University of Southern California Dissertations and Theses

Targeting tumor-initiating stem-like cells through metabolic and epigenetic regulations

(USC Thesis Other)

Targeting tumor-initiating stem-like cells through metabolic and epigenetic regulations

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

TARGETING TUMOR-INITIATING STEM-LIKE CELLS THROUGH
METABOLIC AND EPIGENETIC REGULATIONS

By

Chia-Lin Chen

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(GENETIC, MOLECULAR AND CELLULAR BIOLOGY)
August 2016

Copyright 2016 Chia-Lin Chen
Dedication

To my dearest family
Thanks for your unconditional support

To my friends and people I have ever met
The reason of what I become today

iii

Acknowledgements

First and foremost, I am most grateful to my advisor Dr. Keigo Machida who provided me
this opportunity to pursue my doctoral degree at University of Southern California. In addition,
I would like to thank my committee members, Dr. J.-H. James Ou and Dr. Bangyan L. Stiles
guide me and expand my scientific horizons. Special thanks to Dr. Jae Jung for his helpful
advice and stimulating discussion at my presentations.
I also would like to thank Dr. Vasu Punj from Norris Bioinformatics Core of University of
Southern California for ChIP-sequencing analysis, Dr. Jun Xu from Pathology of University of
Southern California for seahorse assay and fatty acid assay, Dr. Yibu Chen and Dr. Meng Li
from Norris Medical library of University of Southern California for RNA sequencing analysis,
Dr. Jayanth Panyam and Dr. Suresh Swaminathan from Pharmacy of University of Minnesota
for nanoparticle production, Dr. San-Duo Chen from National Chung Hsing University for
pathological assessment of tumor histology, and Dr. Chi-Der Chen from University of
Pennsylvania for mesenchymal stem cell culture. This great work would not have been
possible without this outstanding team work.
I would like to acknowledge California Institute Regenerative Medicine (CIRM) training
grant for financial support during my last two years of PhD training. I am also thankful the
facility core of University of Southern California, including Flow Cytometry Core Facility of USC
Stem Cell center, as well as USC Research Center for Liver Diseases Cell Biology Core
Laboratory.
My dearest father Dr. San-Duo Chen illustrated me the amazing world under the
microscope when I was a little girl. At that moment, doing research became my destiny. At this
moment, I deeply appreciate my family and friends’ unconditional support and the venerable
pioneer’s, Dr. Michael M. C. Lai, encouragement to help me find my own path in life.

iv

Table of Contents

Page Number
Dedication ii
Acknowledgements iii
Table of Contents iv
List of Figures vii
Abstract ix

Chapter One: Introduction 1
1.1 Hepatocellular Carcinoma 1
1.2 Tumor-initiating stem-like cell population 2
1.3 TLR4 signaling activates NANOG to suppress TGF-β pathways in TICs 3

Chapter Two: Targeting the TIC population via metabolic regulation 8
2.1 Introduction 8
2.2 Results 8
2.2.1 Identification of Nanog targets in TICs 8
2.2.2 NANOG suppresses mitochondrial OXPHOS 10
2.2.3 NANOG promotes mitochondrial FAO 14
2.2.4 NANOG orchestrates mitochondrial metabolic reprogramming 17
2.2.5 NANOG orchestrates TIC metabolic reprogramming via AMPK
pathway
19
2.2.6 NANOG prevents mitochondrial ROS production and maintains self-
renewal ability of TICs
21
2.2.7 NANOG orchestrates oncogenicity and therapeutic resistance
mechanisms via mitochondrial metabolic reprogramming
24
2.3 Discussion 26

Chapter Three: Targeting the TIC population via epigenetic regulation 28
3.1 Introduction 28
3.2 Results 28
3.2.1 Identification of FDA-approved drug(s) that can specifically target 28
v

tumor-initiating stem-like cells
3.2.2 ATRA and SAHA combination induces cell apoptosis pathways and
reduces the self-renewal ability of TIC in vitro
34
3.2.3 Genome-wide transcriptome analysis reveals the mechanism for
ATRA and SAHA combination targeting of TICs
37
3.2.4 ATRA + SAHA combination treatment targets the TIC population via
suppression of miR-22
42
3.2.5 The ATRA+SAHA treatment induces the TIC growth arrest and
apoptosis via the PTEN-FOXO pathway
44
3.2.6 The ATRA+SAHA treatment alters the DNA methylation pattern of
NANOG via regulation of miR-22 and TET2
47
3.2.7 Dual drug combination treatments attenuates tumor growth in vivo 52
3.3 Discussion 56

