Article
Roles of H3K27me2andH3K27me3Examined during
ryonic Stem Cells
Highlights
d H3K27me2/H3K27me3 are redistributed during ESC
differentiation
d H3K27 methylation recapitulates changes during
differentiation in Ezh2-edited ESCs
d In pluripotent culture conditions, Ezh2-edited ESCs activate
the neurogenic program
d Ezh2-edited ESCs are refractory to a 2i-medium-induced
Authors
Aster H. Juan, Stan Wang,
Kyung Dae Ko, ..., Rafael Casellas,
Jizhong Zou, Vittorio Sartorelli
Correspondence
juana2@mail.nih.gov (A.H.J.),
sartorev@mail.nih.gov (V.S.)
modify H3K27 methylation states
(H3K27me2/H3K27me3) and report that
such perturbation influences the ability of
ESCs to differentiate into preferential cell
lineages and acquire a ‘‘ground state.’’Juan et al., 2016, Cell Reports 17, 1369–1382
October 25, 2016
http://dx.doi.org/10.1016/j.celrep.2016.09.087‘‘ground state’’Graphical AbstractFate Specification of EmbAccession Numbers
GSE85717Juan et al. use genome editing to subtlyIn Brief
Cell Reports
during Fate Specification of E
ss
ez
,5
tit
t R
Q
c
es
P
te
y
u
r
vmethylation (Margueron and Reinberg, 2011) and exerts impor-
tant functions in ESCs and induced pluripotent stem cells
(iPSCs; Boyer et al., 2006; Chamberlain et al., 2008; Lee et al.,
et al., 2015; Souroullas et al., 2016), results in markedly improved
H3K27 trimethylase activity, possibly via a Phe/Tyr switchmech-
anism (Collins et al., 2005).2006; Leeb et al., 2010; Montgomery et al., 2005; Pasini et al.,
2007; Shen et al., 2008; Onder et al., 2012; Riising et al., 2014).
Here, we exploit these biochemical properties and introduce
the Y641F mutation in one of the two Ezh2 alleles of ESCs viaSUMMARY
The polycomb repressive complex 2 (PRC2) methyl-
ates lysine 27 of histone H3 (H3K27) through its
catalytic subunit Ezh2. PRC2-mediated di- and tri-
methylation (H3K27me2/H3K27me3) have been
interchangeably associated with gene repression.
However, it remains unclear whether these two
degrees of H3K27 methylation have different func-
tions. In this study, we have generated isogenic
mouse embryonic stem cells (ESCs) with a modified
H3K27me2/H3K27me3 ratio. Our findings document
dynamic developmental control in the genomic dis-
tribution of H3K27me2 and H3K27me3 at regulatory
regions in ESCs. They also reveal that modifying
the ratio of H3K27me2 and H3K27me3 is sufficient
for the acquisition and repression of defined cell line-
age transcriptional programs and phenotypes and
influences induction of the ESC ground state.
INTRODUCTION
Polycomb repressive complex 2 (PRC2) is essential for H3K27
Their genome-wide distribution and distinct enrichment suggest
different functions for the individual H3K27 methylation states
(Cui et al., 2009; Steiner et al., 2011; Azuara et al., 2006; Bern-
stein et al., 2006; Boyer et al., 2006; Lee et al., 2006; Shen
et al., 2009). While H3K27me3 and H3K27me1 have been asso-
ciated with gene repression and activation, respectively (Mar-
gueron and Reinberg, 2011; Cui et al., 2009; Ferrari et al.,
2014), the role of H3K27me2 has remained less explored. Recent
studies conducted in temperature-sensitive mutant E(z)61
Drosophila EZ2-2 cells (Lee et al., 2015) and in PRC2 (Eed
/
)
knockout ESCs attribute a role to H3K27me2 in preventing inap-
propriate enhancer activation (Ferrari et al., 2014). The proposed
function of H3K27me2 in restricting random and unscheduled
transcription throughout the genome predicts its developmen-
tally regulated distribution at distinct regulatory regions. How-
ever, the temporal dynamics of H3K27me2 in ESCs are unclear.
Thus, the question of whether maintenance of an appropriate
balance between H3K27me2 and H3K27me3 is relevant for the
regulation of gene expression and acquisition of distinct cell
fates remains unanswered. Perturbation of the H3K27me2/
H3K27me3 ratio, without altering PRC2 stoichiometry, may
directly address the function of these two degrees of H2K37
methylation. Mutation of the PRC2 subunit Ezh2 tyrosine 641
to phenylalanine (Ezh2Y641F), occurring in B cell lymphomas
and melanomas (Yap et al., 2011; McCabe et al., 2012; BarsottiAster H. Juan,1,10,* Stan Wang,1,2,3,10,11 Kyung Dae Ko,1 Ho
Karinna O. Vivanco,1,4 Anthony M. Ascoli,1 Gustavo Gutierr
Roger A. Pedersen,3,8 Benjamin A. Garcia,6 Rafael Casellas
1Laboratory of Muscle Stem Cells and Gene Regulation, National Ins
National Institutes of Health, Bethesda, MD 20892, USA
2Wellcome Trust/Cancer Research UK Gurdon Institute, Tennis Cour
3Department of Surgery, University of Cambridge, Cambridge CB2 0
4Department of Biological Engineering, Massachusetts Institute of Te
5Genomics and Immunity, NIAMS, National Institutes of Health, Beth
6Epigenetics Program, Department of Biochemistry and Biophysics,
Philadelphia 19104 PA, USA
7Synaptic Function Section, The Porter Neuroscience Research Cen
National Institutes of Health, Bethesda, MD 20892, USA
8The Anne McLaren Laboratory for Regenerative Medicine, Universit
9iPSC Core Facility, Center for Molecular Medicine, National Heart, L
MD 20892, USA
10Co-first author
11Present address: Columbia University College of Physicians and Su
12Lead Contact
*Correspondence: juana2@mail.nih.gov (A.H.J.), sartorev@mail.nih.go
http://dx.doi.org/10.1016/j.celrep.2016.09.087Article
Roles of H3K27me2 and H3KThis is an open access article under the CC BY-N27me3 Examined
mbryonic Stem Cells
ein Zare,1 Pei-Fang Tsai,1 Xuesong Feng,1
-Cruz,1 Jordan Krebs,5 Simone Sidoli,6 Adam L. Knight,7
Jizhong Zou,9 and Vittorio Sartorelli1,12,*
ute of Arthritis, Musculoskeletal and Skin Diseases (NIAMS),
oad, Cambridge CB2 1QN, UK
Q, UK
hnology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA
da, MD 20892, USA
erlman School of Medicine, University of Pennsylvania,
r, National Institute of Neurological Disorders and Stroke,
of Cambridge, Downing Street, Cambridge, CB2 3EH, UK
ng and Blood Institute, National Institutes of Health, Bethesda,
geons, 630 West 168th Street, New York, NY 10032, USA
(V.S.)Cell Reports 17, 1369–1382, October 25, 2016 1369
C-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
transcription activator-like effector nuclease (TALEN)-mediated
genome editing. We document specific and dynamic genome-
wide distribution of H3K27me2 and H3K27me3 in undifferenti-
ated and differentiating ESCs and find that modification of the
H3K27me2/H3K27me3 ratio results in the acquisition and
repression of defined cell lineages, as well as influencing the
induction of the ESC ground state.
