Content uploaded by Heiner Niemann
Author content
All content in this area was uploaded by Heiner Niemann on Apr 03, 2020
Content may be subject to copyright.
Gene-specific profiling of DNA methylation and
mRNA expression in bovine oocytes derived from
follicles of different size categories
F. Mattern
A
,J. Heinzmann
B
,
C
,
D
,D. Herrmann
B
,A. Lucas-Hahn
B
,T. Haaf
A
and H. Niemann
B
A
Institute of Human Genetics, Julius Maximilians University, 97070 Wu
¨rzburg, Germany.
B
Institute of Farm Animal Genetics, Friedrich-Loeffler-Institut, Federal Research Institute of
Animal Health, Mariensee, 31535 Neustadt, Germany.
C
Present address: Gynemed GmbH and CO. KG, 23738 Lensahn, Germany.
D
Corresponding author. Email: heiner.niemann@fli.de
Abstract. Epigenetic changes, such as DNA methylation, play an essential role in the acquisition of full developmental
competence by mammalian oocytes during the late follicular growth phase. Here we used the bovine model to investigate
the DNA methylation profiles of seven candidate genes (imprinted: bH19,bSNRPN; non-imprinted: bZAR1,bDNMT3A,
bOCT4,bDNMT3 Lo and bDNMT3 Ls) and the mRNA expression of nine candidate genes (imprinted: bSNRPN,bPEG3,
bIGF2R; non-imprinted: bPRDX1,bDNMT1B,bDNMT3A,bZAR1,bHSF1 and bNLRP9) in oocytes from antral follicles
of three different size classes (#2 mm, 3–5 mm, $6 mm) to unravel the epigenetic contribution to this process.
We observed an increased number of aberrantly methylated alleles in bH19,bSNRPN and bDNMT3 Lo of oocytes from
small antral follicles (#2 mm), correlating with lower developmental competence. Furthermore, we detected an increased
frequency of CpG sites with an unclear methylation status for DNMT3 Ls, specifically in oocytes from follicles $6 mm,
predominantly at three CpG positions (CpG2, CpG7 and CpG8), of which CpG7 is a potential regulatory site. No major
differences in mRNA expression were observed, indicating that the transcriptional machinery may not yet be active during
the follicular growth phase. Our results support the notion that a follicle diameter of ,2 mm is a critical stage for
establishing DNA methylation profiles and indicate a link between DNA methylation and the acquisition of oocyte
developmental competence.
Additional keywords: abnormally methylated alleles, CpG sites, epigenetics, imprinted genes, limited dilution assay,
pyro-sequencing.
Received 19 August 2016, 14 December 2016, published online 3 February 2017
Introduction
The relationship between follicle size and oocyte developmental
competence is well established in several mammalian species,
i.e. human (Tsuji et al. 1985), pig (Motlik et al. 1984), sheep
(Szo¨llo¨si et al. 1988), cow (Pavlok et al. 1992), goat (Crozet
et al. 1995), monkey (Schramm et al. 1993) and horse (Goudet
et al. 1997). The proportion of blastocysts was significantly
higher when bovine oocytes from larger ($2 mm) follicles had
been used rather than oocytes isolated from small (#2 mm)
follicles (Pavlok et al. 1992;Lonergan et al. 1994;Blondin and
Sirard 1995;Hagemann et al. 1999;Lequarre et al. 2005).
Meiotic progression of mammalian oocytes requires the well-
orchestrated expression of genes to achieve full developmental
competence. De novo synthesis of RNA has been observed up to
the germinal vesicle (GV) stage and ceases after germinal ves-
icle breakdown (GVBD; Memili et al. 1998;Tomek et al. 2002).
Genes involved in epigenetic reprogramming and chromatin
modifications are thought to play a crucial role during oocyte
maturation (De La Fuente 2006;Oliveri et al. 2007).
Epigenetic changes have been identified as a fundamental
mechanism to ensure the well-orchestrated expression of genes,
initially in the oocyte genome and subsequently in the biparental
embryonic genome. The necessity for the presence of both a
male and female pronucleus for normal embryogenesis had been
demonstrated by generating diploid biparental gynogenetic and
androgenetic mouse embryos (McGrath and Solter 1984;Tsuji
et al. 1985;Szo¨llo¨siet al. 1988;Schramm et al. 1993;Lonergan
et al. 1994;Blondin and Sirard 1995;Crozet et al. 1995;
Goudet et al. 1997;Memili et al. 1998;Hagemann et al. 1999;
Tomek et al. 2002;Oliveri et al. 2007). The asymmetric
genomic contribution of the pronuclei is achieved by methyla-
tion of specific DNA regions to ensure parental-allele-specific
CSIRO PUBLISHING
Reproduction, Fertility and Development, 2017, 29, 2040–2051
https://doi.org/10.1071/RD16327
Journal compilation ÓCSIRO 2017 www.publish.csiro.au/journals/rfd
gene expression (Reik and Walter 2001). Errors in genomic
imprinting are usually associated with abnormal development.
Up to now, nine human imprinting syndromes have been identi-
fied, of which three (Beckwith–Wiedemann syndrome (BWS),
Angelman syndrome and the maternal hypomethylation syn-
drome) show strong links to assisted reproduction technologies
(ART; Amor and Halliday 2008). The large offspring syndrome
(LOS) has a similar phenotype as BWS and has been reported in
ruminants after ART, specifically after somatic cell nuclear
transfer (SCNT; Robbins et al. 2012;Chen et al. 2013). In
addition to imprinted genes, other developmentally important
genes, such as the pluripotency marker OCT4, are regulated by
DNA methylation before and after fertilisation (Scho¨ler 1991;
Pesce and Scholer 2001;Favaedi et al. 2012). Previous studies in
mice and cattle revealed an increasing number of maternal
imprints with progression of folliculogenesis, thus indicating a
size-dependent establishment of maternal imprints (Hiura et al.
2006;O’Doherty et al. 2012).
Previously, we reported the status of two distinct histone
modifications, i.e. H3K9 me2 and H4K1ac, and global DNA
methylation for bovine oocytes derived from two different
classes of follicles (#2mm and $2–8 mm) during in vitro
maturation. While global DNA methylation remained stable
during the entire maturation period, H3K9 me2 and H4K1ac
displayed stage-specific profiles, suggesting an essential
involvement in the acquisition of oocyte developmental
capacity (Racedo et al. 2009). Bos taurus is a uniparous
species, has a gestation length of 9 months and has very
similar follicular dynamics to humans (Me´ne´zo and He´rubel
2002). Usually, one follicle emerges out of a cohort of antral
follicles, becomes dominant and develops into a Graafian
follicle (McGee and Hsueh 2000;Kanitz 2003). Moreover,
humans and cattle also seem to share many features in the
establishment of methylation imprints during oogenesis, as
indicated by BWS and LOS (Robbins et al. 2012;Chen et al.
2013).
The goal of the present study was to gain insight into the
establishment of DNA methylation marks and mRNA expres-
sion profiles using a panel of genes known to be critically
involved in the regulation of growing female germ cells. We
used the bovine model to investigate the DNA methylation
profiles of seven candidate genes (imprinted: bH19,bSNRPN;
non-imprinted: bZAR1,bDNMT3A,bOCT4,bDNMT3 Lo and
bDNMT3 Ls) and the mRNA expression of nine candidate genes
(imprinted: bSNRPN,bPEG3,bIGF2R; non-imprinted:
bPRDX1,bDNMT1B,bDNMT3A,bZAR1,bHSF1 and bNLRP9)
in oocytes from antral follicles of three different size classes
(#2 mm, 3–5 mm, $6 mm). The different size categories corre-
spond to different developmental capacities of bovine oocytes
(Pavlok et al. 1992;Lonergan et al. 1994).
H19 encodes an RNA that might serve as a primary micro-
RNA precursor downregulating specific mRNAs during verte-
brate development. The promotor of H19 is unmethylated in the
oocyte and methylated in spermatozoa (Cai and Cullen 2007;
Ideraabdullah et al. 2008;Hansmann et al. 2011). The small
nuclear ribonucleoprotein N (SNRPN) is a splicing factor that is
regulated by a differentially methylated region (DMR). The
DMR of SNRPN is methylated in the oocyte and is linked to
Prader–Willi and Angelman syndromes, caused by opposite
methylation defects (Lucifero et al. 2006).
The octamer-binding transcription factor 4 (Oct4) is one of
the most valid markers used for studies on epigenetic repro-
gramming and pluripotency. Through its Pit-Oct-Unc (POU)
domain within the Oct4 protein, target genes are transcription-
ally regulated by binding to an octamer recognition sequence
within specific promoter or enhancer regions (Scho¨ler 1991;
Pesce and Scholer 2001;Favaedi et al. 2012). Oct4 is unmethy-
lated and expressed in pluripotent cells, including embryonic
stem (ES) cells, embryonic carcinoma (EC) cells and oocytes.
