University of Southern California Dissertations and Theses

Roles of Klf4 in embryonic stem cells

(USC Thesis Other)

Roles of Klf4 in embryonic stem cells

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

ROLES OF KLF4 IN EMBRYONIC STEM CELLS

by

Rosemary G. Andrianakos

_____________________________________________________________

A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(BIOCHEMISTRY)

May 2008

Copyright 2008 Rosemary G. Andrianakos
ii
Acknowledgments
Special thanks to Chunming Liu for his generous donation of original Klf4 constructs
plasmid DNA and siRNA. Special thanks to DR. Qui-Long Ying for providing Oct-4 and Sox-2
mouse plasmid DNA and Mouse Embryonic Stem Cells. I am eternally grateful for the help received
from my lab mates and their enthusiasm for carrying on my project. Special thanks to, Peilin Zhang
for performing Taqman ®Mouse Stem Cell Pluripotency Array assay on Klf4- siRNA infected mES
cells, Houtan Noushmehr for Teaching me use the RSAT and Q-PCR Biostatistical software, Yang
Yang for finishing up mapping of Oct-4 and Sox-2 clones with Klf4, and for her continued effort in
establishing endogenous interactions of transcription factors. Thanks to Junmook Lyu and Vicky
Yamamoto for training and me. Thanks to Dr. Lu for being a wonderful and supporting mentor, and
giving me the opportunity to branch out and work on a project that I found very interesting and
rewarding. Very special thanks to my committee members for taking the time to review this thesis.
iii
Table of Contents
Acknowledgments ii
List of Figures iv
Abstract vii
Chapter 1. Introduction 1
Chapter 2. Materials and Methods 7
Chapter 3. Results 29
Chapter 4. Discussion 79
References 96
Appendix. Abbreviations 100
iv
List of Figures
Figure 1: Structural Representation of Klf4 Domain Deletions 3

Figure 2: Lysates of 293T Cells Infected With 2µl klf4-WT and Klf4-ΔC 13

Figure 3: FUIGW Virus Test One Day Post Infection 14

Figure 4: Cloning Klf4 Constructs into FUIPW Vector 19

Figure 5: Amino Acid Sequences of Each Segment Sox-2 20

Figure 6: Amino Acid Sequences of Each Segment Oct-4 21

Figure 7: Sox-2 Mapping 22

Figure 8: Oct-4 Mapping 23

Figure 9: Map and Features of PCR Blunt 25

Figure 10: Cloning of Sox-2 Constructs into FUIGW Vector 26

Figure 11: Cloning Oct-4 Constructs into FUIGW Vector 28

Figure 12a: Enrichment of (Oct/ Sox) DNA Binding Sites in Pluripotency 30
Linked Genes

Figure 12b: Enrichment of Klf4 DNA Binding Sites in Pluripotency 30
Linked Genes

Figure 13: Human Embryonic Stem Cells, Klf4 / Oct-4 / Nuclear 63x 31

Figure 14: Human Embryonic Stem Cells, Oct-4 / Sox-2 / Nuclear 40x 32

Figure 15: Human Embryonic Stem Cells, Klf4 / Oct-4 / Nuclear 20x-Magnified 33

Figure 16: Human Embryonic Bodies, Klf4 / Nuclear 34

Figure 17: Human Embryonic Bodies, Oct-4 / Nuclear 35

Figure 18: Comparative Expression of hES and EB 36

Figure 19: Lysate Controls: Klf4 and β-catenin Binding, a Mono-Flag 38
v
Figure 20: Immunoprecipitation: Klf4 and β-catenin Binding 39

Figure 21: Binding Between Klf4 (ΔC, ΔM, and WT) & Sox2 40

Figure 22: Binding Between Klf4 (ΔC, ΔM, and WT) & Oct-4 41

Figure 23: Binding of Sox-2 to Klf4-ΔZF (429,454) 42

Figure 24: Binding of Oct-4 to Klf4-ΔZF (429,454) 43

Figure 25: Competition between β-catenin and Oct-4 for Klf4 Binding 44

Figure 26: Lysate Controls: Competition between Sox-2 and Oct-4 for 45
Klf4 Binding

Figure 27: Western Blots of mES Whole Cell Lysates 46

Figure 28: Immunoprecipitation: Binding Between Oct-4 and Sox-2 in mES 47

Figure 29: Binding Between Sox-2 and Oct-4 in 293T Cells 48

Figure 30: Immunoprecipitation: Binding of Klf4 and Oct-4 in mES 49

Figure 31: Immunoprecipitation: Binding Between Klf4 and Sox-2 in mES 50

Figure 32: Alkaline Phosphatase Staining Klf4 mES Infections 52

Figure 33: Alkaline Phosphatase Staining Klf4-siRNA 53

Figure 34: Virus Infected mES Cells SSEA1 FACS Analysis 55

Figure 35: SSEA1 Staining Controls 56

Figure 36: SSEA1 Staining of mES Infected FUIPW-Klf4 Mutants and 57
PLKO-Klf4 siRNA

Figure 37: Q-PCR Results for First Passage 59

Figure 38: Q-PCR Results for Second Passage 60

Figure 39: Q-PCR Results for Third Passage 61

Figure 40: RT-PCR Results 63
vi

Figure 41: Prosite Domain Prediction for Oct-4 64

Figure 42: Oct-4 Cloning Sequence Alignment with Full Length Oct-4 66

Figure 43: Prosite Domain Prediction for Sox-2 67

Figure 44: Sox-2 Cloning Sequence Alignment with Full Length Sox-2 68

Figure 45: β-catenin with Increasing Amounts of Klf4 Luciferase Assay 70

Figure 46: β-catenin with Increasing Amounts of Oct-4 Luciferase Assay 71

Figure 47: β- catenin with Increasing Amounts of Sox-2 Luciferase Assay 72

Figure 48: β-catenin with Increasing Amounts of Oct-4 and Klf4 74
Luciferase Assay

Figure 49: β-catenin with Increasing Amounts of Oct-4 and Klf4 and Sox-2 75
Luciferase Assay

Figure 50: Luciferase Assay of β-catenin with Increasing Amounts of (Klf4, 76
Oct-4, and Sox-2) (A)

Figure 51: Luciferase Assay of β-catenin with Increasing Amounts of (Klf4, 77
Oct-4, and Sox-2) (B)

Figure 52: Klf4 C-terminal Binding Interactions 82

Figure 53: Modified Diagram of Stem Cell Maintenance Circuitry 84
vii
Abstract
Human and mouse ES cells are defined by their ability to confer self-renewal and
differentiate into various cell types, making them an invaluable tool for investigation of Stem cell
maintenance, differentiation, and reprogramming. Human ES cells have infinite potential for
therapeutic applications. Unfortunately, acquiring ES cells is problematic, and reprogramming
somatic cells to Induced Pluripotent Stem Cells occurs by an unknown mechanism.
This thesis establishes Klf4’s role in ES cell maintenance. Klf4 interacts with core ES cell
maintenance transcription factors, Oct-4 and Sox-2, which bind to Klf4’s C-terminal domain zinc
fingers, and this binding is essential for the collective function of the ES-Cell maintenance network.
Moreover, this study contributes to the knowledge of dose dependant cross-regulation of various
factors in the stem cell maintenance network, and provides a model for understanding how these
factors may confer ES cell stability, self renewal, and reprogramming.

1
Chapter 1. Introduction
Mouse Embryonic Stem cells (mES) are indispensable as an animal model that provides
valuable insight into their human counterpart. Furthermore, mES cells can be cultured in a
completely defined medium without feeder layer support, which provides the advantage of allowing
investigation of molecular interactions pertinent to ES cell biology without the added confusion of
culturing in unknown factors. Over the last few decades mES cells have contributed to a new era in
investigation involving mutants and how they affect ES cell stability and our ability to preferentially
direct differentiation of ES cells into various tissue types. This research has in turn spawned a new
era in the exploration of the potential of hES cells in therapeutic applications.
The basal human drive to preserve the quality of life has been met with ethical and political
controversy. Consequently, scientists have met this challenge with innovative and practical
techniques that have awed the world and inspired much enthusiasm about the future cumulative health
of mankind. Scientists have pioneered a field previously unaddressed by modern science, namely, the
reprogramming of differentiated cells back to their most primitive form, ES cells. ES cells are unique
in their ability to confer self-renewal and pluripotency. The successful and complete reprogramming
of mouse and human somatic cells to ES cells, which can be maintained without becoming
tumorgenic, is further complicated by the mode through which these transcription factors essential for
ES cell identity are delivered. Current techniques utilized for reprogramming employ lentiviral and
retroviral constructs for delivery of these factors into somatic cells. While this approach is usefully
applied to the investigation of reprogramming itself, it is not applicable for therapeutic applications
because these viruses can be spontaneously` re-activated and lead to cancerous formations in the
recipient of such therapy (33).
The ultimate goal of research in the stem cell field is to develop small molecular drugs that
can trigger the cells internal stem cell maintenance machinery such that, the cell can overcome its
differentiated state by achieving proper levels of each transcription factor in the stem cell maintenance
network. Unfortunately, such a goal is currently far fetched unless we first improve our
2
understanding of how ES cell identity is established and maintained on a molecular level. Gaining
such knowledge will provide productive insights into fundamental mechanisms present in ES cells
and provide alternative options for reprogramming terminally differentiated cells to ES cells. While
much of the previous research that has been conducted has paved the way to accomplish this feat,
much still remains to be discovered. Many previous studies focused on the ability of the ES cell’s
transcriptional factors to activate transcription of downstream gene targets involved in pluripotency
and self-renewal, as well as their ability to mediate auto- and cross-transcriptional regulation of genes
encoding the transcriptional factors themselves.
While this is not the focus of this thesis, it also served to address some of these issues.
Alternativly, one must consider that establishing the functional relationships between proteins is
critical for understanding the stem cell maintenance mechanism because a prominent theme
throughout scientific history has proven that protein complexes mediate the majority of cellular
processes in all aspects of cell biology. Consequently, this thesis aims at investigating the protein-
protein interactions that Kruppel-like Factor 4 (Klf4) makes with other stem cell factors and further
defining if these interactions functionally contribute to the maintenance of mES cells.
Klf4 gene is located on chromosome
#
9, and the Klf4 open reading frame encodes a protein
of 470 amino acids that contains several functional domains. These domains include: an activation
domain, a transcriptional inhibitory domain, a zinc finger DNA-binding domain, a nuclear localization
signal, and a potential PEST sequence (dictates the rapid degradation of Klf4 protein) (35) (Figure 1).
3

Figure 1: Structural Representation of Klf4 Domain Deletions
Schematic diagram used to show which functional domains are deleted for each Klf4 mutant. Klf4-
Wt has all domains as it is full length and contains all 470 amino acids. Klf4-ΔM (deletion of amino
acids 155-139) is missing the Nuclear Localization Sequence and Transcriptional Inhibitory Domain.
Klf4-ΔC (deletion of amino acids 402-440) is missing the zinc finger DNA Binding Domain.
Individual Zinc finger locations predicted by Prosite indicated fingers exist. Only the last two C-
terminal Zinc finger deletions were available for use in this thesis.

Previous research on Klf4 in conjunction with the WNT/ β-catenin signaling pathway proves
that Klf4 regulates normal intestinal homeostasis and tumor repression (40). In the presence of Wnt
stimulation GSK-3β mediated degradation of β-catenin is inhibited, which leads to the stabilization
and subsequent nuclear translocation of β-catenin, where it complexes with T-cell Factor and activates
target genes which regulate cell fate and growth (40). In this context Klf4 acts as a tumor suppressor
protein that functions to bind to β-catenin trans-activation domain and inhibit β-catenin mediated
transcription, thus leading to the alternative differentiation response (40). In fact, up until recently,
most research concerning Klf4, involved investigation of many non-stem cell type cell lineages and
4
predominantly supports the theory that Klf4 functions mainly as a tumor suppressor. For instance,
Klf4 has been shown to inhibit proliferation of cells via, blockage of the G1 / S progression of the cell
cycle, repression of Cyclin-D1 promoter activity, and even blocking entry into mitosis following
DNA damage (8, 9, 10). However, strictly labeling Klf4 as an oncogene or tumor suppressor gene is
not accurate as Klf4’s function seems to depend not only on the cellular context but the timing of
expression in development. While Klf4 has been proven to be essential for terminal differentiation in
certain cell types appearing later in development, this thesis suggests a role in earlier development in
the context of stem cells where it functions as a transcription factor involved in maintaining
pluripotency and self renewal of mES cells.
This thesis aims to investigate the network of transcription factors involved in maintaining
ES cell identity and illuminate their contribution in maintaining pluripotency and self-renewal of mES
cells. In this context we propose Klf4 as a stem cell factor not a differentiation factor. Klf4’s possible
role in induction of pluripotency and stem cell maintenance had not been suggested until
reprogramming studies performed by Yamanaka et al., which utilized this Klf4 factor for the
reprogramming of somatic cells to stem cells. Unfortunately, their first publication deemed Klf4 as a
dispensable factor in reprogramming (33). However this thesis proves that Klf4 factor is required for
the maintenance of ES cell self-renewal and pluripotency. Within their discussions in the Yamanaka
paper, the authors have tried to reconcile their observations with existing knowledge and hypothesized
that Klf4 function could be replaced by other oncogenes. Investigators found that Klf4 was a tumor
related factor that could not be replaced by other oncogenes namely T-Ras, Tcl1, β-catenin, and
STAT3, proving that Klf4 did indeed have a unique function in stem cells never anticipated before
(33). This finding was interesting because LIF has been shown to activate β-catenin, which mediates
activation of the STAT3 pathway whose downstream target is C-myc (6). This pathway is well
documented as providing ES cell maintenance in the mouse model. However, the mystery remained
concerning how Klf4 was inadvertently acting as an oncogene that promoted stem cell-like
characteristics in this cellular context of reprogramming. The fact that Klf4 function could not be

