Identifying protein co factors of oct4, an essential stemness transcription factor, by affinity purification and mass spectrometry

Part of this feature of ES cells is dependent upon the expression levels of Oct4 within the cell – beyond a given range; ES cells do not retain their pluripotency, but begin to display s

IDENTIFYING PROTEIN PARTNERS OF OCT4,

AN ES-CELL SPECIFIC TRANSCRIPTION FACTOR,

BY GENE-TAGGING & PROTEOMICS APPROACHES

B.A (Cum Laude), Cornell University

A T HESIS S UBMITTED F OR T HE D EGREE O F

D OCTOR O F P HILOSOPHY

NUS GRADUATE SCHOOL FOR INTEGRATIVE SCIENCES AND ENGINEERING

2008

A CKNOWLEDGEMENTS

Is there anything of which one can say,

"Look! This is something new"?

It was here already, long ago;

It was here before our time

Ecclesiastes 1:10

To finally complete this work takes a village

Many thanks to my committee, especially Dr Thomas Lufkin, my main supervisor, for research

direction, patience and keeping things in perspective while I was exploring and scrabbling things

together

Drs Tsai Heng Hang, Patricia Ng & Rhonda Ponnampalam have been instrumental in various

ways, especially with protein work, mass spectrometry, and random jokes

To members of the Lufkin lab past and present, especially Mathia, Sumantra, Serene, Song Jie,

Yun, Gerry, Val, Max, Sook Peng, thank you for the scientific and non-scientific talk, coffee and

muffins, and for making the past few years go by in a flash I‟ve really enjoyed myself at lab

Others in GIS have provided timely assistance and guidance I‟d like to thank Steph for handling

the MS samples, Yun for the microinjections and Vega for walking us patiently through various

data analyses methods

To my family, for loving me and dealing with my ups and downs, and believing that I could

finish more than I myself thought I could Same to DP, for love, patience, and all those fun