Gene structure and sequence, localisation of the gene, regulatory
regions and even expression patterns are highly conserved
between the mouse, human and bovine orthologues (Nordhoff
et al. 2001;Kurosaka et al. 2004;Kuhtz et al. 2014).
In cattle, ZAR1 is expressed throughout preimplantation
embryo development, as well as in the ovaries, testis, heart
and muscle. Knockout experiments revealed a crucial role for
oocyte-to-embryo transition. The translated protein of this
oocyte-specific maternal-effect gene contains an evolutionary
conserved plant homeo-domain (PHD), indicating a role in
transcription regulation (Wu et al. 2003a,2003b;Uzbekova
et al. 2006).
DNA methyltransferases (DNMTs), including DNMT1B,
DNMT3A and DNMT3 L, are enzymes that are critically
involved in the establishment and maintenance of the correct
methylation patterns at imprinted and non-imprinted loci in
early development. DNMT3A, in combination with DNMT3 L,
is essential for establishing the initial CpG methylation pattern
(de novo methylation) during gametogenesis, whereas
DNMT1B is mainly involved in maintaining this pattern during
chromosome replication and repair (Mortusewicz et al. 2005;
Chen and Li 2006). In the complex of DNMT3A and DNMT3 L,
the latter has no catalytic activity and serves as a leading
coenzyme. DNMT3 L binds to permissive histone modifications
and recruits DNMT3A. Studies in mice found that DNMT3 L is
regulated by three different germline-specific promoters, pro-
ducing altered transcripts in stem cells (DNMT3 Ls), oocytes
(DNMT3 Lo) and adult testis (DNMT3 Lat). The transcript of
DNMT3 Lat is not translated, whereas DNMT3 Ls and DNMT3
Lo produce full-length proteins regulated by DNA methylation
(Suetake et al. 2004;O’Doherty et al. 2011;Saitou et al. 2012).
Materials and methods
Follicle isolation and oocyte collection
Oocytes were isolated from slaughterhouse ovaries as previ-
ously described (Wrenzycki et al. 2001;Heinzmann et al. 2011).
The donor animals were Holstein–Friesian dairy cows of
unknown age. Only ovaries showing a physiological morphol-
ogy and physiological structures and free of prominent corpora
lutea and pathological conditions such as cysts, were included in
this study. Briefly, ovaries were collected at a local abattoir on
13 different experimental days and transported to the laboratory
within 1.5–2 h at 308C in 0.9% NaCl containing 6 mgmL
1
penicillin and 50 mgmL
1
streptomycin and were washed three
times with 0.9% NaCl supplemented with penicillin and strep-
tomycin. Each experimental day, follicles from five ovaries
Epigenetics in oocytes related to follicle size follicles Reproduction, Fertility and Development 2041
were dissected with fine scissors and separated according to
size into three groups: Group A #2 mm (referred to as small),
Group B 3–5mm (referred to as medium) and group C $6mm
(referred to as large). Follicles $10 mm were discarded to avoid
bias of the results from cystic structures and potential atresia.
Follicles were then distributed into Petri dishes containing
phosphate-buffered saline (PBS) solution. Cumulus–oocyte
complexes (COCs) were isolated from follicles of the three size
categories by puncturing and pressing of the follicles with a bent
preparation needle followed by double washing in Dulbecco’s
PBS (Sigma Aldrich) supplemented with 0.33 mM Na-pyruvate,
5.56 mM glucose, 0.9 mM calcium chloride dihydrate,
50 mgmL
1
streptomycin, 6 mgmL
1
penicillin G, 4 IU L
1
heparin and 1 mg mL
1
bovine serum albumin (BSA; fraction V;
Sigma Aldrich). COCs were screened for a homogenous cyto-
plasm and at least three layers of surrounding cumulus cells and
were collected in TCM-air, pH 7.2 (TCM199; Sigma Aldrich)
containing 50 mgmL
1
gentamycin sulfate (Sigma Aldrich),
0.2 mM Na-pyruvate (Sigma Aldrich), 4.2mM NaHCO
3
(Roth)
and 1 mg mL
1
BSA (FAF; Sigma Aldrich). Cumulus cells were
removed by incubation in PBS containing 1 mg mL
1
BSA
(Fraction V; Sigma Aldrich) and 0.1% hyaluronidase (Sigma
Aldrich) for 5 min at 398C followed by vortexing (Vortex Genie)
at 1400 rpm for 5 min. The remaining cumulus cells were
removed by careful pipetting. Finally, immature oocytes were
washed three times in PBS containing 0.1% polyvinyl alcohol
(PVA) and frozen at 808C either in groups of 5–10 for meth-
ylation analysis or singly for subsequent quantitative reverse
transcription polymerase chain reaction (RT-qPCR).
DNA methylation analysis
The limiting dilution bisulfite (pyro)sequencing is a method for
analysing DNA methylation patterns of single DNA molecules,
using minimal amounts of DNA. Dilution of bisulfite-treated
DNA down to a single molecule before PCR enables the analysis
of single alleles and avoids amplification bias (El Hajj et al.
2011). These technical features render the limiting dilution
bisulfite (pyro)sequencing a suitable method for analysing low
numbers of oocytes at the level of single alleles.
Briefly, the DNA of pools with 5–10 oocytes was isolated
and bisulfite-converted with the EZ DNA Methylation Direct
Kit (Zymo Research Corporation). Bisulfite-treated DNA was
eluted with 14 mL elution buffer into a 1.5-mL LoBind tube
(Eppendorf). The internal coating of the LoBind tube consists of
special material that reduces sample-to-surface binding to avoid
possible loss of DNA due to interaction with the surface of the
tube. The multiplex PCR reaction amplified amplicons for all
seven genes (H19,SNRPN,ZAR1,OCT4,DNMT3A,DNMT3 Lo
and DNMT3 Ls) and contained all seven primer combinations.
The following second gene-specific nested PCR generated the
final amplicons for subsequent sequencing. For all PCR reac-
tions, the Fast Start Taq polymerase (Roche) was used in
combination with primers for bisulfite-converted DNA
designed with Pyrosequencing Assay Design Software (Qiagen;
Table 1). The mastermix for the multiplex PCR reaction was
prepared for 26 samples (including two spare samples to
compensate for pipetting errors) and was prepared for each pool
of oocytes separately. This mastermix contained 65 mL
10PCR Reaction Buffer (with 20 mM MgCl2), 13 mL PCR
Grade Nucleotide Mix, 26 mL of each primer (10 mM), 5.2 mL
FastStart Taq DNA Polymerase (5 U mL
1
) and 176.8 mL RN-
ase-free water. The mix was split keeping 100 mL in a separate
reaction tube as negative control and adding the remaining
volume to the bisulfite-converted oocyte DNA in the LoBind
tube. To ensure proper dispensation of single DNA molecules,
the solution was mixed thoroughly by extensive up and down
pipetting. The mix was distributed in volumes of 25 mL across 20
separate wells with the same pipette tip to avoid possible loss of
DNA caused by surface interaction. Thus each well contained
,0.25–0.5 oocyte equivalents. The PCR reaction was per-
formed according to the manufacturer’s instructions with an
annealing temperature of 548C and 34 cycles. For amplification
of each gene separately, 1 mL of the multiplex PCR product was
used as template in the subsequent nested PCR. The standard
PCR procedure was also applied for the nested PCR; only the
annealing temperature and cycle number varied between certain
genes (H19:658C, 27 cycles; SNRPN and ZAR1:578C, 30
cycles; OCT4 and DNMT3 Lo:548C, 32 cycles; DNMT3A and
DNMT3 Ls:648C, 31 cycles).
The products of nested PCR were then subjected to either
(pyro)sequencing or Sanger sequencing. The choice of the
sequencing method depended on the density and distribution
of CpG sites and length of the target region. Based on our
experimental experience, the maximum length of sequences to
analyse is 70 bp for (pyro)sequencing and 300 bp for Sanger
sequencing. Amplicons of DNMT3 Lo and OCT4 were pyrose-
quenced on a Pyromark Q96MD system (Qiagen) and analysed
with the Pyromark Q-CpG software. The genes H19,SNRPN,
ZAR1,DNMT3A and DNMT3 Ls were directly sequenced on an
Applied Biosystems 3130/3130xl Genetic Analyzer and ana-
lysed with BiQ Analyzer Version 2.00 (Bock et al. 2005).
General criteria and yields of methylation analysis
Molecular analysis of germ cells is prone to somatic cell con-
tamination. As control for somatic cell contamination the panel
of genes for methylation analysis contained imprinted genes
(H19 and SNRPN) and developmentally important, non-
imprinted genes (DNMT3 Lo and OCT4). H19 is paternally
imprinted and non-methylated in oocytes whereas SNRPN is
maternally imprinted and thus fully methylated in oocytes. The
methylation pattern of somatic cells would consist of a mixture
of methylated and non-methylated alleles for both genes.
