Elucidation of transcription factors controlling mouse ES cell pluripotency and differentiation

1.1 Early embryo development and stem cell fate decision 2 1.3 Molecular mechanisms in regulating ES cells pluripotency and differentiation 7 1.3.2 Intrinsic determination of Pluripoten

CHARACTERIZATION OF ZFP206 AND REST

YU HONGBING

NATIONAL UNIVERSITY OF SINGAPORE

2009

AN INVESTIGATION OF THE REGULATORY

NETWORK IN EMBRYONIC STEM CELLS THROUGH

CHARACTERIZATION OF ZFP206 AND REST

2009

My sister, who admires me so much that I cannot disappoint her

My supervisor, Dr Lawrence Stanton, who gave me the valuable opportunity to pursue the stem cell research with the cutting edge technology in his lab He provided me with the freedom to grow and develop, but always there with guidance

to keep me on the correct track I was impressed with his encouragement when my first project was scooped Because of his encouragement and patient guidance, I quickly regained my confidence and moved forward Special thanks also go to him for his great effort in editing my papers and correcting this thesis I love his American style of guidance!!!

Dr Thomas Lufkin and his group members, Petra Kraus and Lim Siew Lan, for

tremendous help in Zfp206 knockout project

Dr Rory Johnson, who is very smart and reads tons of papers He gave me a many good suggestions and was a fantastic collaborator His help was like a glowing ember in the snow that warmed me when I needed it most Thank you very much,

my friend!

Galih Kunarso, the future Doctor, so smart and so efficient Thanks very much for your great bioinformatics analyses It was always a pleasure to work with him The memorable photos taken by him were a lot of fun

Hong Huimei Felicia, the future Singapore Star, her help really speeded up my project and relieved my burden

I would also like to thank Khaw Swea Ling, Wong Kee Yew, and other lab mates for their help

Thanks to Ralf Jauch and Choo Siew Hua for their help with EMSA experiments

I would also like to thank these colleagues: Andrew Hutchins, Soh Boon Seng, Luo Wenglong, Wang Caoyang, and Sun Lili

And all of those who have helped me!!!

1.1 Early embryo development and stem cell fate decision 2

1.3 Molecular mechanisms in regulating ES cells

pluripotency and differentiation 7

1.3.2 Intrinsic determination of Pluripotency 17

1.3.5 Transcriptional regulatory network in ES cells 25

1.4.1 The SCAN domain family of zinc finger

Chapter II: In vitro role of Zfp206 in mouse embryonic stem cells

and in vivo role in mouse development……….53

2.2.3 Zfp206 selectively activates or represses target genes 70

2.2.4 Zfp206 physically interacts with Oct4 and Sox2 72

2.2.5 Genome wide mapping of Zfp206 in ES cell with ChIP-seq 74

2.3 In vivo knockout of Zfp206 in mouse 84

2.3.1 Zfp206 Gene trap cell line generation 84 2.3.2 Genotyping of Zfp206 gene trap cell line with PCR 86 2.3.3 Southern-blot Genotype strategy for Zfp206

knockout mouse from the 285B6 gene trap cell line 94 2.3.4 Generation of the Zfp206 knockout mouse 96

2.3.5 Generation of recombinant Zfp206 knockout mouse 99 2.3.6 Generation of double knockout Zfp206 mouse 102

2.4 Discussion 104

Chapter III: The role of REST and its cofactors in ES cells ………110

3.1 Introduction 110

3.2 Result 113

3.2.1 The role of REST and its cofactors in ES Cells 113 3.2.1.1 Dynamic expression profile of REST cofactors in ES Cell 113

3.2.1.2 Depletion of REST cofactors in ES cells 116 3.2.2 Genome wide mapping of REST and its cofactors’ binding sites in ES Cells 121

3.2.2.1 Generation of Stable REST cofactors over expressing cells 121

3.2.2.2 Genome wide mapping of REST and its cofactors 125 3.2.3 Motif analysis of REST and its cofactors 129

3.2.4 Co-occupancy analysis of REST and its cofactors with other TFs and epigenetic histone marks 133

3.2.5 Derepression of REST targets correlate with co-occupancy of REST and its cofactors 137

3.3 Discussion 149

3.3.1 The role of REST and its cofactors in ES Cells 149 3.3.2 Genome wide mapping of REST cofactors’ binding sites 150 3.3.3 The role of REST cofactors in REST mediated gene repression 152

Chapter IV : Conclusion and Perspectives…… ………155

4.1 Conclusion 155

4.2 Perspectives 159

Chapter V: Materials and Methods……… 161

5.2 Expansion and mitotic inactivation of MEF cells 161

5.5 RNA interference (shRNA) 162 5.6 LIF withdrawal and Retinoic acid induced cell differentiation 163

5.16 ChIP-chip assays, data processing, and statistical analysis 168 5.17 Functional annotations using the Panther database 169

5.19 Statistical analysis of microarray data 170

5.22 Transformation of chemically competent cells 171 5.23 PCR analysis of transformants 172 5.24 Isolation of plasmid DNA from bacteria 172 5.25 Preparation of bacterial stocks 173 5.26 Isolation of genomic DNA from cell lines 173