5
replaced by these oncogenes led me to speculate that Klf4 was either a downstream factor of STAT3
or that Klf4 was functioning in another parallel pathway that induces stem cell self-renewal and
pluripotency independently. Alternatively, to reconcile their findings in the Yamanaka paper, the
authors postulated that there must be an important balance at play between C-myc and Klf4. C-myc
expression serves to repress p21 which, normally serves to activate Klf4’s anti-proliferative function.
This theory was formulated to coincide with previous literature that had established Klf4’s role as a
tumor suppressor gene. However, a more recent paper by Yamanaka proves that induced pluripotent
mouse stem cells can be produced in the absence of C-myc, which may disprove the importance of the
C-myc and Klf4 interplay (21). Although this publication occurred well after the onset of this thesis
investigation, the findings inspired me to confidently pursue the possibility that Klf4 was a key player
in a parallel pathway in which its relationship with other transcription factors was not only essential
but sufficient for the maintenance of mES cell identity.
Instead of investigating how extrinsic factors induce pluripotency in mES cells, this thesis
hypothesizes that Klf4 is a member of a parallel pathway which either duplicates or undermines the
effects of the extrinsic factors. This approach really serves to illuminate the most primitive and
essential core of the stem cell maintenance circuitry, thus providing a model that may be applicable to
hES in the future. Nanog has been viewed as the master regulator of the stem cell maintenance
circuitry and studies show that Nanog overexpression induces ES cell maintenance in the absence of
extrinsic factor LIF, thus negating the need for activation of C-myc, another factor implicated in stem
cell maintenance (6). Extrinsic factor BMP on the other hand can lead to expression of Nanog
through another pathway. If Klf4 is part of another, more potent pathway, then mutation of the
proteins functional domains or depletion of the protein in the presence of LIF and BMP should share
that same end result of differentiation of these altered stem cells despite the pro-stem cell factors that
these extrinsic factors provide. This thesis proves that this is indeed the case. Other core factors, Oct-
4 and Sox-2, have been shown to transcriptionally regulate the Nanog gene. Therefore, this thesis
hypothesizes that Klf4 functions on the same level or one level above Oct-4 and Sox-2 in the ES cell
6
maintenance hierarchy. It seems to be a common trend among nuclear transcription factors that
factors that bind each other can induce transcriptional or functional regulation of each other. This
thesis proposes that Klf4 binds Oct-4 and Sox-2, and that this binding is essential for ES cell
maintenance.
7
Chapter 2. Materials and Methods
Selection of genes involved in maintenance of pluripotency and self-renewal of MES cells
were obtained from Applied Biosystems gene signature array, Taqman ®Mouse Stem Cell
Pluripotency Array. A total of seven genes were assayed. Three of the genes assayed (Nanog,
Pou5f1, and Sox-2) fit under the category of gene products that are functionally related to
maintenance of pluripotency. Six of the genes (Nanog, Pou5f1 (Oct-3/4), Tdgf1, Dnmt3b, Gabrb3,
and Gdf3) fit under the category of genes expressed in undifferentiated cells. Note Nanog and Oct-4
appears in both categories, and (Tdgf1, Dnmt3b, Gabrb3, and Gdf3) are genes that are
developmentally regulated in humans, which serve as useful markers for the identification of
pluripotent stem cells but their role in maintenance of ES cells is unclear. The Genes were looked up
on ENSEMBLE to obtain identification numbers for future analysis. Regulatory Sequence Analysis
Tools Website was used for the experiment. Gene identification numbers were used to retrieve
sequences of genes, 10,000 bp upstream and 200 bp downstream from the transcriptional start site of
each gene, from Mus Musculus Ensemble Link. The matrixes of Oct/Sox and Klf4 were obtained
form literature (37, 38). Both the Oct/Sox matrix and Klf4 matrix were assayed for the seven genes of
interest. For controls ten separate sets of seven random genes were assayed for Oct/ Sox matrix and
Klf4 matrix. For the two separate experiments, each random set gave a number of hits, these numbers
were entered in a Microsoft excel sheet and averages and standard deviations were obtained.
Observed numbers in selected genes and average numbers in corresponding random sets were
obtained and used for calculation of the Z value from equation,
n sd
observed ected
z
/
exp !
= . For
analysis of Oct/Sox DNA binding motif presence in selected and random set, and analysis of Klf4
DNA binding motif presence in selected and random sets, values n=10 and DF=9 were used for
calculation of Z-value and resulting p values were provided.
For immunostaining of hES, cells fixation was carried out for 30 minutes in 2%
Paraformaldahyde, washed 10 minutes continuously with PBS, permeablized with 0.1% Triton X for
8
10 minutes, and washed with blocking buffer (10% goat serum in PBS) continuously for 15 minutes.
Primary antibodies used from Santa Cruz included: Oct ¾ (H-134) Rabbit Polyclonal, Oct ¾ (C-10)
Mouse Monoclonal, Klf4 (H-180) Rabbit Polyclonal, and Sox-2 (H-65) Rabbit Polyclonal. The
concentrations of all antibodies from Santa Cruz were 200 µg / ml. Mouse β-catenin was from BD
transduction laboratories with a concentration of 250 µg / ml. Each well prepped for immunostaining
was incubated in the primary antibody for 24 hours at 4°C at a dilution of (antibody / 1.5% Goat
Serum in PBS) of, 1:50 for Santa Cruz antibodies, and 1:40 for BD antibodies. The next day wells
were washed 3x with blocking buffer, then incubated with secondary antibodies Rabbit α-Cy3 or
Rabbit α-Cy3 and Rabbit α-Cy2 or Mouse α−Cy2, diluted 1:200 in blocking solution and incubated
at 20°C in the dark for 1.5 hours. Finally wells were washed 3x for 5 minutes in PBS and underwent
final incubation in Hoechst Nuclear Staining diluted at 1:1,500 in blocking buffer for 20 minutes at
20°C in the dark.
Immunostaining of mES cells was carried out with only SSEA1 antibody from University of
Iowa Hybridoma bank. For staining, a dilution of 1:400 of primary antibody / 0.5% goat serum in
PBS (a concentration of 1 µg / ml), was used for incubation for the duration of 2 hours at 20°C.
Samples were then washed 3x in 0.5% goat serum in PBS. Then wells were incubated in secondary
antibody Mouse α-Cy3 at a dilution of 1:1,000 in PBS for 1 hour at 20°C in the dark. Finally wells
were incubated in Hoechst Nuclear Staining at a dilution of 1:2,000 for 20 minutes at 20°C in the
dark.
For all immunostaining, final washes were carried out and slides were set with Floromount
and cover slips. Proper controls of primary and secondary antibodies alone were performed to
eliminate the suspicion of background staining. Results were viewed on a Zeiss inverted fluorescent
microscope and pictures were all calibrated for same exposure time and light intensity per set.
Plasmid DNA amplification was accomplished by transformation into DH5α chemico-
competent cells following the recommended protocol from Invitrogen. Drug selection depended on
the type of vector that the plasmid DNA was ligated into; typically the vectors utilized in this project
9
had regions that conferred resistance to either Amphacillin or Kanamycin. Bacterial cells were then
plated on LB-agar plates infused with drugs for selection at a concentration of 100 mg drug / L and
allowed to incubate at 37°C for no longer than 18 hours. The following day, several colonies were
selected and individually inoculated into a 3 ml cocktail of LB and proper selection drugs at a
concentration of 100 µg drug / ml LB. The samples were placed in shaker at a setting of 220 RPM
and 37°C for 18-21 hours. Further amplification was accomplished by diluting the starter culture
1/250 into drug selective LB medium at a concentration of 100 µg drug / ml LB and growing in
shaker at a setting of 220 RPM and 37°C for 18-21 hours.
Plasmid DNA purification was accomplished by harvesting the bacterial cells by
centrifugation with, Becman JLA-16.250 rotor at 6,500 x g for 30 minutes at 4°C. Excess supernatant
was promptly removed and Qiagen Plasmid Maxi Kit was used for extraction and purification of
plasmid DNA. The gravity flow protocol in the Qiagen plasmid purification handbook was followed
and DNA pellet was washed briefly with 70% ethanol and air dried. Pellet size dictated the amount of
water used to re-suspend the plasmid DNA pellet. To determine the yield, DNA concentration was
analyzed by the spectrophotometry feature of the Nanodropper at a wavelength of 260 nm. Purity was
determined by the A
260
/ A
280
ratio, which in all cases was between 1.8 and 2.0.
Transfection of plasmid DNA was carried out in 293T cells that were cultured in defined
medium of, 1 x DMEM, 10% FBS, 1% penicillin-streptomycin, and 1% glutamine. For 6 cm tissue
culture dishes cells were plated at 0.5 x 10^6 cells per dish in a total volume of 4 ml media 24 hours
prior to transfection. For 10 cm tissue culture dishes cells were plated at 1 x 10^6 cells per dish in a
total volume of 10 ml media 24 hours prior to transfection. Media was changed and cells were treated
with 0.5 mM chloroquine approximately 15 minutes prior to transfection to increase efficiency.
Calcium chloride (CaCl) transfection method was used with 2 molar CaCl and transfection buffer at a
pH of 7.15. Volume of transfection cocktail for 6 cm dish transfection was 0.5 ml, while for the 10
cm dishes the transfection cocktail volume was 1 ml. Plasmid DNA concentration ranged from 0.01
µg to 4 µg for Luciferase assays, 6 µg to 12 µg for overexpression studies, and was approximately 28
10
µg for all virus production. Total volume of media and transfection cocktail for 6 cm dishes was 4 ml
and for 10 cm dish total volume was 7 ml. Calcium chloride facilitated DNA precipitation was
allowed to occur for 6 hours in the incubator, after which media was changed, and for overexpression
studies transfection efficiency was optimized by adding 1 molar sodium butyrate. Then cells were
further incubated for 42 hours prior to lysis or viral collection.
For Luciferase assays 0.5 x 10^ 6 cells were plated in 12 well dishes and duplicate dishes
were made for each transfection condition. No chloroquine was used to increase transfection
efficiency and media was changed on the morning of transfection to 0.8 ml per well. Each
transfection mixture was calibrated with PCDNA to make the total concentration of DNA the same
for each reaction. Total transfection mixture volume including DNA was 0.5 ml, but only 0.2 ml was
added to each dish giving a final total volume of 1 ml while CaCl transfection reaction occurred, and
6 hours later media was changed to fresh media, 2 ml per well. Cells were washed in PBS 3x and
lysed in 1 x Passive Lysis Buffer for 15 minutes at 20°C, 24 hours post transfection. Supernatant was
collected and loaded in 96 well plates at 20 µl per well. The Spectra Max M5, Molecular Devises,
Microplate Reader was used for sequential assay of Firefly and Renilla Luciferases. First Firefly
Luciferase activity was assayed using Lars II Buffer, followed analysis of Renilla Luciferase activity
by Stop and Glow Buffer. Controls included cells not transfected with any DNA designated blank,
cells transfected with only PCDNA, and cells transfected with Topflash and Renilla to expose
background Luciferase activity that would later be subtracted out of data when analyzed. Data
analysis methods subtracted out background Luciferase activity and ratio Firefly / Renilla of
Luciferase activity was re-calibrated to give the proper fold change between samples. Duplicate wells
were analyzed and averages of each duplicate data point were taken, as well as standard deviation for
each data point. β-catenin induced Luciferase activity is accomplished by the ability of β-catenin to
bind to Tcf factor. TCF in turn binds to its promoter and drives Luciferase gene expression.
For overexpression studies 0.5 µg of FUIGW, GFP control was added to each transfection
cocktail facilitating the assessment of transfection efficiency. Approximately 48 hours post-
11
transfection, 70 % - 80 % of cells expressed GFP confirming high transfection efficiency. Cells were
washed 3x with PBS to remove excess media, then lysis was carried out with either, KLB or RIPA
lysis buffers and a protease inhibitor cocktail for 30 minutes at 4°C, after which dishes were scraped
and transferred to Eppendorf tubes and centrifuged for 15 minutes at 14,000 RPM and 4°C. Protein
lysate concentrations were then determined using the Spectramax M5, molecular devices, Microplate
readers BSA protein quant assay. Protein standard curve was made using BSA Protein standard from
Pierce, and working reagent from Pierce was used to dilute the protein lysates 20:1 for analysis. For
each experiment all amounts of protein lysates were calibrated to be equal for traditional western
blotting and immunoprecipitation studies.
Ten percent acrylamide gel was used for Western blots of protein lysates and
immunoprecipitation of protein lysates. Acrylamide gel was transferred to Immunobilon-P transfer
membrane at 4°C for 3 hours at 250 Amps, blocked for 1 hour in 5% milk / TBST, incubated with
primary antibody overnight, washed 3x for 10 minutes, incubated in secondary antibody for 1 hour,
washed 5x for 10 minutes, incubated in Santa Cruz Western Blot Luminescent Reagent for 2, minutes
exposed to film, and developed by Kodak X-Omat 2,000A Processor.
Immunoprecipitation was performed on protein lysates of overexpression studies and
endogenous level studies. For endogenous protein interactions 1,000 µg -1,500 µg of protein lysates
were incubated with primary antibody at 4°C on rotator for 4 hours, then 10 µl of Pierce Immobilized
Protein A/G beads were added and incubated at 4°C on rotator over night. For overexpression studies
all constructs cloned into vectors were attached to an experimentally introduced tag, which further
facilitated protein isolation via commercially available tag conjugated beads. For Flag tagged
constructs Mouse derived α-flag M2 agarose beads from Sigma were used. Once these beads were
activated via washing several times with KLB, 10 µl were added to each Eppendorf tube containing
500 µg protein lysate and put on rotator overnight at 4°C. The next day beads were washed 8x with
KLB, then beads were re-suspended with 1 x SDS and put in 95°C heat block for 5 minutes.
Supernatant was promptly used for western blotting. After primary and secondary antibody
12
incubation followed by film development to view preliminary results, blots were re-hydrated in PBS
and washed in TBST. Restore Western Blot Stripping Solution from Pierce was applied to blot for 20
minutes in 37°C incubator and then for an additional 10 minutes on shaker at 20°C. Stripped blots
were then re-assayed to confirm that the beads used were indeed pulling down the proper protein.
Appropriate antibodies were used to blot for corresponding protein presence.
For virus concentration three dishes of each plasmid DNA type were transfected and
collected 48 hours post transfection, filtered with 0.45 Micron Cellulose Acetate filter and were ultra-
centrifuged for 1.5 hours at 28,000 RPM. Virus pellet was re-suspended in 100 µl of PBS, 2 µl of
which was used to test the virus titer. Post Virus purification the Klf4-WT pellet size was smaller
than that of Klf4-ΔC and Klf4-ΔM. The pellet size for Klf4-ΔM and Klf4-ΔC were the same size so
only one was used to test titer needed to infect mES cells. These constructs were flag tagged,
infection of 293T cells was carried out for two days and cells were lysed and western blotted for
detection of flag tag (Figure 2).
Western Blot shows that the flag tagged virus constructs were detected at a predicted location
relative to ladder, and that in fact, the Klf4-WT infection was weaker. To ameliorate this discrepancy,
twice the amount of Klf4-WT virus was used to infect mES cells as compared to the other two
viruses. To further test virus infection efficiency and condition, FUIGW lentivirus was also made and
used to infect a dish of mES cells at the same time infection with other Klf4 constructs occurred. GFP
expression was detected 1 day post infection and the virus was strong enough to infect almost 40% of
the cells (Figure 3). This percentage is typical for mES cells as they are quite difficult to infect
compared to other cell types. Additionally, Chunming Liu tested Klf4-siRNA efficiency prior to
donating it to us by assaying Klf4 RNA knock down with several siRNA constructs.
13

Figure 2: Lysates of 293T Cells Infected With 2µl Klf4-WT and Klf4-ΔC
293T cells infected with 2µl of each virus type, incubated for 2 days, Lysed with Ripa buffer,
concentrations analyzed by BCA protein quant. assay, and 30µg protein lysate loaded in 10%
Acrylamide gel. Immuno-blotted with α-Flag-HRP antibody.
14

Figure 3: FUIGW Virus Test One Day Post Infection
GFP expression 24 hours post-infection of MES cells with FUIGW Virus Control. Light Field view
of MES cells (left), GFP filter (right).
15
Due to the difficulty of infecting CF1 feeder free mES cells, a higher titer was used for
infection of this cell type typically 30 µl–40 µl of virus was used. Virus infection was facilitated by
adding 8 µg / ml of Polybrene, and media was changed the following morning. Klf4 mutant
constructs were cloned into FUIPW vectors which confers Puromyocin resistance, which, allowed for
Puromyocin destruction of mES cells that were not sufficiently infected with lentivirus. In addition,
lentivirus of the FUIPW vector alone was used as a control to test if virus infection induced
differentiation and for later calibration for RNA studies. Two days post viral infection of mES cells
Puromyocin selection was started at 1µg / ml per day in single cell suspension and continued 3 days
past when negative control was entirely killed off. Additionally, PLKO-siRNA-Klf4 virus and PLKO
vector control virus were Puromyocin selected throughout the experiment. LIF and serum were
provided the entire time of mES cell culture. Cells were expanded to three or more 6 well dishes and
analyzed for 3 passages. For each passage, 1.5 x 10^6 cells were plated in 6 well tissue culture dishes
for alkaline phosphatase analysis, 0.1 x 10^6 cells were plated on 8 well slides the day before the rest
of the cells would be analyzed, such that they could be fixed on the same day cells were collected for
RNA extraction and FACS sorting. The next day duplicate dishes of cells were passaged for the next
round of analysis and a fraction was collected for RNA extraction and for fixation for future FACS
sorting, thus giving results for each passage.
Mouse ES cell defined culture media contained, 1x GMEM, 15%FBS, 1% nonessential
amino acids, 1% sodium pyruvate, 1% glutamate, 1% penicillin streptomycin, 0.1% 2-β-
mercaptoethanol. Tissue culture dishes were coated with 0.1% gelatin. LIF was provided to prevent
spontaneous differentiation of mES cells. Changing of media was required every other day to provide
fresh and potent LIF. Additionally, passage of mES cells was maintained every other day to inhibit
over crowding of mES cell colonies which could also lead to spontaneous differentiation.
Mouse Derived SSEA-1 (mc-480) was obtained from the University of Iowa, Developmental
Studies Hybridoma Bank. Mouse ES cells were washed 2x in PBS to remove dead cells, then
trypsinized in 0.25% 1x Trypsin for 3 minutes. Cellular pellet was washed 2x in PBS and fixed in 1%
16
Paraformaldahyde for 20 minutes, and re-suspended in FACS buffer (5% FBS in PBS). For SSEA1
staining cells were counted and 5x10^4 cells were used per vile. SSEA1 primary antibody incubation
was accomplished by incubation at 20 µg / ml for 5 minutes at 20°C, followed by washing 3x with
FACS buffer, secondary incubation with a 0.75 µg / ml dilution of Mouse α-Cy2 for 5 minutes, and
finally samples were washed and re-suspended in FACS buffer. Proper negative and positive controls
were used but isotope controls were not required. Negative controls included healthy MES cultured
and passaged at the same time as viral infected cells that were incubated with primary antibody alone
or secondary antibody alone. The positive controls were these healthy mES incubated in both
antibodies together, and was used for calibration of the FACS sorting BD ISA 2 special order system
machine parameters and set up analysis gate prior to running other samples. Analysis software used
was BD Facsdiva and samples were analyzed with a FITC filter.
Alkaline phosphatase detection kit from Sigma / Chemicon International was used initially
for preliminary results of PLKO-siRNA-Klf4 infections giving a visible red staining to positive
colonies. Then later Alkaline Phosphatase Substrate 3 from Vector Laboratories was used giving a
high sensitivity visible blue staining for positive colonies. Cells were fixed with 4%
paraformaldehyde in PBS for 20 minutes at 20°C. Then cells were washed with TBS and 100 mM
Tris-HCl at pH 8.5, as PBS is known to interfere with alkaline phosphatase enzymatic activity post
fixation. Then the reaction mixture was allowed to incubate for 30 minutes at 20°C in the dark.
Further washing with TBST removed excess staining residue and results were visualized using
inverted florescent microscope with a colored camera feature.
A fraction of cells were washed 3x in PBS and re-suspended in Trizol and pumped through a
syringe with a 24 gauge needle to homogenize cell lysates and immediately frozen on dry ice and
stored in -80°C freezer. Upon thaw approximately 200 µl of chloroform was added to tube and
vortexed for 15 seconds, and then centrifuged for 50 minutes at 4,000 RPM and 4°C. The uppermost
layer containing the RNA was removed and ethanol was added such that final concentration of
ethanol was 35%. Then supernatant and precipitate was applied to a Qiagen RNeasy mini column and
17
protocol from the RNeasy mini handbook for animal cell RNA purification was followed. The RNA
concentration was assayed at 280 nm using the Nanodropper. The A
260
/ A
280
ratio was between 1.9
and 2.1 for all RNA samples, confirming good RNA purity and stability.
CDNA was produced from 0.6 µg of RNA for q-PCR, and 0.9 µg of RNA for RT-PCR. The
protocol from Invitrogen Super Script III First-Strand Synthesis System for RT-PCR was followed
using random primers. The cDNA synthesis program was performed in the Eppendorf PCR machine
at serial settings of 50 minutes at 50°C, 10 minutes at 25°C, and 50 minutes at 50°C, 5 minutes at
85°C, and finally destruction of residual RNA was carried out for 20 minutes at 37°C. Triplicates of
every condition were made for each primer pair and Gapdh was the endogenous control used.
FUIPW from each passage was used to calibrate the log fold changes observed in that days respective
FUIPW-Klf4-WT and FUIPW-Klf4-ΔC infected mES cell q-PCR readings. Primer designs were
obtained from previous publications (9, 12). Syber green super mix from Biorad was added to each
well along with 10 m moles of each proper primer per set. The Applied Biosystems 7900 HT Fast
Real Time PCR System was utilized for q-PCR with serial PCR settings of 10 minutes at 95°C, 50
cycles of 15 seconds at 95°C, and 1 minute at 60°C, with data collection occurring only at 60°C. Data
was analyzed with RQ manager software and was calibrated for internal control, Gapdh. Q-PCR
results showed the largest differences in slopes of FUIPW, FUIPW-Klf4-WT, and FUIPW-Klf4-ΔC at
around 25 cycles of PCR. This information was used to select the number of cycles used for RT-PCR
to gather data before plateau is reached at 50 cycles of PCR. The 25 cycle condition was used for RT-
PCR of the following primers Gapdh, and Neuro-d. Alternatively 35 cycles were used for, Fgf8,
Brachyury, Fox a2, and Nkx2.5. Taq-DNA polymerase was used for amplification of target cDNA.
The Eppendorf PCR machine settings were as follows: 3 minutes for 95°C, followed by 25 or 35
cycles of 30 seconds at 94°C, 30 seconds at 58°C, 30 seconds at 72°C, and 4 minutes at 72°C.
Amplification products were loaded in 1% agarose gel with 0.5 mg/ml ethidium bromide in TAE
Buffer. Products were analyzed in an Alpha Innotech Gel box.
18
Klf4 constructs, (Klf4-WT, Klf4-ΔC, Klf4-ΔM, Klf4 Δ-ZF
#
429, Klf4 Δ-ZF
#
454) plasmid
DNA were a generous donation from Dr. Chunming Liu. Unfortunately, constructs were in a plasmid
vector not useful for lentivirus production, so they were cloned into the vector FUIPW. Insertion site
of constructs into vector was, after the Ubiquitin promoter, with an internal ribosomal entry site after
the insert, to drive Puromyocin resistance expression in infected cells. Klf4 constructs were digested
out of the existing vector in two steps to make a single blunt end ligation. Constructs were first
digested with EcoR1 and EcoR1 buffer for 2 hours at 37°C, then 100 µ Moles DNTP’s and T4
polymerase were added and incubated at 16°C for 15 minutes to create a blunt end cut site.
Purification was accomplished by adding phenol-chloroform and spinning such that finally DNA was
contained in the top layer. This layer was removed, chloroform was added, and top layer was
collected again. Finally precipitation was accomplished by adding 1/10
th
the volume of sodium
acetate (NaAc) and 2.5x the volume ethanol and incubating at -20°C for 30 minutes. Pellets were
obtained, speed-vac dried, and re-suspended in water.
The second step of digestion was then performed with BamH1 in Buffer
#
2 and BSA for 3
hours at 37°C. Then a small volume was analyzed on a 1% Agarose gel to verify the proper size of
Klf4 inserts. Meanwhile, FUIPW vector was digested with BamH1, Hpa1 and shrimp alkaline
phosphatase in Buffer
#
4 for 3 hours at 37°C. Then all samples were precipitated again, this time
using 5 µg Glycogen, 1/10
th
volume NaAc, and 2.5 x Ethanol and incubating for 30 minutes at -20°C.
Pellets were obtained, speed-vac dried and re-suspended in water.
Concentrations were taken with Nanodropper and approximately, 50 ng vector FUIPW, and
250 ng KLf4 insert were used for overnight ligation with Promega T4 Ligase at 16°C. The ligation
products were then transformed into DH5α chemico-competent cells and plated on Amphacillin LB
plates overnight at 37°C. Colonies were picked and inoculated into drug infused LB and plasmid
DNA was extracted using QIA Spin Mini-prep Kit Protocol. A single blunt end and single sticky end
ligation tactic was used to ensure insert would be correctly oriented. For blunt-blunt end ligation site
Hpa1 made a blunt end cut in the vector, and T4 polymerase was added to fill in the sticky end created
19
by EcoR1 in the Klf4 insert constructs. For the sticky end ligation the vector was cut with BamH1
and the insert was also cut with BamH1 (Figure 4).

Figure 4: Cloning Klf4 Constructs into FUIPW Vector
Cut sites utilized in the vector were Xba1, BamH1, PME1, and Hpa1. Cut sites utilized for digestion
of Klf4 constructs out of the PCDNA 2 vector were, BamH1 and EcoR1. Sticky end of vector and
insert BamH1sites were ligated. EcoR1 cut site was artificially blunted with T4 polymerase and
DNTP’s. Vector was cut with Hpa1creating, a blunt end. Blunt ends (modified EcoR1 and Hpa1 cut
sites) were ligated. Inserts were later digested out with PME1 and Xba1.
20
However, for digestion of candidate FUIPW-Klf4-constructs mini-prepped DNA, enzymes
PME 1 and Xba 1 in Buffer
#
4 were used. This enzyme combination excised part of the vector as
well, such that all inserts would be shifted upward on the agarose gel approximately 3 Kb. Proper
clones were sequenced with w009 Ubiquitin primers at the USC DNA Core Facility and results were
verified using NCBI Blast and Alignment Software.
Unfortunately, Sox-2 and Oct-4 were ligated into unknown vector so primers were designed
to produce various truncation mutants of Sox-2 and Oct-4 (Figure 5, 6). Additionally, tags were
introduced into the C-terminus of these constructs, (HA for Sox-2 and MYC for Oct-4), to facilitate
future isolation and identification of these exogenously introduced constructs (Figure 7, 8).

Figure 5: Amino Acid Sequences of Each Segment Sox-2
Sox-2 primers were designed such that, different combinations of these primers would give the
following protein regions. DNA sequences of each segment were predicted according to primer pair
used and translated with Prosite Software. Segments 1-6 were created. The segments are designed as
progressive serial truncation mutants, starting with truncation of the C-terminus of Sox-2 and going to
full length, then by serial truncation of the N-terminus. Colors represent each fragment of Sox-2,
Orange is the N-terminal fragment, blue is the middle fragment, and Gray is the C-terminal fragment.
21

Figure 6: Amino Acid Sequences of Each Segment Oct-4
Oct-4 primers were designed such that, different combinations of these primers would give the
following protein regions. DNA sequences of each segment were predicted according to primer pair
used and translated with Prosite Software. Segments 1-7 were created. The segments are designed as
progressive serial truncation mutants, starting with truncation of the C-terminus of Oct-4 and going to
full length, then by serial truncation of the N-terminus. Colors represent each fragment of Oct-4,
Orange is the N-terminal fragment, fragment 2 is red, fragment 3 is green and the C-terminal fragment
is purple.
22

Figure 7: Sox-2 Mapping
The segments are designed to isolate the three fragments, 1) N-terminal fragment is orange, 2) middle
fragment is blue, and 3) C-terminal fragment is gray. 4) Is the full length and has all three fragments.
5) Contains fragment one and two. 6) Contains fragment two and three. HA tags are introduced at the
new C-terminus of each clone, and experimentally introduces cut sites are indicated.
23

Figure 8: Oct-4 Mapping
The segments are designed as progressive serial truncation mutants, starting with truncation of the C-
terminus of Oct-4 and going to full length, then by serial truncation of the N-terminus. N-terminal
fragment is red, fragment 2 is green, fragment 3 is blue and C-terminal fragment is purple. 1)
Contains only the N-terminal fragment one. 2) Contains fragment one and two. 3) Contains fragment
one, two, and three. 4) Is the full length and has all four fragments. 5) Contains fragment two, three,
and four. 6) Contains fragment three and four. 7) Contains only the C-terminal fragment four. MYC
tags are introduced at the new C-terminus of each clone, and experimentally introduces cut sites are
indicated.
24
KOD Hot Start Polymerase Protocol from Navogen was followed and proper primers sets were mixed
with plasmid DNA of Sox-2 and Oct-4. PCR amplification was carried out on Eppendorf PCR
machine with sequential cycles of, (Polymerase activation at 95°C for 2 minutes, and 25 cycles of
(denaturation at 95°C for 20 seconds, annealing at 55°C for 10 seconds, and extension at 70°C for 15
seconds)). PCR Products were loaded in 1% agarose gel, bands were excised, and purification was
accomplished by following QIA-quick Gel Extraction Kit Protocol.
KOD Polymerase generated blunt ended PCR products. PCR Products for Sox-2 were
directly ligated overnight with T4 Ligase at 16°C into a zero-blunt end vector from Invitrogen (Figure
9). This ligation mixture was transformed into chemico-competent cells and plated on LB dishes
infused with 50 µg / ml Kanamycin, and pre-treated for 3 hours at 37°C with, 0.8 µg X-gal diluted in
DMF ( dimethylsormamide) and 0.8 µg of IPTG diluted in H
2
O, to identify possible recombinants.
White colonies were picked and inoculated into LB with 100 µg / ml Kanamycin and amplified in a
shaker at 37°C. Samples were mini-prepped and a small portion was digested with EcoR1 in EcoR1
Buffer for 1 hour at 37°C, and loaded into 1% agarose gel for analysis. Once clones with proper
insert were identified, the rest of the sample was digested and loaded in gel, and proper insert was
excised and gel purified. Similarly, FUIGW vector was digested with EcoR1 in EcoR1 Buffer for 3
hours at 37°C, then treated with shrimp alkaline phosphatase for 30 minutes and loaded in 1% agarose
gel, and proper bands were excised and gel purified as well. Ligation of 50 ng of vector FUIGW and
250 ng of Sox-2 inserts were performed overnight at 16°C with Promega T4 Ligase (Figure 10). The
ligation reaction mixture was transformed into chemico-competent cells and resulting colonies were
inoculated into LB infused with 100 µg / ml Amphacillin for amplification. DNA was promptly
isolated via mini-prep, and digested with Hpa-1 and BamH1 for 1 hour at 37°C in Buffer
#
4 to check
proper insert ligation. The size of observed bands were the same as previously predicted for PCR,
because BamH1cut site is directly next to EcoR1 cut sight in FUIGW vector.
25

Figure 9: Map and Features of PCR Blunt
Map features of Blunt End Sub -Cloning Vector. Blunt ended insert is ligated into vector such that, it
is in between the EcoR1 cut sites of the vector, sites are still intact and are utilized later to remove the
Insert, leaving the insert with sticky ends. Vector confers Kanamycin resistance and Zeocin
resistance, facilitating drug selection. Ligation of blunt PCR Fragment disrupts expression of the
LacZα-ccdB gene fusion, permitting growth of only positive recombinants. Host Ecoli strains that do
not contain LacZI, lack constitutive activity of the Lac promoter, therefore, IPTG and X-gal are need
to identify possible recombinants.
26
Furthermore, Hpa1 cut site was experimentally introduced directly after the HA tag of the
insert through primer design. Additionally, this cloning design checked for orientation in one
digestion step, as the absence of a band would indicate incorrect orientation of insert in vector (Figure
10).

Figure 10: Cloning of Sox-2 Constructs into FUIGW Vector
Cut site utilized to digest the vector was EcoR1. Similarly, Sox-2 inserts were isolated by digesting
them out of the Zero Blunt End Sub-Cloning Vector, with EcoR1. EcoR1 sticky ends of vector and
insert were ligated. To confirm proper orientation of insert cut sites BamH1 of the vector and
experimentally introduced Hpa1 were utilized. Improper orientation of the insert would give no
detectable band upon digestion, because it would produce a 13 bp fragment. Segment size would be
the same as indicated from PCR because digestion scheme only adds 13bp to the insert, an
undetectable difference.
27
Proper clones were sequenced with w009 Ubiquitin primers at the USC DNA core facility
and results were verified using NCBI blast and Alignment Software. Oct-4 KOD PCR products were
digested with enzymes Bgl2 and EcoR1 in EcoR1 Buffer for 3 hours at 37°C, digestion products were
then loaded in 1% agarose gel, excised, and gel purified. FUIGW vector was digested with BamH1
and EcoR1 in EcoR1 Buffer for 3 hours at 37°C, then treated with SAP for 30’ at 37°C. Finally,
FUIGW was loaded in 1%EDTA Agarose gel, excised, and gel purified. Ligation of 50 ng vector
FUIGW and 250 ng Oct-4 inserts were performed overnight at 16°C with Promega T4 Ligase (Figure
11). The ligation reaction mixture was transformed in chemico-competent cells and resulting colonies
were inoculated into LB infused with 100 µg / ml Amphacillin for amplification. DNA was promptly
isolated by mini-prep and digested with Xba1 in Buffer
#
2. Because Oct-4 was suspected to have a
BamH1 cut sight at the N-terminus Bgl2 was included in the N-terminal primer design. Once Bgl2 is
cut, its 5’ overhang can be ligated with BamH1 5’overhang. However, once this ligation occurs the
sight is not viable for further enzymatic digestion, therefore, Xba1 was used as it cuts the vector in
only two specific spots up and downstream of the insert (Figure 11).
Digestion with Xba1 creates a fragment of 1.4 Kd, so each predicted insert size was shifted
upward by this amount. After first round of sequencing and alignment it was discovered that no
BamH1 cut site existed in the N-terminus of Oct-4 plasmid DNA we received, so a combination of
Xba1 and EcoR1 were used for future digestion of cloning candidates. This should create three
segments, vector fragment from cut sites Xba1-Xba1 at 9.2 Kb, another vector fragment from cut sites
Xba1- EcoR1 at 1.3 Kb, and inserts should show up at same size predicted from PCR because Xba1
cut site of vector is only 6 bp upstream from BamH1 cut site of vector (Figure 11). Proper clones
were sequenced with w009 Ubiquitin primers at the USC DNA core facility and results were verified
using NCBI blast and Alignment Software.
28

Figure 11: Cloning Oct-4 Constructs into FUIGW Vector
Cut site utilized in the vector were Xba1, BamH1, and EcoR1. Cut sites utilized for digestion of Oct-
4 constructs was Bgl 2, and EcoR1. Sticky ends of vectors BamH1site and Inserts Bgl 2 site were
ligated. Vector and insert were both cut with EcoR1 and these two sticky ends were ligated. Inserts
were digested out by one of two methods. First method was Xba1 alone, which gives a vector
fragment of 9.2 kb, and an insert that is shifted upward 1.4 kb. Second method was Xba1with EcoR1,
which gives two vector fragments at 1.4kb and 9.3 kb, and inserts are the same size as predicted by
PCR.
29
Chapter 3. Results
Biostatistical queries were answered using RSAT to investigate the enrichment of Klf4 and
(Oct / Sox) DNA-binding elements in upstream regions of well known ES cell maintenance and
pluripotency genes. Genes used were, Gdf3, Dmt3b, Oct-4, Tdgf1, Gabrb3, Sox-2, and Nanog. For
controls ten separate sets of seven random genes were assayed for Oct/ Sox matrix and Klf4 matrix.
Results indicated that Klf4 was significantly more enriched in this gene set as compared to
several sets of random genes, p<0.001 ( Figure12b). In addition the Oct4/Sox2 DNA-binding motif
was also significantly enriched in these same genes, p<0.001 (Figure 12a).
Immunostaining studies were carried out on Human Embryonic Stem Cells (hES-2) cultured
on feeder layer. Staining with antibodies for Klf4, Oct-4 and Sox-2 were performed to investigate the
expression and localization of these transcription factors. All three transcription factors had a very
strong expression in the nucleus (Figure 13, 14). Klf4 was expressed throughout the nucleus with
considerable concentrated expression in the nucleoli (Figure 13). There was an overlap in expression
for Oct-4 and Klf4, confirming their co-expression; yet even more interesting was the co-localization
of Klf4 and Oct-4. It appears that there were nuclei in which nuclear Hoechst staining was visible
thus confirming the presence of a nucleus, yet this region lacked both Klf4 and Oct-4 expression
(Figure 15). Additionally co-expression studies of Oct-4 and Sox-2 were carried out and an overlap
of Oct-4 and Sox -2 expression was observed (Figure 14).
30

Figure 12 a,b Enrichment Studies
A) Enrichment of (Oct/ Sox) DNA Binding Sites in Pluripotency Linked Genes B) Enrichment of
Klf4 DNA Binding Sites in Pluripotency Linked Genes. Genes used were, Gdf3, Dmt3b, Oct-4,
Tdgf1, Gabrb3, Sox-2, and Nanog. Both the Oct/Sox matrix and Klf4 matrix were assayed for the
seven genes of interest. For controls ten separate sets of seven random genes were assayed for Oct/
Sox matrix and Klf4 matrix. For the two separate experiments, each random set gave a number of hits;
these numbers were used to acquire standard deviations. Observed number in selected genes and
average number in corresponding random sets were obtained and used for calculation of the Z value
form equation
n sd
observed ected
z
/
exp !
= . For Oct/sox Matrix analysis n=10, and DF=9, SD =
4.34, Ex=18, and Ob=35. So Z= ((18-35 / (4.34/ sq rt 10)) = -12.39, with DF= 9, this gives a p-value
(p < 0.001). Motifs per gene=Hits / number of genes, Ob=35/7=5 motifs per gene, and Ex=18/7= 2.5
motifs per gene b) KLf4 Matrix analysis, n=10, and DF=9, SD = 11 Ex=63.4, and Ob=176. So Z=
((63.4-176 / (11/ sq.rt 10)) = - 32.45, with DF= 9, this gives a p-value (p < 0.001). Motifs per
gene=Hits/ number of genes, Ob=176/7=25 motifs per gene, and Ex=63.5/7=9 motifs per gene.
31

Figure 13: Human Embryonic Stem Cells, Klf4 / Oct-4 / Nuclear 63x
Primary antibodies used from Santa Cruz included, Oct ¾ (C-10) Mouse Monoclonal, Klf4 (H-180)
Rabbit Polyclonal. Secondary antibodies used were Rabbit α-Cy2 (green) and Mouse α−Cy3 (red).
For nuclear detection, Hoechst Nuclear Staining was used. Inverted Florescent microscope was used
at a magnification of 63x. Filters used show Klf4, Oct-4 and nuclear staining individually, and in
overlay.
32

Figure 14: Human Embryonic Stem Cells, Oct-4 / Sox-2 / Nuclear 40x
Primary antibodies used from Santa Cruz included, Oct ¾ (C-10) Mouse Monoclonal, and Sox-2 (H-
65) Rabbit Polyclonal. Secondary antibodies used were Rabbit α-Cy2 (green) and Mouse α−Cy3
(red). For nuclear detection, Hoechst Nuclear Staining was used. Inverted Florescent microscope
was used at a magnification of 40x. Filters used here show Oct4, Sox-2, and nuclear staining
individually, and in overlay.
33

Figure 15: Human Embryonic Stem Cells, Klf4 / Oct-4 / Nuclear 20x-Magnified.
Primary antibodies used from Santa Cruz included, Oct ¾ (C-10) Mouse Monoclonal, Klf4 (H-180)
Rabbit Polyclonal. Secondary antibodies used were Rabbit α-Cy2 (green) and Mouse α−Cy3 (red).
For nuclear detection, Hoechst Nuclear Staining was used. Inverted Florescent microscope was used
at a magnification of 20x. Filters used show Klf4, Oct-4, and nuclear staining individually, and in
overlay. Yellow and White boxes serve to draw attention of areas of co-localization of Oct-4 and
Klf4, these regions are amplified in the connecting figure.
34
Furthermore, staining of Embryonic bodies (EB) with Klf4 and Oct-4 showed undetectable expression
even at high exposure (Figure 16, 17).

Figure 16: Human Embryonic Bodies, Klf4/ Nuclear
Primary antibody used from Santa Cruz, Rabbit Polyclonal, Klf4 (H-180). Secondary antibody used
was Rabbit α-Cy3 (red). For nuclear detection, Hoechst Nuclear Staining was used. Inverted
Florescent microscope was used at a magnification of 20x and 63x. Filters used show Klf4 and
nuclear staining individually, and in overlay. Exposure is turned up much higher than usual to ensure
no Klf4 staining is detected

35

Figure 17: Human Embryonic Bodies, Oct-4/ Nuclear
Primary antibody used from Santa Cruz was Oct ¾ (C-10) Mouse Monoclonal.. . Secondary antibody
used was Mouse α-Cy3 (red). For nuclear detection, Hoechst Nuclear Staining was used. Inverted
Florescent microscope was used at a magnification of 63x. Filters used show Oct-4 and nuclear
staining individually, and in overlay. Exposure is turned up much higher than usual to ensure no Oct-
4 staining is detected.