surprises that kept me sane

Thanks God, for keeping me very well and alive

T ABLE OF C ONTENTS

Acknowledgements ii

Table of Contents iii

Summary ix

List of Tables & Figures xi

Abbreviations xiii

1 Introduction 15

1.1 Transcriptional Regulation in Mammalian Development 15

1.1.1 Gene-Specific Transcription Factors 15

1.1.1.1 Mechanism of Action 15

1.1.1.2 Helix-Turn-Helix Domain Transcription Factors & Octamer Proteins 16

1.1.2 DNA Binding Sites of Transcription Factors 17

1.1.2.1 Analysis of Transcription Factor Binding Sites 17

1.2 Early Mouse Development 19

1.2.1 ES Cells Are a Model for ICM Pluripotency 19

1.2.2 Pluripotency is controlled on multiple fronts 21

1.2.2.1 Known Signaling networks 21

1.2.2.2 Protein-Protein Interactions 24

1.2.2.3 Epigenetic Regulation of Pluripotency 24

1.2.3 Key Transcription Factors Regulate Pluripotency 27

1.2.3.1 Oct4/Pou5f1 27

1.2.3.2 Sox2 28

1.2.3.3 Nanog 28

1.2.4 Large scale studies on pluripotency 29

1.3 Role of Oct4 in Maintenance of Pluripotency 34

1.3.1 Regulation of Oct4 Expression 35

1.3.2 Oct4 Structure & Domains 36

1.3.3 Known Protein Interaction Partners of Oct4 37

1.4 Finding protein-protein interactions 44

1.4.1 Early Affinity Chromatography 44

1.4.2 Immunoprecipitation and Co-Immunoprecipitation 45

1.4.3 Affinity Purification/Mass Spectrometry 46

1.4.3.1 Tandem Affinity Purification 49

2 Project Goals 52

2 Chapter: Materials & Methods 53

2.1 DNA Manipulation 53

2.1.1 Plasmids 53

 Targeting Plasmids 53

2.1.2 Bacterial Strains & Antibiotics 54

2.1.3 Genomic DNA Extraction 54

 ES Cell & Mouse Tail Tip Genomic DNA 54

2.1.4 Southern Blotting 55

 DIG Probe Design 55

 Hybridization & Washing 55

 Detection 56

2.2 RNA Manipulation 57

2.2.1 RNA Extraction from ES Cells 57

2.2.2 RNA to cDNA Reverse Transcription 57

2.2.3 siRNA Knockdown 57

2.2.4 Illumina Bead Chip Gene Expression Assay 58

 RNA Amplification 58

 Illumina Bead Chip Hybridization & Data Analysis 58

2.3 Protein Manipulation 59

2.3.1 Protein Extraction 59

 Total Protein Extraction 59

 Nuclear/Cytoplasmic Protein Extraction 59

2.3.2 Affinity Purification of Protein Complexes 60

 His6 Tag 60

 Flag Tag 60

 S Tag 61

 CBP Tag 61

 BAP Tag 63

2.3.3 Buffer Exchange & Desalting of Proteins 64

2.3.4 Acetone Precipitation 64

2.3.5 Detection of Proteins 64

 Western Blotting 64

 Primary Antibodies 65

 Secondary Antibodies 65

 Coomassie Blue and Silver Staining 66

2.3.6 Mass Spectrometry 66

 Precautions Against Keratin Contamination 66

 1D Gel Separation 66

 In Gel Digestion 67

 LC-MS/MS 68

 Peptide and Protein Identification 68

 Co-Immunoprecipitation 68

2.4 Tissue Culture 69

2.4.1 Embryonic Stem Cells 69

 ES Cell Maintenance 69

 Electroporation of ES cells 69

 Homologous Recombination Targeting Vectors 70

 Cre-expressing Vector 70

 Alkaline Phosphatase (AP) Staining 70

 lacZ Staining Protocol 70

2.4.2 HEK-293 Cells 71

 Cell Culture & Transfection 71

2.5 Animal Work 71

2.5.1 Blastocyst Microinjection 71

2.5.2 Genotyping by PCR 72

3 Epitope Tagged-Oct4 Embryonic Stem Cells 73

3.1 Introduction 74

3.2 Design and Generation of Tagged Mouse Embryonic Stem Cells 74

3.2.1 Choice of Affinity Purification Tags 74

3.2.1.1 Tag Size 76

3.2.1.2 Binding affinity and elution 77

3.2.1.3 Localization of tags 77

3.2.1.4 Orthogonal Tandem Purifications 81

3.2.2 Design and construction of targeting vectors 81

3.2.2.1 Endogenous tagging of bait by homologous recombination in mouse ES cells 81

3.2.2.2 Design of N and C-terminal targeting cassettes 83

3.2.2.3 BAC Recombineering for generation of targeting constructs 84

3.2.2.4 Generation of tagged-Oct4 ES cell lines 86

3.2.2.5 Tagged-Oct4 Expression on Removal of Antibiotic Selection Cassette 88

3.2.2.6 NBH-Oct4 ES Cells require an additional targeting step 91

3.2.2.7 Generation of Tagged-Oct4 Mice Evaluate the Effect of Tag on Oct4 Function 93 3.2.3 Discussion 96

4 Affinity Purification 100

4.1 Basis of Affinity Purification 100

4.2 Protein Extraction 102

4.3 Histidine6 (His) Purification 103

4.4 FLAG (F) Purification 108

4.5 Biotin Acceptor Peptide (BAP) Purification 110

4.6 Calmodulin Binding Peptide (CBP) Purification 114

4.7 S-Tag (S) Purification 116

4.8 Tandem Affinity Purifications 117

5 Preparing Proteins for Mass Spectrometry 121

5.1 Identification of Purified Proteins by Mass Spectrometry 121

5.1.1 How MS and MS/MS work 121

5.1.2 Preparing Proteins for Mass Spectrometry 122

5.1.3 Peptide Ionization Methods 122

5.1.4 Protein Identification 123

5.2 Improving Peptide Identification by Sample Preparation - Strategies for protein complex preparation 124

5.2.1 Enriched Oct4 complexes: Multiple Friends and Freeloaders 124

5.2.2 In solution tryptic digestion & LC-Separation 125

5.2.3 Gel Separation and In-gel tryptic digestion 126

5.2.4 1D SDS-PAGE and in-gel digestion is appropriate for low abundance complexes 127 5.2.5 Future improvements at parallel processing require less labor intensive approaches 128 5.2.6 Mass Spectrometry by LC-ESI-MS/MS 129

6 Identifying Proteins by Tandem Mass Spectrometry 130

6.1 Introduction 130

6.1.1 Raw Spectra to Peptide Identification 130

6.1.2 Peptide Identification by de novo sequencing 130

6.1.3 Peptide Identification by hybrid approaches 131

6.1.4 Peptide Identification by database searches 132

6.2 From Peptides to Proteins 132

6.2.1 Choice of Sequence Database and Peptide Modifications 133

6.2.2 Two heads are better than one 133

6.2.3 Initial Search – SEQUEST 134

6.2.4 Moving on – Scaffold 134

6.3 Assessing Purifications by Mass Spectrometry 139

6.3.1 Problem of common proteins 139

6.3.2 Semi-Quantitative Proteomics by Spectral Counting 141

7 Discovering Putative Interactors of Oct4 143

7.1 Identification of Proteins 144

7.2 Semi-quantitative Proteomics Revisited 145

7.3 Proteins Unique to the Tagged-Oct4 Samples are Putative Oct4 Interactors 149

7.4 Proteins Common to Tagged-Oct4 and Wildtype ES Cell Samples 153

7.4.1 Finding a normalizer 153

7.4.1.1 Normalization of Total Spectral Counts 154

7.4.1.2 Normalization with a Known Non-interactor Protein 155

7.4.2 Alternatively Speaking - Finding Proteins That Trend with Oct4 156

7.4.3 Additional Known Oct4 Interactors are Found by Correlation 158

7.4.4 New Oct4 Interactors are Discovered by the Correlation Method 160

7.5 Co-Immunoprecipitation Analyses 164

7.5.1 Krueppel-like Factor 5 (Klf5) 164

7.5.2 Estrogen-related receptor beta (Esrrb) 165

7.5.3 Lysine-specific Histone Demethylase (Lsd1)/Amine Oxidase (Flavin-containing) Domain 2 (Aof2) 168

7.6 Oct4 is associated with transcriptional regulators 170

7.6.1 Oct4 as a repressor 170

7.6.1.1 Oct4 is involved in transcriptional silencing through BHC complexes 170

7.6.1.2 Oct4 and members of NuRD complexes 172

7.6.2 Oct4 as an Activator of Gene Expression 173

7.7 Discussion 174

8 Conclusion and Future Perspectives 180

8.1 Conclusion 180

8.2 Future Applications 182

8.2.1 Extended Application 1: Examining the DNA bound protein complex 182

8.2.2 Extended Application 2: Chromatin Immunoprecipitation using Epitope Tag Antibodies 183

8.2.3 Extended Application 3: Finding Oct4 partners in different cellular contexts 184

9 Appendices 186

10 References 197

S UMMARY

The demands of embryonic development require tight transcriptional regulation in order to

manage multiple signals and outcomes in the organism Sequence-specific transcription factors

play central roles in this coordinative procedure, acting as a molecular switchboard for the control

of expression While ongoing developments have facilitated extensive studies of protein-DNA

interactions, the protein-protein interactions that surround these central players can provide

fundamental clues to the recruitment of factors necessary for transcriptional regulation

Oct4 is a key factor governing the pluripotency of embryonic stem cells, which are feted for their

capacity to both self-renew and differentiate to all cell lineages of the embryo proper Part of this

feature of ES cells is dependent upon the expression levels of Oct4 within the cell – beyond a

given range; ES cells do not retain their pluripotency, but begin to display signs of differentiation

to specific lineages To gain perspective on the role of Oct4 through its protein interactions, I

have engineered ES cell lines expressing epitope tagged Oct4 from the endogenous locus at

biologically relevant levels These tagged-Oct4 cell lines were used for affinity purification to

enrich for Oct4 complexes via the epitope tag, and isolated proteins then identified by mass

spectrometry Known partners of Oct4 were reaffirmed in this work; in addition, new interactions

were established which pointed towards a spectrum of roles for Oct4 in transcriptional regulation

The establishment of this serves a dual purpose – one is the realization of protein interactions

mediated by a transcription factor central to the maintenance of pluripotency This study is the

first to demonstrate an extensive breadth of interactions, which are validated in part by other

experimental approaches both in this work, and in published studies, and serve as a platform for

further insights into each of the complexes served by Oct4

Additionally, the prevailing knowledge surrounding Oct4 provided this study with guiding posts

in developing this technique as a generic technology suitable for the discovery of protein-protein

interactions from less abundant proteins such as transcription factors Epitope tagged proteins

generated by the knock-in of tags to the endogenous locus are versatile in purpose, and

higher-throughput studies can be made with a common optimized protocol for multiple proteins with

similar tags The biological context of these interactions is not sacrificed because of endogenous

expression levels Furthermore, tagged cells derived by this approach can be used for the

generation of transgenic mice, to obtain previously inaccessible cell and tissue samples for the

discovery of protein-protein interactions

L IST OF T ABLES & F IGURES

Figure 1 Origin of embryonic stem cell lines 21

Figure 2 Atomic Model of the IFN-β Enhanceosome 32

Figure 3 Overview of Experimental Approach 73

Figure 4 N-terminal Affinity Tag Combinations 80

Figure 5 Targeted epitope tag insertion at the 5' and 3’ end of the Oct4 (Pou5f1) locus 84