Summary

Embryonic stem (ES) cells can be maintained in undifferentiated states for indefinite passages and yet retain the potential to differentiate into all cell types An intricate transcriptional regulatory network is, in part, responsible for maintaining such a

pluripotency state In vitro depletion of components of this network, such as Oct4, Sox2, Nanog, and Zic3, induce distinct cellular differentiation responses In this

thesis, I provide additional details of the regulatory network in ES cell pluripotency

through characterization of the transcriptional regulators, REST and Zfp206, which

have emerged as regulators of ES cell pluripotency

REST has been shown to repress neuronal gene expression in neuronal stem cells (NSC) and non neuronal cells Our group has recently shown that REST regulates

distinct regulatory pathways in ES cells and NSC By using genome wide mapping

of the binding sites for REST and 5 of its cofactors, as well as gene expression profiling upon loss of REST and each cofactors, I found that the REST complex

regulates ES cell pluripotency through recruitment of its cofactors

In addition, using genome wide mapping techniques, I have identified a Zfp206 regulatory network and established a physiological interaction of Zfp206 with Oct4 and Sox2 to further expand our understanding of this transcriptional regulatory network Genome wide mapping of Zfp206 binding sites with ChIP-seq shows that Zfp206 binding targets are enriched in developmental process, transcription regulation, and embryogenesis Finally, a knockout of Zfp206 in mice was generated,

though phenotyping is still ongoing

LIST OF TABLES

Table 1 Gene ontology of Zfp206 target genes

62 Table 2 Common transcription factors targets of Oct4, Sox2 and

Table 3 Gene ontology of Zfp206 ChIP-seq target genes

75 Table 4 Summary of the ChIP-seq libraries of REST and its cofactors

Figure 1.4 Intracellular signaling pathways activated through FGFRs 16 Figure 1.5 Core Transcriptional Regulatory Network in Human ES Cells 26 Figure 1.6 Transcriptional regulatory networks in ES Cells 28

Figure 1.8 Model of the structural features of the mouse SCAN domain

Figure 1.9 Conserved domains in some of the SCAN domain family

Figure 1.12 Schematic Models for the Differential Regulation of REST

Figure 2.1 Mapping of Zfp206 binding sites at the Oct4 promoter 58 Figure 2.2 Validation of selected Zfp206 target genes 61 Figure 2.3 Consensus DNA motif for Zfp206 binding sites 64 Figure 2.4 Frequent co-targeting of genes by Zfp206 and other

Figure 2.15 Alignment analysis of PCR product 1 sequence with intron1

Figure 2.17 Southern blot genotyping for non-recombinant and

Figure 3.1 Alignment of CoREST homologues by consensus ClustalW

Figure 3.5 Lineage markers expression change upon knockdown of

Figure 3.6 Over-expression construct for REST cofactors 123 Figure 3.7 Morphology of stable REST cofactor’s overexpressing cells 124 Figure 3.8 Gene expression profiles of REST cofactors in stable cells 124 Figure 3.9 REST cofactors ChIP enrichment with REST targets 126 Figure 3.10 Sin3A ChIP is enriched with REST targets 126 Figure 3.11 Normalization of REST and its cofactors ChIP-seq data with

Figure 3.12 Identification of enriched motifs by using a de novo approach 130 Figure 3.13 Identification of enriched motifs by using a de novo approach 131 Figure 3.14 ChIP-seq sites overlapping RE1 motif for REST and its

Figure 3.15 Multiple transcription factor-binding Loci Associated with

Figure 3.16 Multiple histones modification-Binding Loci associated with

Figure 3.17 Gene expression profile of REST knockdown with REST

List of ABBREVIATIONS

Symbol Definition

ml milli-litre

mM milli-molar

Chapter I Introduction

1 Brief introduction

Embryonic stem (ES) cells are derived from the inner cell mass (ICM) of early preimplantation embryos (Evans and Kaufman, 1981) (Martin, 1981) (Thomson et al., 1998) These cells have attracted broad interest due to their ability to self-renew indefinitely when cultured under proper conditions and their capacity to differentiate into virtually all cell and tissue types These fundamental features have made ES cells a valuable resource for cellular replacement therapy and tissue engineering In addition, pluripotent ES cells are an excellent ex vivo model for studying early embryonic development of mammals (Smith, 2001)

Disclosing the underlying regulatory mechanisms that control ES cell pluripotency and differentiation will greatly advance the use of ES cells for cell-based therapy and also shed light on early embryonic development In recent decades, much effort has been expended to understand ES cell pluripotency and significant progress has been made Several master regulators of ES cell pluripotency, such as Oct4, Sox2, and Nanog, have been described , however, the details remain unclear regarding how these transcription factors execute the molecular processes that maintain ES cell pluripotency (Mitsui et al., 2003; Stanton and Bakre, 2007) (Niwa, 2001) (Boyer et al., 2005b; Brandenberger et al., 2004b) (Avilion et al., 2003b; Wei et al., 2005a) (Kim et al., 2008)

Genome wide mapping of transcription factors’ binding sites in ES cells has improved our understanding in ES cell pluripotency by revealing some of the regulatory circuitry that operates in these cells Recently, many details of a comprehensive transcriptional regulatory network, responsible for ES cell pluripotency, have been reported However, our understanding of this network is still limited My thesis will build upon our current knowledge of the transcriptional regulatory network by mapping the binding sites of additional transcription factors: Zfp206, which was recently identified as a pluripotency regulator (Wang, 2007); and

several cofactors of REST Also, the in vivo role of Zfp206 in mouse development

will be explored

In this chapter, I will review early embryonic development, the origins of ES cells, and our current understanding of the molecular mechanisms regulating ES cell pluripotency and differentiation

