Genome-wide mapping of 5-hydroxymethylcytosine in embryonic stem cells

Article abstract of DOI:10.1038/nature10102

5-hydroxymethylcytosine (5hmC) is a modified base present at low levels in diverse cell types in mammals. 5hmC is generated by the TET family of Fe(II) and 2-oxoglutarate-dependent enzymes through oxidation of 5-methylcytosine (5mC). 5hmC and TET proteins have been implicated in stem cell biology and cancer, but information on the genome-wide distribution of 5hmC is limited. Here we describe two novel and specific approaches to profile the genomic localization of 5hmC. The first approach, termed GLIB (glucosylation, periodate oxidation, biotinylation) uses a combination of enzymatic and chemical steps to isolate DNA fragments containing as few as a single 5hmC. The second approach involves conversion of 5hmC to cytosine 5-methylenesulphonate (CMS) by treatment of genomic DNA with sodium bisulphite, followed by immunoprecipitation of CMS-containing DNA with a specific antiserum to CMS. High-throughput sequencing of 5hmC-containing DNA from mouse embryonic stem (ES) cells showed strong enrichment within exons and near transcriptional start sites. 5hmC was especially enriched at the start sites of genes whose promoters bear dual histone 3 lysine 27 trimethylation (H3K27me3) and histone 3 lysine 4 trimethylation (H3K4me3) marks. Our results indicate that 5hmC has a probable role in transcriptional regulation, and suggest a model in which 5hmC contributes to the poised chromatin signature found at developmentally-regulated genes in ES cells.

Full text of DOI:10.1038/nature10102

LETTER
doi:10.1038/nature10102
Genome-wide mapping of 5-hydroxymethylcytosine
in embryonic stem cells
1
,2
1,3
2
1,2
4
2
5
William A. Pastor *, Utz J. Pape *, Yun Huang *, Hope R. Henderson , Ryan Lister , Myunggon Ko , Erin M. McLoughlin ,
6
2
7
1
5
3
Yevgeny Brudno , Sahasransu Mahapatra , Philipp Kapranov , Mamta Tahiliani {, George Q. Daley , X. Shirley Liu ,
4
7
5
1,2
Joseph R. Ecker , Patrice M. Milos , Suneet Agarwal & Anjana Rao
5
-hydroxymethylcytosine (5hmC) is a modified base present at low were very similar (Fig. 1c): even at low densities of 5hmC, more than
1–5
levels in diverse cell types in mammals . 5hmC is generated by the 90% of DNA fragments calculated to contain a single 5hmC were
TETfamilyofFe(II)and2-oxoglutarate-dependentenzymesthrough precipitated after GLIB treatment. Anti-CMS pull-down showed
oxidation of 5-methylcytosine (5mC)
1,2,4–7
. 5hmC and TET proteins increased density dependence compared to GLIB, but had very low
1,4,5,8,9
have been implicated in stem cell biology and cancer
, but background, such that there was still a strong preference for precipi-
information on the genome-wide distribution of 5hmC is limited. tation of sparsely hydroxymethylated amplicons over unmodified
Here we describe two novel and specific approaches to profile the ones (Fig. 1d). The performance of a commercial polyclonal anti-
genomic localization of 5hmC. The first approach, termed GLIB hmC antiserum was inferior to that of anti-CMS, in terms of higher
(
glucosylation, periodate oxidation, biotinylation) uses a combina- background pull-down of unmodified DNA (3.0% versus 0.06%) as well
tion of enzymatic and chemical steps to isolate DNA fragments con- as greater density dependence (Fig. 1e). By testing PCR amplicons with
taining as few as a single 5hmC. The second approach involves varying 5mC, weconfirmed that the methyl-DNAimmunoprecipitation
conversion of 5hmC to cytosine 5-methylenesulphonate (CMS) by (MeDIP) technique, which uses a monoclonal antibody to 5mC, is
treatment of genomic DNA with sodium bisulphite, followed by extremely density-dependent (Fig. 1f).
