The FASEB Journal • Research Communication
Both the folate cycle and betaine-homocysteine
methyltransferase contribute methyl groups for
DNA methylation in mouse blastocysts
Baohua Zhang,*,† Michelle M. Denomme,‡,§ Carlee R. White,‡,§ Kit-Yi Leung,{ Martin B. Lee,*,†
Nicholas D. E. Greene,{ Mellissa R. W. Mann,‡,§ Jacquetta M. Trasler,k,# and Jay M. Baltz*,†,1
*Ottawa Hospital Research Institute, Ottawa, Ontario, Canada; †Departments of Obstetrics and
Gynecology, and Cellular and Molecular Medicine, University of Ottawa Faculty of Medicine, Ottawa,
Ontario, Canada; ‡Department of Obstetrics and Gynecology, and Biochemistry, Schulich School of
Medicine and Dentistry, Western University, London, Ontario, Canada; §Children’s Health Research
Institute, London, Ontario, Canada; {Developmental Biology and Cancer Program, University College
London Institute of Child Health, London, United Kingdom; kResearch Institute of the McGill University
Health Centre, Montréal Children’s Hospital, Montréal, Quebec, Canada; and #Departments of Human
Genetics, Pediatrics, and Pharmacology and Therapeutics, McGill University, Montréal, Quebec, Canada
The embryonic pattern of global DNA methylation is first established in the inner cell mass (ICM) of the
mouse blastocyst. The methyl donor S-adenosylmethionine
(SAM) is produced in most cells through the folate cycle,
but only a few cell types generate SAM from betaine (N,
N,N-trimethylglycine) via betaine-homocysteine methyltransferase (BHMT), which is expressed in the mouse
ICM. Here, mean ICM cell numbers decreased from 18-19
in controls to 11-13 when the folate cycle was inhibited by
the antifolate methotrexate and to 12-14 when BHMT
expression was knocked down by antisense morpholinos.
Inhibiting both pathways, however, much more severely
affected ICM development (7–8 cells). Total SAM levels in
mouse blastocysts decreased significantly only when both
pathways were inhibited (from 3.1 to 1.6 pmol/100 blastocysts). DNA methylation, detected as 5-methylcytosine
(5-MeC) immunofluorescence in isolated ICMs, was minimally affected by inhibition of either pathway alone but
decreased by at least 45–55% when both BHMT and the
folate cycle were inhibited simultaneously. Effects on cell
numbers and 5-MeC levels in the ICM were completely
rescued by methionine (immediate SAM precursor) or
SAM. Both the folate cycle and betaine/BHMT appear to
contribute to a methyl pool required for normal ICM development and establishing initial embryonic DNA methylation.—Zhang, B., Denomme, M.M., White, C. R., Leung,
K.-Y., Lee, M. B., Greene, N. D. E., Mann, M. R. W., Trasler,
J. M., Baltz, J. M. Both the folate cycle and betainehomocysteine methyltransferase contribute methyl groups
ABSTRACT
Abbreviations: 5-MeC, 5-methylcytosine; BHMT, betainehomocysteine methyltransferase; BSA, bovine serum albumin;
Con MO, control morpholino; E, embryonic day; hCG, human
chorionic gonadotropin; H+T, hypoxanthine and thymidine
(at 1 mM each, except as noted); ICM, inner cell mass; ICR,
imprinting control region; KSOM, potassium-supplemented
simplex-optimized medium; MO, morpholino; MTHFR,
methylenetetrahydrofolate reductase; MTX, methotrexate;
SAM, S-adenosylmethionine; TE, trophectoderm
0892-6638/15/0029-1069 © FASEB
for DNA methylation in mouse blastocysts. FASEB J.
29, 1069–1079 (2015). www.fasebj.org
Key Words: embryo • epigenetics • inner cell mass • methionine • S-adenosylmethionine
THE FINAL STAGE OF THE MAMMALIAN preimplantation embryo
before it implants in the uterus is the blastocyst. The blastocyst develops from an undifferentiated morula by formation of a fluid-filled blastocoel cavity and differentiation
of 2 initial cell lineages: the inner cell mass (ICM) and the
trophectoderm (TE). The TE encloses the blastocoel and
will differentiate into extraembryonic tissues, including the
fetal portion of the placenta and the outermost trophoblast
layer of the fetal membranes. The ICM lies within the
blastocoel on the inner wall of the TE. Just before implantation, the ICM itself diverges into 2 lineages: the
epiblast that will give rise to the fetus, umbilicus, and innermost fetal membranes; and the primitive endoderm
that becomes the intermediate layers of the extraembryonic membranes (1–3).
During early embryogenesis, genome-wide epigenetic
reprogramming occurs wherein the gamete-specific patterns of DNA methylation are globally erased and replaced
with the initial embryo-specific epigenome. Global DNA
demethylation begins shortly after fertilization, with
methylation reaching low levels in the morula and early
blastocyst in mouse and human embryos (4–7). There are
important exceptions to this general scheme, including
imprinted genes and repetitive sequences that escape
the postfertilization erasure and maintain methylation (8,
9), but most of the genome becomes hypomethylated.
1
Correspondence: The Ottawa Hospital Research Institute,
501 Smyth Rd., Mailbox 411, Ottawa, ON, Canada, K1H 8L6.
E-mail: jbaltz@ohri.ca
doi: 10.1096/fj.14-261131
This article includes supplemental data. Please visit http://
www.fasebj.org to obtain this information.
1069
Widespread de novo methylation then reappears in the
epiblast after implantation (4–7). The precise developmental stage where DNA methylation begins to increase
is not fully known. However, the initial indication of increased methylation that has been reported in mouse is
the reappearance of 5-methylcytosine (5-MeC) immunofluorescence at the blastocyst stage, where it is restricted to the nuclei of the ICM (7, 10, 11). There may be
species differences in this timing because in bovine embryos, it reappears in the late cleavage stages instead (12).
Recently, direct sequencing of the preimplantation and
early postimplantation embryo methylome (4, 6) confirmed
the general pattern of erasure followed by de novo methylation indicated by changes in 5-MeC immunofluorescence.
Most methylated sequences appeared in postimplantation
epiblast, however, indicating that a major portion of de novo
methylation continues postimplantation. The identities of
the initial methylated regions that are detected in the ICM by
5-MeC immunofluorescence remain unknown.
DNA methylation involves the addition of methyl groups
to cytosine residues by DNA methyltransferases (3). The
universal methyl donor for these and most other methyltransferases is S-adenosylmethionine (SAM) (13), which is
produced through either of 2 known 1-carbon metabolic
pathways in mammals (14, 15). Almost all cells utilize the
folate cycle to supply 1-carbon units as methyl groups.
However, a few cell types also employ a mechanism that uses
betaine (N,N,N-trimethylglycine) as the methyl donor. In
the folate cycle, methyl groups are shuttled to homocysteine
via 5-methyltetrahydrofolate. The transfer of a methyl
group converts homocysteine to methionine that is further
processed by addition of adenosine to yield SAM (16–18).
One-carbon groups carried on reduced folates are also required for the synthesis of thymidylate and purines (17, 19).
In the betaine-dependent pathway, the methyl group is
instead donated by betaine and transferred directly to
homocysteine by the enzyme betaine-homocysteine
methyltransferase (BHMT; EC2.1.1.5) (14, 15, 20).
A methyl pool is almost certainly produced by preimplantation embryos. Exogenously supplied methionine is
converted to SAM by both mouse and bovine preimplantation
embryos (21, 22). Preimplantation embryos also likely have
a fully functional folate cycle (15, 23) because all the relevant
enzymes of 1-carbon metabolism are expressed in the embryos of several species (24, 25), and inhibiting the folate cycle
using the antifolate, methotrexate (MTX), causes preimplantation developmental arrest in vitro due to depletion of
thymidylate and purines (25, 26). Endogenous folates appear
to be sufficient for preimplantation embryogenesis because
exogenous folate is not required for the development of
viable blastocysts from fertilized mammalian eggs in vitro (25,
26). Nevertheless, preimplantation mouse embryos are able
to take up folates via folate receptor-mediated endocytosis
and thus replenish their folate pool (27).
