University of Southern California Dissertations and Theses

Pleotropic potential of Stat3 in determining self-renewal, apoptosis, and differentiation in mouse embryonic stem cells

(USC Thesis Other)

Pleotropic potential of Stat3 in determining self-renewal, apoptosis, and differentiation in mouse embryonic stem cells

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content PLEOTROPIC POTENTIAL OF STAT3 IN DETERMINING
SELF-RENEWAL, APOPTOSIS, AND DIFFERENTIATION IN
MOUSE EMBRYONIC STEM CELLS

By
Chih-I Tai

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(GENETIC, MOLECULAR, AND CELLULAR BIOLOGY)
Dec 2014

2
Dedication:

To my family for their unconditional support and faith in me

3
Acknowledgements

This work cannot be finished without the support and guidance provided
by my advisor, Dr. Qi-Long Ying, who first brought me to the field of stem
cell biology and established himself as a true scientist and a lifetime
mentor. I express my deepest gratitude to him for his criticism and
encouragement during last four years.

Additionally, I also want to thank members of my dissertation committees,
Dr. Andy McMahon, Dr. Cheng-Ming Chuong, Dr. Francesca Mariani, and Dr.
Robert Maxson for their helpful comments and assistance. At last, I thank
all the members in Ying’s lab for technical support and suggestions.

4
Table of Contents

Chapter I: Background of Embryonic Stem Cells

I. An overview of embryonic stem cells ........................................................................................ 7
II. Leukemia Inhibitor Factor (LIF) .................................................................................................. 9
III. LIF/Stat3 signaling .......................................................................................................................... 11
IV. LIF/ERK/MEK and LIF/PI3K/AKT signaling ....................................................................... 12
V. LIF/Stat3 targets .............................................................................................................................. 14

Chapter II: The role of Stat3 in ESC Self-Renewal

I. Introduction ....................................................................................................................................... 16
II. Results
a. Identification of genuine Stat3 targets ................................................................................... 17
b. Overexpression of Gbx2 maintains mESC self-renewal in the absence of LIF ..... 20
c. Gbx2 is a direct downstream target of the LIF/Stat3 pathway ............................... 26
d. Overexpressing Gbx2 sustains Stat3
−/−
ESCs in an undifferentiated state ............ 29
e. Gbx2 exhibits a redundant role in LIF-mediated self-renewal ................................ 30
f. Overexpression of En2 maintains mESC self-renewal in the absence of LIF ....... 32
g. En2 is a downstream target of the LIF/Stat3 pathway ............................................. 34
h. Overexpressing En2 sustains Stat3
−/−
ESCs in an undifferentiated state .............. 35
i. En2 exhibits a redundant role in LIF-mediated self-renewal .................................. 37
j. Gbx2 is specifically expressed in naï ve ESCs ............................................................... 38
k. Gbx2 enhances reprogramming of MEFs to a state of naï ve pluripotency by using
Oct4, Sox2, Klf4, c-Myc ...................................................................................................... 39
l. Gbx2 promotes reprogramming of EpiSCs to a state of naï ve pluripotency ........ 40
III. Discussion ............................................................................................................................................ 42

5
Chapter III: Stat3 signaling in promoting apoptosis and differentiation

I. Introduction ........................................................................................................................................ 46
II. Results

a. Enhancing Stat3 activity liberates B6 mESCs from the culture requirement of
feeders ................................................................................................................................................. 48
b. Stat3 induces differentiation of mESCs when its activation level exceeds certain
thresholds .......................................................................................................................................... 53
c. Mimicking Stat3 hyperactivation using a tyrosine phosphatase inhibitor ............. 59
d. Blocking Stat3 phosphorylation negative feedback regulation by knocking down
the expressions of PIAS3 and Socs3 ........................................................................................ 61
e. Hyperactivation of Stat3 in ESCs promotes trophectoderm lineage ......................... 64
f. Stat3 Cdx2 is the key factor that mediates TE differentiation of ESCs induced by
Stat3 hyperactivation .................................................................................................................... 67
g. Tfap2c, a direct downstream target of Stat3, is the key factor in mediating
trophectoderm differentiation induced by Stat3 hyperactivation ............................. 70
h. Blocking MEK/ERK signaling failed to prevent Stat3 induced trophectoderm
differentiation .................................................................................................................................. 75
i. Identify potential candidates blocking Stat3 induced TE differentiation by protein
kinase inhibitors screening ......................................................................................................... 76
j. GCSF induced Stat3 hyperactivation promotes apoptosis in ESCs ............................. 79
k. Enhancing Stat3 signaling induces tumor cells apoptosis ............................................. 80
l. The mechanism behind the apoptosis induced by Stat3 ................................................. 82
III. Discussion ............................................................................................................................................ 84

6
Chapter IV: Summary and Perspectives

I. Summary ................................................................................................................................................... 88
II. Potential mechanism for the roles of Gbx2 and En2 in ESC self-renewal
a. Gbx2 may act as a repressor of Otx2 to block ESC differentiation ............................ 89
b. En2 may enhance ESC self-renewal through its oncogenic properties ................... 92
c. Gbx2 is a major downstream effector of LIF/Stat3 signaling, while En2 may be a
common target for LIF/Stat3 and -Catenin pathways ................................................. 93
III. Potential mechanism for ESC apoptosis and differentiation induced by Stat3
hyperactivation
a. The expression level and sustainability of Stat3 activity determine whether ESCs
will differentiate into the TE lineage ..................................................................................... 95
b. Stat3 targets, Gbx2 and Klf4, are downregulated promptly after inducing Stat3
hyperactivation .............................................................................................................................. 96
c. Kinase inhibitor screening assay indicates that NF B pathways may also
participate in Stat3 induced TE differentiation ................................................................. 99
d. GCSF triggers B6-S3Y118F mESCs apoptosis in a JAK/ROCK1 dependent manner
............................................................................................................................................................ 100
IV. Improving and optimizing the culture conditions for ESCs derived from non-
responsive strains or species ......................................................................................................... 102
V. Conclusions ........................................................................................................................................... 104

Materials and Methods ............................................................................................................................. 105
References ......................................................................................................................................................... 111

7
Chapter I
Background of Embryonic Stem Cells

I. An overview of embryonic stem cells
Embryonic stem cells (ESCs) are tissue culture artifacts that represent a group of highly
potent cells at a very early stage of development before implantation. The
developmental course begins at the time of fertilization, an event that gives rise to an
embryo comprising cells that divide several times to form a morula (Figure 1). The
embryo then undergoes a morphology change following the redistribution of the
amplified cells inside the embryo and become a blastocyst, which possesses a
blastocoele (the cavity), an outer layer of trophoblast cells, and an inner cell mass. After
implantation, the trophoblast cells later combine with maternal endometrium and form
the placenta. Inner cell mass cells, on the other hand, give rise to primitive endoderm
and epiblast, and ultimately develop into an organism. ESCs are originated from the
inner cell mass.

Figure 1: Early Embryogenesis and Embryonic Stem Cell Derivation.

Adapted from Nature Reviews Molecular Cell Biology 6, 919-928

Nature Reviews Molecular Cell Biology 6, 919-928

8
There are two fundamental (or defining) properties of ESCs: the ability to infinitely
propagate, called self-renewal, and the capacity to differentiate into any cell type,
termed pluripotency. Since the derivation of mouse ESCs (mESCs) in 1981 (1, 2), these
particular characteristics have attracted the attention of researchers who work on cell
lineage differentiation and regenerative medicine. This pluripotency property also
allows scientists to manipulate and gene-modify ESCs to create animal models. With the
advent of gene modification approaches such as homologous recombination, zinc finger
nuclease and CRISPR/Cas systems (3-5), mouse ESCs (mESCs) have become a powerful
tool for the study of gene function. In 2007, for the discovery of mESCs and the
approach to create gene-targeted animals via homologous recombination using mESCs,
Sir Martin Evans, Mario Capecchi and Oliver Smithies were awarded the Nobel Prize in
physiology or medicine, highlighting the importance of ESC research and the translation
of these ideas to in vivo systems.
ESCs have long been considered tissue culture artifacts for several reasons. The
cells from the inner cell mass do not proliferate in vivo and show no long term self-
renewal. Moreover, ESCs in culture are exposed to numerous extrinsic signals from the
serum and, therefore, acquired some novel properties that enabled proliferation in an
undifferentiated state. However, these culture artifact ESCs possess the pluripotency to
develop into a whole organism when injected into a blastocyst, suggesting that ESCs
represent a frozen developmental state of inner cell mass cells. It is inevitable that cells
gain some functions from in vitro culture, which is not a recapitulation of the in vivo
environment, but ESCs are most related to the in vivo cell type in a blastocyst.

9
II. Leukemia Inhibitor Factor (LIF)
ESCs were first derived from mouse embryos in 1981 (1, 2). At the time, it was observed
that a substratal layer of mitotically inactive mouse fibroblast cells (so-called “feeders”)
was crucial for maintaining mESCs in an undifferentiated state. Subsequently, scientists
sought to identify and isolate the key molecule(s) responsible for this maintenance
effect. In 1988, A.G. Smith et al. and R. L. William et al found that feeders are
unnecessary when a soluble factor, identified as leukemia inhibitory factor (LIF), is
present (6, 7).
LIF is required to sustain the inner cell mass cells if blastocysts undergo embryo
diapause, which is a protective phenomenon in which suckling mammalians trigger a
delay in blastocyst implantation when the environment poses a risk to embryonic
development (8). However, LIF signaling knockout mouse models, such as GP130, LIFR,
JAK1, and Stat3, show normal early embryogenesis before implantation (9-12). This
implies that LIF is only required to sustain the cell population in the inner cell mass in a
halted phase of development, such as in vivo diapause or in vitro tissue culture. The
phenotypes of LIF signaling knockout mouse models also suggest that ESCs represent a
cellular population that exists only transiently before implantation.
In the 25 years since the discovery of LIF, scientists have been studying how LIF
promotes ESC self-renewal. LIF belongs to the interleukin-6 (IL-6) cytokine family, which
includes cardiotrophin-1, oncostatin M (OSM), and ciliaryneurotrophic factor (CNTF).
These cytokines activate signal transduction through glycoprotein 130 (GP130) and

10
show similar ability to support mESCs self-renewal (13-18), confirming that GP130
signaling is the key to sustaining mESC pluripotency (Figure 2). However, activation of
GP130 receptor complexes induces pleiotropic outcomes in various cell types in vivo and
in vitro. In addition to ESC self-renewal, stimulation of GP130 causes differentiation and
apoptosis in myeloid leukemia cells (19), induction of acute phase gene expression in
hepatocytes (20), proliferative and hypertrophic responses in cardiomyocytes(21),
cholinergic differentiation of sympathetic neurons (22), survival of motor neurons (23),
and astrocyte differentiation of neuroepithelial precursors (24).

Figure 2: IL-6 Family Cytokine Complexes

Each IL-6 family cytokine stimulates various combinations of receptor subunits. The signaling-
transduction subunit GP130 is the common component among all the cytokine complexes. This
implies that GP130 is crucial for the self-renewal mechanism in mESCs.

Picture adapted from Nature Reviews Neuroscience 8, 221-
232

11
III. LIF/Stat3 signaling
LIF binds to LIF receptor (LIFR), which forms a heterodimer with GP130 (Figure 3).
LIFR/GP130 dimerization leads to phosphorylation and activation of GP130 bound-JAK
tyrosine kinase (25, 26). Among the four JAKs, JAK1 is particularly important for the self-
renewal of mESCs, as indicated by the finding that higher LIF dosage is needed to
maintain mESCs with JAK1 expression knocked down than is needed for wild-type
mESCs(27). The cytoplasmic domain of GP130 contains several docking sites that are
phosphorylated by the activated JAKs. These phosphorylated sites recruit Src homology-
2 (SH2) domain-containing proteins such as signal transducer and activator of
transcription 3 (Stat3), which later becomes a target of the activated JAKs. Stat3 is
activated by JAKs through phosphorylation of its tyrosine 705 residue and forms a
homodimer with another activated Stat3. Dimerized Stat3 translocates into nucleus,
where it binds to promoters or enhancer regions of its target genes and functions as a
transcriptional activator (28).

Figure 3: LIF dependent activation of
Stat3 promotes self-renewal

LIF induces the association of LIFR and
gp130 cytokine receptors. Activated
JAKs cause the recruitment, tyrosine
phosphorylation and dimerization of
STAT3. The STAT3 dimers then
translocate to the nucleus, where they
control the transcription of
pluripotency genes. The importance of
STAT3 is confirmed by reports that a
conditionally regulated form of the
transcription factor, STAT3ER, can,
when activated, eliminate the need for
added LIF. ER: estrogen receptor.

12
Stat3 is the main effector of the activation of GP130 receptor complex by IL-6, LIF,
and CNTF. In myeloid leukemia M1 cells, activation of Stat3 appears to be the main
effector of the differentiation response to IL-6 or LIF (29). Stat3 activation has also been
shown to mediate CNTF or LIF-induced differentiation of neuroepithelial progenitors
into astrocytes (24).
Consistent with these findings, Stat3 also has been reported to be essential in LIF
mediated self-renewal and sufficient, in a constitutively active form, to eliminate the
need for LIF in the maintenance of mESC self-renewal. Niwa and colleagues created a
Stat3 dominant-negative mutant and found that expressing the mutant in mESCs
abrogates the self-renewal effect of LIF and promotes differentiation (30). They also
found that LIF dependency, self-renewal and clonal colony-formation are highly related
to Stat3 expression and activation (30). In another study, it was found that artificial,
hormone-dependent activation of chimeric Stat3 could eliminate the need for LIF and
maintain mESCs in an undifferentiated state (31) (Figure 3). Together, Stat3 is the crucial
component in LIF mediated self-renewal.

IV. LIF/ERK/MEK and LIF/PI3K/AKT signaling
Dimerization of LIFR/GP130 is also associated with phosphorylation of the tyrosine 118
residue (Y118) in the cytoplasmic domain of GP130. Phosphorylation of this tyrosine
allows the cytoplasmic domain to associate with the protein tyrosine phosphatase SHP-
2 (32). Parallel with the activation of Stat3 signaling, the binding of LIF to LIFR/GP130

13
leads to the activation of the mitogen-activated protein kinase (MAPK) and the
phosphatidylinositol-3 phosphate kinase (PI3K) pathways through SHP-2 (Figure 4).
SHP-2 interacts with growth-factor-receptor-bound protein 2 (Grb2) and son of
sevenless (SOS) to activate the MAPK pathway via the Ras/RAF/MEK/ERK cascade. The
MAPK pathway regulates many cellular responses and processes in somatic cells,
particularly proliferation and differentiation. In mESCs, the Ras/RAF/MEK/ERK cascade
can be activated by LIF and autocrine FGF4. Embryos with genetic ablation of Grb2 and
Shp-2 showed impaired early differentiation in the epiblast (33, 34), suggesting that LIF
and FGF4 induce MAPK signaling to provide differentiation cues for later lineage
determination after implantation. Specific attenuation of MAPK signaling either by
inhibition of MEK activity or forced expression of ERK phosphatases blocks
differentiation and further facilitates self-renewal (35), confirming that MAPK activation
possesses a pro-differentiation effect, and removing the differentiation cue enhances
mESCs self-renewal.

Figure 4: LIF activates MAPK pathway
thought SHP-2

LIF-induced activation of JAKs leads to the
recruitment and subsequent
phosphorylation of SHP2. SHP2 then
interacts with the Grb2 (growth-factor-
receptor-bound protein 2)–SOS (son of
sevenless) complex to activate the MAPK
pathway, which acts through the
Ras/RAF/MEK/ERK cascade.

14
Activated SHP-2 also induces the PI3K/AKT pathway. The serine/threonine kinase
AKT regulates multiple biological processes, including cell survival, proliferation, growth,
and glycogen metabolism (36). Blocking PI3K/AKT signaling by a chemical inhibitor,
LY294002, increases ERK phosphorylation and, therefore, in the long-term, reduces the
ability of LIF to promote self-renewal (37). The expression of constitutively activated
AKT in mESCs has been reported to promote self-renewal in the absence of LIF (38).
Niwa and colleagues proposed that T-box3 (Tbx3) is the main downstream target of PI3k
in the context of LIF-independent mESC self-renewal (39). However, the PI3K/AKT
pathway never showed significance in promoting mESC self-renewal in our experiments.
Unlike previous reports, suppressing or enhancing PI3K/AKT signaling showed no
phenotypes associated with the maintenance of mESC self-renewal, suggesting that
PI3K/AKT might just play a supportive, but dispensable role.

V. LIF/Stat3 targets
Current evidence shows that multiple targets of Stat3, such as Klf4 and Klf5, are involved
in ESC self-renewal (40-42). Overexpressing these factors individually bypasses LIF and
feeder requirements for mESC maintenance. However, Hall et al. reported that the total
colony number of Klf4 and Klf5 transfectants in the absence of LIF was greatly reduced
compared with culture in LIF (43), suggesting that overexpressing Klf4 and Klf5 is not
sufficient to recapitulate the effect of LIF. In addition, neither expression of Klf4 nor Klf5
can sustain self-renewal in Stat3 knockout mESCs (43). In the loss-of-function assays, it

15
was observed that mESCs can remain undifferentiated in the presence of LIF after knock
down of expression of both Klf4 and Klf5 (44). These findings support that LIF/Stat3
triggers multiple downstream factors by inducing different signals to maintain mESCs in
an undifferentiated state. Besides Klf4 and Klf5, several reports also show that Stat3
regulates Tfcp2l1, Pim1, Pim3, Pramel7, and c-Myc (42, 45-51). However, I could not
reproduce certain results in these studies. For example, I found LIF does not induce the
gene expression of Tfcp2l1, Pim3, Pramel7 and c-Myc in the mESCs that we have in our
lab. Therefore, likely, LIF induces the expression of these genes only in certain mESCs
strains. Currently, It is unclear how LIF/Stat3 contributes to mESC pluripotency.
Therefore, identifying genuine downstream effectors of Stat3 clarifies the mechanisms
how Stat3 maintains ESC self-renewal.

16
Chapter II
The role of Stat3 in ESC Self-Renewal

I. Introduction
As I mentioned earlier in Chapter I, it is still unclear how LIF/Stat3 mediates mESC self-
renewal. According to previous reports (43, 44), LIF/Stat3 might trigger multiple
downstream targets to maintain mESCs in an undifferentiated state. Therefore,
identifying genuine Stat3 targets that promote mESC self-renewal would help elucidate
the mechanism how Stat3 contributes to mESC pluripotency.
Previous attempts to identify LIF/Stat3 downstream targets responsible for
mediating mESC self-renewal have been hampered because LIF activates not only Stat3
but also many other pathways, including the MEK/ERK pathway, which negatively
regulates mESC self-renewal (35). In order to identify novel self-renewal-associated
genes induced by LIF, I focused on direct downstream targets of Stat3 by stimulating the
LIF/Stat3 pathway while blocking activation of the MEK/ERK pathway, and screened for
factors crucial for mESC self-renewal. Here, I identified Gastrulation Brain Homeobox 2
(Gbx2) and Engrailed 2 (En2) as downstream targets of Stat3, that when overexpressed
was sufficient to sustain mESC self-renewal in the absence of added LIF. I also found that
Gbx2 expression distinguishes naï ve state ESCs from primed-state epiblast stem cells
(EpiSCs), and promotes the conversion of EpiSCs to ESCs.

17
II. Results

a. Identification of genuine Stat3 targets
To facilitate the identification of Stat3 targets that contribute to mESC self-renewal, I
exploited the GRgp-Y118F chimeric receptor, which can activate JAK/Stat3 signaling
upon stimulation with granulocyte colony-stimulating factor (GCSF) (35). Substitution of
phenylalanine for tyrosine 118 in the gp130 segment of the receptor prevents it from
inducing MEK/ERK signaling, thereby enabling a high-fidelity examination of Stat3-
specific effects. I introduced GRgp-Y118F into B6 ESCs derived from the C57BL/6 mouse
strain. As expected, addition of GCSF to B6-Y118F ESCs strongly activated Stat3 without
affecting ERK1/2 phosphorylation (Fig. 5).