Human ES cells and EB’s were lysed, and resulting protein lysates of each sample were
calibrated to be equal for western blots. BCA Protein Quantification Assay was performed for
determination of protein concentrations. Western blots of whole cell lysates of hES and EB showed
higher amounts of stem cell factor Sox-2 in hES cells as compared to EB, whereas Oct-4 western blot
did not show a profound difference in the Oct-4A version of the protein but the smaller version Oct-
4B form appeared to be higher in EB (Figure 26). Interestingly Klf4 seemed to have an alterative
form of the protein in EB that was smaller than the full length. Unfortunately, this version was not
isolated for sequencing and we cannot speculate as to what its function may be or where it is localized
(Figure 18).
36

Figure 18: Comparative Expression of hES and EB
Comparison of transcription factors expression between hES cells and Embryonic Bodies. Primary
antibodies used from Santa Cruz included, Oct ¾ (H-134) Rabbit Polyclonal, Klf4 (H-180) Rabbit
Polyclonal, and Sox-2 (H-65) Rabbit Polyclonal. Secondary antibody used for each blot was Goat α-
Rabbit, and 20µg of each type of protein lysate were used for each lane. Blots were stripped and
stained for Mouse Monoclonal Actin, as a loading control.
37
Overexpression studies were carried out in 293T cells because they do not have endogenous
expression of any of the factors assayed in this thesis. They were transfected with appropriate DNA
for each experimental parameter and lysed two days post-transfection. Resulting protein lysates were
calibrated such that each sample of western blot or Immunoprecipitation has equal amounts of protein
lysate. Calibration of protein lysate was facilitated by the BCA Protein Quantification Assay. In each
case, PCDNA was used as a negative control, and a means of equalizing the transfection conditions
for each experiment. Overexpression studies of Flag fused Klf4 constructs in 293T cells revealed
interaction with transcription factors essential for stem cell self-renewal and pluripotency. Klf4
deletion mutant Klf4-ΔM was missing amino acids 155-399 which coded for nuclear localization
sequence and the transcriptional inhibitory domain. Klf4 Deletion Mutant Klf4 ΔC was missing
amino acids 402-470 which coded for the DNA binding zinc finger DNA binding domain (Figure 1).
Test run of immunoprecipitation of Klf4 deletion mutants and β-catenin revealed that β-catenin can
still bind to Klf4 mutant constructs, Klf4-ΔC, Klf4-ΔM, Klf4-ΔZF-429, and Klf4-ΔZF 454 (Figure 19,
20).
38

Figure 19: A) Lysate Controls: Klf4 and β-catenin Binding, a Mono-Flag
Overexpression of Klf4 (3 µg) constructs with β-catenin (3 µg). Lysate controls were immuno-
blotted with Mouse-Monoclonal α- Flag antibody to show the expression of various flag tagged Klf4
deletion mutant constructs. Secondary antibody used was Goat α- Mouse. B) Lysate Controls:
Klf4 and β-catenin Binding, a Poly-HA Overexpression of Klf4 (3 µg) constructs with β-
catenin (3 µg). Lysate controls were immuno-blotted with Rabbit-Polyclonal α-HA antibody to show
the expression of HA tagged β-catenin constructs. Goat α-Rabbit was used for secondary antibody

39

Figure 20: Immunoprecipitation: Klf4 and β-catenin Binding
Overexpression of Klf4 (3 µg) constructs with β-catenin (3 µg). A) Immunoprecipitation with Mouse
derived α-flag M2 Agarose beads to pull down Klf4 constructs, and Immuno-blotted with Rabbit-
Polyclonal α-HA to detect bound β-catenin. . Goat α-Rabbit was used for secondary antibody. B)
Blot stripped and re-blotted for Polyclonal α-Flag HRP antibody to confirm Klf4 was pulled down by
beads.

40
Furthermore, immunoprecipitation of the KLf4 ΔC and Klf4 ΔM constructs revealed that the
C-terminus of Klf4 is required for binding to both transcription factors Oct-4 and Sox-2 (Figure 21,
22).

Figure 21: Binding Between Klf4 (ΔC, ΔM, and WT) & Sox-2
Overexpression of Klf4 (3 µg) constructs with Sox-2 (3 µg). A) Lysate controls were immuno-blotted
with Mouse-Monoclonal α-Flag antibody to show the expression of various flag tagged Klf4 deletion
mutant constructs. Secondary antibody used was Goat α- Mouse. B) Lysate controls were immuno-
blotted with Rabbit-Polyclonal α-Sox-2 antibody to show the expression of Sox-2. Goat α-Rabbit
was used for secondary antibody.
C) Immunoprecipitation with Mouse derived α-flag M2 Agarose beads to pull down Klf4 constructs,
and immuno-blotted with Rabbit-Polyclonal α-Sox-2 to detect bound Sox-2. Goat α-Rabbit was used
for secondary antibody was used. D) Blot stripped and re-blotted for Polyclonal α-Flag HRP
antibody to confirm Klf4 was pulled down by beads.
41

Figure 22: Binding Between Klf4 (ΔC, ΔM, and WT) & Oct-4
Overexpression of Klf4 (3 µg) constructs with Oct-4 (3 µg). A) Lysate controls were immuno-
blotted with Mouse-Monoclonal α-Flag antibody to show the expression of various flag tagged Klf4
deletion mutant constructs. Secondary antibody used was Goat α- Mouse. B) Lysate controls were
immuno-blotted with Rabbit-Polyclonal α-Oct-4 antibody to show the expression of Oct-4. Goat α-
Rabbit was used for secondary antibody. C) Immunoprecipitation with Mouse derived α-flag M2
Agarose beads to pull down Klf4 constructs, and Immuno-blotted with Rabbit-Polyclonal α-Oct-4 to
detect bound Oct-4. Goat α-Rabbit was used for secondary antibody was used. D) Blot stripped and
re-blotted for Polyclonal α-Flag HRP antibody to confirm Klf4 was pulled down by beads.
42
The C-terminus of Klf4 has three zinc fingers with the first zinc finger domain located in the amino
acid span of 391-415, the second 421-445, and the third 451-473 (Figure 1). Further mapping of Oct-
4 and Sox-2 interactions with Klf4 were accomplished by overexpression studies of the last two zinc
finger domains of the C-terminus Klf4-ΔZF-429 and Klf4-ΔZF-454. Klf4-ΔZF-429 was required for
binding to Sox-2 (Figure 23).

Figure 23: Binding of Sox-2 to Klf4-ΔZF (429,454)
Overexpression of Klf4 (2 µg) constructs with Sox-2 (2 µg). A) Lysate controls were immuno-blotted
with Rabbit α-Polyclonal Sox-2 antibody, to show the expression of Sox-2. Goat α-Rabbit was used
for secondary antibody. B) Lysate controls were immuno-blotted with Mouse Monoclonal α-Flag
antibody to show the expression of various flag tagged Klf4 zinc finger deletion mutant constructs.
Secondary antibody used was Goat α- Mouse. C) Immunoprecipitation with Mouse derived α-flag
M2 Agarose beads to pull down Klf4 constructs, and immuno-blotted with Rabbit Polyclonal α-Sox-2
to detect bound Sox-2. Goat α-Rabbit was used for secondary antibody was used. B) Blot was
stripped and re-blotted with Mouse-Monoclonal α-Flag antibody to confirm Klf4 was pulled down by
beads. Goat α-Mouse was used for secondary antibody was used.
43
Conversely, Oct-4 was still able to bind to Klf4-ΔZF-429 and Klf4-ΔZF-454 (Figure 24).
Through a process of elimination we concluded that Oct-4 must bind to the first zinc finger of the C-
terminus located from amino acids 391 to 415.

Figure 24: Binding of Oct-4 to Klf4-ΔZF (429,454)
Overexpression of Klf4 (2 µg) constructs with Oct-4 (2 µg). A) Lysate controls were immuno-blotted
with Rabbit Polyclonal α-Oct-4 antibody to show the expression of Oct-4. Goat α-Rabbit was used
for secondary antibody. B) Lysate controls were immuno-blotted with Mouse Monoclonal α-Flag
antibody to show the expression of various flag tagged Klf4 zinc finger deletion mutant constructs.
Secondary antibody used was Goat α-Mouse. C) Immunoprecipitation with Mouse derived α-flag
M2 Agarose beads to pull down Klf4 constructs, and Immuno-blotted with Rabbit Polyclonal α-Oct-4
to detect bound Oct-4. Goat α-Rabbit was used for secondary antibody was used. D) Blot stripped
and re-blotted with Mouse Monoclonal α-Flag antibody to confirm Klf4 was pulled down by beads.
Goat α-Mouse was used for secondary antibody was used.

44
Overexpression studies were also used to investigate the possibility that, transcription factors
that bind to Klf4 compete with each other. Results indicated that Klf4’s ability to bind β-catenin is
not inhibited by incremental increases of Oct-4 expression (Figure 25).

Figure 25: Competition Between β-catenin and Oct-4 for klf4 Binding. Various types of
plasmid DNA were overexpressed in 293T cells, Flag tagged Klf4 (2 µg), HA tagged β-catenin (2ug),
with incremental increases of 1µg, 2µg, and 4 µg of Oct-4. A) Lysate control immuno-blotted with
Rabbit-Polyclonal α-HA antibody to show the expression of β-catenin. Goat α-Rabbit was used for
secondary antibody. B) Lysate control immuno-blotted with Rabbit-Polyclonal α-Oct-4 antibody to
show the incremental increase in expression of Oct-4. Goat α-Rabbit was used for secondary
antibody. C) Lysate controls were immuno-blotted with Rabbit-Polyclonal α-Klf4 antibody to show
the expression of Klf4. Goat α-Rabbit was used for secondary antibody. D) immunoprecipitation
with Mouse derived α-flag M2 Agarose beads to pull down Klf4 constructs, and immuno-blotted with
Rabbit-Polyclonal α-HA to detect amount of β-catenin bound to Klf4. Goat α-Rabbit was used for
secondary antibody. E) Blot was stripped and immuno blotted with Rabbit-Polyclonal α-KLf4 to
confirm that flag beads pulled down Klf4. Goat α-Rabbit was used for secondary antibody. F) Blot
was stripped and immuno-blotted with Rabbit-Polyclonal α-Oct-4 to detect amount of Oct-4 bound to
Klf4. Goat α-Rabbit was used for secondary antibody.
45
Similarly Klf4’s ability to bind Oct-4 is not inhibited by incremental increases of Sox-2
expression (Figure 26).

Figure 26: Lysate Controls: Competition between Sox-2 and Oct-4 for Klf4 Binding
Various types of plasmid DNA were overexpressed in 293T cells, Flag tagged Klf4 (2 µg), Oct-4
(2ug), with incremental increases of 2µg, and 4 µg of Sox-2. A) Lysate control immuno-blotted with
Rabbit-Polyclonal α-Oct-4 antibody to show the expression of Oct-4. Goat α-Rabbit was used for
secondary antibody. B) Lysate control immuno-blotted with Mouse-Monoclonal α-Flag antibody to
show the expression of Klf4. Goat α-Rabbit was used for secondary antibody. C) Lysate control
immuno-blotted with Rabbit-Polyclonal α-Sox-2 antibody, to show the incremental increase in
expression of Sox-2. Goat α-Rabbit was used for secondary antibody. D) Immunoprecipitation with
Mouse derived α-flag M2 Agarose beads to pull down Klf4 constructs, and immuno-blotted with
Rabbit-Polyclonal α-Oct-4 to detect amount of Oct-4 bound to Klf4. Goat α-Rabbit was used for
secondary antibody. E) Blot was stripped and immuno-blotted with Mouse Monoclonal α-Flag to
confirm that flag beads pulled down flag tagged Klf4. Goat α-Mouse was used for secondary
antibody. F) Blot was stripped and immuno-blotted with Rabbit-Polyclonal α-Sox-2 to detect amount
of Sox-2 bound to Klf4. Goat α-Rabbit was used for secondary antibody.

46
Endogenous expression of transcription factors Klf4, Oct-4 and Sox-2 were detected in CF1
mES whole cell lysates (Figure 27).

Figure 27: Western Blots of mES Whole Cell Lysates
Western blots show expression of various transcription factors in mES cells. Primary antibodies used
from Santa Cruz included, α-Oct ¾ (H-134) Rabbit Polyclonal, α-Sox-2 (H-65) Rabbit Polyclonal,
and α-Klf4 (H-180) Rabbit Polyclonal. Secondary antibody used for each blot was Goat α-Rabbit,
and 15µg of each type of protein lysate were used for each lane.
47
Immunoprecipitation studies in mES cells revealed that Oct-4 interacts with Sox-2 when cells were
lysed with Ripa buffer. However, earlier immunoprecipitation studies of mES cell endogenous
protein lysates obtained via KLB lysis did not show such an interaction between Oct-4 and Sox-2
(Figure 28).

Figure 28: Immunoprecipitation: Binding Between Oct-4 and Sox-2 in mES
Pierce Immobilized Protein A/G beads were experimentally conjugated to either control IGG Mouse
(left lane) or Mouse Monoclonal α- Sox-2 primary antibody (right lane) were used for
immunoprecipitation of Sox-2 out of 1,000µg of whole cell lysate. Samples were immuno-blotted
with Rabbit Polyclonal α-Oct-4. Blots were stripped and blot with Rabbit polyclonal α-Sox-2 to
confirm that beads pulled down Sox-2. Secondary antibody used for each blot was Goat α-Rabbit.
This experiment was performed with two different lysis buffers.

48
Nonetheless, it is plausible that they bind each other because they bind to Klf4 zinc fingers
that are in close juxtaposition. Overexpression studies of 293T cells with just Oct-4 and Sox-2 were
used for immunoprecipitation to see it these two proteins interact. There was no interaction detected
between Oct-4 and Sox-2, but these data are inconclusive as Klf4 is not endogenously expressed in
the 293T cell line (Figure 29).

Figure 29: Binding Between Sox-2 and Oct-4 in 293T Cells
Overexpression of Oct-4 (2 µg) constructs with Sox-2 (2 µg). A) Lysate controls were immuno-
blotted with Mouse Monoclonal α-Oct-4 antibody to show the expression of Oct-4. Secondary
antibody used was Goat α- Mouse B) Lysate controls were immuno-blotted with Rabbit-Polyclonal
α-Sox-2 antibody to show the expression of Sox-2. Goat α-Rabbit was used for secondary antibody.
Pierce Immobilized Protein A/G beads were experimentally conjugated to Mouse Monoclonal α-Oct-
4 primary antibody and used for immunoprecipitation of Oct-4 out of 80 µg of whole cell lysate. C)
Samples were immuno-blotted with Rabbit Polyclonal α-Oct-4, to confirm that beads pulled down
Oct-4. Secondary antibody used was Goat α-Rabbit. D) Blots were stripped and blot with Rabbit
Polyclonal α-Sox-2 to detect if Sox-2 was bound to Oct-4. Secondary antibody was Goat α-Rabbit.

49
Endogenous interactions of Klf4 with Oct-4 were complicated to decipher as small amounts
of Klf4 protein were still visible in the IGG Mouse control lane indicating that washing of beads was
not complete, none the less the beads were washed the same number of times and the Klf4 band
appears higher in the lane that Oct-4 was pulled down (Figure 30).

Figure 30: Immunoprecipitation: Binding of Klf4 and Oct-4 in mES
Pierce Immobilized Protein A/G beads were experimentally conjugated to either control IGG Mouse
(left lane) or Mouse Monoclonal α-Oct-4 primary antibody (right lane) were used for
immunoprecipitation of Oct-4 out of 1,000µg of whole cell lysate. A) Samples were immuno-blotted
with Rabbit polyclonal α-Klf4 to detect if Klf4 was bound to Oct-4. Secondary antibody was Goat α-
Rabbit. B) Blots were stripped and blotted with Rabbit Polyclonal α-Oct-4, to confirm that beads
pulled down Oct-4. Secondary antibody used was Goat α-Rabbit.

50
For endogenous interaction between Klf4 and Sox-2 the antibody heavy chain overlaps the
detection of Klf4 at 55 Kd and makes the ability to establish an interaction impossible (Figure 31).

Figure 31: Immunoprecipitation: Binding Between Klf4 and Sox-2 in mES
Pierce Immobilized Protein A/G beads were experimentally conjugated to either control IGG Mouse
(left lane) or Mouse Monoclonal α-Sox-2 primary antibody (right lane) were used for
immunoprecipitation of Sox-2 out of 1,500µg of whole cell lysate. A) Samples were immuno-blotted
with Rabbit Polyclonal α-Klf4 to detect if Klf4 was bound to Sox-2. Secondary antibody was Goat α-
Rabbit B) Blots were stripped and blot with Rabbit Polyclonal α-Sox-2, to confirm that beads pulled
down Sox-2. Secondary antibody used was Goat α-Rabbit.
51
Cloning of Klf4 mutants into FUIPW vector was successful, and aided the creation of Klf4
WT, Klf4-ΔC and Klf4-ΔM lentiviruses. Preliminary testing of viruses in 293T cells via infection
followed by western blot indicated that viruses were viable (Figure 2). To further test virus infection
efficiency and condition, FUIGW lentivirus was also made and used to infect a dish of mES cells at
the same time infection with other Klf4 constructs occurred, and GFP expression was detected (Figure
3). These constructs were cloned into a FUIPW vector that allowed for Puromyocin selection which
facilitated the purification of the population of infected cells. FUIPW was used as a positive control,
and non-infected mES cells were used a negative control for Puromyocin selection.
Mouse ES cells were infected with lentiviral constructs FUIPW, FUIPW-Klf4-WT, FUIPW-
Klf4-ΔC, and FUIPW-Klf4-ΔM. Subsequent experiments were performed on only FUIPW, FUIPW-
KLf4-WT and FUIPW-Klf4-ΔC because, mES cells infected with FUIPW-Klf4-ΔM (that have
deletion of the nuclear localization sequence and transcriptional inhibitory domain) appeared to be
cell cycle arrested and could not be expanded enough for RNA analysis. Also presented are various
trials of PLKO and PLKO-Klf4-siRNA infections.
Positive alkaline phosphatase staining confirms ES cell identity. Infected mES cells were
plated in equal numbers and alkaline phosphatase staining was performed for two passages. FUIPW
control confirms that mES cells maintained their ES cell identity despite the harsh conditions of
infection and Puromyocin selection. Viral overexpression of FUIPW-Klf4-WT revealed that ES cells
maintain their ES cell identity. Alternatively, viral overexpression of FUIPW-Klf4-ΔC, which lacks
the ability to bind to stem cell factors Oct-4 and Sox-2, lost expression of alkaline phosphatase,
confirming that these cells were differentiated to another cell type (Figure 32). It appears that Klf4 is
required to maintain ES cell Identity and deletion of the functional domain that interacts with other
stem cell factors, and obliterates the ability to maintain ES cell characteristics. Morphology of
FUIPW-Klf4−ΔC is not shown but cells appeared to be differentiated into endo-mesoderm cell
lineages. Despite the inability of FUIPW-Klf4-ΔM infected MES to be expanded past the first few
passages, the cells analyzed were negative for alkaline phosphatase staining indicating that they were
52
differentiated. Alkaline phosphatase staining of PLKO-siRNA-Klf4 virally infected mES cells
revealed results that were consistent with that of FUIPW-Klf4-ΔC infection, namely, the FUIPW
control and FUIPW-Klf4-WT were positive for alkaline phosphatase staining while LKO-siRNA-
KLf4 were negative for staining (Figure 33).