Although the folate cycle is nearly ubiquitous, the
betaine-BHMT pathway had been considered significant
in rodents only in the liver (28). Unexpectedly, we recently
found that BHMT is also active in the ICM of the mouse
blastocyst and that mouse embryos accumulate high levels
of betaine from the female tract (29, 30). It is not known
whether the DNA remethylation that apparently begins in
the ICM requires methyl groups that are synthesized in the
late preimplantation embryo either through the folate
1070
Vol. 29
March 2015
cycle or from betaine, nor whether this methyl pool is
critical to blastocyst lineage and postimplantation development. Restricted methyl availability may contribute to
the unexplained deleterious effects on the epigenome and
on offspring that can occur from various stresses during the
preimplantation period (9, 31–33), particularly if DNA
methylation is affected.
MATERIALS AND METHODS
Chemicals and media
Chemicals were obtained from Sigma-Aldrich (St. Louis, MO,
USA) unless specified. Embryo culture media were modified
versions of KSOM (potassium-supplemented simplex-optimized
medium) and HEPES-KSOM (34) with glutamine omitted and
polyvinyl alcohol (1 mg/ml; cold-water soluble, MW 30–70 K)
substituted for bovine serum albumin (BSA), producing basal
embryo culture media containing no amino acids or nucleosides,
except where these were added (see below). KSOM was equilibrated with 5% CO2/air. Fresh stocks were prepared in KSOM
(1 mM MTX, 10 mM thymidine, 100 mM methionine, and 5 mM
SAM) or 1 M NaOH (10 mM hypoxanthine).
Mouse embryos
Animal protocols were approved by the Animal Care Committees
of the Ottawa Hospital Research Institute or the University of
Ottawa Faculty of Medicine. Eight-cell embryos were obtained
from female CF1 mice (4–7 wk old; Charles River Laboratories,
Saint-Constant, QC, Canada) superovulated with eCG (5 IU) and
human chorionic gonadotropin (hCG; 5 IU) ;47 h later, and
caged overnight individually with B6D2F1 males (Charles River
Laboratories). Eight-cell embryos were flushed from excised
oviducts with HEPES-KSOM at ;67 h post-hCG and cultured for
48 h in KSOM drops under mineral oil in 5% CO2/air at 37°C
(34), which corresponds to late blastocysts with ;80 cells (see
RESULTS). Images of embryos were recorded at the end of culture. In-vivo-produced 1-cell embryos, morulae, or blastocysts
were obtained at ;21, ;76, or ;94 h post-hCG, respectively.
Morpholino and MTX treatments
Two nonoverlapping antisense morpholinos (MOs) (Gene Tools
Limited Liability Company, Philomath, OR, USA) targeted Bhmt.
BhmtMO1 (59-GTGCCATCTTTCCGGTGTAGTGAGT-39) targeted the start codon (underlined) region of murine Bhmt, and
BhmtMO2 (GTGCCATCTTTCCGGTGTAGTGAGT) targeted
72–48 nt 59 from the start codon. A control morpholino (Con
MO) (GTcCCATgTTTgCGGTcTAcTGAGT) had 5 mismatches
(lowercase) compared to BhmtMO1. MO procedures and
knockdown validation (;70% decrease) were previously described (29). Eight-cell stage embryos were cultured for 48 h with
MO (200 mM) and Lipofectin (6.25 ml/ml) as previously described (29, 35). MTX at the concentrations specified was added
to inhibit the folate cycle. Where indicated, hypoxanthine and
thymidine (H+T) were added to overcome the effects of MTX on
purine and thymidylate synthesis, as described below.
Differential cell counting and lineage determination
in blastocysts
TE and ICM cells were counted as previously described (29, 36).
Blastocysts were exposed for 10 s to 0.1% Triton X-100 and
100 mg/ml propidium iodide to label TE nuclei and then were
The FASEB Journal x www.fasebj.org
ZHANG ET AL.
fixed overnight (100% ethanol with 25 mg/ml bisbenzimide at
4°C) to label all nuclei with bisbenzimide. ICM (bisbenzimide
only) and TE (bisbenzimide and propidium iodide) nuclei were
visualized in slightly compressed blastocysts using conventional
fluorescence microscopy (Leica DMLB with Chroma filter set
31000 for bisbenzimide and 41002 for propidium iodide; Leica
Microsystems Incorporated, Buffalo Grove, IL, USA).
OCT4 (POU5FL, ICM specific) and NANOG (epiblast specific) were detected by immunofluorescence. Blastocysts were
fixed (3.7% paraformaldehyde for 1 h), permeabilized (0.1%
Triton X-100/0.3% BSA for 30 min), and blocked (0.3% BSA/
0.01% Triton X-100 for 1 h) in PBS at room temperature and
labeled with primary antibodies [1:400 OCT4 mouse monoclonal
IgG, C-10 (Santa Cruz Biotechnology, Incorporated, Dallas, TX,
USA); and 1:200 NANOG rabbit polyclonal IgG, RECRCAB0002P-F (Cosmo Bio Company, Limited, Tokyo, Japan)]
in blocking solution overnight at 4°C, followed by secondary
antibodies [Alexa 594 goat anti-mouse IgG A-11005 (Invitrogen,
Carlsbad, CA, USA), and Alexa 488 goat anti-rabbit IgG A-11034,
both 1:400] and 10 mg/ml bisbenzimide. A Zeiss Axioimager M1
microscope (Jena, Germany) with 365 nm excitation, 445/50 nm
emission (bisbenzimide), 545/25 excitation, 605/70 emission
(Alexa 594), and 470/40 excitation, 525/50 emission (Alexa 488)
filter sets was used for imaging.
ICM isolation
ICMs were isolated from blastocysts by immunosurgery as previously described (37, 38). Zonae pellucidae were removed with
Acid Tyrode and TE lysed by exposure to heat-inactivated (56°C
for 30 min) rabbit anti-mouse serum (Sigma-Aldrich; M-5774)
followed by guinea pig complement (Sigma-Aldrich; S-1639, 1:5
in KSOM), each for 10 min at 37°C. The ICM was isolated by
repeated aspiration through a narrow-bore pipette.
5-MeC immunofluorescence
5-MeC was detected by immunofluorescence essentially as previously described (10). Embryos or ICMs were fixed in 4%
formaldehyde in PBS (15 min), permeabilized (0.2% Triton X100 in PBS for 30 min), depurinated (4 N HCl and 0.1% Triton
X-100 for 10 min), blocked overnight (2% BSA in PBS), and
incubated with anti-5-MeC antibody (1:500, ABI-100; Eurogentec,
AnaSpec, Incorporated, Fremont, CA, USA) in blocking solution
(4°C overnight) followed by goat anti-mouse Alexa 488 (1:500 for
1 h at room temperature in the dark) and bisbenzimide (10 mg/ml
for 10 min at room temperature). Imaging was done with a Leica
DMLB microscope using Chroma filter sets 31000 (bisbenzimide)
and 41001 (Alexa 488). Whole-mount fluorescence intensity
was quantitated using ImageJ 1.4 (National Institues of Health,
Bethesda, MD, USA) to draw a perimeter enclosing each blastocyst or isolated ICM and determine the average intensity within
the enclosed area. The average intensity was used to normalize
for variation in cell number. In each repeat, the average intensity
of the control group was set to 100 to normalize for variability in
immunofluorescence levels between repeats.