1.1 Early embryo development and stem cell fate decision

During development of the early mouse embryo, a series of multipotential cells appear transiently in the embryo Several of these have been isolated and cultured providing convenient sources of stem cells from these early embryos, such as embryonic stem cells (ES cells), trophoblast stem cells (TSC), extraembryonic stem cells (XEN) (Ralston and Rossant, 2005) Somatic (also known as adult or tissue) stem cells, which were derived from a specific late embryo stage or adult tissue

origin, have also been derived such as hematopoietic stem cells (HSC) and neuronal stem cells (NSC) (Weissman et al., 2001) (Rietze and Reynolds, 2006)

Following the fertilization of the oocyte, the zygote develops into a complete embryo, from which different embryo-derived stem cells appear at specific stages (Figure 1.1) The zygote starts the first cell cleavage after 24 hours, and then two more successive rounds of cell division leading to formation of the early morula At this stage, every cell is totipotent, which means that these cells can form all cell types of the organism The first cellular differentiation occurs at the late morula stage with the formation of the trophectoderm and inner cell mass (ICM) Trophectoderm stem cells can be derived from trophectoderm (Tanaka et al., 1998), and ES cells can be derived from the inner cell mass (ICM) (Evans and Kaufman, 1981) (Martin, 1981) The second differentiation event occurs in the late blastula stage with the formation of the epiblast (also called primitive ectoderm) and primitive endoderm (PE) (also called hypoblast) (Rossant, 2007) (Yamanaka et al., 2006) The epiblast will give rise to the whole embryo However the PE, from which XEN cells can be generated experimentally, will become extraembryonic endoderm and finally give rise to the yolk sac, which provides nutrients and patterning information to the embryo during subsequent development (Bielinska et al., 1999)

It is noteworthy that key regulators have been identified for lineage specification during early embryo development The POU domain transcription factor Oct4 and the

caudal-related transcription factor Cdx2 have been identified as key regulators of the first embryo differentiation steps (Nichols et al., 1998) In the absence of Oct4, mice die around the time of implantation, and only trophectoderm cells can be observed (Nichols et al., 1998) (Strumpf et al., 2005) Cdx2 deficient mice die prior to implantation stages due to the lack of trophectoderm development (Strumpf et al.,

2005) In vitro, Oct4 is expressed only in ES cells, not in TS or XEN cells, and disruption of Oct4 leads to differentiation of ES cell into TS cells (Tanaka et al., 1998)

(Kunath et al., 2005) (Niwa et al., 2000; Palmieri et al., 1994) Conversely, Cdx2 is expressed only in TS cells not ES cells (Tanaka et al., 1998) (Niwa et al., 2005)

Nanog and Gata6, two other transcription factors, have been confirmed as key

regulators of the second stage of embryonic differentiation Without Nanog, mouse embryos develop trophectoderm and primitive endoderm normally, but lack epiblast (Mitsui et al., 2003) Gata6 is expressed only in the primitive endoderm and extraembryonic endoderm cells and not epiblast cells (Kunath et al., 2005) Forced expression of Gata6 in ES cells causes them to differentiate into extraembryonic endoderm (Fujikura et al., 2002)

Figure 1.1 Origin of stem cells in the mammalian embryo In this figure, the pluripotent cells of the embryo are tracked in green From left to right, the morula-stage mouse embryo (embryonic day 2.5; E2.5) holds a core of pre-ICM (inner cell mass) cells that turn into ICM cells at cavitation/blastulation (E3–E4) At this stage, embryonic stem cell (ES cell) and trophoblast stem cell

(TSC) lines can be derived in vitro, and implantation occurs in vivo As the blastocyst fully expands (and undergoes implantation in vivo), the ICM

delaminates giving rise to a primitive ectoderm and a primitive endoderm layer

At this stage, pluripotent cell lines that are known as embryonic carcinoma cells (ECCs) can be derived from the primitive ectoderm — whether they are distinct from ES cell has not been resolved At E6 and subsequent stages, the experimental ability to derive ES cells, TSCs and ECCs from the mouse

embryo is progressively lost, and the in vivo embryo will start gastrulating

(Modified from Boiani M., 2005)

1.2 Genomics studies in the embryonic stem cells

With the derivation of ES cells, much effort has been put into their molecular characterization, which is essential for ES cell-based therapy and understanding of the developmental biology Genomics studies have uncovered numerous molecular signatures that are unique to ES cells with high-throughput technologies such as gene expression arrays, massively parallel signature sequencing (MPSS), serial analysis of gene expression (SAGE) and gene identification signature paired-end ditags (Stanton and Bakre, 2007) (Ivanova et al., 2002) (Brenner et al., 2000; Ng et al., 2005; Wei et al., 2004)

Gene expression arrays have been used to identify ES cell specific molecular signatures by comparing differentiated cells with other stem cells (Ivanova et al., 2002) (Fortunel et al., 2003) With MPSS technology, genes involved in the LIF pathway, which is essential for mouse but not human ES cell pluripotency, are found

to be expressed in mouse, but not in human ES cells Also, some common

pluripotent genes were found in both mouse and human ES cells such as Oct4, Sox2 and Nodal etc (Wei et al., 2005a) With the SAGE method, well-known pluripotency transcription factor Nanog was found to be enriched in pluripotent ES cells, and

proved to be a key regulator of ES cell pluripotency (Mitsui et al., 2003)

ES cell pluripotency is maintained by a unique transcription factor network, which can bind to regulatory DNA sequences and direct target gene expression Gene

expression arrays can identify the genes regulated by these transcription factors, but cannot tell whether these genes are direct downstream targets Using the ChIP-chip, ChIP-PET and ChIP-seq technologies, one can now uncover the genome-wide binding sites of transcription factors These technologies have successfully identified Oct4, Sox2 and Nanog transcriptional networks in mouse and human ES cells (Boyer et al., 2005b) (Loh et al., 2006a) Recently, a more detailed and comprehensive mapping of the ES cell transcription network has been reported, which disclose a comprehensive regulatory network in ES cells (Kim et al., 2008) (Chen et al., 2008)

1.3 Molecular mechanisms in regulating ES cell pluripotency and

differentiation

ES cells are derived from the ICM of the developing blastocyst, so they are in vitro

equivalents of epiblast cells, which gives rise to the whole embryo (Evans and Kaufman, 1981; Martin, 1981) (Thomson et al., 1998) Hence, ES cells are a good source for developmental study, especially for human development, which cannot be

studied in vivo due to ethical considerations In addition, ES cells are immortal cells

and can be differentiated into many somatic cells, which strongly suggests their great potential in clinical application To achieve the final goal of clinical application, ES cells need to be completely differentiated into a specific cell type, and all of the immortal cells need to be eliminated, which is currently not possible Hence, it is important to understand the molecular mechanism underlying control of ES cell

pluripotency maintenance and differentiation With a more clear understanding of these molecular details it should be possible to differentiate ES cells into cell types

of clinical utility Some details of the complicated molecular pathways underlying

ES cell pluripotency and differentiation have begun to emerge These include: 1) extrinsic signaling by growth factors such as leukemia inhibitor factor(LIF)/Stat3, bone morphogenetic protein (BMP), wingless-type MMTV integration site (Wnt), and FGF; 2) intrinsic signaling by key transcription factors like Oct4, Sox2, Nanog, Sall4, Zfp206, Zfp281, Zic3 and REST; 3) microRNA such as miR-134 and miR-21; and 4) epigenetic changes through histone modifications mediated by jmjd1a, jmjd2c, NuRD, polycomb group( YY1, Rnf2, and Rybp) and SWI/SNF chromatin remodeling complexes

1.3.1 Extrinsic signaling pathway

The LIF signaling cascade ES cells can only be expanded under proper culture

conditions such as co-culture with a mouse fibroblast feeder layer or in the presence

of added cytokines such as LIF These proper conditions are required or ES cells will undergo differentiation (Gough et al., 1989) (Smith, 2001) (Williams et al., 1988) Standard tissue culture media with typical supplements are not sufficient to support

ES cell self-renewal At first, the co-culture with a mouse fibroblast feeder layer was thought to be essential It was later shown that LIF-deficient fibroblasts are not capable of supporting ES cell self-renewal, which suggested that the LIF production

is a key function of feeder cells (Stewart et al., 1992) Similarly, upon withdrawal of

LIF, the ES cells cannot self-renew beyond a few days, which is due to the effect of LIF and a few cytokines that act via the gp130 receptor (Yoshida et al., 1994) Basically, LIF receptor (LIFR) is activated by LIF to form a heterodimer with gp130, which leads to the activation of Janus tyrosine kinases (Jak) (Ernst et al., 1996) The activated Jak recruits and activates Stat3 Stat3 is a signal transducer and transcription factor, whose over-expression is sufficient to support self-renewal (Matsuda et al., 1999) (Niwa et al., 1998)

However, the dependency for the LIF/gp130/Stat3 pathway is not conserved between mouse and human ES cells Human ES cells do not need LIF for self-renewal

(Reubinoff et al., 2000) Furthermore, in vivo work in the mouse does not agree with the in vitro experimental result that LIF/gp130/Stat3 pathway is essential for ES cell

pluripotency Knockout mice for LIF/LIFR, gp130 and Stat3 retain the capacity to develop past gastrulation (Stewart et al., 1992) (Ware et al., 1995) This suggests that there may be alternative pathways that can maintain ES cell pluripotency

The BMP/Smad signaling pathway The bone morphogenetic proteins (BMPs)

play crucial roles in embryogenesis and tissue homeostasis (Chen et al., 2004) (De Robertis and Kuroda, 2004) (Varga and Wrana, 2005) (Nohe et al., 2004) BMPs have been reported to be able to enhance hematopoitic, mesodermal, and epidermal differentiation in mouse embryos and in mouse ES cells (Wiles and Johansson, 1999) (Adelman et al., 2000) (Adelman et al., 2002) (Kawasaki et al., 2000) Different

concentrations of BMPs have been shown to induce human ES cells to differentiate into trophoblast lineages (Xu et al., 2002).Similarly to LIF, BMP4 is essential for mouse ES cells to maintain pluripotency (Ying et al., 2003a) BMP4 deficient embryos develop past the ICM stage (Fujiwara et al., 2001) BMP signaling appears

to cross-talk with the LIF signaling cascade In the presence of LIF, BMP4 takes a role to enhance mouse ES cell pluripotency by activating the gene encoding SMAD4 (similar to mothers against decapentaplegic homologue-4), which will activate downstream target genes, including members of the Id gene family On the contrary, when cultured without LIF, BMP4 interacts with different SMAD transcription factors such as SMAD1 and SMAD5), which inhibit rather than activate Id genes

As a result, ES cells will differentiate into neuronal lineages (Ying et al., 2003b) Interestingly, ES cells over-expressing Id are able to self-renew in serum free media

in the presence of LIF, which confirm Id as a downstream target of BMP4 signaling pathway (Chen et al., 2004) These results indicate that BMP4 and LIF signaling must be balanced to regulate ES cell pluripotency, though the detailed mechanism of this cross talk remains unclear

Figure 1.2 Combinatorial Signaling pathways regulating the pluripotency of ES cells Cell-surface receptors from three different pathways initiate signals that are conveyed to the nucleus and affect key

pluripotency transcription factors like Oct4, Sox2, Nanog, and Stat3 (Adapted

from Boiani M., 2004)