Figure 32: Alkaline Phosphatase Staining Klf4 mES Infections
Alkaline Phosphatase Substrate 3 from Vector Laboratories was used giving a high sensitivity visible
blue staining for ES cell positive colonies. For each passage 1.5 x 10^6 cells were plated in 6 well
tissue culture dishes the day before for Alkaline Phosphatase analysis. A) Passage one, analysis of
FUIPW, FUIPW-Klf4-WT, FUIPW-Klf4-ΔC, and FUIPW-Klf4-ΔM. B) Passage two, analysis of
FUIPW, FUIPW-Klf4-WT, and FUIPW-Klf4-ΔC.
53

Figure 33: Alkaline Phosphatase Staining Klf4-siRNA
Alkaline Phosphatase detection kit from Sigma / Chemicon International was used giving a visible red
staining to positive colonies. Figure shows pictures taken when samples FUIPW, FUIPW-Klf4-WT
and PLKO-siRNA-Klf4 infections of MES Cells were dried out. Additionally, a picture of PLKO-
siRNA-Klf4 infection in PBS, obtained directly after staining is provided.
54
SSEA1 expression is a specific marker for mES cells. FACS Sorting confirmed preliminary
results of alkaline phosphatase staining, and provided us with a mean percentage of mES cells
positive for SSEA1 expression. Healthy non-virus treated mES that were culturally maintained
throughout the experiment were used as a positive control to calibrate the machine and 97% of the
population was positive for SSEA1 staining. Conversely, when healthy MES cells were incubated in
primary or secondary antibody alone then these served as negative controls for SSEA1 mediated
FACS Sorting. Results show that there is a negligible population shift in SSEA1 staining for FUIPW
and FUIPW-Klf4-WT virally infected cells for passage one (P1) and passage two (P2), as FUIPW
expression population dropped only 3.6%, and FUIPW-Klf4-WT increased 1.6% between passages.
Conversely, there was a significant reduction in mean population expression of SSEA1 for mES cells
infected with FUIPW-Klf4-ΔC for both passages with a value of 59.5% for P1, and 38.3% for P2,
percent dropped was 21.2 %. Reduction in SSEA1 expression in cells infected with FUIPW-Klf4-ΔM
was not as profound as FUIPW-Klf4-ΔC. Population of SSEA1 positive expression for FUIPW-Klf4-
ΔM infected cells was 66.8% for P1, and 70.5 % for P2, with a between passage increase of 3.7 %
(Figure 34).
To investigate the possibility that antibody titer used for FACS sorting was too high or that
FACS machine was overly sensitive, SSEA1 antibody immuno-staining was performed. PLKO
virally infected cells were used as a positive control for PLKO-Klf4-siRNA. Pictures corroborated
the previous evidence obtained from alkaline phosphatase staining. FUIPW, FUIPW-Klf4-WT, and
PLKO were all positive for SSEA1 staining. Whereas, FUIPW-Klf4-ΔC, FUIPW-Klf4 ΔM, and
PLKO-Klf4-siRNA were all negative for staining (Figure 35, 36).
55

Figure 34: Virus Infected mES Cells SSEA1 FACS Analysis
Mouse Derived SSEA-1 (mc-480) from University of Iowa Developmental Studies Hybridoma Bank.,
secondary used was Mouse α-Cy2. Negative controls included healthy MES cultured and passaged at
same time as viral infected cells, that were incubated with primary antibody alone or secondary
antibody alone. The positive control was these healthy MES incubated in both antibodies together,
and was used for calibration of the FACS sorting BD ISA 2 special order system machine parameters,
and to set up analysis gate prior to running other samples. Analysis software used was BD Facs diva
and samples were analyzed with a FITC filter. Values were use to create a graph that shows values of
mean population expression of SSEA1 for virally infected MES cells. Y-axis shows values of mean
population of MES cells positive for SSEA1 staining. X-axis shows passage number and condition
used for viral infection of MES cells
56

Figure 35: SSEA1 Staining Controls
Immunostaining of mES cells was carried out with only SSEA1 antibody from University of Iowa
Hybridoma bank, and 0.1 x 10^6 cells were plated on 8 well slides the day before analysis. For
Staining antibody was adjusted to a concentration of 1µg / ml. Secondary antibody used was Mouse
α-Cy3, and nucleus was detected with Hoechst Nuclear Staining. Inverted Florescent microscope was
used at a magnification of 20x. Filters used show SSEA1 staining (red), and nuclear staining (blue),
Light field, and overlay of all three. This figure shows controls for staining process, primary and
secondary antibody alone, as well as, vector controls for viral infection FUIPW and PLKO.
57

Figure 36: SSEA1 Staining of mES Infected FUIPW-Klf4 Mutants and PLKO-Klf4
siRNA. Immunostaining of mES cells was carried out with only SSEA1 antibody from University
of Iowa Hybridoma bank, and 0.1 x 10^6 cells were plated on 8 well slides the day before analysis.
For Staining antibody was adjusted to a concentration of 1µg / ml. Secondary antibody used was
Mouse α-Cy3, and nucleus was detected with Hoechst Nuclear Staining. Inverted Florescent
microscope was used at a magnification of 20x. Filters used show SSEA1 staining (red), and nuclear
staining (blue), Light field, and overlay of all three. This figure shows staining for, FUIPW-Klf4-
WT, FUIPW-Klf4-ΔC, and FUIPW-Klf4-ΔM, and PLKO-siRNA-Klf4.
58
This confirms that alkaline phosphatase staining is a more sensitive means of assaying stem
cell characteristics, which would explain why it is extensively used in the stem cell field.
To investigate what cell lineages infected cells progressively differentiated into, analysis of
RNA was performed. Q-PCR was carried out on CDNA products of RNA extracted from these cells
for the duration of 3 passages. Triplicates of every condition were made for each primer pair and
Gapdh was the endogenous control used. RNA from FUIPW infected cells served as a positive
control that represented healthy mES cells. FUIPW for each day was used to calibrate the Log fold
changes observed in that day’s respective FUIPW-Klf4-WT and FUIPW-Klf4-ΔC infected MES cell
q-PCR readings. Differentiation markers tested for q-PCR were, endoderm marker Fgf8, Ectoderm
marker Foxa2, neural ectoderm marker Neuro-d, and mesoderm marker Brachyury (T). The first
passage results showed that all differentiation markers were higher in FUIPW-Klf4-ΔC infected mES
cells as compared to FUIPW-Klf4-WT infected cells. However, reliability of this data set is
questionable as the standard deviations of the control were very high (Figure 37).
Analysis of the second passage revealed, all differentiation markers were higher in FUIPW-
Klf4-ΔC infected mES except Neuro-d (Figure 38).
The final passage results revealed differentiation markers were all higher in FUIPW-Klf4-ΔC
infected cells as compared to FUIPW-Klf4-WT infected cells (Figure 39). Collectively, q-PCR
results of all three passages indicate that differentiation markers are consistently expressed at higher
levels for all three days in FUIPW-Klf4-ΔC infected mES cells, and differentiation into mesoderm
and ectoderm lineages are most prominently represented. Interestingly, Neuro-d seems to be down
regulated in all cases when compared to FUIPW control.
59

Figure 37: Q-PCR Results for First Passage
For passage 1, triplicates of every condition were made for each primer pair and Gapdh was the
endogenous control used. FUIPW for each day was used to calibrate the Log fold changes observed
in that days respective FUIPW-Klf4-WT and FUIPW-Klf4-ΔC infected mES cell q-PCR readings.
Differentiation markers tested for q-PCR were, endoderm marker Fgf8, Ectoderm marker Foxa2,
neural ectoderm marker Neuro-d, and mesoderm marker Brachyury (T). Y-axis shows expression as
Log 10 relative quantities. X-axis indicates virus used for infection of mES cells and the primer set
used. Standard deviations were obtained from analysis with SDS RQ manager software.
60

Figure 38: Q-PCR Results for Second Passage
For passage 2, triplicates of every condition were made for each primer pair and Gapdh was the
endogenous control used. FUIPW for each day was used to calibrate the Log fold changes observed
in that days respective FUIPW-Klf4-WT and FUIPW-Klf4-ΔC infected MES cell Q-PCR readings.
Differentiation markers tested for Q-PCR were, endoderm marker Fgf8, Ectoderm marker Foxa2,
neural ectoderm marker Neuro-d, and mesoderm marker Brachyury (T). Y-axis shows expression as
log 10 relative quantities. X-axis indicates virus used for infection of MES cells and the primer set
used. Standard deviations were obtained from analysis with SDS RQ manager software.
61

Figure 39: Q-PCR Results for Third Passage
For passage 3, triplicates of every condition were made for each primer pair and Gapdh was the
endogenous control used. FUIPW for each day was used to calibrate the Log fold changes observed
in that days respective FUIPW-Klf4-WT and FUIPW-Klf4-ΔC infected MES cell Q-PCR readings.
Differentiation markers tested for Q-PCR were, endoderm marker Fgf8, Ectoderm marker Foxa2,
neural ectoderm marker Neuro-d, and mesoderm marker Brachyury (T). Y-axis shows expression as
log 10 relative quantities. X-axis indicates virus used for infection of MES cells and the primer set
used. Standard deviations were obtained from analysis with SDS RQ manager software.
62
Because q-PCR is very sensitive and rather large standard deviations were observed, RT-
PCR was attempted. Retesting of passage 1 was imperative because it had the largest standard
deviations for the FUIPW control, so a comparison of FUIPW and FUIPW-Klf4-ΔC was conducted.
Results of RT-PCR were consistent with data obtained from q-PCR and differentiation factors Fgf8,
Brachyury, and Foxa2 were higher in cells infected with FUIPW-Klf4-ΔC. Another differentiation
marker whose primers were not suitable for q-PCR because of its different annealing temperature was
added to this analysis, Nkx 2.5. Nkx 2.5 is a cardiac muscle marker, and results show that it is highly
expressed in FUIPW-Klf4-ΔC infected mES cells (Figure 40).
A more in depth analysis was carried out for the last passage because we expected that these
cells had reached their optimal differentiation and stability. A comparison of FUIPW with FUIPW-
Klf4-WT, and a comparison of FUIPW and FUIPW-Klf4-ΔC were carried out. Collectively, these
allow us to make inferences about the relationship between FUIPW-Klf4-WT and FUIPW-Klf4-ΔC
infected cells. Results were consistent with q-PCR results and indicate that Fgf8, Brachyury, and
Foxa2 are all detected at higher levels in the cells infected with FUIPW-Klf4-ΔC then their FUIPW
and FUIPW-Klf4-WT counterparts. Large standard deviations were also observed in the q-PCR
analysis of Neuro-d. RT-PCR allowed us to analyze if Neuro-d was in fact down regulated in
FUIPW-Klf4-WT infected cells when compared to FUIPW control. RT-PCR results confirm that this
is indeed the case (Figure 40).
To map which domains of Oct-4 interact with Klf4 C-terminus, cloning experiments were
conducted. Prosite software analysis gave predictions of domains present in Oct-4 sequence, two
domains were identified, Pou and Homeobox -2. The Pou domain of Oct-4 is located between amino
acids 131 and 205, and the Homeo-domain of Oct-4 is located between amino acids 221 and 281
(Figure 41).
63

Figure 40: RT-PCR Results
Information from q-PCR was used to select the number of cycles used for RT-PCR to gather data
before plateau is reached at 50 cycles of PCR. The 25 cycle condition was used for RT-PCR of the
following primers Gapdh, and Neuro-d, and Nkx 2.5. Alternatively 35 cycles were used for, Fgf8,
Brachyury, Foxa2, and Nkx 2.5. Note: an additional differentiation was added to this analysis, Nkx
2.5, a cardiac muscle marker. Gapdh was used as an internal control. A) Passage one, compared
expression between FUIPW and FUIPW-Klf4-ΔC. Primer sets used were, Fgf8, Brachyury, Nkx 2.5,
and Foxa2. B) For passage three, a comparison of expression between FUIPW and FUIPW-Klf4-WT
was conducted. Primer sets used were, Fgf8, Brachyury, Foxa2, Nkx 2.5, and Neuro-d. C)
Additionally, passage three, compared expression between FUIPW and FUIPW-Klf4-ΔC. Primer sets
used were, Fgf8, Brachyury, and Foxa2.
64

Figure 41: Prosite Domain Prediction for Oct-4
CDNA sequence of Mouse Oct-4 was analyzed by Prosite domain prediction software. A) Shows the
cDNA sequence converted to its respective amino acid sequence. B) Diagram of the domain hits
found in the software. C) Location of domain within the amino acid sequence.
65
The Pou domain consists of four α-helices packed together to enclose a hydrophobic core
and has an unusual helix turn helix structure, and has a highly charged amino acid region (155-162)
that bind selectively to the DNA Octomer motif ATGCAAAT. The Homeobox domain is composed
of 60 amino acids, and has a very highly conserved a helix-turn-helix structure that is very important
in DNA binding, in yeast, proteins that have this motif act as master switches in differentiation and
gene expression (Prosite). Oct-4 was divided into four segments. Segments
#
1 (1-131) and segment
#
4 (283-378) conferred no hits for domain identification. Segment
#
2 (132-282) contained the Pou
domain, and segment
#
3 (225-282) contained the Homeobox domain (Figure 42).
Mapping was designed by creating seven clones of Oct-4 had progressive C-terminal and N-
terminal deletions, providing useful constructs for functional analysis in the future (Figure 6, 8).
Furthermore, constructs were designed with primers that introduced a MYC tag at the C-terminus and
were finally cloned into the FUIGW vector that gives positive GFP identification of transfected and
infected cells.
Prosite domain prediction software gave one hit for Sox-2 analysis, the HMG box, located
between amino acids 43 and 111 (Figure 43). These High Mobility Group chromosomal proteins
have low molecular weights and are commonly transcription factors. They are unique in that they
have unusual DNA-binding activity; they bind DNA without sequence specificity at distorted DNA
sites (Prosite). Sox-2 was divided into 3 segments, segment
#
1 (1-112) which contained that HMG
box, segment
#
2 (113-215) and segment
#
3 (216-320) which both had no hits during domain prediction
(Figure 44). This protein was cloned using progressive N-terminal and C-terminal mapping, as well
as individual domain cloning (Figure 5, 7). These Sox-2 constructs were designed with primers that
introduced a HA tag at the C-terminus, and were also cloned into the FUIGW vector.
66

Figure 42: Oct-4 Cloning Sequence Alignment with Full Length Oct-4
Primer combinations of Oct-4 were manipulated to show what resulting cDNA sequence would define
the boundaries of each fragment. These four fragments were translated to amino acid sequences and
aligned with amino acid sequence of the full length protein. User Sequence one (top) is the segment
boundaries selected for cloning. User Sequence two (bottom) is the portion of the full length amino
acid sequence of Oct-4 that aligns with User Sequence 1. User sequence one of each fragment was
entered into Prosite domain prediction soft ware to confirm that predicted domains were present in
proper segments.
67

Figure 43: Prosite Domain Prediction for Sox-2
cDNA sequence of Mouse Sox-2 was analyzed by Prosite domain prediction software. A) Shows the
cDNA sequence converted to its respective amino acid sequence. B) Diagram of the domain hits
found in the software. C) Location of domain within the amino acid sequence.
68

Figure 44: Sox-2 Cloning Sequence Alignment with Full Length Sox-2
Primer combinations of Sox-2 were manipulated to show what resulting cDNA sequence would
define the boundaries of each fragment. These four fragments were translated to amino acid
sequences and aligned with amino acid sequence of the full length protein. User Sequence one (top)
is the segment boundaries selected for cloning. User Sequence two (bottom) is the portion of the full
length amino acid sequence of Sox-2 that aligns with User Sequence 1. User sequence one of each
fragment was entered into Prosite domain prediction soft ware to confirm that predicted domains were
present in proper segments.
69
Previous studies offer compelling evidence that Klf4 acts in the nucleus to represses β-
catenin mediated transcription in a HT29 colon cancer cell line through a cross-talk mechanism which
ultimately leads to differentiation (40). In this cellular context Klf4 behaves as a tumor suppressor
and is necessary for these cells to accomplish terminal differentiation. Consequently, investigation of
the relationship between various transcription factors and β-catenin were performed. β-catenin has
Luciferase activity when its promoter is activated. Repression or enhancement of this Luciferase
activity upon co-transfection of 293T cells with Oct-4, Sox-2, and Klf4 was analyzed. In all assays
the positive control was β-catenin alone leading to standard activation of Luciferase activity and
negative control was stem cell transcription factors alone which should not have had a significant
activation of Luciferase activity. For Luciferase assays equal number of cells were used in each well,
and duplicate dishes were made for each transfection condition. Duplicates allowed for acquisition of
averages and standard deviations for each condition within an experiment. Each transfection
condition was calibrated with PCDNA to make total concentration of DNA the same for each reaction
within an experiment. For Luciferase activity analysis, controls included, cells not tranfected with
any DNA designated blank, cells and transfected with PCDNA, and cells transfected with Topflash
and Renilla to expose background Luciferase activity that would later be subtracted out of data when
analyzed. Klf4 and β-catenin co-transfections confirm previous results found in colon cancer, that of
Klf4’s potent inhibition of β-catenin promoter activity (40) (Figure 45). Similarly, when transcription
factors involved in stem cell maintenance, Oct-4 and Sox-2, were individually tested in conjunction
with β-catenin, results indicated that they were also potent inhibitors of β-catenin promoter activity
(Figure 46, 47). Comparison of the graphs of each individual transcription factors ability to inhibit β-
catenin promoter induced Luciferase activity, showed relative inhibition potency of Klf4 > Oct-4 >
Sox-2.
70

Figure 45: β-catenin with Increasing Amounts of Klf4 Luciferase Assay
Graph of β -catenin promoter induced Luciferase activity. 293T cells were transfected with β-catenin
(1µg), and incremental increases of Klf4-WT at concentrations of 0.5µg, 1µg and 2µg. Positive
control was transfection of β -catenin alone. Negative control was transfection of Klf4-WT alone.
Duplicates of every condition were performed to make a graph of the averages and provide standard
deviations. Y-axis shows Luciferase activity. X-axis shows condition used for transfection.
71