Determination of methylation status of Snrpn and H19
imprinting control regions
Methylation within the H19 and Snrpn imprinting control regions
(ICRs) was measured in individual immunosurgically isolated
ICMs by bisulfite mutagenesis as previously described for single
oocytes (39) with modifications as follows. In brief, ICMs were
simultaneously lysed and embedded in 3% low-melting point
agarose (Sigma-Aldrich) beads, and bisulfite treatment was
METHYL SOURCES FOR BLASTOCYST DNA REMETHYLATION
performed as described previously (39). Amplification of the
Snrpn and H19 ICRs was performed as described (31), except
that reactions were multiplexed with 2 sets of gene-specific
primers. Negative controls (no ICM) were processed alongside
each bisulfite reaction. First-round product (5 ml) was added
into separate Snrpn and H19 second-round PCRs. PCR amplification and cloning were performed as previously described
(39). Approximately 30 ml of colony PCR was sent to Bio-Basic
Incorporated (Markham, ON, Canada) for sequencing. Percent methylation was calculated as the number of methylated
CpGs divided by total CpGs. Normally, Snrpn is methylated on
the maternal allele and H19 on the paternal allele. Because the
maternal and paternal alleles could not be distinguished in
these assays, the expected level of methylation is ;50%, with
methylation levels between 40 and 60% considered normal.
Measurement of SAM
A total of 100 blastocysts were frozen in each microcentrifuge tube
using liquid nitrogen and stored at 280°C. Thawed samples were
sonicated in 60 ml ice-cold aqueous buffer containing 4 mM ammonium acetate, 0.1% formic acid, and 0.1% heptafluorobutyric
acid (pH 2.5). SAM was quantified by HPLC (40 ml sample injection volume) followed by mass spectrometry, as described
previously (40). S-Adenosylhomocysteine was undetectable above
background in these small samples.
Embryo transfer
Female CD-1 recipients (8–15 wk old; Charles River Laboratories)
were mated with vasectomized CD-1 males. Blastocysts (12–18)
were transferred into these pseudopregnant females 2.5 d postcoitus using the NSET (Non-Surgical Embryo Transfer Device;
ParaTechs, Lexington, KY, USA), according to the manufacturer’s
instructions. Recipients were killed on embryonic d 10.5 (E10.5).
Data analysis
Data were graphed and analyzed with Prism 5 (GraphPad Software, Incorporated, San Diego CA, USA). Data are presented as
the mean 6 SEM. N indicates the number of independent repeats,
and n indicates the total embryos where applicable. Comparisons
were made by t test, ANOVA followed by the Tukey’s multiple
comparison test, or 2-way ANOVA. P , 0.05 was considered
significant.
RESULTS
Effect of BHMT and folate cycle inhibition on
blastocyst development and cell lineage allocation
We have previously validated antisense MOs that effectively
knock down BHMT expression and activity in blastocysts
(29). Here, we used MTX, a well-characterized antifolate
that targets the key folate cycle enzyme dihydrofolate reductase (41), to disrupt the folate cycle. This rapidly depletes
intracellular tetrahydrofolate and 5-methyltetrahydrofolate
(42, 43) and thus blocks homocysteine remethylation.
Inhibition of the folate cycle with MTX also blocks purine
and thymidylate synthesis. This lethal effect can be circumvented with exogenous thymidine and purine (hypoxanthine) while still rapidly depleting folates and selectively
targeting SAM synthesis (44), as previously shown for
1071
BHMT knockdown alone (BhmtMO1 at 0 mM MTX)
decreased development by a moderate amount, as previously shown (29). Both treatments simultaneously
(BhmtMO1 with 1 mM MTX), however, further decreased
development significantly (Fig. 2A).
MTX and BHMT knockdown had distinct effects on the
ICM and TE cell lineages. TE cell number determined by
differential cell counting was reduced with increasing
amounts of MTX but was not affected by BHMT knockdown. In contrast, ICM cell number was decreased by either MTX or BHMT knockdown independently and
significantly further decreased by both treatments together
(Fig. 2B). Comparable results (Fig. 2C, D) were obtained
using the second Bhmt MO (BhmtMO2). As an independent means of counting ICM cells, we used OCT4
immunofluorescence and also counted the subset of ICM
cells in the epiblast lineage by NANOG immunofluorescence (Fig. 2E, F). The number of OCT4-positive nuclei
was essentially the same as the number of ICM cells determined by differential cell counting (cf. Fig. 2B, D). The
decrease in the number of OCT4-positive cells with
BhmtMO2 and further decrease with both BhmtMO2 and
MTX together (Fig. 2F) mirrored the results with differential cell counting. The NANOG-positive subset of ICM
cells was also decreased by BhmtMO2 and further decreased by simultaneous MTX and BhmtMO2 (Fig. 2F),
indicating that epiblast was affected.
murine, bovine, and ovine embryos (25, 26). We had also
attempted to knock down expression of the key folate cycle
enzyme specific to SAM synthesis, methylenetetrahydrofolate
reductase (MTHFR) (45), with antisense MOs, but protein levels were only slightly decreased (data not shown),
probably due to the persistence of maternal MTHFR
protein in blastocysts. Thus, our strategy here was to use
MTX to target the folate cycle in conjunction with BHMT
knockdown using established antisense MOs.
We first optimized conditions for targeting the folate cycle
with MTX while rescuing the effects on purine and thymidylate synthesis using exogenous H+T. Up to 1 mM each of
H+T did not affect blastocyst development or TE and ICM
cell numbers (Fig. 1A, B). Blastocyst development was not
reduced by up to 1 mM MTX when 1 mM each of H+T was
present but was completely suppressed by 10-fold less MTX
without H+T (Fig. 1C). The effect on both ICM and TE cell
number was also much less severe with H+T than without (Fig.
1D). Thus, 1 mM MTX with 1 mM H+T was used in further
experiments to inhibit SAM synthesis by the folate cycle.
Because neither knocking down BHMT in blastocysts
(29) nor inhibiting the folate cycle with MTX in the presence of H+T (Fig. 1) had a profound effect on blastocyst
development by themselves, we blocked both pathways simultaneously to investigate whether they might function in
parallel. Blastocyst development from the 8-cell stage was
not affected by MTX in the presence of Con MO (Fig. 2A).
B
TE
80
60
40
20
20
40
15
10
20
5
0
0
0
2
4
6
8
10
0
0
[H+T], µM
50
0
10-2
6
8
10
0
10-1
[MTX], µM
100
2
4
6
8
10
[H+T], µM
TE
ICM
25
60
100
10-3
4
D
Without H+T
With H+T
150
2
[H+T], µM
Number of cells
% Development
C
ICM
25
60
100
Number of cells
% Development
A 120
20
15
40
10
20
5
0
10-3
10-2
10-1
[MTX], µM
100
0
10-3
10-2
10-1
100
[MTX], µM
Figure 1. Effect of H+T, and MTX, on blastocyst development and cell numbers. A) Development following culture of 8-cell
embryos for 48 h in increasing concentrations of H+T. H+T was added in equal concentrations ([H+T]). The number of embryos
developing to the blastocyst stage was scored for N = 3 separate cultures (n = 25 embryos per group). Each point represents the
mean 6 SEM of the percent blastocyst development from the 8-cell stage. B) Effect of H+T on blastocyst cell allocation in TE or
ICM. The left panel shows mean TE cell numbers and the right panel ICM cell numbers, determined by differential cell
counting. The numbers of blastocysts assessed were n = 13–20 per group, from N = 3 independent cultures in each condition.
Each data point represents the mean 6 SEM of cell number. C) Culture of 8-cell embryos in increasing concentrations of MTX in
the presence or absence of 1 mM [H+T]. The number of embryos developing to the blastocyst stage was scored for N = 3
independent repeats (n = 19 embryos per group in each repeat). D) Effect of increasing dosage of MTX with or without 1 mM [H+T]
on blastocyst cell allocation. The left panel shows mean TE cell numbers and the right panel ICM cell numbers. The numbers
of blastocysts assessed were n = 12–17 per group in each repeat, with N = 3 repeats. In (C) and (D), each data point represents
the mean 6 SEM as in (A) and (B).