Figure 5: GCSF activates Stat3
without affecting ERK1/2
phosphorylation in B6-Y118F ESCs.

Stimulation by GCSF failed to induce
phosphorylation of ERK1/2. Addition of
PD0325901, a selective MEK inhibitor,
abolished the phosphorylation of ERK1/2
induced by LIF. Stimulation by either LIF or
GCSF induced Tyrosine 705 phosphorylation
of Stat3.
Fig. 5

18
To examine transcriptional changes associated with Stat3 activation, I starved B6-
Y118F ESCs in serum-free medium for 6 hours and then treated them with GCSF for one
hour to stimulate transcription of primary Stat3 targets but limit non-direct downstream
transcription. I treated a separate culture with LIF and the MEK inhibitor PD0325901, as
I reasoned this condition would mimic the effect of GCSF stimulation and thereby
corroborate results obtained with GCSF (Figure. 6).

Figure 6: Schematic illustration of the strategy to identify genuine Stat3
targets. B6 ESCs engineered to stably express a chimeric receptor (GRgp-Y118F) were able to
activate Jak/Stat3 but not MEK/ERK pathways in the presence of GCSF. PD0325901 was exploited
to prevent induction of MEK/ERK signaling by LIF. Treatment of the engineered B6 ESCs with GCSF
and LIF/PD0325901 specifically induced the Jak/Stat3 pathway.

Fig. 6

19
RNA from the treated cells was extracted and analyzed by microarray. Genes
upregulated in both GCSF and LIF/PD0325901 treatments were considered candidate
primary Stat3 targets. Based on the fold change values from microarray analysis and the
functional role of each gene as reported in the relevant scientific literature, 19 genes
were selected and their expression was confirmed by quantitative real-time PCR (qRT-
PCR) (Table 1). The microarray data are accessible through GEO series accession number
GSE38719.

PROBEID SYMBOL GC_vs_CN
microarray
LD_vs_CN
microarray
GC_vs_CN
qRT-PCR
LD_vs_CN
qRT-PCR
10356484 Gbx2 1.60 1.58 2.790 ± 0.085 2.224 ± 0.061
10385518 Tgtp 1.51 2.08 2.770 ± 0.136 3.928 ± 0.230
10580282 Junb 1.50 1.50 2.287 ± 0.101 3.367 ± 0.077
10560685 Bcl3 1.46 2.55 1.871 ± 0.025 4.254 ± 0.106
10426110 Pim3 1.43 1.62 2.208 ± 0.124 3.198 ± 0.193
10583519 Icam1 1.42 1.95 1.700 ± 0.240 1.704 ± 0.259
10502359 Dapp1 1.36 1.85 1.441 ± 0.060 1.770 ± 0.037
10405211 Gadd45g 1.35 1.79 2.980 ± 0.072 1.546 ± 0.095
10413393 Hesx1 1.30 2.16 1.801 ± 0.053 3.660 ± 0.115
10407350 Fgf10 1.28 1.72 1.663 ± 0.139 2.589 ± 0.097
10399360 Rhob 0.98 2.43 0.927 ± 0.003 3.028 ± 0.050
10513008 Klf4 1.23 1.89 1.994 ± 0.031 3.618 ± 0.056
10361091 Atf3 1.12 1.64 5.210 ± 0.370 3.025 ± 0.066
10408557 Serpinb1a 1.31 1.63 1.176 ± 0.057 2.531 ± 0.110
10520368 En2 1.34 1.61 1.440 ± 0.028 1.649 ± 0.059
10371662 Spic 1.04 1.47 1.249 ± 0.077 2.524 ± 0.107
10372781 Irak3 1.21 1.50 1.114 ± 0.082 1.624 ± 0.141
10520965 Yes1 1.12 1.51 1.373 ± 0.049 1.741 ± 0.133
10493474 Muc1 1.10 1.50 2.404 ± 0.093 3.178 ± 0.055

Table 1
1
Table 1: Candidate genes identified by microarray. The list of candidate genes
selected based on the fold change values and their biological functions. The fold change values from
microarray analysis were validated by qRT-PCR. Data represent mean ± SD of four experimental
replicates. GC, GCSF; CN, control; LD, LIF/PD0325901.

20
b. Overexpression of Gbx2 maintains mESC self-renewal in the absence of LIF
To screen genes that can promote mESC self-renewal, I artificially expressed each of the
19 candidate genes in 46C ESCs and cultured them in the absence of LIF. 46C ESCs are
feeder-independent ESCs derived from the 129 mouse strain (52). Among the 19 genes
screened, Gbx2 was found capable of supporting mESC self-renewal when
overexpressed. Specifically, 46C ESCs transduced to overexpress a Gbx2 transgene
formed colonies exhibiting typical undifferentiated morphology in the absence of LIF
(Fig. 7a, b). To further evaluate their status, I fixed the cells and stained them for
alkaline phosphatase (AP) activity and classified the colonies as differentiated,
undifferentiated or mixed. Approximately 25% of the Gbx2 transfectant colonies were
undifferentiated and 50% were mixed while 90% of the EGFP transfectant colonies were
fully differentiated (Fig. 8). Overexpression of Gbx2 supported long-term maintenance
of undifferentiated 46C ESCs in the absence of added LIF (Fig. 9a, b). The results were
validated in another mESC line, R1, to exclude the possibility that the phenotype was
attributed to the specific genetic background of 46C ESCs (Fig. 10). shRNA knock down
of Gbx2 in Gbx2-overexpressing R1 ESCs (Fig. 11a) rendered them unable to retain an
ESC identity in the absence of LIF (Fig. 11b, c), confirming the observed self-renewal-
promoting effect is indeed attributed to Gbx2 overexpression.

21

a
EGFP
Gbx2
Relative expression level
WT
Gbx2
0
10
20
30
b Figure 7: Overexpressing
Gbx2 sustains mESC self-
renewal.
(a) Phase-contrast images of
46C ESCs overexpressing
EGFP or Gbx2 after 5 days’
culture in the absence of LIF.
(b) qPCR analysis of Gbx2
mRNA expression in 46CESCs
transduced to express EGFP
and Gbx2.
AP staining positive
colonies
EGF
P
Gbx
2
EGFP
Gbx2
Figure 8: Overexpressing Gbx2 sustains mESC self-
renewal.
AP staining and quantification of colonies of 46C ESCs
overexpressing EGFP or Gbx2 after 7 days without LIF
supplementation.
Fig. 7
Fig. 8

22

a
b
Phase
Phase
Phase
Oct4
Nanog
Klf4
DAPI
DAPI
DAPI
Phase
Phase
Phase
Oct4
Nanog
Klf4
DAPI
DAPI
DAPI
Figure 9: Overexpressing Gbx2 sustains mESC self-renewal.
(a) AP staining of 46C ESCs overexpressing Gbx2 cultured more than 40 days
(passage>9) in the absence of LIF supplementation. (b) Immunofluorescence
staining of Oct4, Nanog, and Klf4 in 46C ESCs overexpressing Gbx2 cultured more
than 40 days (passage>9) in the absence of LIF supplementation. Scale bars = 50
m.
Fig. 9

23

Fig. 10
R1-EGFP R1-Gbx2
a
b
Phase
Phase
Oct4
Nanog
DAPI
DAPI
C
Figure 10: Overexpression of
Gbx2 sustains R1 ESC self-
renewal in the absence of
LIF.
(a) Colony morphology of R1
ESCs overexpressing EGFP or
Gbx2 cultured for 5 days in the
absence of LIF
supplementation. (b) AP
staining of R1 ESCs
overexpressing Gbx2 cultured
more than 3 weeks
(passage>4) in the absence of
LIF supplementation. (c)
Immunofluorescence staining
of Oct4, Nanog, and Klf4 in R1
ESCs overexpressing Gbx2
cultured more than 3 weeks
(passage>4) in the absence of
LIF supplementation. Scale
bars = 50 m

24

Phase Klf4 DAPI
b
a
b
c
Figure 11: The self-renewal phenotype of Gbx2-overexpressing ESCs is
attributed to the elevated expression of Gbx2.
(a) Expression of Flag-tagged Gbx2 in Gbx2-overexpressing R1 ESCs transduced to
express control and Gbx2 shRNAs was examined by western blot using anti-Flag
antibody. (b) AP staining of Gbx2-overexpressing R1 ESCs transduced to express
control and Gbx2 shRNAs. Scale bars: 25 m. (c) Quantitative analysis of AP
staining results for Gbx2-overexpressing R1 ESCs transduced to express control
and Gbx2 shRNAs.
Fig. 11

25
To determine whether ESCs maintained by Gbx2 overexpression retain
pluripotency, I exploited a loxP-based excisable vector containing the Gbx2 transgene.
Transient expression of Cre recombinase results in excision of the Gbx2 transgene and
simultaneously brings GFP under the CAG promoter. ESCs transfected with the excisable
vector were clonally selected and then maintained in serum without added LIF for 20
days. The cells were then transiently transduced to express Cre. Gbx2 excision
transfectants enriched by FACS sorting reacquired LIF dependence (Figure. 12a). The
revertant cells were injected into 12 mouse blastocysts and gave rise to three chimeric
fetuses with widespread GFP expression (Figure. 12b), suggesting that Gbx2
overexpressing ESCs retain their pluripotency.

Fig. 12
46C-Gbx2
46C-Gbx2 + Cre
a b
Figure 12: Forced expression of Gbx2
maintains mESCs pluripotency
(a) Phase contrast images of Gbx2-
overexpressing cells and Cre-excision derivatives
in serum medium without LIF. (b) The three
chimeric embryos (E11.5) with widespread GFP,
generated from Gbx2 transfectants subjected to
Cre-mediated excision of Gbx2.

26
c. Gbx2 is a direct downstream target of the LIF/Stat3 pathway
Jak1-mediated phosphorylation of Stat3 results in homodimerization and activation of
Stat3 (53). To further confirm that Gbx2 is downstream of the LIF/Stat3 pathway, I
evaluated Gbx2 gene induction in 46C cells treated with a Jak1 inhibitor in the absence
of the inhibitor, Gbx2 transcription showed a prompt induction within one hour of
continuous LIF treatment, whereas the induction was abolished in the presence of the
inhibitor (Figure 13a). Moreover, overexpression of Stat3 in mESCs was associated with
higher basal and LIF-induced expression of Gbx2 than was detected in control mESCs
expressing Stat3 at the endogenous level (Figure 13b), indicating that the expression
level of Stat3 is correlated with the expression level of Gbx2. The induction
of Gbx2expression by LIF was similar with or without cycloheximide (Figure 13c), an
inhibitor of protein biosynthesis, which excludes the possible regulation of Gbx2 by
other LIF-induced downstream targets of Stat3.

27

I next exploited Stat3 null (Stat3
− /−
) ESCs to further test the requirement for
Stat3 in Gbx2 induction. Stat3
− /−
ESCs are routinely maintained in the 2i condition
containing Gsk3 inhibitor CHIR99021 and MEK/ERK inhibitor PD0325901 (54). Expression
of Socs3, a direct downstream target of Stat3, was upregulated in 46C ESCs upon 1 hour
of LIF treatment, but not in equivalently treated Stat3
− /−
ESCs (Figure 14a). Likewise, LIF
Fig. 13
a b
a

a
c
b
a

a
Figure 13: Gbx2 is directly regulated by LIF/Stat3 and its overexpression is
sufficient to sustain Stat3
−/−
ESC self-renewal. (a) qRT-PCR analysis of Gbx2 mRNA
expression in 46C ESCs after LIF treatment for 1 hour. 46C ESCs were starved for 6
hours in serum-free medium before LIF stimulation. Jak1 inhibitor (Calbiochem; 10
µM) was administered 1 hour before LIF. (b) qRT-PCR analysis of Gbx2 mRNA
expression in B6 ESCs overexpressing EGFP or Stat3 after LIF treatment for 1 hour.
These cells were starved for 6 hours in serum-free medium before LIF stimulation. (c)
qRT-PCR analysis of Gbx2 mRNA expression in 46C ESCs after LIF stimulation in the
presence or absence of cyclohexamide (CHX; 50 µg/ml) for 1 or 3 hours.

28
failed to induce Gbx2 expression in Stat3
−/−
ESCs (Figure 14b). A chromatin
immunoprecipitation (ChIP) assay performed on mESCs by using antibodies against
Stat3 indicated an association between Stat3 and Gbx2 loci (55, 56). Taken together,
these data convincingly demonstrate that Gbx2 is a direct target of the LIF/Stat3
signaling pathway.

Fig. 14
a
b
Figure 14: LIF failed to induce Gbx2 expression in Stat3
−/−
ESCs
(a) Socs3, a downstream target of Stat3, is induced by LIF in 46C ESCs but not in
Stat3
−/−
ESCs. qRT-PCR analysis of endogenous Socs3 mRNA expression in 46C
and Stat3
−/−
ESCs with or without LIF stimulation for 1 hour.. (b) qRT-PCR
analysis of Gbx2 mRNA expression in Stat3
−/−
ESCs after LIF treatment for 1 hour.
Error bars represent s.d. (n=4)

29
d. Overexpressing Gbx2 sustains Stat3
−/−
ESCs in an undifferentiated state
Having established a role for Gbx2 as a downstream agent of LIF/Stat3-mediated self-
renewal, I next sought to determine whether overexpression of Gbx2 would suffice to
maintain the undifferentiated state in Stat3
−/−
ESCs. In serum medium without LIF or 2i,
Stat3
−/−
ESCs transduced to overexpress Gbx2 could be propagated and passaged while
retaining typical ESC morphology, AP activity and expression of Oct4, Nanog and Klf4
(Figure 15a–c). In contrast, control Stat3
−/−
ESCs transduced to express EGFP died or
differentiated within three days in the same condition (Figure 15a). These results imply
that overexpression of Gbx2 can compensate for the absence of Stat3 to maintain the
self-renewal phenotype.

30
Fig. 15

e. Gbx2 exhibits a redundant role in LIF-mediated self-renewal
To ascertain whether Gbx2 is essential for LIF/Stat3-mediated ESC self-renewal, I
infected 46C ESCs with shRNAs designed to knock down endogenous Gbx2 expression,
and collected RNA samples 5 days later. Expression of Gbx2 was 70% to 80% reduced as
measured by qRT-PCR (Figure 16a). In the presence of LIF, a greater percentage of
partially differentiated colonies was observed for Gbx2 knockdown 46C ESCs than cells
a
b
c
Figure 15: Forced Gbx2
expression promotes self-
renewal of Stat3
−/−
ESCs (a)
Colony morphology of Stat3
−/−

ESCs overexpressing Gbx2,
cultured for 6 days in serum-
only medium. (b) AP staining of
Stat3
−/−
ESCs overexpressing
Gbx2 cultured for more than 3
weeks (passage >6). (c)
Immunofluorescence staining
of Oct4, Nanog and Klf4 in
Stat3
−/−
ESCs overexpressing
Gbx2, cultured for more than 3
weeks (passage >6).

31
infected with scramble control shRNA (Figure 16b). However, knockdown of Gbx2 was
incapable of totally impairing ESC self-renewal: More than 50% of Gbx2 knockdown
colonies remained undifferentiated and only 5–15% of colonies fully differentiated
(Figure 16c).

Fig. 16

Figure 16: Gbx2 is redundant in LIF-mediated self-renewal.
(a) qRT-PCR analysis of endogenous Gbx2 mRNA expression in 46C ESCs
transduced to express control or Gbx2 shRNAs. Error bars represent the s.d.
(n = 4). *P<0.01. (b) AP staining of 46C ESCs transduced to express control and
Gbx2 shRNAs. Scale bars: 50 µm. (c) Quantitative analysis of AP staining in 46C
ESCs transduced to express control and Gbx2 shRNAs.

32
These results suggest that while Gbx2 plays a role in promoting mESC self-renewal,
its expression is not essential for LIF/Stat3-mediated self-renewal. This is consistent with
the finding that knockout of Gbx2 results in no morphological abnormalities in the
blastocyst that would indicate a defect in self-renewal (57). Moreover, knockdown of
two other Stat3 targets, Klf4 and Klf5, does not lead to mESC differentiation in the
presence of LIF (44). Collectively, these results imply that LIF/Stat3 supports mESC self-
renewal through redundant mechanisms involving multiple factors downstream of Stat3,
such as Gbx2, Klf4 and Klf5.

f. Overexpression of En2 maintains mESC self-renewal in the absence of LIF
Besides Gbx2, En2 was also found capable of supporting mESC self-renewal when
overexpressed. Colonies of 46C ESCs transduced to overexpress an En2 transgene
showed typical ESC morphology in the absence of LIF (Figure 17a). To further validate
their status, AP staining was performed and I found ESCs transfected to express En2
showed more undifferentiated and mixed colonies than the EGFP transfectant (Figure
17b). Similar phenotype was also observed in R1 ESCs (Figure 17d). Overexpression of
En2 supported long-term maintenance of 46C and R1 ESCs in the absence of LIF (Figure
17c, e), suggesting that the self-renewal promoting phenotype is not due to the specific
genetic background of 46C ESCs.

33
Fig. 17

a b c
Figure 17: Overexpression of En2 sustains 46C and R1 ESC self-renewal
in the absence of LIF.
(a) Colony morphology of 46C ESCs overexpressing EGFP or En2 cultured for 5
days in the absence of LIF supplementation. (b) AP staining of 46C ESCs
overexpressing En2 cultured more than 3 weeks (passage>4) in the absence of
LIF supplementation. (c) Immunofluorescence staining of Oct4, Nanog, and
Klf4 in 46C ESCs overexpressing En2 cultured more than 3 weeks (passage>4)
in the absence of LIF supplementation.

34

g. En2 is a downstream target of the LIF/Stat3 pathway
I next confirmed whether En2 is downstream of the LIF/Stat3 pathway by treating 46
mESCs with a Jak1 inhibitor. The inhibitor can block the induction of En2 transcription
after one hour of LIF stimulation (Figure 18a). Moreover, En2 induction has no response
to LIF in Stat3
− /−
ESCs (Figure 18b). This suggests En2 is regulated by LIF/Stat3 signaling.

e
(d) Colony morphology of R1 ESCs overexpressing vector alone or En2 cultured
for 5 days in the absence of LIF supplementation. (e) Immunofluorescence
staining of Oct4, Nanog, and Klf4 in R1 ESCs overexpressing En2 cultured more
than 3 weeks (passage>4) in the absence of LIF supplementation
d

35
Fig. 18

h. Overexpressing En2 sustains Stat3
−/−
ESCs in an undifferentiated state
Having established a role for En2 as a downstream target of Stat3, I next asked whether
overexpression of En2 can compensate the loss of Stat3 and maintain Stat3
−/−
mESCs
undifferentiated. In serum medium without LIF or 2i, control Stat3
−/−
ESCs transduced to
express EGFP died or differentiated within three days in the same condition (Figure 19a).
However, Stat3
−/−
ESCs transduced to overexpress En2 could be propagated and
passaged while retaining typical ESC morphology and expression of Oct4, Nanog and
Klf4 (Figure 19a, b). These results imply that, like Gbx2, overexpression of another Stat3
a
b
Figure 18: En2 is regulated by LIF/Stat3 and its overexpression is sufficient
to sustain Stat3
−/−
ESC self-renewal. (a) qRT-PCR analysis of En2 mRNA
expression in 46C ESCs after LIF treatment for 1 hour. 46C ESCs were starved for
6 hours in serum-free medium before LIF stimulation. Jak1 inhibitor
(Calbiochem; 10 µM) was administered 1 hour before LIF. (b) qRT-PCR analysis
of En2 mRNA expression in Stat3
−/−
ESCs after LIF treatment for 1 hour. Error
bars represent s.d. (n=4)

36
target, En2, can compensate for the absence of Stat3 to maintain the self-renewal
phenotype.
Fig. 19

Vec_day6
En2-F_day6
OCT4
NANOG
KLF5
DAPI
DAPI
DAPI
Figure 19: Overexpression of
En2 sustains Stat3
−/−
ESC self-
renewal in the absence of LIF.
(a) Colony morphology of
Stat3
−/−
ESCs overexpressing
vector only or En2 cultured for 5
days in the absence of LIF
supplementation. (b)
Immunofluorescence staining of
Oct4, Nanog, and Klf4 in Stat3
−/−

ESCs overexpressing En2
cultured more than 3 weeks
(passage>4) in the absence of LIF
supplementation.
a
b

37
i. En2 exhibits a redundant role in LIF-mediated self-renewal
En2 has shown to contribute LIF/Stat3-mediated ESC self-renewal; therefore, I want to
ascertain if En2 is also essential for LIF/Stat3-mediated ESC self-renewal. I knocked
down endogenous En2 expression by En2 shRNAs. Expression of En2 was 40% to 70%
reduced as measured by qRT-PCR (Figure 20a). In the presence of LIF, I observed a
higher percentage of partially differentiated colonies in En2 knockdown 46C ESCs than
ESCs transfected to express scramble control shRNA (Figure 20b). However, overall
knocking down the endogenous expression of En2 was incapable of abrogating ESC self-
renewal (Figure 20b). These En2 knockdown ESCs can still be continuously passaged.

Fig. 20

a
b
Figure 20: En2 is redundant in
LIF-mediated self-renewal.
(a) qRT-PCR analysis of
endogenous En2 mRNA
expression in 46C ESCs
transduced to express control or
Gbx2 shRNAs. Error bars
represent the s.d. (n = 4). *P<0.01.
(b) Quantitative analysis of AP
staining in 46C ESCs transduced
to express control scramble and
En2 shRNAs.

38
j. Gbx2 is specifically expressed in naï ve ESCs
I next examined the expression of Gbx2 in undifferentiated and differentiating 46C ESCs.
Gbx2 mRNA expression diminished rapidly during embryoid body formation (Figure 21a),
and was only negligibly detectable in epiblast stem cells (EpiSCs) differentiated from 46C
ESCs (Figure 21b, c). These results indicate that loss of the ESC state coincides with
downregulation of Gbx2 following removal of the LIF/Stat3 signal.
Fig. 21

a
b
a
c
a
Figure 21: Gbx2 expresses
only in the ESC state and is
downregulated after
differentiation
(a) qRT-PCR analysis of Gbx2
mRNA expression at different
time points of embryoid body
differentiation. (b) qRT-PCR
analysis of Gbx2 mRNA
expression in 46C ESCs and
EpiSCs. (c) qRT-PCR analysis of
Fgf5, Rex1 and Oct4 mRNA
expression to distinguish 46C
ESCs and EpiSCs. Error bars
represent SD (n=4).

39
k. Gbx2 enhances reprogramming of MEFs to a state of naï ve pluripotency by
using Oct4, Sox2, Klf4, c-Myc
Because Gbx2 is a transcription factor and is specifically expressed in the ESCs, I asked
whether it could function as a reprogramming factor analogous to Oct4, Sox2, Klf4, and
c-Myc (58, 59). I transduced mouse embryonic fibroblasts (MEFs) to express all five of
these factors and found that they exhibited higher reprogramming efficiency than
controls expressing Oct4, Sox2, Klf4, c-Myc and EGFP, as indicated by the number of
induced pluripotent stem cell (iPSC) colonies that formed (Figure 22a, b). This result
implies that Gbx2 synergistically cooperates with the other four factors in the
reprogramming process.

40
Fig. 22

l. Gbx2 promotes reprogramming of EpiSCs to a state of naï ve pluripotency
Recent reports have established that primed-state EpiSCs can be converted to naï ve-
state ESCs by activation of Stat3 signaling and overexpression of its downstream targets,
such as Klf4 and Nanog (60-62). To test whether Gbx2 is sufficient to mediate
reprogramming of EpiSCs to a naï ve pluripotent state, I used the lentiviral system to
introduce the Gbx2 transgene into 129 mouse strain-derived E3 EpiSCs carrying an Oct4-
GFP transgene (63). After selection, the cells were cultured atop feeders in LIF plus 2i
a
b
Figure 22: Gbx2 Enhances
reprogramming efficiency of MEF or
iPCs using four Yamanaka factors
(a) AP staining conducted on MEFs on
day 20 of OSKM reprogramming with
EGFP or Gbx2. O, Oct4; S, Sox2; K, Klf4;
M, c-Myc. (b) Quantitative analysis of
colony formation of OSKM+EGFP- and
OSKM+Gbx2-expressing cells. Error bars
represent the s.d. (n = 2). *P<0.05.

41
and within 5 days formed compact, dome-shaped colonies exhibiting strong AP activity,
while wild-type E3 EpiSCs differentiated or died in the same condition (Figure 23a). qRT-
PCR analysis of the Gbx2 transfectants showed Oct4 expression and a typical ESC marker
profile, with upregulation of Rex1, Stella and Klf4. Conversely, the expression of Fgf5, a
marker for EpiSCs, was lost. The mRNA expression level of Gbx2 in the transfectants was
4- to 14-fold higher than in ESCs (Figure 23b). These data demonstrate that forced
expression of Gbx2 can convert EpiSCs back to naï ve pluripotency.
Fig. 23

Figure 23: Gbx2
promotes
reprogramming of
EpiSCs to a state of
naï ve pluripotency
(a) Oct4 expression and
AP activity in E3 EpiSCs
transduced to express
EGFP or Gbx2. The images
were taken 5 days after
transfer to LIF plus 2i-
supplemented medium.
Scale bars: 50 µm. (b)
qRT-PCR analysis of
expression of key markers
in ESCs, EpiSCs and ESC-
like cells. C1, C2, C4, C7
and C9 are the individual
colonies of ESC-like cells.
The y-axis is the relative
expression normalized to
that of ESCs. Error bars
represent the s.d. (n = 4).
*P<0.01.
a
b

42
III. Discussion
In this project, a novel approach was applied to specifically reveal direct downstream
targets of Stat3, and I successfully identified known and novel factors related to LIF-
mediated pluripotency. Among these downstream targets, Gbx2 showed a unique
lineage-specific expression pattern in mESCs and sufficiency to sustain mESCs self-
renewal in the absence of LIF. The presence of Gbx2 thus imparts the pluripotency and
differentiation potential of ESCs. The finding provides an insight for understanding the
mechanisms of ESCs self-renewal.
To specifically focus on direct targets of Stat3 induced by LIF, I blocked MEK/ERK
signaling by using a genetically engineered mouse ESC line and a selective MEK inhibitor.
The duration of stimulation is a key factor to identify direct targets of Stat3. For example,
Socs3, a Stat3 target, can be activated within 30 mins after addition of LIF (57).
Therefore, the stimulatory agents in the study were applied for one hour to avoid the
indirect transcription signals activated by Stat3 downstream genes. Cartwright et al.
reported that c-Myc is a target of Stat3, based on the results of a ChIP assay and an
increase in c-Myc mRNA level after LIF stimulation for 24 hours (58). In my microarray
analysis, however, I did not observe upregulation of c-Myc under those condtions. I
propose that c-Myc is indirectly regulated by Stat3 on the following grounds: First,
activation of c-Myc required prolonged LIF stimulation (2). Second, the ChIP data
reported by Cartwright does not suffice to indicate that c-Myc is a direct target of Stat3.
Stat3 is a key regulatory factor for many genes in ESCs, and it has been reported that
58.3% of Stat3-binding sites are shared with Oct4, Sox2, and/or Nanog (49). Therefore, it

43
is likely that Stat3 works as a cofactor and indirectly binds to the promoter region of c-
Myc. My results indicate that not all direct targets of Stat3 can be revealed by detecting
gene expression levels at one hour post-LIF administration. For instance, at one hour
following LIF or GCSF administration, I did not observe induced expression of Pramel7 in
ESCs, which has been reported to be upregulated 10 minutes after LIF stimulation is first
applied (59).
Among the candidate genes summarized by microarray analysis, Gbx2 and En2
were identified to promote and maintain mESCs self-renewal in the absence of LIF.
Moreover, Klf4, Klf5 and Pim1, reported as Stat3 targets and crucial factors for LIF
mediated self-renewal (40, 43, 60), were also observed to be upregulated in the
microarray analysis (Klf5 showed 1.47- and 1.35-fold change in GCSF versus control and
LIF/PD versus control, respectively). Hence, the data further suggest that this novel
approach provides a platform for identifying the genuine direct pluripotency targets of
LIF/Jak/Stat3.
I found that mESCs overexpressing Gbx2 or En2 can survive after single-cell
dissociation and can be propagated indefinitely. Paradoxically, both Gbx2 and En2 are
associated with nervous system development. In mice, Gbx2 and En2 are involved in
brain development, specifically, determination of the mid/hindbrain border (61, 62).
Gbx2 overexpression has been indicated to be a crucial for the clonogenicity and
tumorigenesis of certain human prostate cancer cell lines (63). En2 was also reported to
be essential for morphological transformation and proliferation in breast cancer cell
lines (64). Moreover, many of the pluripotency factors are indicated to be highly

44
correlated to tumorigenesis (65, 66). These reports further echo my finding that Gbx2
and En2 are involved in mESC self-renewal.
Gbx2 was first reported to be associated to ESC pluripotency based on its stage-
specific expression during blastocyst formation (67). The promoter of Gbx2 is co-
occupied and potentially regulated by several pluripotency factors, such Stat3, Oct4,
Tcf3, Sox2 and Nanog (48, 68, 69). The expression level of Gbx2 is highly correlated to
the expression level of Stat3, and cannot be induced by LIF in Stat3 null ESCs. In addition,
overexpression of Gbx2 can compensate for the loss of Stat3, allowing maintenance of
Stat3 null cells in the serum condition. Taken together, the data suggest that Gbx2
locates immediately downstream of Stat3 and contributes to ESC pluripotency.
Some reports suggest Gbx2 is a pluripotency marker in mESCs (70, 71). Due to its
function in the development of central neural system, Gbx2 has also been considered a
neuroectoderm marker in human ESCs (72-74). Wang et al reported that ectopic
expression of Gbx2 facilitated the differentiation of human ESCs (74), implying that Gbx2
plays different roles in naï ve- versus primed-state ESCs. My result indicates that
expression of Gbx2 decreases sharply during in vitro differentiation and diminishes
when ESCs differentiate into epiblast stem cells (70). Taken together, these data indicate
that Gbx2 is a lineage marker of the naï ve state ESCs and possesses different functions
in different lineage contexts.
Yamanaka and Takahashi used four factors to overcome the developmental
restriction, converting mouse somatic cells to ESC-like (51). The four factors, Oct4, Sox2,
Klf4, and c-Myc (OSKM) are transcription factors highly expressed and related to ESCs

45
self-renewal. Based on the results in this chapter, the finding that Gbx2 was shown to
promote self-renewal implies its potential role as a reprogramming factor. Indeed,
addition of Gbx2 enhanced OSKM-mediated reprogramming efficiency of fibroblasts to
iPSCs. The result is consistent with the hypothesis that pluripotency-promoting factors
might be primary effectors of reprogramming (51).
Multiple downstream targets of LIF/Stat3 have been identified as crucial to ESC
self-renewal (43, 59, 60). In this study, overexpression of Gbx2, En2, or Klf4 was able to
substitute for the effect of LIF to maintain ESC self-renewal in the serum condition;
however, these gene–modified ESC lines gave rise to fewer undifferentiated colonies
than did the control ESCs cultured in the presence of LIF. This suggests that individually
expressing these factors cannot fully mimic the effect of LIF. Knock down of Gbx2, En2
and Klf4 expression showed no impact on differentiation in the presence of LIF (En2 and
Klf4 data not shown). Although I cannot rule out the effects of remaining gene
expressions after siRNA knockdown, there is no report indicating that the knockout
model of these three genes showed any abnormal phenotypes during the blastocyst
stages (62, 75, 76). Therefore, the functions of these factors in self-renewal may be
dispensable and compensated for by other pluripotency factors regulated by LIF/Stat3.
Moreover, consistent with the previous finding (43), overexpression of Klf4 could not
support Stat3 null ESCs for multiple passages, while Gbx2 could, suggesting Klf4 and
Gbx2 regulate self-renewal by distinct mechanisms. These lines of evidence imply that
LIF/Stat3 supports ESC self-renewal through redundant self-renewal mechanisms
involving multiple downstream factors of Stat3.

46
Chapter III
Stat3 signaling in promoting apoptosis and
differentiation

I. Introduction
ESCs were originally derived by explanting mouse blastocysts onto a layer of mitotically
inactivated fibroblasts (‘feeders’) in medium containing fetal calf serum (FCS) (1, 2).
Under this condition, mESCs lines can be derived and subsequently propagated
indefinitely. LIF can replace feeder cells in serum-containing medium for derivation of
ESCs from the 129 strain of mice (6, 7). LIF maintains mESC self-renewal through
activation of the Stat3. Intriguingly, strains of mice vary greatly in the ease with which
their embryos can be directed to give rise to mESCs under the LIF+FCS condition (64),
and the underlying mechanism remains largely unknown. The 129 strain has been found
to be the most permissive for ESC derivation (65). 129 mESCs can be derived and
maintained in the absence of feeders in medium supplemented with LIF and FCS (66).
To date, no ESC lines have been derived from any non-129 strains of mice in the LIF+FCS
condition without the use of feeders.
Stat3 is the crucial component in LIF mediate self-renewal. LIF/Stat3 signaling
activates multiple downstream targets in mESCs. LIF/Stat3-mediated mESC self-renewal
can be partially recapitulated by overexpressing some of these downstream targets,
including Tfcp2l1, Gbx2, Klf4, Klf5, Pim1, Pim3, Pramel7, and c-Myc (42, 45-51).

47
Current understanding of Stat3-mediated mESC self-renewal has been dominated
by the view that LIF/Stat3 signaling functions in a binary “on/off” manner. It is not clear
whether the Stat3 activation level is critical to ESC self-renewal and whether non-129
strain mESCs can be maintained under feeder-free condition through manipulation of
Stat3 activity.
Here, I used chemical and genetic approaches to modulate Stat3 activity in mESCs
and determined the corresponding effect on self-renewal and differentiation. I
observed that Stat3 exhibits a dose-dependent effect in mESC self-renewal and
differentiation.

48
II. Results

a. Enhancing Stat3 activity liberates B6 mESCs from the culture requirement of
feeders
Embryonic stem cells (ESCs) were originally derived by explanting mouse blastocysts on
to a layer of mitotically inactivated fibroblasts (‘feeders’) in medium containing fetal calf
serum (FCS) (1, 2). Under these conditions ESC lines can be derived and subsequently
propagated indefinitely. Leukemia inhibitory factor (LIF) can replace feeder cells in
serum-containing medium for derivation of ESCs from the 129 strain of mice (6, 7). LIF
maintains mESC self-renewal through activation of Stat3. Intriguingly, strains of mice
vary greatly in the ease with which their embryos can be directed to give rise to ESCs
under the LIF+FCS condition (64), and the underlying mechanism remains largely
unknown. The 129 strain has been found to be the most permissive for ESC derivation
(65). 129 mESCs can be derived and maintained in the absence of feeders in medium
supplemented with LIF and FCS (66). To date, no ESC lines have been derived from any
non-129 strains of mice in LIF+FCS condition without the use of feeders.
Current understanding of Stat3-mediated ESC self-renewal has been dominated by
the view that LIF/Stat3 signaling function in a binary “on/off” manner. It is not clear
whether Stat3 activation level is critical to ESC self-renewal and whether non-129 strain
mESCs can be maintained under feeder-free condition through manipulation of Stat3
activity. B6 mESCs were derived from the C57BL/6 strain of mouse and are routinely

49
maintained by co-culture with feeders in the presence of LIF and FCS. When removed
from feeders and cultured in ESC medium supplemented with LIF and FCS, these B6
mESCs died or differentiated and could not be continuously propagated (Figure 24a). I
examined total and phosphorylated Stat3 levels in LIF-stimulated B6 mESCs and found
that they were both markedly lower than that of LIF-stimulated 46C mESCs, an feeder-
independent ESC line derived from 129 strain of mouse (Figure 24b, c) (52).

50
Fig. 24

a
a
a
a
b
c
Figure 24: Feeder dependent B6 mESC showed lower Stat3 expression and
activation in the condition of LIF compared to feeder independent 46C
mESCs.
(a) Phase contrast (upper panel) and AP staining images (lower panel) of B6 ESCs
cultured in mESC medium supplemented with LIF with or without feeders for 7
days. (b) Western blot analysis of phospho-Stat3 (Tyr705) and total Stat3
expression levels in B6-WT and 46C mESCs treated with or without LIF for 30min.
B6-WT and 46C mESCs were starved in serum free medium for overnight before
the treatment. (c) Relative expression levels of total and phosphorylated Stat3 in
B6-WT and 46C mESCs treated with LIF for 30 mins. The Stat3 expression levels
were normalized to -tublin without LIF stimulation. (c) Quantitative result of (b)
normalized with the expression of -tublin

51
Next, I asked whether enhancing Stat3 activity can maintain B6 mESC self-renewal
in the feeder-free condition. I introduced a Stat3 transgene into B6 mESCs to increase
overall Stat3 expression (hereafter called B6-Stat3). I also engineered B6 mESCs to
express a gp130-Y118F chimeric receptor (B6-Y118F). The mutated Y118F receptor
consists of the extracellular domain of the granulocyte colony-stimulating factor (GCSF)
receptor fused to the transmembrane and cytoplasmic region of gp130 and contains a
tyrosine-for-phenylalanine substitution at residue 118 (35), which can prevent a Stat3
negative feedback regulator, suppressor of cytokine signaling 3 (Socs3), from blocking
de novo Stat3 activation. The Y118F ESCs allow me to activate Stat3 in a GCSF dose-
dependent manner, circumventing the endogenous LIF/gp130 receptors that might be
limited in ESCs. As expected, increased total and phosphorylated Stat3 levels were
observed in B6-Stat3 mESCs compared to B6 mESCs. Phosphorylated Stat3 levels in B6-
Y118F mESCs treated with GCSF were also significantly higher than that of LIF-treated B6
mESCs (Figure 25c). Both B6-Stat3 and B6-Y118F mESCs could be maintained without
feeders in the presence of LIF and GCSF, respectively (Figure 25a, b), suggesting that
enhancing Stat3 activity can liberate B6 mESCs from the dependence of feeders.

52
Fig. 25

Figure 25: Enhancing Stat3 activity liberates
B6 ESCs from the dependence of feeders
(a) Phase contrast images of B6-WT, B6-Stat3,
and B6-Y118F mESCs cultured in mESC medium
supplemented with either LIF or GCSF in the
absence of feeders for 7 days. ESCs were plated
into 0.1% gelatin-coated 6-well plates at a
density of 1000 cells/well. (b) AP staining result
of (a). Scale bars: 50 µm
a
b
c

(c) Western blot analysis of phospho-Stat3 (Tyr705) and total Stat3levels in wild-type
B6 (B6-WT), B6-Stat3 (left panel) and B6-Y118F mESCs (right panel) treated with LIF or
GCSF for the indicated times. NT: no treatment.

53
b. Stat3 induces differentiation of mESCs when its activation level exceeds certain
thresholds
To further define how Stat3 activation level can affect ESC fate, I introduced both Stat3
and gp130-Y118F transgenes into B6 mESCs (hereafter called B6-S3Y118F mESCs) so that
I can broadly manipulate Stat3 activation level. B6-S3Y118F mESCs remained
undifferentiated without feeders in the presence of LIF (Figure 26a). Surprisingly, B6-
S3Y118F mESCs, instead of maintained self-renewal, differentiated right after
administration of 50ng/ml GCSF for 24 h (Figure 26b). Administration of LIF activates
JAK/Stat3 as well as PI3K/AKT and MEK/ERK signaling pathways, while administration
GCSF activates JAK/Stat3 but not PI3K/AKT and MEK/ERK pathways in B6-S3Y118F
mESCs due to the lack of SHP2 docking site in the chimeric gp130-Y118F chimeric
receptor (35). To determine whether lack of PI3K/AKT or/and MEK/ERK activation is the
cause of the differentiation phenotype, selective chemical inhibitors against these two
pathways (LY03934 and PD0325901) were applied to the cells along with LIF. I found
that B6-S3Y118F mESCs remained undifferentiated and showed no apoptotic phenotype
after administration of the inhibitors along with LIF (Figure 26c, d), suggesting that lack
of PI3K/AKT or MEK/ERK signaling is unlikely the cause of the rapid differentiation
phenotype.

54
Fig. 26

Next, I asked whether the phenotype is correlated with Stat3 activation level. JAK1
activity is required for Stat3 activation through regulation of Stat3 phosphorylation (27),
therefore, I used a selective JAK1 inhibitor (JAK1i) to attenuate Stat3 activity in mESCs.
Indeed, GCSF-induced differentiation in B6-S3Y118F mESCs could be prevented by
applying 1 g/ml JAK1i (Figure 27a). When co-administrated with 1 g/ml JAK1i, B6-
S3Y118F mESCs could be continuously propagated and remained undifferentiated in the
presence of 50 ng/ml GCSF, a dose of GCSF otherwise would induce rapid differentiation.
Contrarily, B6-S3Y118F mESCs could not be maintained in LIF plus 1 g/ml JAK1i (Figure
27b). Taken together, these results imply that excess Stat3 activation is the main cause
of B6-S3Y118F mESC differentiation induced by GCSF.
a b c d
Figure 26: GCSF induces B6-S3Y118F mESCs death and differentiation
(a) Phase contrast images of B6-S3Y118F mESCs cultured in mESC medium
supplemented with LIF or GCSF for 24 hours (b). (c) Phase contrast images of B6-
S3Y118F mESCs cultured in the presence of LIF plus 1 µM PD0325901 (MEK
inhibitor) or 5 µM LY294002 (PI3K inhibitor) (d) for 48 hours.

55
Fig. 27

To further confirm that Stat3 activity is correlated to the rapid differentiation
phenotype, I manipulated Stat3 activity by applying different concentrations of GCSF
and JAK1i. I found out that reduced concentration of GCSF can maintain B6-S3Y118F
ESCs self-renewal and that an adequate dosage of JAK1 inhibitor can also prevent the
cells from the differentiation induced by GCSF (Figure 28a). I examined Stat3
phosphorylation levels in B6-S3Y118F mESCs treated with various concentrations of
GCSF and JAK1i (Figure 28b). The result indicated that Stat3 can be significantly
activated in B6-S3Y118F ESCs by GCSF comparing the cells in the presence of LIF (Figure
28c). Stat3 activation level can be modulated via applying varying concentrations of
GCSF and JAK1 inhibitor. Combining the emergence of differentiation phenotype with
the quantitated Stat3 activation level normalized with -tublin expression level, a
a
b
Figure 27: GCSF induces B6-S3Y118F mESCs death and
differentiation
(a) Phase contrast images of B6-S3Y118F mESCs cultured in the
presence of GCSF, or GCSF plus 1 µg/ml JAK1i for 48 hours. (b) AP
staining of B6-S3Y118F mESCs cultured in the presence of LIF or GCSF
with or without JAK1i for 7 days.

56
threshold can be drawn to distinguish whether the B6-S3Y118F mESCs can be
maintained self-renewal or induced differentiation (Figure 28d), suggesting that Stat3
signaling promotes differentiation instead of self-renewal once its activation exceeds
certain level.
Fig. 28

a b
c
Figure 28: GCSF induces B6-S3Y118F mESCs death and differentiation
(a) Phase contrast images of B6-S3Y118F mESCs cultured in the presence of various
doses of GCSF for 3 days. (b) Phase contrast images of B6-S3Y118F mESCs cultured in
the presence of 50 ng/ml GCSF plus various doses of JAK1i for 3 days. (c) Western blot
analysis of phospho-Stat3 (Tyr705) levels of B6-S3Y118F mESCs treated with
different doses of GCSF and JAK1i for 24 hours. (d) Quantification of the normalized
phospho-Stat3 (Tyr705) levels against endogenous -tublin from the result in (c).
Red bar indicates the conditions that mESCs undergo differentiation. Blue bar
indicates the conditions that mESCs can be maintained undifferentiated. Scale bars:
50 µm
d

57
I tested this rapid differentiation phenotype in another mESC line to confirm it is
a general phenomenon in mESCs. The similar differentiation phenotype can be observed
when I introduced both Stat3 and chimeric gp130-Y118F transgenes into E14 mESCs, a
widely-used MEF-independent mESC cell line. Both E14-Y118F and E14-S3Y118F mESCs
could be maintained in LIF condition as well (Figure 29a). However, E14-S3Y118F ESCs
underwent an apparent rapid differentiation following application of GCSF while E14-
Y118F ESCs remained undifferentiated (Figure 29a). The Stat3 phosphorylation in E14-
S3Y118F mESCs could be induced higher and more sustained by GCSF than LIF (Figure
29b, c). The phenotype can also be reversed by giving JAK1 inhibitor (Figure 29d).
Fig. 29

a
Figure 29: Co-
expressing Stat3
transgene and
chimeric gp130-
Y118F in E14 mESCs
phenocopies the
crisis phenotype in
B6-S3Y118F mESCs
(a) Phase contrast
images of E14-Y118F,
E14-Stat3, and E14-
S3Y118F mESCs
cultured in mESC
medium only or mESC
medium supplemented
with LIF or GCSF.
Images were taken 6
days after plating.

58

(b) Western blot analysis of phospho-Stat3 (Tyr705)
levels in E14-Stat3 and E14-S3Y118F mESCs treated
with LIF or GCSF for 30 mins. Both E14-Stat3 and
E14-S3Y118F mESCs were starved in the serum free
medium for overnight before the treatment. (c)
Western blot analysis of phospho-Stat3 (Tyr705)
levels in E14-S3Y118F ESCs treated with LIF or GCSF
for 30 mins or 6 hours. The mESCs were starved in
the serum free for overnight medium before the
treatment. (d) Phase contrast images of E14-
S3Y118F ESCs cultured in GCSF or GCSF plus JAK1i
for 3 days.
b
c
d

59
c. Mimicking Stat3 hyperactivation using a tyrosine phosphatase inhibitor
Stat3 phosphorylation is strictly regulated by negative feedback regulators or a process
of de-phosphorylation, which is mediated by
protein tyrosine phosphatases (PTPs). A PTP
inhibitor, sodium vanadate (Na
3
VO
4
), was
applied to B6-S3Y118F mESCs to mimic the
effect of GCSF in inducing Stat3
phosphorylation. mESCs formed more
compact and spherical colonies and showed
more cell death when treated with sodium
vanadate than control. (Figure 30a, upper
panel). Moreover, mESCs transfected to over
express Stat3 showed more sensitivity to sodium vanadate (Figure 30a, lower panel).
Moreover, mESCs transfected to express additional Stat3 showed more sensitive to
sodium vanadate (Figure 30a, lower panel). The apoptosis phenotype could be rescued
by giving JAK1i (Figure 30b). However, unlike GCSF induced ESC crisis, TE differentiation
was not observed and the Stat3 activity showed similar with or without SV for 24 hours
(Figure 30c). The results suggest that either Stat3 or JAK1 activation can induce
apoptosis. A selective Stat3 inhibitor can be applied to confirm if the apoptosis is
triggered through Stat3 signaling or other pathways induced by JAK1. A similar result
was observed when I used SHP1/2 inhibitor to replace SV. Addition of SHP1/2 inhibitor

60
to B6-S3Y118F mESCs induced apoptosis (Figure 30d) without affecting Stat3
phosphorylation (Figure 30e).
Fig 30

I.

a
Figure 30: Mimicking the effect of Stat3 hyperactivation by using a protein tyrosine
phosphatase inhibitor, sodium vanadate (SV) and a SHP1/2 inhibitor
(a) Phase contrast image of B6-Y118F and B6-S3Y118F mESCs in the presence of LIF plus
the indicated SV concentrations (b) Phase contrast image of B6-S3Y118F mESCs in the
presence of LIF and 12 mM SV with or without JAK1i. (c) Western blot result of phospho-
Stat3 (Tyr705) and -tublin in B6-S3Y118F mESCs in the indicated conditions. (d) Western
blot result of phospho- -tublin in B6-S3Y118F mESCs in the indicated
concentration of the SHP1/2 inhibitor. (e) Phase contrast image of B6-S3Y118F mESCs in
the presence of LIF and the indicated concentration of SHP1/2 inhibitor.

b
a
c
b
a
d
c
b
a
e
d
c
b
a

61

d. Blocking Stat3 phosphorylation negative feedback regulation by knocking down
the expressions of PIAS3 and Socs3
Suppressor of cytokine signaling 3 (Socs3) is a negative feedback regulator for Stat3
phosphorylation turnover. Socs3 exerts its inhibitory function through the recruitment
of SHP2 to gp130 (67). Stat3 activation induced by GCSF in B6-S3Y118F mESCs is not only
higher but also more sustained than it is in B6 wild type and B6-Stat3 mESCs. Stat3
phosphorylation through the gp130-Y118F chimeric receptor can bypass the Socs3
negative feedback loop, as gp130-Y118F lacks a suitable SHP2 docking site. Therefore, I
examined whether reducing Socs3 expression in B6-Stat3 mESCs could phenocopy the
result in B6-S3Y118F mESCs. However, knock down of endogenous Socs3 expression
could not induce B6-Stat3 mESCs to undergo apoptosis or differentiation. Instead, B6-
Stat3 mESCs transfected to express Socs3 shRNA gave rise to colonies that were more
compact than those of untransfected cells (Figure 31a). The level of Stat3 activation 24
hours after LIF stimulation was slightly higher in B6-Stat3 Socs3 shRNA cells than in
controls (Figure 31b). The result suggests that blocking Socs3 negative feedback
regulation can sustain Stat3 activity and induce the cells to form colonies with more
spherical and compact morphology. Stat3 activation is controlled by the amount of
ligands, LIFR/gp130 receptors, Stat3, and Stat3 phosphorylation in the turnover system.
To achieve Stat3 hyperactivation, B6-Stat3 mESCs require not only an impaired SHP2
negative feedback loop, but also a high abundance of gp130 receptor units.

62
Fig 31

Protein inhibitor of activated STAT3 (PIAS3) is a transcriptional repressor that
inhibits the DNA binding ability of Stat3 (68). Based on the same rationale I applied in
the experiments with Socs3, I asked whether blocking the Stat3 inhibitor, PIAS3, could
recapitulate the crisis effect. However, knock down of PIAS3 expression did not enhance
Stat3 signaling; instead, it reduced Stat3-mediated transcription as judged by the
a
b
a
b
a
Figure 31: Mimicking the effect of Stat3 hyperactivation by knocking down
the endogenous expression of Socs3.
(a) Western blot result of phospho-Stat3 (Tyr705) and -tublin in B6-Stat3 and
B6-Stat3 mESCs transfected to express Socs3 shRNA in the indicated conditions
(b) Phase contrast images of B6-Stat3 and B6-Stat3 mESCs transfected to express
Socs3 shRNA in the condition of LIF

63
expression levels of its direct targets, Gbx2 and En2 (Figure 32a). Reducing PIAS3
expression had no effect on the crisis phenotype induced by Stat3 hyperactivation
(Figure 32b). The result contradicts previous reports that PIAS3 acts as a Stat3 inhibitor.
A more thorough study should be done to ascertain whether PIAS3 functions as a
transcriptional cofactor of Stat3.
Fig 32

1
st
lane: B6-S3-118
2
nd
lane: B6-S3-118 + PIAS3 shRNA #1#4
3
rd
land: B6-S3-118 + PIAS3 shRNA #1#5
Figure 32: Mimicking the effect of Stat3 hyperactivation by knocking
down the endogenous expression of PIAS3.
(a) qPCR analysis of PIAS3, Gbx2, En2, and Klf4 expression in B6-S3Y118F
mESCs transfected with control or PIAS3 shRNA (b) Phase contrast images of
B6-S3Y118F mESCs transfected with control or PIAS3 shRNA in the condition
of 10 ng/ml GCSF for 4 days

b
a
a
b
a

64
e. Hyperactivation of Stat3 in ESCs promotes trophectoderm lineage
B6-S3Y118F ESCs underwent rapid differentiation morphologically when Stat3 is
hyperactivated by GCSF (Figure 33a). I performed RT-PCR analysis to determine which
cell lineages are induced by Stat3 hyperactivation. Total RNA was collected from B6-
S3Y118F ESCs treated with GCSF at different time points and the expression of key
differentiation makers was compared to the cells maintained in LIF. Expression of the
ESC markers, Gbx2, Klf4, Oct4, Sox2, and Esrrb was significantly downregulated after
treatment with 50 ng/ml GCSF within 24 hours (Figure 33b), confirming that ESC
underwent differentiation.

65
Fig. 33

Next, I examined the expression levels of various germ layer markers in B6-
S3Y118F mESCs treated with GCSF. Surprisingly, GCSF treatment did not upregulated
any of the three somatic germ layer markers tested (Figure 34a). Instead, it significantly
induced the expression of trophectoderm (TE) markers (Figure 34b). These results
a
b
Figure 33: Stat3 hyperactivation
downregulates pluripotency
markers and promotes
differentiation
(a) Phase contrast images of B6-
S3Y118F mESCs maintained in LIF or
treated with GCSF for the indicated
times. (b) qPCR analysis of the
expression of pluripotency genes in
B6-S3Y118F mESCs maintained in
LIF or treated with GCSF for the
indicated times.

66
suggest that Stat3 hyperactivation promotes ESCs' exit from the naive state into TE
lineage.
Fig. 34

b
a
Figure 34: Stat3 hyperactivation promotes B6-S3Y118F mESCS to differentiate
into trophectoderm lineage

(a) qPCR analysis of gene expression in B6-S3Y118F mESCs maintained in LIF or treated
with GCSF for the indicated times. Endoderm markers: Gata4, FoxA2, Sox17; ectoderm
markers: Nestin, Sox1; mesoderm markers: MixL1, T. (b) qPCR analysis of gene
expression in B6-S3Y118F mESCs maintained in LIF or treated with GCSF for the
indicated times. Eomes, Cdx2, Dlx3, Esx1, Gata3, and Psx1 are markers of TE.

67
f. Stat3 Cdx2 is the key factor that mediates TE differentiation of ESCs induced by
Stat3 hyperactivation
Among the TE markers, Cdx2 and Gata3 were highly upregulated within 24 hours of
GCSF treatment in B6-S3Y118F mESCs (Figure 35a). Notably, both Cdx2 and Gata3 have
been reported to be key regulators in TE differentiation and overexpressing either one is
sufficient to promote ESC differentiation towards TE lineage (69-72). Therefore, I asked
whether early induction of Cdx2 or Gata3 contributes to TE differentiation triggered by
Stat3 hyperactivation.

68
Fig. 35

II.

a
b
c
d
e
Figure 35: GCSF induced Cdx2 expression mediates the TE differentiation
phenotype
(a) ) qPCR analysis of Cdx2 expression in B6-S3Y118F mESCs treated with GCSF or
LIF for the indicated times. B6-S3Y118F mESCs were starved overnight in serum
free medium prior to the treatments. (b) qRT-PCR analysis of Cdx2 expression in B6-
S3Y118F mESCs expressing scramble or Cdx2 shRNAs. (c) qRT-PCR analysis of Cdx2
expression in scramble or Cdx2 shRNAs-expressing B6-S3Y118F mESCs treated with
LIF or GCSF for 24 hours. (d) Phase contract images of scramble or Cdx2 shRNAs-
expressing B6-S3Y118F mESCs cultured in the presence of GCSF for 7 days. (e) AP
staining of scramble or Cdx2 shRNAs-expressing B6-S3Y118F mESCs cultured in the
presence of GCSF for 7 days.

69
I first examined Cdx2 expression profile of B6-S3Y118F mESCs treated with either
GCSF or LIF and found that expression of Cdx2 was induced by GCSF but not by LIF
(Figure 35b). shRNA-mediated knockdown of Cdx2 reduced both endogenous and
stimulated expression level of Cdx2 in LIF and GCSF conditions (Figure 35c, d). Cdx2
knockdown could prevent B6-S3Y118F ESCs from entering TE differentiation (Figure 35e,
f). Gata3, likes Cdx2, was also induced by GCSF but not LIF in B6-S3Y118F mESCs (Figure
36a). shRNA-mediated Gata3 knockdown (Figure 36b, c), however, was not sufficient to
block TE differentiation induced by GCSF (Figure, 36d). Together, these results suggest
that Cdx2 mediates TE differentiation of ESCs induced by Stat3 hyperactivation.
Fig. 36

a
c d
Figure 36: GCSF induced Gata3 expression is not required for the TE differentiation
phenotype (a) qPCR analysis of Gata3 expression in B6-S3Y118F mESCs treated with
GCSF or LIF for the indicated times. B6-S3Y118F mESCs were starved overnight in serum
free medium prior to the treatments. (b) qRT-PCR analysis of Gata3 expression in scramble
or Gata3 shRNAs-expressing B6-S3Y118F mESCs maintained in LIF. (c) qRT-PCR analysis of
Gata3 expression in scramble or Gata3 shRNAs-expressing B6-S3Y118F mESCs treated with
GCSF for 24 hours. (d) Phase contract images of scramble or Gata3 shRNAs-expressing B6-
S3Y118F mESCs cultured in the presence of GCSF for 7 days. Scale bars: 50 µm. Error bars
represent the s.d. (n=4).
b

70
g. Tfap2c, a direct downstream target of Stat3, is the key factor in mediating
trophectoderm differentiation induced by Stat3 hyperactivation
Cdx2 expression has been shown to be regulated by the WNT/ -catenin (73) and the
Hippo/TEAD4 pathways (74). I next asked whether these two pathways are involved in
Cdx2 induction by Stat3 hyperactivation. I applied selective inhibitors to activate or
inhibit WNT/ -catenin signaling and found no change in Cdx2 expression following GCSF
treatment (Figure 37a). These inhibitors did not prevent GCSF-induced TE differentiation
either (Figure 37b).
Fig. 37

Figure 37: GCSF induced TE differentiation
is in a WNT/ -catenin signaling
independent manner
(a) qRT-PCR analysis of Cdx2 expression in B6-
S3Y118F mESCs cultured in the indicated
conditions. (b) Phase contrast images of B6-
S3Y118F mESCs cultured in the indicated
conditions for 5 days.
a
b

71
To examine the role of Tead4, I knocked down its expression via multiple shRNAs
(Figure 38a). Tead4 knockdown partially downregulated the induction of Cdx2 and
Gata3 by GCSF (Figure 38b, c). However, Tead4 knockdown could not prevent the TE
differentiation (Figure 38d). Taken together, these results imply that both WNT/ -
catenin and Hippo/TEAD4 signaling are not the significant regulators of Cdx2 induction
caused by Stat3 hyperactivation.
Fig. 38

b a
b
c
a
b
Figure 38: Tead4 can partially regulate Cdx2 and Gata3 expression induced
by GCSF but is not required for Stat3 hyperactivation mediated TE
differentiation
(a) qRT-PCR analysis of Tead4 expression in scramble or Tead4 shRNAs-expressing
B6-S3Y118F mESCs. (b) qRT-PCR expression of Cdx2 expression in scramble or
Tead4 shRNAs-expressing B6-S3Y118F mESCs treated with GCSF for 24 hours. (c)
qRT-PCR analysis of Gata3 expression in scramble or Tead4 shRNAs-expressing B6-
S3Y118F mESCs treated with GCSF for 24 hours. (d) Phase contrast images of
scramble or Tead4 shRNAs-expressing B6-S3Y118F mESCs cultured in the
presence of GCSF for 7 days.
d
a
b

72
Among the Stat3 direct targets I identified in microarray analysis (GSE38719), I
found that Tfap2c is a potential candidate linking Stat3 to TE differentiation. Tfap2c is
required for embryonic development and proliferation of extraembryonic TE cells (75).
Forced expression of Tfap2c promotes TE differentiation in ESC through associating with
Cdx2 protein or its loci (76, 77). Tfap2c expression was continuously upregulated in B6-
S3Y118F mESCs treated with GCSF whereas LIF-induced expression of Tfap2cplateaused
after 2 hours post-stimulation (Figure 39a). To further validate that GCSF induces an
increased Tfap2c expression as compared to LIF, I examined Tfap2c expression by
sequentially giving GCSF to B6-S3Y118F ESCs maintained in LIF. Similarly, GCSF can break
the limit of basal Tfap2c expression activated by LIF and continuously induce its
expression (Figure 39b).

73
Fig. 39

I next used a de novo protein synthesis inhibitor, cycloheximide (CHX), to verify
whether Tfap2c is directly regulated by Stat3. I observed that GCSF induces Tfap2c
within one hour of application and Tfap2c expression level was similar with or without
CHX treatment (Figure 39c). Additionally, a previous study has indicated an association
between Stat3 and Tfap2c loci (56). These evidences suggest that Tfap2c is a direct
downstream target of Stat3 and that Stat3 hyperactivation can sustain Tfap2c induction.
To determine whether Stat3 hyperactivation induces TE differentiation through
upregulation of Tfap2c, I knocked down Tfap2c expression by shRNA (Figure 40a).

Figure 39: Tfap2c can be directly regulated by Stat3 and the expression can
continuously be induced when Stat3 is hyperactivated
(a) qPCR analysis of Tfap2c expression in B6-S3Y118F mESCs treated with GCSF or LIF for
the indicated times. B6-S3Y118F mESCs were starved overnight in serum free medium prior
to the treatments. (b) qPCR analysis of Tfap2c expression in B6-S3Y118F mESCs treated
with GCSF for the indicated times. B6-S3Y118F mESCs were maintained in the LIF condition
prior to GCSF treatment. (c) qPCR analysis of Tfap2c expression in B6-S3Y118F mESCs
treated with GCSF or GCSF plus cycloheximide (CHX) for 1 hour. B6-S3Y118F mESCs were
starved overnight in serum free medium prior to the treatment.
a
b
c
b
b
b

74
Tfap2c knockdown reduced GCSF-mediated Cdx2 induction (Figure 40b) and prevented
TE differentiation (Figure 40c, d). My results suggest that Tfap2c and Cdx2 are the two
key factors that mediate TE differentiation of mESCs induced by Stat3 hyperactivation.

Fig. 40

a
b
b
b

a
b
c
b
d
b
Figure 40: Knocking down the endogenous expression of Tfap2c can reduce the
induction level of Cdx2 by GCSF and rescue the differentiation phenotype.
(a) qRT-PCR analysis of Tfap2c expression in scramble or Tfap2c shRNAs-expressing
B6-S3Y118F mESCs. (b) qRT-PCR analysis of Cdx2 expression in scramble or Tfap2c
shRNAs-expressing B6-S3Y118F mESCs treated with GCSF for 24 hours. (c) Phase
contrast images of scramble or Tfap2c shRNAs-expressing B6-S3Y118F mESCs
cultured in the presence of GCSF for 7 days. (d) AP staining of scramble or Tfap2c
shRNAs-expressing B6-S3Y118F mESCs cultured in the presence of GCSF for 7 days.

75
h. Blocking MEK/ERK signaling failed to prevent Stat3 induced trophectoderm
differentiation
In previous studies, the MAPK pathway has been shown to play a pivotal role in
regulating the emergence of extraembryonic lineage (78, 79). B6-S3Y118F ESCs treated
with PD0325901 to block MAPK signaling showed no difference in the first 24 hours of
GCSF treatment when comparing to the control without PD0325901. However, reduced
numbers of differentiated cells were observed when treated with PD0325901 after two
days (Figure 41). This suggests that Stat3 hyperactivation-induced TE differentiation
occurs in a MAPK pathway-independent manner and the MAPK pathway signaling is
required for viability of the differentiated cells.
Fig 41

Figure 41: MAPK pathway is required for the viability of the differentiated
induced by GCSF but is not essential for TE differentiation mediated by Stat3
hyperactivation
Phase contrast images of B6-S3Y118F mESCs in the condition of GCSF with or
without PD for the indicated time

76
i. Identification of potential candidates that can block Stat3 induced TE
differentiation by chemical compound library screening
In order to understand how Stat3 hyperactivation induces mESC apoptosis and
differentiation, I performed a small molecular library screening to identify the
compounds that prevent the crisis. B6-S3Y118F mESCs were plated onto eight 96-well
plates at a density of 1000 cells/well. A total of 800 protein kinase inhibitors were
screened. Thirteen inhibitors were identified to block differentiation (Figure 42a) and
one inhibitor to block apoptosis (Figure 42b). ROCKi has no effect on regulating the fate
of B6-S3Y118F mESCs but it blocks apoptosis induced by Stat3 hyperactivation (Figure
42b). More follow up validation experiments need to be performed to confirm the
screening results.

77
Fig 42

Figure 42: Screening protein kinase
inhibitors that block the TE
differentiation and apoptosis
induced by Stat3 hyperactivation
(a) Phase contrast image of B6-
S3Y118F mESCs treated with the
inhibitors which block TE
differentiation (b) Phase contrast image
of B6-S3118F mESCs with or without
ROCKi in the indicated concentration of
GCSF

a
b
a
b
a

78
I conducted additional tests with IKK2 inhibitor to ascertain whether the NF B
pathway is related to the TE differentiation. IKK2i (IKK2 inihibitor VI, EMD Millipore)
efficiently blocked TE differentiation induced by GCSF in B6-S3Y118F mESCs (Figure 43a).
However, western blot result suggests that GCSF does not activate NF B signaling in B6-
S3Y118F mESCs (Figure 43b). Furthermore, the IKK2 inhibitor can block Stat3
phosphorylation (Figure 43c), suggesting that the inhibitor either has off-target activity
on JAK1i or blocks Stat3 activity in a non-canonical manner.
Fig 43

a
b
a
b
a
b
a
c
b
a
b
a
Figure 43: IKK2 inhibitor prevents ESC from GCSF induced TE differentiation by
blocking Stat3 phosphorylation
(a) Phase contrast image of B6-S3Y118F mESCs in the condition of LIF or GCSF with
or without IKK2 inhibitor (b) Western blot analysis of phospho-p65 of B6-S3Y118F
mESCs in the indicated conditions (c) Western blot analysis of Stat3 phosphor-705
and 727 of B6-S3Y118F mESCs in the indicated conditions

79
j. GCSF-induced Stat3 hyperactivation promotes apoptosis in ESCs
I found that Stat3 hyperactivation induces not only TE differentiation but apoptosis as
well. Annexin V/PI staining revealed that the proportion of apoptotic B6-S3Y118F ESCs
was greater when treated with GCSF for 24 hours than when treated with LIF (27% vs.
7%) (Figure 44a, b). Inducing Stat3 hyperactivation by using 50 ng/ml GCSF resulted in
46C-S3Y118F cells exhibiting a greater proportion of apoptotic cells than did B6-S3Y118F
mESCs (Figure 44c). Addition of JAK1i can prevent GCSF-induced TE differentiation as
well as apoptosis (Figure 44d).
Fig. 44
a

Figure 44: Apoptosis can be induced by Stat3 hyperactivation in
ESC and can be rescued by JAK1i
(a) Annexin V/PI analysis of B6-S3Y118F in the presence of LIF (left
panel) and after GCSF treatment for 24 hours (right panel). X-axis
represents the intensity of Annexin V staining. Y-axis indicates the
intensity of PI staining.

80
b c

d

k. Enhancing Stat3 signaling induces tumor cells apoptosis
We next confirmed this apoptosis phenotype by utilizing two Stat3 mutant cancer cell
lines, HT29 and DLD1, which are reported to express constitutively active Stat3. We
transduced them to express grGP130-Y118F to further induce Stat3 activation (Figure
45a, b). We saw reduced cell proliferation and increased cell death in both cell lines
(b) Images of phase contrast and Immunpostaining of Annexin V of B6-S3Y118F in the
presence of LIF (upper panel) and after GCSF treatment for 24 hours (lower panel). (c) Phase
contrast images of 46C- and B6-S3Y118F mESCs 2 days after culturing in the condition LIF or
GCSF. (d) Phase contrast images of 46C-S3Y118F 2 days after culturing in the condition of LIF,
GCSF or GCSF plus JAK1i.

81
after inducing Stat3 activation with GCSF (Figure 45c); in contrast, wild type lines were
indifferent to GCSF (Figure 45d). This result is consistent what I saw earlier in mESCs.
This part of work was conducted by previous lab member, Eric Schulze.
Fig. 45
a b

c d

Figure 45: Enhancing Stat3 signaling induce apoptosis in HT29 and DLD1 cancer
lines
(a) Western blot analysis of phospho-Stat3 S705 expression level in DLD1-Y118F cells with
or without GCSF. (b) Western blot analysis of phospho-Stat3 S705 expression level in HT29-
Y118F cells with or without GCSF. (c) Phase contrast images of DLD1-Y118F and HT29-
Y118F cells with or without GCSF. (d) Phase contrast images of DLD1-WT and HT29-WT
cells with or without GCSF.

82
l. The mechanism behind the apoptosis induced by Stat3
It is still unclear how Stat3 hyperactivation triggers apoptosis. Excess Stat3 activity
may upregulate expression of P21waf1 and initiate cell cycle arrest (80). Therefore, I
checked the expression pattern of P21waf1 after inducing Stat3 hyperactivation. I found
P21waf1 expression did not increase until 48 hours after the start of GCSF treatment
(Figure 46a). Since the apoptosis phenotype emerged within 24 hours after the addition
of GCSF, I exclude P21waf1 as the cause of the apoptotic phenotype. In the drug
screening result, I found that an inhibitor of Rho-associated, coiled-coil containing
protein kinase (ROCK) has no effect on regulating the fate of B6-S3Y118F mESCs but it
significantly blocks apoptosis induced by Stat3 hyperactivation (Figure 46b). ROCK
inhibitor has been shown to increase clonal efficiency and improve survival in human
embryonic stem cells (hESCs) (81). It is unclear how the Rho/ROCK pathway is connected
to the low survival rate in dissociated hESCs. The link between the Rho/ROCK pathway
and Stat3-induced apoptosis might signify that the cell death-susceptibility of hESCs is
connected with Stat3 signaling.

83
Fig. 46
a b

B6-S3Y118F
P21 expression
Figure 46: Apoptosis induced by Stat3 hyperactivation
(a) qPCR analysis of p21 expression in B6-S3Y118F mESCs maintained in LIF or
treated with GCSF for the indicated times (b) Phase contrast image of B6-S3118F
mESCs with or without ROCKi in the indicated concentration of GCSF

84
III. Discussion:
In the second project, I describe a previously unknown function for Stat3 in mESCs,
specifically, that Stat3's self-renewal function is context and dose-dependent. I also
address a long-standing technical and practical question regarding mESC culturing: how
to promote feeder-free self-renewal in non-permissive (derivation refractory) mESC
strains through modulation of Stat3 activity. LIF-induced Stat3 activity is not sufficient to
maintain non-129 ESC self-renewal under feeder-free conditions, and enhancing Stat3
activity through overexpression of a Stat3 transgene or a gp130 chimeric receptor is
necessary and sufficient to obviate the need for a feeder layer and maintain an
undifferentiated ESC state. However, elevating Stat3 activity over a certain threshold
will switch Stat3’s self-renewal promoting effect into one that induces ESC
differentiation toward the TE lineage.
How does Stat3 play such contradictory roles in the same pluripotent stem cells?
Upon stimulation by cytokines, Stat3 is first recruited to the receptors via its Src-
homology-2 (SH2) domain, and then phosphorylated on tyrosine 705, leading to
dimerization and translocation to the nucleus, where it binds specific DNA sequences
and activates target gene transcription. I hypothesize that Stat3 recruits distinct co-
activators and activates distinct gene programs depending on its activation level. Stat3
not only regulates the expression of pluripotency-related genes such as Tfcp2l1, Gbx2,
Klf4, Klf5, Pim1, Pim3, Pramel7, and c-Myc (42, 45-51), but also induces the expression
of factors that promote TE differentiation. The outcome of Stat3 activation in mESCs,

85
whether to promote self-renewal or induce differentiation, is likely determined by the
balance of expression level between pluripotency- and differentiation-related genes
both induced by Stat3 (Figure 47). When Stat3 is regulated by LIF, Tfap2c expression can
be temporarily induced but remains constantly low (Figure 47). The pluripotency factors
parallelly induced by Stat3 then promote self-renewal of mESCs. However, when Stat3 is
highly and sustainably activated by GCSF in B6-S3Y118F mESCs, the expression of Tfap2c
can be continuously induced and consequently upregulates Cdx2 expression. Both
Tfap2c and Cdx2 later override the pluripotency established by Stat3 and induce mESC
differentiation toward the TE lineage (76, 77).
Fig. 47

Figure 47: Model of mESCs fate regulated by Stat3 activation level.

86
Using the mESC model, I demonstrate that a TE lineage factor, Tfap2c, can be
directly regulated by LIF/Stat3 signaling. Several pieces of evidence indicate that
LIF/Stat3 signaling is involved in TE development during implantation. Blastocyst TE
gives rise to the mammalian placenta, which plays a significant role during implantation
and gestation. In vivo, expression of LIF is essential for the mammalian endometrium
during blastocyst implantation (9), and LIF receptor-null mutant mice show abnormal
placental architecture and succumb to prenatal death (10). In addition, Stat3 activity is
reported to be necessary for trophoblast differentiation during the implantation process
(82-84). Additionally, enhanced and prolonged Stat3 activity resulting from a lack of its
negative regulator, SOCS3, promotes differentiation of trophoblast stem cells (85).
These physiological results echo my finding that TE differentiation can be regulated by
the LIF/Stat3 pathway, implying that Tfap2c might also be a key component in LIF/Stat3-
mediated TE and placenta development.
The degree of Stat3 activation by extrinsic factors might be different among ESCs
from different species, and these differences might account for the failure to establish
authentic ESCs from non-rodent species under the LIF condition. This hypothesis is
supported by my finding that artificially elevated Stat3 signaling supports self-renewal of
ESCs derived from the derivation-refractory B6 mouse strain as well the rat (86). By
studying the role of LIF/Stat3 signaling in isolation and in context within mESCs, we
might gain insights into why LIF is not sufficient to promote self-renewal in ESCs from
other species, and ascertain whether a universal latent mammalian self-renewal
mechanism exists. I anticipate that the differentiation-related genes upregulated by

87
Stat3 must also be finely controlled to ensure ESC self-renewal and that controlled
modulation of Stat3 function can facilitate the establishment of authentic non-rodent
ESCs. The knowledge we gained from this study might eventually be applied for
optimizing condition for the derivation and expansion of ESCs/iPSCs from non-rodent
species, such as rabbits, chickens, pigs, and cows, in which authentic ESCs have not yet
been established.

88
Chapter IV
Summary and Perspectives
I. Summary
Since their first isolation in 1981 by Sir Martin J. Evans and Matthew Kaufman and
independently by Gail R. Martin, mESCs have contributed greatly to the knowledge of
stem cell biology and regenerative medicine (1, 2). Both groups reported the
requirement of feeders for mESCs to be derived and maintained in vitro. The major
contribution of feeders in supporting mESC self-renewal is by producing LIF. LIF
promotes mESC self-renewal through activation of Stat3. In the past four years, I have
been working on understanding the molecular mechanism underlying ESC fate
regulation mediated by Stat3. I identified several novel Stat3 targets in mESCs. Among
these targets, Gbx2 and En2 were found to have a novel function in promoting mESC
self-renewal. This finding broadens our knowledge of how LIF/Stat3 simultaneously
induces the expression of multiple pluripotency factors and mediates mESC pluripotency
using redundant mechanisms. In addition to pluripotency factors, a key differentiation
factor of TE lineage, Tfap2c, was identified to be directly regulated by LIF/Stat3. Tfap2c
expression induced by Stat3 hyperactivation is capable of inducing mESC differentiation
towards the TE lineage even in the presence of self-renewal cues (76, 77). Moreover, in
addition to TE differentiation, Stat3 hyperactivation also triggers apoptosis in mESCs.
These findings implicate that Stat3 promotes self-renewal or differentiation in a dose-

89
dependent manner. My work has generated novel concepts about how LIF/Stat3
promotes self-renewal and how Stat3 activity level is crucial in determining mESC fates.
Here, I will provide my perspectives on the underlying mechanisms.

II. Potential mechanism for the roles of Gbx2 and En2 in ESC self-
renewal

a. Gbx2 may act as a repressor of Otx2 to block ESC differentiation
Gbx2 is a transcription factor involved in regulating central neuron system development
and mid/hind brain patterning (87). ChIP-seq analysis performed in the human PC3 cell
line suggests that more than 50% of the total 286 Gbx2 target genes are related to
neural development (88). Among these 286 candidates, I did not identify any genes
related to ESC self-renewal or pluripotency. This suggests that the Gbx2-associated gene
expression profile varies in the cellular context of epigenetics. In prostate cancer cells,
Gbx2 expression was found to correlate with the expression of IL-6 (89), which belongs
to the same ligand family as LIF and triggers Stat3 activity via gp130, thereby acting as a
cue for self-renewal. In the same study, Gbx2 protein was shown to physically bind to
the promoter of the IL-6 gene. This suggests that Gbx2 may promote ESC self-renewal by
acting as a transcription activator to induce the expression of the self-renewal cue, IL-6.
However, mESCs do not express IL-6 receptors and have no response to IL-6 ligand (18).
Furthermore, in my study, overexpressing Gbx2 promotes Stat3 knockout mESC self-

90
renewal, implying that the self-renewal phenotype promoted by Gbx2 is not through
enhancing the activity of Stat3 mediated by IL-6.
On the other hand, as a typical homeobox gene, Gbx2 can also act as a
transcription repressor. During brain development, several findings indicate that a
reciprocal repression between Gbx2 and Otx2 establishes the mid-hind brain boundary
(90, 91). Interestingly, besides brain development, Otx2 also plays an important role in
ESC differentiation. Otx2 is expressed in pre-implantation embryos. Knocking down the
endogenous expression of Otx2 prevents mESC differentiation and retains mESC self-
renewal (92). I performed a similar experiment to confirm this result by knocking down
the expression of Otx2. Consistent with the finding, 46C mESCs transduced to express
Otx2 shRNA can be maintained undifferentiated in the absence of LIF (Fig. 48a).
Moreover, in these Otx2 shRNA transfectants, I observed an induction of Gbx2 but not
Klf4 (Fig. 48b). The result echoes the reciprocal antagonism between Gbx2 and Otx2
during brain development.

91
Fig. 48
a

b

Together, it seems that Gbx2 and Otx2 may repress each other to maintain a
balance between self-renewal and differentiation. A chicken-or-the-egg question arises:
which one comes first? Does the Otx2 repression promote ESC self-renewal by inducing
the expression of Gbx2? Or does the Gbx2 induction prevent differentiation cues by
suppressing Otx2 transcription? The latter hypothesis is more plausible because LIF can
induce the expression of Gbx2 even though Otx2 is generally expressed in mESCs (92).
To see if Otx2 inhibition is required for Gbx2 mediated ESC self-renewal, we can deplete
or knockdown Groucho/Tle, which is the corepressor that Gbx2 interacts with to
suppress Otx2 transcription (93). We can also ectopically express truncated Gbx2 lacking
the transcriptional activation domain to confirm if the transcription activator function is
Figure 48: Knocking down the endogenous
expression of Otx2 promotes 46C mESCs
self-renewal
(a) Phase contrast image of 46C mESCs
transfected with control or Otx2 shRNA in the
absence of LIF for 5 days (b) qPCR analysis of
the expression of Otx2, Bbx2, En2, and Klf4.
First bar: control shRNA, 2
nd
-5
th
bar: Otx2
shRNA #1-4.