immunoprecipitation of CMS-containing DNA with a specific anti-
We applied the GLIB and anti-CMS techniques to enrich 5hmC-
5
serum to CMS . High-throughput sequencing of 5hmC-containing containing regions in genomic DNA using genomic DNA with low,
DNA from mouse embryonic stem (ES) cells showed strong enrich- intermediate and high levels of 5hmC (Supplementary Fig. 1c, left
ment within exons and near transcriptional start sites. 5hmC was panel). For the GLIB and anti-CMS pull-downs, the amount of specif-
especially enriched at the start sites of genes whose promoters bear ically precipitated genomic DNA was proportional to the relative
dual histone 3 lysine 27 trimethylation (H3K27me3) and histone 3 amount of 5hmC (Supplementary Fig. 1c). The GLIB technique did
lysine 4 trimethylation (H3K4me3) marks. Our results indicate that notproducemutations(SupplementaryTable3), thebiotinylatedDNA
5
hmC hasa probable role in transcriptionalregulation, and suggest a could be efficiently eluted by heating with formamide (Supplementary
model in which 5hmC contributes to the ‘poised’ chromatin sig- Fig. 1d), and the biotinylated adduct had a minimal inhibitory effect
nature found at developmentally-regulated genes in ES cells.
on PCR at 5–25% hmC density (delay of approximately 0.1 cycles per
We developed two independent methods for precipitation of5hmC in converted 5hmC residue; Supplementary Fig. 1f). There was no PCR
genomic DNA. The GLIB method (Fig. 1a) entails addition of a glucose delay with CMS-containing PCR amplicons except at very high CMS
3
molecule to each 5hmC with T4 phage b-glucosyltransferase (BGT)
levels (Supplementary Fig. 1e), consistent with our previous report
(
Supplementary Fig. 1a). The glucose moiety is oxidized with sodium that CMS inhibits PCR predominantly at biologically irrelevant
periodate, which converts the vicinal hydroxyl groups to aldehydes , sequences where multiple CMS adducts occur in a row .
10
13
and further modified with aldehyde-reactive probe, which adds two
We investigated the genome-wide localization of 5hmC in murine
biotin molecules to each 5hmC (Fig. 1a). A related strategy, which uses V6.5 ES cells. For GLIB-treated DNA, we chose Helicos single mol-
a custom-synthesized UDP-glucose analogue (UDP-6-N3-glucose), was ecule DNA sequencing, which does not require an amplification step
11
14,15
recently used to profile 5hmC distribution in mouse brain . The second and thus avoids PCR bias . For CMS-enriched genomic DNA, we
method uses an antibody against cytosine 5-methylenesulphonate used an Illumina instrument, as longer read lengths are needed for
5
16
(
(
CMS) , produced by reaction of 5hmC with sodium bisulphite efficient alignment of bisulphite-treated DNA to the genome . With
12
Fig. 1b) . Anti-CMS antibodies are more sensitive and less density- the GLIB method, 119,600 regions of the genome, averaging 1,422 bp
dependent than anti-5hmC in DNA dot blot assays . Both methods are in length, showed a substantially higher density of reads in the 1BGT
specific for DNA containing 5hmC (Supplementary Fig. 1b) .
5
5
as opposed to the –BGT sample; with the CMS method, comparison of
We examined the ability of GLIB-treated (biotinylated) and enriched to input DNA identified 109,264 enriched regions (average
bisulphite-treated 5hmC-containing DNA to be pulled down by strep- length 1,168 bp). There was high overlap in the enriched regions, here
tavidin and anti-CMS antisera, respectively. Using varying ratios of designated 5hmC-enriched regions of the genome (HERGs) (Fig. 1g).
dCTP:dhmCTP, we generated 201 base pairs PCR amplicons with Comparing the number of HERGs retrieved by using different frac-
differing incorporation of cytosine and 5hmC in identical sequence tions of aligned reads yielded a curve that approached an asymptote,
contexts (Supplementary Table 1). At each dhmCTP:dCTP ratio, the suggesting that a majority of hydroxymethylated regions had been
fraction of amplicons that contain no 5hmC, and therefore should not identified (Supplementary Fig. 2a).