1072
Vol. 29
March 2015
The FASEB Journal x www.fasebj.org
ZHANG ET AL.
B
Con MO
BhmtMO1
80
60
40
20
60
40
20
10-1
100
c
60
40
20
C
-
NANOG
MO2/-
C/-
OCT4
MO2
-
C
+
10-1
100
10-2
0
MO2
+
DNA
60
b
b
40
20
0
MO:
MTX:
Merge
20
a
a
MO2
-
C
-
F
20
15
b
c
5
C
-
MO2
-
MO2
+
C
+
15
b
b
c
5
0
C
-
b
10
0
MO:
MTX:
MO2
+
C
+
10
MO:
MTX:
100
ICM
a
a
15
10-1
[MTX], µM
TE
80
OCT4 + cells
b
Number of cells
% Development
a
0
E
5
0
10-2
0
D
a
80
MO:
MTX:
10
[MTX], µM
120
100
15
NANOG + cells
10-2
[MTX], µM
C
20
0
0
ICM
25
Number of cells
100
0
TE
80
Number of cells
% Development
120
Number of cells
A
MO2
-
C
+
a
c
5
0
MO:
MTX:
MO2
+
a,b
b
10
C
-
MO2
-
C
+
MO2
+
a
a
60
TE cells
MO2/+
C/+
80
b
40
b
20
0
MO:
MTX:
C
-
MO2
-
C
+
MO2
+
Figure 2. Effect of BHMT knockdown and MTX on blastocyst development and cell allocation to the ICM and TE. A) Blastocyst
development as a function of MTX concentration in the presence of BhmtMO1 or Con MO. Each point represents the mean 6
SEM of N = 4 independent repeats. The effects of MO (P , 0.0001), MTX dose (P = 0.0005), and interaction between MTX and
MO (P = 0.0019) were significant by 2-way ANOVA. B) TE and ICM cell numbers, as indicated (n = 15–23 per group, N = 3
independent repeats; mean 6 SEM). For TE, only the effect of MTX was significant (P , 0.0001, 2-way ANOVA). For ICM, the
effects of MTX and MO were each significant (both P , 0.0001). C) Same as (A), except BhmtMO2 was used and MTX was used
at 1 mM only (n = 40 per group, N = 3 independent repeats; mean 6 SEM). C, control. D) Same as (B), except MO and MTX were
as in (C) (n = 26–33 per group, N = 3). For both (C) and (D), means are significantly different where bars do not share letters
(P , 0.01, ANOVA with Tukey’s test). E) Representative immunofluorescence images of OCT4, NANOG, and DNA
(bisbenzimide) of blastocysts with the same 4 treatment groups as in (C) and (D). Scale bar, 50 mm. F) Quantification of OCT4- and
NANOG-positive cells and TE cells (total minus OCT4). Means are significantly different where bars do not share letters (P , 0.0001,
except b vs. c for NANOG, P , 0.05, ANOVA with Tukey’s test; N = 3, n = 33–44 blastocysts per group).
SAM levels in blastocysts
Control blastocysts contained approximately 3 pmol SAM/
100 blastocysts (Fig. 3). Blastocysts treated with BHMT MO
or MTX alone had SAM levels that did not differ significantly from controls, although MTX caused a trend toward
a decrease. However, total SAM content was significantly
lowered (by ;50%) in blastocysts in which BHMT had
METHYL SOURCES FOR BLASTOCYST DNA REMETHYLATION
been knocked down and the folate cycle inhibited with
MTX simultaneously (Fig. 3).
5-MeC immunofluorescence in blastocysts
We next determined whether BHMT knockdown or
inhibition of SAM synthesis via the folate cycle affected
5-MeC immunofluorescence at the blastocyst stage.
1073
(N = 10; P = 0.65; Supplemental Fig. S1B). Thus, DNA
methylation at these loci in the ICM as assessed by bisulfite
sequencing was not perturbed by simultaneous BHMT
knockdown and folate cycle inhibition.
4
a
3
a,b
2
b
Rescue by remethylation products
It was possible that the effects on blastocyst development
and 5-MeC levels were due to incomplete rescue by H+T or
C
MO1
MTX:
-
-
+
+
1074
Vol. 29
March 2015
MO1
+
MO:
5-MeC
DNA
Merge
B
140
Normalized fluorescence
Knocking down BHMT with BhmtMO1 or inhibiting folate metabolism with MTX had no noticeable effect on
the appearance of 5-MeC in the ICM when compared to
control embryos (Fig. 4A). However, with simultaneous
BHMT knockdown and folate cycle inhibition, 5-MeC
staining of the ICM became nearly indistinguishable
from the lower level of fluorescence in TE (Fig. 4A).
Quantification of the mean fluorescence intensity over
the whole blastocyst indicated that 5-MeC was significantly decreased in blastocysts with MTX and BhmtMO1
together, but not with either alone (Fig. 4B).
Because 5-MeC levels were most severely affected in the
ICM, we assessed 5-MeC in the ICM that had been isolated
by immunosurgery. Knocking down BHMT or inhibiting
SAM synthesis via the folate cycle alone in blastocysts only
slightly decreased 5-MeC levels in the isolated ICM (Fig. 5).
In contrast, 5-MeC was again substantially decreased in the
ICM when both pathways were perturbed. Each antisense
Bhmt MO had a similar effect (Fig. 5B).
We also assessed the methylation status of the ICRs of 2
well-characterized imprinted genes, Snrpn and H19, whose
methylation was previously shown to be susceptible to
perturbation by in vitro culture (31), in immunosurgically
isolated ICMs from control blastocysts (Con MO) and
blastocysts where BHMT had been knocked down and the
folate cycle inhibited (BhmtMO1 and MTX). Normal
methylation levels were observed for Snrpn in 4 of 7 control
and 6 of 10 BhmtMO1 plus MTX-treated ICMs and for H19
in 7 of 8 control and 7 of 10 BhmtMO1 plus MTX-treated
ICMs. The remaining ICM in both groups exhibited some
methylation loss, likely attributable to the effect of in vitro
embryo culture (31). The mean methylation at the Snrpn
ICR was 41 6 3% (mean 6 SEM, N = 7) for control ICM and
40 6 2% (N = 10) for ICM where both pathways were
perturbed (P = 0.85, t test; data sets in Supplemental Fig.
S1A). Likewise, the mean methylation at the H19 ICR was
46 6 2% (N = 8) for control vs. 45 6 2% for the treated ICM
MTX:
C
Figure 3. Effect of BHMT knockdown and folate cycle
inhibition on SAM in blastocysts. SAM was measured in groups
of 100 blastocysts that had been cultured from the 8-cell stage.
There were 4 treatment groups assessed as indicated at the
bottom, where BHMT had been knocked down with
BhmtMO1 (MO1) or treated with control mismatched MO
(C), and where MTX was either present (+) or absent (2).
Bars indicate the mean 6 SEM of N = 6 (C,MTX2 and MO1,
MTX2) or N = 3 (C,MTX+ and MO1,MTX+) independent
measurements. Bars that do not share the same letter indicate
significantly different means (ANOVA, Tukey’s, P , 0.05).