RESULTS
Distinct H3K27 Methylation States Are Enriched at
Functionally Defined Genomic Regulatory Regions of
Different Classes of Genes in ESCs
We surveyed the genome-wide distribution of the three H3K27
methylation states (H3K27me1, H3K27me2, and H3K27me3) in
mouse ESCs by chromatin immunoprecipitation followed by
sequencing (ChIP-seq). Antibody specificities were documented
by dot blot and immunoprecipitation-immunoblot assays with a
battery of antibodies recognizing different degrees of H3K27
methylation (Figure S1A). The number of replicates, computed
peaks for the histone marks, and RNA sequencing (RNA-seq)
reads are reported in Figure S1B. Analysis of the ChIP-seq data-
sets revealed that27%of H3K27me3,8%of H3K27me2, and
20% of H3K27me1 peaks were enriched at promoter regions,
respectively (Figure S2A). H3K27me1 was enriched at transcrip-
tion start sites (TSSs) and in gene bodies (Figure S2B). Different
percentages of H3K27 methylation were also observed at intra-
genic regions: 64% of H3K27me1, 44% of H3K27me2, and
30% of H3K27me3 signals (Figure S2A). Similar percentages of
H3K27me3 and H3K27me2 (42% and 47%, respectively)
were found at intergenic regions, whereas this genomic
compartment hosted only 15% of H3K27me1 peaks (Fig-
ure S2A). To gain initial insight into the regulatory regions
occupied by H3K27me2 and H3K27me3, we positioned their
respective averaged normalized tag densities at TSSs or
genomic regions characterized by the presence of DNase-I-
hypersensitive sites (DHSs) derived from a collection of publicly
available DHS datasets in ESCs, different somatic cells, and
tissues (see Experimental Procedures). TSSs were excluded
from the considered DHSs, which therefore are likely to encom-
pass prospective active regulatory regions. The H3K27me3
signal was mainly enriched at regions immediately surrounding
the TSS (±1 kb) (Figure S2C, blue trace), whereas H3K27me2
preferentially accumulated at prospective DHS regions (Fig-
ure S2D, red trace). To further define the genomic location of
H3K27 methylation, we intersected H3K27me1, H3K27me2, or
H3K27me3 peaks with either H3K4me3 at TSSs (H3K4me3+/
TSSs) or H3K4me1, a histone mark preferentially enriched at
distant regulatory regions (Heintzman et al., 2007). The bulk of
H3K27me3+/TSSs (88%) were H3K4me3+, whereas compa-
rably fewer H3K27me2+/TSSs (55%) were co-occupied by
H3K4me3. More than 90% of the H3K27me1 signal occurred
at H3K4me3+/TSSs (Figure S2E). At H3K4me1+ regions, the per-
+ +centages of H3K27me3 regions and H3K27me2 regions were
similar (44% and 48%, respectively), whereas H3K27me1 was
less represented (27%) (Figure S2F). To evaluate whether the
catalytic PRC2 subunit Ezh2 was differentially enriched at
H3K27me3 and H3K27me2 regions, we conducted Ezh2 ChIP-
1370 Cell Reports 17, 1369–1382, October 25, 2016seq experiments. Scatterplot analysis revealed a positive corre-
lation for H3K27me3 and negative correlation for H3K27me2
with Ezh2 binding (Figure S3A). Both H3K27me3 and
H3K27me2 regions were equally enriched for guanines and cyto-
sines and displayed enrichment for a similar set of DNA-binding
motifs (Figures S3B and S3C).
Next, we investigated the transcriptional state of H3K27me3+,
H3K27me2+, and H3K27me1+ genes. The majority of
H3K27me3+ genes were repressed or expressed at low levels
(reads per kilobase of transcription per million [RPKM] < 1)
(53%), and the remaining 47% were expressed at higher
levels (RPKM > 1). H3K27me3+ genes with higher expression
(RPKM > 1) were enriched for early developmental processes,
whereas lowly expressed genes (RPKM<1) encompassed those
involved in cell-cell signaling, neuron differentiation, and cell fate
commitment (Figure S2G; Tables S1 and S2). H3K27me2+ genes
were mainly expressed at higher levels (64%, RPKM > 1) and
corresponded to metabolic processes (Figure S2H; Tables S1
and S2). Lowly expressed (RPMK < 1) H3K27me2+ genes iden-
tified terms related to immune responses (Figure S2H; Tables
S1 andS2).More than 90%of theH3K27me1+ geneswere highly
expressed and corresponded to those involved in non-coding
RNA metabolic processes (rRNA and tRNA), cellular responses
to stress, and cell-cycle control (Figure S2I). Overall, these
data indicate that in ESCs, H3K27me3, H3K27me2, and
H3K27me1 are distributed at distinct genomic regions and are
associated with genes related to discrete biological processes
in different transcriptional states.
Developmental Dynamics of H3K27 Methylation States
in ESCs
To assess the developmental dynamics of H3K27 post-transla-
tional modifications (PTMs), we conducted ChIP-seq for
H3K27me3, H3K27me2, H3K27me1, and H3K27 acetylation
(ac) in ESCs cultured in conditions either maintaining pluripo-
tency (high serum, on feeders, with leukemia inhibitory factor
[LIF]) or inducing differentiation (cultured for 24 hr at low density,
without serum and without LIF). Regions with the highest
H3K27me3 enrichment (top quartile) in pluripotent ESCs prefer-
entially retained this histone modification in differentiating
ESCs, whereas H3K27me2+ regions were less likely to acquire
H3K27me3 (Figure 1A). H3K27me3 was intersected with
H3K4me3+/TSSs and H3K4me1+ maps. Approximately 53% of
bivalent genes (H3K27me3+/H3K4me3+) retained H3K27me3+
upon differentiation (Figure 1B, top). The percentage of
H3K27me3+ retention in differentiating ESCs was higher at
repressed regions (64%at H3K27me3+/H3K4me3
 regions; Fig-
ure 1B, middle) and poised enhancers (61% at H3K27me3+/
H3K4me1+ regions; Figure 1B, bottom). In differentiating ESCs,
H3K27me3 was retained at genomic regions associated
(±20 kb from TSSs) with repressed genes related to develop-
mental processes (Figure 1C; Table S3). In contrast to
H3K27me3 (Figure 1D, blue trace), regions with the highestH3K27me2 enrichment in pluripotent ESCs were more prone
to acquire H3K27ac during differentiation (Figure 1D, red trace).
H3K27me2 displayed more dynamic behavior than H3K27me3,
with 43% of H3K27me2+/H3K4me3+/TSSs and 28%
H3K27me2+/H3K4me3
 regions becoming H3K27ac (Figure 1E,
(legend on next page)
Cell Reports 17, 1369–1382, October 25, 2016 1371
top and middle, respectively). Intriguingly, a similar percentage
of poised enhancers (16%; Figure 1B bottom) and
H3K27me2+/H3K4me1+ regions (18%; Figure 1E, bottom)
transitioned to H3K27ac in differentiating ESCs. Thus, while
occurring at enhancers (Ferrari et al., 2014), an H3K27me2-to-
H3K27ac switch was more prevalent at promoters. Regulatory
regions undergoing this switch were located in the vicinity
(±20 kb from TSSs) of upregulated genes related to metabolic
processes (Figure 1F; Table S3). Compared to pluripotent
ESCs, the average normalized tag density of intergenic
soluble and insoluble (chromatin), fractions and found decreased
H3K27me2 and increased H3K27me3 histones in Y641F com-
pared to wild-type (WT) ESCs in the chromatin nuclear-insoluble
fraction (NI) (Figure 2B). These findings were substantiated by
quantitative liquid chromatography-tandem mass spectrometry
(LC-MS/MS) analysis documenting diminished H3K27me2 in
Y641F ESCs (Figure S4A). Increased H3K27me3 in Y641F ESCs
occurredat thehistonevariantH3.3 (FigureS4A),which is required
for proper establishment of H3K27me3 at the promoters of devel-
opmentally regulated genes of ESCs (Banaszynski et al., 2013).
S
S
is
3+
hm
enH3K27me3 peaks was increased in both ESC differentiated for
24 hr (ESC-24h) and more differentiated somatic cells (skeletal
myoblast [MB]) (Figure 1G, left). Conversely, the intergenic
H3K27me2 signal was decreased in ESC-24h and, more pro-
foundly, in skeletal myoblasts (Figure 1G, right). Almost half
(48%) of H3K27me1+/H3K4me3+/TSS regions acquired
H3K27ac in differentiating ESCs (Figure S3D, top) whereas
similar H3K27ac percentages were observed for H3K27me1+/
H3K4me3
 and H3K27me1+/H3K4me1+ regions, respectively
(Figure S3D, middle and bottom). Average normalized
H3K27me1 tag density was comparable in pluripotent and
differentiating ESCs (Figure S3E). Collectively, these data indi-
cate that during ESC differentiation, H3K27me3, H3K27me2,
and H3K27me1 are dynamically redistributed at regulatory re-
gions of distinct classes of genes.
TALEN-Mediated Genome Editing of Ezh2 Modifies
H3K27me2 and H3K27me3 in ESCs
To perturb H3K27 methylation, we generated isogenic ESCs via
TALEN-mediated genome editing (Bogdanove and Voytas,
2011) of the Ezh2 coding region, causing the replacement of
Tyr641 with phenylalanine (Y641F) (Figure 2A). This mutation, de-
tected in a subset of B cell lymphomas andmelanomas, results in
the formation of Ezh2Y641F-containing PRC2 complexes with
increased enzymatic activity on H3K27me2 substrates, thereby
shifting the steady state of H3K27 to favor in vivo trimethylation
(Yap et al., 2011; McCabe et al., 2012; Barsotti et al., 2015; Sour-
oullas et al., 2016). Maintenance of one Ezh2 wild-type allele en-
sures generation of H3K27 mono- and dimethylated substrates
upon which Ezh2Y641F can act (Yap et al., 2011). To confirm
that Ezh2 genome editing in ESCs would recapitulate H3K27
methylation modifications observed in neoplastic cells, we per-
formed immunoblot analysis of cytoplasmic, along with nuclear-
Figure 1. Dynamic Distribution of H3K27 Methylation States during E
(A) Average normalized tag density profiles depicting H3K27me3 enrichment in E
(blue trace) or H3K27me2+ (red trace) in pluripotent ESCs.
(B) Transition of H3K27me3 across bivalent versus repressed promoters and po
(ESC-24h).
(C) GO-biological processes associated with genes within ±20 kb of H3K27me
area). Expression cutoffR 1.5-fold.