Wnt signaling The WNT family of proteins are secreted glycoproteins that are

highly conserved and involved in multiple developmental events during early embryogenesis (Cadigan and Nusse, 1997) (Logan and Nusse, 2004) These pathways have been shown to regulate cellular communication between neighboring cells to coordinate cell fate specification, proliferation, differentiation, survival, apoptosis and cellular migration (Logan and Nusse, 2004) The canonical WNT signaling pathway (Figure 1.3) is activated by binding of a WNT ligand to one of several Frizzled receptors on the cell membrane These receptors then transduce a signal to several

intracellular proteins including Dishevelled(Dsh), glycogen synthase kinase-3β

(GSK-3), Axin, Adenomatous Polyposis Coli(APC), and the transcriptional regulator,

β-catenin Cytoplasmic β-catenin levels are normally kept low through continuous

proteasome-mediated degradation, which is controlled by a complex containing GSK-3/APC/Axin When cells receive Wnt signals, the degradation pathway is

inhibited, and consequently β-catenin accumulates in the cytoplasm and nucleus The accumulated β-catenin will translocate to the nucleus and associate with T-cell

factor/lymphoid enhancer factor (Tcf/Lef) transcription factors, and regulate target genes expression

Figure 1.3 The canonical Wnt signaling pathway. In cells not exposed to

a Wnt signal(left panel), β-catenin is degraded through interactions with Axin, APC, and the protein kinase GSK-3 Wnt proteins (right panel) bind to the Frizzled/LRP receptor complex at the cell surface These receptors transduce

a signal to Dishevelled (Dsh) and to Axin, which may directly interact (dashed lines) As a consequence, the degradation of β-catenin is inhibited, and this protein accumulates in the cytoplasm and nucleus β-catenin then interacts with TCF to control transcription Negative regulators are outlined in black Positively acting components are outlined in color (Adapted from Logan, C.Y., 2004)

It has been reported that the WNT signaling pathway plays a role in maintaining ES cell pluripotency Sato has demonstrated that the activation of the WNT signaling pathways by GSK3β inhibitor treatment can maintain both mouse and human ES cell pluripotency (Sato et al., 2004) In addition, genetic modification of WNT signaling in

ES cells by either inactivation of the APC complex or introducing a dominant negative β-catenin will inhibit neural differentiation (Haegele et al., 2003) Furthermore, activation of the WNT pathway downstream target genes that include

cyclins and c-Myc showed a similar effect (He et al., 1998) Interestingly, Tcf3, which

is a downstream target of WNT pathway, has recently been shown to negatively regulate ES cell pluripotency (Tam et al., 2008) (Yi et al., 2008)

Fibroblast growth factor (FGF) signaling FGF family members are highly

conserved in gene sequence and structure, and FGF signaling is important in mouse embryo development Interference with the FGF signaling pathway by disruption of FGFRs causes severe defects in early mouse embryo development (Deng et al., 1994) (Yamaguchi et al., 1994) In brief, FGFs activate the FGFRs tyrosine kinase through binding and subsequent receptor dimerization (McKeehan et al., 1998) The FGF signal can be transduced by three signaling kinase cascades: the Ras/MAP kinase pathway, the PLCy/Ca2+ pathway and the PI3K pathway (Figure 1.4)

Different from mouse ES cells, human ES cells require FGF signaling for maintaining pluripotency Transcriptome interrogation of human ES cells has shown that the FGF

signaling pathway is enriched in undifferentiated human ES cells (Brandenberger et al., 2004c) Inhibition of the FGF signaling pathway induces human ES cell differentiation into a flattened phenotype further confirming that FGF signaling is essential to maintain human ES cell pluripotency (Dvorak et al., 2005) Some downstream targets of FGF signaling such as PI3K have been shown to be involved in pluripotency (Pyle et al., 2006), although the details of this remain unclear

Figure 1.4 Intracellular signaling pathways activated through FGFRs The activated FGF pathway involves three signaling cascades: the Ras/MAP kinase pathway (in blue), PI3 kinase/Akt pathway (in green) and the PLCγ/Ca2+ pathway (in yellow) Proteins connected with two pathways are striped Members of the FGF synexpression group are illustrated in red

(Adapted from Bottcher, 2005)

1.3.2 Intrinsic determination of pluripotency

In order to maintain ES cells pluripotency, finely tuned regulation of intrinsic signaling pathways by key transcription factors, such as Oct4, Sox2, Nanog, Sall4, Zfp206 and REST are essential

Oct4 and Sox2 Oct4 (encoded by Pou5f1) is one of the master regulators of ES cell

pluripotency It is a transcription factor that recognizes a highly conserved octamer

element (Scholer et al., 1990) (Chew et al., 2005a) Its expression pattern in vitro and

in vivo indicate its essential role in ES cell pluripotency In vitro, Oct4 is highly

expressed in undifferentiated pluripotent cells like ES cells, EG cells, EC cells but

disappears upon differentiation (Nichols et al., 1998) In vivo, its expression was

restricted to blastomeres, pre-implantation stage, PGCs and the ICM of blastocysts

(Scholer et al., 1990) (Rosner et al., 1990) (Yeom et al., 1996) In vitro depletion of Oct4 in ES cells leads to dramatic differentiation, which proves its importance in ES cell pluripotency (Loh et al., 2006b) In vivo knockout of Oct4 leads to disruption of