Figure 6 BAC Recombineering Steps 86

Figure 7 Southern Blot Screening for Positive Homologous Recombinants 87

Figure 8 Cre-mediated Removal of Antibiotic Selection Cassette 89

Figure 9 Expression of Tagged Oct4 Following Removal of Neo Cassette 90

Figure 10 Alkaline Phosphatase Staining of Tagged Oct4 ES Cell Lines 91

Figure 11 Targeted insertion of hBirA at the Rosa26 locus 92

Figure 12 lacZ Positive Staining of NBH-Oct4 ES cells with hBirA 93

Figure 13 Biotinylation of the BAP tag with hBirA 93

Figure 14 Schematic of a typical affinity purification 101

Figure 15 Evaluating His Purification Steps 105

Figure 16 His Purification for NBH-Oct4 complexes 106

Figure 17 His Purification for NBH-Oct4 and NFH-Oct4 complexes 107

Figure 18 FLAG Purification for recovery of NFH-Oct4 complexes 109

Figure 19 Activation of BAP tag 111

Figure 20 Retrieving NBH-Oct4 Complexes Off The Beads 113

Figure 21 CBP Purification for NSC-Oct4 complexes 115

Figure 22 Capture of NSC-Oct4 complexes with an S purification 117

Figure 23 Tandem Purification with NBH-Oct4 ES cells 119

Figure 24 Possible Approaches to Identification of Protein Complex Components 124

Figure 27 Approach Taken to Further Reduce Sample Complexity Before MS 128

Figure 28 A Single Search Method Misses Real Proteins 135

Figure 29 Improved Protein Identification with Scaffold 136

Figure 30 Finalized Workflow to Obtain High Confidence Protein Lists 138

Figure 31 The Majority of Proteins Identified in Tagged and Mock Purifications Overlap 140

Figure 32 Enriched Proteins Are Not Readily Distinguished by Gel Staining 140

Figure 33 Flowchart for Identification of Putative Oct4 Interactors 143

Figure 34 Confident Protein Identifications 145

Figure 35 What are Spectral Counts? 147

Figure 36 Linear Relationship between SCxC and Protein Abundance 148

Figure 37 Enriched GO Terms Associated with Oct4 Interactors Found Exclusively in the Tagged-Oct4 Sample 150

Figure 38 Distinguishing Oct4-trending Proteins from Noise 158

Figure 39 Enriched GO Terms among Putative Oct4 Interactors found by Correlation Method 161

Figure 40 CoIP; Oct4 and Klf5 165

Figure 41: Co-IP: Oct4 and Esrrb 167

Figure 42 CoIP; Oct4 and Lsd1/Aof2 169

Figure 43 Distribution of Proteins Identified in this Study 176

Figure 44 Mini-Interactome of Oct4 in ES Cell Biology 178

Table 1 Oct4 Acts as a Transcriptional Activator 38

Table 2 Oct4 Acts as a Transcriptional Repressor 40

Table 3 Additional Oct4 Interactors 42

Table 4 Overview of Epitope Tags 76

Table 5 Considerations for Tandem Tag Combinations 79

Table 6 ES cell targeting frequencies and Tagged ES lines obtained 88

Table 7 Mice Obtained from Oct4 +/TAG Matings 95

Table 3 Select Putative Oct4 Interactors Uniquely Identified in the Tagged Sample 152

Table 4 Select Oct4 Interactors Identified by the Correlation Method 163

A BBREVIATIONS

E (e.g E3.5) Embryonic day (mouse)

emPAI Exponentially Modified Protein Abundance Index

FACS Fluorescent activated cell sorting

His, His6 Histidine-6 (tag)

MALDI Matrix Assisted Laser Desorption Ionization

MuDPIT Multi-Dimensional Protein Identification Technology

MS/MS Tandem mass spectrometry

V V6.4 wildtype ES cell sample, in context - as untagged control

1 I NTRODUCTION

The functional workings of life and development relies much on the action and regulation of

various gene products expressed, modified and regulated in a highly orchestrated manner

throughout various spatial and temporal time points In eukaryotes, basal transcription of genes is

carried out by different types of nuclear RNA Polymerases (RNAP), depending upon the specific

type of product involved (RNAPI ribosomal RNA, RNAPII messenger RNA, RNAPIII small

RNAs) In the synthesis of protein products from a gene, the minimal requirements necessary for

expression involve the assembly of basal transcription factors (TFII factors) with RNAPII on the

promoter region of a gene, to allow the formation of a pre-initiation complex (PIC) and

subsequent transcription of downstream sequence into messenger RNA for later translation into

proteins

1.1.1 Gene-Specific Transcription Factors

1.1.1.1 Mechanism of Action

To integrate and regulate the transcriptional activities at the many possible promoters across the

genome, an additional level of control is enabled through the use of „switchboard‟ like proteins

which bind to cis-regulatory promoter and enhancer DNA sequences These gene-specific

transcription factors represent a level above that of the basal transcription machinery, and

function as gears to activate or repress particular genes Transcription factors may act

synergistically or exclusively on multiple fronts to alter the rate of recruitment, assembly and

processing of basal transcriptional machinery,counter various active or repressive marks left by

other factors and to aid in chromatin remodeling activities that change the accessibility of DNA

to the transcriptional machinery This can be done both directly, and through the recruitment of

the relevant co-factors, into a fully assembled complex structure at the enhancer

(“enhanceosome”) The enhanceosome models the modular approach by which complex

eukaryotic organisms are able to integrate and fine-tune the expression of individual genes across

a sea of promoters

The majority of transcription factors (TF) are able to bind to specific DNA recognition sequences

through a distinct DNA –binding domain that is specific for a sequence motif These DNA

binding domains are functionally independent of the other domains within the TF Structurally,

the sequence motif presents a surface which is recognizable by the DNA binding domain and the

greater the specificity of the motif which the TF recognizes, the more likely it functions as a niche

factor in very specialized cellular regulation TFs can potentially bind to the major or minor

groove of type B double-helical DNA, although domains that bind to the major groove are able to

do so in a more specific manner, because all four nucleotides, A, T, G and C are seen as

chemically different, whereas from the minor groove angle, nucleotides are viewed as only

purines (A, G) and pyrimidines (C, T) To improve on DNA binding specificity, many TFs also

group together in hetero or homodimers and multimers in order to extend the region of

recognition necessary for specific regulation at the correct locus As such, consensus sequences

for TF binding to DNA demonstrate only a fraction of the possible binding sites which can be

occupied by TFs, and a clearer understanding of the function of a TF across the genome can only

be seen in the light of the associated factors which also accumulate at that particular promoter

1.1.1.2 Helix-Turn-Helix Domain Transcription Factors & Octamer Proteins

Many variants of DNA binding domains exist, and a widely studied area involves eukaryotic

Helix-Turn-Helix (HTH) containing TFs, so known because of the particular configuration of two

α-helices in a distinct turn from one another HTH-containing TFs include the well-known

homeodomain proteins, which are involved in a wide range of key stages in development

(Aravind et al., 2005)

Within the HTH configuration, the C-terminal α-helix binds to DNA via the major groove and as

the recognition helix, confers sequence specificity to TF binding A suitable example of this is

found in octamer proteins containing a POU domain, which can be seen as two subdomains,