Figure 46: β-catenin with Increasing Amounts of Oct-4 Luciferase Assay
Graph of β -catenin promoter induced Luciferase activity. 293T cells were transfected with β-catenin
(1µg), and incremental increases of Oct-4 at concentrations of 0.5 µg, 1µg and 2µg. Positive control
was transfection of β -catenin alone. Negative control was transfection of Oct-4 alone. Duplicates of
every condition were performed to make a graph of the averages and provide standard deviations. Y-
axis shows Luciferase activity. X-axis shows condition used for transfection.
72

Figure 47: β- catenin with Increasing Amounts of Sox-2 Luciferase Assay
Graph of β -catenin promoter induced Luciferase activity. 293T cells were transfected with β-catenin
(1µg), and incremental increases of Sox-2 at concentrations of 0.5 µg, 1µg and 2µg. Positive control
was transfection of β -catenin alone. Negative control was transfection of Sox-2 alone. Duplicates of
every condition were performed to make a graph of the averages and provide standard deviations. Y-
axis shows Luciferase activity. X-axis shows condition used for transfection.
73
Another assay was done in which, Oct-4 and Klf-4 were co-transfected in 293T cells to
investigate if these two factors had a synergistic repressive effect on β-catenin Luciferase activity. At
0.5 µg of Oct-4 and Klf4 individually with β-catenin, Klf4 is a better repressor of β-catenin.
However, a synergistic repression that was greater than any one factor alone was observed when these
two factors were combined at 0.5 µg each (Figure 48). Analysis of all three factors with β-catenin
was performed to investigate if a synergistic repressive effect would be observed. To set up the
parameters for this experimental analysis, 0.02 µg, 0.1 µg, 0.5 µg, and 1 µg of each transcription
factor was transfected individually with β-catenin. The potency trend seen earlier, in which
repression of β-catenin activity was Klf4> Oct4> Sox-2, was still prominent. However when a 0.02
µg and 0.1 µg combination of all three transcription factors with β-catenin was investigated, the
repressive activity was less than with each individual factor alone with β-catenin at these
concentrations. Conversely, when a combination of 0.5 µg and 1 µg of all three transcription factors
with β-catenin was investigated, the repression activity of β-catenin was more than with each
individual factor alone with β-catenin at these concentrations (Figure 49). This indicates that there
may be a dose dependant correlation to repression.
To investigate if a combination of just Oct4 and Sox-2 have an ability to repress β-catenin
Luciferase activity to the same extent as all three factors together did, another experiment was carried
out. The Luciferase activity for β-catenin in this study was exceptionally high making it impossible to
investigate the synergistic repressive activity of the transcription factors (Figure 50). To ameliorate
this problem, removal of the β-catenin plot point from the graph allowed us to investigate what was
happening on a smaller scale (Figure 51). At 0.1 µg of transcription factors with β-catenin the same
trend seen earlier of klf4> Oct-4> Sox-2 occurred again; however, at 0.5µg of transcription factors
Oct-4 took a small lead in repressive activity over Klf4. This is not very troublesome as this
experiment had been performed many times, and generally Klf4 has always been greater repressor of
β-catenin Luciferase activity as compared to Oct-4. Furthermore, the difference in repression shown
in this graph is rather small. Interpretation of 0.1 µg transfections revealed that, 0.1 µg (Klf4 + Oct-4)
74
> 0.1 µg (Klf4 + Sox-2) > 0.1 µg (Oct-4 + Sox-2). Furthermore, 0.1 µg (Oct-4+ Sox-2) had lower
repressive activity than 0.1 µg (Klf4+ Oct-4 + Sox-2).

Figure 48: β-catenin with Increasing Amounts of Oct-4 and Klf4 Luciferase Assay
Graph of β -catenin promoter induced Luciferase activity. 293T cells were transfected with β-catenin
(1µg), and incremental increases of Oct-4 and Klf4 at concentrations of 0.25µg and 0.5 µg. Positive
control was transfection of β -catenin alone. Negative control was transfection of Oct-4 and Klf4
alone. Duplicates of every condition were performed to make a graph of the averages and provide
standard deviations. Y-axis shows Luciferase activity. X-axis shows condition used for transfection.
75

Figure 49: β-catenin with Increasing Amounts of Oct-4 and Klf4 and Sox-2
Luciferase Assay
Graph of β -catenin promoter induced Luciferase activity. 293T cells were transfected with β-catenin
(1µg), and incremental increases of Oct-4, Klf4, and Sox-2 at concentrations of 0.2µg, 0.1µg, 0.5 µg,
and 1µg. Positive control was transfection of β -catenin alone. Negative control was transfection of
Oct-4, Klf4, and Sox-2 alone at each concentration tested. Duplicates of every condition were
performed to make a graph of the averages and provide standard deviations. Y-axis shows Luciferase
activity. X-axis shows condition used for transfection.
76

Figure 50: Luciferase Assay of β-catenin with Increasing Amounts of
(Klf4, Oct-4, and Sox-2). Graph of β -catenin promoter activity induced Luciferase activity.
293T cells were transfected with β-catenin (1µg), and incremental increases of Oct-4, Klf4, and Sox-2
at concentrations of 0.5 µg, and 1µg. Positive control was transfection of β -catenin alone. Negative
control was transfection of Oct-4, Klf4, and Sox-2 alone at each concentration tested. Every possible
combination of two factors was tested for each concentration. Additionally, two combinations of all
three factors were tested, at concentrations of 0.5 µg, and 1µg. Duplicates of every condition were
performed to make a graph of the averages and provide standard deviations. Y-axis shows Luciferase
activity. X-axis shows condition used for transfection
77

Figure 51: Luciferase Assay of β-catenin with Increasing Amounts of
(Klf4, Oct-4, and Sox-2). Same Figure

50, but β-catenin expression bar is removed from the
graph, to allow analysis of other bars on Graph of β -catenin promoter induced Luciferase activity.
293T cells were transfected with β-catenin (1µg), and incremental increases of Oct-4, Klf4, and Sox-2
at concentrations of 0.5 µg, and 1µg. Positive control was transfection of β -catenin alone. Negative
control was transfection of Oct-4, Klf4, and Sox-2 alone at each concentration tested. Every possible
combination of two factors was tested for each concentration. Additionally, two combinations of all
three factors were tested, at concentrations of 0.5 µg and 1µg. Duplicates of every condition were
performed to make a graph of the averages and provide standard deviations. Y-axis shows Luciferase
activity. X-axis shows condition used for transfection
78
These results indicate that Klf4 repressive activity is greater when it is paired with Oct-4 as
compared to when it is paired with Sox-2, and this trend holds true for analysis at 0.5 µg also.
Surprisingly, we see a shift of repression ability of the Oct-4 and Sox-2 combination in the 0.5 µg
analysis. The Oct-4/ Sox-2 combination was the least repressive at 0.1 µg but became a potent
inhibitor at 0.5 µg, so much so that, this combination is even more repressive then the 0.5µg
combination of all three transcription factors. This is an important discovery because a combination
of Oct-4 and Sox-2 alone is sufficient to inhibit β-catenin promoter induced Luciferase activity.
79
Chapter 4. Discussion
Suspicions that Klf4 was part of the ES cell maintenance circuitry arose initially via analysis
of Bio-statistical queries acquired by using RSAT to investigate the enrichment of Klf4 DNA binding
elements in upstream regions of well known ES cell maintenance and pluripotency genes. Results
indicated that Klf4 was significantly more enriched in this gene set as compared to several sets of
random genes. It is well established that, many ES cell specific genes are transcriptionally regulated
by cooperative binding of Oct-4 and Sox-2 to their conjoined Oct-Sox element on target genes (17).
Consequently, we assayed for the combined Oct4/Sox2 DNA-binding motif and found that it was also
significantly enriched in these same ES cell maintenance and pluripotency genes. These data
indicated that Klf4 may indeed play a more predominant role than ever suspected in ES cell
maintenance.
Further statistical analysis of upstream regions will be useful in determining if Klf4 DNA
binding hits are in close proximity to the Oct-Sox hits. We were also planning to perform chip-
sequencing analysis to answer the question, does Klf4-WT, Oct-4, and Sox-2 all bind to the same
promoter? Alternativly, if we suspect that Klf4 acts at a great distance from Oct-4 and Sox-2 we
could utilize a new technique coined, 3 C’s. This technique allows a researcher to show how DNA
elements, such as enhancers can communicate over long distances with their target genes. This
method takes advantage of the physical occurrence of DNA elements that loop out of the chromatin to
interact with their respective genes (32). It generally works by over expressing your proteins in a cell
line such that they will form a complex on the DNA, applying DNAse to digest the unbound DNA,
and immunoprecipitation of protein bound DNA. This DNA is further purified and DNA fragments
are ligated, from this you make cDNA and perform PCR. You will get a mixture of products from
this because the ligation step is random but the most represented product is the one that is further
analyzed (32). For example, if we hypothesized that the ES protein network binds the promoter of a
gene on one chromosome and then members of the protein complex bind somewhere entirely
80
different, perhaps even on another chromosome, then when we did 3C’s method we would get a
mixture of results but the majority would show a representation of the two chromosomes.
Preliminary expression experiments were performed on hES cells to get an idea of
expression levels and localization of various stem cell transcription factors. Western blots were used
to assay endogenous protein expression levels in hES cells and differentiated EB cells (Figure 18). As
expected hES cells had a higher expression of Sox-2 than EB cells. Oct-4 showed minimal change in
a full length version of the protein, which may be explained in part if we take into consideration that
profiles of gene expression of mice show that at the early stages of differentiation cells do not show a
down regulation of Oct-4 until much later than other ES transcription factors (29). Oct-4 western blot
also showed an increase in the smaller version of Oct-4 in EB. This may be explained by the fact that
there are two structurally similar isoforms of Oct-4, Oct-4A and Oct-4B. Oct-4B is smaller than Oct-
4A, and although they have identical DNA binding and C-terminal domains, Oct-4B lacks the
functional transactivation domain, and is mostly localized in the cytoplasm. Furthermore, “Oct-4B
cannot induce or antagonize the transcription of genes under control of Oct-4A, and it is not sufficient
to maintain stem cell self-renewal and permits them to display differentiated ES cell phenotypes,”
(16). The alternate version of Klf4 that naturally occurs in EB cells is smaller than the one found in
hES. It would be interesting to isolate and sequence this smaller version of the protein to see if
certain functional domains are missing.
Immunostaining studies showed co-expression of Oct-4 with Sox-2 and Klf4 with Oct-4 in
hES cells (Figure 13, 14, 15). Co-localization was shown between Klf4 and Oct-4 indicating that
these two transcription factors functions may be interrelated and possibly dependant on each other
(Figure 15). However, lack of monoclonal Klf4 antibody availability made direct analysis of Klf4 and
Sox-2 impossible. Our only alternative was to use deductive thinking to obtain conclusions from the
observations that Klf4 was co-expressed with Oct-4, Oct-4 was co-expressed with Sox-2, and so Klf4
must be co-expressed with Sox-2. Alternativly, immunostaining studies of EB were devoid of Klf4
81
and Oct-4 staining, which further established the role of these transcription factors as key players in
pluripotent stem cells as opposed to differentiated cells (Figure 16, 17).
While genome based regulatory and occupancy studies imply that there are various cross-
regulatory and auto-regulatory circuits, they fail to reveal how various transcription factors act in a
protein network. To this end, Klf4 deletion mutants were informative in not only establishing its
binding partners but in solidifying its role in self renewal and maintenance of mES cells. Chunming
Liu found that Klf4 binds to the C-terminus of β-catenin, and thereby inhibits β-catenin signaling.
The C-terminus of β-catenin contains the transcativation domain (40). A figure from this paper shows
that even the construct lacking zinc finger domain (Klf4-Δc), was sufficient for binding to β-catenin.
Although Klf4-ΔC construct could physically bind β-catenin, it was not functionally able to inhibit the
β-catenin transactivation domain as well as the Klf4-WT version could in Luciferase assays (40).
This finding was used to test our immunoprecipitation conditions, to assure that our method and
products could faithfully produce correct results (Figure 19, 20). Immunoprecipitation test run of
Klf4 deletion mutants and β-catenin revealed that, β-catenin can still physically bind to Klf4 when,
the middle domain was deleted, the C-terminus was deleted, and individual zinc fingers 429 and 454
were deleted (Figure 20). This is consistent with Chunming Liu’s figure, and the study goes further to
show that the middle region is not required for binding either. This leads us to speculate that it is
likely to be the N-terminal region of Klf4 that binds to the C-terminal region of β-catenin.
Additional immunoprecipitation studies prove that Klf4 C-terminus is required for its
interaction with transcription factors Oct-4 and Sox-2 via binding to specific zinc fingers in this
domain (Figure 21, 22, 23, 24). This data proves that Oct-4 and Sox-2 form a complex on the C-
terminal domain of Klf4 (Figure 52). I further hypothesized that this complex of transcription factors
is involved in maintaining stem cell phenotype.

82

Figure 52: : Klf4 C-terminal Binding Interactions
Diagrams the location of core Stem cell Transcription factors binding to individual zinc fingers
present in the C-terminus of Klf4. Triangles represent the Zinc fingers.

Another publication proves that β-catenin physically binds with Oct-4 (34). This reveals that
Klf4 is involved in binding both β-catenin and Oct-4, and that β-catenin and Oct-4 also bind to each
other. I used this information and went further to investigate if overexpression of Oct-4 adversely
affected β-catenin binding. Results reveal binding of Klf4 to β-catenin was not diminished upon
incremental overexpression of Oct-4 indicating that these factors do not compete for Klf4 binding
(Figure 25). Additionally, Oct-4 binding to Klf4 is not diminished upon incremental over expression
of Sox-2 indicating that C-terminal mediated binding of these factors is non-competitive (Figure 26).
Collectively, transcription factors that potentially bind different domains (β-catenin and Oct-4) or
same domains (Oct-4 and Sox-2) of Klf4 do not compete with each other for binding.
83
The Klf4 zinc fingers involved in the binding of Oct-4 and Sox-2 are in close enough vicinity
to support the hypothesis that Oct-4 and Sox-2 interact. Endogenous binding of Oct-4 and Sox-2 was
detected in mES cell lysates that were lysed with Ripa buffer, a result that has been verified several
times (Figure 28). Alternatively, earlier immunoprecipitation studies of mES cells endogenous
protein lysates obtained via KLB lysis did not clearly show such an interaction between Oct-4 and
Sox-2 (Figure 28). Ripa buffer is a more potent lysis buffer that completely lyses the nucleus. We
can speculate that the interaction of Oct-4 and Sox-2 preferentially occurs in the nucleus as most
transcriptional factor interaction occurs in this location, but sub-cellular fractionation and subsequent
immunoprecipitation would be needed to support such a claim. Overexpression studies of 293T cells
with transfection of only Oct-4 and Sox-2 were used for immunoprecipitation to see it these two
proteins interact. There was no interaction detected between Oct-4 and Sox-2, but this data is
inconclusive, as Klf4 is not endogenously expressed in the 293T cell line (Figure 29). Perhaps Klf4’s
C-terminus is a loading dock for Oct-4 and Sox-2 and only in its presence are the proteins able to
interact (Figure 28, 52).
Endogenous interactions of Klf4 with, Oct-4 and Sox-2, were difficult to obtain because the
antibody heavy chain for polyclonal Klf4 showed up rather close to Klf4 protein band location. To
ameliorate this problem we will invest in a monoclonal Klf4 antibody. This approach will work
because this new antibody can be used for immunoprecipitation of endogenous Klf4 and subsequent
immuno-blotting of Oct-4 and Sox-2, whose protein expression bands are located far from heavy and
light chains observed on western blots (Figure 30, 31).
Collectively, with the results presented in this thesis and previous finding in the field, we are
for the first time able to establish a model for the ES cell maintenance protein network, in which β-
catenin binds both Oct-4 and Klf4, while Klf4’s C-terminus is also involved in binding of Oct-4 and
Sox-2, and Oct-4 and Sox-2 bind with each other (Figure 53, blue dotted lines).
84

Figure 53: Modified Diagram of Stem Cell Maintenance Circuitry
This diagram collectively represents findings of this thesis and current knowledge about the stem cell
maintenance network. Purple rectangle represents β-catenin gene, and the green rectangle represents
the Nanog gene. All other shapes represent proteins of the network and transcriptional regulation of
their respective genes. Blue dotted lines represent protein- protein interactions. Curved red arrows
represent auto-regulation of a protein in its own transcription. Red lines represent proteins ability to
inhibit gene transcription. Green lines represent the ability of proteins to induce gene transcription.
85
Cloning of various domains of Oct-4 and Sox-2 into FUIGW vector with experimentally
introduced MYC and HA tags, respectively was accomplished. Introducing tags will facilitate
immunoprecipitation and identification of externally provided overexpressed protein constructs in
future experiments. It should also be noted that cloned domains that showed no characteristic hits in
the domain prediction software should not be assumed to lack functional activity. Not much is known
about these transcription factors and these clones may be very useful for binding assays and functional
analysis of protein–protein interaction of the ES cell maintenance network. Current experiments seek
to map which domains of Oct-4 and Sox-2 bind to the Klf4 C-terminus. We may also further define
which domain of Oct-4 is involved in binding β-catenin. Additionally, we can define which domains
of Oct-4 and Sox-2 are involved in their binding to each other. Once we obtain Nanog plasmid DNA,
we can investigate how Nanog interacts with the protein Network we have established thus far. This
will prove to be an essential step because Nanog’s transcription network has been shown to regulate
pluripotency in mES cells, and it is speculated to be the master regulator by many scientists in the
stem cell field (18). Future experiments will aim at determining if Klf4, Oct-4, and Sox-2 bind to
Nanog. If such interactions are found we have the invaluable tools that can further define the
respective domains involved in these interactions.
Lentiviral introduction of Klf4 mutants into mES cells proved that expression of functional
Klf4 domains, are essential for maintaining mES cell self-renewal and maintenance. FUIPW was
used as a positive control and was used to calibrate experimental parameters and confirmed that mES
cells maintain their ES cell identity despite the harsh conditions imposed upon them during infection
and Puromyocin selection. FUIPW-Klf4-ΔM infected mES cells served as a good negative control
because deletion of the M-domain abolished the nuclear localization signal, inhibiting translocation of
Klf4 to where its functional activity is implemented. Unfortunately, culturing of these cells was
problematic because they seemed to be cell cycle arrested every time viral infection was attempted.
For each experiment only small amounts of cells were obtained from each passage and were only used
for, SSEA1 expression analysis, alkaline phosphatase staining, and FACS sorting. As expected
86
FUIPW-Klf4-ΔM infected mES cells showed negative staining for ES cell markers SSEA1 and
alkaline phosphatase, and had a slight population shift upon FACS sorting (Figures 32, 34, 35, 36).
Viral Over expression of FUIPW-Klf4-WT reveals that the ES cells maintain their ES cell identity
(Figures 32, 34, 35, 36). Alternatively, viral overexpression of FUIPW-Klf4-ΔC, which lacks ability
to bind to Stem cell factors Oct-4 and Sox-2; lost expression of Alkaline Phosphatase and SSEA1
staining, showed a population shift in SSEA1 FACS sorting, and Q-PCR in addition RT-PCR results
confirmed that these cells were differentiated to various cell lineages confirming loss of pluripotency
(Figures 32, 34, 35, 36). Similarly PLKO-KLf4-siRNA viral infection was performed several times
and had been shown in this thesis to lack Alkaline Phosphatase and SSEA1 staining (33, 36).
Collectively, these data confirm that Klf4 C-terminus is an essential functional domain that controls
ES cell maintenance once inside the nucleus of mES cells. We conclude this because the same
phenotype was seen for functional domain deletion, and knockdown of functional protein.
Our lab is currently repeating the PLKO-Klf4-siRNA experiment to obtain sufficient RNA
amount for use of Taqman ®Mouse Stem Cell Pluripotency Array analysis which will be useful for
more inclusive gene profiling of genes linked to, pluripotency, self renewal, and determination of
specific cell lineage differentiation. Quantitative expression of Oct-4 has been shown to produce
three different responses with very precise thresholds for each response. A perfect level of Oct-4 will
maintain pluripotency, less than a two fold increase in expression leads to differentiation to endoderm
and mesoderm lineages, whereas, repression of Oct-4 leads to differentiation to trophectoderm
lineages (14). Q-PCR and RT-PCR results of this paper focused on expression of differentiation
markers of three germ layers, Fgf8 (endoderm), Foxa2 (ectoderm), Neuro-d (neural ectoderm), and
Brachyury (mesoderm). Additionally, Nkx2.5 a cardiac muscle marker was attempted in RT-PCR.
Results indicate that in the FUIPW-Klf4-ΔC infected mES cells all the differentiation markers were
expressed at higher levels in cells than FUIPW-Klf4-WT infected cells. Interestingly, FUIPW-Klf4-
WT infected mES cells showed repressed expression of Neuro-d. Unfortunately, trophoectoderm
markers were not available for analysis. Nonetheless, q-PCR and RT-PCR results show an array of
87
differentiation potentials indicating that, perhaps Klf4 may not only regulate Oct-4 levels, but that
changes in Klf4 levels itself can lead to differentiation. In fact, our labs preliminary results from,
Taqman ®Mouse Stem Cell Pluripotency Array analysis of Klf4-siRNA not only mirrors the results I
found for FUIPW-Klf4-ΔC, but showed an increase in tropoectoderm markers as well. This further
gene profiling analysis done by others will also prove to be more reliable, as it analyzes various
markers and has several internal controls other than Gapdh, such as 18s ribosomal RNA and β-actin.
Future experiments should include a functional analysis of Oct-4 and Sox-2 binding to Klf4,
to confirm that Klf4 acts through Oct-4 and Sox-2 to induce self- renewal, reprogramming and
pluripotency. Means to accomplish this feat are established in this thesis. Recall, that the zinc fingers
of the Klf4 C-terminus bind to Oct-4 and Sox-2. Therefore, selective overexpression of the zinc
finger domain or zinc fingers individually, should serve to block the self renewal function imposed by
Oct-4 and Sox-2, thereby further establishing that Klf4 acts through these factors to maintain ES cell
identity. However, this may prove problematic if they cannot localize in the nucleus to bind Oct-4
and Sox-2. Alternatively, a peptide that is fused to poly arginine could be used, such that it can go to
the nucleus and block protein transduction.
Overexpression studies of FUIPW and FUIPW-Klf4-WT in mES cells should be tested in
two culturing conditions, BMP (serum alone), and LIF alone. Pervious studies show that LIF alone is
insufficient in maintaining mES cell self-renewal and cells differentiate preferentially into neural
phenotypes (20, 38). BMP is an anti-neurogenesis factor that has been shown to contribute to
differentiation; “it patterns differentiation of mesoderm and is able to generate cells with an immature
hepatocyte phenotype” (11, 37). Ying et al. show that a combination of LIF and BMP in culture work
together to maintain self-renewal (38). Furthermore, if you initially culture cells in BMP and LIF, and
then remove LIF there is a switch in BMP action that no longer supports self-renewal but instead
promotes differentiation (38). Two pathways for maintenance of ES cells exist here. LIF is a
cytokine that binds the gp130 receptor that serves to activate transcription factor STAT3, which goes
to the nucleus to promote ES cell self-renewal (38). Recently this LIF pathway has been further
88
defined to prove that, LIF up regulates β-catenin, which activates the STAT3 pathway, and leads to
the ultimate transcriptional regulation of stem cell maintenance factor C-myc (6). Another
downstream target of this pathway is ERK which is not required for self-renewal because it is a pro-
differentiation factor, in fact, if you reduce ERK expression then this would serve to promote self-
renewal (5). Ying et al. discovered that BMP did not serve to indirectly increase LIF’s activation of
STAT3, activate STAT3 itself, or down regulate ERK, which indicated that the LIF and BMP
pathways were not interrelated (38). BMP has a much more intricate pathway, and overexpression of
BMP leads to differentiation. BMP’s activate SMAD via phosphorylation, this initial activity primes
a BHLH neural differentiation factor called mash, if this were to continue uninterrupted then mash
activation itself would lead to differentiation. Alternativly, activation of SMAD has two other
options: one leads to activation of p38 MAP-kinase (a death factor), and the other leads to stimulation
of self renewal. How SMAD activates self-renewal is through the activation of the Inhibition of
Differentiation genes (ID), this activation creates negative BHLH factors, that serve to inhibit the
primed mash, thus steering the pathway away form differentiation and allowing ID genes to activate
self renewal (38).
One way of establishing the hierarchy in stem cell network is to show how many pathways a
factor can bypass. For instance, Nanog is considered the master regulator by scientists not only
because it is linked to more genes involved in proliferation and self-renewal than Oct and Sox both
alone and combined, but because it can bypass requirements for the initial steps of self maintenance
pathways (4). If we were to culture mES cells in the absence of LIF and BMP we would see
differentiation of these cells within a day, the signals that these extrinsic factors provide have been
proven to be absolutely essential for ES cell maintenance in culture. However, when Nanog is
overexpressed in mES cells then these cells can maintain self-renewal in the absence of LIF or BMP
signals. Nanog bypasses the requirement for STAT3 in serum containing media although full stem
cell renewal efficiency is optimized by a combination of overexpression and LIF. Furthermore,
Nanog has been shown to bypass the requirement of BMP as Nanog is able to directly or indirectly
89
induce the expression of ID (7). Conversely, Oct-4 overexpression is not sufficient in maintaining an
undifferentiated phenotype in the absence of LIF. This proves that Oct-4 cannot bypass the signal
provided by LIF stimulation (23).
Characterization of Klf4 over expression in these two culturing conditions will be an
essential experiment in the future. If we hypothesize that Klf4 is a downstream of STAT3 then we
should be able to maintain stem cells in the absence of LIF, conversely if Klf4 activates ID then we
should be able to maintain stem cells in the absence of BMP (serum). Further characterization of
these cells could be accomplished by a combination of, SSEA1 staining or FACS sorting, Alkaline
Phosphatase staining, morphology characterization, and STAT3 expression or assaying activation of
ID genes.
Additionally the canonical WNT signaling cascade has also been established as playing a
role in maintaining stem cells; it has been shown that LIF enhances nuclear levels of β-catenin in the
same manner as Wnt (34). Historically, this pathway has had a long established role in Cancer
progression. In the absence of Wnt signaling, β-catenin is phosphorylated by a complex that is
composed of GSK3β and APC (acts as a scaffold to AXIN 1 and 2). Phosphorylation of β-catenin
results in ubiquitination of this molecule which, subsequently, undergoes proteosome-mediated
degradation. Wnt signaling inhibits the phosphorylation of β-catenin, thus allowing it to accumulate
in the cytoplasm and translocate into the nucleus. Nuclear β-catenin binds to molecules of the Tcf
family and stimulates the expression of the Wnt responsive genes (28).
Publications show that APC specific mutations that allow β-catenin to accumulate in the
nucleus block the stem cell’s ability to differentiate (13). Additionally, expression of constitutively
active mutant form of β-catenin establishes long term proliferation of ES cells even in the absence of
LIF (34). The paper suggests that both the Wnt and LIF pathways contribute to the maintenance of
mES cells, because they both can up-regulate β-catenin nuclear expression. This lays out another
pathway option for β-catenin other than the STAT pathway. This alternate pathway stipulates that
nuclear β-catenin up-regulates Nanog, although not directly, it does so through physical interaction
90
with Oct-4 (34). These findings are quite interesting because Octamer and Sox binding elements exist
on the Nanog gene, and binding of Oct-4 and Sox-2 to these elements has been shown to regulate
Nanog transcription (14). As discussed earlier Nanog over expression negates the requirement of LIF
as it can provide activation of ID in a SMAD parallel pathway. Similarly constitutively active mutant
form of β-catenin lacks the need for LIF. The logic of this would clearly suggest that perhaps when
β-catenin is overexpressed, it binds to Oct-4 and thereby increases Nanog transcription which
subsequently leads to activation of ID’s that serve to maintain ES cells in the absence of any initial
LIF stimulus. If one accepts the model proposed in this thesis then this pathway integrates quite well.
The proposed protein network stipulates that β-catenin binds both Klf4 and Oct-4. Klf4 binds β-
catenin, Oct-4, and Sox-2. Oct-4 and Sox-2 bind each other and have the added responsibility of
transcriptionally regulating Nanog (Figure 54). Taking this into consideration Klf4 plays a central
and crucial role in the stem cell maintenance protein network.
When analyzing the protein network proposed by this thesis, one can start to ascribe
functions to individual proteins in the network. We have already discussed how the extrinsic factors
such as LIF, BMP, and Wnt all serve to support self renewal and pluripotency, either through the
activation of the downstream master regulator, Nanog, or through activation of other downstream
targets that regulate pluripotent genes. Our understanding of how the network functions is even more
convoluted when we take into account, how protein levels of each factor poses stimulatory or
inhibitory effects on transcriptional levels of other factors in the network. Understanding how levels
of each factor maintain ES cell identity will help us illuminate the transcriptional dynamics that are at
play. Overexpression of several proteins in this pathway, have been shown to lead to differentiation,
further exemplifying the essential need of cross-regulatory and auto-regulatory signals in this
network. Overexpression or under expression of Oct-4 as well as Sox-2, lead to differentiation of ES
cells (15, 17, 23). Publications confirm that Oct-4, Sox-2 and Klf4 each have individual feedback
loops to help these factors auto-regulate themselves (3, 19, 25, 36). It is well established that many
ES cell specific genes are transcriptionally regulated by cooperative binding of Oct-4 and Sox-2 to
91
their conjoined Oct-Sox element on target genes (17). For example, UTF1 a transcriptional co-
activator and FGF4 enhancer both, recruit the Oct-4 Sox-2 complex differentially by a one base pair
difference in their Octomer-Sox binding motifs (1, 22). The synergistic regulation imposed by the
Oct-4 Sox-2 complex, depends on a very specific spatial arrangement of these factor binding sites (1).
Furthermore, crystal structure of the POU/HMG/ DNA complex suggests that Oct-4 and Sox-2 bind in
different ways to DNA. Oct-4 binds in a DNA specific manner while Sox-2 binds in relatively non-
specific manner and is dependant on the preliminary binding of Oct-4 (26). Similarly, the synergistic
binding of Oct-4 and Sox-2 is needed to regulate the Oct-3/4 gene (24). More recent literature further
establishes that transcriptional regulation of both Oct-4 and Sox-2 genes is accomplished by
synergistic binding of this Oct-4/Sox-2 complex (17). Oct-4/ and Sox-2 also acts on the downstream
master regulator, Nanog. Octomer-Sox elements are required for the cis transcriptional regulation of
the Nanog gene (14).
Another example of dynamic control of transcription involves Nanog itself. Nanog is
involved in another independent feedback loop with β-catenin. As mentioned earlier, β-catenin up
regulates Nanog expression via its ability to physically bind to Oct-4 (34). Conversely,
overexpression of Nanog can repress β-catenin transcriptional activity (2). It seems unnecessary that
Nanog would have the ability to inhibit β-catenin because overexpression studies of β-catenin and
Nanog individually, indicate that higher levels of these proteins reinforce ES cell identity and
discourage differentiation. An interesting development revealed in this thesis was, that the
overexpression of the Klf4-WT construct failed to make MES cells differentiate either. It seems that
levels of Oct-4 and Sox-2 are tantamount in triggering the decision to differentiate. One might
speculate that Nanog’s imposed inhibitory action on β-catenin is present more so to maintain stringent
control of Oct-4 and Sox-2 levels to prevent differentiation than to, induce differentiation by shutting
off β-catenin. Unfortunately, this is just speculation as no study has established β-catenin’s ability to
transcriptionally regulate Oct-4 and Sox-2 yet.
92
This thesis addressed what would happen to β-catenin promoter induced Luciferase activity
when we overexpressed various transcription factors. In β-catenin Luciferase co-transfection assays
we saw that individually Klf4, Oct-4 and Sox-2 transcription factors could repress β-catenin
Luciferase activity and that a combination of Oct-4 and Sox-2 alone was also sufficient in repressing
β-catenin Luciferase activity to the same extent or more as Klf4 alone and in combination with Oct-4
and Sox-2 (Figures 45-51). This is an interesting development because expression of transcription
factors such as Oct-4, Sox-2 and β-catenin have been linked to maintaining ES identity, and now we
see that there is an additional level of control in which differentiation response can be carried out. We
can propose that when Oct-4 and Sox-2 are overexpressed they try to auto-regulate themselves to
prevent onset of differentiation, but upon some threshold level they activate Nanog, which can inhibit
β-catenin transcription. Additionally as found in this thesis Oct-4 and Sox-2 can directly inhibit β-
catenin activity as well. Collectively, these multiple levels of feed back loops that inhibit β-catenin
halt the stem cell maintenance circuitry, which may ultimately lead to differentiation (Figure 54).
Similarly, Klf4 expression acts independently as a very potent inhibitor of β-catenin mediated
promoter activity, suggesting overlapping roles of transcription factors in the network. It should also
be noted that these regulation loops for β-catenin discussed above, may also be imposed to prevent the
transformation of this well orchestrated stem cell maintenance network to an uncontrollable cancerous
circuit. In fact, β-catenin over expression via the Wnt pathway has been documented in colon cancer
cells (28, 40).
We are in the process of acquiring a Luciferase gene that can be driven by Nanog promoter
to test the relationship between Nanog promoter activity and overexpression of Klf4. As a positive
control we expect, a combination of Oct-4 and β-catenin, and a combination of Oct-4 and Sox-2 will
activate Nanog promoter. As a negative control, we can use β-catenin without Oct-4, which
according to previous literature should not activate Nanog promoter. If an activation of Nanog
promoter is found with expression of Klf4-WT alone, we could test if KLf4-ΔC is able to activate
Nanog as well. If Klf4-WT induces some basal level of Nanog promoter activity, we can test if this
93
activation is increased by combinatory transfection with (Oct-4 and Sox-2) or (β-catenin and Oct-4)
or (Oct-4, Sox-2, and β-catenin). Additionally, we can further establish that the C-terminus of Klf4 is
involved in ES-cell maintenance through indirect activation of Nanog (made possible by its physical
interaction with Oct-4 and Sox-2). We could assay this indirect activation of Nanog by comparing
Klf4-WT and Klf4-ΔC individually partnered with, (Oct-4 and Sox-2) or (β-catenin and Oct-4) or
(Oct-4, Sox-2, and β-catenin). Furthermore, the previously published finding concerning β-catenin
mediated activation of Nanog via its interaction with Oct-4 can be further defined if we utilize the
Oct-4 deletion mutants that we created for mapping in this thesis to determine the functional domain
of Oct-4. Once when we receive Nanog plasmid DNA, we could compare the inhibition potency of
Nanog and Klf4 on β-catenin, and investigate if they have a synergistic repression. Ultimately, these
proposed experiments can illuminate the dosage dependant interplay that occurs between factors in
the stem cell maintenance network.
It will be crucial to evaluate Klf4, Oct-4 and Sox-2 binding and transcriptional effect on
Nanog because, we are in the process of applying our ES cell network model to hES cells, to
determine how reprogramming of somatic cells is accomplished. Previous papers by Yamanaka
prove that reprogramming of mouse fibroblast cells can be accomplished by viral introduction of Oct-
4, Sox-2, C-myc, and Klf4, and that in this context Klf4 was dispensable. This thesis proves that Klf4
is not dispensable and is essential for self renewal of mES cells. Moreover, recent papers by
Yamanaka et al. attempted reprogramming of mouse and human fibroblast cells, to produce Induced
Pluripotent stem cells, with just 3 factors (Klf4, Sox-2, and Oct-4). Researchers found that this
combination was sufficient to produce high quality stem cells with less background than the previous
four factor combination (21). In this scenario C-myc is dispensable.
One may speculate that for reprogramming and stem cell maintenance that C-myc is
dispensable because Klf4 possibly has an over lapping role with it. If we take into consideration that
there are various pathways in stem cells that lead to the same terminal event of maintenance, and are
willing to recognize that there are differences in these pathways when we consider mouse and human
94
models, we can begin to speculate as to why C-myc is dispensable. Human ES cells, unlike mES cells
are grown in the absence of LIF. In mES cells it has been discovered that the key target of the LIF
induced self renewal pathway is C-myc. LIF’s activation of β-catenin, and subsequent activation
STAT3, leads to transcriptional regulation of C-myc. Furthermore, if there is removal of LIF then C-
myc protein is phosphorylated on theronine residues, and undergoes GSK3β mediated degradation
(6). We hypothesized earlier that Klf4 overexpression may negate the necessity of LIF, as Klf4 may
be a down stream target of STAT3, which could mean that Klf4 would have an alternate means of
activating C-myc in the absence of LIF induced β-catenin. If we consider the alternate proposed
pathway for β-catenin, in which β-catenin acts through Oct-4 to activate Nanog, then we can see how
important making a connection between Klf4 and Nanog is. If we could prove that Nanog induced
Luciferase activity is up regulated when we overexpress Klf4 alone, or in combination with Oct-4 and
Sox-2, this could explain how self-renewal is maintained, through Nanog’s activation of ID via the
SMAD parallel pathway. Both scenarios can bypass the need for C-myc either because it activates it
through another downstream target or, hES predominantly utilizes another pathway in which Klf4,
Oct-4, and Sox-2 are all that are needed to up regulate Nanog expression.
This thesis proves that Klf4 is involved in mES cell maintenance and is a central component
of the ES cell protein network. It further exemplifies how proteins in the network cooperate to
maintain ES cell Identity. However this interdependency is double edged because, it renders ES cells
susceptible to differentiation when inactivation or changes in expression levels of any of its number of
components occurs. There seems to be hope regarding the reprogramming of somatic cells because
some of the essential factors in the ES cell maintenance hierarchy do not cause differentiation upon
over expression for example, Klf4 and Nanog. The future may aim at finding small molecular drugs
that can induce Klf4 expression. One such drug that exists for induction of Nanog in mES is retinol,
the alcohol form of vitamin A. However research proves that the mechanism of this drug is unknown.
Unfortunately, the possibility that retinol can act in the LIF, Wnt/b-catenin, and Oct-Sox pathways has
been disproved (8). Nonetheless, this drug inhibits differentiation of mES. Although this thesis
95
aimed at defining the most primitive core of ES stem cell factors that are possibly upstream form
Nanog, one cannot ignore that other factors may exist that have overlapping roles with Klf4, Oct-4,
Sox-2, and Nanog. Furthermore, future investigations of temporal and quantitative expression of the
proteins that comprise the ES cell network, as well as, any protein modifications they undergo could
prove to be essential to our development of drugs that maintain ES cell identity or induce
reprogramming of somatic cells.
96
References
1) Ambrosetti, D.C., Basilico, C., and Dailey, L. (1997) Synergistic Activation of Fibroblast Growth
Factor 4 Enhancer by Sox-2 and Oct-3/4 Depends on Protein -Protein Interactions Facilitated
by a Specific Spatial Arrangement of Factor Binding Sites, Molecular and Cellular Biology,
VOL 17, NO 11, pp 6321-6329.

2) Anton, R., Kestler, H.A., Kuhl, M. (2007) β-catenin Signaling Contributes to Stemness and
Regulates Early Differentiation in Murine Embryonic Stem Cells, Federation of European
Biochemical Societies VOL 581, pp 5247-5254.

3) Boer, B., Kopp, J., Mallanna, S., Desler, M., Chakrevarty, H., Wilder, P.J., Bernadt, C., and
Rizzino, A. (2007) Elevating the Levels of Sox2 in Embryonal Carcinoma Cells Inhibits The
Expression of Sox2: Oct3/4 Target Genes, Nucleic Acide Research, VOL 35, NO 6, pp
1773-1789.

4) Boyer, L.A., Lee, T.I., Cole, M.F., Johnstone, S.E., Levine, S.S., Zucker, J.P., Guenther, M.G.,
Kumar, R.M., Murray, H.L., Jenner, R.G., Gifford, D.K., Melton, D.A., Jaenisch, R., and
Young, R.A. (2005) Core transcriptional Regulatory Circuitry in Human Embryonic Stem
Cells, Cell, VOL 122, pp 947-956.

5) Burdon, T., Stracey, C., Chambers, I., Nichols, J., and Smith, A.(1999) Suppression of SHP-2
and ERK Signaling Promotes Self-Renewal of Mouse Embryonic Stem
Cells,.Developmental Biology, VOL 210, pp 30-43.

6) Cartwright, P., McLean, C., Sheppard, A., Rivett, D., Jones, K., and Dalton, S. (2004) LIF/STAT
Controls ES cell Self-Renewal and Pluripotency by a Myc Dependent Mechanism,
Development, VOL 132, 885-896.

7) Chambers, I. (2004) The Molecular Basis of Pluripotency in Mouse Embryonic Stem Cells,
Cloning and Stem Cells, VOL 6, NO 4, pp 386-391.

8) Chen, L., Yang, M., Dawes, J., and Khillian, J.S. (2007) Suppression of ES Cell Differentiation by
Retinol ( Vitamin A) via the Overexpression of Nanog, Differentiation, VOL 75, pp 682-
693.

9) Chen, S., Do, J.T., Zhang, Q., Yao, S., Yan, F., Peters, E.C., Scholer, H.R., Schultz, P.G., and
Ding, S.(2006) Self-Renewal of Embryonic Stem Cells By a Small Molecule, PNAS, VOL
103, NO 46, pp 17266-17271.

10) Chen, X., Johns, D.C., Geiman, D.E., Marban, E., Dang, D.T., Hamlin, G., Sun, R., and Yang,
V.W. ( 2001) Kruppel-like Factor 4 ( Gut Enriched Kruppel-like factor) Inhibits Cell
Proliferation by Blocking G
1
/S Progression of the Cell Cycle, The Journal of Biological
Chemistry, VOL 276, NO 32, pp 30423-30428.

11) Gouon-Evans, V., Boussemart, L., Gadue, P., Nierhoff, D., Klehler, C.I., Kubo, A., Shafritz, D.A.
(2006) BMP-4 is Required for Hepatic Specification of Mouse Embryonic Stem Cell-
Derived Definitive Endoderm, Nature Biotechnology, VOL 24, NO 11, pp 1402-1411.

12) Kaji, K., Caballero, I.M., MacLeod, R., Nichols, J., Wilson, VA., and Hendrich, B. (2006) The
NuRD Component Mbd3 is Required For Pluripotency of Embryonic Stem Cells, Nature
Cell Biology VOL 8, NO 3, pp 285-292.
97
13) Kielman, M.F., Rindapaa, M., Gaspar, C., Poppel, N.V., Breukel, C., Leeuwen, S.V., Taketo,
M.M., Roberts, S., Smits, R., and Fodde, R. (2002) APC Modulates Embryonic Stem-Cell
Differentiation by Controlling the Dosage of β-catenin Signaling, Nature Genetics, VOL 32,
pp 594-605.

14) Kuroda, T., Tada, M., Kubota, H., Hatano, S.Y., Suemori, H., Nakatsuji, N., and Tada, T. (2005)
Octamer and Sox Elements are Required for Transcriptional cis Regualtion of Nanog Gene
Expression, VOL 25, NO 6, pp 2475-2485.

15) Kopp, J.L., Ormsbee, B.D., Desler, M., and Rizzino, A. (2008), Small Increases in the Level of
Sox-2 Trigger the Differentiation of Mouse Embryonic Stem Cells, Stem Cells, DOI:
10,1634/stemcells.2007-0951.

16) Lee, J., Kim, H., Rho, J.Y., Han, Y.M., and Kim, J. (2006) The Human OCT-4 Isoforms Differ in
Their Ability to Confer Self-renewal, The Journal of Biological Chemistry (JBC), VOL 281,
NO 44, pp 33554-33565.

17) Lin, J., Loh, Y.H., Zhang, W., Chen, X., Tam, W.L., Yeap, L.S., Li,P., Ang, Y.S., Lim, B.,
Robson, P., and Ng, H.H. (2005) Reciprocal Transcriptional Regulation of Pou5f1 and Sox-2
via The Oct4/ Sox2 Complex in Embryonic Stem Cells, Molecular and Cellular Biology, VOl
25, NO, 14, pp 6031-6046.

18) Loh, Y.H., Wu, Q., Chew, J.L., Vega, V.B., Zhang, W., Chen, X., Borque, G., George, J., Leong,
B., Liu, J., Wong, K.W., Sung, K.W., Lee, C.W.H., Zhao, X.D., Chiu, K.P., Lipovich, L.,
Kuznetsov, V.A., Robson, P., Stanton, L.W., Wei, C.L., Ruan, Y., Lim,B., and Ng, H.H.
(2006) The Oct and Nanog Transcription Network Regulates Pluripotency in Mouse
Embryonic Stem Cells, Nature Genetics, VOL 38, NO 4, pp 431-440.

19) Mahatan, C.S., Kaestner, K.H., Gieman, D.E., and Yang, V.W. (1999) Characterization of the
Structure and Regulation of the Murine Gene Encoding Gut-enriched Kruppel- Like Factor
(Kruppel-like Factor 4), Nucleic Acids Research, VOL.27, NO 23. (Figure
#
1 and
#
35 of this
thesis use Figure
#
1 of this paper).

20) Martin-ibanez, R., Noelia Urban, Sergent-Tanguy, S., Pineda, R.J., Garrido-Clua,N., Alberch, J.,
and Canals, J.M. (2007) Interplay of Lukemia Inhibitory Factor and Retinoic Acid on Neural
Differentiation of Mouse Embryonic Stem Cells, Journal Of Neuroscience Research VOL 85,
pp 2686-2701.

21) Nakagawa, M., Koyanagi, M., Tanabe, K., Takahash, K., Ichisaka, T., Aoi, T., Okita, K.,
Mochiduki, Y., Takazawa, N., and Yamanaka, S. (2008) Generation of Induced Pluripotent
Stem Cells Without MYC From Mouse and Human Fibroblasts, Nature Biotechnology VOL
26, NO 1, pp 101-106.

22) Nishimoto, M., Fukushima, A., Okuda, A., and Muramatsu, M. (1999) ) The Gene for the
Embryonic Stem Cell Coactivator UTF1 Carries a Regulatory Element Which Selectively
Interacts with a Complex Composed of Oct-4 and Sox-2, Molecular and Cellular Biology,
VOL 19, NO 8, pp 5453-5465.

23) Niwa, H., Miyazaki ,J.I., and Smith, A.G.(2000) Quantitative Expression of Oct3/4 Defines
Differentiation, Dedifferentiation or Self-Renewal of ES cells, Nature Genetics, VOL 24,
Letter, pp 372-676.

98
24) Okumura-Nakanishi, S., Saito, M., Niwa, H., and Ishikawa, F. (2004) Oct-3/4 and Sox-2 Regulate
Gene in Embryonic Stem Cells, The Journal of Biological Chemistry, VOL 280, NO 7, pp
5307-5317.

25) Pan, G., Li, J., Zhou, Y., Zheng, H., and Pir, D. (2006) A Negative Feedback Loop of
Transcription Factors That Controls Stem Cell Pluripotency and Self-Renewal The FASEB
Journal, VOL 20, pp E1094-E1102, supplementary pp 1730-1732.

26) Remenyi, A., Lins, K., Nissen, L.J., Reinbold, R., Scholer, H.R., and Wilmanns, H. (2003)
Crystal Structure of the POU/HMG/DNA Ternary Complex Suggests Differential Assembly
of Oct-4 and Sox-2 on Two Enhancers, Genes & Development, VOL 17, pp 2048-2059.

27) Rowland, B.D., Bernards, R., and Peeper, D.S. (2005) The Klf4 Tumor Suppressor is a
Transcriptional Repressor of p53 That Acts as a Contex-dependant Oncogene, Nature Cell
Biology, VOL 7, NO 11, pp 1074-1082.

28) Roya, T., and Clevers, H. (2005) WNT Signaling in Stem Cells and Cancer, Nature Genetics VOL
434, NO 14, pp 843-850.

29) Sene, K.H., Porter, C.J., Palidwor, G., Perez-Iratxeta, C., Muro, E.M., Campbell, P.A., Rudnicki,
M.A., and Andrade- Navarro, M.A. (2007) Gene Function in Early Mouse Embryonic Stem
Cell Differentiation, BMC Genomics, VOL 8, NO 85 pp 1-21 (Micro array data is available
for public use and includes additional genes not discussed in paper) www.biomedical .com
/1471-2164/8/85

30) Shie,G.J., Chen, Z.Y., Fu, M., Pestell, R.G., and Tseng, C.C. (2000) Enriched Krupple-Like
Factor Represses Cyclin D1 Promoter Activity Through Sp1 Motif, Nucleic Acids Research,
VOL 28, NO 15, pp 2969-2976.

31) Shields, J.M., and Yang, V.W. (1998) Identification of the DNA Sequence That Interacts With the
Gut-Enriched Kruppel-Like Factor, Nucleic Acids Research, VOL 26, NO 3, pp 96-802.

32) Simonis, M., Kooren, J., and De Laat, W. (2007) An Evaluation of 3C-Based Methods to Capture
DNA Interactions, Nature Methods, VOL 4, NO 11, pp 895-901.

33) Takahahi, K., and Yamamaka, S. (2006) Induction of Pluripotent Stem Cells From Mouse
Embryonic and Adult Fibroblast Cultures by Defined Factors, Cell, VOL 131 pp 861-872.

34) Takao, Y., Yokota, T., and Koide, H. (2007) Beta Catenin Up-regulates Nanog Expression
Through Interaction With Oct3/4 in Embryonic Stem Cells, Biochemical and Biophysical
Research Communications, NO 253, pp 699-705.

35) Wei, D., Kanai, M., Huang, S., and Xie, K.(2005) Emerging Role of KLf4 in Human
Gastrointestinal Cancer, Carcinogenesis, VOL 27, NO 1, pp 23-31.

36) Wiebe, M.S., Nowling, T.K., and Rizzino, A. (2003) Identification of Novel Domains Within
Sox-2 and Sox-11 Involved in Autoinhibition of DNA Binding and Partnership Specificity,
The Journal of Biological Chemistry, VOL 278, NO 20, pp 17901-17911.

37) Willems, E., Leyns, L. (2007) Patterning of Mouse Embryonic Stem Cell-Derived Pan-Mesoderm
by Activin A/Nodal and BMP4 Requires Fibroblast Growth Factor Activity, Differentiation,
original Article Journal Compilation, pp 1-15. DOI:10.1111/j.1432-0436.2007.00257.x
99
38) Ying, Q.L., Nichols, J., Chambers, I., and Smith, A. (2003) BMP Induction of Id Proteins
Suppresses Differentiation and Sustains Embryonic Stem Cell Self-Renewal in Collaboration
With Stat3, Cell, VOL 115, pp 281-292.

39) Yoon, H.S., and Yang, V.W. ( 2004) Requirement of Kruppel-like Factor 4 in Preventing Entry
into Mitosis Following DNA Damage, The Journal of Biological Chemistry, VOL 279, NO
6, pp 5035-5041.

40) Zhang, W., Chen, X., Kato, Y., Evans, P.M., Yuan, S., Yang,J., Rychahou, P.G., Yang, V.W.,
He, X., Evers, B.M., Liu, C.( 2006) Novel Cross Talk of Kruppel-like Factor 4 and β-Catenin
Regulated Normal Intestinal Homeostasis and Tumor Repression, Molecular and Cellular
Biology, VOL 26, NO 6, pp 2055-2064.
100
Appendix. Abbreviations
mES- Mouse Embryonic Stem cells

hES- Human Embryonic Stem cells

EB- Embroid Bodies, differentiated hES cells

Klf4- Krupple-like Factor 4

Klf4-ΔC- Krupple-like Factor 4 with Carboxy terminus deleted

Klf4-ΔM- Krupple-like Factor 4 with middle region deleted

Klf4- Δzf 429- Krupple-like Factor 4 with second zinc finger located at amino acid 429 deleted

Klf4-Δzf 454- Krupple-like Factor 4 with third zinc fingr located at amino acid 454 deleted

FUIPW- Vector with Ubiquitin promoter and a internal ribosomal entry site that drives Puromyocin
resistance.

FUIGW- Vector with Ubiquitin promoter and a internal ribosomal entry site that drives Green
florescent protein expression

LIF- Leukemia Inhibitory Factor

C-myc-Cellular Myelocytomatosis Oncogene

BMP- Bone Morphogenic Protein

Oct-4- a.k.a Pouf51- Octomer-Binding Transcroption Factor 4

Sox-2- SRY-box Containing Gene 2

Nanog- Nanog Homeobox

RSAT- Regulatory Sequence Analysis Tool

Tdgf1- Teratocarcinoma Derived Growth Factor

Dnm3b- DNA Methyl Transferase 3b

Gabrb3- Gamma Aminobutyric Acid Receptor Subunit Beta 3

Gdf3-Growth Differentiation Factor 3

SSEA1- Stage Specific Antigen 1

FBS- Fetal bovine Serum

GFP- Green Fluorescent Protein
101
CF1- feeder free mES cell line

293T cells- Somatic cell line

Gapdh- Glyceraldehyde-3-phosphate dehydrogenase

Fgf8- Fibroblast Growth Factor 8

T- Brachyury

Foxa2- Forkhead Box A2

Nkx 2.5- NK@ Transcription Factor Related, Locus 5

Neuro-d- Neurogenic Differentiation

Cdx2- Caudal Type Homeobox 2

q-PCR- Quantitiative Polymerase Chain Reaction

RT-PCR- Reverse Transcriptase Polymerase Chain Reaction

Abstract (if available)

Abstract Human and mouse ES cells are defined by their ability to confer self-renewal and differentiate into various cell types, making them an invaluable tool for investigation of Stem cell maintenance, differentiation, and reprogramming. Human ES cells have infinite potential for therapeutic applications. Unfortunately, acquiring ES cells is problematic, and reprogramming somatic cells to Induced Pluripotent Stem Cells occurs by an unknown mechanism.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

Deletion of Krüppel-like factor 4 (KLF4) affects NPCs to embark on differentiative progression and leads to delayed neurogenesis in mouse embryonic cortex

PDF

The mechanisms of somatic cell reprogramming

PDF

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

PDF

Exploring stem cell pluripotency through long range chromosome interactions

PDF

Hepatic differentiation in human naïve stem cells compared to human embryonic stem cells

PDF

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

PDF

Novel roles for Maf1 in embryonic stem cell differentiation and adipogenesis

PDF

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

PDF

Derivation and characterization of human embryonic stem (hES) cells and human induced pluripotent stem (hiPS) cells in clinical grade conditions

PDF

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

PDF

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

PDF

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

PDF

The noncanonical role of telomerase in prostate cancer cells: exploring a non-telomeric signaling role for telomerase protein (TERT) in a cancer cell line

PDF

Interaction of epigenetics and SMAD signaling in stem cells and diseases

PDF

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

PDF

Glucocorticoids and Runx2 synergistically stimulate Wif1 and compromise pre-osteoblats in vitro

PDF

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

PDF

IPS and CNS cell models of peroxisomal disorders

PDF

Studies on the role of a novel protein, TMEM 56 in tumorigenic growth for MCF-7 cells

Asset Metadata

Creator Andrianakos, Rosemary G. (author)

Core Title Roles of Klf4 in embryonic stem cells

School Keck School of Medicine

Degree Master of Science

Degree Program Biochemistry

Publication Date 04/29/2008

Defense Date 03/15/2008

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

Tag differentiation of Es cells,ES cell maintenance and self renewal,Klf4,OAI-PMH Harvest,Oct-4,pluripotency,reprogramming somatic cells,Sox-2,stem cells

Language English

Advisor Lu, Wange (committee chair), Stellwagen, Robert H. (committee member), Tokes, Zoltan A. (committee member)

Creator Email roseandrianakos@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1210

Unique identifier UC148941

Identifier etd-Andrianakos-20080429 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-70866 (legacy record id),usctheses-m1210 (legacy record id)

Legacy Identifier etd-Andrianakos-20080429.pdf

Dmrecord 70866

Document Type Thesis

Rights Andrianakos, Rosemary G.

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu