-condensed chromatin does not surround the nucleolus and gene

expression is active. In addition, the expression levels of various enzymes that catalyze histone modifications were shown by

RT-PCR to increase with oocyte growth. Taken together, the results show that all of the epigenetic modifications increased in a

concerted manner during oocyte growth, and suggest that these increases are not associated with gene expression.

Reproduction (2007) 133 85–94

Introduction

Generation of gametes during oogenesis is a crucialprocess in the creation of new life for the nextgeneration. Oocytes that are arrested at the prophaseof the first meiosis grow and increase in size in theovaries of female mouse pups after birth, and theseoocytes reach the germinal vesicle (GV) stage withmaximum size at around 3 weeks after birth. Thesegrowing oocytes express several oocyte-specific genes,such as c-mos and zp3. However, as they approach theGV stage, transcription stops and remains silent untilafter the oocytes are fertilized. In mice, transcriptionstarts at mid S/G2 phase in the one-cell stage, whenoocyte-specific genes are not transcribed, while somegenes not previously expressed in oocytes are tran-scribed (Schultz 1993, Aoki et al. 1997, Kigami et al.2003). Therefore, the gene expression pattern is alteredglobally so as to transform the differentiated oocytes into

q 2007 Society for Reproduction and Fertility

ISSN 1470–1626 (paper) 1741–7899 (online)

totipotent embryos, which suggests that global genomeremodeling occurs during these periods.

There is evidence to support the idea that chromatinstructure is altered globally during oocyte growth.Mouse oocytes isolated from antral follicles show twoessentially different chromatin configurations (Debeyet al. 1993, Zuccotti et al. 1995). These are termed SN, inwhich the chromatin is highly condensed and isconcentrated in the area around the nucleolus, andNSN, in which the chromatin is less condensed and doesnot surround the nucleolus. Initially, all the oocytes havethe NSN-type chromatin configuration (Mattson &Albertini 1990, Debey et al. 1993). As the oocytesgrow, the configuration shifts to the SN-type for someoocytes, while it remains as the NSN-type for the others.The percentage of SN-type oocytes increases with theoocyte size (Wickramasinghe et al. 1991, Debey et al.1993, Zuccotti et al. 1995). At the GV stage, the SN-typeconfiguration is observed for a large proportion of the

DOI: 10.1530/REP-06-0025

Online version via www.reproduction-online.org

86 S Kageyama and others

oocytes. The SN-type configuration is strictly correlatedwith the cessation of transcription. In contrast, transcrip-tional activity is detected in NSN-type oocytes of any age(Bouniol-Baly et al. 1999, De La Fuente & Eppig 2001,Liu & Aoki 2002). Therefore, the change in the chromatinconfiguration is likely to play an essential role in thealteration of gene expression during oocyte growth andseems to be involved in global genome remodeling,although the mechanism underlying this change remainsto be elucidated.

Recent studies have revealed that epigenetic modifi-cations, such as DNA methylation and histone modifi-cations, play important roles in the regulation ofchromatin structure and gene expression (Robertson &Wolffe 2000, Jenuwein & Allis 2001, Reik et al. 2001).When cells change their characteristics, e.g. duringdifferentiation or cancellation, genome-wide alterationsin these modifications occur. Indeed, genome-widealterations of epigenetic modifications have beenreported in both early and late oogenesis. At aroundE8.0, germ cells concomitantly and significantly reduceboth di-methylation of lysine 9 on histone H3(H3K9me2) and DNA methylation (Seki et al. 2005).Furthermore, although all the N-terminal lysine residuesof the histones in the nucleus are acetylated in GV-stageoocytes, they are prominently deacetylated after germ-inal vesicle breakdown (Kim et al. 2003, Endo et al.2005). Therefore, genome-wide alterations of DNA andhistone modifications may also be involved in genomeremodeling during oocyte growth. However, little isknown about these modifications at this stage. In thepresent study, using immunocytochemistry, we con-ducted a comprehensive analysis of the genome-widealterations of epigenetic modifications during oocytegrowth in mice. In this analysis, the involvement ofepigenetic modifications in the differentiation of chro-matin configuration, i.e. NSN- to SN-type, was alsoexamined. In addition, we examined the expression ofthe enzymes that catalyze histone modifications ingrowing oocytes.

Materials and Methods

Collection of oocytes

Growing oocytes were collected from the ovaries of 5-,10-, and 15-day-old female BDF1 mice (SLC, Shizuoka,Japan), as described below. Ovaries were removed withfine scissors and freed from the surrounding tissues with a27G needle under the operation microscope. Each ovarywas placed in a 200 ml drop PBS, transferred to 0.5%trypsin–EDTA (Gibco-BRL), and incubated at 38 8C withagitation. After 10 min, the ovary was washed withWhitten’s medium (Whitten 1971), and oocytes withdiameters of 30–40, 40–50, and 50–65 mm for the 5-, 10-,and 15-day-old mice respectively, which were regardedas being of normal size for their respective stages


of development, were collected (Bao et al. 2000,Kageyama et al. 2005).

Forty-six hours after injection of 5 IU of pregnantmare’s serum gonadotropin (Sankyo Co., Ltd., Tokyo,Japan), GV-stage oocytes were collected from 4-week-old BDF1 mice (SLC) by puncturing the follicles with asharp needle. The oocytes were placed in Whitten’smedium that was supplemented with 20 mM HEPES and0.2 mM 3-isobutyl-1-methylxanthine. The oocytes wereliberated from the surrounding cumulus cells by gentlepipetting through a narrow-bore glass pipette. Onlythose oocytes with diameter O70 mm were used asGV-stage oocytes.