(D) Average normalized tag density profiles depicting H3K27 acetylation enric
H3K27me3+ (blue trace) or H3K27me2+ (red trace) in pluripotent ESCs.
(E) Transition of H3K27me2 across H3K4me3+/
 promoters and H3K4me1+(ESC-24h).
(F) GO-biological processes associated with genes within ±20 kb of H3K27me2+ r
cutoffR 1.5-fold.
(G) Boxplot of H3K27me3 (left) and H3K27me2 (right) peak intensities for ±5-kb in
ESC developmental stages and in skeletal myoblasts (MBs). p values were deter
1372 Cell Reports 17, 1369–1382, October 25, 2016H3K27me1 was not significantly affected by Y641F (Figures S4A
and S4C). The levels of the PRC2 subunits Ezh2, Suz12, and
Eed were comparable in WT and Y641F ESCs (Figures 2B and
2C). Similar biochemical features were observed in two indepen-
dently isolated Y641F ESC clones (Figure S4B). To study their
genomic distribution, we performed H3K27me2 and H3K27me3
ChIP-seq in WT and Y641F ESCs. Since H3K27me1 was not
modified in Y641F ESCs, we did not further pursue this methyl-
ation state. Averaged tag density of H3K27me3 was increased
in Y641F ESCs at a genomic interval encompassing an area ±5
kb surrounding the TSS (Figure 2D), whereas within the same
genomic intervals, H3K27me2 was decreased (Figure 2E). A
similar analysis performed by centeringH3K27methylation peaks
at prospective DHS regions confirmed increased H3K27me3 and
decreased H3K27me2 in Y641F ESCs (Figures 2F and 2G).
H3K27me2 and H3K27me3 States of Ezh2-Edited ESCs
Resemble Those of Differentiated ESCs
We next sought to compare the genomic distribution of
H3K27me2 and H3K27me3 in pluripotent and differentiating
ESC-24h, embryoid bodies (EBs), and Y641F ESCs. For this,
WT and Y641F ESCs were cultured in conditions favoring main-
tenance of pluripotency. Alternatively, serum and LIF were with-
drawn for 24 hr from ESCs (ESC-24h), and EBs (day 8) were
derived from WT ESCs.
Principal-component analysis (PCA) was employed to mea-
sure and visualize differences in H3K27me2 and H3K27me3
maps between different ESC states in distinct genomic windows
(PCA1, PCA2, and PCA3). This analysis permitted an effective
segregation of the different ESC states, as depicted by 3D pro-
jections (Figures 3A and 3C). H3K27me2 PCA maps revealed
that Y641F ESCs cultured in pluripotent conditions clustered
closer to ESC-24h than to pluripotent ESCs (Figure 3A).
C Differentiation
C-24h for prospective DHS regions within the top quartile of either H3K27me3+
ed enhancers in ESC to different H3K27 states during early ESC differentiation
regions in ESCs that maintain their H3K27me3 status in ESC-24h (shadowed
ent in ESC-24h for prospective DHS regions within the top quartile of either
hancers in ESCs to different H3K27 states during early ESC differentiationegions in ESCs that acquire H3K27ac in ESC-24h (shadowed area). Expression
tergenic genomic regions surrounding the summit of each peak at two different
mined by Student’s t test.
Approximately half (50%) of the genes with reduced H3K27me2
in Y641F ESCs corresponded to those acquiring H3K27 acetyla-
tion (Figure 3B). When correlated with gene expression, Gene
Ontology (GO) analysis of upregulated genes returned terms
related to metabolic processes (Figure 3B; Table S4), similar to
those identified in ESC-24h (Figure 1F; Table S3). H3K27me3
PCA maps indicated that Y641F ESCs clustered closer to EB8
and ESC-24h than to pluripotent ESCs (Figure 3C). H3K27me3
PCA maps of all ESC states and EBs were markedly segregated
from those of somatic cells and tissue types (Figure 3C). Thema-
jority (21%) of the regions displaying de novo H3K27me3 in
Y641F ESCs had no detectable H3K27me2 in WT ESCs, over-
lapped with regions featuring increased H3K27me3, and were
assigned by proximity (±20 kb) to genes whose expression
was repressed in ESC-24h (Figure 3D). GO analysis of this
gene list identified terms related to developmental processes
(Figure 3D; Table S4), similar again to those identified in ESC-
24h (Figure 1C; Table S3). During developmental lineage speci-
fication of mouse and human ESCs, H3K27me3 increases at
intergenic regions (Hawkins et al., 2010; Pauler et al., 2009;
Zhu et al., 2013). Our analysis further indicates that compared
is in line with previou
ESCs have lower H3K2
tiated cells (Hawkins e
2013).
Ezh2-Edited ESCs P
Fate Program and Ar
Naive Ground State
The distinct PRC2-me
served prompted us to
WT and Y641F ESCs. I
against the pluripoten
1998; Avilion et al., 20
cific embryonic antige
in WT and Y641F ES
cence-activated cell so
gauge SSEA-1 express
and Y641F ESCs (Figur
showed an equivalent
(Figure 4B, right). Withi
are phenotypically indi
Cell Repo(F and G) Averaged normalized tag density pro-
files of H3K27me3 (F) or H3K27me2 (G) for
genomic regions surrounding (±2 kb) prospective
DHS regions.
to WT ESCs, the H3K27me3 signal in
both ESC-24h and Y641F ESCs spreads
laterally from the TSS and invades
intergenic regions (Figures 3E–3G). In
agreement with the H3K27me3 PCA re-
sults (Figure 3A), genomic H3K27me3
distribution in ESC-24h was more similar
to that of Y641F ESCs than WT ESCs,
with reduced representation at pro-
moters and increased occupancy of in-Figure 2. TALEN-Mediated Genome Edit-
ing of Ezh2 Modifies H3K27me2 and
H3K27me3 in ESCs
(A) Illustration of TALEN-mediated genome editing
creating the Ezh2Y641F allele in ESCs.
(B) Western blot for Ezh2, tubulin, H3K27me2
(me2), H3K27me3 (me3), and total histone H3
in WT or Y641F ESC chromatin fractions. NI,
nuclear-insoluble fraction; NS, nuclear-soluble
fraction; C, cytoplasmic fraction.
(C) Western blot for PRC2 components Ezh2,
Suz12, and Eed in WT or Y641F ESC whole-cell
lysates.
(D and E) Boxplots of H3K27me3 (D) and
H3K27me2 (E) peak intensities for genomic re-
gions surrounding (±5 kb) TSSs in WT and Y641F
ESCs. p values were determined by Student’s
t test.tergenic regions (Figures 3F and 3G).
Altogether, these findings reveal that
H3K27me2 and H3K27me3 maps of
Y641F ESCs resemble those of ESCs un-
dergoing differentiation. This observation
s reports documenting that pluripotent
7me3 levels compared to more differen-
t al., 2010; Pauler et al., 2009; Zhu et al.,
referentially Activate the Neural
e Refractory to the 2i-Induced
diated H3K27 methylation patterns ob-
investigate potential differences between
mmunostaining with an antibody directed
t markers Pou5f1 (Oct4) (Nichols et al.,
03) and mouse ESC-specific stage-spe-
n 1 (SSEA-1) indicated similar reactivity
Cs (Figure 4A). We employed fluores-
rting (FACS) analysis to more accurately
ion and found it to be comparable in WT
e 4B, left). Lastly, a growth curve analysis
proliferation rate for WT and Y641F ESCs
n these parameters, WT and Y641F ESCs
stinguishable.
rts 17, 1369–1382, October 25, 2016 1373
Figure 3. H3K27me2 and H3K27me3 States in Ezh2-Edited Y641F ESCs Resemble Those of Differentiating ESCs
(A) Three-dimensional principal component plots of H3K27me2 in ESC-WT, ESC-24h, and ESC-Y641F.
(B) GO-biological processes associated with genes within ±20 kb of regions with decreased H3K27me2 in ESC-Y641F and increased H3K27ac in ESC-24h
(shadowed area) and withR1.5-fold increased expression in Y641F compared to WT ESCs.
(C) Three-dimensional principal-component plots of H3K27me3 in ESC-WT, ESC-24h, ESC-Y641F, day 8 embryoid bodies (EB8), and various cell or tissue types.
(D) GO-biological processes associated with genes within ±20 kb of regions with increased H3K27me3 in ESC-Y641F and increased H3K27me3 in ESC-24h
(shadowed area) and withR1.5-fold decreased expression in Y641F compared to WT ESCs.
(E) Unsupervised clustering heatmap of H3K27me3 at genomic regions surrounding (±5 kb) TSSs in ESC-WT, ESC-24h, and ESC-Y641F.
(F) Bar graphs depicting the distribution of H3K27me3 across genomic regions in ESC-WT, ESC-24h, and ESC-Y641F.
(G) H3K27me3 signal tracks for representative loci Hey1 and T in ESC-WT, ESC-24h, and ESC-Y641F.
1374 Cell Reports 17, 1369–1382, October 25, 2016
Next, we performed RNA-seq on WT and Y641F ESCs
cultured in conditions favoring maintenance of pluripotency.
PCA of three independent RNA-seq biological replicates segre-
gated WT and Y641F ESCs into two distinct expression profiles,
indicating specific gene expression patterns in the two cell pop-
ulations (Figure 4C). When considering developmental pro-
Figure 4. Ezh2-Edited ESCs Preferentially Activate the Neural Fate Pro
(A) Morphology (bright field) and pluripotent marker (POU5F1 and SSEA-1) immu
(B) FACS analysis of SSEA-1 expression (left) and growth curve (right) of WT and
(C) Two-dimensional principal-component plots of three RNA-seq biological rep
(D) GO-biological processes associated with genes with R1.2-fold decreased
WT ESCs.
(E) Fold-change expression (RPKM) from three RNA-seq biological replicates
mesodermal genes (T and Des), and endodermal genes (Gata4 and Gata6) betwcesses, GO analysis of downregulated transcripts in Y641F
ESCs returned terms related to mesoderm (blood vessel and
skeletal) and epithelium development. Instead, forebrain devel-
opment was a term represented in the categories of upregulated
transcripts (Figure 4D; Table S4). We compared published
mesodermal and neuronal gene signatures with those of control
gram
nostaining of WT and Y641F ESCs.
Y641 ESCs.
licates depicting gene expression profiles of WT and Y641F ESCs.
(green bars) or increased (orange bars) expression in Y641F compared to
of pluripotent genes (Pou5f1 and Nanog), neural genes (Sox1 and Pax6),
een WT and Y641F ESCs. Data are presented as mean ± SD.
Cell Reports 17, 1369–1382, October 25, 2016 1375
and Y641F ESCs by gene set enrichment analysis (GSEA) (Sub-
ramanian et al., 2005). While the gene set from control ESCs
(ES-WT) was anti-correlated, the Y641F ESCs gene set was
positively correlated with cell-lineage-associated genes, espe-
cially with the signature for neural stem cells (Figure S5A). Moti-
vated by these observations, we queried the expression of
individual mesodermal and neural-specific transcripts. In
concordance with immunostaining results, Pou5f1 (Oct4) tran-
scripts were similarly represented in WT and Y641F ESCs.
Expression of Nanog, another core pluripotency factor (Mitsui
et al., 2003; Chambers et al., 2003), was also comparable in
WT and Y641F ESCs (Figure 4E; Table S1). In contrast, tran-
scripts forSox1 andPax6, two of the earliest neural markers (Ma-
stick et al., 1997; Schwarz et al., 1999; Pevny et al., 1998; Wood
and Episkopou, 1999), were increased, whereas mesodermal
markers brachyury (T) and desmin (Des) and endodermal
Gata4 and Gata6 transcripts were decreased in Y641F ESCs
(Figure 4E; Table S1). Thus, while retaining expression of plurip-
otency factors, Y641F ESCs initiate expression of early neural
genes and repress that of mesodermal and endodermal genes,
even when cultured in a pluripotency-promoting medium.
Culturing ESCs in defined medium containing mitogen-acti-
vated protein/extracellular signal-related kinase (MEK) and
glycogen synthase kinase 3 (GSK3) inhibitors (2i) reduces
morphological heterogeneity and mosaic expression of pluripo-
tency factors, mimics the environment of thematuremouse inner
cell mass, and has allowed derivation of germline-competent
ESCs from rats (Kalkan and Smith, 2014; Marks et al., 2012;
Buehr et al., 2008). A lower expression of lineage-affiliated genes
is consistent with acquisition of the naive ground state of 2i ESCs
(Marks et al., 2012). To evaluate whether the transcriptome could
be reprogrammed by 2i, we transferred serum-cultured control
and Y641F ESCs to 2i medium and allowed them to adapt to
the new culture conditions for two or three passages. Next, we
generated and analyzed RNA-seq datasets from control and
Y641F ESCs either cultured in serum or transferred to 2i.
Pathway KEGG (Kyoto Encyclopedia of Genes and Genomes)
analysis was employed to identify global transcriptional changes
(Marks et al., 2012). From this analysis, it emerged that pathways
enriched in 2i-control and Y641F ESCs were distinct (Figure 5A;
Table S5). It was also evident that 2i-Y641F ESCs were enriched
for pathways identified in ESCs cultured in either serum (path-
ways to cancer, focal adhesion) or 2i (glycolysis, propionate
metabolism) (Table S5) (Marks et al., 2012). The most enriched
pathway (axon guidance) was unique to 2i-Y641F ESCs (Fig-
ure 5A). Since expression of developmental regulators is
affected by 2i (Marks et al., 2012), we queried this category of
genes. Relative to serum conditions, naive factors Nanog, Esrrb,
Klf5, and Zfp42/Rex1, along with pluripotent factors Pou5f1
(Oct4) and Sox2, were modestly reduced in 2i Y641F ESCs (Fig-
ure 5B). Mesodermal and endodermal developmental regulators
were overall decreased in Y641F ESCs cultured in either serum
or 2i conditions. In contrast to control ESCs (Table S5) (Markset al., 2012), expression of ectodermal regulators was generally
increased in both serum and, with the exception of Sox1, 2i-
Y641F ESCs (Figure 5B). When cultured in 2i medium, ESCs
generate uniform, compact, and small dome-shaped refractile
colonies. In contrast, ESCs in serum are heterogeneous in
1376 Cell Reports 17, 1369–1382, October 25, 2016morphology and flattened, with irregular borders and discern-
able individual cells (Wray et al., 2011; Marks et al., 2012).
Accordingly, both control and Y641F ESCs cultured in serum
or LIF and on feeders formed flattened colonies with discernable
individual cells, whereas 2i conditions conferred morphological
characteristics of the naive ground state only to WT ESCs.
Y641F ESCs were partially refractory to 2i-induced morpholog-
ical changes (Figure 5C). Altogether, enrichment of pathways
upregulated in serum conditions, reduced expression of naive
and pluripotent factors, increased expression of ectodermal reg-
ulators, and resistance to 2i-induced morphological changes
suggest that Y641F ESCs, even when cultured in 2i conditions,
retain transcriptional and morphological characteristics associ-
ated with a more primed or differentiated state.
The Neuronal Program Is Retained in Ezh2-Edited
ESC-Derived Embryoid Bodies
We next asked whether the bias toward neural commitment dis-
played by Y641F ESCs was maintained at later stages of differ-
entiation. AdherentWT andY641F ESCswere cultured for 2 days
in a medium (N2B27) supporting ESC differentiation to neural
precursors (Ying et al., 2003). While Pou5F1 levels decreased,
expression of the neural markers Sox1 and Pax6 remained
more elevated in Y641F ESCs (Figure S5B). To further investigate
the developmental fate of Y641F ESCs, we platedWT and Y641F
ESCs in conditions preventing adherence and isolated floating
EBs after culturing them in suspension for 8 and 13 days, respec-
tively. GO analysis of RNA-seq datasets identified genes
transcripts involved in neuronal differentiation among those up-
regulated in both 8- and 13-day Y641F ESC-derived EBs (Fig-
ure 5D; Table S5). Finally, EBswere plated and allowed to adhere
on a tissue culture dish for 4 days in standard culture conditions.
EBs derived from WT ESCs displayed clear signs of mesoderm
differentiation, as indicated by the presence of several fields of
beating cardiomyocytes (Movie S1). Beating cardiomyocytes
were instead not detectable in cultures of Y641F ESC-derived
EBs (Figure 5E). In contrast, structures resembling neuronal pro-
jections and immunostaining positive for Tuj1, an antibody
against neuron-specific class III b-tubulin, were prevalent in
Y641F ESC-derived EBs (Figure 5E). Finally, we injected WT
and Y641F ESCs into severe combined immunodeficiency
(SCID) mice to induce teratoma formation. Histological analysis
of teratomas revealed the presence of ectoderm, mesoderm,
and endoderm derivatives (Figure S5C), indicating that Y641F
ESCs retained developmental competency. Expression of neural
transcripts Sox1, Hes5, and Ngn1 was increased, while meso-
dermal brachyury (T) and endodermal Gata4, Gata6, and
Sox17 transcripts were decreased in teratomas derived from
Y641F ESCs (Figure S5D). Overall, these findings document a
propensity for ESCs with an increased H3K27me3/H3K27me2
ratio to acquire the neural cell fate.
IncreasedH3K27me3 and Transcriptional Repression of
the TGF-b Pathway Underlies Neural Commitment of
Ezh2-Edited ESCs
Vertebrate neuronal differentiation from uncommitted ESCs is
initiated via cell-intrinsic mechanisms involving signals inhibiting
the activity of pro-neuronal pathways (Hemmati-Brivanlou and
Melton, 1997). We reasoned that studying the pathways
repressed in Y641F ESCs might provide novel insight into these
mechanisms. Pathway KEGG analysis of RNA-seq datasets re-
vealed that the transforming growth factor b (TGF-b) pathway
was repressed in both ESC-24h and Y641F ESCs (Figures S6A
Figure 5. Ezh2-Edited ESCs Are Refractory to the 2i-Induced Naive G
Embryoid Bodies
(A) Pathway KEGG analysis for genes withR1.2-fold increased (top) or decrease
(B) Heatmap illustrating the RNA expression fold change between Y641F and W
(C) Morphology of WT and Y641F ESCs cultured in 2i medium.
(D) GO-biological processes associated with genes withR1.5-fold increased ex
(E) Percentage of ESC colonies positive for beating cardiomyocytes (top left) or ne
of adherent EBs grown from WT or Y641F ESCs.and 6A; Table S5). Inhibition of the TGF-b family of growth factors
induces neuralization and brain formation in amphibians (Hem-
mati-Brivanlou and Melton, 1994; Reversade et al., 2005) and
enhances ESC neuronal differentiation (Tropepe et al., 2001).
Thus, we queried expression of TGF-b family members TGFb1,
round State and Retained Their Neural Program in Differentiating
d (bottom) expression in Y641F compared to WT ESCs cultured in 2i medium.
T cultured in serum and 2i medium for selected genes of different lineages.
pression in Y641F ESCs, EB8, or EB13 compared to WT EB8 or EB13.
uronal projections (bottom left), along with neural marker Tuj1 immunostaining
Cell Reports 17, 1369–1382, October 25, 2016 1377
Figure 6. Increased H3K27me3 and Transcriptional Repression of the TGF-b Pathway Underlie Neural Commitment of Ezh2-Edited ESCs
Cultured in Serum
(A) Pathway KEGG analysis for genes withR1.2-fold increased expression in Y641F compared to WT ESCs.
(B) Fold-change expression (RPKM) from three RNA-seq biological replicates for TGF-b family members between WT and Y641F ESCs. Data are presented as
mean ± SD.
(C) H3K27me3 and H3K27ac signal tracks and RNA-seq traces at TGF-b1 and BMP4 loci in WT and Y641F ESCs.
(D and E) Fold-changeSox1 expression before or after the addition of TGF-b1 (D) or BMP4 (E) recombinant proteins inWT andY641F ESCs. Data are presented as
mean ± SD (n = 3).
1378 Cell Reports 17, 1369–1382, October 25, 2016
TGFb2, and TGFb3, along with bone morphogenetic protein
1(BMP1) and BMP4, and found their transcripts to be reduced
in Y641F ESCs (Figure 6B). ChIP-seq and RNA-seq profiles of
the TGFb1 and BMP4 loci documented increased and de novo
acquiredH3K27me3, decreasedH3K27ac, and reduced expres-
sion in Y641F ESCs (Figure 6C). Reduced TGFb1 and BMP4
expression was also observed in ESC-24h (Figure S6B), sug-
gesting a physiologic role for this phenomenon.
To establish whether repression of the TGF-b pathway was
causally linked to increased expression of the neural marker
Sox1 (Figure 4E), we cultured WT and Y641F ESCs in the pres-
ence of either TGFb1 or BMP4 recombinant protein. When
exposed to either TGF-b factors, Sox1 expression was specif-
ically and significantly decreased in Y641F ESCs (Figures 6D
and 6E). Thus, increased H3K27me3 at both the TGFb1 and
BMP4 loci is associated with their repression and Sox1 activa-
tion in Y641F ESCs, while re-establishment of TGFb1 and
BMP4 signaling results in Sox1 levels comparable to those
observed in WT ESCs.
DISCUSSION
In ESCs, H3K27me3 is mainly enriched at bivalent promoters
carrying H3K4me3 (Azuara et al., 2006; Bernstein et al., 2006).
During the chromatin reorganization process accompanying
ESC differentiation into somatic cells and tissues, H3K27me3
is re-distributed to much larger genomic areas (Pauler et al.,
2009; Hawkins et al., 2010; Zhu et al., 2013). H3K27me3 recon-
figuration may be related to the need for silencing inappropriate
gene expression required for cell lineage specification. In addi-
tion to H3K27me3, PRC2 also catalyzes formation of the precur-
sor H3K27me2 with a higher efficiency, a phenomenon reflected
by the higher abundance of H3K27me2 compared to H3K27me3
in ESCs (Jung et al., 2010; Voigt et al., 2012).
In this study, we found the genome-wide distribution of
H3K27me3 and H3K27me2 to differ in pluripotent and differenti-
ating ESCs, with the two degrees of H3K27 methylation being
enriched at functionally distinct genomic regulatory regions of
different classes of genes. In pluripotent ESCs, H3K27me3+ reg-
ulatory regions are associated with developmental genes (Lee
et al., 2006; Boyer et al., 2006), which tend to remain
H3K27me3+ and repressed during initial ESC differentiation.
On the other hand, we found H3K27me2 to be preferentially en-
riched at metabolic genes in pluripotent ESCs and replaced by
acetylation in differentiating ESCs. Metabolic reprogramming
of pluripotent stem cells is emerging as key regulator of ener-
getics and epigenetics (Zhang et al., 2012; Ryall et al., 2015),
and our findings indicate that H3K27me2 may play an important
role in setting the transcriptional switch leading to metabolic
reprogramming.
Without modifying the stoichiometry of the PRC2 subunits, we
introduced, via TALEN-mediated genome editing, a single point
mutation in one of the two Ezh2 alleles of ESCs, causing Ezh2 topreferentially trimethylate H3K27 and thus effectively altering the
H3K27me3/H3K27me2 ratio. Genome-wide comparison of
H3K27me3 and H3K27me2 maps in pluripotent and differenti-
ating ESCs, along with Ezh2-edited (Y641F) ESCs, revealed
several common features between differentiating ESCs andY641F ESCs. Previous reports have shown that H3K27me3 do-
mains expand during ESC differentiation (Zhu et al., 2013; Wu
et al., 2015). We found that, similar to ESC differentiation,
H3K27me3 domains are also expanded in Y641F ESCs cultured
in conditions favoring maintenance of pluripotency. While the
majority of H3K27me3 domains were retained during ESC differ-
entiation, H3K27me2 domains were preferentially resolved to
either H3K27 acetylation or H3K27me3, suggesting a more plas-
tic role of this histone mark as compared to H3K27me3. Impor-
tantly, preferential enrichment of H3K27me2 in pluripotent ESCs
at metabolic genes destined to be H3K27 acetylated and acti-
vated in differentiating ESCs was recapitulated in Y641F ESCs,
further suggesting the possibility that H3K27me2 might have a
distinct and specific role in ESC metabolic reprogramming.
That the H3K27 PTMs experimentally generated by Ezh2
Y641Fmutation do occur during unperturbed ESC differentiation
makes this model system of value for further investigation.
While our findings clearly document a decreased H3K27me2/
H3K27me3 ratio in both differentiating and Y641F ESCs, com-
pared to pluripotent ESCs, a normalization strategy, such as
that provided by ChIP with reference exogenous genome
(ChIP-RX; Orlando et al., 2014), would be required to better
assess quantitative H3K27 methylation dynamics.
The H3K27 methylation commonalities observed in differenti-
ating ESCs and undifferentiated Y641F ESCs translated into
transcriptome similarities between the two ESC. Even when
maintained in culture conditions favoring pluripotency, Y641F
ESCs initiated transcription of genes, such as Sox1 and Pax6,
indicative of early neural commitment. Because of the inherent
tendency of ESCs cultured in serum to upregulate develop-
mental regulators, we cultured the cells in 2i conditions which
repress lineage-affiliated gene expression, rendering ESCs tran-
scriptionally similar to pre-implantation naive epiblast cells (Bor-
oviak et al., 2014; Ficz et al., 2013; Habibi et al., 2013). 2i Y641F
ESCs retained increased expression of ectodermal regulators
andwere refractory to 2i-inducedmorphological changes, which
are characteristics suggestive of a primed or differentiating
state. The TGF-b pathway was enriched in both control serum
ESCs and 2i Y641F ESCs but downregulated in serum Y641F
ESCs (Figures 5 and 6) (Marks et al., 2012). A possible interpre-
tation of these findings is that upregulation of pathways that drive
differentiation (e.g., TGF-b) is indicative of initial acquisition of
differentiation potential in control serum and 2i Y641F ESCs.
Further specification toward a more specialized neurogenic
cell fate in serum Y641F ESCs would result in repression of the
TGF-b pathway. In this sense, the bias toward neural commit-
ment of serum Y641F ESCs was retained during differentiation,
resulting in the formation of Y641F ESC-derived EBs with an
overt neuronal phenotype and reduced representation of meso-
dermal derivatives. The observation that dissociated Xenopus
laevis ectodermal cells acquire neural identity in the absence of
instructive signals led to the proposal of a ‘‘neural default’’
model, which posits that the inhibition of an inhibitor drives neur-alization of the ectoderm (Mun˜oz-Sanjua´n and Brivanlou, 2002).
As in amphibians, the default state of mouse ESCs is neural, and
inhibition of TGF-b signaling promotes the transition from ESCs
to primitive neural stem cells (Tropepe et al., 2001). Moreover,
human ESCs can be differentiated into neuroectoderm-like
Cell Reports 17, 1369–1382, October 25, 2016 1379
progenitors positive for Pax6 by inhibiting TGF-b and BMP (Gif-
ford et al., 2013). In differentiating ESCs and Y641F ESCs,
H3K27me3 at several members of the TGF-b superfamily was
increased and their expression reduced compared to undifferen-
tiated ESCs. Reactivation of the TGF-b pathway by exposing
Y641F ESCs to either TGFb1 or BMP4 returned Sox1 levels to
those observed in control ESCs. Thus, modification of the
H3K27me3/H3K27me2 ratio in Y641F ESCs results in gene
expression mimicking that occurring during physiologic ESC
differentiation.
Collectively, the observations reported here document
genome-wide H3K27me2 and H3K27me3 maps of pluripotent
and differentiating ESCs and ascribe a differential role to
H3K27me3 and H3K27me2 in regulating the cell fate specifica-
tion of ESCs. Since H3K27me1 was not affected in Y641F
ESCs, it seems reasonable to propose that the differences
observed in control and Y641F ESCs can be ascribed rather to
a modified H3K27me2/H3K27me3 ratio. Ezh2 is present in
distinct PRC2 complexes (Kuzmichev et al., 2004), and the pres-
ence or absence of defined PRC2-associated subunits influ-
ences its methylation efficiency (Sarma et al., 2008; Peng
et al., 2009; Shen et al., 2009; Pasini et al., 2010; Li et al.,
2010; Landeira et al., 2010; Sanulli et al., 2015). Thus, PRC2
composition may influence the degree of H3K27 methylation
and, in doing so, participates in regulating ESC biology. If this
is the case, in addition to sequential modulation of signaling
pathways to direct the differentiation of pluripotent stem cells
(Ben-Zvi and Melton, 2015), reagents affecting discrete H3K27
methylation statesmay be envisaged as tools to coax ESCdiffer-
entiation into defined cell lineages.
EXPERIMENTAL PROCEDURES
Tissue Culture
Embryonic day 14 (E14) ESCs were grown initially on CF-1 MEF feeder
(GlobalStem) and then passaged feeder-free on gelatin-coated dishes in ES
medium (DMEM [GIBCO], 15% HyClone fetal bovine serum [FBS] [GE Health-
care], GlutaMAX [Gibco], MEM nonessential amino acids [Gibco], sodium
pyruvate [Gibco], LIF [Millipore; ESGRO], 0.1 mM b-mercaptoethanol, and
antibiotics). For 24-hr differentiation, feeder-free ESCs were plated in low den-
sity (13 106/10-cm plate) for 24 hr in ESmedium (described above) without LIF
and serum. For EB culture, 1 3 106 feeder-free ESCs were passaged onto
10-cm ultra-low-attachment culture dishes (Sigma). Serum-cultured ESCs
were converted to ground state by culturing and passaging the cells in
ESGRO-2i medium (Millipore, #SF016-100) following the manufacturer’s in-
structions. EBs were maintained in 15% HyClone FBS, MEM nonessential
amino acids, and sodium pyruvate. For adherent culture, three EBs were
collected and plated per well in gelatin-coated six-well plates. Adherent EBs
were differentiated in 10% HyClone FBS, 5% horse serum (Gibco), GlutaMAX,
MEM nonessential amino acids, and 0.1mM b-mercaptoethanol. For culture in
N2B27 neural differentiation media, 33 105 feeder-free ESCs were passaged
onto gelatin-coated 10-cm culture dishes. The following day, dishes were
changed to N2B27 media (1:1 ratio of Neurobasal:DMEM/Ham’s F-12 [Life
Technologies] supplemented with 1:200 N2 [Life Technologies], 1:100 B27
[Life Technologies], and 0.1 mM b-mercaptoethanol). C2C12 myoblasts
were maintained in DMEM medium with 20% FBS (Gibco) as previously
described (Mousavi et al., 2012).Culture with Recombinant Proteins
ESCs were seeded feeder-free at a concentration of 1.8 3 106 per 10-cm
gelatin-coated dish. The following day, normal ES media was supplemented
with either recombinant mouse Tgfb1 (R&D Systems, 240-B-010) or Bmp4
1380 Cell Reports 17, 1369–1382, October 25, 2016(R&D, 5020-BP-010) protein at a concentration of 30 ng/mL or 40 ng/mL,
respectively.
Analysis of Cell Growth
To generate cell growth curves, ESCs (3 3 105) were seeded in triplicate onto
six-well gelatin-coated dishes on day 0. Cells were harvested and counted via
Cellometer K2 Image Cytometer (Nexcelom Bioscience) every day for 3 days.
Teratoma Formation and Analysis
A description of teratoma formation and analysis is provided in Supplemental
Experimental Procedures.
RNA Expression Analysis
A description of RNA expression analysis by qPCR is provided in Supple-
mental Experimental Procedures.
Western Blot Analysis
A description of western blot analysis is provided in Supplemental Experi-
mental Procedures.
Antibody Specificity Analysis
A description of antibody specificity analysis is provided in Supplemental
Experimental Procedures.
Immunofluorescence
A description of immunofluorescence of ESC and adherent EBs is reported in
Supplemental Experimental Procedures.
Flow Cytometry
A description of flow cytometry of ESC is provided in Supplemental Experi-
mental Procedures.
Mass Spectrometry
A description of mass spectrometry of histone H3 is provided in Supplemental
Experimental Procedures.
RNA-Seq
A description of RNA-seq procedures is provided in Supplemental Experi-
mental Procedures.
ChIP-Seq
A description of ChIP-seq procedures is provided in Supplemental Experi-
mental Procedures.
ChIP-Seq and RNA-Seq Data Analysis
Descriptions of data processing and bioinformatic analyses, including tag den-
sity, 6-mer analysis, motif enrichment, PCA, and heatmap analysis are pro-
vided in Supplemental Experimental Procedures.
TALEN Generation and Genome Editing
Adescription of TALENandEzh2Y641Fdonor construction, alongwithgenome
editing in ESCs, is provided in Supplemental Experimental Procedures.
Antibodies
A detailed list of antibodies is available in Table S1.
ACCESSION NUMBERS
The accession number for the ChIP-seq and RNA-seq datasets reported in this
paper is GEO: GSE85717.SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures,
six figures, five tables, and one movie and can be found with this article online
at http://dx.doi.org/10.1016/j.celrep.2016.09.087.
AUTHOR CONTRIBUTIONS
A.H.J., S.W., and V.S. conceived and designed the overall project and wrote
the manuscript with input from the other authors. A.H.J. and S.W. performed
most of the experiments. K.D.K. and H.Z. designed and conducted computa-
tional analysis. P.-F.T., X.F., K.O.V., A.A., G.G.-C. J.K., S.S., and A.L.K. per-
formed tissue culture, immunostaining, and mass spectrometry experiments.
J.Z. performed TALEN experiments. R.A.P., B.A.G., and R.C. provided advice
and technical support. All authors have read and approved the manuscript.
ACKNOWLEDGMENTS
We thank Prof. J.B. Gurdon and members of his laboratory in Cambridge, UK
for hosting S.W. during the initial part of this work. We also thank the personnel
of the NIAMS Sequencing, Light Imaging, and Flow Cytometry Facilities and of
the NHLBI Transgenic and Pathology Cores at the National Institutes of Health
(NIH). S.W. was supported in part by fellowships from the Gates Cambridge
Trust and NIH-Cambridge MD/PhD Program. S.S. and B.A.G. were supported
by a grant from the NIH (GM110174). This research was supported in part by
the Intramural Research Program of the NIAMS at the NIH.
Received: December 23, 2015
Revised: September 7, 2016
Accepted: September 26, 2016
Published: October 25, 2016
REFERENCES
Avilion, A.A., Nicolis, S.K., Pevny, L.H., Perez, L., Vivian, N., and Lovell-Badge,
R. (2003). Multipotent cell lineages in early mouse development depend on
SOX2 function. Genes Dev. 17, 126–140.
Azuara, V., Perry, P., Sauer, S., Spivakov, M., Jørgensen, H.F., John, R.M.,
Gouti, M., Casanova, M., Warnes, G., Merkenschlager, M., and Fisher, A.G.
(2006). Chromatin signatures of pluripotent cell lines. Nat. Cell Biol. 8, 532–538.
Banaszynski, L.A., Wen, D., Dewell, S., Whitcomb, S.J., Lin, M., Diaz, N.,
Elsa¨sser, S.J., Chapgier, A., Goldberg, A.D., Canaani, E., et al. (2013). Hira-
dependent histone H3.3 deposition facilitates PRC2 recruitment at develop-
mental loci in ES cells. Cell 155, 107–120.
Barsotti, A.M., Ryskin, M., Zhong, W., Zhang, W.G., Giannakou, A., Loreth, C.,
Diesl, V., Follettie, M., Golas, J., Lee, M., et al. (2015). Epigenetic reprogram-
ming by tumor-derived EZH2 gain-of-function mutations promotes aggressive
3D cell morphologies and enhances melanoma tumor growth. Oncotarget 6,
2928–2938.
Ben-Zvi, D., and Melton, D.A. (2015). Modeling human nutrition using human
embryonic stem cells. Cell 161, 12–17.
Bernstein, B.E., Mikkelsen, T.S., Xie, X., Kamal, M., Huebert, D.J., Cuff, J., Fry,
B., Meissner, A., Wernig, M., Plath, K., et al. (2006). A bivalent chromatin struc-
ture marks key developmental genes in embryonic stem cells. Cell 125,
315–326.
Bogdanove, A.J., and Voytas, D.F. (2011). TAL effectors: customizable pro-
teins for DNA targeting. Science 333, 1843–1846.
Boroviak, T., Loos, R., Bertone, P., Smith, A., and Nichols, J. (2014). The ability
of inner-cell-mass cells to self-renew as embryonic stem cells is acquired
following epiblast specification. Nat. Cell Biol. 16, 516–528.
Boyer, L.A., Plath, K., Zeitlinger, J., Brambrink, T., Medeiros, L.A., Lee, T.I.,
Levine, S.S., Wernig, M., Tajonar, A., Ray, M.K., et al. (2006). Polycomb com-
plexes repress developmental regulators in murine embryonic stem cells.
Nature 441, 349–353.
Buehr, M., Meek, S., Blair, K., Yang, J., Ure, J., Silva, J., McLay, R., Hall, J.,Ying, Q.L., and Smith, A. (2008). Capture of authentic embryonic stem cells
from rat blastocysts. Cell 135, 1287–1298.
Chamberlain, S.J., Yee, D., and Magnuson, T. (2008). Polycomb repressive
complex 2 is dispensable for maintenance of embryonic stem cell pluripo-
tency. Stem Cells 26, 1496–1505.Chambers, I., Colby, D., Robertson, M., Nichols, J., Lee, S., Tweedie, S., and
Smith, A. (2003). Functional expression cloning of Nanog, a pluripotency sus-
taining factor in embryonic stem cells. Cell 113, 643–655.
Collins, R.E., Tachibana, M., Tamaru, H., Smith, K.M., Jia, D., Zhang, X.,
Selker, E.U., Shinkai, Y., and Cheng, X. (2005). In vitro and in vivo analyses
of a Phe/Tyr switch controlling product specificity of histone lysine methyl-
transferases. J. Biol. Chem. 280, 5563–5570.
Cui, K., Zang, C., Roh, T.Y., Schones, D.E., Childs, R.W., Peng, W., and Zhao,
K. (2009). Chromatin signatures in multipotent human hematopoietic stem
cells indicate the fate of bivalent genes during differentiation. Cell Stem Cell
4, 80–93.
Ferrari, K.J., Scelfo, A., Jammula, S., Cuomo, A., Barozzi, I., St€utzer, A.,
Fischle, W., Bonaldi, T., and Pasini, D. (2014). Polycomb-dependent
H3K27me1 and H3K27me2 regulate active transcription and enhancer fidelity.
Mol. Cell 53, 49–62.
Ficz, G., Hore, T.A., Santos, F., Lee, H.J., Dean, W., Arand, J., Krueger, F.,
Oxley, D., Paul, Y.L., Walter, J., et al. (2013). FGF signaling inhibition in
ESCs drives rapid genome-wide demethylation to the epigenetic ground state
of pluripotency. Cell Stem Cell 13, 351–359.
Gifford, C.A., Ziller, M.J., Gu, H., Trapnell, C., Donaghey, J., Tsankov, A., Sha-
lek, A.K., Kelley, D.R., Shishkin, A.A., Issner, R., et al. (2013). Transcriptional
and epigenetic dynamics during specification of human embryonic stem cells.
Cell 153, 1149–1163.
Habibi, E., Brinkman, A.B., Arand, J., Kroeze, L.I., Kerstens, H.H., Matarese,
F., Lepikhov, K., Gut, M., Brun-Heath, I., Hubner, N.C., et al. (2013). Whole-
genome bisulfite sequencing of two distinct interconvertible DNA methylomes
of mouse embryonic stem cells. Cell Stem Cell 13, 360–369.
Hawkins, R.D., Hon, G.C., Lee, L.K., Ngo, Q., Lister, R., Pelizzola, M., Edsall,
L.E., Kuan, S., Luu, Y., Klugman, S., et al. (2010). Distinct epigenomic land-
scapes of pluripotent and lineage-committed human cells. Cell Stem Cell 6,
479–491.
Heintzman, N.D., Stuart, R.K., Hon, G., Fu, Y., Ching, C.W., Hawkins, R.D.,
Barrera, L.O., Van Calcar, S., Qu, C., Ching, K.A., et al. (2007). Distinct and pre-
dictive chromatin signatures of transcriptional promoters and enhancers in the
human genome. Nat. Genet. 39, 311–318.
Hemmati-Brivanlou, A., and Melton, D.A. (1994). Inhibition of activin receptor
signaling promotes neuralization in Xenopus. Cell 77, 273–281.
Hemmati-Brivanlou, A., and Melton, D. (1997). Vertebrate neural induction.
Annu. Rev. Neurosci. 20, 43–60.
Jung, H.R., Pasini, D., Helin, K., and Jensen, O.N. (2010). Quantitative mass
spectrometry of histones H3.2 and H3.3 in Suz12-deficient mouse embryonic
stem cells reveals distinct, dynamic post-translational modifications at Lys-27
and Lys-36. Mol. Cell. Proteomics 9, 838–850.
Kalkan, T., and Smith, A. (2014). Mapping the route from naive pluripotency to
lineage specification. Philos. Trans. R. Soc. Lond. B Biol. Sci. 369, 20130540.
Kuzmichev, A., Jenuwein, T., Tempst, P., and Reinberg, D. (2004). Different
EZH2-containing complexes target methylation of histone H1 or nucleosomal
histone H3. Mol. Cell 14, 183–193.
Landeira, D., Sauer, S., Poot, R., Dvorkina, M., Mazzarella, L., Jørgensen, H.F.,
Pereira, C.F., Leleu, M., Piccolo, F.M., Spivakov, M., et al. (2010). Jarid2 is a
PRC2 component in embryonic stem cells required for multi-lineage differen-
tiation and recruitment of PRC1 and RNA Polymerase II to developmental reg-
ulators. Nat. Cell Biol. 12, 618–624.
Lee, T.I., Jenner, R.G., Boyer, L.A., Guenther, M.G., Levine, S.S., Kumar, R.M.,
Chevalier, B., Johnstone, S.E., Cole, M.F., Isono, K., et al. (2006). Control of
developmental regulators by Polycomb in human embryonic stem cells. Cell
125, 301–313.
Lee, H.G., Kahn, T.G., Simcox, A., Schwartz, Y.B., and Pirrotta, V. (2015).Genome-wide activities of Polycomb complexes control pervasive transcrip-
tion. Genome Res. 25, 1170–1181.
Leeb,M., Pasini, D., Novatchkova, M., Jaritz, M., Helin, K., andWutz, A. (2010).
Polycomb complexes act redundantly to repress genomic repeats and genes.
Genes Dev. 24, 265–276.
Cell Reports 17, 1369–1382, October 25, 2016 1381
Li, G., Margueron, R., Ku, M., Chambon, P., Bernstein, B.E., and Reinberg, D.
(2010). Jarid2 and PRC2, partners in regulating gene expression. Genes Dev.
24, 368–380.
Margueron, R., and Reinberg, D. (2011). The Polycomb complex PRC2 and its
mark in life. Nature 469, 343–349.
Marks, H., Kalkan, T., Menafra, R., Denissov, S., Jones, K., Hofemeister, H.,
Nichols, J., Kranz, A., Stewart, A.F., Smith, A., and Stunnenberg, H.G.
(2012). The transcriptional and epigenomic foundations of ground state plurip-
otency. Cell 149, 590–604.
Mastick, G.S., Davis, N.M., Andrew, G.L., and Easter, S.S., Jr. (1997). Pax-6
functions in boundary formation and axon guidance in the embryonic mouse
forebrain. Development 124, 1985–1997.
McCabe, M.T., Graves, A.P., Ganji, G., Diaz, E., Halsey, W.S., Jiang, Y., Smi-
theman, K.N., Ott, H.M., Pappalardi, M.B., Allen, K.E., et al. (2012). Mutation of
A677 in histone methyltransferase EZH2 in human B-cell lymphoma promotes
hypertrimethylation of histone H3 on lysine 27 (H3K27). Proc. Natl. Acad. Sci.
USA 109, 2989–2994.
Mitsui, K., Tokuzawa, Y., Itoh, H., Segawa, K., Murakami, M., Takahashi, K.,
Maruyama, M., Maeda, M., and Yamanaka, S. (2003). The homeoprotein
Nanog is required for maintenance of pluripotency in mouse epiblast and ES
cells. Cell 113, 631–642.
Montgomery, N.D., Yee, D., Chen, A., Kalantry, S., Chamberlain, S.J., Otte,
A.P., and Magnuson, T. (2005). The murine polycomb group protein Eed is
required for global histone H3 lysine-27 methylation. Curr. Biol. 15, 942–947.
Mousavi, K., Zare, H., Wang, A.H., and Sartorelli, V. (2012). Polycomb protein
Ezh1 promotes RNA polymerase II elongation. Mol. Cell 45, 255–262.
Mun˜oz-Sanjua´n, I., and Brivanlou, A.H. (2002). Neural induction, the default
model and embryonic stem cells. Nat. Rev. Neurosci. 3, 271–280.
Nichols, J., Zevnik, B., Anastassiadis, K., Niwa, H., Klewe-Nebenius, D.,
Chambers, I., Scho¨ler, H., and Smith, A. (1998). Formation of pluripotent
stem cells in the mammalian embryo depends on the POU transcription factor
Oct4. Cell 95, 379–391.
Onder, T.T., Kara, N., Cherry, A., Sinha, A.U., Zhu, N., Bernt, K.M., Cahan, P.,
Marcarci, B.O., Unternaehrer, J., Gupta, P.B., et al. (2012). Chromatin-modi-
fying enzymes as modulators of reprogramming. Nature 483, 598–602.
Orlando, D.A., Chen, M.W., Brown, V.E., Solanki, S., Choi, Y.J., Olson, E.R.,
Fritz, C.C., Bradner, J.E., and Guenther, M.G. (2014). Quantitative ChIP-Seq
normalization reveals global modulation of the epigenome. Cell Rep. 9,
1163–1170.
Pasini, D., Bracken, A.P., Hansen, J.B., Capillo, M., and Helin, K. (2007). The
polycomb group protein Suz12 is required for embryonic stem cell differentia-
tion. Mol. Cell. Biol. 27, 3769–3779.
Pasini, D., Cloos, P.A., Walfridsson, J., Olsson, L., Bukowski, J.P., Johansen,
J.V., Bak, M., Tommerup, N., Rappsilber, J., and Helin, K. (2010). JARID2 reg-
ulates binding of the Polycomb repressive complex 2 to target genes in ES
cells. Nature 464, 306–310.
Pauler, F.M., Sloane, M.A., Huang, R., Regha, K., Koerner, M.V., Tamir, I.,
Sommer, A., Aszodi, A., Jenuwein, T., and Barlow, D.P. (2009). H3K27me3
forms BLOCs over silent genes and intergenic regions and specifies a histone
banding pattern on a mouse autosomal chromosome. Genome Res. 19,
221–233.
Peng, J.C., Valouev, A., Swigut, T., Zhang, J., Zhao, Y., Sidow, A., and Wy-
socka, J. (2009). Jarid2/Jumonji coordinates control of PRC2 enzymatic activ-
ity and target gene occupancy in pluripotent cells. Cell 139, 1290–1302.
Pevny, L.H., Sockanathan, S., Placzek, M., and Lovell-Badge, R. (1998). A role
for SOX1 in neural determination. Development 125, 1967–1978.
Reversade, B., Kuroda, H., Lee, H., Mays, A., and De Robertis, E.M. (2005).
Depletion of Bmp2, Bmp4, Bmp7 and Spemann organizer signals inducesmassive brain formation in Xenopus embryos. Development 132, 3381–3392.
Riising, E.M., Comet, I., Leblanc, B., Wu, X., Johansen, J.V., and Helin, K.
(2014). Gene silencing triggers polycomb repressive complex 2 recruitment
to CpG islands genome wide. Mol. Cell 55, 347–360.
1382 Cell Reports 17, 1369–1382, October 25, 2016Ryall, J.G., Cliff, T., Dalton, S., and Sartorelli, V. (2015). Metabolic reprogram-
ming of stem cell epigenetics. Cell Stem Cell 17, 651–662.
Sanulli, S., Justin, N., Teissandier, A., Ancelin, K., Portoso, M., Caron, M.,
Michaud, A., Lombard, B., da Rocha, S.T., Offer, J., et al. (2015). Jarid2
methylation via the PRC2 complex regulates H3K27me3 deposition during
cell differentiation. Mol. Cell 57, 769–783.
Sarma, K., Margueron, R., Ivanov, A., Pirrotta, V., and Reinberg, D. (2008).
Ezh2 requires PHF1 to efficiently catalyze H3 lysine 27 trimethylation in vivo.
Mol. Cell. Biol. 28, 2718–2731.
Schwarz, M., Alvarez-Bolado, G., Dressler, G., Urba´nek, P., Busslinger, M.,
and Gruss, P. (1999). Pax2/5 and Pax6 subdivide the early neural tube into
three domains. Mech. Dev. 82, 29–39.
Shen, X., Liu, Y., Hsu, Y.J., Fujiwara, Y., Kim, J., Mao, X., Yuan, G.C., and
Orkin, S.H. (2008). EZH1 mediates methylation on histone H3 lysine 27 and
complements EZH2 in maintaining stem cell identity and executing pluripo-
tency. Mol. Cell 32, 491–502.
Shen, X., Kim, W., Fujiwara, Y., Simon, M.D., Liu, Y., Mysliwiec, M.R., Yuan,
G.C., Lee, Y., and Orkin, S.H. (2009). Jumonji modulates polycomb activity
and self-renewal versus differentiation of stem cells. Cell 139, 1303–1314.
Souroullas, G.P., Jeck, W.R., Parker, J.S., Simon, J.M., Liu, J.Y., Paulk, J.,
Xiong, J., Clark, K.S., Fedoriw, Y., Qi, J., et al. (2016). An oncogenic Ezh2 mu-
tation induces tumors through global redistribution of histone 3 lysine 27 trime-
thylation. Nat. Med. 22, 632–640.
Steiner, L.A., Schulz, V.P., Maksimova, Y., Wong, C., and Gallagher, P.G.
(2011). Patterns of histone H3 lysine 27 monomethylation and erythroid cell
type-specific gene expression. J. Biol. Chem. 286, 39457–39465.
Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L., Gil-
lette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S., and Me-
sirov, J.P. (2005). Gene set enrichment analysis: a knowledge-based approach
for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. USA
102, 15545–15550.
Tropepe, V., Hitoshi, S., Sirard, C., Mak, T.W., Rossant, J., and van der Kooy,
D. (2001). Direct neural fate specification from embryonic stem cells: a primi-
tive mammalian neural stem cell stage acquired through a default mechanism.
Neuron 30, 65–78.
Voigt, P., LeRoy, G., Drury, W.J., 3rd, Zee, B.M., Son, J., Beck, D.B., Young,
N.L., Garcia, B.A., and Reinberg, D. (2012). Asymmetrically modified nucleo-
somes. Cell 151, 181–193.
Wood, H.B., and Episkopou, V. (1999). Comparative expression of the mouse
Sox1, Sox2 and Sox3 genes from pre-gastrulation to early somite stages.
Mech. Dev. 86, 197–201.
Wray, J., Kalkan, T., Gomez-Lopez, S., Eckardt, D., Cook, A., Kemler, R., and
Smith, A. (2011). Inhibition of glycogen synthase kinase-3 alleviates Tcf3
repression of the pluripotency network and increases embryonic stem cell
resistance to differentiation. Nat. Cell Biol. 13, 838–845.
Wu, J., Okamura, D., Li, M., Suzuki, K., Luo, C., Ma, L., He, Y., Li, Z., Benner,
C., Tamura, I., et al. (2015). An alternative pluripotent state confers interspecies
chimaeric competency. Nature 521, 316–321.
Yap, D.B., Chu, J., Berg, T., Schapira, M., Cheng, S.W., Moradian, A., Morin,
R.D., Mungall, A.J., Meissner, B., Boyle, M., et al. (2011). Somatic mutations at
EZH2 Y641 act dominantly through a mechanism of selectively altered PRC2
catalytic activity, to increase H3K27 trimethylation. Blood 117, 2451–2459.
Ying, Q.L., Stavridis, M., Griffiths, D., Li, M., and Smith, A. (2003). Conversion
of embryonic stem cells into neuroectodermal precursors in adherent mono-
culture. Nat. Biotechnol. 21, 183–186.
Zhang, J., Nuebel, E., Daley, G.Q., Koehler, C.M., and Teitell, M.A. (2012).
Metabolic regulation in pluripotent stem cells during reprogramming and
self-renewal. Cell Stem Cell 11, 589–595.Zhu, J., Adli, M., Zou, J.Y., Verstappen, G., Coyne, M., Zhang, X., Durham, T.,
Miri, M., Deshpande, V., De Jager, P.L., et al. (2013). Genome-wide chromatin
state transitions associated with developmental and environmental cues. Cell
152, 642–654.