A
-
MO1
C
C
-
0
MO:
MO1
1
+
pmoles/100 blastocysts
a
b
120
a
a
100
80
c
60
40
20
0
MO:
C
MO1
C
MO1
MTX:
-
-
+
+
Figure 4. 5-MeC immunofluorescence in blastocysts. A) Representative images of 5-MeC immunofluorescence (left) and DNA
(bisbenzimide, middle), or merged images (right) in blastocysts
cultured from the 8-cell stage with Con MO (C) or BHMT MO
(BhmtMO1), with (+) or without (2) MTX as indicated at left of
panel. The position of the ICM is indicated by white arrows in
each merged image. Scale bar, 50 mm. B) Mean fluorescence intensity
normalized to control (mean for the control group within each
repeat was set to 100) for the same treatments as in (A) (as
indicated at the bottom). Each bar represents the mean 6 SEM
of N = 6 independent sets of measurements (total blastocysts: 66
C,MTX2; 60 BhmtMO1,MTX2; 59 C,MTX+; and 65 BhmtMO1,
MTX+). Bars not sharing letters are significantly different (ANOVA
with Tukey’s test: a vs. b, P , 0.05; a vs. c, P , 0.01; b vs. c, P , 0.001).
The FASEB Journal x www.fasebj.org
ZHANG ET AL.
B
MO1
-
C
-
Normalized fluorescence
A
120
a
100
b
b
80
c
60
40
20
MO1
+
C
+
Normalized fluorescence
0
MO:
C
MO1
C
MO1
MTX:
-
-
+
+
120
a
b
80
60
c
40
20
MO:
MTX:
0
5-MeC
DNA
Merge
MO:
C
MO2
C
MO2
MTX:
-
-
+
+
MTX toxicity. We therefore attempted to rescue the detrimental effects with exogenous methionine, the direct
product of homocysteine remethylation. Met is taken up
via active transport into both TE and ICM (46) and is
converted into SAM in blastocysts (21). Methionine (0.2 mM)
rescued blastocyst development and reversed the decrease in ICM and TE cell numbers (Fig. 6A, B). Most
strikingly, 5-MeC immunofluorescence in the ICM was
completely restored by methionine (Fig. 6C). We next
attempted rescue by supplying exogenous SAM (0.2 mM).
SAM similarly rescued development from the 8-cell to
blastocyst stages when both BHMT and the folate cycle
a
a
80
a
b
60
40
20
0
120
Postimplantation development
Embryos cultured from the 8-cell to blastocyst stages in
4 treatment groups (BhmtMO1 antisense MO or Con
MO, each with or without MTX) were transferred to
a
a
100
80
60
b
40
20
0
MO:
C
MO2
C
MO2
MO:
C
MO2
MO2
MTX:
-
+
-
+
MTX:
-
+
+
MET:
-
-
+
+
MET:
-
-
+
TE
80
60
a
a,c
c
b
40
ICM
20
20
0
Number of cells
% Development
100
Number of cells
were perturbed (Fig. 7A). SAM completely restored total
ICM (OCT4-positive) and epiblast (NANOG-positive) cell
numbers and partially restored TE cell number (Fig. 7B).
Finally, 5-MeC levels assessed in isolated ICMs were also
restored by SAM (Fig. 7C).
C
Normalized fluorescence
A
B
a,b
100
Figure 5. 5-MeC immunofluorescence in isolated ICMs. A)
Representative images of 5-MeC
immunofluorescence in isolated
ICMs from blastocysts treated as
in Fig. 4. Scale bar, 20 mm. B)
Mean fluorescence intensity normalized to the control group for
the ICM from blastocysts in
which BHMT had been knocked
down with BhmtMO1 (upper)
or BhmtMO2 (lower). Each bar
represents the mean 6 SEM of
N = 5 independent measurements with BhmtMO1 (total
ICMs: 48 C,MTX2; 35 MO1,
MTX2; 39 C,MTX+; and 32
MO1,MTX+) and N = 3 with
BhmtMO2 (25 C,MTX2; 21
MO1,MTX2; 26 C,MTX+; and
21 MO1,MTX+). Bars not
sharing letters are significantly
different (ANOVA with Tukey’s
test: upper panel, all P , 0.001;
lower panel, a vs. b, P , 0.01,
and a or b vs. c, P , 0.001).
a
a
15
a
b
10
5
0
MO:
C
MO2
C
MO2
MO:
C
MO2
C
MO2
MTX:
-
+
-
+
MTX:
-
+
-
+
MET:
-
-
+
+
MET:
-
-
+
+
METHYL SOURCES FOR BLASTOCYST DNA REMETHYLATION
Figure 6. Rescue of blastocyst development
and 5-MeC levels by methionine. A) Effect of
methionine on blastocyst development from the
8-cell stage with BhmtMO2 (MO2) or without
MO (C), and with (+) or without (2)MTX.
Methionine(Met, 0.2 mM) was added to culture
medium in the presence or absence of MTX
plus BhmtMO2. The decreased development
seen with both BhmtMO2 and MTX was
completely rescued by methionine (mean 6
SEM; ANOVA with Tukey’s test: a vs. b, P ,
0.01; N = 3). B) Effect of methionine on
blastocyst cell numbers in TE (left panel) and
ICM (right), assessed in blastocysts shown in (A)
by differential cell counting. Rescue by methionine was complete for the ICM (a vs. b, P ,
0.001) and nearly to control levels for TE (a vs.
b, P , 0.001; a or b vs. c, P , 0.05). C) Effect of
methionine on 5-MeC immunofluorescence in
isolated ICMs. Methionine restored fluorescence levels essentially to control levels (a vs.
b, P , 0.001).
1075
A
B
60
40
20
0
MO2
MO2
MTX:
-
+
+
SAM:
-
-
+
C
b
10
5
a
10
a
b
5
0
120
a
a
100
80
a
80
c
60
TE cells
b
60
40
20
0
b
40
20
0
MO:
C
MO2
MO2
MO:
C
MO2
MO2
MTX:
-
+
+
MTX:
-
+
+
SAM:
-
-
+
SAM:
-
-
+
Both the folate cycle and the betaine-BHMT pathway appear to contribute to SAM production that is required for
de novo global DNA remethylation in the mouse ICM as
detected by 5-MeC immunofluorescence. Inhibiting either
pathway alone produced only minimal reductions in 5MeC, but when both pathways were disrupted, 5-MeC levels
in the ICM were substantially decreased. These effects appear to be specifically due to inhibition of SAM production,
rather than nonspecific or toxic effects, because addition
March 2015
a
15
15
NANOG + cells
C
DISCUSSION
Vol. 29
a
0
pseudopregnant recipients (5–9 recipients per treatment,
12–18 blastocysts per recipient). Fetuses, placentas, and
resorption sites were examined on E10.5, when fetuses and
placentas could be assessed separately. The implantation
rate was significantly decreased only when both pathways
were perturbed (Fig. 8A, H). The fraction of transferred
blastocysts that resulted in fetuses on E10.5 was not significantly decreased by BMHT knockdown alone but was
lower in both MTX groups (Fig. 8B, H). Decreased fetal
numbers were due mainly to increased resorption of MTXtreated embryos that had implanted (Fig. 8C, D). Fetal
weight, crown-rump length, and placental weight were
determined for all treatment groups, although very few
fetuses were available in the BhmtMO plus MTX group
(;4% of embryos transferred survived to become fetuses
on E10.5; Fig. 8B), providing limited material for analysis.
Nevertheless, when the surviving fetuses were assessed,
those in all 3 treatment groups exhibited significantly decreased mean fetal weights, crown-rump lengths, and placental weights compared with control (Fig. 8E–G).
1076
OCT4 + cells
b
80
20
a
MO:
Normalized fluorescence
Figure 7. Rescue of blastocyst development and
5-MeC levels by SAM. Blastocysts were treated as
in Fig. 6, but in the presence or absence of 0.2 mM
SAM. A) The decrease in blastocyst development
with BhmtMO2 and MTX was rescued by SAM
added to the medium (mean 6 SEM; ANOVA with
Tukey’s test, P , 0.01; N = 3). B) Panels show
mean numbers of OCT4-positive nuclei (upper
panel), NANOG-positive epiblast nuclei (middle
panel), and TE (lower panel, calculated as
bisbenzimide positive but OCT4 negative). The
numbers of OCT4- and NANOG-positive nuclei
were restored by SAM in the medium (P , 0.001),
whereas the number of TE nuclei was partially
restored (a vs. b vs. c, P , 0.01; n = 36–46 in total
from the N = 3 independent sets shown in (A). C)
Effect of SAM on 5-MeC immunofluorescence in
isolated ICMs. SAM fully restored 5-MeC (a vs. b,
P , 0.001; n = 42–48 ICMs in total collected in
N = 3 independent sets).
% Development
100
a
of either the direct SAM precursor methionine or SAM to
the medium rescued global 5-MeC levels in the ICM. This
interpretation is consistent with the effects in bovine embryos of directly blocking methionine conversion to SAM
with ethionine, where 5-MeC levels were also reduced (22).
Therefore, we propose that, unlike virtually all nonhepatic cells, the mouse ICM has 2 pathways for SAM
production that are at least partly redundant. A robust
capacity to produce SAM may be necessitated by the large
demand imposed by global DNA remethylation in the
mouse epiblast. In addition, there may be other increases
in demand on the methyl pool at the blastocyst stage because new histone methylation patterns become established in both the ICM and TE at the blastocyst stage (47)
and are involved in ICM lineage allocation (48). The increased metabolic activity in blastocysts also may require
methyl groups to supply many of the approximately 1% of
proteins in mammals that are predicted to be methyltransferases (49).
The effects of BHMT knockdown and folate cycle inhibition on development to the blastocyst stage and on cell
numbers in each lineage were also consistent with the hypothesis that the folate cycle is active in both the TE and
ICM, whereas BHMT is restricted only to the ICM because
only MTX had an effect on TE, whereas both MTX and
BHMT knockdown affected the ICM. Decreased blastocyst
development and reduced cell numbers are likely independent of any perturbations to global DNA remethylation
because remethylation in the ICM occurs late in the blastocyst stage after the ICM has formed, does not occur in
the TE, and knockout of zygotic DNA methyltransferases
Dnmt1, Dnmt3a, and Dnmt3b does not prevent formation
of the cell lineages in blastocysts (50). Instead, a restricted
The FASEB Journal x www.fasebj.org
ZHANG ET AL.
a
60
40
b
20
0
MO:
MTX:
D
a,b
C
-
MO1
-
C
+
b
100
resorption
(% of implantations)
MO1
+
E
a
a
20
60
a
40
b
20
b
0
MO1
-
C
+
MO1
-
C
+
MO1
+
G
a
80
a
60
b
40
b
20
MO:
MTX:
MO1
+
F
0.03
b
0.02
b
b
0.01
0
C
-
100
0
C
-
0.04
0
MO:
MTX:
a
a
60
40
80
MO:
MTX:
b
80
C
100
fetuses
(% of implantations)
80
fetal weight (g)
Implantation rate (%)
a
placental weight (g)
B
100
Fetuses per ET (%)
A
C
-
MO1
-
C
+
MO1
+
b
b
a,b
MO1
-
C
+
MO1
+
0.08
a
0.06
0.04
0.02
0
MO:
MTX:
C
-
MO1
-
C
+
MO1
+
MO:
MTX:
C
-
H
crown-rump length
(mm)
8
a
6
b
b
b
P
F
2
R
F
R
R
F
R
0
MO:
MTX:
P
P
4
C
-
MO1
-
C
+
MO1
+
MO:
MTX:
C
-
MO1
-
C
+
MO1
+
Figure 8. Postimplantation development of blastocysts after embryo transfer. Development was analyzed on E10.5 following
transfer of blastocysts that had been cultured from the 8-cell stage for 48 h with control (C) or Bhmt antisense (BhmtMO1) MO,
and with (+) or without (2) 1 mM MTX, as indicated at the bottom of each panel. A) The number of blastocysts that implanted in
the recipient uterus and continued to develop sufficiently to be detected on E10.5 was assessed, and implantation rate was
calculated as the number of fetuses plus resorption sites divided by number of embryos transferred. B) Development of fetuses
from blastocysts on E10.5 (% fetuses/blastocysts transferred). C) Fetal development (% fetuses/total implantations). D)
Resorption sites evident on E10.5 (% resorptions/total implantations). E) Weights of fetuses recovered in grams. F) Weights of
placentas recovered in grams. G) Crown-rump length of fetuses recovered in millimeters. In (A)–(G), means (6SEM) represented
by bars that do not share the same letters are significantly different: ANOVA, Tukey’s: P , 0.05 for (A), (B), and (F); P , 0.01 for
(C) and (D); and P , 0.001 for (E) and (G). There were N = 5–9 recipients per group as follows: Con MO, no MTX, 5 recipients
with 39 fetuses and 40 placentas successfully recovered for fetal weight, crown-rump length measurement, and placental weight;
BhmtMO1, no MTX, 5 recipients, 28 fetuses, and 29 placentas; Con MO, plus MTX, 7 recipients, 14 fetuses, and 14 placentas; and
BhmtMO1 plus MTX, 9 recipients, 6 fetuses, and 6 placentas. Note that a larger number of embryo transfers was performed for
the latter 2 groups to attempt to obtain enough fetuses and placentas for measurements, although the low implantation rate and
development of fetuses per embryo transferred (A, B) still resulted in a smaller number being available for measurements. H)
Examples of uteri before dissection (upper panels), and fetuses (F), placentae (P), and resorption sites (R) (lower panels) for
each treatment. Each lower panel shows the material recovered from the uterus in the upper panel. With nonsurgical embryo
transfer, embryos are transferred into only 1 uterine horn. Scale bars, 1 cm.
supply of SAM may impair the development of each
lineage by reducing the methyl pool available for the
various other methyl-requiring biochemical pathways in
the blastocyst.
Embryo transfer experiments indicated that inhibition
of both the BHMT and folate-dependent SAM-generating
pathways substantially decreased implantation as assessed
by the number of resorption sites and fetuses present on
E10.5, but there was no significant effect of inhibiting each
pathway alone. This is consistent with the more severe
METHYL SOURCES FOR BLASTOCYST DNA REMETHYLATION
effects of inhibition of both pathways on both cell number
and 5-MeC levels in the ICM. However, once embryos had
implanted and developed long enough to be present as
either fetuses or resorption sites on E10.5, embryos from
both groups where the folate cycle had been inhibited with
MTX were much more likely to be resorbed. This may
reflect impairment of placental function because our
results indicate that TE, which gives rise to the fetal compartment of the placenta, relies on the folate cycle alone,
and TE cell number was particularly sensitive to MTX.
1077
Mice lacking a functional Bhmt gene are fertile (51).
Similarly, mice lacking the Mthfr gene and thus unable to
produce SAM through the folate cycle are viable on some
genetic backgrounds, and these are fertile (52). Thus,
there does not appear to be an absolute requirement for
either a functional folate cycle or BHMT activity alone for
embryos to develop through the blastocyst stage in vivo and
develop normal fetuses and placentae. This is consistent
with our conclusion here that BHMT and the folate cycle
can each compensate at least in part when the other is
perturbed. It is also possible that external methionine may
be taken up by the embryo from the maternal reproductive
tract in vivo to compensate for impaired homocysteine
remethylation, similar to the methionine rescue that we
demonstrated in vitro.
Relatively mild stress during preimplantation development can result in substantial dysregulation of a
number of imprinted genes that are differentially methylated between the maternal and paternal alleles, leading
to aberrant hypomethylation of the normally methylated allele (8). DNA methyltransferase activity is required to maintain imprinted gene methylation during
preimplantation development (53–55), implying a requirement for SAM in imprinting maintenance. Surprisingly, the methylation status of H19 and Snrpn was
unaffected in the ICM even when both pathways for SAM
production were inhibited, despite the apparent loss of
overall 5-MeC immunofluorescence. These results potentially indicate either that the methylation patterns of at least
these 2 imprinted genes are stable in the blastocyst or that
DNA methylation at imprinted domains is preferentially
maintained when there is reduced availability of SAM.
Because deletion or depletion of DNMT1o/DNMT1s
during preimplantation development resulted in the loss
of imprinted methylation (55, 56), the latter explanation
is favored, although this remains to be investigated.
Despite the recent focus of investigations on imprinted
genes, a more global epigenetic instability may result from
suboptimal conditions during preimplantation embryogenesis that is not restricted just to the subset of differentially methylated imprinted genes (8). We propose that
during late preimplantation and early postimplantation
development, it may be DNA remethylation in the ICM
and early epiblast, which probably requires a very substantial flux of methyl groups through the SAM pool, that is
most susceptible to conditions that restrict SAM production.
The resulting perturbation of embryonic epigenetic marks
in the genome may underlie some of the reported negative
sequelae of embryo culture stress, techniques of assisted
reproductive technologies, and adverse maternal nutrition
during the peri-implantation period.
This work was supported by Canadian Institutes of Health
Research Operating Grant MOP97972 (to J.M.B. and J.M.T.).
B.Z. was partially supported by a Lalor Foundation postdoctoral
fellowship. M.M.D. and C.R.W. were partially supported by
CIHR Training Program in Reproduction, Early Development,
and the Impact on Health (TGF96122) studentships and
laboratory exchange visit awards. K.Y.L. and N.D.E.G. were
supported by the Medical Research Council, London, United
Kingdom (J003794). J.M.T. is a James McGill Professor at
McGill University and a member of the Research Institute of
the McGill University Health Centre, which is supported in part
by the Fonds de la Recherche du Québec-Santé.
1078
Vol. 29
March 2015
REFERENCES
1. Biggers, J. D., Bell, J. E., and Benos, D. J. (1988) Mammalian
blastocyst: transport functions in a developing epithelium. Am. J.
Physiol. 255, C419–C432
2. Rossant, J., and Tam, P. P. L. (2009) Blastocyst lineage formation,
early embryonic asymmetries and axis patterning in the mouse.
Development 136, 701–713
3. Hudson, Q. J., Kulinski, T. M., Huetter, S. P., and Barlow, D. P.
(2010) Genomic imprinting mechanisms in embryonic and
extraembryonic mouse tissues. Heredity (Edinb) 105, 45–56
4. Smith, Z. D., Chan, M. M., Humm, K. C., Karnik, R., Mekhoubad,
S., Regev, A., Eggan, K., and Meissner, A. (2014) DNA methylation
dynamics of the human preimplantation embryo. Nature 511,
611–615
5. Guo, H., Zhu, P., Yan, L., Li, R., Hu, B., Lian, Y., Yan, J., Ren, X.,
Lin, S., Li, J., Jin, X., Shi, X., Liu, P., Wang, X., Wang, W., Wei, Y.,
Li, X., Guo, F., Wu, X., Fan, X., Yong, J., Wen, L., Xie, S. X.,
Tang, F., and Qiao, J. (2014) The DNA methylation landscape of
human early embryos. Nature 511, 606–610
6. Smith, Z. D., Chan, M. M., Mikkelsen, T. S., Gu, H., Gnirke, A.,
Regev, A., and Meissner, A. (2012) A unique regulatory phase of
DNA methylation in the early mammalian embryo. Nature 484,
339–344
7. Reik, W., Dean, W., and Walter, J. (2001) Epigenetic reprogramming
in mammalian development. Science 293, 1089–1093
8. Denomme, M. M., and Mann, M. R. W. (2012) Genomic imprints
as a model for the analysis of epigenetic stability during assisted
reproductive technologies. Reproduction 144, 393–409
9. Market Velker, B. A., Denomme, M. M., and Mann, M. R. (2012)
Loss of genomic imprinting in mouse embryos with fast rates of
preimplantation development in culture. Biol. Reprod. 86, 143
10. Santos, F., and Dean, W. (2006) Using immunofluorescence to
observe methylation changes in mammalian preimplantation
embryos. Methods Mol. Biol. 325, 129–137
11. Santos, F., Hendrich, B., Reik, W., and Dean, W. (2002) Dynamic
reprogramming of DNA methylation in the early mouse embryo.
Dev. Biol. 241, 172–182
12. Dobbs, K. B., Rodriguez, M., Sudano, M. J., Ortega, M. S., and
Hansen, P. J. (2013) Dynamics of DNA methylation during early
development of the preimplantation bovine embryo. PLoS One 8,
e66230
13. Lu, S. C. (2000) S-Adenosylmethionine. Int. J. Biochem. Cell Biol.
32, 391–395
14. Selhub, J. (1999) Homocysteine metabolism. Annu. Rev. Nutr. 19,
217–246
15. Ikeda, S., Koyama, H., Sugimoto, M., and Kume, S. (2012) Roles
of one-carbon metabolism in preimplantation period—effects on
short-term development and long-term programming—. J. Reprod.
Dev. 58, 38–43
16. Tibbetts, A. S., and Appling, D. R. (2010) Compartmentalization
of Mammalian folate-mediated one-carbon metabolism. Annu.
Rev. Nutr. 30, 57–81
17. Fowler, B. (2001) The folate cycle and disease in humans. Kidney
Int. Suppl. 78, S221–S229
18. Fox, J. T., and Stover, P. J. (2008) Folate2mediated one-carbon
metabolism. In Vitamins & Hormones, Folic Acid and Folates
(Gerald, L., ed.), pp. 1–44, Academic Press, New York
19. Bolusani, S., Young, B. A., Cole, N. A., Tibbetts, A. S., Momb, J.,
Bryant, J. D., Solmonson, A., and Appling, D. R. (2011) Mammalian
MTHFD2L encodes a mitochondrial methylenetetrahydrofolate
dehydrogenase isozyme expressed in adult tissues. J. Biol. Chem. 286,
5166–5174
20. Garrow, T. A. (1996) Purification, kinetic properties, and cDNA
cloning of mammalian betaine-homocysteine methyltransferase.
J. Biol. Chem. 271, 22831–22838
21. Menezo, Y., Khatchadourian, C., Gharib, A., Hamidi, J.,
Greenland, T., and Sarda, N. (1989) Regulation of S-adenosyl
methionine synthesis in the mouse embryo. Life Sci. 44, 1601–1609
22. Ikeda, S., Sugimoto, M., and Kume, S. (2012) Importance of
methionine metabolism in morula-to-blastocyst transition in bovine preimplantation embryos. J. Reprod. Dev. 58, 91–97
23. Steegers-Theunissen, R. P., Twigt, J., Pestinger, V., and Sinclair,
K. D. (2013) The periconceptional period, reproduction and
long-term health of offspring: the importance of one-carbon
metabolism. Hum. Reprod. Update 19, 640–655
The FASEB Journal x www.fasebj.org
ZHANG ET AL.
24. Ikeda, S., Namekawa, T., Sugimoto, M., and Kume, S. (2010)
Expression of methylation pathway enzymes in bovine oocytes
and preimplantation embryos. J. Exp. Zool. A Ecol. Genet. Physiol.
313, 129–136
25. Kwong, W. Y., Adamiak, S. J., Gwynn, A., Singh, R., and Sinclair,
K. D. (2010) Endogenous folates and single-carbon metabolism
in the ovarian follicle, oocyte and pre-implantation embryo. Reproduction 139, 705–715
26. O’Neill, C. (1998) Endogenous folic acid is essential for normal
development of preimplantation embryos. Hum. Reprod. 13,
1312–1316
27. Kooistra, M., Trasler, J. M., and Baltz, J. M. (2013) Folate
transport in mouse cumulus-oocyte complexes and preimplantation embryos. Biol. Reprod. 89, 63
28. Delgado-Reyes, C. V., Wallig, M. A., and Garrow, T. A. (2001)
Immunohistochemical detection of betaine-homocysteine Smethyltransferase in human, pig, and rat liver and kidney.
Arch. Biochem. Biophys. 393, 184–186
29. Lee, M. B., Kooistra, M., Zhang, B., Slow, S., Fortier, A. L.,
Garrow, T. A., Lever, M., Trasler, J. M., and Baltz, J. M. (2012)
Betaine homocysteine methyltransferase is active in the mouse
blastocyst and promotes inner cell mass development. J. Biol.
Chem. 287, 33094–33103
30. Anas, M. K., Lee, M. B., Zhou, C., Hammer, M. A., Slow, S.,
Karmouch, J., Liu, X. J., Bröer, S., Lever, M., and Baltz, J. M.
(2008) SIT1 is a betaine/proline transporter that is activated in
mouse eggs after fertilization and functions until the 2-cell stage.
Development 135, 4123–4130
31. Market-Velker, B. A., Fernandes, A. D., and Mann, M. R. (2010)
Side-by-side comparison of five commercial media systems in
a mouse model: suboptimal in vitro culture interferes with imprint maintenance. Biol. Reprod. 83, 938–950
32. Fleming, T. P., Kwong, W. Y., Porter, R., Ursell, E., Fesenko, I.,
Wilkins, A., Miller, D. J., Watkins, A. J., and Eckert, J. J. (2004)
The embryo and its future. Biol. Reprod. 71, 1046–1054
33. Sinclair, K. D., and Singh, R. (2007) Modelling the developmental
origins of health and disease in the early embryo. Theriogenology 67,
43–53
34. Lawitts, J. A., and Biggers, J. D. (1993) Culture of preimplantation
embryos. Methods Enzymol. 225, 153–164
35. Siddall, L. S., Barcroft, L. C., and Watson, A. J. (2002) Targeting gene
expression in the preimplantation mouse embryo using morpholino
antisense oligonucleotides. Mol. Reprod. Dev. 63, 413–421
36. Thouas, G. A., Korfiatis, N. A., French, A. J., Jones, G. M., and
Trounson, A. O. (2001) Simplified technique for differential
staining of inner cell mass and trophectoderm cells of mouse and
bovine blastocysts. Reprod. Biomed. Online 3, 25–29
37. Hewitson, L. C., Martin, K. L., and Leese, H. J. (1996) Effects of
metabolic inhibitors on mouse preimplantation embryo development
and the energy metabolism of isolated inner cell masses. Mol. Reprod.
Dev. 43, 323–330
38. Solter, D., and Knowles, B. B. (1975) Immunosurgery of mouse
blastocyst. Proc. Natl. Acad. Sci. USA 72, 5099–5102
39. Denomme, M. M., Zhang, L., and Mann, M. R. W. (2012) Single
oocyte bisulfite mutagenesis. J. Vis. Exp. 64, 4046
40. Burren, K. A., Mills, K., Copp, A. J., and Greene, N. D. (2006) Quantitative analysis of s-adenosylmethionine and s-adenosylhomocysteine
in neurulation-stage mouse embryos by liquid chromatography
tandem mass spectrometry. J. Chromatogr. B Analyt. Technol. Biomed.
Life Sci. 844, 112–118
41. Huennekens, F. M. (1994) The methotrexate story: a paradigm
for development of cancer chemotherapeutic agents. Adv.
Enzyme Regul. 34, 397–419
42. Allegra, C. J., Fine, R. L., Drake, J. C., and Chabner, B. A. (1986)
The effect of methotrexate on intracellular folate pools in
METHYL SOURCES FOR BLASTOCYST DNA REMETHYLATION
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
human MCF-7 breast cancer cells. Evidence for direct inhibition
of purine synthesis. J. Biol. Chem. 261, 6478–6485
Baram, J., Allegra, C. J., Fine, R. L., and Chabner, B. A. (1987)
Effect of methotrexate on intracellular folate pools in purified
myeloid precursor cells from normal human bone marrow.
J. Clin. Invest. 79, 692–697
Fiskerstrand, T., Ueland, P. M., and Refsum, H. (1997) Folate
depletion induced by methotrexate affects methionine synthase
activity and its susceptibility to inactivation by nitrous oxide.
J. Pharmacol. Exp. Ther. 282, 1305–1311
Chen, Z., Karaplis, A. C., Ackerman, S. L., Pogribny, I. P., Melnyk, S.,
Lussier-Cacan, S., Chen, M. F., Pai, A., John, S. W. M., Smith, R. S.,
Bottiglieri, T., Bagley, P., Selhub, J., Rudnicki, M. A., James, S. J., and
Rozen, R. (2001) Mice deficient in methylenetetrahydrofolate
reductase exhibit hyperhomocysteinemia and decreased
methylation capacity, with neuropathology and aortic lipid
deposition. Hum. Mol. Genet. 10, 433–443
Miller, J. G. O., and Schultz, G. A. (1985) Amino acid transport
in mouse blastocyst compartments. J. Embryol. Exp. Morphol. 89,
149–158
Shi, L., and Wu, J. (2009) Epigenetic regulation in mammalian
preimplantation embryo development. Reprod. Biol. Endocrinol. 7, 59
Torres-Padilla, M. E., Parfitt, D. E., Kouzarides, T., and
Zernicka-Goetz, M. (2007) Histone arginine methylation
regulates pluripotency in the early mouse embryo. Nature 445,
214–218
Katz, J. E., Dlakić, M., and Clarke, S. (2003) Automated
identification of putative methyltransferases from genomic
open reading frames. Mol. Cell. Proteomics 2, 525–540
Sakaue, M., Ohta, H., Kumaki, Y., Oda, M., Sakaide, Y.,
Matsuoka, C., Yamagiwa, A., Niwa, H., Wakayama, T., and
Okano, M. (2010) DNA methylation is dispensable for the
growth and survival of the extraembryonic lineages. Curr. Biol.
20, 1452–1457
Teng, Y. W., Mehedint, M. G., Garrow, T. A., and Zeisel, S. H.
(2011) Deletion of betaine-homocysteine S-methyltransferase in
mice perturbs choline and 1-carbon metabolism, resulting in
fatty liver and hepatocellular carcinomas. J. Biol. Chem. 286,
36258–36267
Lawrance, A. K., Racine, J., Deng, L., Wang, X., Lachapelle, P., and
Rozen, R. (2011) Complete deficiency of methylenetetrahydrofolate
reductase in mice is associated with impaired retinal function
and variable mortality, hematological profiles, and reproductive
outcomes. J. Inherit. Metab. Dis. 34, 147–157
Kurihara, Y., Kawamura, Y., Uchijima, Y., Amamo, T., Kobayashi,
H., Asano, T., and Kurihara, H. (2008) Maintenance of genomic
methylation patterns during preimplantation development
requires the somatic form of DNA methyltransferase 1. Dev.
Biol. 313, 335–346
Howell, C. Y., Bestor, T. H., Ding, F., Latham, K. E., Mertineit, C.,
Trasler, J. M., and Chaillet, J. R. (2001) Genomic imprinting
disrupted by a maternal effect mutation in the Dnmt1 gene. Cell
104, 829–838
Cirio, M. C., Martel, J., Mann, M., Toppings, M., Bartolomei, M.,
Trasler, J., and Chaillet, J. R. (2008) DNA methyltransferase 1o
functions during preimplantation development to preclude
a profound level of epigenetic variation. Dev. Biol. 324, 139–150
Hirasawa, R., Chiba, H., Kaneda, M., Tajima, S., Li, E., Jaenisch,
R., and Sasaki, H. (2008) Maternal and zygotic Dnmt1 are
necessary and sufficient for the maintenance of DNA methylation
imprints during preimplantation development. Genes Dev. 22,
1607–1616
Received for publication October 14, 2014.
Accepted for publication November 3, 2014.
1079