All procedures described here were reviewed andapproved by the University of Tokyo Institutional AnimalCare and Use Committee and were performed inaccordance with the Guiding Principles for the Careand Use of Laboratory Animals.

Immunofluorescence confocal microscopy

Oocytes were fixed for 1 h with 3.7% paraformaldehydein PBS. After washing with PBS/0.1% BSA, the oocyteswere permeabilized with 0.5% Triton X-100 in PBS for15 min at room temperature and then immunostainedwith antibodies against acetylated lysines 5 and 12 ofhistone H4 (Upstate Biotechnology, Charlottesville, VA,USA), lysines 9 (Cell Signaling Technology Inc., Beverly,MA, USA) and 18 (Upstate Biotechnology) of histoneH3, di-methylated lysine 4 (Upstate Biotechnology) andtri-methylated lysine 4 (Abcam, Cambridge, MA, USA) ofhistone H3, di-methylated lysine 9 (Upstate Bio-technology) and tri-methylated lysine 9 (Abcam) ofhistone H3, and 5-methylated cytosine (Eurogentec,Seraing, Belgium). To detect 5-methyl-cytosine (5-MeC),the cells were pretreated with 2 N HCl at roomtemperature for 30 min and then neutralized for 20 minwith PBS/0.1% PBS, before treatment with the antibody.Treatment with the primary antibodies (1:100 dilutions)was performed at 4 8C overnight. The cells wereincubated with fluorescein (FITC)-conjugated anti-rabbitor anti-mouse secondary antibody (Jackson ImmunoR-esearch Laboratories, West Grove, PA, USA; 1:50dilution) at room temperature for 45 min. To visualizethe DNA, the cells were counterstained with propidiumiodide. The cells were mounted on glass slides withVectaShield (Vector Laboratories, Burlingame, CA, USA)and observed under the Carl Zeiss 510 laser-scanningconfocal microscope (Carl Zeiss MicroImaging GmbH,Oberkochen, Germany). The oocytes and embryos at allstages were subjected to immunocytochemistry andmounted on glass slides.

Fluorescence was detected using a laser-scanningconfocal microscope (LSM501, Carl Zeiss, Tokyo). Fordetection, the laser power was set to the level at whichthe oocyte with the strongest fluorescence intensityshowed an almost saturated signal in order to adjust the

www.reproduction-online.org

Epigenetic modifications in oocytes 87

signal intensity such that the signals of all of the oocyteswere below the saturated level. Semi-quantitativeanalysis of the fluorescence intensities from the imagesobtained by laser-scanning confocal microscopy wasconducted using the NIH Image program (NationalInstitutes of Health, Bethesda, MD, USA). The pixelvalue/unit area was measured for the nucleus, and theaverage value for two different regions of the cytoplasmwas subtracted as background. This value was multipliedby the thickness of nucleus to correct for differences innucleus size between oocytes. Nuclear thickness wasmeasured by confocal microscopy. In each experiment,the average value calculated for the nuclei of GV-stageoocytes was set at 100% and the values for the growingoocytes were expressed relative to this value. Incomparisons with the SN and NSN types, the averagevalue for the SN-type oocytes was set at 100%.

RT-PCR

Total RNA samples were isolated from oocytes andembryos using ISOGEN (Nippon Gene, Tokyo, Japan), asdescribed previously (Kageyama et al. 2004). The RNAwas reverse-transcribed in a 20 ml reaction mixture thatcontained 5 U ReverScript II (Wako, Osaka, Japan) and0.5 mg oligo(dT)12–18 primer (Invitrogen Corp.) at 42 8Cfor 1 h, followed by 51 8C for 30 min. The templatemRNA was digested with 60 U RNase H (TaKaRa, Shiga,Japan) at 37 8C for 20 min. As the external control, 50 pgof rabbit a-globin RNA was added to each tube beforethe isolation of total RNA.

PCR was performed using the iCycler (Bio-Rad). Thereaction mixture consisted of template cDNA that wasderived from four oocytes or embryos, 0.2 mM of eachprimer, 300 mM dNTPs, 3 mM MgCl2, and 0.05 U/mlExTaq DNA polymerase (TaKaRa). The sequences of thePCR primers used are shown in Supplemental Table S1.PCR was performed for 32 cycles for rabbit a-globin,ESET, and Smyd3, and for 40 cycles for the other genes.Each cycle consisted of denaturation at 95 8C for 15 s,annealing for 15 s, and extension at 72 8C for 20 s. ThePCR products were separated by electrophoresis in a 2%agarose gel and stained with ethidium bromide. The gelimage was obtained using the DT-20MP u.v. illuminator(ATTO, Tokyo, Japan) and the relative amounts of thePCR products were determined by measuring thedensities of the bands using the NIH image software.The values for the transcripts of the target genes werenormalized with those for the rabbit a-globin mRNA.

Results

Changes in histone acetylation during oocyte growth

Global alterations in histone acetylation during oocytegrowth were examined by immunocytochemistry. Theacetylation levels at K9 and K18 on histone H3 (H3K9ac


and H3K18ac respectively), and at K5 and K12 onhistone H4 (H4K5ac and H4K12ac respectively) wereexamined in oocytes collected from 5-, 10-, and 15-day-old mice, and in GV-stage oocytes (Fig. 1). The results offluorescence quantification showed that the levels ofacetylation in any of these lysine residues increased withthe age of the mice or the growth of the oocytes. Thepatterns of increased acetylation levels were similar forall the residues, with the sole exception of H4K5ac(Fig. 2). In these residues, the acetylation levels were lowin oocytes from 5-day-old mice and were slightlyincreased in oocytes from 10-day-old mice. Theacetylation levels increased dramatically in 15-day-oldmice and increased further in GV-stage oocytes. There-fore, the acetylation levels of these lysine residues wereinversely correlated with transcriptional activity. In aprevious report, using a reporter gene, active transcrip-tion was detected when the plasmid was microinjectedinto oocytes with diameters !50 mm, which correspondto the 5- and 10-day-old oocytes in the present study,while transcription was much reduced in oocytes withdiameters O50 mm, which correspond to the 15-day-oldand GV-stage oocytes used in the present study (Worradet al. 1994). In the case of H4K5, the acetylation levelincreased slightly up to day 15, and subsequentlyincreased dramatically in the GV-stage oocytes.

Changes in histone methylation during oocyte growth

The di- and tri-methylation of H3K4 (H3K4me2 andH3K4me3 respectively), and H3K9 (H3K9me2 andH3K9me3 respectively) were examined in the growingoocytes (Fig. 3). It is known that methylation of H3K4and H3K9 is involved in the activation and suppressionof gene expression respectively (Jenuwein & Allis 2001,Grewal & Moazed 2003). The levels of H3K4me2,H3K4me3, and H3K9me2 increased slightly on day 10and then increased significantly on day 15 (P!0.0001,by Student’s t-test). In the GV-stage oocytes, thesemethylation levels increased further (Fig. 4). However,H3K9me3 remained at a low level until day 15 or slightlyincreased on day 15, and then increased prominently inthe GV-stage oocytes.

In the nuclei of oocytes from mice of any age, with theexception of GV-stage oocytes, dense staining forH3K4me3 and H3K9me3 was detected (Fig. 3). Thesesignals co-localized with those for the DNA, with intensefluorescence that presumably represents heterochroma-tin regions, since the dotted signals for tri-methylatedlysines were completely superimposed on those for DNAin the merged images (Supplemental Figs S1 and S2).In the GV-stage oocytes, the fluorescence signals fortri-methylated lysines co-localized uniformly with theDNA signals. On the other hand, in oocytes at any stage,the signals for di-methylated lysines co-localizeduniformly with the DNA signals and few of the dotted


5 days 10 days 15 days GV

H3K9ac

DNA

H3K18ac

DNA

H4K5ac

DNA

H4K12ac

DNA

Figure 1 Alterations in global histone acetylationduring oocyte growth. Growing oocytes, collectedfrom 5-, 10-, and 15-day-old mice, and GV-stageoocytes were immunostained with antibodies againstacetylated lysine 9 on histone H3 (H3K9ac), acetyl-ated lysine 18 on histone H3 (H3K18ac), acetylatedlysine 5 on histone H4 (H4K5ac), and acetylatedlysine 12 on histone H4 (H4K12ac). DNAwas stainedwith propidium iodide. Three independent experi-ments were performed and similar results wereobtained. BarZ20 mm.


signals could be superimposed on those for DNA withintense fluorescence (Supplemental Figs S1 and S2).

Changes in DNA methylation during oocyte growth

The level of 5-MeC increased with oocyte growth, with aslight increase on day 10 and marked increases there-after, until the GV stage (Fig. 5). As observed for tri-methylated H3K4 and H3K9, several of the dottedsignals for 5-MeC could be superimposed on those forDNA in oocytes from all stages, with the exception of theGV stage (Fig. 5A and Supplemental Fig. S3). In the caseof the GV-stage oocytes, the 5-MeC signal was spreadover the entire region in which DNA was detected.


Different epigenetic modifications in two typesof GV-stage oocytes with different chromatinconfigurations

SN- and NSN-type oocytes were distinguished accordingto the chromatin configurations deduced from images ofDNA-stained, GV-stage oocytes. The levels of the variousepigenetic modifications, i.e. H4K5 and H4K12 acetyl-ation, H3K9 methylation, and 5-MeC, were comparedbetween these two oocyte types (Fig. 6A). Quantificationof the immunofluorescence signals revealed that thelevels of all these modifications were higher in SN-typeoocytes than in NSN-type oocytes (Fig. 6B). Thesedifferences did not come from the different patterns ofchromatin condensation, which could cause the


0

50

100

150

5 days 10 days 15 days GVRel

ativ

e am

ount H3K9ac

0

50

100

150


ativ

e am

ount H3K18ac

020406080

100120


ativ

e am

ount H4K5ac

020406080

100120


ativ

e am

ount H4K12ac

Figure 2 Quantification of histone acetylation in growing oocytes.Growing oocytes, collected from 5-, 10-, and 15-day-old mice, andGV-stage oocytes were immunostained with antibodies againstacetylated lysine 9 on histone H3 (H3K9ac), acetylated lysine 18 onhistone H3 (H3K18ac), acetylated lysine 5 on histone H4 (H4K5ac),and acetylated lysine 12 on histone H4 (H4K12ac), followed bytreatment with FITC-conjugated secondary antibodies. The fluor-escence levels of the oocytes were measured as described in theMaterials and Methods section. The average value for GV-stage oocytesis set at 100%, and the values for the growing oocytes are expressedrelative to this value. Bars represent S.E.M. In each group, 17–20 oocyteswere analyzed.


immunofluorescence to be measured incorrectly,because there was no significant difference betweenSN and NSN oocytes in the intensity of fluorescence ofpropidium iodide, which stains DNA (data not shown).Although the distribution of DNA in the nuclei of NSN-type oocytes was similar to that of 15-day-old oocytes,that of 5-MeC was different. Several dotted signals for5-MeC were observed in the nuclei of the 15-day-oldoocytes, whereas few such signals were detected in theNSN-type oocytes (Fig. 6A).