Chapter Four: Experimental procedures 60
4.1 Experimental procedures related to Chapter 2
60
4.1.1 ChIP-seq sample preparation and bioinformatics analysis
60
4.1.2 XF24 extracellular flux analyzer for measurement of cellular OCR
and ECAR
61
4.1.3 Stable-isotope carbon labeling is traced for flux analysis
62
4.1.4 Fatty acid β-oxidation assay
62
4.1.5 Mitochondria labeling and measurement of ROS levels
62
4.1.6 ATP production measurements
63
4.1.7 Cox6a2 and Acadvl promoter luciferase assay
63
4.1.8 Quantitative real-time PCR (qPCR)
63
4.1.9 In vivo rescue experiments of OXPHOS gene inhibition and FAO by
implantation of TICs into immunocompromised mice
64
4.2 Experimental procedures related to Chapter 3
65
4.2.1 TICs isolation
65
vi

4.2.2 Cell culture
65
4.2.3 Chemical screening and analysis
65
4.2.4 Cell viability assay
66
4.2.5 Nanog-GFP screening
66
4.2.6 Determination of combination dose
67
4.2.7 Annexin V staining
67
4.2.8 Caspase activity analysis
68
4.2.9 TUNEL staining assay
68
4.2.10 Tumor spheroid formation assay
69
4.2.11 Anchorage-independent growth assay
69
4.2.12 All-trans Retinoic acid nanoparticles conjugated with CD133
69
4.2.13 In vivo tumorigenicity experiments
70
4.2.14 RNA sequencing
71
4.2.15 Bisulfite sequencing
72
4.2.16 Chromatin immunoprecipitation assays (ChIP-qPCR)
72
Bibliography 74

vii

List of Figures

Page number
Figure 1-1 Surveillance results from the American Cancer Society 1
Figure 1-2 TLR4 activates NANOG through E2F regulation 5
Figure 1-3 Ectopic activation of TLR4 induced IGF2BP3 mediated
YAP1 phosphorylation
7
Figure 2-1 Identification of Nanog downstream target genes using
ChIP-sequencing
9
Figure 2-2 Nanog suppressed the OXPHOS pathways. 10
Figure 2-3 Nanog suppressed oxygen consumption in TICs. 11
Figure 2-4 Nanog directly regulated the OXPHOS gene, Cox6a2. 13
Figure 2-5 Nanog promoted fatty acid oxidation. 14
Figure 2-6 Nanog directly regulated expression of the FAO gene,
Avadvl.
16
Figure 2-7 Nanog suppressed fatty acid synthesis. 18
Figure 2-8 Nanog metabolically reprogrammed TICs through AMPK
pathway.
20
Figure 2-9 Nanog prevented ROS production in TICs. 22
Figure 2-10 Nanog regulated expression of OXPHOS and FAO genes
to prevent ROS production by TICs.
23
Figure 2-11 Suppression of tumor growth and drug resistance through
metabolic regulation.
25
Figure 3-1 Identification of selective inhibitors for CD133 (+) cells. 30
Figure 3-2 Identification of selective Nanog inhibitors. 31
Figure 3-3 Schematic diagram of drug combination screenings. 33
Figure 3-4 The ATRA-SAHA combination induces TIC apoptosis. 35
Figure 3-5 Combination treatment inhibited the self-renewal ability of
TICs.
36