Control
shRNA
Otx2 shRNA
#3
Otx2 shRNA
#4

92
essential for Gbx2 to promote ESC self-renewal. This will clarify whether Gbx2 promotes
self-renewal by repressing Otx2 or by activating other pluripotency factors.

b. En2 may enhance ESC self-renewal through its oncogenic properties
En2 is also a homeobox gene and contains both transcriptional activation and repression
domains (94, 95). Molecular studies indicate that En2 protein mainly functions as a
transcription repressor, which is mediated by engrailed homology-1 (EH1) and EH2
domains. Both domains bind to Groucho via its WD40 domain (96, 97). Therefore, like
Gbx2, En2 may also block the expression of Otx2 to promote ESC self-renewal. The
regulation on Otx2 should be verified by using an inducible system to modulate the
expression En2. Also, we can delete or truncate EH1 or EH5 domain to confirm whether
En2 mediates ESC self-renewal as a transcription repressor.
So far, there is no ChIP analysis of En2 conducted in the context of a mammalian
or ESC genome. Therefore, it is still unclear whether En2 promotes ESC self-renewal as
an activator or a repressor. However, in Xenopus axons, En2 attracts the growth cone of
nasal axons by inducing protein phosphorylation for translation initiation and ultimately
increasing protein synthesis (98). Moreover, En2 expression has been shown to highly
correlate with the incidence of human prostate, breast, ovary, and bladder cancers (99-
101). Consistent with what I found earlier, ectopic expression of En2 prevents neuron
precursors from differentiating (102). Thus, En2 may act similarly to other pluripotency
factors involved in tumorigenesis and promote ESC self-renewal by triggering protein
synthesis to enhance cellular proliferation or by inducing the expression of histone

93
modifiers to suppress the genes involved in ESC differentiation (103). Interestingly, En2
is one of the rare transcription factors that can be secreted from transportation vesicles
and internalized by cells (104, 105). This unique property might explain how En2
bypasses Stat3 to promote wild-type or Stat3 null mESC self-renewal. To confirm this,
we would have to verify if En2 can be secreted into the medium. A reliable antibody is
required to examine the existence of En2 protein by western blot or ELISA from the
culture medium of En2 overexpressing mESCs. It is also possible that secreted En2 may
be posttranscriptionally modified or truncated for cellular exportation and
internalization. Therefore, we can test whether the conditioned medium collected from
En2 overexpressing mESCs can promote or enhance wild-type mESC self-renewal. The
possible applications would be significant if we can sustain ESC self-renewal just by
adding recombinant En2 protein to the media.

c. Gbx2 is a major downstream effector of LIF/Stat3 signaling, while En2 may be a
-Catenin pathways
In terms of promoting self-renewal, the major difference between Gbx2 and En2 is the
ability to reprogramming EpiSCs to a naï ve pluripotent state. Overexpressing Gbx2 alone
is sufficient to revert EpiSCs back to ESC-like cells when giving ESC culture condition,
LIF/2i. Therefore, Gbx2 may actively lead to a process of erasing or redistributing the
epigenetic marks that favor ESCs. It is also possible that Gbx2 is capable of repressing
the factors that are important for EpiSC differentiation, such as Otx2 (92). This may

94
explain why reprogramming via Gbx2 still requires LIF or LIF/2i in the culture medium to
induce other pluripotency factors to finish the reprogramming process.
In a recent report in Science (106), the authors exploited a computational
approach to derive a set of interaction combinations to explain ESC behavior under
culture conditions including LIF, CHIR, and PD. Within the meta-model of the
pluripotency network, Gbx2 is considered as a major contributor under LIF/Stat3
signaling based on its sufficiency to promote ESC self-renewal and its ability to
reprogram EpiSCs back to a naï ve state. LIF/Stat3 and -catenin have been shown to be
two major but independent pathways to promote ESC self-renewal (54). In the meta-
model, Tcfcp2l1 is the only common target between the LIF/Stat3 and -catenin
signaling pathways (46, 47). In Xenopus, the En2 promoter has been indicated to
contain three LEF/TCF binding motifs (107). Thus one possibility is that, in addition to
LIF/Stat3, En2 can also be regulated by -catenin. To test this I set up an experiment to
determine if En2 expression can be induce by CHIR. However, upon treatment with CHIR
for two hours, I did not observe an induction of the En2 transcript, suggesting that the
regulation requires other co-factors that do not exist in the mESC context or that En2 is
indirectly regulated by -catenin and longer duration of CHIR treatment is necessary to
observe the induction.

95
III. Potential mechanisms for ESC apoptosis and differentiation induced
by Stat3 hyperactivation

a. The expression level and sustainability of Stat3 activity determine whether ESCs
will differentiate into the TE lineage
In addition to pluripotency factors, LIF/Stat3 also simultaneously induces the expression
of differentiation factors. The expression of Tfap2c, a TE differentiation factor, can be
induced via Stat3 by LIF or GCSF in B6 or B6-S3Y118F mESCs. Tfap2c induction reaches a
plateau level shortly after stimulation and then reduces to a steady expression level.
Tfap2c can be continuously induced beyond the plateau level to promote ESC
differentiation into the TE lineage only when Stat3 is highly and sustainably activated. I
have shown that the Hippo/TEAD4 and -catenin pathways are not required for the
induction of Tfap2c. This implies that either Stat3 has a novel function in regulating
Tfap2c alone or it can compensate for the insufficient amount of cofactors and drive
transcription alone.
Different Stat3 activation levels can lead to different ESC fates. Stat3 may not
only bind to factors that promote ESC pluripotency but also bind to factors that induce
the expression of differentiation-inducing genes. I hypothesize that Stat3 may have
higher binding affinity to cofactors that promote self-renewal than cofactors that induce
differentiation. When Stat3 is normally activated, due to differential binding affinities,
Stat3 will favor to bind to cofactors that promote self-renewal than to cofactors that

96
induce differentiation. On the other hand, when Stat3 is hyperactivated, more activated
Stat3 translocates into nucleus than when it is normally activated. Stat3 will then first
occupy cofactors that promote self-renewal and the excess Stat3 then starts to bind to a
different set of cofactors that induce differentiation. The expression of the
differentiation-inducing genes will then override the self-renewal promoting genes and
lead ESCs to differentiation. This hypothesis explains why, in mESCs, Stat3 has to be
hyperactivated to induce TE differentiation. Besides binding to cofactors, it is also
possible that Stat3 itself has higher binding affinity toward to the promoters of the
pluripotency genes than to those of differentiation genes. In literatures (9, 10), LIF is
critical to induce TE maturation in endometrium, this may be because endometrium
cells do not express the Stat3 self-renewal cofactors or the promoters of pluripotency
genes regulated by Stat3 are epigenetically silenced. Therefore, LIF-activated Stat3 can
induce TE differentiation via interacting with cofactors that promote differentiation or
binding to the promoters of differentiation genes.

b. Stat3 targets, Gbx2 and Klf4, are downregulated promptly after inducing Stat3
hyperactivation
Interestingly, the Stat3 pluripotency target, Gbx2, whose expression can be induced two
hours after GCSF stimulation, is downregulated at the eight hours time point to a level
that is even lower than the steady expression level seen with LIF treatment (Figure 49).
This suggests that a quick repression of Gbx2 occurs between two and eight hours after
Stat3 is hyperactivated. Dominant negative effect caused by excess Stat3 activity is

97
excluded because Gbx2, En2, and Socs3 expression can still be induced 0.5-2 hr after
GCSF treatment. The repression may be related to Tfap2c, which has been shown to
override the pluripotency of ESCs even in the presence of LIF (77). To test this, we have
to determine if Tfap2c protein expression correlates with the repression of Gbx2
transcription by using an inducible system to modulate the expression Tfap2c.

Fig 49

It is also possible that the repression is not related to Tfap2c and is triggered by
other repressors. Stat3 hyperactivation first triggers the repression of pluripotency
factors to drive ESC exit from pluripotency and then the induction of Tfap2c directs the
ESCs to differentiate towards the TE lineage. To identify the factors involved in the
repression mechanism, a Stat3 pull down assay should be performed to compare the
binding partners in B6-S3Y118F mESCs cultured with LIF and GCSF. MALDI-TOF mass
0
1
2
3
4
5
6
7
CN G2 G8 G12 G18 G24
Figure 49: Expression pattern of
Gbx2 after GCSF treatment in B6-
S3Y118F mESCs
qPCR analysis of the expression
pattern of Gbx2 in B6-S3Y118F mESCs
in the indicated times after GCSF
treatment. CN: control; G#: GCSF
treated for the indicated hours. Y-axis
represents the relative expression level
normalized to CN group

98
spectrometry analysis will next be performed after identifying unique protein bands on
SDS-PAGE.
Surprisingly, unlike Gbx2 the expression of the other Stat3 target, Klf4, shows a
prompt reduction after Stat3 hyperactivation (Fig. 50). Opposite to Gbx2, Klf4
transcription is suppressed by 50% two hours after applying GCSF, while the expression
can be upregulated with LIF. This suggests that hyperactivated Stat3 may selectively
block its specific targets instead of activating them. The next question is whether this
prompt suppression of Klf4 is different from the repression of Gbx2. To verify that, we
have to make sure that Klf4 is not suppressed within the first two hours after the
addition of GCSF. Real-time PCR for Klf4 expression should be performed after giving
GCSF for 30 minutes and one hour. However, based on the dissimilarity in gene
repression between Gbx2 and Klf4, the repression mechanisms seem to act differently.

Fig 50

0
0.5
1
1.5
2
2.5
CN 2hr 8hr 24hr
LIF
Figure 50: Expression
pattern of Klf4 after
treatment with LIF or GCSF
in B6-S3Y118F mESCs
qPCR analysis of the expression
pattern of Klf4 in B6-S3Y118F
mESCs in the indicated time
after LIF or GCSF treatment. Y-
axis represents the relative
expression level normalized to
the no-treatment control (CN)
group.

99

I believe that Gbx2 and Klf4 are not the only two genes suppressed by the Stat3
hyperactivation-induced repression mechanisms. Studies have shown that knocking
down Gbx2 or Klf4 expression does not impair LIF mediated ESC self-renewal (44, 45).
To induce ESCs to exit pluripotency, other Stat3 targets must also be inhibited.
Therefore, it would be interesting to identify genes whose expression can be induced by
normal Stat3 activity but are downregulated when Stat3 is hyperactivated. By
comparing and categorizing the timing of the repression of applicable genes, we may
find some common properties, such as a Stat3 binding sequence or Stat3 binding
partners, among the genes which can be suppressed either immediately or after two
hours. Charactering these attributes further can elucidate the repression mechanisms.

c. Kinase inhibitor screening assay indicates that NF B pathways may also
participate in Stat3 induced TE differentiation.
To further decipher how Stat3 hyperactivation induces TE differentiation of ESCs, I
sought to find out what other signaling pathways might also be involved. A chemical
compound-screening assay was performed to identify the potential kinase inhibitors
that can block TE differentiation. Among the compounds I found that can prevent TE
differentiation of ESCs induced by hyperactivation of Stat3, IKK-2 seems to be an
interesting candidate. IKK-2 only participates in the canonical NF B pathway. Even
though in a different cellular context, NF B has been shown to be a critical activator for
the transcription of Cdx2 and Gata3 (108, 109). From a microarray analysis study(110), it

100
was shown that Stat3 and NF B control an overlapping group of genes in a human
mammary epithelial cell line. Therefore, NF B may cooperate with Stat3 to activate the
expression of Tfap2c or other differentiation factors. It is also possible that Tfap2c
requires NF B as a cofactor to induce the expression of Cdx2. However, from my
western blot results, the IKK-2 inhibitor seems to block TE differentiation by inhibiting
Stat3 phosphorylation (Figure 43). Different IKK-2 inhibitors should be tested to confirm
whether the inhibitor is blocking JAK1 activity due to an off-target effect. It is also
possible that IKK-2 participates in activation of Stat3 with an unknown mechanism.

d. GCSF triggers B6-S3Y118F mESCs apoptosis in a JAK/ROCK1 dependent manner

Besides differentiation, Stat3 hyperactivation also triggers ESC apoptosis. Reducing the
endogenous expression of Tfap2c or Cdx2 prevents ESCs from differentiating into TE,
but not from apoptosis (Figure 51). On the other hand, blocking JAK1 activity rescues
ESCs from both TE differentiation and apoptosis, suggesting that apoptosis is mediated
through either Stat3 or JAK1. A similar apoptotic phenotype was also observed in the
two cancer cell lines when Stat3 is hyperactivated via the chimeric gp130 receptor.
Therefore, understanding the mechanism of how Stat3 triggers apoptosis instead of
survival may provide an alternative approach for cancer therapy.

101

Fig. 51

In the compound library-screening assay, a ROCK1 inhibitor was found not to
affect ESC fate but instead to block ESCs from apoptosis. The ROCK1 inhibitor was
indicated to improve survival of hESCs upon single-cell dissociation (111). Although
ROCK1 is a substrate of Caspase-3 and triggers one of the apoptotic hallmarks, blebbing
(112, 113), there is no direct evidence for how the ROCK1 inhibitor blocks apoptosis. The
screening result implies that there is interplay between ROCK1 and JAK/Stat3.
Interestingly, ROCK1 can regulate leptin mediated Stat3 activation (114). Leptin belongs
to the IL-6 family and binding to its receptor results in the activation of JAK2, followed
by the phosphorylation of Stat3 (115). A similar finding has also been reported that JAK
Figure 51: A schematic summary of how Stat3
hyperactivation triggers ESC apoptosis and TE
differentiation

102
signaling can activate ROCK1, which later enhances Stat3 phosphorylation (116). Since
the ROCK1 inhibitor only blocks apoptosis without affecting ESC fate, it is unlikely that
the inhibitor attenuates Stat3 activity. Therefore, the apoptosis in B6-S3Y118F mESCs
may be triggered by excess ROCK1 activity induced by JAK1. This explains why both JAK1
and ROCK1 inhibitors can rescue the apoptosis phenotype.

IV. Improving and optimizing the culture conditions for ESCs derived
from non-responsive strains or species
Following the first ESC derivation from mice in 1981, multiple ESC or ESC-like lines were
derived from several mammalian species. However, so far only murine and rat ESCs can
be cultured in the naï ve pluripotent state and possess the ability to generate germline-
transmitting chimeras. Current knowledge is limited not only to isolate and maintain
ESCs from the species other than mice and rats, but also to ESCs from different strains of
mice.
Mouse ESCs derived from refractory mouse strains, such as B6, are maintained
with both LIF and feeders, while mESCs from 129 strains can grow in LIF without
feeders. 46C mESCs were shown to have higher Stat3 expression and activity when
stimulated by LIF than B6 mESCs. B6 mESCs can be liberated from the requirement of
feeders when Stat3 signaling is enhanced. However, in addition to Stat3, it is possible
that feeders may provide other unknown factors to support ESC self-renewal. Indeed,
giving other self-renewal cues, such CHIR or MAPK inhibitors, can also maintain B6 ESCs

103
in a feeder-free condition (117). More statistical analyses of Stat3 expression among
mESCs from refractory or non-refractory strains are required to claim that the
requirement of feeders in refractory mESCs is due to the poor Stat3 activity.
Furthermore, to understand why mESCs from certain strains have lower derivation rates
and germline transmission efficiencies than from the 129 strain, a genome wide gene
expression study should be conducted by comparing pluripotency factors among mESCs
from different mouse strains.
LIF/Stat3 signaling in mESCs can induce the expression of pluripotency-related
genes as well as differentiation-inducing genes such as Tfap2c and Cdx2. Interestingly,
rat ESCs cultured in LIF only differentiate mainly into Cdx2 positive TE lineage cells [(118)
and unpublished results]. Also, knocking down the expression of Cdx2 causes the colony
morphology of mESCs to be more compact and spherical (unpublished data). I speculate
that rat ESCs could be maintained by LIF if the induction of differentiation-related genes
by LIF/Stat3 signaling were suppressed. More differentiation genes can be identified by
performing expression microarray or RNA-SEQ analysis to compare the expression
profile difference between LIF and GCSF treatment conditions in ESCs cotransfected
with Stat3 and the chimeric receptor gp130. Ultimately, LIF/Stat3 signaling may also be
sufficient to maintain the self-renewal of ESCs from other species if the induction of
differentiation-related genes were specifically blocked.

104
V. Conclusions
In conclusion, my results suggest that LIF/Stat3 signaling maintains ESC pluripotency in a
redundant mechanism. Stat3 regulates the expression of multiple pluripotency factors,
such as Klf4, Gbx2, En2, etc., which promote ESC self-renewal by facilitating cell cycle
progression, inhibiting differentiation, or controlling epigenetic regulation. However,
LIF/Stat3 also induces detrimental effects for ESC self-renewal. Besides the pluripotency
factors, LIF/Stat3 also regulates differentiation factors, such as Tfap2c and Cdx2, to
prime ESCs for cell fate decisions. Moreover, apoptosis can also be triggered by ROCK1
via LIF/GP130R-induced JAK activation. These distinct roles of Stat3 could be explained
by the different levels of activated Stat3 in the cell. These different levels could generate
different outcomes by differential affinity of Stat3 to its pluripotent cofactors or to the
promoter of its self-renewal genes.
My work not only clarifies how LIF/Stat3 promotes ESC self-renewal, but also
reveals potential causes for why enhancing LIF/Stat3 signaling is not sufficient to derive
ESCs from different species because enhancing LIF/Stat3 may have the unintended
consequence of stimulating differentiation and/or apoptosis. Although the models
proposed are consistent with the available data, in-depth systematic analyses await to
further advance our knowledge of LIF/Stat3 in ESC biology. I envision that future studies
will uncover a different combination of pathway inhibitors accompanied by LIF to derive
or maintain ESCs from new species.