DNMT3 Lo acts as oocyte-specific promoter of the DNMT3 L
gene and is non-methylated in the oocyte similar to the plur-
ipotency marker OCT4 (O’Doherty et al. 2011). DNMT3 Lo and
OCT4 are fully methylated in somatic tissues. All four genes can
serve as epigenetic markers for somatic contamination, but
could also indicate putative abnormal methylated alleles. We
defined a somatic contamination as the occurrence of one or
more abnormally methylated alleles (.50% abnormally meth-
ylated CpGs) in at least two of the four genes within the same
oocyte pool. By applying these criteria, we excluded 2 of 19
pools of oocytes extracted from small follicles and none of
the pools of oocytes from medium-sized (11 pools) and large
(eight pools) follicles due to somatic contamination.
2042 Reproduction, Fertility and Development F. Mattern et al.
Most of DNA molecules from bH19,bSNRPN,bZAR1,
bDNMT3A,bOCT4 and bDNMT3 Lo showed alleles that were
.80% or ,20% methylated (Fig. 1). Nevertheless, we defined a
single allele as methylated if more than 50% of the CpGs were
methylated and as non-methylated if less than 50% of the CpGs
were methylated. Single CpGs with methylation values .80%
were classified as methylated CpGs, whereas methylation
values ,20% were counted as non-methylated CpGs. CpGs
with methylation values between 20% and 80% were defined as
CpGs with an unclear methylation state, most likely due to a
mixture of at least two DNA molecules containing a methylated
and non-methylated CpG or a technical artefact. The amplicons
sequenced by direct bisulfite Sanger sequencing were assumed
to represent more than one molecule per reaction when an
apparently heterozygous signal for cytosine and thymine was
observed at the same sequencing position. For assessment of the
efficiency of the limiting dilution bisulfite (pyro)sequencing, we
calculated the allele recovery rate as the number of sequenced
alleles divided by the number of alleles in the starting sample.
Each oocyte comprised zona pellucida and germinal vesicle and
thus provided four amplifiable alleles for each gene. Hence, in a
pool of 10 oocytes, a maximum of 40 alleles could be amplified.
However, the high amount of DNA degradation caused by
bisulfite conversion and different PCR efficiencies reduces the
number of recovered alleles empirically to less than half of the
applied amount. In our study the overall allele recovery rate was
5.7%, ranging from 0.3% for ZAR1 in oocytes from follicles
smaller than 2 mm to 15.0% for DNMT3 Lo in the same category
of oocytes.
Messenger RNA isolation and reverse transcription (RT)
The isolation of poly(A)
þ
RNA from single oocytes (n¼8–13
oocytes per size group) was performed by using Dynabeads
mRNA DIRECT Kit (Life Technologies; Niemann et al. 2010;
Heinzmann et al. 2011). Briefly, immature oocytes were lysed
by adding 40 mL of lysis–binding buffer (100 mM Tris-HCl
pH 8.0, 500 mM LiCl, 10 mM ethylenediamine tetraacetic acid
(EDTA), 1% lithium dodecyl sulfate (LiDS), 5 mM dithio-
threitol (DTT)) and incubated at room temperature for 10 min.
Here, we used 0.5 pg rabbit globin mRNA (BRL) as exogenous
standard that was added to each preparation. For binding the
poly(A)
þ
RNA, 3 mL previously washed Dynabeads Oligo d(T)
25 were added to the lysate and incubated at 248C for 10 min on a
shaker. The beads with the linked poly(A)
þ
RNA were separated
Table 1. Primers used in multiplex PCR, nested PCR and (pyro)sequencing for bH19,bSNRPN,bZAR1,bDNMT3A,bDNMT3 Ls,bOCT4 and
bDNMT3 Lo in bovine oocytes from antral follicles with a diameter of #2 mm, 3–5 mm and $6mm
The promoter regions of bH19,bSNRPN,bZAR1,bDNMT3A and bDNMT3 Ls were directly sequenced by Sanger sequencing. The amplicons of DNMT3 Lo
and bOct4 were analysed by (pyro)sequencing. The inner forward primer of DNMT3 Lo and the inner reverse primer of bOct4 were biotinylated
Gene Primer Sequence (50–30) Amplicon length (bp) Chromosomal localisation Number of CpGs
bH19 Outer forward GAGGGGTATTGAGAGGTTGT 18
Outer reverse CAAACATAAAAATCCCTCAATATCCC 279 29:50147816–50148095
Inner forward AGAGGTTGTGGGTGTGGAGATA 230 29:50147827–50148057
Inner reverse TCCTCTCCCACCTTCAACAA
bSNRPN Outer forward GGGGTGGGGTAGATATTATTTT 30
Outer reverse AAAAAAAAAAAATATTACCCACCACAC 299 21:25818–26117
Inner forward GGTTTTTTTGTTTGAGAGAG 276 21:25818–26094
Inner reverse AAAAAAAAAAAATATTACCCAC
bZAR1 Outer forward ATTTGGGGTAAGTTTATTTTTAGATTAGT 18
Outer reverse CCCCAAAACAACCATCAATA 323 6:68820478–68820801
Inner forward GGGTGTGGAATATTTTTATATTAAGGT 230 6:68820567–68820797
Inner reverse AAAACAACCATCAATATACCCCTAC
bDNMT3A Outer forward GGGTATTTAATTTTATTTGGATATTT 16
Outer reverse CAAAACCCACTACACTCT 262 11:73963160–73963422
Inner forward TTTTAAGGGAGGTTTAATTAATAGAGGT 211 11:73963208–73963419
Inner reverse AACCCACTACACTCTACCT
bDNMT3 Ls Outer forward TAGGGGGGTTTTTGAGAA 8
Outer reverse CAACAAAAACACCCCACA 304 1:145717979–145718283
Inner forward AGGGGGGTTTTTGAGAAGAAGA 270 1:145717980–145718250
Inner reverse TCCACCCAAAACCCAATAACACTAA
bDNMT3 Lo Outer forward GTTTAAGTTTAGGGGATGG 5
Outer reverse CCTCCTTCTCCTTCAAAC 158 1:145750153–145750311
Inner forward TAAGTTTAGGGGATGGATATAG 88 1:145750156–145750244
Inner reverse TCCAAACCCCACTTCCTA
Pyro GGGATGGATATAGTTGTTT
bOct4 Outer forward GGGAGGTTTTTGGAAGTTTAT 8
Outer reverse TTCCCCCCACCCATCCAA 262 23:27769756–27770018
Inner forward GGGAGGTTTTTGGAAGTTTAT 143 7:51481766–51481909
Inner reverse AAATACCTTCCTTCCCCATA
Pyro CCTTCCTTCCCCATAA
Epigenetics in oocytes related to follicle size follicles Reproduction, Fertility and Development 2043
using a Dynal MPC-E-1 magnetic separator. After washing the
beads and the bound mRNA three times, the mRNA was eluted
in 11 mL of sterile water by incubation at 658C for 2.5 min. The
eluted mRNA was immediately used for reverse transcription
(RT; Heinzmann et al. 2011).
RT was carried out in a total volume of 20 mL with 10RT
reaction buffer, 5 mM MgCl
2
, 1 mM dNTP solution, 2.5 mM
random hexamer primers, 20 Units of RNAin and 50 U murine
leukaemia virus (MuLV) reverse transcriptase (all Life Tech-
nologies) and the entire mRNA sample. For RT reaction the
tubes were incubated in a thermal cycler at 258C for 10 min for
primer annealing, followed by 1 h at 428C for elongation and a
final step of 5 min at 958C.
The cDNA generated (0.05 oocyte equivalents per mL) was
diluted to a concentration of 0.025 oocyte equivalents per mL for
PEG3,IGF2R,PRDX1 and DNMT1B,DNMT3A and was left
undiluted (0.05 oocyte equivalents per mL) for SNRPN,HSF1,
NLRP9 and ZAR1 analysis.
qPCR
For quantification of the gene-specific cDNA, quantitative real-
time PCR was performed as described (Heinzmann et al. 2011;
Ulloa et al. 2015). Briefly, the reaction volume of 20 mL con-
tained 10 mL2Power SYBRGreen PCR Master Mix (Life
Technologies), 0.8 mL of each of the forward and reverse pri-
mers (5 mM), 2 mL cDNA and 6.4 mLdH
2
O. The sequences of all
primers used for qPCR are listed in Table 2. The qPCR samples
were measured on an ABI 7500 Fast Real-Time System
(Applied Biosystems) with a protocol of 10 min at 958C,
40 cycles of 15 s at 958C and 1 min at 608C, followed by a slow
heating cycle for obtaining a dissociation curve. A cDNA
dilution standard of pooled blastocyst mRNA was included on
every plate to provide standard curves for each individual gene.