POU-specific (POU-S) and POU-homeodomain (POU-HD) in sequence, separated by a linker

region The POU-HD domain is highly conserved among various octamer proteins, and

demonstrates HTH binding to a specific DNA sequence This is further elaborated upon by the

presence of the POU-S domain which varies between octamer proteins, and adds a degree of

specificity to the targets which the particular octamer protein may bind at

1.1.2 DNA Binding Sites of Transcription Factors

1.1.2.1 Analysis of Transcription Factor Binding Sites

Clearly, a means of understanding the molecular switchboards that regulate gene expression

might be possible if we could get at the potential DNA binding sites for proteins of interest

Chromatin Immunoprecipitation has been used as a means of detecting where proteins of interest

might bind to specific genomic regions in vivo Briefly, protein-DNA interactions are captured in

time by use of a cross-linker, typically formaldehyde, on cells of interest Thereafter, a whole

cell-extract is prepared and long chromatin strands are sonicated to shear DNA into small

fragments associated with their cross-linked proteins An immunoprecipitation is carried out

against the protein of interest, and this enriches for DNA fragments that are associated with the

protein Enriched genomic fragments can be identified by quantitative PCR for regions already

likely to be implicated with the protein, or by other larger scale means such as PET,

ChIP-chip, or ChIP-seq With the genome-wide approaches offered by these larger scale approaches,

identification of these enriched genomic regions requires no a priori information on the potential

binding sites, and offers a better overview to conduct association studies on the types of genes

potentially regulated by a single TF However, a couple of caveats remain First, TFs often work

in tandem with other TFs and co-regulators to achieve a particular regulatory outcome, and the

vast number of binding sites identified by global ChIP analyses represent an overestimate of the

actual number that may be relevant to a particular regulatory network Hence, demonstration of

TF binding to the DNA is not by itself a sufficient indication of its function at that particular

locus, since many other factors are required for it Additionally, TFs may serve as both activators

and repressors, depending upon the context in which it is found, a characteristic not captured by

ChIP alone

1.2 Early Mouse Development

1.2.1 ES Cells Are a Model for ICM Pluripotency

The first differentiation event occurs early in mammalian development, and is a defining moment

where the embryo diverges to form two separate components, the embryonic inner cell mass, and

the extraembryonic trophectoderm (Gardner, 1983) The inner cell mass (ICM) develops as the

embryo transitions from a morula to a blastocyst stage, and is the origin of all subsequent tissues

of the embryo proper In culture, the ICM can be isolated from surrounding tissue, and grown

independently as embryonic stem cells (ES cells) (Brook and Gardner, 1997; Evans and Kaufman, 1981) Like the ICM from which they are derived, ES cells have the similar potential to

differentiate into cells from any of the three germ layers, ectoderm, mesoderm and endoderm

This is demonstrated on several fronts, the directed differentiation with specific cell culture

conditions and growth factors; the spontaneous differentiation of ES cells into multiple

differentiated cell types in embryoid bodies (Koike et al., 2007), and most significantly, the

contribution to all tissue types of the adult organism when reintroduced into a blastocyst, and the

subsequent germline transmission of these ES-cell derived cells to subsequent generations

(Beddington and Robertson, 1989; Bradley et al., 1984) The use of mouse ES cells has propelled

genetic approaches, facilitating the development of knock-out phenotype cell lines and animal

models for a variety of research purposes (Nagy and Rossant, 2001; Rossant et al., 1993)

Because of the pluripotency of ES cells, they have been subject to intense study as a proxy for the

ICM Unlike the majority of cell lines available, they require no transformation or

immortalization to adapt to long term cell culture, and under the right conditions, are able to

maintain their normal karyotype for repeated passages (Loh et al., 2008) Prior to the derivation

of ES cells, the related embryonic carcinoma (EC) cell line was first isolated from

teratocarcinomas in adult mice,(Andrews, 2002) EC cells are phenotypically similar to ES cells,

but show a lower propensity to contribute reliably to chimeras, or differentiate in culture

Interestingly, this points towards the cells of the ICM as a unique developmental time point that

can be largely captured through the study of ES cells

In recent years, the potential of ES cells to generate all possible tissue types has led to extensive

research on the directed differentiation of ES cells, and has fuelled public promise for the

generation of cells and tissues for transplantation and repair in various complex diseases Of late,

seminal breakthroughs have been made in this direction, by the generation of ES-like cells

(Induced Pluripotent Stem cells, iPS cells) from differentiated cell types with the addition of 2 to

4 key factors, the original set which consisted of Oct4, Sox2, Klf4 and c-Myc (Eminli et al., 2008;

Nakagawa et al., 2008; Park et al., 2008; Takahashi et al., 2007; Takahashi and Yamanaka, 2006)

While this opens doors towards future goals in transplantation medicine, it also sparks a more

immediate interest in the regulation of pluripotency, and the reprogramming of differentiated cell

types to a pluripotent state

Figure 1 Origin of embryonic stem cell lines

Figure from Boiani, M and Scholer, H., 2005 Pluripotent ES cells are

obtained from the ICM at stage E3.5, and are able to self-renew or

differentiate in culture.Oct4 is expressed at early embryonic stages in the

morula and ICM of the pre-implantation embryo, and also in the primordial

germ cells (Boyer, L et al 2006)

1.2.2 Pluripotency is controlled on multiple fronts

1.2.2.1 Known Signaling networks

A fundamental understanding of the signaling networks at play in the pluripotent ES cell is

necessary for our understanding of the maintenance of pluripotency or subtle cues that direct ES

cells to downstream lineages An early observation in the culture of ES cells was the need for a

mouse embryonic fibroblast (MEF) feeder layer, without which the self-renewing phenotype

could not be maintained Later on, the crucial factor from the feeder cells was found to be

leukemia inhibitory factor (LIF), a cytokine required for the activation of STAT3 via the gp130

cell surface receptor which functions as a transcription factor in ES cells (Heinrich et al., 2003;

Smith et al., 1988) Interestingly, LIF is not required for normal mouse development, and Lif-/-

mice are able to survive past the E3.5 blastocyst stage at which ES cells are derived However,

more recent evidence suggests that LIF is still required in vivo, as Lif -/- mice embryos subject to

diapause are unable to resume development (Bhatt et al., 1991; Stewart et al., 1992)

Human ES cells do not require LIF, and this may represent inherent differences between these

two species, or perhaps point towards technical differences in the developmental stage at which

human and mouse ES cells are routinely derived from, suggesting that alternative pathways may

be of primary importance across ES cells from different species (Okita and Yamanaka, 2006)

Together with LIF, the BMP family of factors (BMP: Bone Morphogenetic Protein) is necessary

for ES cell pluripotency In the absence of serum in the media, BMPs and LIF are sufficient to

maintain pluripotency (Ying et al., 2003) In this scenario, BMP activates SMAD transcription

factors that in turn stimulate genes of the Id (Inhibitor of Differentiation) family However, BMP

factors without LIF support differentiation, suggesting a fine balance between self-renewal and

differentiation that are regulated by both external and internal signaling events (Ruzinova and