be precipitated, can be calculated using the binomial equation (Sup-
To determine whether HERGs overlapped with methylated DNA
plementary Table 2). Observed and calculated pull-down efficiencies regions, we identified 62,991 5mC-enriched regions of the genome
1
2
Harvard Medical School, Immune Disease Institute and Program in Cellular and Molecular Medicine, Children’s Hospital Boston, Boston, Massachusetts 02115, USA. La Jolla Institute for Allergy &
3
Immunology, La Jolla, California 92037, USA. Department of Biostatistics and Computational Biology, Dana-Farber Cancer Institute and Harvard School of Public Health, Boston, Massachusetts 02115,
4
5
USA. Genomic Analysis Laboratory, The Salk Institute for Biological Studies, La Jolla, California 92037, USA. Division of Hematology/Oncology, Children’s Hospital Boston; Dana-Farber Cancer Institute;
Harvard Stem Cell Institute, Boston, Massachusetts 02115, USA. Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts 02138, USA. Helicos BioSciences
6
7
Corporation, Cambridge, Massachusetts 02139, USA. {Present address: Department of Biochemistry, New York University Langone Medical Centre, New York, New York 10016, USA.
*These authors contributed equally to this work.
3
9 4 N A T U R E V O L 4 7 3 1 9 M A Y 2 0 1 1
|
|
|
©
2011 Macmillan Publishers Limited. All rights reserved

LETTER RESEARCH
a
a2.0
GLIB
2.0
Anti-CMS
Sodium
T4 phage BGT
UDP-glucose
periodate
11
11
1.5
1.5
1.0
17
5
17 19
2
5
19
4
8
10
4
8
5hmC
10 15
2
7
1
.0
1
813
16
9
7
1
813 15
9
Biotin
12
16
12
14 1 3
6
1
6
3
14
Aldehyde reactive
probe (ARP)
0.5
0.5
r2 = 0.59
r
2
= 0.58
X
X
Y
Y
0.5
1.0
1.5
2.0
0.5
1.0
Gene density
1.5
2.0
Biotin
Gene density
5hmC
2
.0
Anti-5mC
b
c
Sodium
bisulphite
1.5
5
19
7
8
0
11
7
1
1
4
1
8
15
2
1
.0
13
9
1
2
d
1
316 6
9
8.4 100.0
100.0 100.0
84.8
76.4
1
4
1
00
100
Anti-CMS
pAb
GLIB
8
7.1
94.0
0.5
r2 = 0.42
8
6
4
2
0
0
0
0
92.0
80
60
40
20
7
2.7
63.1
4.6
X
63.9
6
2.3
Y
5
9.1
4
0.5
1.0
1.5
2.0
3
9.9
39.2
22.2
1.8
35.8
28.2
13.5
Gene density
2
2.4
Calculated
Observed
Calculated
Observed
1
2
0.0
7.6
0
0.7 01.0
0.08 0.06 2.8
GLIB HERGs
Randomized GLIB HERGs
CMS HERGs
Randomized CMS HERGs
MeDIP MERGs
Randomized MeDIP MERGs
(
mC)
0
0.25 0.5
1
2
4
81
(mC) 0 0.25 0.5
1
2
4
20 81
b
c
Average 5hmCs per amplicon
Average 5hmCs per amplicon
6
0
*
Enriched regions associated
with feature (P < 0.05)
e
f
*
*
1
00.0 100.0
100.0 100.0
*
*
1
00
100
Anti-5hmC
pAb
Anti-5mC
mAb
84.8
84.8
8
81.6
80
80
*
62.3
40
*
62.3
*
6
4
0
0
50.4
60
Calculated
Observed
6
4
*
*
40
2
6.5
22.3
22.2
22.2
1
9.0
*
20
Calculated
Observed
20
12.5
9.2
0.26 1.5
4
4
.7
20
1.6
0
0
3.0
0
0.12
0
(
mC)
0
0.5
2
4
20 81
0.5
2
20
81
2
Average 5hmCs per amplicon
Average 5mCs per amplicon
GLIB bases CMS bases
g
GLIB HERGs
CMS HERGs
Transcribed regions
5′ UTR
TSS
Enhancer
TSS
30,312
89,288
19,976
d
TSS
67.0 Mb
103.1 Mb
24.5 Mb
1,000
Observed GLIB
3
2
,000
,400
HERG distribution
Observed anti-CMS
HERG distribution
8
6
00
00
Observed MERG
distribution
Figure 1
|
Comparison of 5hmC enrichment methods. a, The GLIB method.