105
Materials and methods:
ESC and EpiSC culture
46C ESCs were routinely maintained on gelatin-coated plates in GMEM (Sigma)
containing 10% FBS (HyClone), 1 mM sodium pyruvate (Gibco), 100 µM non-essential
amino acids (Gibco), 2 mM GLUTAMAX (Gibco), 1 µM β-Mercaptoethanol, and 1000
U/ml LIF (Stemgent), a formulation hereinafter referred to as ‘mESC medium’. R1 and B6
ESCs were maintained in the same condition but with feeders. Stat3
−/−
ESCs were
maintained with feeders in serum-free N2B27 medium supplemented with 3 µM
CHIR99021 (Stemgent) and 1 µM PD0325901 (Stemgent). EpiSC derivation and culture
were conducted as described previously.
Microarray analysis
B6-Y118F ESCs starved in serum-free medium for 6 hours were treated with 30 ng/ml
GCSF (Preprotech) or 1000 U/ml LIF plus 1 µM PD0325901 for 1 hour. Total RNA was
then extracted by TRIzol (Invitrogen) and submitted to the University of California, Los
Angeles DNA Microarray core facility for analysis. The samples were hybridized to
GeneChip Mouse Gene 1.0 ST Array. The data analysis was performed by using the
Microarray Я US program provided by the USC Norris Medical Library.
Gene transfection
The identified genes from microarray analysis were cloned individually into the pSIN-
EF2-PURO lentiviral vector (Addgene). Stat3 and grgp130-Y118F transgenes were cloned

106
into the pCAG-IRES-IP expression vector or the pSIN-EF2 lentiviral vector. The details on
how to generate grgp130-Y118F transgene have been described in a previous study (35).
pCAG-Stat3 and pCAG-grgp130-Y118F plasmids were introduced into ESCs using
Lipofectamine LTX and Plus reagent (Invitrogen). The primers used for cloning each gene
are shown in supplementary material Table S2. 293T cells were cultured in DMEM
containing 10% FBS. mESCs 80–90% confluent on 10 cm plates were transfected with 5
µg pSPAX2 (Addgene), 3 µg pVSVG (Addgene) and 8 µg of each pSIN-EF2 construct via
LTX (Invitrogen). Six hours after transfection, the medium was replaced with mESC
culture medium. The viruses were then collected and filtered after incubation for 48
hours.
Self-renewal assay
For lentiviral infection, mESCs were seeded 105 cells per well in a 24-well plate, in 600 µl
of medium composed of 300 µl filtered virus medium and 300 µl mESC medium with 4
µg/ml Polybrene (Millipore). After 12–18 hours, the cells were trypsinized and replated
onto a six-well plate and cultured in the presence of 1 µg/ml puromycin for 48 hours.
After selection, transduced ESCs were plated 104 cells per 6 cm dish and subsequently
cultured in mESC medium without LIF. Undifferentiated ESC colonies were isolated and
propagated for further characterization.
Gbx2 reversion and chimera production
46C ESCs were electroporated with linearized pPyCAG-loxP-Gbx2-IRES-pac-STOP-loxP-
eGFP. After electroporation, the cells were plated onto 6 cm culture dishes, in serum

107
without LIF and selected in 1 µg/ml puromycin. Self-renewing colonies were isolated
and then maintained for at least 20 days in the absence of added LIF. The Gbx2-IRES-
pac-STOP cassette was then excised by transfecting the cells with a Cre recombinase
expression plasmid. After 48 hours, GFP-positive cells were enriched by FACS sorting.
Complete excision of the Gbx2 cassette was determined by the reacquisition of LIF
dependence. The GFP-positive colonies were expanded and then microinjected into
C57BL/6 blastocysts. Microinjected blastocysts were transferred to pseudo-pregnant
CD1 mice.
Immunostaining and AP staining
Immunostaining was performed according to a standard protocol. Primary antibodies
used were the following: Oct4 (C-10, Santa Cruz, 1 ∶200), SSEA-1 (480, Santa Cruz,
1 ∶200), Klf4 (R&D Systems, 1 ∶200) and Nanog (R&D Systems, 1 ∶200). Alexa Flour
fluorescent secondary antibodies (Invitrogen) were used at a 1 ∶2000 dilution. Nuclei
were visualized with DAPI. AP staining was performed with an alkaline phosphatase kit
(Sigma) according to the manufacturer's instructions.
Gene expression knockdown
shRNA-expressing plasmids were generated according to Addgene PLKO.1 protocol. The
target-specific shRNA sequences used in this study are as follows: Control shRNA: 5′-
AATTCTCCGAACGTGTCACGT-3′; Gbx2 shRNA#1: 5′-GGTTCGCTATTCGAAGTCATT-3′; Gbx2
shRNA#3: 5′-GAGAGCGATGTGGATTACA-3′. Cdx2 shRNA: GGACAGAAGATGAGTGGAATT;
Gata3 shRNA#1: GCCTGCGGACTCTACCATAA A; Gata3 shRNA#2:

108
ATTGCTGAACATTGCATATAA; Gata3 shRNA#3: CAGTTGTTTGATG CATTTAAA; Tead4
shRNA#1: CCGCCAAATCTATGACAAGTT ; Tead4 shRNA#2: GC TGAAACACTTACCCGAGAA;
Tead4 shRNA#3: CCCTCT CTGTGAGTACATGAT; Tfap2c shRNA:
AGCCGCTCTGCAAGTCTAATA. After lentiviral infection, the cells were incubated with 1
µg/ml puromycin, 10 µg/ml Blasticidin S deaminase, or 100 µg/ml hygromycin for 48–72
hours. The drug-resistant cells were then replated onto a six-well plate, 5000 cells/well,
and incubated for 5–7 days before AP staining.
Reprogramming
MEFs were transfected with the four factors using the piggyBac transposon system and
LTX. The MEFs expressing the four factors were selected in 1 µg/ml puromycin for 2
days and then treated with Gbx2 or EGFP virus-containing medium for 18 hours before
the medium was replaced. After incubation for 10 days in mESC medium supplemented
with LIF, the cells were stained for AP and the number of positive colonies was tallied.
EpiSCs expressing EGFP or Gbx2 following lentiviral transduction were selected with 1
µg/ml puromycin for 2 days and then plated at 3×104 cells/well onto a six-well plate, in
mESC medium supplemented with LIF plus 2i.
qRT-PCR
Total RNA was extracted with TRIzol (Invitrogen). cDNA was synthesized with 1 µg of
total RNA, using the QuantiTech Rev. Transcription Kit (Qiagen). qRT-PCR was performed
with Power SYBR Green PCR Master Mix (Applied Biosystems) according to the
manufacturer's instructions. Signals were detected with an ABI7900HT real-time PCR

109
System (Applied Biosystems). The relative expression level was determined by the 2-ΔCT
method and normalized against Gapdh. The primers used for qRT-PCR are listed in Table
2, 3.
Table 2
Symbol Primer sequence for cloning (5’ -> 3’) Primer sequence for RT-PCR (5’ -> 3’)
Gbx2
F:CACACTAGTATGAGCGCAGCGTTCCCGCC
R: CACATGCATTCAGGGTCGGGCCTGCTCCA
F:TCGCTGCTCGCTTTCTCT
R: GGGTCATCTTCCACCTTTGA
Tgtp
F:CACGAATTCATGGCTTGGGCCTCCAGCTT
R: CACGGATCCCTAAGCTTCCCAGTACTCGG
F:CCCTAAGAGGAAAGCCATCAC
R: GAATGCATCAAAGCTGGAGG
Junb
F:CACACTAGTATGTGCACGAAAATGGAACAGCC
R: CACATGCATTCAGAAGGCGTGTCCCTTGA
F:ATCCCTATCGGGGTCTCAAG
R: CCTGTGTCTGATCCCTGACC
Bcl3
F:CACGAATTCATGCCCCGATGCCCCGCGGG
R: CACGGATCCTCAGCTGCTTCCTGGAGCTG
F:CCGGAGGCCCTTTACTACC
R: GAGTAGGCAGGTTCAGCAGC
Pim3
F:CACACTAGTATGCTGCTGTCCAAGTTCGG
R: CACATGCATTCACAAGCTCTCACTGCTGG
F:CTTCTTCGCGCAGGTGCT
R: AGGTCCACCAGCAGGTTCT
Icam1
F:CACGAATTCATGGCTTCAACCCGTGCCAA
R: CACGGATCCCTAGGGAGGTGGGGCTTGTC
F:GTGATGCTCAGGTATCCATCCA
R: CACAGTTCTCAAAGCACAGCG
Dapp1
F:CACACTAGTATGGGCAGAGCAGAACTTCT
R: CACATGCATCTATTTGAAGATAAATGACC
F:CACAGACCTGCTCCAGGATT
R: AGCTACCATCACGTCCATTG
Gadd45g
F:CACGAATTCATGACTCTGGAAGAAGTCCG
R: CACGGATCCTCACTCGGGAAGGGTGATGCT
F:GGATAACTTGCTGTTCGTGGA
R: AAGTTCGTGCAGTGCTTTCC
Hesx1
F: CACGAATTCATGTCTCCCAGCCTTCGGGA
R: CACGGATCCCTATTTCAGAAGATCTGGGTTG
F: GTAAGACCCCACAGACCCTG
R: TCTCACTGGGAAGATCTGGG
Fgf10
F: CACGAATTCATGTGGAAATGGATACTGACACA
R: CACGGATCCCTATGTTTGGATCGTCATGG
F: TTTGGTGTCTTCGTTCCCTGT
R: TAGCTCCGCACATGCCTTC
Rhob
F: CACACTAGTATGGCGGCCATCCGCAAGAA
R: CACATGCATTCATAGCACCTTGCAGCAGTTGATG
F: GTGCCTGCTGATCGTGTTCA
R: CCGAGAAGCACATAAGGATGAC
Klf4
F: CACGAATTCATGAGGCAGCCACCTGGCGA
R: CACGGATCCTTAAAAGTGCCTCTTCATGT
F: CGAACTCACACAGGCGAGAA
R: CGGAGCGGGCGAATTT
Atf3
F: CACGAATTCATGATGCTTCAACATCCAGG
R: CACGGATCCTTAGCTCTGCAATGTTCCTT
F: GAGGATTTTGCTAACCTGACACC
R: TTGACGGTAACTGACTCCAGC
Serpinb1a
F: CACACTAGTATGGAGCAGCTGAGTTCAGCCA
R: CACATGCATCTATGGGGAACAAACCCTGC
F: ACATCCATTCACGCTTCCAAA
R: GGCCAAGTCAGCACCATACAT
En2
F: CACACTAGTATGGAGGAGAAGGATTCCAAGCC
R: CACATGCATCTACTCGCTGTCCGACTTGC
F: TTCTTCAGGTCCCAGGTCCCGA
R: GCAGTGAAGGCTGTGCGAGGC
Spic
F: CACGAATTCATGACTTGTTGTATTGATCA
R: CACGGATCCTCAGCTCTGGTAACTGCCGT
F: AAACATTTCAAGACGCCATTGAC
R: CTCTGACGTGAGGATAAGGGT

110
Irak3
F: CACACTAGTATGGCCGGCCGGTGCGGGGCCC
R: CACATGCATTCACTGCTTTTTGGACTGTT
F: CTGGCTGGATGTTCGTCATATT
R: GGAGAACCTCTAAAAGGTCGC
Yes1
F: CACACTAGTATGGGCTGCATTAAAAGTAA
R: CACATGCATTTATAAATTTTCTCCTGGTT
F: AGTCCAGCCATAAAATACACACC
R: TGATGCTCCCTTTGTGGAAGA
Muc1
F: CACGAATTCATGACCCCGGGCATTCGGGCT
R: CACGGATCCCTACAAGTTGGCAGAAGTGG
F: TACCAAGCGTAGCCCCTATG
R: TGCTCCTACAAGTTGGCAGA

Table 3
Symbol Primer sequence for RT-PCR (5’ -> 3’) Symbol
Primer sequence for RT-PCR (5’ ->
3’)
Gbx2
F: TCGCTGCTCGCTTTCTCT
R: GGGTCATCTTCCACCTTTGA
Nestin
F: CTCGAGCAGGAAGTGGTAGG
R: TTGGGACCAGGGACTGTTAG
Klf4
F: CGAACTCACACAGGCGAGAA
R: CGGAGCGGGCGAATTT
Sox1
F: GGCCGAGTGGAAGGTCATGT
R: TCCGGGTGTTCCTTCATGTG
Esrrb
F: AACAGCCCCTACCTGAACCT
R: CTCATCTGGTCCCCAAGTGT
Mixl1
F: TTGAATTGAACCCTGTTGTCCC
R: GAAACCCGTTCTCCCATCCACC
Oct4
F: GAAGCAGAAGAGGATCACCTTG
R: TTCTTAAGGCTGAGCTGCAAG
T
F: CCGGTGCTGAAGGTAAATGT
R: CCTCCATTGAGCTTGTTGGT
Sox2
F: ATGGGCTCTGTGGTCAAGTC
R: CCCTCCCAATTCCCTTGTAT
Eomes
F: GGCAAAGCGGACAATAACAT
R: AGCCTCGGTTGGTATTTGTG
Gata4
F: TCTCACTATGGGCACAGCAG
R: GCGATGTCTGAGTGACAGGA
Cdx2
F: ACCGGAATTGTTTGCTGCTGT
R: TCCCGACTTCCCTTCACCAT
Gata6
F: TCCTCCCCTGCCGAAGTC
R: AGGGCCAGAGCACACCAA
Dlx3
F: TACTCGCCCAAGTCGGAATA
R: AGTAGATCGTTCGCGGCTTT
FoxA2
F: CCTCAAGGGAGCAGTCTCAC
R: TTTCTCCTGGTCCGGTACAC
Esx1
F: GAGCTGGAGGCCTTTTTCCA
R: ACACCCACAGGGGGACTCAT
Sox17
F: AGCCATTTCCTCCGTGGTGT
R: AACACTGCTTCTGGCCCTCAG
Gata3
F: GGGCTACGGTGCAGAGGTAT
R: TGGATGGACGTCTTGGAGAA
Psx1
F: CGATGGATGGGTGTGGATGA
R: TGACAGGGCTGGCACTCAAG
Tfap2c
F: ATCCCTCACCTCTCCTCTCC
R: CCAGATGCGAGTAATGGTCGG
Gapdh
F: TGTGAGGGAGATGCTCAGTG
R: TGTTCCTACCCCCAATGTGT

111
References:
1. M. J. Evans, M. H. Kaufman, Establishment in culture of pluripotential cells from
mouse embryos. Nature 292, 154 (Jul 9, 1981).
2. G. R. Martin, Isolation of a pluripotent cell line from early mouse embryos
cultured in medium conditioned by teratocarcinoma stem cells. Proceedings of
the National Academy of Sciences of the United States of America 78, 7634 (Dec,
1981).
3. Y. G. Kim, J. Cha, S. Chandrasegaran, Hybrid restriction enzymes: zinc finger
fusions to Fok I cleavage domain. Proceedings of the National Academy of
Sciences of the United States of America 93, 1156 (Feb 6, 1996).
4. P. Horvath, R. Barrangou, CRISPR/Cas, the immune system of bacteria and
archaea. Science 327, 167 (Jan 8, 2010).
5. K. R. Thomas, M. R. Capecchi, Site-directed mutagenesis by gene targeting in
mouse embryo-derived stem cells. Cell 51, 503 (Nov 6, 1987).
6. A. G. Smith et al., Inhibition of pluripotential embryonic stem cell differentiation
by purified polypeptides. Nature 336, 688 (Dec 15, 1988).
7. R. L. Williams et al., Myeloid leukaemia inhibitory factor maintains the
developmental potential of embryonic stem cells. Nature 336, 684 (Dec 15,
1988).
8. E. Hondo, C. L. Stewart, Profiling gene expression in growth-arrested mouse
embryos in diapause. Genome biology 6, 202 (2005).
9. C. L. Stewart et al., Blastocyst implantation depends on maternal expression of
leukaemia inhibitory factor. Nature 359, 76 (Sep 3, 1992).
10. C. B. Ware et al., Targeted disruption of the low-affinity leukemia inhibitory
factor receptor gene causes placental, skeletal, neural and metabolic defects and
results in perinatal death. Development 121, 1283 (May, 1995).
11. K. Takeda et al., Targeted disruption of the mouse Stat3 gene leads to early
embryonic lethality. Proceedings of the National Academy of Sciences of the
United States of America 94, 3801 (Apr 15, 1997).
12. S. J. Rodig et al., Disruption of the Jak1 gene demonstrates obligatory and
nonredundant roles of the Jaks in cytokine-induced biologic responses. Cell 93,
373 (May 1, 1998).
13. J. C. Conover et al., Ciliary neurotrophic factor maintains the pluripotentiality of
embryonic stem cells. Development 119, 559 (Nov, 1993).
14. T. M. Rose, D. M. Weiford, N. L. Gunderson, A. G. Bruce, Oncostatin M (OSM)
inhibits the differentiation of pluripotent embryonic stem cells in vitro. Cytokine
6, 48 (Jan, 1994).
15. E. Wolf, R. Kramer, I. Polejaeva, H. Thoenen, G. Brem, Efficient generation of
chimaeric mice using embryonic stem cells after long-term culture in the
presence of ciliary neurotrophic factor. Transgenic research 3, 152 (May, 1994).

112
16. K. Yoshida et al., Maintenance of the pluripotential phenotype of embryonic
stem cells through direct activation of gp130 signalling pathways. Mechanisms of
development 45, 163 (Feb, 1994).
17. D. Pennica et al., Cardiotrophin-1. Biological activities and binding to the
leukemia inhibitory factor receptor/gp130 signaling complex. The Journal of
biological chemistry 270, 10915 (May 5, 1995).
18. J. Nichols, I. Chambers, A. Smith, Derivation of germline competent embryonic
stem cells with a combination of interleukin-6 and soluble interleukin-6 receptor.
Experimental cell research 215, 237 (Nov, 1994).
19. M. Tomida, Y. Yamamoto-Yamaguchi, M. Hozumi, Purification of a factor
inducing differentiation of mouse myeloid leukemic M1 cells from conditioned
medium of mouse fibroblast L929 cells. The Journal of biological chemistry 259,
10978 (Sep 10, 1984).
20. H. Baumann, G. G. Wong, Hepatocyte-stimulating factor III shares structural and
functional identity with leukemia-inhibitory factor. J Immunol 143, 1163 (Aug 15,
1989).
21. K. Yoshida et al., Targeted disruption of gp130, a common signal transducer for
the interleukin 6 family of cytokines, leads to myocardial and hematological
disorders. Proceedings of the National Academy of Sciences of the United States
of America 93, 407 (Jan 9, 1996).
22. T. Yamamori et al., The cholinergic neuronal differentiation factor from heart
cells is identical to leukemia inhibitory factor. Science 246, 1412 (Dec 15, 1989).
23. M. Li, M. Sendtner, A. Smith, Essential function of LIF receptor in motor neurons.
Nature 378, 724 (Dec 14, 1995).
24. A. Bonni et al., Regulation of gliogenesis in the central nervous system by the
JAK-STAT signaling pathway. Science 278, 477 (Oct 17, 1997).
25. M. Narazaki et al., Activation of JAK2 kinase mediated by the interleukin 6 signal
transducer gp130. Proceedings of the National Academy of Sciences of the United
States of America 91, 2285 (Mar 15, 1994).
26. N. Stahl et al., Association and activation of Jak-Tyk kinases by CNTF-LIF-OSM-IL-
6 beta receptor components. Science 263, 92 (Jan 7, 1994).
27. M. Ernst, A. Oates, A. R. Dunn, Gp130-mediated signal transduction in embryonic
stem cells involves activation of Jak and Ras/mitogen-activated protein kinase
pathways. The Journal of biological chemistry 271, 30136 (Nov 22, 1996).
28. J. N. Ihle, STATs: signal transducers and activators of transcription. Cell 84, 331
(Feb 9, 1996).
29. M. Minami et al., STAT3 activation is a critical step in gp130-mediated terminal
differentiation and growth arrest of a myeloid cell line. Proceedings of the
National Academy of Sciences of the United States of America 93, 3963 (Apr 30,
1996).
30. H. Niwa, T. Burdon, I. Chambers, A. Smith, Self-renewal of pluripotent embryonic
stem cells is mediated via activation of STAT3. Genes & development 12, 2048
(Jul 1, 1998).

113
31. T. Matsuda et al., STAT3 activation is sufficient to maintain an undifferentiated
state of mouse embryonic stem cells. The EMBO journal 18, 4261 (Aug 2, 1999).
32. T. Fukada et al., Two signals are necessary for cell proliferation induced by a
cytokine receptor gp130: involvement of STAT3 in anti-apoptosis. Immunity 5,
449 (Nov, 1996).
33. C. K. Qu, G. S. Feng, Shp-2 has a positive regulatory role in ES cell differentiation
and proliferation. Oncogene 17, 433 (Jul 30, 1998).
34. A. M. Cheng et al., Mammalian Grb2 regulates multiple steps in embryonic
development and malignant transformation. Cell 95, 793 (Dec 11, 1998).
35. T. Burdon, C. Stracey, I. Chambers, J. Nichols, A. Smith, Suppression of SHP-2 and
ERK signalling promotes self-renewal of mouse embryonic stem cells.
Developmental biology 210, 30 (Jun 1, 1999).
36. D. P. Brazil, Z. Z. Yang, B. A. Hemmings, Advances in protein kinase B signalling:
AKTion on multiple fronts. Trends in biochemical sciences 29, 233 (May, 2004).
37. N. R. Paling, H. Wheadon, H. K. Bone, M. J. Welham, Regulation of embryonic
stem cell self-renewal by phosphoinositide 3-kinase-dependent signaling. The
Journal of biological chemistry 279, 48063 (Nov 12, 2004).
38. S. Watanabe et al., Activation of Akt signaling is sufficient to maintain
pluripotency in mouse and primate embryonic stem cells. Oncogene 25, 2697
(May 4, 2006).
39. H. Niwa, K. Ogawa, D. Shimosato, K. Adachi, A parallel circuit of LIF signalling
pathways maintains pluripotency of mouse ES cells. Nature 460, 118 (Jul 2, 2009).
40. P. Y. Bourillot et al., Novel STAT3 target genes exert distinct roles in the
inhibition of mesoderm and endoderm differentiation in cooperation with Nanog.
Stem Cells 27, 1760 (Aug, 2009).
41. M. Ema et al., Kruppel-like factor 5 is essential for blastocyst development and
the normal self-renewal of mouse ESCs. Cell Stem Cell 3, 555 (Nov 6, 2008).
42. Y. Li et al., Murine embryonic stem cell differentiation is promoted by SOCS-3
and inhibited by the zinc finger transcription factor Klf4. Blood 105, 635 (Jan 15,
2005).
43. J. Hall et al., Oct4 and LIF/Stat3 additively induce Kruppel factors to sustain
embryonic stem cell self-renewal. Cell Stem Cell 5, 597 (Dec 4, 2009).
44. J. Jiang et al., A core Klf circuitry regulates self-renewal of embryonic stem cells.
Nature cell biology 10, 353 (Mar, 2008).
45. C. I. Tai, Q. L. Ying, Gbx2, a LIF/Stat3 target, promotes reprogramming to and
retention of the pluripotent ground state. Journal of cell science 126, 1093 (Mar
1, 2013).
46. S. Ye, P. Li, C. Tong, Q. L. Ying, Embryonic stem cell self-renewal pathways
converge on the transcription factor Tfcp2l1. The EMBO journal 32, 2548 (Oct 2,
2013).
47. G. Martello, P. Bertone, A. Smith, Identification of the missing pluripotency
mediator downstream of leukaemia inhibitory factor. The EMBO journal 32, 2561
(Oct 2, 2013).

114
48. S. Parisi et al., Klf5 is involved in self-renewal of mouse embryonic stem cells.
Journal of cell science 121, 2629 (Aug 15, 2008).
49. I. Aksoy et al., Self-renewal of murine embryonic stem cells is supported by the
serine/threonine kinases Pim-1 and Pim-3. Stem Cells 25, 2996 (Dec, 2007).
50. E. A. Casanova et al., Pramel7 mediates LIF/STAT3-dependent self-renewal in
embryonic stem cells. Stem Cells 29, 474 (Mar, 2011).
51. P. Cartwright et al., LIF/STAT3 controls ES cell self-renewal and pluripotency by a
Myc-dependent mechanism. Development 132, 885 (Mar, 2005).
52. Q. L. Ying, M. Stavridis, D. Griffiths, M. Li, A. Smith, Conversion of embryonic
stem cells into neuroectodermal precursors in adherent monoculture. Nat
Biotechnol 21, 183 (Feb, 2003).
53. N. Stahl et al., Choice of STATs and other substrates specified by modular
tyrosine-based motifs in cytokine receptors. Science 267, 1349 (Mar 3, 1995).
54. Q. L. Ying et al., The ground state of embryonic stem cell self-renewal. Nature
453, 519 (May 22, 2008).
55. Q. Zhou, H. Chipperfield, D. A. Melton, W. H. Wong, A gene regulatory network
in mouse embryonic stem cells. Proc Natl Acad Sci U S A 104, 16438 (Oct 16,
2007).
56. X. Chen et al., Integration of external signaling pathways with the core
transcriptional network in embryonic stem cells. Cell 133, 1106 (Jun 13, 2008).
57. K. M. Wassarman et al., Specification of the anterior hindbrain and
establishment of a normal mid/hindbrain organizer is dependent on Gbx2 gene
function. Development 124, 2923 (Aug, 1997).
58. K. Takahashi, S. Yamanaka, Induction of pluripotent stem cells from mouse
embryonic and adult fibroblast cultures by defined factors. Cell 126, 663 (Aug 25,
2006).
59. J. Yu et al., Induced pluripotent stem cell lines derived from human somatic cells.
Science 318, 1917 (Dec 21, 2007).
60. G. Guo et al., Klf4 reverts developmentally programmed restriction of ground
state pluripotency. Development 136, 1063 (Apr, 2009).
61. J. Yang et al., Stat3 activation is limiting for reprogramming to ground state
pluripotency. Cell Stem Cell 7, 319 (Sep 3, 2010).
62. T. W. Theunissen et al., Nanog overcomes reprogramming barriers and induces
pluripotency in minimal conditions. Curr Biol 21, 65 (Jan 11, 2011).
63. B. Greber et al., Conserved and divergent roles of FGF signaling in mouse epiblast
stem cells and human embryonic stem cells. Cell Stem Cell 6, 215 (Mar 5, 2010).
64. E. Kawase et al., Strain difference in establishment of mouse embryonic stem (ES)
cell lines. The International journal of developmental biology 38, 385 (Jun, 1994).
65. F. A. Brook, R. L. Gardner, The origin and efficient derivation of embryonic stem
cells in the mouse. Proceedings of the National Academy of Sciences of the
United States of America 94, 5709 (May 27, 1997).
66. A. Nagy, J. Rossant, R. Nagy, W. Abramow-Newerly, J. C. Roder, Derivation of
completely cell culture-derived mice from early-passage embryonic stem cells.

115
Proceedings of the National Academy of Sciences of the United States of America
90, 8424 (Sep 15, 1993).
67. J. Schmitz, M. Weissenbach, S. Haan, P. C. Heinrich, F. Schaper, SOCS3 exerts its
inhibitory function on interleukin-6 signal transduction through the SHP2
recruitment site of gp130. The Journal of biological chemistry 275, 12848 (Apr 28,
2000).
68. C. D. Chung et al., Specific inhibition of Stat3 signal transduction by PIAS3.
Science 278, 1803 (Dec 5, 1997).
69. D. Strumpf et al., Cdx2 is required for correct cell fate specification and
differentiation of trophectoderm in the mouse blastocyst. Development 132,
2093 (May, 2005).
70. H. Niwa et al., Interaction between Oct3/4 and Cdx2 determines trophectoderm
differentiation. Cell 123, 917 (Dec 2, 2005).
71. S. Ray et al., Context-dependent function of regulatory elements and a switch in
chromatin occupancy between GATA3 and GATA2 regulate Gata2 transcription
during trophoblast differentiation. The Journal of biological chemistry 284, 4978
(Feb 20, 2009).
72. A. Ralston et al., Gata3 regulates trophoblast development downstream of
Tead4 and in parallel to Cdx2. Development 137, 395 (Feb, 2010).
73. Y. Chen, K. Blair, A. Smith, Robust Self-Renewal of Rat Embryonic Stem Cells
Requires Fine-Tuning of Glycogen Synthase Kinase-3 Inhibition. Stem cell reports
1, 209 (2013).
74. R. Yagi et al., Transcription factor TEAD4 specifies the trophectoderm lineage at
the beginning of mammalian development. Development 134, 3827 (Nov, 2007).
75. U. Werling, H. Schorle, Transcription factor gene AP-2 gamma essential for early
murine development. Molecular and cellular biology 22, 3149 (May, 2002).
76. P. Kuckenberg et al., The transcription factor TCFAP2C/AP-2gamma cooperates
with CDX2 to maintain trophectoderm formation. Molecular and cellular biology
30, 3310 (Jul, 2010).
77. K. Adachi et al., Context-dependent wiring of Sox2 regulatory networks for self-
renewal of embryonic and trophoblast stem cells. Molecular cell 52, 380 (Nov 7,
2013).
78. W. Yang et al., An Shp2/SFK/Ras/Erk signaling pathway controls trophoblast stem
cell survival. Developmental cell 10, 317 (Mar, 2006).
79. C. W. Lu et al., Ras-MAPK signaling promotes trophectoderm formation from
embryonic stem cells and mouse embryos. Nature genetics 40, 921 (Jul, 2008).
80. B. Barre, S. Avril, O. Coqueret, Opposite regulation of myc and p21waf1
transcription by STAT3 proteins. The Journal of biological chemistry 278, 2990
(Jan 31, 2003).
81. K. Watanabe et al., A ROCK inhibitor permits survival of dissociated human
embryonic stem cells. Nature biotechnology 25, 681 (Jun, 2007).
82. J. S. Fitzgerald et al., Signal transduction in trophoblast invasion. Chemical
immunology and allergy 88, 181 (2005).

116
83. J. S. Fitzgerald, T. G. Poehlmann, E. Schleussner, U. R. Markert, Trophoblast
invasion: the role of intracellular cytokine signalling via signal transducer and
activator of transcription 3 (STAT3). Human reproduction update 14, 335 (Jul-Aug,
2008).
84. T. G. Poehlmann et al., Trophoblast invasion: tuning through LIF, signalling via
Stat3. Placenta 26 Suppl A, S37 (Apr, 2005).
85. Y. Takahashi et al., Leukemia inhibitory factor regulates trophoblast giant cell
differentiation via Janus kinase 1-signal transducer and activator of transcription
3-suppressor of cytokine signaling 3 pathway. Mol Endocrinol 22, 1673 (Jul, 2008).
86. P. Li et al., Germline competent embryonic stem cells derived from rat
blastocysts. Cell 135, 1299 (Dec 26, 2008).
87. G. Chapman et al., The mouse homeobox gene, Gbx2: genomic organization and
expression in pluripotent cells in vitro and in vivo. Genomics 46, 223 (Dec 1,
1997).
88. D. A. Roeseler et al., Elongation factor 1 alpha1 and genes associated with Usher
syndromes are downstream targets of GBX2. PloS one 7, e47366 (2012).
89. A. C. Gao, W. Lou, J. T. Isaacs, Enhanced GBX2 expression stimulates growth of
human prostate cancer cells via transcriptional up-regulation of the interleukin 6
gene. Clinical cancer research : an official journal of the American Association for
Cancer Research 6, 493 (Feb, 2000).
90. T. Katahira et al., Interaction between Otx2 and Gbx2 defines the organizing
center for the optic tectum. Mechanisms of development 91, 43 (Mar 1, 2000).
91. A. L. Joyner, A. Liu, S. Millet, Otx2, Gbx2 and Fgf8 interact to position and
maintain a mid-hindbrain organizer. Current opinion in cell biology 12, 736 (Dec,
2000).
92. D. Acampora, L. G. Di Giovannantonio, A. Simeone, Otx2 is an intrinsic
determinant of the embryonic stem cell state and is required for transition to a
stable epiblast stem cell condition. Development 140, 43 (Jan 1, 2013).
93. T. Heimbucher et al., Gbx2 and Otx2 interact with the WD40 domain of
Groucho/Tle corepressors. Molecular and cellular biology 27, 340 (Jan, 2007).
94. C. Logan et al., Cloning and sequence comparison of the mouse, human, and
chicken engrailed genes reveal potential functional domains and regulatory
regions. Developmental genetics 13, 345 (1992).
95. E. N. Tolkunova, M. Fujioka, M. Kobayashi, D. Deka, J. B. Jaynes, Two distinct
types of repression domain in engrailed: one interacts with the groucho
corepressor and is preferentially active on integrated target genes. Molecular
and cellular biology 18, 2804 (May, 1998).
96. G. Jimenez, Z. Paroush, D. Ish-Horowicz, Groucho acts as a corepressor for a
subset of negative regulators, including Hairy and Engrailed. Genes &
development 11, 3072 (Nov 15, 1997).
97. B. H. Jennings et al., Molecular recognition of transcriptional repressor motifs by
the WD domain of the Groucho/TLE corepressor. Molecular cell 22, 645 (Jun 9,
2006).

117
98. I. Brunet et al., The transcription factor Engrailed-2 guides retinal axons. Nature
438, 94 (Nov 3, 2005).
99. N. L. Martin, M. K. Saba-El-Leil, S. Sadekova, S. Meloche, G. Sauvageau, EN2 is a
candidate oncogene in human breast cancer. Oncogene 24, 6890 (Oct 20, 2005).
100. S. K. Bose, R. S. Bullard, C. D. Donald, Oncogenic role of engrailed-2 (en-2) in
prostate cancer cell growth and survival. Translational oncogenomics 3, 37
(2008).
101. R. Morgan et al., Expression of Engrailed-2 (EN2) protein in bladder cancer and
its potential utility as a urinary diagnostic biomarker. Eur J Cancer 49, 2214 (Jun,
2013).
102. R. Benayed et al., Support for the homeobox transcription factor gene
ENGRAILED 2 as an autism spectrum disorder susceptibility locus. American
journal of human genetics 77, 851 (Nov, 2005).
103. F. Neri et al., Myc regulates the transcription of the PRC2 gene to control the
expression of developmental genes in embryonic stem cells. Molecular and
cellular biology 32, 840 (Feb, 2012).
104. A. Joliot et al., Association of Engrailed homeoproteins with vesicles presenting
caveolae-like properties. Development 124, 1865 (May, 1997).
105. A. Maizel, O. Bensaude, A. Prochiantz, A. Joliot, A short region of its
homeodomain is necessary for engrailed nuclear export and secretion.
Development 126, 3183 (Jun, 1999).
106. S. J. Dunn, G. Martello, B. Yordanov, S. Emmott, A. G. Smith, Defining an essential
transcription factor program for naive pluripotency. Science 344, 1156 (Jun 6,
2014).
107. L. L. McGrew, K. Takemaru, R. Bates, R. T. Moon, Direct regulation of the
Xenopus engrailed-2 promoter by the Wnt signaling pathway, and a molecular
screen for Wnt-responsive genes, confirm a role for Wnt signaling during neural
patterning in Xenopus. Mechanisms of development 87, 21 (Sep, 1999).
108. H. Kazumori, S. Ishihara, M. A. Rumi, Y. Kadowaki, Y. Kinoshita, Bile acids directly
augment caudal related homeobox gene Cdx2 expression in oesophageal
keratinocytes in Barrett's epithelium. Gut 55, 16 (Jan, 2006).
109. J. Das et al., A critical role for NF-kappa B in GATA3 expression and TH2
differentiation in allergic airway inflammation. Nature immunology 2, 45 (Jan,
2001).
110. J. Yang et al., Unphosphorylated STAT3 accumulates in response to IL-6 and
activates transcription by binding to NFkappaB. Genes & development 21, 1396
(Jun 1, 2007).
111. K. Watanabe et al., A ROCK inhibitor permits survival of dissociated human
embryonic stem cells. Nature biotechnology 25, 681 (Jun, 2007).
112. M. L. Coleman et al., Membrane blebbing during apoptosis results from caspase-
mediated activation of ROCK I. Nature cell biology 3, 339 (Apr, 2001).
113. M. Sebbagh et al., Caspase-3-mediated cleavage of ROCK I induces MLC
phosphorylation and apoptotic membrane blebbing. Nature cell biology 3, 346
(Apr, 2001).

118
114. H. Huang et al., Rho-kinase regulates energy balance by targeting hypothalamic
leptin receptor signaling. Nature neuroscience 15, 1391 (Oct, 2012).
115. H. Baumann et al., The full-length leptin receptor has signaling capabilities of
interleukin 6-type cytokine receptors. Proceedings of the National Academy of
Sciences of the United States of America 93, 8374 (Aug 6, 1996).
116. V. Sanz-Moreno et al., ROCK and JAK1 signaling cooperate to control actomyosin
contractility in tumor cells and stroma. Cancer cell 20, 229 (Aug 16, 2011).
117. S. Ye et al., Pleiotropy of glycogen synthase kinase-3 inhibition by CHIR99021
promotes self-renewal of embryonic stem cells from refractory mouse strains.
PloS one 7, e35892 (2012).
118. M. Buehr et al., Rapid loss of Oct-4 and pluripotency in cultured rodent
blastocysts and derivative cell lines. Biology of reproduction 68, 222 (Jan, 2003).

Abstract (if available)

Abstract Activation of signal transducer and activator of transcription 3 (Stat3) by leukemia inhibitory factor (LIF) maintains mouse embryonic stem cell (mESC) self-renewal and also facilitates reprogramming to ground state pluripotency. Exactly how LIF/Stat3 signaling exerts these effects, however, remains elusive. I identified gastrulation brain homeobox 2 (Gbx2) and engrailed-2 (En2) as LIF/Stat3 downstream targets that, when overexpressed, allows long-term expansion of undifferentiated mESCs in the absence of LIF/Stat3 signaling. Elevated Gbx2 expression also enhanced reprogramming of mouse embryonic fibroblasts to induced pluripotent stem cells. Moreover, overexpression of Gbx2 was sufficient to reprogram epiblast stem cells to ground state ESCs. My first part of thesis reveal a novel function of Gbx2 and En2 in mESC reprogramming and LIF/Stat3-mediated self-renewal. ❧ Stat3 is essential for mouse embryonic stem cell (mESC) self-renewal mediated by LIF/gp130 receptor signaling. Current understanding of Stat3-mediated ESC self-renewal mechanisms is very limited, and has heretofore been dominated by the view that Stat3 signaling functions in a binary “on/off” manner. Here, in contrast to this binary viewpoint, I demonstrate a contextual, rheostat-like mechanism for Stat3's function in mESCs. Activation and expression levels determine whether Stat3 functions in a self-renewal or a differentiation role in mESCs. I also show that Stat3 induces rapid differentiation of mESCs toward the trophectoderm (TE) lineage when its activation level exceeds certain thresholds. Stat3 induces this differentiation phenotype via induction of Tfap2c and its downstream target Cdx2. My findings provide a novel concept in the realm of Stat3, self-renewal signaling, and pluripotent stem cell biology. Ultimately, this finding may facilitate the development of conditions for the establishment of authentic non-rodent ESCs.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Role of STAT3 phosphorylation in mouse embryonic stem cell self-renewal and differentiation

PDF

Investigating the role of STAT3 in mouse and rat embryonic stem cell self-renewal and differentiation

PDF

Role of beta-catenin in mouse epiblast stem cell, embryonic stem cell self-renewal and differentiation

PDF

Molecular basis of mouse epiblast stem cell and human embryonic stem cell self‐renewal

PDF

Roles of Klf4 in embryonic stem cells

PDF

The role of ERK1/2 in mouse embryonic stem cell fate control

PDF

Interaction of epigenetics and SMAD signaling in stem cells and diseases

PDF

Exploring stem cell pluripotency through long range chromosome interactions

PDF

The function of BS69 in mouse embryogenesis and embryonic stem cell differentiation

PDF

Novel roles for Maf1 in embryonic stem cell differentiation and adipogenesis

PDF

Leucine-rich amelogenin peptide induces osteogenesis in mouse embryonic stem cells

PDF

Elucidating the role of neural crest specific Stat3 signaling in maintaining coronal suture patency during embryonic development

PDF

Harnessing the power of stem cell self-renewal pathways in cancer: dissecting the role of BMI-1 in Ewing’s sarcoma initiation and maintenance

PDF

Characterization of human embryonic stem cell derived retinal pigment epithelial cells for age-related macular degeneration

PDF

Epigenetic plasticity of cultured female human embryonic stem cells and regulation of gene expression and chromatin by PR-SET7 mediated H4K20me1

PDF

Roles of epithelial-mesenchymal transition and niche in tumorigenesis of tumor-initiating cells

PDF

Polycomb repressive complex 2 subunit stabilizes NANOG to maintain self-renewal in hepatocellular carcinoma tumor-initiating stem-like cells

PDF

The role of Prkci in stem cell maintenance and cell polarity using a 3-D culture system

PDF

Molecular role of EZH2 overexpression in Colorectal Cancer progression

PDF

Role of the bone marrow niche components in B cell malignancies

Asset Metadata

Creator Tai, Chih-I (author)

Core Title Pleotropic potential of Stat3 in determining self-renewal, apoptosis, and differentiation in mouse embryonic stem cells

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 09/15/2014

Defense Date 06/10/2014

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Cdx2,differentiation,embryonic stem cell,En2,engrailed 2,epiblast stem cell,EpiSC,ESC,gastrulation brain homeobox 2,Gbx2,LIF,LIF receptor,OAI-PMH Harvest,PIAS3,pluripotency,reprogramming,ROCKi,self renewal,Stat3,Stat3 activation,Stat3 hyperactivation,Tfap2c,trophoblast,trophoectoderm

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Ying, Qi-Long (committee chair), Chuong, Cheng-Ming (committee member), Mariani, Francesca (committee member), Maxson, Robert E., Jr. (committee member), McMahon, Andrew P. (committee member)

Creator Email chihitai@usc.edu,jimmyzxcd@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-478540

Unique identifier UC11287671

Identifier etd-TaiChihI-2944.pdf (filename),usctheses-c3-478540 (legacy record id)

Legacy Identifier etd-TaiChihI-2944.pdf

Dmrecord 478540

Document Type Dissertation

Format application/pdf (imt)

Rights Tai, Chih-I

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA