1. No category

Resource
Sox17-Mediated XEN Cell Conversion Identifies Dynamic
Networks Controlling Cell-Fate Decisions in EmbryoDerived Stem Cells
Graphical Abstract
Authors
Angela C.H. McDonald, Steffen Biechele,
Janet Rossant, William L. Stanford
Correspondence
[email protected] (J.R.),
[email protected] (W.L.S.)
In Brief
McDonald et al. present a dynamic model
of extraembryonic endoderm specification in ESCs by overexpressing Sox17.
Using time-series RNA-seq and promoter
occupancy data, the authors identify
endoderm-fate
regulators
including
Nr5a2, which represses endoderm fate
in ESCs. This model provides insight
into how lineage-determining transcription factors drive cell fate.
Highlights
Accession Numbers
Sox17 converts ESCs into XEN cells that are functional in vivo
GSE61102
Sox17 regulates dynamic GRNs in order to specify cell fate
Sox17 participates in autoregulatory and feedforward network
motifs
Nr5a2 acts in a GRN motif repressing ExEn gene expression in
ESCs
McDonald et al., 2014, Cell Reports 9, 1–14
October 23, 2014 ª2014 The Authors
http://dx.doi.org/10.1016/j.celrep.2014.09.026
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
Cell Reports
Resource
Sox17-Mediated XEN Cell Conversion Identifies
Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells
Angela C.H. McDonald,1,2 Steffen Biechele,1,6 Janet Rossant,1,5,* and William L. Stanford2,3,4,5,*
1Program in Developmental and Stem Cell Biology, Hospital for Sick Children Research Institute, 686 Bay Street, Toronto, ON M5G 0A4,
Canada
2Institute of Biomaterials and Biomedical Engineering, University of Toronto, 164 College Street, Toronto, ON M5S 3G9, Canada
3Sprott Centre for Stem Cell Research, Ottawa Hospital Research Institute, 501 Smyth Road, Ottawa, ON K1L 8L6, Canada
4Cellular and Molecular Medicine, Faulty of Medicine, University of Ottawa, 75 Laurier Avenue East, Ottawa, ON K1N 6N5, Canada
5Co-senior author
6Present address: Eli and Edythe Broad Center of Regeneration Medicine and Stem Cell Research, University of California at San Francisco
School of Medicine, 35 Medical Center Way, San Francisco, CA 94143, USA
*Correspondence: [email protected] (J.R.), [email protected] (W.L.S.)
http://dx.doi.org/10.1016/j.celrep.2014.09.026
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
SUMMARY
Little is known about the gene regulatory networks
(GRNs) distinguishing extraembryonic endoderm
(ExEn) stem (XEN) cells from those that maintain
the extensively characterized embryonic stem cell
(ESC). An intriguing network candidate is Sox17, an
essential transcription factor for XEN derivation and
self-renewal. Here, we show that forced Sox17
expression drives ESCs toward ExEn, generating
XEN cells that contribute to ExEn when placed
back into early mouse embryos. Transient Sox17
expression is sufficient to drive this fate change
during which time cells transit through distinct intermediate states prior to the generation of functional
XEN-like cells. To orchestrate this conversion process, Sox17 acts in autoregulatory and feedforward
network motifs, regulating dynamic GRNs directing
cell fate. Sox17-mediated XEN conversion helps
to explain the regulation of cell-fate changes and
reveals GRNs regulating lineage decisions in the
mouse embryo.
INTRODUCTION
The mammalian preimplantation embryo consists of three
distinct lineages, the epiblast (EPI), which forms the embryo
proper, and two extraembryonic tissues, the trophectoderm
and the primitive endoderm (PE). The trophectoderm is the first
lineage specified by embryonic day (E) 3.5 when it segregates
from the inner cell mass (ICM). As development proceeds, the
trophectoderm contributes to the trophoblast layers of the
placenta. ICM cells initially coexpress markers characteristic of
both PE and EPI but show mutually exclusive PE- or EPI-specific
expression in a ‘‘salt-and-pepper’’ distribution by E3.5. Spatial
segregation of the PE and preimplantation EPI lineages occurs
by E4.5 (Chazaud et al., 2006; Gardner and Rossant, 1979; Plusa
et al., 2008). Although the preimplantation EPI goes on to form all
fetal tissues after implantation, the PE principally forms ExEn
tissues of the yolk sac that surround and pattern the embryo,
collectively described as the ExEn.
Specification of the trophectoderm, PE, and preimplantation
EPI represents the first cell-fate changes that occur during
embryonic development. Lineage-specific transcription factors
driving these cell-fate changes have been identified by gene
expression and functional analysis in embryos, but elucidating
the GRNs in which these transcription factors act is challenging
due to limited cell numbers in the early embryo. The derivation of
three stem cell lines from the preimplantation embryo, trophoblast stem cells (Tanaka et al., 1998), XEN cells (Kunath et al.,
2005), and ESCs (Evans and Kaufman, 1981; Martin, 1981) isolated from the trophectoderm, PE, and preimplantation EPI,
respectively, provides powerful tools to study the GRNs directing early embryonic cell-fate decisions.
The vast majority of work characterizing GRNs maintaining
these stem cell lines has focused on the regulation of ESC selfrenewal and differentiation. ESCs rely on a core GRN regulating
self-renewal and differentiation, made up of Oct4 (Morrison and
Brickman, 2006; Nichols et al., 1998), Nanog (Chambers et al.,
2003; Mitsui et al., 2003), and Sox2 (Masui et al., 2007). This
core GRN is regulated in a feedforward manner by an extended
pluripotency network (Walker et al., 2010). In addition, some
ESC self-renewal factors regulate differentiation toward specific
lineages (Loh and Lim, 2011), including ExEn. Prdm14 protects
ESCs from acquiring ExEn fate by directly repressing ExEn genes
and simultaneously activating ESC genes (Ma et al., 2011). Similarly, Tbx3 maintains ESC self-renewal, whereas loss of Tbx3 in
ESCs induces ExEn differentiation (Ivanova et al., 2006; Lu
et al., 2011; Niwa et al., 2009). Sall4, on the other hand, has
dual positive roles in ESC and XEN cell fate. In ESCs, Sall4 acts
in an interconnected GRN with Oct4, Sox2, and Nanog (SakakiYumoto et al., 2006; Wu et al., 2006; Zhang et al., 2006). In XEN
Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors 1
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
cells, Sall4 directly regulates XEN genes and loss of Sall4 in XEN
cells is incompatible with XEN self-renewal (Lim et al., 2008).
Although some overlap in the nodes of ESC and XEN cell
GRNs have been identified, GRNs distinguishing XEN cells
from those characterized in ESCs are not well understood.
When overexpressed in ESCs, the PE-specific genes Gata6
and Gata4 drive XEN cell commitment, and the loss of these
genes in XEN cells is incompatible with self-renewal (Fujikura
et al., 2002; Lim et al., 2008; Shimosato et al., 2007). SOX transcription factors have also been implicated in maintaining XEN
cell identity. Loss of either Sox7 or Sox17 disrupts XEN cell
self-renewal (Lim et al., 2008). Although Sox17 / mutants
generate PE, XEN cell lines cannot be derived from these embryos (Artus et al., 2011; Kanai-Azuma et al., 2002; Niakan
et al., 2010). Overexpression studies using SOX transcription
factors have produced disparate results. SOX7 overexpression
in human ESCs activates ExEn genes (Se´guin et al., 2008),
whereas Sox7 overexpression in mouse ESCs has little effect
on ExEn gene expression (Futaki et al., 2004). Sox17 overexpression has been reported to induce both ExEn and definitive endoderm genes during mouse embryoid body differentiation (Qu
et al., 2008; Shimoda et al., 2007). However, Sox17 overexpression in mouse ESCs failed to prevent differentiation of all three
germ layers in teratoma assays (Niakan et al., 2010). Intriguingly,
Sox17+ cells isolated from mouse ESC cultures are committed
to XEN cell fate, contributing to the ExEn region of chimeric
mouse embryos (Niakan et al., 2010). In contrast, we have previously shown that SOX17 overexpression in human ESCs induces
mesendoderm fate (Se´guin et al., 2008).
In this study, we aimed to resolve the role of Sox17 in the XEN
cell GRN. By overexpressing Sox17 in mouse ESCs, we have established stable, transgene-independent XEN-like (Sox17-XEN)
cells that contribute to ExEn in vivo, highlighting the ability of
Sox17 to drive cell-fate changes. We further show that Sox17
is a potent inducer of XEN cell fate, only requiring transient overexpression to initiate the XEN GRN. Using RNA sequencing
(RNA-seq) time course analyses in combination with genomewide transcription factor binding data, we developed a dynamic
regulatory model of the GRNs involved in XEN cell specification
and identified putative regulators of XEN cell identity. As predicted by our model, we confirmed that Nr5a2 negatively regulates XEN cell genes, thereby repressing ExEn differentiation in
ESCs. Together, our findings confirm Sox17 as a key node in
the ExEn GRN and that Sox17-mediated XEN conversion is a
useful model to elucidate GRNs directing early mammalian
cell-fate decisions.
Lama1, Sox7, Gata4, and Gata6) were induced within 24 hr
and Collagen IV, Laminin, and Gata4 proteins were expressed
within 48 hr (Figures S2A–S2C). Extended culture of two
Sox17-ESC clones in dox under ESC self-renewal conditions
(serum and LIF) promoted morphological changes resulting in
dispersed cells, demonstrating flat morphology around the
periphery of ESC colonies by day 6 (Figure 1A). By day 12, a
homogenous population of small, round, refractile cells, indistinguishable from embryo-derived XEN cells, was established
(Figure 1A).
Following Sox17 activation, expression of ExEn genes
continued to increase, stabilizing at levels equivalent to XEN
cells by day 12 (Figure 1B). Concomitantly, the expression of
ESC genes Oct4 and Nanog was reduced (Figure 1C). Definitive
endoderm (Gsc and Cxcr4), mesoderm (T and Tbx6), and ectoderm (Sox1 and Pax6) genes were not significantly induced, suggesting that Sox17 drives differentiation toward ExEn (Figures
1D–1F).
To determine whether the XEN cell-like phenotype acquired by
ESCs in response to Sox17 overexpression is stable in the
absence of transgene expression, we cultured Sox17-XEN cells
without dox for at least 30 days. Sox17-XEN cells maintained
expression of XEN proteins Collagen IV, Laminin, Gata4,
Gata6, and Sall4 (Figures 1G and 1H), whereas ESC transcription
factors Oct4 and Nanog were not detected (Figure 1H). Sox17XEN cells maintained a XEN cell-surface profile (Rugg-Gunn
et al., 2012), expressing CD140a, Robo2, and CD26 while lacking expression of ESC markers CD81, CD31 (Figures 1I and 1J),
and definitive endoderm maker CD184 (Figure 1J). Sox17-XEN
cells maintained a XEN-like phenotype for greater than 90 days
in the absence of dox.
When placed back into the embryo, XEN cells contribute to the
parietal endoderm layer of mouse postimplantation embryos
(Kunath et al., 2005). Based on the similarity of Sox17-XEN cells
to embryo-derived XEN cells, we hypothesized that Sox17-XEN
cells would function as ExEn in vivo. To test this, Sox17-XEN
cells ubiquitously expressing mCherry were injected into wildtype host blastocysts and analyzed for embryonic contribution
at E6.5, E7.5, and E8.5. All chimeric embryos observed contained Sox17-XEN cells exclusively in the parietal endoderm
(Figures 1K and 1L).
Taken together, these findings support a model in which
Sox17 is a key ExEn lineage-determining transcription factor.
Upon overexpression in ESCs, Sox17 directs XEN gene activation and pluripotency repression, thereby driving bona fide
XEN cell commitment, as determined by chimera formation
assays.
RESULTS
Establishment of Functional XEN-like Cells from ESCs
by Sox17 Overexpression
To test the ability of Sox17 to drive XEN cell fate in mouse ESCs,
we used doxycycline (dox)-inducible Sox17-ESCs (Rugg-Gunn
et al., 2012) to activate the expression of Sox17 and eGFP.
Following induction, Sox17 was significantly activated in
Sox17-ESCs while remaining in physiological range, at levels
less than 20% of Sox17 expression in embryo-derived XEN cells
(Figure S1). Upon dox treatment, ExEn genes (Col4a1, Col4a2,
2 Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors
Transient Expression of Sox17 Is Sufficient to Drive XEN
Cell Fate in ESCs
The ability of Sox17-XEN cells to maintain a stable phenotype independent of transgene expression suggests that a transition
from exogenous transgene control to endogenous control of
cell fate occurs during XEN conversion. To determine transgene
kinetics during this process, we evaluated GFP expression using
flow cytometry and found that transgene silencing began before
day 6 of dox culture (Figure 2A; Figure S3A). Next, day 6 cells
were sorted (Figure 2B; Figure S3B), and gene expression of
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
Figure 1. Sox17 Overexpression in ESCs Induces Functional XEN-like Cells
(A) Photomicrographs of Sox17-ESCs (clone 4J) cultured in dox. Scale bar, 200 mm.
(B–F) Gene expression during Sox17-ESC dox culture. Two independent clones (3F [pink] and 4J [blue]) are shown at indicated time points relative to uninduced
Sox17-ESCs alongside embryo-derived XEN cells (red triangle) or embryoid bodies (EBs) (orange circles). Error bars = SEM of three independent experiments.
ND, not detected.
(G and H) Immunostaining of Sox17-ESCs and transgene-independent Sox17-XEN cells (clone 4J). Scale bar, 45 mm.
(I and J) Flow cytometric analysis of Sox17-XEN cells (clone 4J) compared to Sox17-ESCs and EB-derived definitive endoderm cells.
(K) E8.5 Sox17-XEN chimeric embryo. Scale bar, 225 mm.
(L) Rate of chimerism detected in recovered embryos following blastocyst injection.
See also Figures S1 and S2.
Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors 3
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
(legend on next page)
4 Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
GFP+ (4J = 76.3% and 3F = 82.8%) and GFP (4J = 23.7% and
3F = 17.2%) cells was evaluated. Oct4 and Nanog expression
was comparable between GFP+ and GFP day 6 cells (Figure 2C), but GFP+ cells expressed significantly higher levels
of XEN genes (Gata4, Gata6, Sox7, Col4a1, Col4a2, and
Lama1) (Figure 2D). When GFP+ and GFP day 6 cells were
cultured for an additional 12 days in dox, both cell populations
acquired comparable XEN-like cell-surface profiles (Figure 2E;
Figure S3C).
These results suggested that transient Sox17 overexpression
might be sufficient to induce XEN cell fate in ESCs. To test this,
we removed dox from cultures after 12, 24, 36, 48, and 72 hr of
treatment and subsequently tracked these cells by flow cytometry at days 6, 12, and 20 (Figure 2F; Figure S3D). We found that a
12 hr pulse of Sox17 overexpression induced CD26 expression
in a subset of cells by day 12, whereas 48 hr of transgene overexpression induced XEN marker expression changes with equivalent kinetics to cells continuously cultured in dox (Figure 2F;
Figure S3D). Sox17-XEN cells derived following 48 hr of transgene overexpression (Sox17(48)-XEN cells) combined with an
additional 30 days of culture maintain a cell-surface profile
indistinguishable from previously established Sox17-XEN cells
(Figures 2G and 2H; Figures S3E and S3F).
To determine whether Sox17(48)-XEN cells were functionally
equivalent to Sox17-XEN cells, we performed chimera assays
and found that Sox17(48)-XEN cells also contributed exclusively
to the parietal endoderm of chimeric embryos (Figures 2I and 2J).
Taken together, these findings show that transient Sox17 overexpression is sufficient to engage the XEN GRN in ESCs, thereby
restricting cell fate to ExEn.
Sox17-Mediated XEN Conversion Is Highly Efficient
To determine the efficiency of Sox17-XEN establishment, we
followed conversion of individual ESCs. As others have in the
field, we defined conversion efficiency as the potential of a
single cell to give rise to reprogrammed daughter cells (Hanna
et al., 2009). CD31+CD140a cells sorted from two independent
Sox17-ESC clones or wild-type ESCs were individually deposited into 96-well plates and subsequently cultured in dox (Figure 3A). XEN cell morphology could be detected by day 14 in
greater than 75% of clones derived from either starting Sox17ESC line (n = 48 clones [3F] and n = 53 clones [4J]) (Figure 3B).
By day 70, greater than 90% of Sox17-ESC clones gave rise to
daughter cells with XEN-like morphology (Figure 3B). No XENlike cells were ever detected in wild-type ESC clones (n = 64
clones) (Figure 3B).
To quantify XEN conversion, expression of CD140a and CD31
were measured at day 70. A detection value of >2% CD140a+
was used as the minimal threshold for XEN cell conversion in
clonal populations, whereas >2% CD31+ cells was used as the
minimal threshold for defining the presence of ESCs. One hundred percent of Sox17-ESC clones gave rise to CD140a+
daughter cells (Figures 3C and 3D), whereas less than 18% of
Sox17-ESC clones contained a subset of CD31+ cells (Figures
3E and 3F). These results demonstrate that most if not all
ESCs have the ability to acquire XEN fate following Sox17 overexpression, albeit at different latencies.
Cells Transit through Distinct Compartments during
Sox17-Mediated XEN Conversion
We previously observed a subset of Sox17-ESCs negative for
both XEN and ESC surface markers when cultured in dox for
6 days (Figure 2F), suggesting that cells may transit through a
double-negative compartment during XEN conversion. To test
this, we sorted cells at day 6 using CD140a and CD31 (Figure 4A;
Figure S4A) and analyzed gene expression of each compartment.
Cells located in the XEN compartment (CD140a+CD31 ) had
significantly higher expression of XEN genes (Col4a1, Col4a2,
Lama1, Gata6, Gata4, and Sox7) compared to CD140a CD31
cells and CD140a CD31+ cells (Figure 4B; Figure S4B). Nanog
expression was significantly reduced in CD140a+CD31 cells
compared to CD140a CD31+ cells; however, no significant differences were observed for Oct4 and Sall4 expression between
compartments (Figure 4B). After continued culture, a subset of
cells from day 6 CD140a CD31 and CD140a CD31+ compartments activated CD140a expression (Figure 4C; Figure S4C),
demonstrating that cells undergoing XEN conversion transit
through a distinct compartment, characterized by the lack of
both ESC marker CD31 and XEN cell marker CD140a.
Sox17-D6 Cells Are Distinct from Sox17-XEN Cells
To determine whether XEN-like cells present by day 6 of Sox17
overexpression (Sox17-D6) (Figure 4A) were equivalent to
stable XEN-like cells, we performed RNA-seq to compare the
expression profiles of Sox17-D6 and Sox17-XEN cells (day 60).
Hierarchical clustering demonstrated that Sox17-D6 cells are
distinct from Sox17-XEN and embryo-derived XEN cells, suggesting that these cells may represent an intermediate state
of XEN conversion (Figure 4D). Although similar to embryoderived XEN cells, stable Sox17-XEN gene expression is
distinct, despite their ability to contribute to the parietal endoderm in vivo.
Figure 2. Transient Sox17 Expression Is Sufficient to Drive XEN Conversion in ESCs
(A) Flow cytometric analysis of GFP expression in Sox17-ESCs (clone 4J) cultured in dox compared to uninduced Sox17-ESCs.
(B) FACS plot of Sox17-ESCs (clone 4J) on day 6 of dox culture.
(C and D) Expression of pluripotency genes (C) and XEN genes (D) in sorted cells (B) relative to uninduced Sox17-ESCs and compared to embryo-derived XEN
cells. Error bars = SEM of six experiments (two clones 3 three independent experiments).
(E) Flow cytometric analysis of sorted Sox17-ESCs (B) at day 18 of dox culture.
(F) Sox17-ESCs (clone 4J) were treated with dox transiently for 12, 24, 36, 48, or 72 hr and tracked at indicated times using flow cytometry.
(G and H) Flow cytometric analysis of Sox17(48)-XEN cells (clone 4J) compared to Sox17-ESCs and EB-derived definitive endoderm (DE).
(I) E8.5 Sox17(48)-XEN chimeric embryo. Scale bar, 225 mm.
(J) Rate of chimerism detected in recovered embryos following blastocyst injection.
See also Figure S3.
Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors 5
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
Figure 3. Sox17-Mediated XEN Conversion
Is Highly Efficient
(A) Single-cell Sox17-mediated XEN conversion
experimental outline.
(B) Cumulative percentage of Sox17-ESC and wildtype ESC clones exhibiting XEN cell morphology at
indicated time points.
(C) Flow cytometric measurement of CD140a
expression in Sox17-ESC and wild-type ESC clones.
(D) Binned data from (C).
(E) Flow cytometric measurement of CD31 expression in Sox17-ESC and wild-type ESC clones.
(F) Binned data from (E).
To identify gene expression patterns across cell states, we
performed K-means clustering, identifying four distinct clusters
(Figure 4E; Table S1). We compiled a list of statistically enriched
pathways and gene ontology (GO) terms for each cluster (Table
S2) and created enrichment maps (Merico et al., 2010) (Figure 4F). Fast-response downregulated genes were associated
with neural processes, trophoblast development, and FGF
signaling. Slow-response downregulated genes were associated with mesoderm and neural development, cell adhesion,
and Wnt signaling. Upregulated genes were associated with
cell motility, endoderm development, and BMP signaling, a
pathway involved in ExEn development (Coucouvanis and Martin, 1999). These results suggest that ectopic Sox17 expression
drives waves of gene activation and repression regulating the
acquisition of ExEn fate.
Interestingly, a cluster of genes uniquely expressed in Sox17D6 cells (Figure 4E) associated with intestinal functions such as
regulation of intestinal cholesterol absorption was also identified
(Figure 4F; Table S2). Given that a subset of the ExEn, the
visceral endoderm (VE), performs intestinal-like functions prior
to placental development (Jollie, 1990), we postulated that
Sox17-D6 cells may be more similar to VE than parietal endoderm, the ExEn tissue that XEN cells resemble most (Artus
et al., 2012; Kunath et al., 2005; Paca et al., 2012). To this end,
we asked whether VE genes were uniquely enriched in Sox17D6 cells using Gene Set Enrichment Analysis (GSEA) (Subramanian et al., 2005) and a VE gene set we constructed using
published microarray data (Figure S5; Table S3) (Paca et al.,
2012). The VE gene set was significantly enriched in Sox17-D6
cells, suggesting that Sox17-D6 cells are more similar to VE
than Sox17-XEN cells (Figure 4G). Moreover, Sox17-D6 cells express VE genes Amn, Cubn, and Lrp2 and express high levels of
E-cadherin while expressing low levels of parietal endoderm
genes Krt23, Nid1, and Krt18 (Figures 4H and 4I). These analyses
demonstrate that Sox17 directs ESCs to undergo a stepwise
XEN fate conversion during which cells transit through intermediate states, including a transient VE-like state.
6 Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors
Generation of a Dynamic Regulatory
Map of Sox17-Mediated XEN
Conversion
To further understand Sox17-activated
gene expression changes during XEN
conversion, we performed an RNA-seq
time series. Given that cell-surface protein
expression does not change within the first 48 hr of dox treatment (Figures S6A and S6B), we performed RNA-seq on unsorted cells after 0, 7, 21, 35, and 49 hr of dox culture. These
data were inputted into the Dynamic Regulatory Events Miner
(DREM) algorithm (Ernst et al., 2007), plotting dynamic gene
expression throughout XEN conversion (Figure 5A). Each path
(line) represents a group of genes sharing similar expression,
and points at which expression diverges among a group of
genes, known as bifurcation points, are indicated in green.
Paths containing known ESC and XEN genes are outlined in
Figure 5B.
The identification of gene activation and repression waves
using this algorithm further supports a model of a Sox17-activated multiphasic XEN conversion process. Next, we performed pathway and GO term enrichment analyses on gene
groups contained in particular expression paths outlined by
our dynamic regulatory map. The earliest activated genes
(path 4) were induced after 7 hr and are associated with developmental processes including endoderm development and
retinoic acid signaling (Figures 5A, 5C, and 5D), a signaling
molecule used to induce XEN fate in ESCs (Cho et al.,
2012). A subset of path 4 endoderm, BMP, and retinoic acid
signaling pathway genes maintained high expression
throughout XEN establishment (path 13) (Figures 5A, 5C, and
5D). By 21 hr, repression of genes associated with ectoderm
and mesoderm development began (path 21) (Figures 5A,
5C, and 5D).
After 49 hr, secondary waves of gene regulation were initiated,
including activation of genes associated with cell movement
(path 24). Genes associated with cellular processes such as
ion transport as well as additional developmental processes
including lung development, which utilizes XEN genes including
Tbx3 and Gata6, were also upregulated during this time (path 34)
(Figures 5A, 5C, and 5D). Simultaneously, secondary waves of
gene repression began after 49 hr, suppressing genes associated with mesoderm, ectoderm, and cell adhesion (path 33)
(Figures 5A, 5C, and 5D).
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
Together, these data support a multiphasic model of ExEn fate
acquisition during which ExEn gene activation is followed by
subsequent repression of ectodermal and mesodermal gene
repression.
Prediction and Testing of GRN Motifs Regulating XEN
Cell Fate
Next, we combined genome-wide transcription factor binding
data for known ESC and XEN cell GRN members (Table S4)
with transcription factor binding data included in DREM and
used this algorithm to annotate DREM gene expression paths
(outlined in Figure 5A) with transcription factors predicted by
DREM to regulate the expression of genes contained in a particular path. Interestingly, Sall4 was predicted to regulate several
bifurcation points in combination with Sox17 (Figure 6A). Both
transcription factors are predicted to regulate early XEN genes
expressed highly by 21 hr (path 13), as well as XEN genes activated by 49 hr after Sox17 overexpression (path 29). Although
Gata4 and Gata6 are established XEN cell-fate regulators (Fujikura et al., 2002; Lim et al., 2008; Shimosato et al., 2007), their
transcriptional targets have not been identified in XEN cells.
Our analysis predicts that Gata4 and Gata6 act as both activators (paths 10, 24, and 34) and repressors (path 16) during XEN
establishment (Figure 6A).
In addition to predicting roles for XEN GRN members, our
analysis also predicted regulators not previously associated
with XEN cell fate including Foxa2, Foxo1, Pou2f1, Hnf4a,
Foxd3, and Nr5a2 (Figure 6A). To determine whether putative
XEN transcription factors are actually active in XEN cells, we reanalyzed our time-series expression data and determined those
transcription factors expressed uniquely in either XEN cells or
ESCs or expressed in both cell types (Figure 6B). We then categorized putative XEN regulators into three network motifs: (1)
XEN repressors expressed in ESCs, (2) XEN activators expressed in XEN cells, and (3) transcription factors expressed in
both ESCs and XEN cells, in which these factors play unique
roles in each cell type (Figure 6C).
An intriguing XEN repressor candidate is Nr5a2. Nr5a2 has
been used to replace Oct4 during induced pluripotent stem
cell (iPSC) reprogramming, and chromatin immunoprecipitation
sequencing (ChIP-seq) results demonstrate that Nr5a2 is bound
to regions regulating numerous XEN genes including Hnf4a,
Col4a1, and Lama5 (Heng et al., 2010). We drafted a GRN motif
in which Nr5a2 directly represses XEN genes in ESCs (Figure 7A)
and hypothesized that the loss of Nr5a2 in ESCs should activate
XEN genes in a similar manner to the known XEN repressor
Prdm14 (Ma et al., 2011). To test our hypothesis, we knocked
down Nr5a2 and Prdm14 using small interfering RNA (siRNA)
(Figures 7B and 7C) and found that even incomplete knockdown of Nr5a2 triggered activation of XEN genes after 48 hr,
comparable to the effect of Prdm14 (Figure 7D). We integrated
this Nr5a2 repression motif into the existing core ESC GRN
(Walker and Stanford, 2009) along with Klf4, Tbx3, and
Prdm14 DNA interactions (Table S4), demonstrating that
Nr5a2 regulates both self-renewal and differentiation in ESCs
(Figure 7E).
To refine our model of XEN conversion, we integrated gene
expression and transcription factor-DNA interaction data, draft-
ing two distinct GRNs regulating the establishment and maintenance of XEN fate. We predicted that Sox17 would directly
activate XEN transcription factors within 48 hr of overexpression
while ESC GRN members are still expressed, resulting in an intermediate state in which members of both ESC and XEN
GRNs are active (Figure 7F). Upon Sox17-XEN establishment,
ESC regulators Nanog, Oct4, Sox2, Nr5a2, Foxd3, and Prdm14
are no longer expressed (Figure 1C; Figure S7), suggesting
that the ESC GRN is inactive in XEN cells. Based on our model,
we predicted that Sox17 maintains its regulation of XEN transcription factors as well its own promoter in a stable XEN state
(Figure 7G).
To test our predicted Sox17 GRN motifs, we performed ChIPqPCR to assess Sox17 binding of XEN genes (Niakan et al.,
2010) and found that Sox17 directly activates XEN GRN members Sox17, Gata4, Gata6, and Sall4 within 48 hr of Sox17
overexpression and that these interactions are maintained in
Sox17-XEN cells (Figure 7H), confirming that Sox17 acts in
both autoregulatory and feedforward motifs (Figures 7I and 7J)
across cell states suggesting that Sox17 activates dynamic
GRNs to orchestrate XEN cell fate in ESCs.
DISCUSSION
Here, we have shown that Sox17 overexpression in mouse ESCs
efficiently drives cell fate to ExEn, highlighting the central role of
Sox17 in the XEN GRN. Sox17-XEN cells maintain a XEN-like
molecular profile and mimic the ability of XEN cells to contribute
to ExEn in vivo. Although Sox17-induced ExEn gene activation in
mouse ESCs has been previously reported (Niakan et al., 2010),
the affinity tagged Sox17 expression construct used in that study
was heterogeneously activated and did not bind to its own promoter or those regulating XEN GRN members Gata4 and Gata6
(Niakan et al., 2010). Hence, this suggests that, if Sox17 is not
sufficiently expressed or is functionally impaired, Sox17 cannot
adequately activate the XEN GRN and drive ESC conversion to
XEN cells.
Our mouse results also contrast with our prior study performed
using human ESCs, in which SOX17 overexpression directed
mesendoderm specification (Se´guin et al., 2008). This discrepancy is likely due to differences in developmental state between
human and mouse ESCs, as human ESCs have been shown to
be more similar to murine postimplantation epiblast stem cells
(Brons et al., 2007; Tesar et al., 2007). Sox17 also plays a critical
role in early blood development (He et al., 2011; Kim et al., 2007),
suggesting that Sox17 may act within unique GRN modules
specifying distinct tissues, depending on developmental
context.
Although embryo-derived XEN cells provide a model of ExEn
maintenance and differentiation, Sox17-mediated XEN conversion provides a unique tool to explore ExEn specification. Using
this model, we found that ESCs transit through distinct states
characterized by rapid activation of ExEn genes and suppression
of ectodermal genes followed by repression of mesodermal
and ESC genes, which results in an intermediate VE-like state.
Upon continued culture, Sox17-XEN cells acquire a parietal
endoderm-like phenotype similar to embryo-derived XEN cells
(Kunath et al., 2005). Interestingly, BMP signaling directs XEN
Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors 7
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
(legend on next page)
8 Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
differentiation to VE-like cells (Artus et al., 2012; Paca et al.,
2012), suggesting that BMP signaling-associated genes active
during XEN conversion may drive an intermediate VE-like state.
We combined our dynamic model of ExEn specification with
transcription factor-DNA interaction data, which allowed us to
predict regulators of XEN specification that can be classified
into three categories: XEN repressors active in ESCs, XEN activators in XEN cells, and transcription factors expressed in both
ESCs and XEN cells, playing a unique role in each cell type. Using this model, we predicted and tested a role for the ESC selfrenewal factor Nr5a2 (Gu et al., 2005; Heng et al., 2010) in the
regulation of XEN fate, demonstrating the ability of Nr5a2 to
repress XEN genes in ESCs. Embryo studies support a role for
Nr5a2 in endoderm regulation as Nr5a2 regulates the expression
of ExEn-associated genes in definitive endoderm lineages (Pare
et al., 2001; Rausa et al., 1999), and its expression is required in
the ExEn for proper patterning of the embryo (Labelle-Dumais
et al., 2006).
The current model suggests that PE is specified when reciprocal inhibition of Nanog and Gata6 is disrupted in favor of
Gata6 (Lanner, 2014). In vitro, overexpression of Gata6 or
Gata4 drives XEN fate (Fujikura et al., 2002); however, only
Gata6 is required for PE specification in vivo (Schrode et al.,
2014). Here, we show that Sox17 directly activates Gata6 and
Gata4 when overexpressed in ESCs, allowing Sox17 to drive
XEN conversion. Others have shown that Sox17 is not required
for Gata6 expression during ExEn differentiation but is required
for normal Gata4 expression (Niakan et al., 2010). These data
suggest that Sox17 GRN motifs reinforce PE fate, placing
Sox17 downstream of Gata6 in the ExEn GRN. However, lossof-function studies suggest that Sox17 is a critical member of
the XEN GRN because XEN cells cannot be derived from
Sox17 / embryos (Niakan et al., 2010) and knockdown of
Sox17 disrupts XEN self-renewal (Lim et al., 2008).
Transient Sox17 overexpression drives highly efficient XEN
conversion in ESCs, suggesting that the GRNs governing XEN
and ESC fates are closely connected, allowing for quick activation or repression of the opposing lineage. In our model, Sox17
overexpression activates XEN genes within 7 hr, followed by
repression of ESC genes beginning within 21 hr. Sox17 directs
these gene expression changes using at least two distinct mechanisms. Within 48 hr, Sox17 activates a positive autoregulatory
loop, which allows Sox17 to increase the rate of its own transcription upon reaching an activation threshold. Such positive
autoregulatory loops generate cell-cell variability (Walker and
Stanford, 2009), which could explain the variable latencies
observed across clones during Sox17-mediated XEN conversion. Second, Sox17 activates positive feedforward loops in
which it directly activates downstream XEN genes while simultaneously activating XEN GRN members. We analyzed Sox17
binding in XEN cells (Niakan et al., 2010) and found that Sox17
directly represses Nr5a2, suggesting a third mechanism in which
Sox17 may prevent the action of XEN repressors. Others have
suggested that Sox17 may compete with the binding of pluripotency transcription factors Nanog and Sox2 at XEN promoters to
regulate ExEn fate (Niakan et al., 2010).
In conclusion, Sox17-mediated XEN conversion provides
insight into how lineage-determining transcription factors drive
cell-fate decisions in embryo-derived stem cells. By integrating
gene regulatory data, we have illustrated a model in which
each stem cell state possesses a GRN that maintains selfrenewal while repressing the self-renewal network of the
opposing stem cell state. When a member of one of these
GRNs is overexpressed, as demonstrated here by Sox17, dynamic GRNs orchestrate cell-fate change. Understanding how
GRNs drive cell-fate decisions is essential for the effective generation of specific cell types for downstream in vitro studies, the
derivation of more mature cell types for regenerative medicine
purposes, or developing novel therapeutics for degenerative
diseases and malignancies.
EXPERIMENTAL PROCEDURES
Cell Culture and Transfection
R1 ESCs (Nagy et al., 1993) were cultured at 37 C and 5% CO2 on mitotically
inactivated mouse embryonic fibroblasts (MEFs) in ESC media containing
Dulbecco’s modified Eagle medium supplemented with 15% fetal bovine
serum (FBS) (HyClone, Lot KSD28666), 2 mM GlutaMAX (Invitrogen),
0.1 mM beta-mercaptoethanol (Sigma-Aldrich), 0.1 mM nonessential amino
acids (Invitrogen), 1 mM sodium pyruvate (Invitrogen), and 1,000 U/ml LIF
(Chemicon). For definitive endoderm differentiation, ESCs were cultured as
embryoid bodies (Kubo et al., 2004) for 2 days after which 100 ng/ml Activin
A (R&D Systems) was added for an additional 2 days.
Sox17-ESCs (Rugg-Gunn et al., 2012) were cultured in ESC media with
100 ng/ml dox (Sigma) to induce Sox17 expression. Sox17-XEN cells were
maintained in ESC media. For chimera experiments, cells were electroporated
with 10 mg of PB-CAG-mCherry and 1 mg pCAG-PBase (Wang et al., 2008)
followed by 150 mg/ml G418 selection (Gibco).
IM8AR XEN cells (Kunath et al., 2005) were cultured in RPMI 1640
(Gibco) supplemented with 15% FBS (HyClone, Lot KSD28666), 1 mM sodium
pyruvate (Gibco), 100 u mM beta-mercaptoethanol (Sigma-Aldrich), and 2 mM
L-glutamine (Gibco) at a ratio of 30% standard media to 70% MEF-conditioned media on 0.1% gelatin-coated plates or MEFs.
For siRNA experiments, ESCs were transfected with siRNA pools (Dharmacon) against Nr5a2 (L-047044-01) or Prdm14 (L-043221-01) at 50 mM final
Figure 4. Sox17-XEN Cells Transit through Distinct Phases during Sox17-Mediated XEN Conversion
(A) FACS plot of Sox17-ESCs on day 6 of dox culture (clone 4J).
(B) Gene expression analysis of each FACS compartment shown in (A). Error bars = SEM of four independent experiments.
(C) Flow cytometric analysis of cells sorted in (A) at day 12 and 18 of dox culture.
(D) Hierarchical clustering of RNA-seq data.
(E) Gene clusters identified by K-means clustering analysis.
(F) Enrichment maps outlining significantly enriched pathways and GO terms for gene clusters outlined in (E).
(G) GSEA comparing the enrichment of a VE gene set (Figure S4) in Sox17-D6 cells versus Sox17-XEN cells. Normalized enrichment score (NES) and nominal
(NOM) p values are included.
(H) RNA-seq expression (FPKM values) of VE genes (Amn, Cubn, and Lrp2) and parietal endoderm genes (Krt23, Nid1, and Krt18).
(I) E-cadherin expression measured by flow cytometry.
See also Figures S4 and S5 and Tables S1, S2, and S3.
Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors 9
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
(legend on next page)
10 Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
Figure 6. Identification of Putative Regulators of ExEn Cell Fate
(A) Dynamic regulatory map (Figure 5A) annotated with putative regulators of bifurcation points. DREM-identified regulators are listed in red and previously
identified ExEn regulators are listed in black.
(B) FPKM values were plotted for putative XEN regulators in three distinct network motif categories.
(C) XEN repressors expressed in ESCs, XEN activators expressed in XEN cells, and transcription factors expressed in both ESC and XEN cells.
See also Table S4.
concentration using Lipofectamine 2000 (Invitrogen) following the manufacturer’s instructions.
Gene Expression Analysis
Real-time PCR gene expression analysis was carried out as previously
described (Se´guin et al., 2008). Primer sequences are included in the Supplemental Experimental Procedures.
Illumina Sequencing and Data Analysis
RNA-seq libraries were prepared and sequenced at the McGill University and
Genome Quebec Innovation Centre using an Illumina HiSeq Genome Analyzer;
100-bp-long reads were obtained. Illumina reads were aligned to the mm9
mouse genome assembly and analyzed using the tuxedo package as previously described (Trapnell et al., 2012).
Immunostaining
Immunostaining was performed as previously described (Se´guin et al., 2008)
using antibodies listed in the Supplemental Experimental Procedures. Images
were captured using a Zeiss Axiovert 200 inverted microscope with a Hamamatsu C9100-13 EM-CCD camera and a Quorum spinning disk confocal scanning head. The Volocity software package (PerkinElmer) was used for image
acquisition.
Flow Cytometry
Generation and immunostaining of single-cell suspensions are outlined in
the Supplemental Experimental Procedures. Flow cytometry was performed using a Becton Dickinson LSR II and Dako Cytomation MoFlow
at the SickKids-UHN Flow Cytometry Facility. Fluorescence-activated
cell sorting (FACS) plots were generated using FlowJo software (Tree
Star).
Chimeras
Chimeric embryos were generated as previously described (Kunath et al.,
2005) with stable Sox17-XEN cells or Sox17(48)-XEN cells. In brief, wildtype blastocysts (C57BL/6J 3 B6D2F1/J) were collected at 3.25 days postcoitum (dpc) and injected with 10–20 mCherry-labeled cells and transferred
into pseudopregnant surrogates (2.5 dpc). Embryos were dissected at 6.5,
7.5, and 8.5 dpc and were imaged using a MZ16F microscope (Leica) equipped with a MicroPublisher 5.0 RTV camera (Qimaging). All animal experiments were performed in accordance with Canadian Council for Animal
Care guidelines.
Chromatin Immunoprecipitation
Chromatin was formaldehyde crosslinked and sonicated to approximately 1
kb lengths and analyzed using the Magna ChIP assay kit (Millipore, 17-611)
according to the manufacturer’s instructions. Chromatin was immunoprecipitated using 10 mg of goat anti-Sox17 (R&D, AB-108-C) or anti-goat immunoglobulin G (IgG) (R&D, AB-108-C). Resulting DNA was used as input for
real-time PCR using previously published primers (Niakan et al., 2010).
Statistical Analysis
Unpaired t tests were used to compare two groups. Results were expressed as
mean ± SEM. *p < 0.05; **p < 0.01. t tests were performed using Prism5
(GraphPad).
ACCESSION NUMBERS
RNA-seq data of R1 ESCs, IM8AR XEN cells, and Sox17-ESC time course
have been deposited at NCBI Gene Expression Omnibus under the accession
number GSE61102.
Figure 5. Construction of a Dynamic Regulatory Map for Sox17-Mediated XEN Conversion
(A) DREM-generated dynamic regulatory map during Sox17-ESC (clone 4J) dox culture. Expression levels relative to uninduced Sox17-ESCs. Gene expression
paths of interest are labeled by path number.
(B) DREM paths containing known ESC and XEN cell markers.
(C and D) Enrichment maps outlining significantly enriched pathways and GO terms for indicated gene expression paths (A).
See also Figure S6.
Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors 11
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
Figure 7. Sox17 Regulates Dynamic Networks Driving XEN Cell Fate
(A) Predicted Nr5a2 repressive motif in ESCs.
(B) Nr5a2 expression 48 hr after Nr5a2 knockdown in wild-type ESCs.
(C) Prdm14 expression 48 hr after Prdm14 knockdown in wild-type ESCs.
(D) Expression of XEN genes in wild-type ESCs 48 hr after knockdown of Nr5a2 or Prdm14. Error bars = SEM of two independent experiments.
(E–G) Schematic of GRN interactions regulating ESC and XEN cell states. Active interactions in Sox17-ESCs are shown in color. Inactive modules are shown in
gray (E). (F) Active interactions in Sox17-ESCs after 48 hr of Sox17 overexpression are shown in color. Inactive modules are shown in gray. (G) Active interactions
in Sox17-XEN cells are shown in color. Inactive modules are shown in gray.
(legend continued on next page)
12 Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures,
seven figures, and four tables and can be found with this article online at
http://dx.doi.org/10.1016/j.celrep.2014.09.026.
ACKNOWLEDGMENTS
We thank Linda Wei from the TCP Transgenic Core Facility for performing blastocyst injections, Jodi Garner at the SickKids ESC Facility for tissue culture,
Sheyun Zhao at the SickKids-UHN Flow Cytometry Facility for cell sorting,
Theodore Perkins, Gareth Palidwor, Chris Porter, and Matt Tanner at the
Sprott Centre for Stem Cell Research for advice and data processing, Andras
Nagy for providing piggyBac plasmids, Kevin Eggan and Kathy Niakan for
sharing ChIP-chip data sets, and David Picketts for critically reading the manuscript. This study was funded by Canadian Institutes of Health Research grant
MOP-97873 to J.R. and Canadian Institutes of Health Research grant MOP89910 to W.L.S. A.C.H.M. was supported by an Ontario Graduate Scholarship
in Science and Technology and a CIHR Banting and Best CGS Doctoral
Research Award. W.L.S. is supported by a Tier 1 Canada Research Chair in Integrated Stem Cell Biology.
Received: June 24, 2014
Revised: August 8, 2014
Accepted: September 14, 2014
Published: October 16, 2014
REFERENCES
Artus, J., Piliszek, A., and Hadjantonakis, A.K. (2011). The primitive endoderm
lineage of the mouse blastocyst: sequential transcription factor activation and
regulation of differentiation by Sox17. Dev. Biol. 350, 393–404.
Artus, J., Douvaras, P., Piliszek, A., Isern, J., Baron, M.H., and Hadjantonakis,
A.-K. (2012). BMP4 signaling directs primitive endoderm-derived XEN cells to
an extraembryonic visceral endoderm identity. Dev. Biol. 361, 245–262.
Brons, I.G.M., Smithers, L.E., Trotter, M.W.B., Rugg-Gunn, P., Sun, B., Chuva
de Sousa Lopes, S.M., Howlett, S.K., Clarkson, A., Ahrlund-Richter, L., Pedersen, R.A., and Vallier, L. (2007). Derivation of pluripotent epiblast stem cells
from mammalian embryos. Nature 448, 191–195.
Chambers, I., Colby, D., Robertson, M., Nichols, J., Lee, S., Tweedie, S., and
Smith, A. (2003). Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells. Cell 113, 643–655.
Chazaud, C., Yamanaka, Y., Pawson, T., and Rossant, J. (2006). Early lineage
segregation between epiblast and primitive endoderm in mouse blastocysts
through the Grb2-MAPK pathway. Dev. Cell 10, 615–624.
Cho, L.T., Wamaitha, S.E., Tsai, I.J., Artus, J., Sherwood, R.I., Pedersen, R.A.,
Hadjantonakis, A.-K., and Niakan, K.K. (2012). Conversion from mouse embryonic to extra-embryonic endoderm stem cells reveals distinct differentiation
capacities of pluripotent stem cell states. Development 139, 2866–2877.
Coucouvanis, E., and Martin, G.R. (1999). BMP signaling plays a role in visceral
endoderm differentiation and cavitation in the early mouse embryo. Development 126, 535–546.
Fujikura, J., Yamato, E., Yonemura, S., Hosoda, K., Masui, S., Nakao, K., Miyazaki Ji, J., and Niwa, H. (2002). Differentiation of embryonic stem cells is
induced by GATA factors. Genes Dev. 16, 784–789.
Futaki, S., Hayashi, Y., Emoto, T., Weber, C.N., and Sekiguchi, K. (2004). Sox7
plays crucial roles in parietal endoderm differentiation in F9 embryonal
carcinoma cells through regulating Gata-4 and Gata-6 expression. Mol. Cell.
Biol. 24, 10492–10503.
Gardner, R.L., and Rossant, J. (1979). Investigation of the fate of 4-5 day postcoitum mouse inner cell mass cells by blastocyst injection. J. Embryol. Exp.
Morphol. 52, 141–152.
Gu, P., Goodwin, B., Chung, A.C.-K., Xu, X., Wheeler, D.A., Price, R.R.,
Galardi, C., Peng, L., Latour, A.M., Koller, B.H., et al. (2005). Orphan nuclear
receptor LRH-1 is required to maintain Oct4 expression at the epiblast stage
of embryonic development. Mol. Cell. Biol. 25, 3492–3505.
Hanna, J., Saha, K., Pando, B., van Zon, J., Lengner, C.J., Creyghton, M.P.,
van Oudenaarden, A., and Jaenisch, R. (2009). Direct cell reprogramming is
a stochastic process amenable to acceleration. Nature 462, 595–601.
He, S., Kim, I., Lim, M.S., and Morrison, S.J. (2011). Sox17 expression confers
self-renewal potential and fetal stem cell characteristics upon adult hematopoietic progenitors. Genes Dev. 25, 1613–1627.
Heng, J.-C.D., Feng, B., Han, J., Jiang, J., Kraus, P., Ng, J.-H., Orlov, Y.L.,
Huss, M., Yang, L., Lufkin, T., et al. (2010). The nuclear receptor Nr5a2 can
replace Oct4 in the reprogramming of murine somatic cells to pluripotent cells.
Cell Stem Cell 6, 167–174.
Ivanova, N., Dobrin, R., Lu, R., Kotenko, I., Levorse, J., DeCoste, C., Schafer,
X., Lun, Y., and Lemischka, I.R. (2006). Dissecting self-renewal in stem cells
with RNA interference. Nature 442, 533–538.
Jollie, W.P. (1990). Development, morphology, and function of the yolk-sac
placenta of laboratory rodents. Teratology 41, 361–381.
Kanai-Azuma, M., Kanai, Y., Gad, J.M., Tajima, Y., Taya, C., Kurohmaru, M.,
Sanai, Y., Yonekawa, H., Yazaki, K., Tam, P.P.L., and Hayashi, Y. (2002).
Depletion of definitive gut endoderm in Sox17-null mutant mice. Development
129, 2367–2379.
Kim, I., Saunders, T.L., and Morrison, S.J. (2007). Sox17 dependence distinguishes the transcriptional regulation of fetal from adult hematopoietic stem
cells. Cell 130, 470–483.
Kubo, A., Shinozaki, K., Shannon, J.M., Kouskoff, V., Kennedy, M., Woo, S.,
Fehling, H.J., and Keller, G. (2004). Development of definitive endoderm
from embryonic stem cells in culture. Development 131, 1651–1662.
Kunath, T., Arnaud, D., Uy, G.D., Okamoto, I., Chureau, C., Yamanaka, Y.,
Heard, E., Gardner, R.L., Avner, P., and Rossant, J. (2005). Imprinted X-inactivation in extra-embryonic endoderm cell lines from mouse blastocysts.
Development 132, 1649–1661.
Labelle-Dumais, C., Jacob-Wagner, M., Pare´, J.-F., Be´langer, L., and Dufort,
D. (2006). Nuclear receptor NR5A2 is required for proper primitive streak
morphogenesis. Dev. Dyn. 235, 3359–3369.
Lanner, F. (2014). Lineage specification in the early mouse embryo. Exp. Cell
Res. 321, 32–39.
Ernst, J., Vainas, O., Harbison, C.T., Simon, I., and Bar-Joseph, Z. (2007).
Reconstructing dynamic regulatory maps. Mol. Syst. Biol. 3, 74.
Lim, C.Y., Tam, W.-L., Zhang, J., Ang, H.S., Jia, H., Lipovich, L., Ng, H.-H.,
Wei, C.-L., Sung, W.-K., Robson, P., et al. (2008). Sall4 regulates distinct transcription circuitries in different blastocyst-derived stem cell lineages. Cell
Stem Cell 3, 543–554.
Evans, M.J., and Kaufman, M.H. (1981). Establishment in culture of pluripotential cells from mouse embryos. Nature 292, 154–156.
Loh, K.M., and Lim, B. (2011). A precarious balance: pluripotency factors as
lineage specifiers. Cell Stem Cell 8, 363–369.
(H) Sox17 occupancy in Sox17-ESCs after 48 hr of dox treatment and in Sox17-XEN cells. Percentage of input and fold enrichment were calculated for Sox17 and
IgG-bound chromatin over a negative control region (Ppil4; Niakan et al., 2010) and presented as relative enrichment over uninduced Sox17-ESCs. Error bars =
SEM of two biological replicates.
(I) Schematic of Sox17 autoregulatory loop active in (F) and (G).
(J) Schematic of Sox17 feedforward loops active in (F) and (G).
See also Figure S7.
Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors 13
Please cite this article in press as: McDonald et al., Sox17-Mediated XEN Cell Conversion Identifies Dynamic Networks Controlling Cell-Fate Decisions
in Embryo-Derived Stem Cells, Cell Reports (2014), http://dx.doi.org/10.1016/j.celrep.2014.09.026
Lu, R., Yang, A., and Jin, Y. (2011). Dual functions of T-box 3 (Tbx3) in the
control of self-renewal and extraembryonic endoderm differentiation in mouse
embryonic stem cells. J. Biol. Chem. 286, 8425–8436.
Ma, Z., Swigut, T., Valouev, A., Rada-Iglesias, A., and Wysocka, J. (2011).
Sequence-specific regulator Prdm14 safeguards mouse ESCs from entering
extraembryonic endoderm fates. Nat. Struct. Mol. Biol. 18, 120–127.
Martin, G.R. (1981). Isolation of a pluripotent cell line from early mouse
embryos cultured in medium conditioned by teratocarcinoma stem cells.
Proc. Natl. Acad. Sci. USA 78, 7634–7638.
Masui, S., Nakatake, Y., Toyooka, Y., Shimosato, D., Yagi, R., Takahashi, K.,
Okochi, H., Okuda, A., Matoba, R., Sharov, A.A., et al. (2007). Pluripotency
governed by Sox2 via regulation of Oct3/4 expression in mouse embryonic
stem cells. Nat. Cell Biol. 9, 625–635.
Rugg-Gunn, P.J., Cox, B.J., Lanner, F., Sharma, P., Ignatchenko, V., McDonald, A.C.H., Garner, J., Gramolini, A.O., Rossant, J., and Kislinger, T.
(2012). Cell-surface proteomics identifies lineage-specific markers of embryo-derived stem cells. Dev. Cell 22, 887–901.
Sakaki-Yumoto, M., Kobayashi, C., Sato, A., Fujimura, S., Matsumoto, Y.,
Takasato, M., Kodama, T., Aburatani, H., Asashima, M., Yoshida, N., and
Nishinakamura, R. (2006). The murine homolog of SALL4, a causative gene
in Okihiro syndrome, is essential for embryonic stem cell proliferation, and
cooperates with Sall1 in anorectal, heart, brain and kidney development.
Development 133, 3005–3013.
Schrode, N., Saiz, N., Di Talia, S., and Hadjantonakis, A.K. (2014). GATA6
levels modulate primitive endoderm cell fate choice and timing in the mouse
blastocyst. Dev. Cell 29, 454–467.
Merico, D., Isserlin, R., Stueker, O., Emili, A., and Bader, G.D. (2010). Enrichment map: a network-based method for gene-set enrichment visualization
and interpretation. PLoS ONE 5, e13984.
Se´guin, C.A., Draper, J.S., Nagy, A., and Rossant, J. (2008). Establishment of
endoderm progenitors by SOX transcription factor expression in human
embryonic stem cells. Cell Stem Cell 3, 182–195.
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.
Shimoda, M., Kanai-Azuma, M., Hara, K., Miyazaki, S., Kanai, Y., Monden, M.,
and Miyazaki, J. (2007). Sox17 plays a substantial role in late-stage differentiation of the extraembryonic endoderm in vitro. J. Cell Sci. 120, 3859–3869.
Morrison, G.M., and Brickman, J.M. (2006). Conserved roles for Oct4 homologues in maintaining multipotency during early vertebrate development.
Development 133, 2011–2022.
Nagy, A., Rossant, J., Nagy, R., Abramow-Newerly, W., and Roder, J.C.
(1993). Derivation of completely cell culture-derived mice from early-passage
embryonic stem cells. Proc. Natl. Acad. Sci. USA 90, 8424–8428.
Niakan, K.K., Ji, H., Maehr, R., Vokes, S.A., Rodolfa, K.T., Sherwood, R.I.,
Yamaki, M., Dimos, J.T., Chen, A.E., Melton, D.A., et al. (2010). Sox17 promotes differentiation in mouse embryonic stem cells by directly regulating
extraembryonic gene expression and indirectly antagonizing self-renewal.
Genes Dev. 24, 312–326.
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.
Shimosato, D., Shiki, M., and Niwa, H. (2007). Extra-embryonic endoderm
cells derived from ES cells induced by GATA factors acquire the character
of XEN cells. BMC Dev. Biol. 7, 80.
Subramanian, A., Tamayo, P., Mootha, V.K., Mukherjee, S., Ebert, B.L.,
Gillette, M.A., Paulovich, A., Pomeroy, S.L., Golub, T.R., Lander, E.S., and
Mesirov, 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.
Tanaka, S., Kunath, T., Hadjantonakis, A.K., Nagy, A., and Rossant, J. (1998).
Promotion of trophoblast stem cell proliferation by FGF4. Science 282, 2072–
2075.
Tesar, P.J., Chenoweth, J.G., Brook, F.A., Davies, T.J., Evans, E.P., Mack,
D.L., Gardner, R.L., and McKay, R.D.G. (2007). New cell lines from mouse
epiblast share defining features with human embryonic stem cells. Nature
448, 196–199.
Niwa, H., Ogawa, K., Shimosato, D., and Adachi, K. (2009). A parallel circuit of
LIF signalling pathways maintains pluripotency of mouse ES cells. Nature 460,
118–122.
Trapnell, C., Roberts, A., Goff, L., Pertea, G., Kim, D., Kelley, D.R., Pimentel,
H., Salzberg, S.L., Rinn, J.L., and Pachter, L. (2012). Differential gene and
transcript expression analysis of RNA-seq experiments with TopHat and
Cufflinks. Nat. Protoc. 7, 562–578.
Paca, A., Se´guin, C.A., Clements, M., Ryczko, M., Rossant, J., Rodriguez, T.A.,
and Kunath, T. (2012). BMP signaling induces visceral endoderm differentiation of XEN cells and parietal endoderm. Dev. Biol. 361, 90–102.
Walker, E., and Stanford, W.L. (2009). Transcriptional networks regulating embryonic stem cell fate decisions. In Regulatory Networks in Stem Cells, V.K.
Rajasekhar and M.C. Vemuri, eds. (New York: Springer), pp. 87–100.
Pare, J.F., Roy, S., Galarneau, L., and Belanger, L. (2001). The mouse fetoprotein transcription factor (FTF) gene promoter is regulated by three GATA
elements with tandem E box and Nkx motifs, and FTF in turn activates the
Hnf3beta, Hnf4alpha, and Hnf1alpha gene promoters. J. Biol. Chem. 276,
13136–13144.
Walker, E., Chang, W.Y., Hunkapiller, J., Cagney, G., Garcha, K., Torchia, J.,
Krogan, N.J., Reiter, J.F., and Stanford, W.L. (2010). Polycomb-like 2 associates with PRC2 and regulates transcriptional networks during mouse embryonic stem cell self-renewal and differentiation. Cell Stem Cell 6, 153–166.
Plusa, B., Piliszek, A., Frankenberg, S., Artus, J., and Hadjantonakis, A.K.
(2008). Distinct sequential cell behaviours direct primitive endoderm formation
in the mouse blastocyst. Development 135, 3081–3091.
Wang, W., Lin, C., Lu, D., Ning, Z., Cox, T., Melvin, D., Wang, X., Bradley, A.,
and Liu, P. (2008). Chromosomal transposition of PiggyBac in mouse embryonic stem cells. Proc. Natl. Acad. Sci. USA 105, 9290–9295.
Qu, X.-B., Pan, J., Zhang, C., and Huang, S.-Y. (2008). Sox17 facilitates the
differentiation of mouse embryonic stem cells into primitive and definitive
endoderm in vitro. Dev. Growth Differ. 50, 585–593.
Wu, Q., Chen, X., Zhang, J., Loh, Y.-H., Low, T.-Y., Zhang, W., Zhang, W., Sze,
S.K., Lim, B., and Ng, H.H. (2006). Sall4 interacts with Nanog and co-occupies
Nanog genomic sites in embryonic stem cells. J. Biol. Chem. 281, 24090–
24094.
Rausa, F.M., Galarneau, L., Be´langer, L., and Costa, R.H. (1999). The nuclear
receptor fetoprotein transcription factor is coexpressed with its target gene
HNF-3beta in the developing murine liver, intestine and pancreas. Mech.
Dev. 89, 185–188.
Zhang, J., Tam, W.-L., Tong, G.Q., Wu, Q., Chan, H.-Y., Soh, B.-S., Lou, Y.,
Yang, J., Ma, Y., Chai, L., et al. (2006). Sall4 modulates embryonic stem cell
pluripotency and early embryonic development by the transcriptional regulation of Pou5f1. Nat. Cell Biol. 8, 1114–1123.
14 Cell Reports 9, 1–14, October 23, 2014 ª2014 The Authors