Expression of histone-modifying enzymes

As described above, various histone modifications arealtered dramatically during oocyte growth (Figs 2 and 4).Given that enzymes that catalyze methylation oracetylation of histones have been identified, weexamined the expression of genes that encode thesehistone-modifying enzymes during oocyte growth toelucidate which enzymes are involved in altering histonemodifications during this period.

Histone acetylation is catalyzed by a class of enzymesknown as histone acetyltransferases (HATs). Since globalhistone acetylation of H3K9, K18, and H4K5, K12prominently increases during oocyte growth (Fig. 2), weexamined the expression of SRC-1, p300, PCAF, CBP,HAT1, and TIP60, which have been shown to acetylatethese lysine residues in somatic cells (Lachner et al.2003, Peterson & Laniel 2004, Kikuchi et al. 2005)during oocyte growth (Fig. 7 and Supplemental Fig. S4).The src-1, p300, and pcaf genes exhibited similarpatterns of expression: relatively low expression in5-day-old oocytes, slightly increased expression in10-day-old oocytes, and markedly increased expressionin 15-day-old oocytes. These changes in the expression


levels corresponded to those of acetylated H3K9,H3K18, and H4K12, but not that of H4K5 (Fig. 2).Since SRC-1, p300, and PCAF catalyze the acetylation ofH3K9 (SRC-1; Peterson & Laniel 2004), H3K9, H3K18,and H4K12 (p300; Lachner et al. 2003), and H3K9(PCAF; Kikuchi et al. 2005) respectively, these enzymesmay be involved in the elevated acetylation of theselysine residues during oocyte growth. Although it isknown that p300 also catalyzes the acetylation of H4K5,and its expression level increased markedly in 15-day-old oocytes, the acetylation level of H4K5 did notincrease prominently at that time point (Fig. 2). There-fore, the role of p300 in oocytes may be different fromthat in somatic cells. On the other hand, no obviouschange in expression level was detected for cbp, hat1 ortip60 during oocyte growth (Fig. 7).

Histone lysine methylation is catalyzed by histonemethyltransferases with the SET domain. SET7, SMYD3,and MLL are known to increase the methylation of H3K4(Wang et al. 2001, Hamamoto et al. 2004, Smith et al.2004). The expression level of set7, which is known to beassociated with di-methylation but not tri-methylation ofH3K4, increased continuously during oocyte growth(Fig. 8 and Supplemental Fig. 5). Since this pattern ofchange is similar to that of the H3K4me2 (Fig. 4), SET7may be involved in this type of methylation throughoutoocyte growth. The expression levels of smyd3 and mll,which are associated with tri-methylation of H3K4, wererelatively low in the 5- and 10-day-old oocytes,markedly increased in the 15-day-old oocytes, anddecreased in the GV-stage oocytes (Fig. 8 and Supple-mental Fig. 5). Since the level of H3K4me3 showed onlya slight increase on day 15 (Fig. 4), it is unclear whetherthese enzymes are associated with tri-methylation ofH3K4 during oocyte growth.

Suppressor of variegation 3–9 homolog 1 (SUV39H),ESET, G9a, and its homolog GLP are known to be H3K9methyltransferases. The expression levels of all theseenzymes increased between days 5 and 10 and thenincreased markedly on day 15 (Fig. 8 and SupplementalFig. 5). Since G9a and GLP are known to catalyzedi-methylation of H3K9, the elevated levels of theseenzymes may be involved in increasing H3K9me2during oocyte growth (Fig. 4). SUV39H and ESET havebeen reported to catalyze tri-methylation of H3K9.However, only a marginal increase in H3K9me3 wasobserved on day 15 (Fig. 4). Therefore, it is unclearwhether SUV39H or ESET is associated with tri-methylation of H3K9 during oocyte growth.

Discussion

In the present study, we examined various histonemodifications in growing oocytes, in which the globalgene expression level decreases with growth. Recentstudies have shown that various histone modificationsplay important roles in the regulation of gene expression



H3K4me2

DNA

H3K4me3

DNA

H3K9me2

DNA

H3K9me3

DNA

Figure 3 Alterations in global histone methylationduring oocyte growth. Growing oocytes, collectedfrom 5-, 10-, and 15-day-old mice, and GV-stageoocytes were immunostained with antibodiesagainst di-methylated lysine 4 on histone H3(H3K4me2), tri-methylated lysine 4 on histoneH3 (H3K4me3), di-methylated lysine 9 on histone H3(H3K9me2), and tri-methylated lysine 9 on histoneH3 (H3K9me3). DNA was stained with propidiumiodide. Three independent experiments wereperformed and similar results were obtained.BarZ20 mm.


in mitotic cells. Some modifications, e.g. acetylation ofmost of the lysine residues in histones H3 and H4 andmethylation of H3K4, up-regulate transcription, whileother modifications, e.g. methylation of H3K9, down-regulate transcription (Jenuwein & Allis 2001, Kurdistaniet al. 2004, Bernstein et al. 2005). However, our resultsshow that the levels of all these histone modificationsincrease during oocyte growth. Although transcriptionalactivity decreased with oocyte growth (Worrad et al.1994), the levels of acetylated histones and methylatedH3K4, which are associated with active geneexpression, increased (Figs 1–4). Therefore, it is possiblethat these increases may be involved in genome-widealteration of chromatin configuration, which is not


associated with transcription. Chromatin configurationis altered globally during oocyte growth; the NSN-typeconfiguration is changed to the SN-type configuration(Debey et al. 1993, Zuccotti et al. 1995). It has beenreported that this change in chromatin configuration isnot involved in decreasing transcriptional activity (De LaFuente et al. 2004).

Comparisons of the epigenetic modifications in SN-and NSN-type oocytes revealed that all of the modifi-cations examined were non-equivalent between thesetwo types of oocytes (Fig. 6). Although it has beenreported that DNA methylation and histone modifi-cations contribute to a mechanism that can alterchromatin structure (Jenuwein & Allis 2001), it is not


020406080

100120


H3K4me3

Rel

ativ

e am

ount

020406080

100120


H3K4me2

Rel

ativ

e am

ount

020406080

100120


H3K9me2

Rel

ativ

e am

ount

020406080

100120


H3K9me3

Rel

ativ

e am

ount

Figure 4 Quantification of histone methylation in growing oocytes.Growing oocytes, collected from 5-, 10-, and 15-day-old mice, andGV-stage oocytes were immunostained with antibodies againstdi-methylated lysine 4 on histone H3 (H3K4me2), tri-methylated lysine4 on histone H3 (H3K4me3), di-methylated lysine 9 on histone H3(H3K9me2), and tri-methylated lysine 9 on histone H3 (H3K9me3),followed by treatment with FITC-conjugated secondary antibodies. Thefluorescence levels of the oocytes were measured as described in theMaterials and Methods section. The average value for GV-stage oocytesis set at 100% and the values for the growing oocytes are expressedrelative to this value. Bars represent S.E.M. In each group, 20–22 oocyteswere analyzed.


clear whether changes in histone modifications causethe alteration from the NSN-type to the SN-typechromatin configuration in oocytes. Nevertheless, ourresults suggest that increases in epigenetic modificationsare associated with the acquisition of meiotic anddevelopmental competences (Eppig & Schroeder 1989,Schultz 2002). Small growing oocytes do not have thecompetence to complete meiotic maturation. Duringgrowth, they acquire this competence, which is firstobserved by day 15 post partum. However, these oocytesstill do not have the ability to accomplish preimplanta-tion development. This competence is acquired by day22 post partum. In our study, most of the epigeneticmodifications showed marked increases in 15-day-old

DNA

5-MeC

5 days 10 days 15 daysA

B

020406080

100120


Rel

ativ

e am

ount

of fl

uore

scen

ce


oocytes and further increases in GV-stage oocytes.Furthermore, GV-stage oocytes with the NSN-typechromatin configuration, which are developmentallyincompetent (Liu & Aoki 2002), showed lower levels ofepigenetic modifications when compared with develop-mentally competent SN-type oocytes (Fig. 6). Therefore,increases in epigenetic modifications, which are associ-ated with changes in chromatin configuration, may playan important role in the acquisition of meiotic anddevelopmental competences during oocyte growth.

Although both H3K9me2 and H3K9me3 increasedduring oocyte growth, the patterns of these changes andtheir intra-nuclear localizations differed (Figs 3 and 4).Studies have suggested that di- and tri-methylation exerta similar function, which is to suppress gene expression(Arney & Fisher 2004). However, very recent reportshave suggested that the functions of di- and tri-methylation are different. H3K9me2 localizes to thesilent domains within euchromatin and facultativeheterochromatin, such as the mammalian inactiveX-chromosome, and functionally represses transcription(Rougeulle et al. 2004, Tachibana et al. 2005). On theother hand, H3K9me3 is associated with the formationof highly condensed regions of the genome, which aretermed constitutive heterochromatin (Arney & Fisher2004). These findings are consistent with our resultsshowing that in growing oocytes, H3K9me2 was evenlylocalized throughout the nucleus, in which euchromatinmay be spread uniformly, and that H3K9me3 co-loca-lized with the condensed DNA domains, which mayrepresent heterochromatin regions (Fig. 3). Therefore,the increases in di- and tri-methylation of H3K9 may beinvolved in the suppression of gene expression in theeuchromatin and the formation of heterochromatin in

GV

Figure 5 Alterations in DNA methylation during oocytegrowth. Growing oocytes, collected from 5-, 10-, and15-day-old mice, and GV-stage oocytes were immunos-tained with antibodies against 5-methyl-cytosine(5-MeC), followed by treatment with FITC-conjugatedsecondary antibody. DNA was stained with propidiumiodide. (A) Confocal microscopy images of oocytes.(B) Quantification of fluorescence. The levels of fluor-escence in the oocytes were measured as described in theMaterials and Methods section. The average value forGV-stage oocytes is set at 100%, and the values for thegrowing oocytes are expressed relative to this value. Barsrepresent S.E.M. In each group, 20 oocytes, which hadbeen obtained from three independent experiments, wereanalyzed. BarZ20 mm.


5-MeC

H4K5ac

H4K12ac

H3K9me2

NSNSN

00.20.40.60.81.01.2

5-MeC H4K5ac

H4K12ac H3K9me2

DNA

DNA

NSNSN

DNA

DNA

A

B

Rel

ativ

e am

ount

00.20.40.60.81.01.2

Rel

ativ

e am

ount

00.20.40.60.81.01.2

Rel

ativ

e am

ount

00.20.40.60.81.01.2

Rel

ativ

e am

ount

SN NSN SN NSN

SN NSN SN NSN

Figure 6 Differential DNA methylation andhistone modifications in two types of GV-stageoocytes. GV-stage oocytes were immunostainedwith antibodies against 5-methyl-cytosine(5-MeC), acetylated lysine 5 on histone H4(H4K5ac), acetylated lysine 12 on histone H4(H4K12ac), and di-methylated lysine 9 onhistone H3 (H3K9me2), followed by thetreatment with a FITC-conjugated secondaryantibody. DNA was stained with propidiumiodide. SN-type oocytes (in which the con-densed heterochromatin is located around thenucleolus) and NSN-type oocytes (in which theheterochromatin does not surround thenucleolus) were distinguished according tochromatin configuration, as observed in theimages of the DNA-stained, GV-stage oocytes.(A) Confocal microscopy images of the oocytes.(B) Quantification of fluorescence. The levels offluorescence in the oocytes were measured asdescribed in the Materials and Methods section.The average value for the SN-type oocytes is setat 100% and the values for the NSN-typeoocytes are expressed relative to this value. Barsrepresent S.E.M. The levels of fluorescence weresignificantly higher for the SN-type than for theNSN-type oocytes, for all of the epigeneticmodifications examined (P!0.05; Student’st-test). In each group, 17–30 oocytes, which hadbeen obtained from 2–3 independent experi-ments, were analyzed. BarZ20 mm.


the oocyte genome, which would lead to the silencing ofthe entire genome during oocyte growth.

Our results show that the level of DNA methylationincreases markedly in 10-day-old growing oocytes(Fig. 5). Since it has been reported that DNA methylationof imprinting genes is established between days 10 and15 in oocytes (Lucifero et al. 2004), the increase ingenome-wide DNA methylation seems to reflect thisphenomenon during this period. However, we alsofound that genome-wide DNA methylation increasedafter day 15 (Fig. 5). It has been suggested that de novoDNA methylation is catalyzed by Dnmt3b and Dnmt3L(Lyko et al. 1999, Suetake et al. 2004). The expression ofdnmt3b and dnmt3L mRNAs was observed in GV-stageoocytes as well as in 15-day-old oocytes (La Salle et al.2004, Lucifero et al. 2004). Therefore, it is possible thatthese enzymes catalyze DNA methylation in regions ofthe genome other than imprinted genes until the GVstage. Indeed, DNA methylation was spread over theentire region in which DNA was detected in the nuclei ofGV-stage oocytes. Although the role of this type of global


DNA methylation is not clear, a recent study has shownthat DNA methylation inhibits the mobilization oftransposons (Kato et al. 2003). Although this system ofgenome transformation is important for the acquisitionof genome diversity, germ cells need to maintain stabilityof their genomes, so as to pass them onto the nextgeneration (Lisch 2002, Kazazian 2004). Therefore,genome-wide DNA methylation may repress, transiently,transposon mobilization during oocyte growth.

We examined the expression of various enzymes thatcatalyze histone modifications and found candidateenzymes possibly associated with an increased abun-dance of histone modifications during oocyte growth(Figs 7 and 8). Although the RT-PCR method is not thebest method for detecting slight differences in theamounts of transcripts, the changes in the amounts ofthese enzymes were sufficiently large to confirm theirincrease during oocyte growth. Their expression levelsincreased markedly on day 15, concomitant withmarked increases for most of the histone modifications.The expression levels of the de novo DNA methylases


020406080

100120

CBP

Rel

ativ

e am

ount


020406080

100120 p300

Rel

ativ

e am

ount


020406080

100120 TIP60

Rel

ativ

e am

ount


20406080

100120 PCAF

Rel

ativ

e am

ount

5 days 10days 15days GV

020406080

100120

SRC-1

Rel

ativ

e am

ount


020406080

100120

HAT1

Rel

ativ

e am

ount


0

Figure 7 Expression of enzymes that catalyze histone acetylation inmouse oocytes. The expression levels of the genes that encode thehistone acetyltransferases SRC-1, CBP, P300, HAT1, PCAF, and Tip60were examined by RT-PCR in growing oocytes obtained from 5-, 10-,and 15-day-old mice, and GV-stage oocytes. Three independentexperiments were performed. The average value was determined foreach type of oocyte; the highest value was set at 100%, and the valuesfor other types of oocyte are expressed relative to this value. Barsrepresent S.E.M.

SET 7

020406080

100120

Rel

ativ

e am

ount

5 days 10 days15 days GV

MLL

020406080

100120

Rel

ativ

e am

ount

5 days 10 days15 days GV

020406080

100120

SMYD3

Rel

ativ

e am

ount


020406080

100120

GLP

Rel

ativ

e am

ount

5 days 10 days 15 days GV0

20406080

100120

SUV39H1

Rel

ativ

e am

ount


020406080

100120

G9a

Rel

ativ

e am

ount

5 days 10 days 15 days GV0

20406080

100120

ESET

Rel

ativ

e am

ount


Figure 8 Expression of enzymes that catalyze histone methylation inmouse oocytes. The expression levels of genes that encode the histonemethyltransferases SET7, MLL, SMYD3, GLP, SUV39H1, G9a, and ESETwere examined by RT-PCR in growing oocytes obtained from 5-, 10-,and 15-day-old mice, and GV-stage oocytes. Three independentexperiments were performed. The average value was determined foreach type of oocyte; the highest value was set at 100% and the valuesfor other types of oocyte are expressed relative to this value. Barsrepresent S.E.M.



dnmt3b and dnmt3L showed prominent increases on day15 (Lucifero et al. 2004). Therefore, the changes inepigenetic modifications seem to be regulated by theexpression of their catalyzing enzymes. By contrast, theexpression levels of smyd3 and mll decreased inGV-stage oocytes and were not consistent with thelevels of H3K4 methylation. However, it is still possiblethat these enzymes are involved in H3K4 methylation inthe oocytes, since some transcripts are not translatedconstantly during oocyte growth. For instance, thetranscripts of tPA and hprt are not translated efficientlyduring oocyte growth (Paynton et al. 1988, Huarte et al.1992). Therefore, the analysis of protein levels for theseenzymes would be important for clarifying theirinvolvement in the changes in H3K4 methylation.

In the present study, we examined the changes invarious epigenetic modifications and in the expression ofthe enzymes that catalyze histone modifications duringoocyte growth. The results of these analyses providevaluable clues as to the roles of epigenetic modificationsin mechanisms that regulate cellular functions duringoocyte growth.

Acknowledgements

The authors declare that there is no conflict of interest thatwould prejudice the impartiality of this scientific work.

References

Aoki F, Worrad DM & Schultz RM 1997 Regulation of transcriptionalactivity during the first and second cell cycles in the preimplantationmouse embryo. Developmental Biology 181 296–307.

Arney KL & Fisher AG 2004 Epigenetic aspects of differentiation.Journal of Cell Science 117 4355–4363.

Bao S, Obata Y, Carroll J, Domeki I & Kono T 2000 Epigeneticmodifications necessary for normal development are establishedduring oocyte growth in mice. Biology of Reproduction 62 616–621.

Bernstein BE, Kamal M, Lindblad-Toh K, Bekiranov S, Bailey DK,Huebert DJ, McMahon S, Karlsson EK, Kulbokas EJ III, Gingeras TR,et al. 2005 Genomic maps and comparative analysis of histonemodifications in human and mouse. Cell 120 169–181.

Bouniol-Baly C, Hamraoui L, Guibert J, Beaujean N, Szollosi MS &Debey P 1999 Differential transcriptional activity associated withchromatin configuration in fully grown mouse germinal vesicleoocytes. Biology of Reproduction 60 580–587.

Debey P, Szollosi MS, Szollosi D, Vautier D, Girousse A & Besombes D1993 Competent mouse oocytes isolated from antral follicles exhibitdifferent chromatin organization and follow different maturationdynamics. Molecular Reproduction and Development 36 59–74.

De La Fuente R & Eppig JJ 2001 Transcriptional activity of the mouseoocyte genome: companion granulosa cells modulate transcriptionand chromatin remodeling. Developmental Biology 229 224–236.

De La Fuente R, Viveiros MM, Burns KH, Adashi EY, Matzuk MM &Eppig JJ 2004 Major chromatin remodeling in the germinal vesicle(GV) of mammalian oocytes is dispensable for global transcriptionalsilencing but required for centromeric heterochromatin function.Developmental Biology 275 447–458.

Endo T, Naito K, Aoki F, Kume S & Tojo H 2005 Changes in histonemodifications during in vitro maturation of porcine oocytes.Molecular Reproduction and Development 71 123–128.



Eppig JJ & Schroeder AC 1989 Capacity of mouse oocytes frompreantral follicles to undergo embryogenesis and development tolive young after growth, maturation, and fertilization in vitro. Biologyof Reproduction 41 268–276.

Grewal SI & Moazed D 2003 Heterochromatin and epigenetic controlof gene expression. Science 301 798–802.

Hamamoto R, Furukawa Y, Morita M, Iimura Y, Silva FP, Li M, Yagyu R& Nakamura Y 2004 SMYD3 encodes a histone methyltransferaseinvolved in the proliferation of cancer cells. Nature Cell Biology 6731–740.

Huarte J, Stutz A, O’Connell ML, Gubler P, Belin D, Darrow AL,Strickland S & Vassalli JD 1992 Transient translational silencing byreversible mRNA deadenylation. Cell 69 1021–1030.

Jenuwein T & Allis CD 2001 Translating the histone code. Science 2931074–1080.

Kageyama SI, Nagata M & Aoki F 2004 Isolation of nascent messengerRNA from mouse preimplantation embryos. Biology of Reproduc-tion 71 1948–1955.

Kageyama SI, Liu H, Nagata M & Aoki F 2005 Stage specific expressionof histone deacetylase 4 (HDAC4) during oogenesis and earlypreimplantation development in mice. Journal of Reproduction andDevelopment 52 99–106.

Kato M, Miura A, Bender J, Jacobsen SE & Kakutani T 2003 Role of CGand non-CG methylation in immobilization of transposons inArabidopsis. Current Biology 13 421–426.

Kazazian HH Jr 2004 Mobile elements: drivers of genome evolution.Science 303 1626–1632.