ICM formation, which further confirms its role in ES cell pluripotency Interestingly,

the loss of Oct4 in the embryo leads to the expression of a caudal type transcription

factor Cdx2 necessary for specifying trophoblast stem (TS) cell fate as well as the formation of the trophectoderm (Nichols et al., 1998) (Niwa et al., 2005) (Strumpf et al., 2005)

The HMG-box transcription factor Sox2, a partner of Oct4, is another essential master regulator of ES cell pluripotency which recognizes a sox-binding element (Yuan et al., 1995) (Nishimoto et al., 1999) (Chew et al., 2005b) Sox2 has a similar role in ES cell

pluripotency in that depletion of Sox2 both in vitro and in vivo lead to trophoblast

lineage differentiation (Chew et al., 2005b) (Avilion et al., 2003b) (Boiani and Scholer, 2005) Surprisingly, Sox2 and Oct4 not only physically interact with each other by forming a heterodimer, which has been shown to be essential for ES cell pluripotency regulation, but also their binding sites are very close to each other for

many targets (Chew et al., 2005b; Nishimoto et al., 1999; Yuan et al., 1995)

Nanog Nanog is a homeodomain-containing transcription factor, and is expressed

exclusively in pluripotent ES, EG and EC cells, but not in differentiated or adult tissues and stem cells (Chambers et al., 2003) (Ramalho-Santos et al., 2002) (Wang et al., 2003) Nanog was reported as a pluripotency factor at the same time by two independent groups (Chambers et al., 2003) (Mitsui et al., 2003) Using functional expression cloning, Chambers found that Nanog over expression confers ES cells with the ability to self-renew and maintain pluripotency in a LIF-independent manner

(Chambers et al., 2003) Using digital differential display, Mitsui found that Nanog

was highly enriched in ES cells Consistent with Chamber’s observation, ES cells

over expressing Nanog are resistant to LIF withdrawal (Mitsui et al., 2003)

Upon Nanog knockout ES cells lose pluripotency and differentiated into the

extraembryonic endoderm lineage, which further confirms its essential role in ES cell

pluripotency (Chambers et al., 2003; Mitsui et al., 2003) Nanog deficient embryos show a lack of epiblast and primitive ectoderm formations, which suggests that Nanog

is required for maintaining ICM pluripotency along with Oct4 and Sox2 (Mitsui et al.,

2003)

Sall4 Since the identification of key pluripotency factors such as Oct4, Sox2 and

Nanog, people are trying to identify upstream regulators as well as downstream targets of the three key pluripotency factors Sall4 was identified as an upstream

regulator of Oct4 in an RNAi screen for genes that regulate Oct4 (Zhang et al., 2006a) Sall4 is a zinc finger transcription factor from the spalt family Sall4 has a similar expression pattern as Oct4, Sox2 and Nanog in the early stage embryo development (Hamatani et al., 2004) (Zhang et al., 2006a) Sall4 directly binds to the Oct4 promoter and regulats Oct4 gene expression in ES cells Similar to the loss of Oct4, depletion of Sall4 triggers ES cells to preferentially differentiate into trophoblast cells

(Zhang et al., 2006a) Sall4 also interacts with Nanog and regulates its expression

through direct co-occupancy at Nanog’s genomic regions (Wu et al., 2006) Recent

genome-wide mapping of Sall4 binding sites suggested that it has different roles in ES and extraembryonic stem (XEN) cells In ES cells, Sall4 is essential for maintaining pluripotency by forming a transcriptional regulation network with Oct4, Sox2 and

Nanog In XEN cells, Sall4 maintains self-renewal by repressing certain lineage genes

such as Gata4, Gata6, Sox7, and Sox17 (Lim et al., 2008)

Zfp281 Zfp281, a zinc finger transcription factor, is essential for ES cell pluripotency

It was first implicated as an ES cell pluripotency regulator because it is highly expressed in ES cells but down regulated upon differentiation (Brandenberger et al.,

2004a) (Wei et al., 2005b) In addition, Zfp281 has been identified as an Oct4, Sox2,

and Nanog downstream target (Chen et al., 2008) Its role as a pluripotency regulator was further confirmed when it was pulled down with the Nanog complex (Wang et al., 2006)

Recently, the knockdown of Zfp281 was shown to induce ES cell differentiation,

which is a direct evidence for its role in ES cell pluripotency (Wang et al., 2008) Promoter activity assays and ChIP experiments have shown that Zfp281 regulates

Nanog expression by direct binding to its promoter region Zfp281 has been shown to

physically interact with Oct4, Sox2, and Nanog Furthermore, Zfp281 shares a large set of common targets with Oct4, Sox2 and Nanog Taken together, Zfp281 is an integral component of the Oct4, Sox2 and Nanog regulatory network (Wang et al., 2008)

Zic3 Zic3 (Zinc finger protein of the cerebellum 3) is another ES cell specific

transcription factor Zic3 has been shown to play an essential role in mouse embryo development Zic3 is widely expressed in a lot of tissues such as embryonic ectoderm

and mesoderm and throughout the tailbud, retina and limb bud (Kitaguchi et al., 2002)

(Elms et al., 2004) (Orkin, 2005) Zic3 mutations and knockout have exhibited a wide

spectrum of abnormal phenotypes (Gebbia et al., 1997) (Purandare et al., 2002) (Aylsworth, 2001)

In vitro, Zic3 is highly expressed in both mouse and human pluripotent ES cells,

while down regulated significantly upon differentiation treatment, suggesting its role

in ES cell pluripotency (Brandenberger et al., 2004c) (Wei et al., 2005b) Zic3 has

been shown to be a regulated by Oct4, Sox2 and Nanog by direct binding to its promoter (Boyer et al., 2005a) (Loh et al., 2006a) Our lab has characterized Zic3 as

an essential ES cell pluripotency regulator active in repressing specifically endoderm

lineage differentiation (Lim et al., 2007)