Standard curves were used for the calculation of the relative
concentration of each target gene to be normalised to the signal
from globin mRNA that had been included into the mRNA
samples as a control. For quantification the Sequence Detection
Software 1.4.1 (Applied Biosystems) was used.
Statistical analysis
The differences in the number of abnormally methylated alleles
between the oocytes extracted from follicles of the three
different size categories (#2 mm, 3–5 mm and $6 mm) were
statistically tested by Freeman–Halton test. Kruskal–Wallis
one-way ANOVA was used to evaluate gene expression data.
Levels of P,0.05 were considered to be statistically
significant.
Results
Increased abnormal methylation in oocytes from small
antral follicles
Overall, we analysed 466 alleles with a total of 6318 CpGs in
343 oocytes. The majority of the detected alleles were either
fully methylated or completely unmethylated (68%) (Table 3,
Fig. 1). Analysis of DNA methylation by limiting dilution
ⱕ2 mm ⱖ6 mm
3–5 mm
bH19
ⱕ2 mm ⱖ6 mm
3–5 mm
bSNRPN
ⱕ2 mm ⱖ6 mm
3–5 mm
bZAR1
ⱕ2 mm ⱖ6 mm
3–5 mm
bOct4
ⱕ2 mm ⱖ6 mm
3–5 mm
bDNMT3A
ⱕ2 mm ⱖ6 mm
3–5 mm
bDNMT3Lo
Fig. 1. Overview of DNA methylation results of single alleles for bH19,bSNRPN,bZAR1,bDNMT3A,bDNMT3
Ls,bOCT4 and bDNMT3 Lo in bovine oocytes from antral follicles with a diameter of #2 mm, 3–5 mm and $6 mm.
One row represents a single allele with single CpGs as boxes. A white box stands for an unmethylated CpG, a black
box for a methylated CpG and a grey box for an unclear methylation state. Alleles containing grey CpGs were
excluded from statistical analysis.
2044 Reproduction, Fertility and Development F. Mattern et al.
bisulfite (pyro)sequencing revealed no major differences
(P.0.05) between oocytes isolated from follicles of #2 mm,
3–5 mm and $6 mm for bH19,bSNRPN,bZAR1,bDNMT3A,
bOCT4 and bDNMT3 Lo (Table 3). In addition to fully meth-
ylated and completely unmethylated alleles, we observed an
increased frequency (32%) of DNA molecules containing CpGs
with an unclear methylation status (Table 3,Fig. 1). When
considering only alleles with clearly methylated and unmethy-
lated CpGs, 3 out of 299 alleles (,1%) displayed aberrant
methylation profiles (i.e. .50% of CpGs with an aberrant
methylation profile), including one aberrant molecule in bH19
and two aberrantly methylated alleles in bSNRPN. Taking into
account all alleles analysed, including CpGs with an unclear
methylation status (n¼466), seven alleles in oocytes of small
follicles (#2 mm) and one allele in oocytes derived from
medium and large size (3–5 mm, $6 mm) follicles showed
aberrant methylation (Table 3,Fig. 1). Oocytes derived from
follicles #2 mm contained two differentially methylated
molecules in the imprinted genes bH19 and bSNRPN and three
aberrant alleles in DNMT3 Lo. Aberrantly methylated alleles
were found in oocytes derived from medium and large size
follicles in bH19 (3–5 mm) and bDNMT3 Lo ($6 mm).
Increased frequency of CpG sites with unclear methylation
profile in bDNMT3 Ls
DNMT3 L is the catalytically inactive coenzyme of DNMT3A
(Hata et al. 2002;Jia et al. 2007). Based on studies in mice
(O’Doherty et al. 2011), we provide here the first methylation
data for the orthologous promoter regions of the two transcripts
of DNMT3 L, DNMT3 Lo and DNMT3 Ls, in immature bovine
oocytes from follicles of different sizes (#2 mm, 3–5 mm and
$6 mm). Analysis of the bovine DNMT3 Ls revealed a mixture
of methylated and non-methylated CpG sites in single alleles for
all three groups of oocytes. In addition to the methylated and
non-methylated status, we observed an increased number of
CpG sites with an unclear methylation status in all categories of
female germ cells (Fig. 2a). Specifically, CpG7 and CpG8 had
an unclear methylation status for single CpG sites in all analysed
alleles of oocytes from large antral follicles ($6 mm). Statistical
testing revealed a significantly increased frequency of this sig-
nal at CpG2, CpG7 and CpG8 in oocytes isolated from large
antral follicles ($6 mm) compared with their counterparts iso-
lated either from small (#2 mm) or medium follicles (3–5mm;
Table 4). The analysed region spans the transcription start site
with 134 bp upstream and 169 bp downstream (Fig. 2b). CpG1–4
belong to the first exon and CpG5–8 are located 110 bp upstream
of the transcription start site. To evaluate if the CpG sites of the
first exon are differentially methylated compared with CpG sites
upstream of the transcription start site, the methylation results of
the neighbouring four CpGs were combined and reanalysed
separately for the two regions of the split amplicon. This com-
parison revealed a significant increase of the number of single
CpG sites with an unclear methylation status for CpG1–4 and
CpG5–8 in oocytes from large follicles ($6 mm) compared with
medium follicles (3–5 mm; P
CpG1–4
,0.002, P
CpG5–8
,0.0001).
This unclear methylation pattern was also significantly increased
in oocytes extracted from small follicles (#2 mm) compared
with oocytes isolated from large follicles ($6 mm) for CpG5–8
(P
CpG5–8
,0.0001).
Reduced transcript levels in oocytes from antral follicles
#2 mm and $6mm
For analysis of relative poly(A)
þ
mRNA abundance in single
oocytes from follicles of #2 mm, 3–5 mm and $6 mm size, we
included genes critically involved in the oxidative stress
response (bPRDX1), DNA methylation (bDNMT1B and
Table 2. Primers used for qPCR
Gene Sequence (50-30) Amplicon length (bp) Accession number
bPRDX1 TCAAGCCTGATGTCCAGAAGAGC 174 NM_174431.1
CCGTCCTGTCCCACACCAC
bDNMT1B CAGTTCACATATCAAAGTACCAGC 240 EU273277
GGCATCAAGATGGACAACC
bDNMT3A ACTATACCGACGTCTCCAACATGA 57 XM_001252215
ACCGGCCCAGCAGTCTCT
bIGF2R CTACGACCTGACCGAGTG 95 NM_174352
TGACAGCCTCCCAGTTG
bPEG3 CCTTGGCAGACATCAGAAGATCT 61 NM_001002887
TCCAAACAGCTTCCCATCATG
bSNRPN CACCAAGAGGTGGTTAAAGCAGTA 80 NM_001079797
AACAATGCAAGCTGAGGCAGAAG
bHSF1 AAGATTCGCCAGGACAGTGTTACC 259 NM_001076809.1
CGCCGTCGTTCAGCATCAGG
bNLRP9 GCGGCGGTGCTGTGTGAAG 173 NM_001024664.1
CTGCGTCTGCCCTCGTCATC
bZAR1 TGCCGAACATGCCAGAAG 170 NM_001076203.1
TCACAGGATAGGCGTTTGC
Globin (rabbit a-1 globin) GCAGCCACGGTGGCGAGTAT 256 Oryctolagus cuniculus
GTGGGACAGGAGCTTGAAAT NM_001082389.2
Epigenetics in oocytes related to follicle size follicles Reproduction, Fertility and Development 2045
bDNMT3A), imprinting (bSNRPN,bPEG3 and bIGF2R) and
maternal effects (bZAR1,bHSF1 and bNLRP9). We observed
reduced transcript levels, albeit not significant for eight genes
(bPRDX1,bDNMT1B,bDNMT3A,bPEG3,bIGF2R,bZAR1,
bHSF1 and bNLRP9) in oocytes from small (#2 mm) and large
follicles ($6 mm) compared with oocytes from medium folli-
cles (3–5 mm; Fig. 3). BSNRPN was expressed at similar levels
in female germ cells from small (#2 mm) and medium follicles
(3–5 mm) and was reduced (P.0.05) in oocytes from large
follicles ($6 mm).
Discussion
Acquisition of mammalian oocyte developmental competence
occurs in a complex, well-orchestrated manner within the
growing follicle, ultimately resulting in ovulation of a mature,
fertilisable oocyte. Epigenetic mechanisms play an essential
role in the development of germ cells and for the early embryo as
well (Monk 2015). Here, we analysed the DNA methylation
profiles of seven candidate genes and mRNA expression of nine
selected genes in oocytes from antral follicles of three different
size categories in the bovine model. Our results uncovered
Table 3. DNA methylation results for bH19,bSNRPN,bZAR1,bDNMT3A,bDNMT3 Ls,bOct4 and bDNMT3 Lo in bovine oocytes from antral
follicles with a diameter of #2 mm, 3–5 mm and $6mm
No significant differences between the size groups in the methylation of whole alleles and single CpG errors were detected for any gene
Gene No. CpG sites
in assay
Parameter Follicle sizes Sum of all groups
#2 mm 3–5 mm $6mm
bH19 18 No. of alleles analysed 33 40 16 89
No. of alleles with unclear CpGs 19 (58%) 26 (65%) 10 (63%) 55 (62%)
No. of abnormal alleles (unclear alleles excluded) 1 (7%) 0 0 1 (3%)
No. of abnormal alleles (unclear alleles included) 2 (6%) 1 (3%) 0 3 (3%)
No. of single CpG errors 3 (1%) 0 0 3 (0%)
Allele recovery rate 0.05 0.10 0.06 0.06
bSNRPN 30 No. of alleles analysed 15 8 8 31
No. of alleles with unclear CpGs 2 (13%) 6 (75%) 4 (50%) 12 (39%)
No. of abnormal alleles (unclear alleles excluded) 2 (15%) 0 0 2 (11%)
No. of abnormal alleles (unclear alleles included) 2 (13%) 0 0 2 (7%)
No. of single CpG errors 2 (1%) 0 2 (2%) 4 (1%)
Allele recovery rate 0.02 0.02 0.03 0.02
bOct4 8 No. of alleles analysed 23 25 28 76
No. of alleles with unclear CpGs 1 (4%) 3 (12%) 3 (11%) 7 (9%)
No. of abnormal alleles (unclear alleles excluded) 0 0 0 0
No. of abnormal alleles (unclear alleles included) 0 0 0 0
No. of single CpG errors 6 (3%) 5 (3%) 3 (2%) 14 (3%)
Allele recovery rate 0.03 0.06 0.10 0.06
bDNMT3A 16 No. of alleles analysed 16 37 21 74
No. of alleles with unclear CpGs 6 (38%) 18 (49%) 11 (52%) 35 (47%)
No. of abnormal alleles (unclear alleles excluded) 0 0 0 0
No. of abnormal alleles (unclear alleles included) 0 0 0 0
No. of single CpG errors 2 (1%) 8 (3%) 4 (3%) 14 (2%)
Allele recovery rate 0.02 0.09 0.08 0.05
bDNMT3 Lo 5 No. of alleles analysed 104 28 26 158
No. of alleles with unclear CpGs 12 (12%) 0 6 (23%) 18 (11%)
No. of abnormal alleles (unclear alleles excluded) 0 0 0 0
No. of abnormal alleles (unclear alleles included) 3 (1%) 0 1 (4%) 4 (3%)
No. of single CpG errors 4 (1%) 0 2 (2%) 6 (1%)
Allele recovery rate 0.15 0.07 0.09 0.12
bZAR 18 No. of alleles analysed 2 27 9 38
No. of alleles with unclear CpGs 2 (100%) 11 (41%) 7 (78%) 20 (53%)
No. of abnormal alleles (unclear alleles excluded) 0 0 0 0
No. of abnormal alleles (unclear alleles included) 0 0 0 0
No. of single CpG errors 0 1 (0%) 0 1 (0%)
Allele recovery rate 0.003 0.07 0.03 0.03
Sum of all genes 95 No. of alleles analysed 193 165 108 466
No. of alleles with unclear CpGs 42 (22%) 64 (39%) 41 (38%) 147 (32%)
No. of abnormal alleles (unclear alleles excluded) 3 (2%) 0 0 3 (1%)
No. of abnormal alleles (unclear alleles included) 7 (4%) 1 (1%) 1 (1%) 9 (2%)
No. of single CpG errors 17 (1%) 14 (1%) 11 (1%) 42 (1%)
Allele recovery rate 0.05 0.07 0.07 0.06
2046 Reproduction, Fertility and Development F. Mattern et al.
highly dynamic changes in DNA methylation related to the
different size categories of antral follicles, indicating a potential
time window for methylation of distinct gene regions in devel-
oping bovine oocytes.
Most of our DNA methylation analysis for bH19,bSNRPN,
bZAR1,bDNMT3A,bOct4 and bDNMT3 Lo showed similar
methylation patterns in all three categories of oocytes. In
contrast to our recent studies using limiting dilution bisulfite
(pyro)sequencing (Heinzmann et al. 2015;Mattern et al. 2016),
here we observed an increased number of CpGs with an unclear
methylation status. Typically, a CpG site with an unclear
methylation status is characterised by a sequencing signal for
cytosine and thymine at the same base position. Depending on
its methylation status, a single DNA molecule analysed by
limiting dilution bisulfite (pyro)sequencing should display
either a signal for cytosine or for thymine at each analysed
CpG site. A signal for both bases at the same site is not consistent
with amplification of one single DNA molecule. In previous
studies, such ambiguous patterns occurred only rarely, mainly
due to failed bisulfite conversion or failed allele separation
(DNA clumping) and usually were considered as technical
artefacts. Here, we observed an increased number of alleles
with unclear methylation patterns, while the frequency of
aberrantly methylated alleles was not significantly different
between follicle groups, regardless of whether alleles with
unclear CpGs were included or not. In both analyses, the number
of aberrantly methylated alleles in oocytes of small follicles
(#2 mm) was increased compared with oocytes from medium
and large (3–5 mm, $6 mm) follicles. It is well known that
bovine oocytes from follicles #2 mm have a significantly lower
competence to undergo in vitro maturation and regular fertilisa-
tion and lack the capability to cleave beyond the eight-cell stage
(Pavlok et al. 1992). Similarly, a higher developmental compe-
tence was reported for oocytes from large ($6 mm) versus
medium follicles (#4 mm; Lequarre et al. 2004,2005).
(a)
(b)
ⱖ2 mm
1
100
2345678
12
234 56781
⫹16912 bp 19 bp 1 bp 137 bp 0 bp 8 bp 12 bp
⫺134
⫹1
1–4 5–8 1–4 5–8
∗
∗
∗
∗
1–4 5–8
345678 12345678 12345678
12345678 12345678
ⱖ6 mm
3–5 mm
80
60
40
Methylation state [%]Methylation state [%]
20
0
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
CGCGCAGGATGACGCAGGGATGAGAGCG
bDNMT3Ls
CREB binding site
Fig. 2. Overview of DNA methylation results for bDNMT3 Ls in bovine
oocytes from antral follicles with a diameter of #2 mm, 3–5 mm and
$6 mm. (a) One row represents a single allele with single CpGs as boxes.
A white box represents an unmethylated CpG, a black box a methylated CpG
and a grey box an unclear methylation status. The bar diagrams show the
relative methylation status for each CpG site and the combined exonic and
upstream located CpGs. The frequency of CpGs with an unclear methylation
status is significantly increased at CpG2, CpG7 and CpG8 in oocytes from
follicles $6 mm (*) compared with oocytes from smaller follicles (#2mm
and 3–5 mm). (b) Analysed region of DNMT3 Ls with the location of the
CpGs on the amplicon and the CREB-binding site at CpG7 (indicated
in red).
Table 4. DNA methylation results for bDNMT3 Ls in bovine oocytes
from antral follicles with a diameter of #2 mm, 3–5 mm and $6mm
M, number of CpGs with a methylated state; U, number of unmethylated
CpGs; M/U, number of unclear methylation states. The frequency of CpGs
with an unclear methylation state is significantly increased at CpG2, CpG7
and CpG8 in oocytes from follicles $6 mm compared with oocytes from
smaller follicles (#2 mm: P
CpG2
,0.03, P
CpG7
,0.0001, P
CpG8
,0.0001
and 3–5 mm: P
CpG2
,0.005, P
CpG7
,0.0008, P
CpG8
,0.0001; Freeman–
Halton test)
Follicle size Methylation state No. CpGs at each CpG position
12345678
#2mm M 516 1 3 811 713
M/U 21011342
U 15 5 21 18 13 8 11 7
3–5 mm M 11 22 0 2 10 14 7 13
M/U 010100110
U 20 8 31 28 21 17 13 18
$6mm M 15033400
M/U 1 5 1 0 1 2 12 12
U 1021198600
Epigenetics in oocytes related to follicle size follicles Reproduction, Fertility and Development 2047
In general, oocyte diameter correlates with follicle size.
However, this correlation is weak and only 10 to 15% of large
follicles also contain large oocytes and vice versa. Nevertheless,
similar to the correlation of developmental competence with
size of the follicle, the developmental capacity of the oocytes
increases with a larger diameter of the oocyte. The highest
increase with regard to meiotic progression, cleavage and
blastocyst formation was observed in oocytes with a diameter
of 100–110 mm and 110–120 mm, compared with other size
categories, indicating a critical size window for acquiring full
developmental competence (Fair 2003). Bovine oocytes acquire
full meiotic competence at a diameter of 110 mm, when they do
not yet show transcriptional activity (Fair et al. 1995). The
oocyte stops growing at a diameter of 120 mm when the follicle
reaches a size of 3 mm. The critical size of 110 mm is attained
at a follicle size of ,2mm (Fair et al. 1995). The increased
number of alleles with aberrant methylation in developmentally
important genes of oocytes from small antral follicles (#2 mm)
did not reach the level of statistical significance compared with
the two larger follicle categories. Consistent with the develop-
mentally critical follicle size of 2 mm, this finding may be related
to the acquisition of methylation imprints.
Previously, DNA methylationprofiles for the imprinted genes
SNRPN,MEST,IGF2R,PEG10 and PLAGL1 were analysed
in different size oocytes, i.e. 101–110 mm, 110–120 mmand
$120 mm(O’Doherty et al. 2012). A gradually increasing level
of size- and gene-specific methylation was found, with the
biggest differences in DNA methylation of SNRPN,PEG10
and PLAGL1 between oocytes smaller and bigger than 110mm.
Female germ cells smaller than 110 mm have a reduced develop-
mental competence and most likely have not yet fully established
the DNA methylation profiles in imprinted genes. A possible
explanation for this observation could be that oocytes from
follicles smaller than 2 mm and germ cells with a diameter
smaller than 110 mm represent a critical intermediate stage of
the developing oocyte pool. The finding of abnormally methyl-
ated alleles of imprinted genes at this growth phase probably
reflects a natural intermediate status of developing oocytes and
does not constitute a pathological condition. Since the process of
epigenetic reprogramming is not yet completed, oocytes of this
size still have the potential to acquire the correct methylation
profile and concomitantly full developmental competence.
The dynamics of DNA methylation during oocyte develop-
ment are probably also reflected by the present methylation
results for DNMT3 Ls. The methylation assay employed in
the present study has been successfully used in a recent study,
where we analysed the effects of extended in vitro maturation in
bovine oocytes on critical epigenetic parameters and found a
significant hypermethylation of DNMT3 Lo in female germ
cells matured for 48 h (extended maturation) compared with
24 h maturation (Heinzmann et al. 2015). Furthermore, we
detected significant hypomethylation of DNMT3 Ls at a poten-
tial CREB (cAMP response element-binding protein) binding
site,located in CpG7, in embryos developed from oocytes
in vitro matured for 48 h. Here, we used this assay to analyse
DNMT3 Ls in immature oocytes from follicles of different size
and determined an increased frequency of CpG sites with an
unclear methylation profile in oocytes from all three follicle size
categories (Fig. 2a). The most dramatic increase of unclear
CpGs was recognised for CpG7 and CpG8 in germ cells from
follicles bigger than 6 mm. A significant increase of these CpGs
was discovered for CpG2, CpG7 and CpG8 in oocytes from
follicles $6 mm compared with oocytes from small and medium
follicles (#2 mm and 3–5 mm).
The observation of unclear signals at defined CpG sites,
including one potential regulatory position, favours a biological
explanation rather than technical reasons, and we assume that
the unclear methylation features were induced by the molecular
machinery responsible for epigenetic regulation. The commonly
used standard protocols for bisulfite sequencing include protein
4.5
3.5
3.0
Relative transcript abundance
2.5
2.0
1.5
1.0
0.5
0
PRDX1
DNMT1B
DNMT3A
IGF2R
PEG3
SNRPN
HSF1
NLRP9
ZAR1
4.0 ⬍2 mm
⬎6 mm
3᎐5 mm
Fig. 3. Relative poly(A)
þ
mRNA expression results for bPRDX1,bDNMT1B,bDNMT3A,bPEG3,bIGF2R,bZAR,bHSF1 and bNLRP9 in bovine oocytes
from antral follicles with a diameter of #2 mm, 3–5 mm and $6 mm. Mean expression s.e.m. is shown. No significant differences could be detected
(P.0.05).
2048 Reproduction, Fertility and Development F. Mattern et al.
removal between DNA isolation and addition of the bisulfite
reagent. In most cases, including the present protocol, this step is
performed by proteinase K treatment. Proteins attached to the
DNA may interfere with the bisulfite reagent and thereby may
influence the extent of conversion. Although we applied pro-
teinase K digestion and did not observe any non-CpG methyla-
tion, we propose that the unclear signals at distinct CpGs were
caused by factors related to DNA methylation that were not
removed by proteinase K. This suggestion is supported by
experimental evidence in DU145 cells, in which a genomic
region around theGSTP1 (Glutathione S-Transferase Pi 1)
promoter was sequenced after treatment with and without
proteinase K digestion before bisulfite conversion. A compari-
son of the sequences of both treatments revealed a high level of
non-methylated CpGs and partial conversion at random sites for
DNA that had not been treated with proteinase K (Warnecke
et al. 2002).
The present findings indicate that the methylation status of
CpG2, CpG7 and CpG8 is crucial for regulation of DNMT3 Ls,
which may undergo methylation changes during oogenesis.
These changes in DNA methylation correlate with increasing
size of the follicle and the oocyte and reach a climax in oocytes
from follicles larger than 6 mm. Thus, we assume that factors of
the DNA methylation machinery are binding to these regulatory
CpGs and interfere with the reagents of the bisulfite mix,
causing the observed unclear signals for single CpGs.
The present results regarding mRNA expression of nine
selected genes revealed reduced transcript levels for eight genes
(bPRDX1,bDNMT1B,bDNMT3A,bPEG3,bIGF2R,bZAR,
bHSF1 and bNLRP9) in oocytes from small (#2 mm) and large
follicles ($6 mm) relative to oocytes from medium follicles
(3–5 mm). BSNRPN was expressed at lower levels in oocytes
from large follicles compared with similar expression levels in
oocytes from small (#2 mm) and medium follicles (3–5 mm).
It has been shown that RNA synthesis is diminished in oocytes
bigger than 110 mm compared with smaller oocytes. The highest
expressionlevel was determined in oocytes ,110 mm, correlating
with a follicle size of ,3mm(Fair et al. 1995). This is in contrast
to the present findings. A possible reason for these diverging
findings might be different approaches by which RNA expres-
sion was measured. Previously, autoradiographic labelling was
used to measure global RNA synthesis, whereas in the present
study mRNA was isolated, reverse transcribed and subjected to
gene-specific quantitative real-time PCR. Previously, mRNA
levels of 13 selected genes were measured with the aid of real-
time PCR in oocytes extracted from follicles with diameters of
,3 mm, 3–5 mm, 5–8 mm and .8 mm and gene-specific find-
ings were made. While four genes showed no differences in
expression among the four categories of oocytes (GDF9,
CCNB1,STK6,BMP15), a sharp increase in oocytes from
follicles .8 mm in size was determined for PSMB2,SKIIP,
CDC5 L,RGS16,PRDX1 and four genes displayed a gradual
increase as follicular size increased (H2A,CKS1,PTTG1,
CCNB2;Mourot et al. 2006). PRDX1 was also included in our
panel of genes and our findings were similar to the previous
study. Similar to our findings, no significant differences in
oocytes isolated from follicles with a size from ,3mm to
8 mm (9 of 13 genes) were observed, which reflects similar
categories of follicles as in our study (#2mm to $6 mm).
Collectively, these findings indicate that the majority of genes,
including the candidates analysed in the present study, are not
differentially expressed in follicles with a size between #2mm
and 8 mm.
In conclusion, in the present study we examined DNA
methylation profiles and mRNA expression of a panel of genes
in oocytes isolated from three different size categories of
growing bovine follicles. We detected a non-significant increase
of aberrantly methylated alleles in the developmentally impor-
tant genes H19,SNRPN and DNMT3 Lo of oocytes isolated from
small antral follicles (#2 mm). In combination with the well-
known reduced developmental competence of oocytes from
small (#2 mm) follicles, this observation indicates that a follicle
diameter of 2 mm is a crucial growth stage for acquisition of the
physiological methylation profile and developmental compe-
tence. Furthermore, we recognised an increased frequency of
CpG sites with an unclear methylation status for DNMT3 Ls,
specifically in oocytes from follicles $6 mm. These unclear
signals were predominantly found at three CpG positions
(CpG2, CpG7 and CpG8), including one potential regulatory
site (CpG7). Overall, no major differences were determined
with regard to mRNA expression, indicating that the transcrip-
tional machinery is silent during major phases of folliculogen-
esis. These results point to a new time window for epigenetic
changes at a regulatory locus with developmental implications
for the growing bovine oocyte.
References
Amor, D. J., and Halliday, J. (2008). A review of known imprinting
syndromes and their association with assisted reproduction technologies.
Hum. Reprod. 23(12), 2826–2834. doi:10.1093/HUMREP/DEN310
Blondin, P., and Sirard, M. A. (1995). Oocyte and follicular morphology as
determining characteristics for developmental competence in bovine
oocytes. Mol. Reprod. Dev. 41(1), 54–62. doi:10.1002/MRD.
1080410109
Bock, C., Reither, S., Mikeska, T., Paulsen, M., Walter, J., and Lengauer, T.
(2005). BiQ Analyzer: visualization and quality control for DNA
methylation data from bisulfite sequencing. Bioinformatics 21(21),
4067–4068. doi:10.1093/BIOINFORMATICS/BTI652
Cai, X., and Cullen, B. R. (2007). The imprinted H19 noncoding RNA is a
primary microRNA precursor. RNA 13(3), 313–316. doi:10.1261/RNA.
351707
Chen, T., and Li, E. (2006). Establishment and maintenance of DNA
methylation patterns in mammals. Curr. Top. Microbiol. Immunol.
301, 179–201. doi:10.1007/3-540-31390-7_6
Chen, Z., Robbins, K. M., Wells, K. D., and Rivera, R. M. (2013). Large
offspring syndrome: a bovine model for the human loss-of-imprinting
overgrowth syndrome Beckwith–Wiedemann. Epigenetics 8(6),
591–601. doi:10.4161/EPI.24655
Crozet, N., Ahmed-Ali, M., and Dubos, M. P. (1995). Developmental
competence of goat oocytes from follicles of different size categories
following maturation, fertilization and culture in vitro.J. Reprod. Fertil.
103(2), 293–298. doi:10.1530/JRF.0.1030293
De La Fuente, R. (2006). Chromatin modifications in the germinal vesicle
(GV) of mammalian oocytes. Dev. Biol. 292(1), 1–12. doi:10.1016/
J.YDBIO.2006.01.008
El Hajj, N., Trapphoff, T., Linke, M., May, A., Hansmann, T., Kuhtz, J.,
Reifenberg, K., Heinzmann, J., Niemann, H., Daser, A., Eichenlaub-
Ritter, U., Zechner, U., and Haaf, T. (2011). Limiting dilution bisulfite
Epigenetics in oocytes related to follicle size follicles Reproduction, Fertility and Development 2049
(pyro)sequencing reveals parent-specific methylation patterns in single
early mouse embryos and bovine oocytes. Epigenetics 6(10),
1176–1188. doi:10.4161/EPI.6.10.17202
Fair, T. (2003). Follicular oocyte growth and acquisition of developmental
competence. Anim. Reprod. Sci. 78(3–4), 203–216. doi:10.1016/S0378-
4320(03)00091-5
Fair, T., Hyttel, P., and Greve, T. (1995). Bovine oocyte diameter in relation
to maturational competence and transcriptional activity. Mol. Reprod.
Dev. 42(4), 437–442. doi:10.1002/MRD.1080420410
Favaedi, R., Shahhoseini, M., and Akhoond, M. R. (2012). Comparative
epigenetic analysis of Oct4 regulatory region in RA-induced differenti-
ated NT2 cells under adherent and non-adherent culture conditions. Mol.
Cell. Biochem. 363(1–2), 129–134. doi:10.1007/S11010-011-1165-Y
Goudet, G., Bezard, J., Duchamp, G., Gerard, N., and Palmer, E. (1997).
Equine oocyte competence for nuclear and cytoplasmic in vitro matura-
tion: effect of follicle size and hormonal environment. Biol. Reprod. 57
(2), 232–245. doi:10.1095/BIOLREPROD57.2.232
Hagemann, L. J., Beaumont, S. E., Berg, M., Donnison, M. J., Ledgard, A.,
Peterson, A. J., Schurmann, A., and Tervit, H. R. (1999). Development
during single IVP of bovine oocytes from dissected follicles: interactive
effects of estrous cycle stage, follicle size and atresia. Mol. Reprod. Dev.
53(4), 451–458. doi:10.1002/(SICI)1098-2795(199908)53:4,451::AID-
MRD11.3.0.CO;2-3
Hansmann, T., Heinzmann, J., Wrenzycki, C., Zechner, U., Niemann, H.,
and Haaf, T. (2011). Characterization of differentially methylated
regions in 3 bovine imprinted genes: a model for studying human
germ-cell and embryo development. Cytogenet. Genome Res. 132(4),
239–247. doi:10.1159/000322627
Hata, K., Okano, M., Lei, H., and Li, E. (2002). Dnmt3L cooperates with the
Dnmt3 family of de novo DNA methyltransferases to establish maternal
imprints in mice. Development 129(8), 1983–1993.
Heinzmann, J., Hansmann, T., Herrmann, D., Wrenzycki, C., Zechner, U.,
Haaf, T., and Niemann, H. (2011). Epigenetic profile of developmentally
important genes in bovine oocytes. Mol. Reprod. Dev. 78(3), 188–201.
doi:10.1002/MRD.21281
Heinzmann, J., Mattern, F., Aldag, P., Bernal-Ulloa, S. M., Schneider, T.,
Haaf, T., and Niemann, H. (2015). Extended in vitro maturation affects
gene expression and DNA methylation in bovine oocytes. Mol. Hum.
Reprod. 21(10), 770–782. doi:10.1093/MOLEHR/GAV040
Hiura, H., Obata, Y., Komiyama, J., Shirai, M., and Kono, T. (2006). Oocyte
growth-dependent progression of maternal imprinting in mice. Genes
Cells 11(4), 353–361. doi:10.1111/J.1365-2443.2006.00943.X
Ideraabdullah, F. Y., Vigneau, S., and Bartolomei, M. S. (2008). Genomic
imprinting mechanisms in mammals. Mutat. Res. 647(1–2), 77–85.
doi:10.1016/J.MRFMMM.2008.08.008
Jia, D., Jurkowska, R. Z., Zhang, X., Jeltsch, A., and Cheng, X. (2007).
Structure of Dnmt3a bound to Dnmt3L suggests a model for de novo
DNA methylation. Nature 449(7159), 248–251. doi:10.1038/
NATURE06146
Kanitz, W. (2003). Follicular dynamic and ovulation in cattle—a review.
Arch. Tierz. Dummerstorf 46(2), 187–198.
Kuhtz, J., Romero, S., De Vos, M., Smitz, J., Haaf, T., and Anckaert, E.
(2014). Human in vitro oocyte maturation is not associated with
increased imprinting error rates at LIT1, SNRPN, PEG3 and GTL2.
Hum. Reprod. 29(9), 1995–2005. doi:10.1093/HUMREP/DEU155
Kurosaka, S., Eckardt, S., and McLaughlin, K. J. (2004). Pluripotent
lineage definition in bovine embryos by Oct4 transcript localization.
Biol. Reprod. 71(5), 1578–1582. doi:10.1095/BIOLREPROD.104.
029322
Lequarre, A. S., Traverso, J. M., Marchandise, J., and Donnay, I. (2004).
Poly(A) RNA is reduced by half during bovine oocyte maturation
but increases when meiotic arrest is maintained with CDK inhibitors.
Biol. Reprod. 71(2), 425–431. doi:10.1095/BIOLREPROD.103.026724
Lequarre, A. S., Vigneron, C., Ribaucour, F., Holm, P., Donnay, I., Dalbies-
Tran, R., Callesen, H., and Mermillod, P. (2005). Influence of antral
follicle size on oocyte characteristics and embryo development in the
bovine. Theriogenology 63(3), 841–859. doi:10.1016/J.THERIOGEN
OLOGY.2004.05.015
Lonergan, P., Monaghan, P., Rizos, D., Boland, M. P., and Gordon, I. (1994).
Effect of follicle size on bovine oocyte quality and developmental
competence following maturation, fertilization, and culture in vitro.
Mol. Reprod. Dev. 37(1), 48–53. doi:10.1002/MRD.1080370107
Lucifero, D., Suzuki, J., Bordignon, V., Martel, J., Vigneault,C., Therrien, J.,
Filion, F., Smith, L. C., and Trasler, J. M. (2006). Bovine SNRPN
methylation imprint in oocytes and day 17 in vitro-produced and somatic
cell nuclear transferembryos. Biol. Reprod. 75(4), 531–538. doi:10.1095/
BIOLREPROD.106.051722
Mattern, F., Herrmann, D., Heinzmann, J., Hadeler, K. G., Bernal-Ulloa,
S. M., Haaf, T., and Niemann, H. (2016). DNA methylation and mRNA
expression of developmentally important genes in bovine oocytes
collected from donors of different age categories. Mol. Reprod. Dev.
83(9), 802–814. doi:10.1002/MRD.22692
McGee, E. A., and Hsueh, A. J. (2000). Initial and cyclic recruitment of
ovarian follicles. Endocr. Rev. 21(2), 200–214.
McGrath, J., and Solter, D. (1984). Completion of mouse embryogenesis
requires both the maternal and paternal genomes. Cell 37(1), 179–183.
doi:10.1016/0092-8674(84)90313-1
Memili, E., Dominko, T., and First, N. L. (1998). Onset of transcription in
bovine oocytes and preimplantation embryos. Mol. Reprod. Dev. 51(1),
36–41. doi:10.1002/(SICI)1098-2795(199809)51:1,36::AID-MRD4.
3.0.CO;2-X
Me´ne´zo, Y. J., and He´rubel, F. (2002). Mouse and bovine models for human
IVF. Reprod. Biomed. Online 4(2), 170–175. doi:10.1016/S1472-6483
(10)61936-0
Monk, D. (2015). Germline-derived DNA methylation and early embryo
epigenetic reprogramming: the selected survival of imprints. Int. J.
Biochem. Cell Biol. 67, 128–138. doi:10.1016/J.BIOCEL.2015.04.014
Mortusewicz, O., Schermelleh, L., Walter, J., Cardoso, M. C., and
Leonhardt, H. (2005). Recruitment of DNA methyltransferase I to
DNA repair sites. Proc. Natl. Acad. Sci. USA 102(25), 8905–8909.
doi:10.1073/PNAS.0501034102
Motlik, J., Crozet, N., and Fulka, J. (1984). Meiotic competence in vitro of
pig oocytes isolated from early antral follicles. J. Reprod. Fertil. 72(2),
323–328. doi:10.1530/JRF.0.0720323
Mourot, M., Dufort, I., Gravel, C., Algriany, O., Dieleman, S., and Sirard,
M. A. (2006). The influence of follicle size, FSH-enriched maturation
medium, and early cleavage on bovine oocyte maternal mRNA levels.
Mol. Reprod. Dev. 73(11), 1367–1379. doi:10.1002/MRD.20585
Niemann, H., Carnwath, J. W., Herrmann, D., Wieczorek, G., Lemme, E.,
Lucas-Hahn, A., and Olek, S. (2010). DNA methylation patterns reflect
epigenetic reprogramming in bovine embryos. Cell. Reprogram. 12(1),
33–42. doi:10.1089/CELL.2009.0063
Nordhoff, V., Hubner, K., Bauer, A., Orlova, I., Malapetsa, A., and Scholer,
H. R. (2001). Comparative analysis of human, bovine, and murine Oct-4
upstream promoter sequences. Mamm. Genome 12(4), 309–317.
doi:10.1007/S003350010279
O’Doherty, A. M., Rutledge,C. E., Sato, S., Thakur, A., Lees-Murdock, D. J.,
Hata, K., and Walsh, C. P. (2011). DNA methylation plays an important
role in promoter choice and protein production at the mouse Dnmt3L
locus. Dev. Biol. 356(2), 411–420. doi:10.1016/J.YDBIO.2011.05.665
O’Doherty, A. M., O’Shea, L. C., and Fair, T. (2012). Bovine DNA
methylation imprints are established in an oocyte size-specific manner,
which are coordinated withthe expression of the DNMT3family proteins.
Biol. Reprod. 86(3), 67. doi:10.1095/BIOLREPROD.111.094946
Oliveri, R. S., Kalisz, M., Schjerling, C. K., Andersen, C. Y., Borup, R., and
Byskov, A. G. (2007). Evaluation in mammalian oocytes of gene
2050 Reproduction, Fertility and Development F. Mattern et al.
transcripts linked to epigenetic reprogramming. Reproduction 134(4),
549–558. doi:10.1530/REP-06-0315
Pavlok, A., Lucas-Hahn, A., and Niemann, H. (1992). Fertilization and
developmental competence of bovine oocytes derived from different
categories of antral follicles. Mol. Reprod. Dev. 31(1), 63–67.
doi:10.1002/MRD.1080310111
Pesce, M., and Scholer, H. R. (2001). Oct-4: gatekeeper in the beginnings of
mammalian development. Stem Cells 19(4), 271–278. doi:10.1634/
STEMCELLS.19-4-271
Racedo, S. E., Branzini, M. C., Salamone, D., Wojcik, C., Rawe, V. Y., and
Niemann, H. (2009). Dynamics of microtubules, motor proteins and 20S
proteasomes during bovine oocyte IVM 5. Reprod. Fertil. Dev. 21(2),
304–312. doi:10.1071/RD08111
Reik, W., and Walter, J. (2001). Genomic imprinting: parental influence on
the genome. Nat. Rev. Genet. 2(1), 21–32. doi:10.1038/35047554
Robbins, K. M., Chen, Z., Wells, K. D., and Rivera, R. M. (2012). Expression
of KCNQ1OT1,CDKN1C,H19, and PLAGL1 and the methylation
patterns at the KvDMR1 and H19/IGF2 imprinting control regions is
conserved between human and bovine. J. Biomed. Sci. 19, 95.
doi:10.1186/1423-0127-19-95
Saitou, M., Kagiwada, S., and Kurimoto, K. (2012). Epigenetic reprogram-
ming in mouse pre-implantation development and primordial germ cells.
Development 139(1), 15–31. doi:10.1242/DEV.050849
Scho¨ ler, H. R. (1991). Octamania: the POU factors in murine development.
Trends Genet. 7(10), 323–329. doi:10.1016/0168-9525(91)90422-M
Schramm, R. D., Tennier, M. T., Boatman, D. E., and Bavister, B. D. (1993).
Chromatin configurations and meiotic competence of oocytes are related
to follicular diameter in nonstimulated rhesus monkeys. Biol. Reprod. 48
(2), 349–356. doi:10.1095/BIOLREPROD48.2.349
Suetake, I., Shinozaki, F., Miyagawa, J., Takeshima, H., and Tajima, S.
(2004). DNMT3L stimulates the DNA methylation activity of Dnmt3a
and Dnmt3b through a direct interaction. J. Biol. Chem. 279(26),
27816–27823. doi:10.1074/JBC.M400181200
Szo¨ llo¨si, D., Desmedt, V., Crozet, N., and Brender, C. (1988). In vitro
maturation of sheep ovarian oocytes. Reprod. Nutr. Dev. 28(4B),
1047–1080. doi:10.1051/RND:19880705
Tomek, W., Torner, H., and Kanitz, W. (2002). Comparative analysis of
protein synthesis, transcription and cytoplasmic polyadenylation of
mRNA during maturation of bovine oocytes in vitro.Reprod. Domest.
Anim. 37(2), 86–91. doi:10.1046/J.1439-0531.2002.00336.X
Tsuji, K., Sowa, M., and Nakano, R. (1985). Relationship between human
oocyte maturation and different follicular sizes. Biol. Reprod. 32(2),
413–417. doi:10.1095/BIOLREPROD32.2.413
Ulloa, S. M. B., Heinzmann,J., Herrmann, D., Timmermann, B., Baulain, U.,
Grossfeld, R., Diederich, M., Lucas-Hahn, A., and Niemann, H. (2015).
Effects of different oocyte retrieval and in vitro maturation systems on
bovine embryo development and quality. Zygote 23(3), 367–377.
doi:10.1017/S0967199413000658
Uzbekova, S., Roy-Sabau, M., Dalbies-Tran, R., Perreau, C., Papillier, P.,
Mompart,F., Thelie, A., Pennetier,S., Cognie, J., Cadoret, V., Royere,D.,
Monget, P., and Mermillod, P. (2006). Zygote arrest 1 gene in pig, cattle
and human: evidence of different transcript variants in male and female
germ cells. Reprod. Biol. Endocrinol. 4, 12. doi:10.1186/1477-7827-4-12
Warnecke, P. M., Stirzaker, C., Song, J., Grunau, C., Melki, J. R., and Clark,
S. J. (2002). Identification and resolution of artefacts in bisulfite
sequencing. Methods 27(2), 101–107. doi:10.1016/S1046-2023(02)
00060-9
Wrenzycki, C., Herrmann, D., Keskintepe, L., Martins, A., Jr, Sirisathien, S.,
Brackett, B., and Niemann, H. (2001). Effects of culture system and
protein supplementation on mRNA expression in pre-implantation
bovine embryos. Hum. Reprod. 16(5), 893–901. doi:10.1093/HUM
REP/16.5.893
Wu, X., Viveiros, M. M., Eppig, J. J., Bai, Y., Fitzpatrick, S. L., and Matzuk,
M. M. (2003a). Zygote arrest 1 (Zar1) is a novel maternal-effect gene
critical for the oocyte-to-embryo transition. Nat. Genet. 33(2), 187–191.
doi:10.1038/NG1079
Wu, X., Wang, P., Brown, C. A., Zilinski, C. A., and Matzuk, M. M. (2003b).
Zygote arrest 1 (Zar1) is an evolutionarily conserved gene expressed in
vertebrate ovaries. Biol. Reprod. 69(3), 861–867. doi:10.1095/BIOLRE
PROD.103.016022
www.publish.csiro.au/journals/rfd
Epigenetics in oocytes related to follicle size follicles Reproduction, Fertility and Development 2051