Kigami D, Minami N, Takayama H & Imai H 2003 MuERV-L is one ofthe earliest transcribed genes in mouse one-cell embryos. Biology ofReproduction 68 651–654.

Kikuchi H, Takami Y & Nakayama T 2005 GCN5: a supervisor in all-inclusive control of vertebrate cell cycle progression throughtranscription regulation of various cell cycle-related genes. Gene347 83–97.

Kim JM, Liu H, Tazaki M, Nagata M & Aoki F 2003 Changes in histoneacetylation during mouse oocyte meiosis. Journal of Cell Biology162 37–46.

Kurdistani SK, Tavazoie S & Grunstein M 2004 Mapping global histoneacetylation patterns to gene expression. Cell 117 721–733.

Lachner M, O’Sullivan RJ & Jenuwein T 2003 An epigenetic road mapfor histone lysine methylation. Journal of Cell Science 1162117–2124.

La Salle S, Mertineit C, Taketo T, Moens PB, Bestor TH & Trasler JM2004 Windows for sex-specific methylation marked by DNAmethyltransferase expression profiles in mouse germ cells. Develop-mental Biology 268 403–415.

Lisch D 2002 Mutator transposons. Trends in Plant Science 7 498–504.Liu H & Aoki F 2002 Transcriptional activity associated with meiotic

competence in fully grown mouse GV oocytes. Zygote 10 327–332.Lucifero D, Mann MR, Bartolomei MS & Trasler JM 2004 Gene-

specific timing and epigenetic memory in oocyte imprinting. HumanMolecular Genetics 13 839–849.

Lyko F, Ramsahoye BH, Kashevsky H, Tudor M, Mastrangelo MA, Orr-Weaver TL & Jaenisch R 1999 Mammalian (cytosine-5) methyl-transferases cause genomic DNA methylation and lethality inDrosophila. Nature Genetics 23 363–366.

Mattson BA & Albertini DF 1990 Oogenesis: chromatin andmicrotubule dynamics during meiotic prophase. Molecular Repro-duction and Development 25 374–383.


Paynton BV, Rempel R & Bachvarova R 1988 Changes in state ofadenylation and time course of degradation of maternal mRNAsduring oocyte maturation and early embryonic development in themouse. Developmental Biology 129 304–314.

Peterson CL & Laniel MA 2004 Histones and histone modifications.Current Biology 14 R546–R551.

Reik W, Dean W & Walter J 2001 Epigenetic reprogramming inmammalian development. Science 293 1089–1093.

Robertson KD & Wolffe AP 2000 DNA methylation in health anddisease. Nature Reviews. Genetics 1 11–19.

Rougeulle C, Chaumeil J, Sarma K, Allis CD, Reinberg D, Avner P &Heard E 2004 Differential histone H3 Lys-9 and Lys-27 methylationprofiles on the X chromosome. Molecular and Cellular Biology 245475–5484.

Schultz RM 1993 Regulation of zygotic gene activation in the mouse.BioEssays 15 531–538.

Schultz RM 2002 The molecular foundations of the maternal to zygotictransition in the preimplantation embryo. Human ReproductionUpdate 8 323–331.

Seki Y, Hayashi K, Itoh K, Mizugaki M, Saitou M & Matsui Y 2005Extensive and orderly reprogramming of genome-wide chromatinmodifications associated with specification and early developmentof germ cells in mice. Developmental Biology 278 440–458.

Smith ST, Petruk S, Sedkov Y, Cho E, Tillib S, Canaani E & Mazo A 2004Modulation of heat shock gene expression by the TAC1 chromatin-modifying complex. Nature Cell Biology 6 162–167.

Suetake I, Shinozaki F, Miyagawa J, Takeshima H & Tajima S 2004DNMT3L stimulates the DNA methylation activity of Dnmt3a andDnmt3b through a direct interaction. Journal of Biological Chemistry279 27816–27823.

Tachibana M, Ueda J, Fukuda M, Takeda N, Ohta T, Iwanari H,Sakihama T, Kodama T, Hamakubo T & Shinkai Y 2005 Histonemethyltransferases G9a and GLP form heteromeric complexes andare both crucial for methylation of euchromatin at H3-K9. Genes &Development 19 815–826.

Wang H, Cao R, Xia L, Erdjument-Bromage H, Borchers C, Tempst P &Zhang Y 2001 Purification and functional characterization of ahistone H3-lysine 4-specific methyltransferase. Molecular Cell 81207–1217.

WhittenWK 1971 Nutrient requirement for the culture of preimplanta-tion embryos. Advances in the Biosciences 6 129–139.

Wickramasinghe D, Ebert KM & Albertini DF 1991 Meioticcompetence acquisition is associated with the appearance ofM-phase characteristics in growing mouse oocytes. DevelopmentalBiology 143 162–172.

Worrad DM, Ram PT & Schultz RM 1994 Regulation of geneexpression in the mouse oocyte and early preimplantation embryo:developmental changes in Sp1 and TATA box-binding protein, TBP.Development 120 2347–2357.

Zuccotti M, Piccinelli A, Giorgi Rossi P, Garagna S & Redi CA 1995Chromatin organization during mouse oocyte growth. MolecularReproduction and Development 41 479–485.

Received 8 May 2006

First decision 8 June 2006

Revised manuscript received 1 September 2006

Accepted 13 October 2006


Documents

Alterations in epigenetic modifications during oocyte growth in mice