Benezra, 2003)

The Wnt signaling pathway is also involved in early embryonic development and maintenance of

ES cell pluripotency Different Wnt proteins are involved at various stages of mouse development

Wnt5a and Wnt11 are expressed in vivo during the morula to blastocyst transition, and Wnt11 is

further involved with estradiol signaling during E4 at implantation, shortly after the E3.5 stage

ICM which ES cells represent (Mohamed et al., 2004) More significantly, Wnt3 deficient

embryos remain undifferentiated, and continue to express Oct4, suggesting that the expression of

Oct4, a key pluripotency factor, is linked in part to the Wnt pathway (Liu et al., 1999) Activation

of the canonical Wnt pathway results in downstream accumulation of beta-catenin (Ctnnb) in the

nucleus This has been demonstrated to allow the continued self-renewal of ES cells in the

absence of LIF, through the maintenance of pluripotency-associated transcription factors such as

Oct4, Nanog and Rex1 (Sato et al., 2004) The Notch signaling pathway often cross-talks with the

Wnt pathway (Hayward et al., 2008), and various components such as Ctnnb and Rbpj are

expressed across the early development window from 2-cell to blastocyst stages (Cormier et al.,

2004) Notch is likely to play a role in the maintenance of ES cells, since a knockdown of Notch

and other members of the pathway resulted in a decrease in ES cell proliferation (Fox et al., 2008), although this early evidence requires follow up insight into the role of Notch in ES cells

1.2.2.2 Protein-Protein Interactions

In order to further understand the key players of pluripotency at their functional level, there is

also an increasing emphasis on the proteome of ES cells Although genetic and genomic analysis

has provided us with many plausible signaling networks, they cannot fully describe the

developmental pathways and networks that are actually at work in the cell Furthermore the

three-dimensional nature of proteins and their possible post-translational modifications creates a

complexity in analysis and predictive arguments that necessitates a direct look at the proteome

(Whetton et al., 2008)

In that light, a few studies have embarked upon the proteome of ES cells, by comparative analysis

of pluripotent to differentiated ES cells (Baharvand et al., 2008) , the use of microscale

two-dimensional liquid chromatography and tandem mass spectrometry (2D LC-MS/MS) to identify

ES cell-specific factors including Oct4, Sox2 and Utf1 which are at relative low abundance

(Nagano et al., 2005), or a more comprehensive overview of the ES cell proteome (Van Hoof et

al., 2006) Because of the huge dynamic range of the proteome, any one assay of the proteome is

in itself, insufficient to capture the larger spatial and temporal interplay between functional

proteins Proteome analyses have also focused specifically on certain cellular compartments, such

as membrane or nucleus, or altered techniques to capture specific post-translational modifications

to these proteins As such, the variability in technical analysis methods have made comparison of

proteomic results far more challenging than that for genomics and transcriptomics Still, the

complementary nature of the data between these two levels (genomic and proteomic) serves to

open up exciting prospects in the future of our understanding of ES cell networks

1.2.2.3 Epigenetic Regulation of Pluripotency

MicroRNAs (miRNA) are small RNA sequences of 19-25 nucleotides in length, and in recent

years, have been found to play an extensive role in the regulation of development These

non-coding RNAs are endogenously expressed in the nucleus and processed by Drosha and Dicer

proteins When bound in a RNA-induced silencing complex (RISC), miRNA are able to bind to

their complementary target sequences in the 3‟UTR of mRNAs, and either block mRNA

translation or induce the cleavage of the targeted mRNA sequence (Bartel, 2004) Determining

the precise effects of miRNA targeting is challenging, given the extensive range of targets

regulated by a small population of known miRNAs Dicer -/- mice are embryonic lethal

(Kloosterman and Plasterk, 2006), and the corresponding ES cells, though able to divide, were

not able to differentiate normally (Kanellopoulou et al., 2005; Murchison et al., 2005), suggesting

the importance of functional miRNA for proper development However, many specific miRNAs

involved in ES cells have not yet been extensively characterized, although in murine ES cells, the

related mir-290 and mir-302 clusters appear to be important for pluripotency, and were linked to

further epigenetic regulation by de novo DNA methyltransferases (Dnmts) and histone

modifications More specifically, Dicer -/- mES cells showed downregulation of the de novo

Dnmt3 family genes, and the promoters of key pluripotency factors Pou5f1 & Nanog in Dicer -/-

ES cells were shown to be deficient in the repressive methylation marks that accumulated on

differentiation induced by retinoic acid This was rescued on transient expression of some or all

of the members of the mir-290 cluster, pointing towards their role in early embryonic

development (Sinkkonen et al., 2008)

Clearly, a complex interplay of epigenetics is at hand in the regulatory networks that govern

development In the more well-understood mechanism of DNA methylation in ES cells, the loss

of Dnmt3a and Dnmt3b also resulted in a loss of methylation at the key promoters of Pou5f1 and

Nanog examined (Li et al., 2007), and a concomitant lack of repression of these genes on

differentiation Dnmt1 is a methyltransferase involved in the maintenance of methylation marks

at CpG-rich regions of promoters, and /- mice are also embryonic lethal, although

Dnmt1-/- ES cells are able to self-renew (Panning and Jaenisch, 1996) It appears that repressive DNA

methylation marks are a necessary phenomenon for the transition from a pluripotent ES state to

downstream lineages

Methylation marks on histones have also demonstrated an additional level of epigenetic

regulation An octamer of histone proteins and DNA form nucleosome complexes that serve to

condense and wind DNA up in the form of chromatin Depending upon the conformational

structure of chromatin, DNA sequences are made accessible or hidden from view to the

transcriptional machinery As such, the assembly, disassembly and modification of these

nucleosomes are integral to regulation While histones can be modified by a variety of chemical

groups, methylation and acetylation are presently the best understood means by which histone

modifications define transcriptional activity (Schulz and Hoffmann, 2007)

Large scale chromatin immunoprecipitation techniques have further enabled this task, and global

histone methylation studies in ES cells have shown general trends with active promoters

associated with H3K4me3 (Histone 3 trimethylation at Lysine 4) and H3K4me2 and actively

transcribed regions with H3K36me3, while other methylation patterns are repressive (H3K27me3) (Swigut and Wysocka, 2007) Histone methyltransferases (HMT) enable the generation of these

marks, which can also be removed by histone demethylases ES cells demonstrate an unusually

active state of chromatin compared to other cell types, and this may reflect in part, the unique

pluripotent property of cells represented at this point in development In ES cells, the repression

of some differentiation-specific genes is marked by H3K27me3, as generated by the HMT, Ezh2,

a member of the polycomb repressor complex During differentiation, Jumanji-domain proteins

(Jmjd 3, Utx1, Uty1) counter this repressive mark by demethylation, allowing for transcription of

other factors should the ES cell begin to differentiate (Bibikova et al., 2008) Interestingly, over

80% of known ES cell promoters also carry the active H3K4me3 mark, leading some to suggest

the possibility of “bivalent chromatin domains” that harbor imminent active marks for lineage

specific genes that could be rapidly transcribed when repression is lifted during differentiation

(Bernstein et al., 2006)

1.2.3 Key Transcription Factors Regulate Pluripotency

In the most direct means of transcriptional regulation, transcription factors bind to the promoter

region along with the basal transcriptional machinery to further regulate gene expression

Pluripotency-associated transcription factors Oct4, Sox2 and Nanog have been shown to also

associate with bivalent chromatin, which further supports the activation and regulation of these

transcription factors with downstream activation of differentiation specific genes

1.2.3.1 Oct4/Pou5f1

Oct4 is a POU domain transcription factor that shows high expression in ES cells and early

embryonic stages, present from the pre-fertilized oocyte up to E4.5 (Adjaye et al., 1999; Palmieri

et al., 1994; Pesce et al., 1998) Because it was generally believed to be downregulated upon

differentiation, Oct4 has been extensively used as a marker for pluripotency, although a small

pocket of Oct4 expression continues in primordial germ cells (E7.5) and in vitro culture of

germline stem cells (Kanatsu-Shinohara et al., 2004) Oct4 is necessary for early embryonic

development, and Oct4-/- mice do not form an ICM More detailed studies involving an Oct4

inducible transgene has also demonstrated that Oct4 levels are responsible for directing cell fate

of ES cells (Niwa et al., 2000) In more recent works surrounding somatic cell reprogramming,

Oct4 was shown to be one of four necessary factors for the derivation of induced pluripotent stem

(iPS) cells (Takahashi and Yamanaka, 2006) Subsequent studies showed that substitutions could

be made for some or all of the other 3 factors (Sox2, Klf4, c-Myc), but the constant requirement

of Oct4 reactivation and the known need for regulated Oct4 expression in the embryo and ES

cells suggest a central role for Oct4 in the maintenance of pluripotency

1.2.3.2 Sox2

Sox2 is a high mobility group (HMG) transcription factor that has been shown to heterodimerize

with Oct4 on DNA to direct transcriptional regulation (Remenyi et al., 2003) Indeed, the

octamer-sox DNA binding element has been defined (Botquin et al., 1998), and is found in

proximity across many genes, including that those that are upregulated on differentiation (Chew

et al., 2005) Sox2 expression is first found at the morula stage, and is well-expressed

specificially in the ICM of E3.5 blastocyst, much like Oct4 However, unlike Oct4 which is

rapidly downregulated, Sox2 continues to be expressed till the mid-streak stage around E7.5,

although these levels are lower than in the ICM, and is largely restricted to the anterior

neurectoderm Targeted disruption of a single Sox2 allele did not result in any overt phenotype

besides reduced male fertility, but attempts to generate Sox2 -/- mice from heterozygous crosses

did not give any homozygous null embryos In situ analyses suggested that the knockout of Sox2

results in embryonic lethality shortly after the blastocyst stage because of differentiation of the

epiblast to trophoblast tissues (Avilion et al., 2003) Reminiscent of the phenotype with an Oct4

knockout, it provides support to the notion of co-regulatory roles of Oct4 and Sox2

1.2.3.3 Nanog

The homedomain transcription factor Nanog was discovered only in the last 5 years, and is a

novel pluripotency factor (Chambers et al., 2003; Mitsui et al., 2003) Overexpression of Nanog

results in a population of ES cells that self-renew even in the absence of LIF In contrast, factors

such as Oct4 and Sox2 require the presence of LIF in the culture medium to activate LIF/Stat3

signaling pathways for pluripotency However, the expression profile and knockout phenotype of

Nanog points to a degree of interconnectedness between the roles of Nanog and Oct4 – an Oct4

transgene knockdown results in differentiation of ES cells to the trophectoderm, and this can be

rescued specifically with Oct4, but not Nanog (Chambers et al., 2003; Niwa et al., 2000), hence

Nanog-mediated self-renewal is dependent on Oct4 presence Comparatively, Nanog expression

is more distinctly associated with the pluripotent phenotype than the other 2 key factors First

expressed in the morula, expression continues to the ICM and epiblast, and is also found in ES

cells in culture However, expression is rapidly downregulated on differentiation, and Nanog

expression is not reactivated in adult tissues, unlike Sox2 and Oct4 which show limited

reactivation in select tissues Nanog -/- mice are also embryonic lethal, due to the loss of

pluripotency, and differentiation of the ICM into extra-embryonic tissue (Mitsui et al., 2003)

Because of the integral role of Nanog to development, a study was carried out to determine the

protein interaction network surrounding Nanog in mouse ES cells (Wang et al., 2006) Using a

transgenic biotin-tagged Nanog expressed at levels mimicking that of the endogenous protein,

their results demonstrated the association of Nanog with nuclear factors that were independently

found to be necessary for the maintenance of ES cell pluripotency While this list included Oct4,

other enriched proteins were found to be associated with known co-repressor pathways, extending

our present understanding of the potential means by which pluripotency is maintained

1.2.4 Large scale studies on pluripotency

In a bid to understand the role of these crucial transcription factors in the maintenance of ES cells,

genome-wide chromatin immunoprecipitation (ChIP) screens have been undertaken Across the

mouse genome, approximately 1000 and 3000 Oct4 and Nanog binding sites were found

respectively In a stringent subset of these binding sites, 345 loci were found to exhibit both Oct4

and Nanog binding, and this was validated for select loci by the differential expression of these

target genes on RNAi mediated knockdown of Oct4 or Nanog (Loh et al., 2008) A similar study

in human ES cells sought to find the binding sites of Oct4, Sox2 and Nanog, with 353 genes

showing common binding of these 3 factors (Boyer et al., 2005) Genes occupied by Oct4 and

Nanog or all three factors, included an overrepresentation of transcription factors comprised of

both active pluripotency engaged genes and repressed genes active only in later development

This result points towards the co-involvement of these 3 factors in the regulation of gene

expression for targets important across development, and includes loci such as Zic3 and Rest,

both which are now known to play roles in self renewal (Lim et al., 2007; Singh et al., 2008) In

complement to these studies, an extended network of genes related to Oct4 was determined by the

use of both ChIP and genome-wide expression profiling with microarrays (Matoba et al., 2006) in

ES cells where the Oct4 transgene level can be modulated The use of a RNAi screen with

transcriptional regulators identified three novel genes that play a role in self-renewal – Esrrb,

Tbx3 and Tcl1 (Ivanova et al., 2006) Because of the well-established role for Oct4 in

pluripotency, the majority of large scale studies profiled the genome using Oct4 as a central node

Another study examined the correlation of gene expression to that of Oct4, and found that the

identified targets favored roles in chromatin structure, DNA repair and the regulation of

pluripotency and lineage specific genes (Campbell et al., 2007), a finding in agreement with

another RNAi focused screen in human ES cells (Babaie et al., 2007) The results of these large

scale screens often function as starting points for additional research into genes that are

significantly associated with stem cell pluripotency in a variety of ways Meta analysis of these

large data sets can be useful in elucidating new connections in our understanding of stem cell

regulation (Zhou et al., 2007) to formulate gene regulatory networks with which we can better

manipulate ES cells

To better unearth the significance of these target genes in a systems-wide approach, it is

necessary to combine the genomic data with proteomic data, both with large scale proteomic

screens in ES cells (Baharvand et al., 2008; Barthelery et al., 2007; Elliott et al., 2004), as well as

directed interactome mappings (Liang et al., 2008; Wang et al., 2006) for which there is notably

less information, because of the inherent complexity of massive proteomic studies Still, such

information is critical to develop our shared understanding of regulation at both the gene and

protein level, although increasingly, the role of epigenetic modifications and small RNAs are

becoming known, and should be included for a wholistic view of stem cell function

To this end, the concept of an enhanceosome is a platform that can be used to integrate various

levels of regulation into a physical signaling center at every gene across the genome The classic

example of this is found at the IFN-β locus, where multiple transcription factors, c-Jun, IRF-3,

IRF-8 and NFκB are known to bind cooperatively to the nucleosome-free region of the IFN-β

promoter The binding of individual factors is insufficient for transcriptional activation, but the

binding of all transcription factors creates a surface amenable for the recruitment of chromatin

remodelers and nucleosome acetylases that serve to reposition the TATA box into a configuration

suitable for transcriptional activation by RNA polymerase II and TATA box binding protein

(TBP) (Panne, 2008) The atomic level structure of the activated IFN-β locus has been determined,

and shows a close association of multiple factors along the DNA strand (Panne et al., 2007)

Intriguingly, although there are overlapping DNA domains and tight binding of the transcription

factors within ~50bp of the promoter region, direct protein-protein interactions between these

transcription factors are few, and suggests that the shared exposed surfaces of these factors are

available for the recruitment of additional regulatory molecules which do not bind directly to the

DNA (Merika and Thanos, 2001)

Figure 2 Atomic Model of the IFN-β Enhanceosome

This model describes the extensive binding of transcription factors to the

DNA strand that creates a new composite surface for interaction and

recruitment of necessary proteins

The notion of an enhanceosome is in line with the apparent contradiction between the association

of individual transcription factors to a large number of genes, the relatively few number of key

pluripotency genes discovered to date, and the complexity of observed gene regulation The

sensitivity of temporal and spatial regulation in development can however, be achieved through a

multilayered regulation of genes even with few factors, by the accumulated regulatory signal

through an enhanceosome at the promoter region A couple of key studies to date are in

agreement with the likely existence of such a module in pluripotent stem cells Chen et al

mapped the binding sites of 13 DNA-binding transcription factors in ES cells (Oct4, Nanog, Sox2, Stat3, Smad1, Zfx, c-Myc, n-Myc, Klf4, Esrrb, Tcfcp2l1, E2f1, Ctcf), and found specific genomic

regions associated by a subset of these factors examined Additionally, they also observed that the

regions with multiple transcription factor binding sites were found associated with the activating

histone methylation mark, H3K4me3 (Chen et al., 2008) A separate study examined a related

population of transcription factors (Oct4, Nanog, Sox2, c-Myc, Klf4, Dax1, Rex1, Zfp281, Nac1),

and found that genes associated with 4 or more of these factors were also more likely to be

associated with H3K4me3 marks, whereas factors bound by fewer known factors were likely to

be repressed in ES cells (Kim et al., 2008) Together, these studies support the existence of an

enhanceosome necessary for the intricacies of transcriptional regulation, and that additional

transcription factors outside all these examined, may be involved at inactive loci to repress these

genes In light of this, there is an apparent need for protein and epigenetic level information to

understand the assembly and roles of factors present in different enhanceosomes

1.3 Role of Oct4 in Maintenance of Pluripotency

A crucial link in the maintenance of pluripotency, Oct4 is a transcription factor of the POU

(Pit-Oct-Unc) family that is expressed in early development from maternal and zygotic transcripts

(Rosner et al., 1990) in mouse embryos until the blastocyst stage at embryonic day 3.5 (E3.5),

where Oct4 expression becomes restricted to the inner cell mass (ICM) Oct4 (also known as

Pou5f1) was first demonstrated to be essential for embryonic development, as homozygous

Oct4-/- mice embryos failed to form an ICM, although the trophoblast continued to develop normally

(Nichols et al., 1998) Later work by Niwa and colleages (Niwa et al., 2000) demonstrated a

dosage dependence of embryonic development on Oct4 Using a tetracycline-dependent transgene

expressed in Oct4+/- mouse ES cells, they showed that specific lineage markers of differentiation

were observed when Oct4 levels varied by more than 50% from normal diploid expression This

agreed with earlier data from Nichols that heterozygous Oct4 +/- mice were able to develop

normally, and suggests that cells destined for the ICM are committed to the trophectoderm on loss

of Oct4, and that Oct4 serves as a repressor of trophectoderm specific genes When Oct4

transgene levels were increased, Oct4 gain-of-function (GOF) mESCs were induced to

differentiate to cells expressing markers of the primitive endodermal and mesodermal lineages,

suggesting an activating role for Oct4 as a transcription factor This correlates well with in vivo

data that shows an increase in Oct4 levels in the primitive endoderm of late blastocysts (E4.5)

compared to the ICM (E3.5) stage at which Oct4 is already abundant (Palmieri et al., 1994) The

dual role played by Oct4 as both an activator and suppressor suggests that other factors might be

important in determining the specific role of Oct4 in directing ES cells towards early specificity

At later developmental stages however, Oct4 is not essential for the maintenance of somatic stem

cells (Lengner et al., 2008)nor other tissues, and is silenced immediately following blastocyst

implantation by G9A-mediated histone methylation and later, promoter methylation by

Dnmt3a/3b (Lagarkova et al., 2006; Lengner et al., 2008)

On gastrulation, Oct4 ceases expression in most tissues and by E7.5, is contained in the

primordial germ cells (Scholer et al., 1990; Yeom et al., 1996) and is necessary for their survival

(Kehler et al., 2004) In adults, Oct4 expression is exclusively seen in the germ lineage in

maturing and mature gametes (Pesce et al., 1998) and has also been implicated in germ cell

tumors (Looijenga et al., 2003) Interest in the role of Oct4 has focused significantly on early

embryonic stages and ES cells, and has taken form through chromatin immunoprecipitation data

(Boyer et al., 2005; Loh et al., 2006), RNA interference screens (Babaie et al., 2007; Ivanova et

al., 2006) and gene expression profiling (Campbell et al., 2007; Matoba et al., 2006; Zhou et al.,

2007) all intended to extend our understanding of the regulation of pluripotency

1.3.1 Regulation of Oct4 Expression

Oct4/Pou5f1 expression is controlled by cis-acting elements located upstream of the Oct4 gene,

as well as epigenetic modifications such as DNA methylation (Ben-Shushan et al., 1993) and

histone modifications (Mikkelsen et al., 2007) Early work identified a proximal and distal

enhancer region (Yeom et al., 1996) that could bind transcription factors that might be cell-type

specific A later paper found that germ cell nuclear factor (GCNF), an orphan nuclear receptor,

was bound to the proximal enhancer region, and repressed Oct4 expression The loss of GCNF

resulted in more spurious expression of Oct4 following gastrulation, in tissues other than the

germ cell lineage to which Oct4 is normally restricted to at this point (Fuhrmann et al., 2001)

Later analysis of the upstream region to Oct4 located four conserved regions (CR1-4) between

human, murine and bovine Oct4 promoter regions (Nordhoff et al., 2001), of which regions

CR2-4 of the human OctCR2-4 promoter were found to positively regulate OctCR2-4 promoter activity, whereas

a region between CR1-2 was found to negatively regulate hOct4 expression (Yang et al., 2005)

Additionally, Yang et al also identified a minimal Oct4 promoter region that contained GC boxes

and hormone response elements (HREs) that were differentially regulated by the binding of Sp1

and Sp3 GC Box-1 (GC-1) was completely conserved between human and murine sequences,

suggesting that a similar binding of Sp1 and Sp3 may regulate mOct4 sequences Presently,

binding sites for 8 transcription factors known to be active in the regulation of ES cell

pluripotency have been located at the Oct4 gene promoter (Oct4, Sox2, Nanog, Rex1, Dax1,

Nac1, Klf4 and Zfp281;(Kim et al., 2008), confirming that Oct4 expression is in part

self-regulatory, and also tightly controlled in a combinatorial fashion

1.3.2 Oct4 Structure & Domains

Oct4 consists of a conserved POU (Pit-Oct-Unc) DNA binding domain, and an N and C terminal

transactivation domain The POU domain binds the canonical ATGCAAT octamer motif via the

major groove of the DNA, and comprises two sub-domains joined by a flexible linker region

Both sub-domains, POU-homeodomain (POUHD) and POU-specific (POUS) are structurally

distinct, but adopt a helix-turn-helix structure that is characteristic of the POU family of

transcription factors (Cleary and Herr, 1995; Klemm et al., 1994)

The N and C-terminal transactivation domains do not bind DNA directly, but have been

demonstrated to work independently of each other, as shown by fusion of the C-terminal domain

to a Gal4 DNA binding domain (Brehm et al., 1997) Additional work on the transactivation

domains suggests some differences in functionality, with the N-terminal domain active in all cell

types, while the C-terminal domain shows more limited cell type functionality (Brehm et al.,

1999; Brehm et al., 1998) Because the C-terminal domain demonstrated its cell-specific activity

in a heterologous binding assay only when a POU domain (from Oct1 or Oct2), but not Gal4 or

Pit-1 domain was used, it suggests a possible, but as yet unanswered role for the POU domain in

serving an alternative role as part of an interaction platform for other transcriptional co-factors

Interestingly, in ES cells however, Niwa et al (Niwa et al., 2000) have shown that the N- and

C-terminal domains share redundant functions, and the combination of either with the POU domain

is sufficient for proper Oct4 function in ES cells

1.3.3 Known Protein Interaction Partners of Oct4

The central importance of Oct4 to ES cell pluripotency has led to the discovery of a fair number

of protein interaction partners of Oct4 through techniques such as co-immunoprecipitation,

yeast-2-hybrids, GST-pulldowns and mass spectrometry; these are reported in the tables below (Table 1, Table 2 & Table 3) From the reported interactions, it appears that the POU domain of Oct4 is

critical for mediating interactions, and Oct4 is associated with both transcriptional activation and

repression The specified function of Oct4 is thus dependent on the interactions it makes

Additionally, Oct4 is seen to interact with a variety of proteins of different roles, including

transcription factors, viral oncoproteins, and co-regulatory molecules While Oct4 is typically

expected to be a DNA binding transcription factor, a few studies (Ezashi et al., 2001; Guo et al.,

2002; Vigano and Staudt, 1996) suggest that Oct4 can influence transcription independent of

direct DNA binding through interactions with DNA bound factors

Table 1 Oct4 Acts as a Transcriptional Activator

Protein

Partner

Protein Domain (if known)

Oct4 Domain (if known)

Experimental Method

(Vigano and Staudt, 1996)

Mnat1/Mat1

Menage a trois

1

Amino acids 115-309 (inclusive

of coil domain)

coiled-POU GST-fusion

pulldown

Transcriptional Activation Phosphorylation of DNA-bound Oct4 (and Oct1, Oct2) for the recruitment of TFIIH

(Inamoto et al., 1997)

-

POU GST-fusion

pulldown Co- immunoprecipitation

Transcriptional Activation Bridging factors for Oct4 at distal binding sites to the promoter

(Brehm et al., 1999; Scholer

Co-Transcriptional Activation Stabilization of Oct4 DNA binding, or interactions with other factors by architectural changes to the DNA on binding of the HMG domain

(Butteroni et al., 2000)

modeling (Oct1 and Sox2)

Transcriptional Activation Co-binding of Oct4 and Sox2 to DNA was required for activation

of Fgf4 expression

(Ambrosetti

et al., 1997;

Remenyi et al., 2003)

EWS

Ewing‟s

NTD GRP1, 2,

3

POU Bacterial two hybrid

GST-fusion

Transcriptional Activation

(Lee et al., 2005a; Wang

et al., 2006)

Sarcoma

Protein

pulldown Co- immunoprecipitation (P19 EC cell extracts) Affinity Purification/MS

Enhanced expression from reporter construct with Oct4 binding sites

on co-expression of EWS

C-POU GST-fusion

TOF

pulldown/MALDI- immunoprecipitaton (overexpressed in 293T cells)

Co-Transcriptional Activation

Enhanced expression from reporter construct with Oct4 binding sites

on co-expression of Pkm2

(Lee et al., 2008)

(Takao et al., 2007)

Table 2 Oct4 Acts as a Transcriptional Repressor

Protein

Partner

Protein Domain (if known)

Oct4 Domain (if known)

Experimental Method

POU

Co-immunoprecipitation (overexpressed in JAr

choriocarcinoma cells)

Transcriptional Silencing Recruitment of Oct4 to Ets2 on IFN-T

promoter Oct4 does not bind DNA directly

(Ezashi et al., 2001)

FoxD3

Forkhead box

D3

DNA binding domain

POU GST-fusion

pulldown

Transcriptional Silencing Silencing of FoxD3 activated genes, FoxA1 and FoxA2 Oct4 does not bind DNA directly

(Guo et al., 2002)

Co-Transcriptional Silencing

(Niwa et al., 2005)

Co-Transcriptional Silencing Silencing of reporter with Oct4 binding sites

on co-expression of Pias4, sequestering of Oct4 to the nuclear periphery on Pias4:Oct4 interaction

(Tolkunova et al., 2007)

Transcriptional silencing NuRD and NODE repressor complexes

(Liang et al., 2008)

Transcriptional silencing NuRD and NODE repressor complexes

(Liang et al., 2008; Wang

et al., 2006)