Esrrb Esrrb (estrogen-related-receptor), a member of the nuclear orphan receptor

family, is essential for ES cell pluripotency Knockdown of Esrrb with shRNA or

siRNA has been shown to induce differentiation of mouse ES cells in the presence of

LIF (Ivanova et al., 2006) (Loh et al., 2006b) It has been shown that Esrrb and Oct4

reciprocally regulate each other to maintain ES cell pluripotency (Zhang et al., 2008) The pull-down of Esrrb with Nanog further confirmed it as an integral component of the core pluripotency circuitry (Wang et al., 2007a)

Recently, Esrrb was reported to be able to replace Klf4, one of the 4 reprogramming factors in iPS induction Esrrb was also shown to be able to rescue the differentiation

induced by triple knockdown of Klf2, Klf4, and Klf5 (Feng et al., 2009) Genome

wide mapping of Esrrb binding sites has shown it preferentially co-localize with Klf4

in a large set of pluripotent gene targets These results further confirmed its role in ES cell pluripotency (Feng et al., 2009)

Tcf3 Tcf3 (T-cell factor 3), a member of the canonical Wnt signaling pathway, has

been shown to be involved in maintenance of mouse ES cell pluripotency, as its loss

delays the ability of these cells to differentiate via the relief of repression of Nanog (Pereira et al., 2006) Tcf3 was also reported to transcriptionally repress Oct4 in ES cells, and loss of Tcf3 by RNA interference (RNAi) knockdown blocks the ability of

ES cells to differentiate (Tam et al., 2008)

Genome wide mapping of Tcf3 binding sites showed that its targets are associated with stem cell pluripotency, developmental processes, signaling pathways, and

oncogenesis Tcf3 is up-regulated upon differentiation, and its target genes are

activated during differentiation (Tam et al., 2008) A co-occupancy analysis of Tcf3 and Oct4, Sox2 and Nanog core circuitry showed that Tcf3 was an integral component

of it The depletion of Tcf3 showed a similar resistant effect to differentiation

treatment as Wnt pathway activation, suggesting Tcf3 is an important mediator of the

balance between pluripotency and differentiation (Cole and Young, 2008)

1.3.3 MicroRNAs in ES cell pluripotency

miRNAs are 20–25 nucleotide (nt) non-coding RNAs that bind to the 3’- untranslated regions (UTR) or coding regions of target mRNAs through imperfect base-pairing and repress or activate their translation and stability (Rana, 2007) (Tay et al., 2008) Recently, miRNAs have emerged as important regulators of ES cell pluripotency Loss of either Dicer or DGCR8, which are essential for miRNA processing, renders

ES cells incapable to differentiate properly (Tomari and Zamore, 2005) (Bernstein et

al., 2003) In vivo, Dicer deficient mice die at early embryonic stage (Bernstein et al.,

2003)

Studies in both human and mouse ES cells have shown that pluripotent ES cells have distinct miRNA signatures (Houbaviy et al., 2003) (Suh et al., 2004) For example, the miR-290-295 cluster and miR-296 are highly expressed in undifferentiated ES cells, but down-regulated upon differentiation, suggesting these miRNAs are involved

in pluripotency (Houbaviy et al., 2003) In contrast, miR-21, miR-22 and miR-134 are

significantly up-regulated upon differentiation, and miR-21 and miR-134 have been shown to drive ES cell differentiation (Singh et al., 2008) (Tay et al., 2008) Genome

wide mapping of the core transcription factors, Oct4, Sox2 and Nanog, also revealed a

group of miRNA targets that may be activated for proper maintenance of pluripotency

or derepressed for appropriate differentiation (Boyer et al., 2005a)

The importance of miRNAs in ES cells has been realized, but their mechanism of action is still unclear Further studies of these miRNAs in ES cell pluripotency and differentiation regulation is needed

1.3.4 Epigenetic modification in ES cells

ES cells must activate a set of genes to maintain pluripotency and silence other genes

to prevent differentiation This is achieved, in part, by modification of chromatin through the interplay of transcription regulators and histone modifiers ES cells have been shown to maintain a hyperdynamic state of chromatin compared with differentiated cells (Meshorer and Misteli, 2006) Epigenetic studies may help to elucidate the mechanisms underlying ES cell pluripotency Epigenetic modifications include DNA methylation and histone modification DNA hypermethylation, especially in CpG Islands, has been found to be associated with long-term gene silencing in differentiated cells (Klose et al., 2006) (Ng et al., 1999) ES cell

differentiation is coupled with hypermethylation of the Oct4 and Nanog promoter regions as well as Oct4 and Nanog silencing (Yamanaka et al., 2006) (Yeo et al.,

2007) Histone modifications include acetylation, methylation, phosphorylation, and ubiquitinylation, which are important for gene regulation (Strahl and Allis, 2000) A comprehensive study by Mikklesen has shown that tri-methylation of histone H3 lysine 9(H3K9), histone H4 lysine 20(H4K20) and histone H3 lysine 27(H3K27) is always associated with gene silencing, but trimethylation of histone H3 lysine 4(H3K4) is always associated with active gene expression (Mikkelsen et al., 2007) Genome wide mapping has shown that key pluripotency factors like Oct4, Sox2 and Nanog bind to a large set of common targets including active and inactive genes (Boyer

et al., 2005b) (Loh et al., 2006b) Recently, it has been shown that Oct4 and Nanog recruit some repression complexes such as NuRD, Sin3A and Pml in ES cells to repress

differentiation These repressors are able to modify chromatin structure for gene silencing Taken together, we can see that chromatin structures play critical roles in the activation and repression of gene expression This is further exemplified by the demonstration that two histone demethylase genes, Jmjd1a and Jmjd2c, are essential for ES cell pluripotency (Loh et al., 2007)

1.3.5 Transcriptional regulatory network in ES cells

A complicated transcriptional regulatory network is necessary for ES cell

pluripotency Initially, Oct4, Sox2, Nanog were thought to be the primary regulators

of ES cell pluripotency, which not only form auto-regulatory feedback loop but also regulate each others (Tomioka et al., 2002) (Rodda et al., 2005a) (Chew et al., 2005a) Surprisingly, genome wide mapping of the binding sites of these three transcription factors in ES cells with ChIP-chip and ChIP-PET shows they co-localize significantly

to large groups of common target genes, thus forming a core transcriptional regulatory network, which include themselves, developmentally important genes, chromatin remodeling genes, ES cell specific transcription factors, and various lineage specific genes (Figure 1.5) (Boyer et al., 2005b) (Loh et al., 2006b) More recently, a larger protein interaction network, which includes Oct4, Sall4, Zfp281, NacI, Dax1 and Rif1, was identified biochemically by pull down of Nanog protein complexes (Wang et al., 2006)

Figure 1.5 Core Transcriptional Regulatory Network in Human ES Cells

A model for the core transcriptional regulatory network was constructed by identifying Oct4, Sox2, and Nanog target genes that encode transcription factors and chromatin regulators and integrating knowledge of the functions of these downstream regulators based on comparison to multiple expression datasets (Supplemental Data) and to the literature A subset of active and inactive genes co-occupied by the three factors in human ES cells is shown here

Regulators are represented by blue circles; gene promoters are represented

by red rectangles; gray boxes represent putative downstream target genes Positive regulation was assumed if the target gene was expressed whereas negative regulation was assumed if the target gene was not transcribed (Adapted from Boyer, 2006)

Recently, Kim et al reported a comprehensive genome wide mapping of nine transcription factors in ES cells with ChIP-chip (Kim et al., 2008) Similarly, Chen et

al published a comprehensive genome wide mapping of 14 transcription factors in ES cells using ChIP-seq technology, which includes transcription factors involved in external signaling pathways such as Smad1 and Stat3, ES cell pluripotency factors such as Oct4, Sox2, Nanog, Klf4 and Esrrb, and two transcription regulators, P300

and Suz12 (Chen et al., 2008) In the Kim et al study, seven pluripotency transcription

factors’ binding targets were classified as one group and two others as another group, indicating these transcription factors form a pluripotency network They also found that genes bound with multiple transcription factors are generally active (Kim et al., 2008) This is consistent with Chen’s observation where they found a subset of multiple transcription factors binding loci act as ES cell enhanceosomes (Chen et al., 2008) Interestingly, the 13 transcription factors form a comprehensive transcriptional regulatory network (Figure 1.6), and their binding sites were classified into two groups: 1) Oct4, Sox2, Nanog, Smad1, and Stat3, suggesting the integration of the signaling pathways with ES cell key pluripotency factors; and 2) c-Myc, n-Myc, Zfx, and E2f1 Furthermore, the p300 coactivators are preferentially recruited to binding

sites of the first group (Chen et al., 2008)

Figure 1.6 Transcriptional regulatory networks in ES Cells. Network of regulatory interactions inferred from ChIP-seq binding assays and from gene-expression changes during differentiation Nodes are ChIP-seq-assayed transcription factors Arrows point from the transcription factor to the target gene Two sets of published experiments were used to define genes that are differentially expressed during differentiation (Ivanova et al., 2006; Zhou et al., 2007) Thick arrows represent interactions inferred from binding data and both expression experiments, whereas thin arrows represent interactions inferred from binding data and only one of the expression experiments Regulatory targets were inferred from the intersection of top-ranked bound genes and top-ranked differentially expressed genes Thresholds for defining top-ranked genes in the two lists were determined empirically by maximizing the number

of genes in the intersection, subject to two constraints: the p value for the enrichment of genes in the intersection had to be 0.001 or better, and there had to be at least twice as many genes in the intersection as expected All regulatory interactions in this network involve higher-level expression in ES cells and lower-level expression during differentiation There were no interactions among the factors in this network when regulation in the opposite direction was evaluated The network was drawn by using cytoscape (Adapted from Chen, 2008)

1.4 Zfp206

Zfp206 encodes a zinc finger transcription factor with a conserved SCAN domain at its N-terminal It has been recently reported that Zfp206 is a regulator of ES cell

pluripotency (Zhang et al., 2006b) (Wang et al., 2007b) I will review the structural

features as well as the function of Zfp206 in this section

1.4.1 The SCAN domain family of zinc finger transcription factors

Approximately 2% of all human genes encode zinc finger proteins (Tupler et al.,

2001) The zinc finger was first reported in 1985 in Xenopus transcription factor IIIA

(TFIIIA) (Miller et al., 1985) The classical zinc finger contains a zinc-binding, Cys2His2 (C2H2) motif (Pavletich and Pabo, 1991) Each finger is composed of 30 amino acids, which coordinate a zinc ion and fold into a compact structure containing β-turn (including cysteines) and α-helix (histidines) (Figure 1.7) The zinc finger motif has been found to mediate DNA or RNA binding, protein-protein interaction and membrane association