1,800
,200
00
Glucose is added to 5hmC by BGT, oxidized with sodium periodate to yield
aldehydes, and reacted with the aldehyde reactive probe (ARP), yielding two
biotins at the site of every 5hmC. b, 5hmC is converted to CMS by sodium
bisulphite. c–f, Precipitation of PCR amplicons containing (1) varying amounts
of 5hmC by GLIB methodology (c), anti-CMS methodology (d), or anti-5hmC
antibody (e); or (2) varying amounts of 5mC by anti-5mC antibody (f). pAb,
polyclonal antibody; mAb, monoclonal antibody. Between 1 and6 independent
experiments per method, mean percentage input precipitated 6 s.d. is
indicated. g, Overlap between HERGs identified by the GLIB and anti-CMS
methodologies. Left panel, number of HERGs;right panel, number of base pairs
contained within HERGs.
400
200
1
Randomized GLIB HERG distribution
Randomized CMS HERG distribution
Randomized MERG
distribution
6
Position relative to TSS
Position relative to TSS
Figure 2
|
Genomic distribution of 5hmC or 5mC enriched regions of the
genome. a, Correlation of HERG or MERG density on each chromosome (y-
axis) with gene density in the same chromosome (x-axis). Density is defined as
frequency divided by chromosome length. b, c, Both HERGs and MERGs are
enriched in transcribed regions (b), whereas HERGs are preferentially enriched
at enhancers and the start sites of genes (c). The percentage of HERGs or
(
MERGs) by MeDIP. The resulting 5mC profile does not represent a MERGs mapping to the indicated genomic feature (darker bar) is compared
complete map of 5mC in mouse ES cells, but rather is biased towards with the percentage of randomly chosen sequences mapping to that feature
(
(
lighter bar). 59 UTR, 59 untranslated region. TSS, transcription start site
2800bp to 1200bp relative to start of transcription). See Supplementary
regions of dense methylation. Statistics pertaining to the GLIB, anti-
CMS and MeDIP enrichments are shown in Supplementary Figs 2b–d,
the corresponding annotations are provided in Supplementary Tables
Methods for detailed definition of how HERGs or MERGs were classified as
mapping to genomic features. d, Distribution of HERGs and MERGs relative to
the TSS. The centre of each HERG was plotted relative to the nearest TSS in
4
–9, and reads and enrichment for the Hoxb locus are provided in
Supplementary Table 10. As expected, both HERGs and MERGs con-
tained a high frequency of CG sequences relative to the genome at large
1
,000 bp increments from 210 kb to 110 kb surrounding the TSS.
(SupplementaryFig. 3a). Intriguingly, HERGs also containedrelatively
high levels of CAG sequences, the most frequent site of non-CpG HERGs and MERGs to the distribution of DNA fragments of equival-
16
methylation in human ES cells , and we confirmed that the TET1 ent length distributed randomly across the genome, both 5hmC and
catalytic domain is capable of hydroxylating 5mC in CHG and CHH 5mC were enriched within transcribed regions, particularly exons,
1
7
(
H 5 A, T or C) contexts in vitro (Supplementary Fig. 3b).
which are known to be sites of high CpG density as well as high
18
Analysis of the GLIB and anti-CMS HERG sets gave very similar DNA methylation (Fig. 2b and Supplementary Fig. 3c). However,
results. We observed a strong correlation between the densities of only 5hmC was enriched at transcription start sites (TSSs) and within
HERGs and genes on a given chromosome; this trend was less pro- the 59 untranslated regions (UTRs) of genes (Fig. 2c). Moreover, 5hmC
nounced for MERGs (Fig. 2a). When we compared the distribution of was relativelymore enriched in enhancers(defined byH3K4me1 inthe
1
9 M A Y 2 0 1 1 | V O L 4 7 3 | N A T U R E | 3 9 5
©
2011 Macmillan Publishers Limited. All rights reserved

RESEARCH LETTER
1
9
absence of H3K4me3) than 5mC, strongly indicating a connection
between 5hmC and regulatory elements (Fig. 2c). Plotting each HERG
as a single point relative to the nearest TSS, we found that 5hmC is
heavily enriched both 59 and 39 of the TSS, whereas 5mC is enriched
primarily 39 of the TSS (Fig. 2d). These results show a unique distri-
bution of 5hmC in regulatory elements of genes, one that is not
explained simply by the distribution of 5mC, the substrate for TET
enzymes.
a
5hmC-TSS genes (GLIB)
Total = 5,353
5hmC-TSS genes (anti-CMS)
Total = 4,766
All genes
Total = 17,762
60
(
number of genes in category
indicated in or above bar)
P < 10-6
*
*
*
40
*
*
*
The enrichment of 5hmC at the TSS suggested a role for 5hmC in
transcriptional regulation. To evaluate this possibility, we used pub-
*
*
*
*
*
20
20,21
22,23
lished data sets on gene expression
profiles in mouse ES cells to compare the sets of genes with 5hmC or
mC at their start sites (Supplementary Tables 11–13) to the set of all
and histone modification
* *
K4 only
K27 only
K4+K27
Ezh2
Suz12
Ring1B
5
Histone H3 trimethylation
PRC components
genes in the genome. 5hmC is preferentially found at promoters with
high or intermediate CpG content (Supplementary Fig. 4a), even
b
1
6,18,24
though high CpG promoters are hypomethylated in ES cells
.
Hydroxymethylated TSS
Hydroxymethylated
gene body (GLIB)
2
1
0
5
(
GLIB)
This distribution is consistent with the possibility that TET proteins
are preferentially recruited to high CpG regions through their CpG-
12.6
12.2 12.512.012.1
1
1.0
11.1
6
,25
10
5
8.37
8.02
binding CXXC domains
.
7
.06
6.96
4
.71
In ES cells, genes with ‘bivalent’ H3K27 and H3K4 trimethylation
are transcriptionally inactive but poised for expression upon differ-
Decile: 1–5
6
7
8
9
10
1–5
6
7
8
9
10
20,26,27
entiation to embryoid bodies
. We found that genes with 5hmC
(average)
(average)
at their start sites were disproportionately likely to contain bivalent
domains at their promoters; likewise, a majority (,60%) of genes
reported to contain bivalent domains have 5hmC at their start sites
Low
High
Low
High
Expression
Expression
Hydroxymethylated TSS
anti-CMS)
Hydroxymethylated
gene body (anti-CMS)
20
(Fig. 3a). 5hmC was less likely to be found at genes with the activating
(
‘
H3K4me3 only’ mark than is predicted by chance. Moreover, genes
15
1
2.9
12.5 12.5
2.0
12.2
1
1
0.7
10.7
with 5hmC at their start sites showed lower expression in murine ES
cells than other genes (Fig. 3b) and were more likely to be upregulated
uponembryoid body differentiation (Fig. 3c). Thecorrelation of 5hmC
with bivalent domains holds even after adjusting for the known rela-
tion between promoter CpG content and bivalency (Supplementary
Fig. 5). Although 5mC at the TSS also correlates with lower gene
expression in murine ES cells (Supplementary Fig. 4b), 5mC is not
10
8.12
7
8.00
7
.06
5.69
9
5
3.85
10
Decile: 1–5
average)
6
8
1–5
6
7
8
9
10
(
(average)
22
Low
High
Low
High
Expression
Expression
2
8
enriched at the promoters of genes with bivalent domains (Sup-
plementary Fig. 4c), and genes with high levels of 5mC did not tend
to be upregulated upon embryoid body differentiation (Supplemen-
tary Fig. 4d). Thus 5hmC is preferentially enriched at the promoters of
genes with bivalent histone marks in ES cells, indicating that 5hmC
may contribute functionally to the ‘poised’ but inactive state of these
genes in ES cells.
Genes with 5hmC at their start sites are also disproportionately
enriched in the set of genes whose promoters bind polycomb repressor
complex (PRC) components, and in a majority of genes with the
c
d
Tet1
negative
regulation
5
hmC-TSS genes (GLIB)
Total = 3,134
5hmC-TSS (GLIB)
5,298
Tet1
positive
regulation
5hmC-TSS genes (anti-CMS)
Total = 2,753
All genes
Total = 17,762
40 33*
22 48
(
number of genes in category
indicated in or above bar)
P < 10–6
*
P < 0.005
*
*
*
30
5hmC-TSS
Tet1
negative
regulation
(anti-CMS)
Tet1
positive
regulation
2
1
0
0
‘
H3K27me3’ only mark (Fig. 3a). There is a statistically significant
4,724
*
*
correlation between genes that had 5hmC at the TSS and genes that
were upregulated upon small interfering RNA-mediated Tet1 deple-
45
28*
14 56
8
tion (therefore, negatively regulated by Tet1) (Fig. 3d), indicating that
Upregulated
ES to EB
Downregulated
ES to EB
5
hmC in the promoter region has a negative role in the transcription of
P < 0.01
*
some genes in ES cells. Unlike 5mC, however, 5hmC is not substan-
tially enriched at sites of heterochromatic H3K9 or H4K20 trimethyla-
Figure 3
|
Properties of HERGs at transcription start sites. a, The percentage
22
of genes with 5hmC at the TSS (blue and red bars) reported to contain histone
H3 trimethylation (left) or PRC components (right) at their promoters is
compared to the fraction of all genes (grey bars) with these promoter marks .
Number of genes in each category is indicated. b, HERGs are enriched at the
TSSs of genes with low expression in ES cells. All genes were ranked by level of
tion (data not shown).
Collectively, our results support a model in which 5hmC and 5mC
2
2
28
have different roles in transcription. Like 5mC , 5hmC at promoters is
predictive of lower levels of gene expression. However, 5hmC is
2
1
uniquely associated with a ‘poised’ chromatin configuration and with expression in ES cells and sorted into deciles from lowest to highest. The per
genes that are upregulated upon differentiation, and may thus be cent of genes withinthe decile category with5-hmC enriched at the TSS (left) or
involved in priming loci for rapid activation in response to appropriate within gene bodies (right) are shown for each methodology. The first five
deciles, which arecomprisedofgenes lackingstatisticallysignificant expression,
are pooled and averaged in this analysis. c, HERGs are enriched at the TSS of
genes upregulated upon differentiation to embryoid bodies (EB) . The
percentage of genes with 5-hmC at their TSS (blue bars) that are substantially
upregulated or downregulated upon differentiation to EB is compared with the
percentage of total genes similarly regulated (grey bars). Number of genes in
signals. Activation of lineage-specific genetic loci upon differentiation
could occur via a postulated 5mC ‘demethylation’ pathway (5mC to
2
6
1
5
hmCtocytosine) orthroughrecruitmentoftranscriptionalregulators
that specifically recognize 5hmC and are activated in response to dif-
ferentiation signals. The ability to profile 5hmC even at sparsely hydro-
xymethylated loci will allow a careful evaluation of these possibilities in each category is indicated. d, Overlap between genes with 5hmC at the TSS and
differentiating cells. genes positively or negatively regulated by Tet1 (ref. 8).
3
9 6 | N A T U R E | V O L 4 7 3 | 1 9 M A Y 2 0 1 1
©
2011 Macmillan Publishers Limited. All rights reserved

LETTER RESEARCH
1
1
6. Lister, R. et al. Human DNA methylomes at base resolution show widespread
epigenomic differences. Nature 462, 315–322 (2009).
7. Saxonov, S., Berg, P. & Brutlag, D. L. A genome-wide analysis of CpG dinucleotides
in the human genome distinguishes two distinct classes of promoters. Proc. Natl
Acad. Sci. USA 103, 1412–1417 (2006).
METHODS SUMMARY
GLIB precipitation. V6.5 ES cells were lysed and proteins digested by treatment
withProteinase K at55uC. DNAwaspurifiedbyphenol-chloroformextraction and
thenprecipitatedwith ethanol. RNA wasremovedwith RNase A (Qiagen). Samples
were treated with 20ng BGT per 1 mg DNA at 30uC for 3 h (50mM HEPES pH 8.0,
18. Feng, S. et al. Conservation and divergence of methylation patterning in plants and
animals. Proc. Natl Acad. Sci. USA 107, 8689–8694 (2010).
2
2
5mM MgCl , 50mM UDPG for 3 h at 30uC), then oxidized with 23mM sodium
1
9. Creyghton, M. P. et al. Histone H3K27ac separates active from poised enhancers
and predicts developmental state. Proc. Natl Acad. Sci. USA doi:10.1073/
pnas.1016071107 (24 November 2010).
periodate 16 h at 22 uC in 0.1M sodium phosphate pH 7.0. Periodate was quenched
by the addition of 46mM sodium sulphite at room temperature for 10 min, then
exchanged into 13 PBS and incubated with 2 mM Aldehyde Reactive Probe
2
0. Boyer, L. A. et al. Polycomb complexes repressdevelopmental regulatorsinmurine
embryonic stem cells. Nature 441, 349–353 (2006).
(
Invitrogen) for 1 h at 37uC. DNA was sequenced with a HeliScope Single
Molecule Sequencer. See Supplementary Methods for detailed protocol.
21. Guttman, M. et al. Ab initio reconstruction of cell type-specific transcriptomes in
mouse reveals the conserved multi-exonic structure of lincRNAs. Nature
Biotechnol. 28, 503–510 (2010).
CMS precipitation. The generation of the anti-CMS antibody is described else-
5
where . DNA fragments were ligated with methylated adaptors and treated with
2
2. Mikkelsen, T. S. et al. Genome-wide maps of chromatin state in pluripotent and
lineage-committed cells. Nature 448, 553–560 (2007).
sodium bisulphite (Qiagen). The DNA was then denatured for 10 min at 95 uC
(
0.4 M NaOH, 10 mM EDTA), neutralized by addition of cold 2 M ammonium
23. Ku, M. et al. Genomewide analysis of PRC1 and PRC2 occupancy identifies two
classes of bivalent domains. PLoS Genet. 4, e1000242 (2008).
acetate pH 7.0, incubated with anti-CMS antiserum in 13 immunoprecipitation
buffer (10mM sodium phosphate pH 7.0, 140mM NaCl, 0.05% Triton X-100) for
2
4. Meissner, A. et al. Genome-scale DNA methylation maps of pluripotent and
differentiated cells. Nature 454, 766–770 (2008).
2
h at 4 uC, and then precipitated with Protein G beads. Precipitated DNA was
2
5. Zhang, H. et al. TET1 is a DNA-binding protein that modulates DNA methylation
and gene transcription via hydroxylation of 5-methylcytosine. Cell Res. 20,
eluted with Proteinase K, purified by phenol-chloroform extraction, and amplified
by 4–6 cycles PCR using Pfu TurboCx hotstart DNA polymerase (Stratagene).
DNA sequencing was carried out using Illumina/Solexa Genome Analyzer II and
HiSeq sequencing systems.
1
390–1393 (2010).
26. Lee, T. I. et al. Control of developmental regulators by Polycomb in human
embryonic stem cells. Cell 125, 301–313 (2006).
2
2
7. Bernstein, B. E. et al. A bivalent chromatin structure marks key developmental
genes in embryonic stem cells. Cell 125, 315–326 (2006).
8. Fouse, S. D. et al. Promoter CpG methylation contributes to ES cell gene regulation
in parallel with Oct4/Nanog, PcG complex, and histone H3 K4/K27 trimethylation.
Cell Stem Cell 2, 160–169 (2008).
Received 4 December 2010; accepted 11 April 2011.
Published online 8 May 2011.
1
2
3
.
.
.
Tahiliani, M. et al. Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in
mammalian DNA by MLL partner TET1. Science 324, 930–935 (2009).
Kriaucionis, S. & Heintz, N. The nuclear DNA base 5-hydroxymethylcytosine is
present in Purkinje neurons and the brain. Science 324, 929–930 (2009).
Szwagierczak, A., Bultmann, S., Schmidt, C. S., Spada, F. & Leonhardt, H. Sensitive
enzymatic quantification of 5-hydroxymethylcytosine in genomic DNA. Nucleic
Acids Res. 38, e181 (2010).
Ito, S. et al. Role of Tet proteins in 5mC to 5hmC conversion, ES-cell self-renewal
and inner cell mass specification. Nature 466, 1129–1133 (2010).
Ko, M. et al. Impaired hydroxylation of 5-methylcytosine in myeloid cancers with
mutant TET2. Nature 468, 839–843 (2010).
Iyer, L. M., Tahiliani, M., Rao, A. & Aravind, L. Prediction of novel families of enzymes
involved in oxidative and other complex modifications of bases in nucleic acids.
Cell Cycle 8, 1698–1710 (2009).
Loenarz, C. & Schofield, C. J. Oxygenasecatalyzed 5-methylcytosine hydroxylation.
Chem. Biol. 16, 580–583 (2009).
Supplementary Information is linked to the online version of the paper at
www.nature.com/nature.
Acknowledgements We thank B. Ren for assistance in next generation sequencing
using the Illumina platform. We thank M. Guttman for making his RNASeq data set
available to us. W.A.P. is supported by a predoctoral graduate research fellowship from
the National Science Foundation, and Y.H. by a postdoctoral fellowship from the
Leukemia and Lymphoma Society. R.L. is supported by a California Institute for
Regenerative Medicine Training Grant. This study was supported by the National
Institute of Health grants RC1 DA028422, R01 AI44432 and 1 R01 HD065812-01A1
and a grant from the California Institute of Regenerative Medicine (to A.R.), a pilot grant
from Harvard Catalyst, The Harvard Clinical and Translational Science Center (NIH
Grant 1 UL1 RR 025758-02) and NIH K08 HL089150 (to S.A.), and a grant from the
Mary. K. Chapman Foundation (to J.R.E.).
4.
5.
6.
7
.
.
Author Contributions W.A.P., Y.B. and S.A. devised the GLIB method. W.A.P., S.A., H.R.H.
and E.M.M. optimized the GLIB method. Y.H. generated the anti-CMS antiserum, and
Y.H. and W.A.P. optimized the anti-CMS pull-down. W.A.P. and Y.H. grew ES cells. W.A.P.
prepared GLIB samples for sequencing, Y.H. prepared CMS samples, H.R.H. performed
MeDIPs. Helicos sequencing and mapping was performed by P.K. and P.M.M., Illumina
sequencing and mapping was performed by R.L. and J.R.E., and U.J.P. was responsible
for bioinformatic analysis. M.K. performed the anti-5hmC dot blot. W.A.P. and M.T.
performed anti-5hmC pull-downs. H.R.H. and S.M. performed and optimized in vitro
tests of Tet substrate specificity. W.A.P., S.A. andA.R. wrote the manuscript. S.A. andA.R.
coordinated research.
8
Koh, K. P. et al. Tet1 and Tet2 regulate 5-hydroxymethylcytosine production and
cell lineage specification in mouse embryonic stem cells. Cell Stem Cell 8,
2
00–213 (2011).
Delhommeau, F. et al. Mutation in TET2 in myeloid cancers. N. Engl. J. Med. 360,
289–2301 (2009).
9
1
.
2
0. Zhang, H., Li, X. J., Martin, D. B. & Aebersold, R. Identification and quantification of
N-linked glycoproteins using hydrazide chemistry, stable isotope labeling and
mass spectrometry. Nature Biotechnol. 21, 660–666 (2003).
1
1
1
1
1
1. Song, C. X. et al. Selective chemical labeling reveals the genome-wide distribution
of 5-hydroxymethylcytosine. Nature Biotechnol. 29, 68–72 (2011).
2. Hayatsu, H. & Shiragami, M. Reaction of bisulfite with the 5-hydroxymethyl group
in pyrimidines and in phage DNAs. Biochemistry 18, 632–637 (1979).
3. Huang, Y. et al. The behaviour of 5-hydroxymethylcytosine in bisulfite sequencing.
PLoS ONE 5, e8888 (2010).
Author Information Data have been deposited at GEO under accession number
GSE28682. Reprints and permissions information is available at www.nature.com/
reprints. The authors declare competing financial interests: details accompany the
full-text HTML version of thepaper at www.nature.com/nature. Readers are welcome to
comment on the online version of this article at www.nature.com/nature.
Correspondence and requests for materials should be addressed to S.A.
(Suneet.Agarwal@childrens.harvard.edu) or A.R. (arao@idi.harvard.edu or
arao@liai.org).
4. Harris, T. D. et al. Single-molecule DNA sequencing of a viral genome. Science 320,
106–109 (2008).
5. Bowers, J. et al. Virtual terminator nucleotides for next-generation DNA
sequencing. Nature Methods 6, 593–595 (2009).
1
9 M A Y 2 0 1 1 | V O L 4 7 3 | N A T U R E | 3 9 7
©
2011 Macmillan Publishers Limited. All rights reserved