Figure 3-6 Genome-wide transcriptome analysis of drug treated TICs. 38
Figure 3-7 Transcriptome analysis of individual drug treatment of
TICs.
39
Figure 3-8 The drug combination suppressed embryonic stem cell
pluripotency pathways and induced apoptosis pathway.
40
Figure 3-9 Ingenuity Pathway Analysis of the unique set of gene in the
combination treatment.
41
Figure 3-10 Drug combination down-regulated miR-22 to suppress the 43
viii

self-renewal of TICs.
Figure 3-11 Drug combination regulated the key pathways of TICs. 44
Figure 3-12 The drug combination activated PTEN, which is down-
regulated in HCC patients.
45
Figure 3-13 The combination treatment activated PTEN-FOXO
pathway.
46
Figure 3-14 Suppression of miR-22 by drug combination activated
TET2.
48
Figure 3-15 Drug combination activates epigenetic regulators, TET2
and DNMT3A.
49
Figure 3-16 Synergistic interaction of TET2 and DNMT3A induced by
the drug combination treatment altered the DNA
methylation pattern of the Nanog promoter.
50
Figure 3-17 The drug combination attenuated tumor growth. 52
Figure 3-18 The combination treatment suppressed the tumor
recurrence and poor survival gene sets.
54
Figure 3-19 Hypothetical Model 55

ix

Abstract

Hepatocellular carcinoma (HCC) is one of the most common cancers and is a leading
cause of deaths worldwide. Conventional chemotherapy and surgery are the current
treatments for HCC patients; however, tumor growth and recurrence often are not eliminated
by these treatments. Tumor-initiating stem-like cells (TICs), also known as cancer stem cells,
a minor population in the bulk tumor, have been identified as a vital population in tumor
recurrence and drug resistance. So far, the biological features of the TIC population and the
role of TICs in disease progression are not fully understood. Accumulating evidence and our
recent studies have shown that NANOG, a key stemness factor, plays a central role in
regulating the self-renewal and pluripotency of the TIC population. NANOG is expressed in
germ cell tumors as well as other types of carcinomas, including breast, cervix, oral cavity,
kidney, ovary, liver, and prostate. More importantly, overexpression of NANOG promotes
tumor cell resistance to apoptosis and therapeutic agents via the AKT pathway. Recently, we,
along with others, showed that down-regulation of NANOG expression significantly attenuated
tumor growth. To determine whether targeting NANOG could be an effective therapeutic
strategy, we focused on the identification of upstream and downstream factors of NANOG.
To elucidate the downstream target genes of NANOG in TICs, we used a NANOG-
specific antibody in a chromatin immunoprecipitation (ChIP)-sequencing assay. We found that
NANOG was highly associated with mitochondrial metabolic pathways, such as the oxidative
phosphorylation pathway (OXPHOS) and lipid metabolism pathway. Cancer cells are known
to rewrite their bioenergetic programs to adapt and enable rapid proliferation and survival
under harsh conditions. Our chromatin immunoprecipitation-sequencing (ChIP-seq) results
suggested that NANOG might play a role in the metabolic transformation of TICs. The causal
roles of NANOG in mitochondrial metabolic reprogramming occurred through the inhibition of
x

OXPHOS to prevent the production of mitochondrial reactive oxygen species (ROS) and
activation of fatty acid oxidation (FAO), which are required for self-renewal and drug resistance.
Restoration of OXPHOS activity and inhibition of FAO rendered TICs susceptible to a standard
of care chemotherapy drug, sorafenib. The findings revealed that NANOG-mediated
generation of TICs, tumorigenesis, and chemoresistance via dysregulation of mitochondrial
functions.
To identify drug candidates that can suppress Nanog expression directly in TICs, we
performed high-throughput screening of an FDA-approved drug library in TICs. Based on the
drug screening results, we found the combination of an all-trans retinoic acid (ATRA) and a
histone deacetylase (HDAC) inhibitor, Vorinostat (SAHA), inhibited the cell growth and self-
renewal abilities of the TIC population in vitro and suppressed tumor growth in vivo. Genome-
wide transcriptome analysis using RNA sequencing showed that the combined treatment
regimen reduced microRNA-22, which induced expression of phosphatase and tensin
homolog (PTEN) and ten-eleven translocation (TET). PTEN-mediated FOXO activation
promoted TIC apoptosis. In addition, TET interacted synergistically with DNA
methyltransferase 3A (DNMT3A) to alter the methylation pattern within the proximal Nanog
promoter region, leading to suppression of Nanog expression. Taken together, the ATRA-
SAHA combination treatment epigenetically targets the TIC population and may serve as a
novel strategy for HCC treatment.
In conclusion, we utilized the identification of the upstream and downstream factors of
NANOG to reveal drug targets that may eradicate the TIC population. The most predominant
anti-cancer therapy is to target the TIC population via epigenetic and metabolic regulation.
This project revealed the fundamental mechanism of epigenetic and metabolic regulation of
Nanog in the TIC population and elucidated a new therapeutic avenue for anti-cancer
treatment.
1

Chapter One: Introduction

1.1 Hepatocellular carcinoma
Hepatocellular carcinoma (HCC) is one of the most common cancers and leading
cause of cancer deaths in the world. Every year, approximately half a million people are
diagnosed with HCC worldwide (Ferlay et al, 2012). The American Cancer Society
estimates that in 2016, more than 39,000 new cases of liver cancer and more than
27,000 liver cancer deaths are expected in the United States (Fig. 1-1). The liver cancer
death rate has been increasing since 1980; from 2003 to 2012, the rate increased by
2.7% per year. The most critical issue for HCC is its high mortality rate. The 5-year
survival rate for HCC is only 17%. In addition, only 31% of patients diagnosed with a
localized stage of HCC survive for more than 5 years. Even though patients receive the
conventional surgical resection and/or chemotherapy, tumor growth and relapse still
cannot be fully eliminated.

Figure 1-1. Surveillance results from the American Cancer Society show that
cancers of the liver and intraphepatic bile duct are among the most common cancers with
high mortality in the US.
2

Risk factors associated with development of HCC include hepatitis virus infection,
excessive alcohol consumption, dietary aflatoxin exposure, and non-alcohol-associated
diseases such as obesity and diabetes. Chronic infection with Hepatitis C virus (HCV) or
Hepatitis B virus (HBV) leads to liver damage and liver fibrosis, which may result in
hepatocarcinogenesis. Previously, our laboratory demonstrated that HCV infection not
only activated nitric oxide synthase, resulting in cellular DNA damage and gene
mutations (Machida et al., 2004), but also triggered mitochondrial ROS production,
leading to DNA damage and STAT3 pathway activation (Machida et al., 2006). Moreover,
increasing evidence has shown that synergistic interactions between HCV and
alcoholism, obesity, and diabetes increase the risk of HCC (Hassan et al., 2002).
Significant synergism has been observed between heavy alcohol consumption, HCV
infection (odds ratio: 59.9), and diabetes mellitus (odds ratio: 9.9) (Stickel, 2015).
Recently, our laboratory showed that overexpression of HCV protein in liver-specific
transgenic mice in high fat diet and alcohol feeding models significantly increased liver
tumor incidence (Chen et al., 2013 and Dinesh et al., 2016).
During chronic liver damage, toll-like receptor (TLR) signaling is significantly
up-regulated (Testro and Visvanathan, 2009). We further demonstrated that HCV
induced toll-like receptor 4 (TLR4) signaling, which resulted in activation of a stemness
gene, Nanog (Machida, 2006 and 2009) and generation of an original TIC population.

1.2 Tumor-initiating stem-like cell population
TICs, also referred to as cancer stem cells, are considered an important
population in tumor recurrence and therapy resistance (Zhou et al., 2009). The TIC
population possesses several key properties of normal stem cells including self-renewal,
unlimited proliferative potential, and the ability to give rise to daughter cells. However,
unlike highly organized normal stem cells, TICs show aberrant regulation of self-renewal
3

and differentiation programs and produce daughter tumor cells that are in various stages
of differentiation. TICs have been isolated from different types of solid tumors using
various cell surface markers, such as CD133 (Ma et al., 2007), CD49f (Rountree et al.,
2008), CD24 (Lee et al., 2011), and CD90 (Ma et al., 2007). Among these cell surface
markers, CD133 was first used for TIC isolation from a Huh7 human HCC cell line by
Suetsugu et al. (2006). CD133 (+) TICs have the ability to proliferate rapidly (in vitro and
in vivo) and have a preferential potential to form spheroids in primary and subsequent
passages. In vivo, CD133 (+) mouse xenograft models exhibited a greater tendency to
develop tumors, even upon serial transplantation. The xenograft mice also exhibited
chemotherapy resistance through the AKT/PKB and Bcl-2 pathways (Ma et al., 2008).
CD133 (+) TICs express high levels of stemness-associated genes such as
NANOG, SOX2, OCT3/4, BMI-1, NOTCH, β-CATENIN, SMO, NESTIN, ABCG2, and
ABCB1 (Ma et al., 2010). NANOG is in the homeobox family of DNA-binding transcription
factors, promotes oncogenesis (Jeter et al., 2009, Sun et al., 2014), and plays an
important role in maintenance of the TIC population. The tumorigenic effects of NANOG
are associated with cellular and molecular changes such as increased expression of
CD133, ALDH1, CXCR4, and IGFBP5 (Jeter et al., 2011). NANOG is not only expressed
in germ cell tumors (Hoei-Hansen, 2008) but also in carcinomas of the breast (Ezeh et al.,
2005), cervix (Ye et al., 2008), oral cavity (Chiou et al., 2008), kidney (Bussolati et al.,
2008), ovary (Zhang et al., 2008), liver (Xu et al., 2010), and prostate (Shen et al., 2011).
More importantly, overexpression of NANOG promoted tumor cell resistance to apoptosis
and therapeutic agents via the AKT pathway (Noh et al., 2012).

1.3 TLR4 signaling activates NANOG to suppress the TGF-β pathway in TICs
Previously, our lab isolated TICs by CD133 and CD49f sorting (Machida et al., 2009).
We found that the CD133 (+)/CD49f (+) TIC population expressed high levels of TLR4
4

and stemness genes, such as Nanog, Oct4, and Sox2. Moreover, silencing of TLR4 or
NANOG significantly attenuated tumor growth (Chen et al., 2016). To understand the
regulation of NANOG expression in TLR4 activation, we investigated the endogenous
NANOG promoter region. Using chromatin immunoprecipitation-quantitative PCR
(ChIP-qPCR) and promoter luciferase assay, we found that E2F1 was enriched in both
the distal enhancer and proximal promoter regions of NANOG.
We further analyzed the adaptor molecules/kinases in the TLR4 signaling cascade,
such as TBK1, TAB1, IRF3, TRAF6, and TAK1 and showed that TLR4 activated TAK1
and TBK1, which resulted in phosphorylation of E2F1 and activation of Nanog (Fig. 1-2)
(Chen et al., 2016).

5

Figure 1-2. TLR4 activates NANOG through E2F regulation. Upon LPS stimulation of
the TLR4-CD14 complex, the adaptor protein MYD88 and TRIF are recruited to the
TRL4-CD14 complex. Subsequently, the TAB1/TAB2/TAK1 complex is recruited to the
TRAF6 complex by ubiquitylation (Ofengeim and Yuan, 2013). Ubiquitylation of TRAF6 by
Pellino 1 (PELI1) induces the recruitment of the Ikk kinase complex, which regulates the
NF-κB-mediated inflammatory response. Ubiquitylation of the TRAF6 complex also
provides docking sites for TAK1 and TAK1-binding proteins (TAB1 and TAB2). In our study,
we showed that the activated TAB1/TAB2/TAK1 complex subsequently phosphorylated
E2F1, leading to nuclear translocation to the distal enhancer and proximal promoter
region of Nanog, resulting in its activation.
6

Moreover, our lab demonstrated that this TLR4/NANOG-dependent TIC population
was defective in the TGF-β tumor suppressor pathway (Chen et al., 2013). By functional
oncogene screening of a TIC cDNA library, we identified Yes-associated protein 1 (YAP1)
and insulin-like growth factor II mRNA binding protein 3 (IGF2BP3) as
NANOG-dependent genes that inactivate TGF-β signaling. Mechanistically, we
determined that YAP1 mediates cytoplasmic retention of phosphorylated SMAD3 and
suppresses SMAD3 phosphorylation/activation by the IGF2BP3/AKT/mTOR pathway.
Silencing of both YAP1 and IGF2BP3 restored TGF-β signaling, inhibited expression of
pluripotency genes and tumorigenesis, and abrogated chemoresistance of TICs (Fig.
1-3). These results suggested that the activated TLR4/NANOG oncogenic pathway is
critical for suppression of the cytostatic TGF-β tumor suppressor pathway and could
potentially serve as a therapeutic target for HCV-related HCC (Chen et al., 2013).
To explore an effective therapeutic strategy for specifically targeting the TIC
population, in this proposal, we focused on the identification of upstream and
downstream factors of NANOG.

7

Figure 1-3. Ectopic activation of TLR4 by LPS stimulation induces the
overexpression of the pluripotency factor NANOG and self-renewal of TICs.
NANOG induces IGF2BP3 and YAP1, which in turn inhibit TGF-β signaling at the level of
SMAD3 phosphoactivation and p-SMAD3 nuclear translocation. The former effect is
dependent on the IGF2BP3/AKT/mTOR pathway, while the latter is caused by
p-YAP1/SMAD7/SMAD3 interactions, which are enhanced by IGF2BP3/AKT–mediated
YAP1 phosphorylation. Restoration of the TGF-β tumor suppressor pathway by silencing
IGF2BP3 and YAP1 down-regulates TLR4 via SMAD3-mediated transcriptional
repression and inhibits TLR4/NANOG–mediated TIC self-renewal and oncogenic activity
while chemosensitizing TICs (Chen et al., 2013).

8

Chapter Two: Targeting the TIC population through metabolic regulation

2.1 Introduction

We have previously showed that TLR4 signaling transactivates NANOG in TICs. More
importantly, down-regulation of Nanog attenuated tumor progression (Chen et al., 2013).
Therefore, an in-depth understanding of the underlying molecular mechanisms of NANOG
regulation in hepatocarcinogenesis is essential for the development of improved cancer
therapies. In this study, we conducted ChIP-seq with a NANOG-specific antibody in TICs
to identify downstream target genes of NANOG in TICs, leading to development of better
therapeutic strategies for cancer treatment.

2.2 Results

2.2.1 Identification of NANOG targets in TICs
To investigate the underlying mechanism of TIC-mediated tumorigenicity, we
conducted a genome-wide transcriptional profiling of NANOG-promoter interactions in
TICs with a ChIP-seq approach using a NANOG-specific antibody. The ChIP-seq data
identified NANOG binding sites enriched in the proximal to transcription start sites (TSS)
of genes in TICs when compared to CD133 (-) control cells (Fig. 2-1A and 2-1B).
Furthermore, an Ingenuity Pathway analysis (IPA) identified that the target genes are
involved in mitochondrial functions, including oxidative phosphorylation
(OXPHOS)-related and FAO (Fig. 2-1C). Interestingly, a bioinformatics analysis of
NANOG-enriched promoter fragments showed the de novo consensus binding sites were
similar to the STAT3-consensus-binding motif (Fig. 2-1D). It is worth mentioning that
NANOG has been shown to physically bind STAT3 (Torres and Watt, 2008). In fact, we
9

recently found that TLR4 signaling activated Twist via NANOG, which cooperated with
STAT3 to promote formation of TICs and tumorigenesis (Uthaya Kumar et al., 2016).

Figure 2-1. Identification of Nanog downstream target genes using
ChIP-sequencing
(A) Nanog-ChIP analysis: comparison of promoter fragments from CD133 (-) and CD133
(+) cell populations.
(B) Nanog enrichment proximal to initiation site of gene promoters in CD133 (+) cells, but
not in CD133 (-) cells.
(C) Summary of gene ontology families identified by Nanog ChIP-seq analysis
(D) De novo Nanog-binding motifs resemble STAT3-binding motifs.
10

2.2.2 NANOG suppresses mitochondrial OXPHOS
Next, we investigated the importance of metabolic genes, especially those
participating in oxidative phosphorylation based on our gene ontology analysis of
NANOG ChIP-seq results (Fig. 2-2A and 2-2B). As shown in Fig. 2-2C, Nanog silencing
in TICs significantly up-regulated OXPHOS-associated genes including Ndufs2, Ndv2,
Uqcrfs1, Cox6a2, Atp5d, Atp5h, Atp6v1g2, and Cox15 in TICs, which suggests that
NANOG may negatively regulate the OXPHOS pathway.

Figure 2-2. Nanog suppressed the OXPHOS pathways.
(A) Representative model of the OXPHOS pathway. The genes marked in red are hits
from Nanog ChIP-seq
(B) Histogram of Nanog ChIP-seq identified the mitochondrial genes that are Nanog
target genes in CD133 (+) cells.
(C) OXPHOS gene expression was up-regulated in Nanog-silenced cells.
11

Figure 2-3. Nanog suppresses oxygen consumption in TICs.
(A) Nanog- or TLR4-silenced cells had a significantly higher OCR, which indicates that
Nanog suppresses mitochondrial respiratory activity. Furthermore, FCCP-induced OCR
was abrogated by FAO inhibitor (ETO, Etomoxir), but not glycolysis inhibitor (2-DG,
2-deoxyglucose) (n = 3, *p < 0.05).
(B) Quantified data from seahorse results.
12

To functionally test whether NANOG regulated mitochondrial respiration, we
conducted the seahorse extracellular flux assay to determine the oxygen consumption
rate (OCR) in TICs (Fig. 2-3A). As shown in Fig. 2-3B and 2-3C, the basal OCR rates
were significantly higher in Nanog- or TLR4-silenced TICs when compared to control
cells, which suggests that NANOG suppressed mitochondrial respiration in TICs.
To test whether NANOG directly regulated OXPHOS gene expression, we analyzed
the binding activity of the cytochrome c oxidase subunit 6A (Cox6a2) gene because it
was the most down-regulated OXPHOS gene in our qRT-PCR analysis (Fig. 2-2C).
As shown in Figure 2-4A, Nanog enrichment was observed in the Cox6a2 promoter
of TICs as measured by ChIP-qPCR analysis. Furthermore, knockdown of Nanog in TICs
resulted in increased transcription from the Cox6a2 promoter (nt −1433 to +17), which
was suppressed in TICs. A deletion in the Cox6a2 promoter from nt -1433 to -518
significantly increased Cox6a2 promoter activity in TICs (Fig. 2-4B), which indicated that
this region had negative-regulatory activity. Mutations in the Nanog binding sites (-1078
and -790) restored Cox6a2 promoter activity in TICs (Fig. 2-4C). These results
demonstrate that suppression of mitochondrial activity in TICs results from
NANOG-mediated repression of OXPHOS genes.
13

Figure 2-4. Nanog directly regulated the OXPHOS gene, Cox6a2.
(A) ChIP-qPCR of NANOG in the Cox6a2 promoter of TICs (n = 3, *p < 0.05).
(B) Truncation of the Nanog promoter identified the region responsive to
NANOG-mediated inhibition (n = 3, *p < 0.05). Promoter activity was increased by
deletion of the promoter segment containing critical cis-element(s).
(C) Mutagenesis of NANOG binding sites (-1078 and -790) stimulated Cox6a2 promoter
activity (n = 3, *p < 0.05).
14

2.2.3 NANOG promotes mitochondrial FAO
Because silencing of NANOG increased OXPHOS levels in TICs (Fig. 2-2C), we
reasoned that NANOG should activate alternative catabolic pathways in order to meet
the cellular energy demands. Based on our gene ontology analysis of the
Nanog-ChIP-seq data, the lipid metabolism pathways appeared to be a critical property
of TICs (Fig. 2-1C and 2-5B), which suggested that fatty acids were an alternative
energy source (Fig. 2-5A).
Indeed, analysis of gene expression by qRT-PCR showed that expression of genes
associated with the FAO pathway (i.e., Acadvl, Echs1, and Acads) was significantly
increased in TICs; Nanog knockdown in TICs down-regulated transcription of
FAO-associated genes (Fig. 2-5C).
Figure 2-5. Nanog promoted fatty acid oxidation.
(A) Hypothetical model of NANOG-mediated fatty acid oxidation.
(B) NANOG ChIP-seq analysis revealed that FAO genes were NANOG-regulated genes
(i.e., Acadvl).
(C) qRT-PCR analysis of NANOG-target FAO genes in sh-Nanog or
Nanog-overexpressing TICs (n = 3, *p < 0.05).
15

To test whether NANOG directly regulated FAO gene expression, we focused our
analysis on Acadvl in TICs because a significant enrichment of NANOG in the Acadvl
promoter was observed by ChIP-seq (Fig. 2-5B), which was further validated by
ChIP-qPCR analyses (Fig. 2-6A). In luciferase promoter reporter assays, the full-length
Acadvl promoter (nt -1067 to +1) and a truncated version (nt -607 to +1) showed
significant decreases in activity when NANOG was silenced in TICs (Fig. 2-6B). In
addition, mutations in two of the three NANOG binding sites further reduced Acadvl
promoter activity (Fig. 2-6C), which indicates that NANOG binding sites proximal to the
transcription start site were essential for Acadvl transactivation.

16

Figure 2-6. Nanog directly regulated expression of the FAO gene, Avadvl.
(A) ChIP-qPCR of NANOG in the Acadvl promoter in TICs (n = 3, *p < 0.05).
(B) Acadvl promoter luciferase constructs were used to map the region responsive to
NANOG-mediated inhibition. (n = 3, *p < 0.05).
(C) Mutations of NANOG binding sites in the Acadvl promoter reduced its activity.
17

2.2.4 NANOG orchestrates mitochondrial metabolic reprogramming

Because NANOG activated FAO activity, we investigated the effect of NANOG on
fatty acid synthesis (Fig. 2-7A). ChIP-seq showed that NANOG was enriched in the Acly
promoter (Fig. 2-7B). Moreover, fatty acid synthesis genes (Scd1, Fasn, and Acly) were
down-regulated in Nanog-silenced TICs compared to control-scrambled TICs (Fig. 2-7C),
which suggests that NANOG may suppress fatty acid synthesis. Indeed, based on the
fatty acid elongation profile determined by GC/MS, the rate of fatty acid elongation [ratio
of oleate/palmitoleate (C18:1/C16:1 fatty acids)] was significantly increased in
Nanog-silenced TICs (Fig. 2-7D), which indicates that NANOG inhibited fatty acid
elongation.
To better profile the fatty acid substrates in TICs, we conducted metabolomic
analysis and found that when compared to normal hepatocytes, where fatty acid
synthesis and elongation had taken place, the TICs contained more short chain fatty
acids and less long-chain fatty acids (Fig. 2-7E). These data indicated that TICs differed
from normal hepatocytes in their ability to metabolize long-chain fatty acids; in fact, TICs
may be more efficient at metabolizing longer chain fatty acids (>22). From these studies,
we elucidated the causal roles of NANOG in mitochondrial metabolic reprogramming
through activation of FAO, which was found to be required for self-renewal of TICs.

18

Figure 2-7. Nanog suppressed fatty acid synthesis.
(A) Schematic diagram of the proposed role of NANOG in mitochondrial metabolic
reprogramming.
(B) NANOG ChIP-seq analysis revealed that FAO elongation genes (i.e., Acly) were
NANOG-regulated genes.
(C) qRT-PCR analysis of representative genes associated with fatty acid synthesis (n = 3,
*p < 0.05).
19

(D) Rate of fatty acid elongation was affected in Nanog-silenced TICs, as measured
using GC-MS with stable isotope
14
C. The relative ratio of C18:1/C16:1
(oleate/palmitoleate) was determined from measured levels (n = 3, *p < 0.05).
(E) Abnormal reduction of unsaturated long-chain or polyunsaturated fatty acids (PUFA)
in TICs compared to normal hepatocytes. (n = 5 per group).

2.2.5 NANOG orchestrates TIC metabolic reprogramming via the AMPK pathway
Next, we investigated how NANOG induced the switch in metabolism pathways.
Depletion of ATP or an increase in the ratio of AMP/ATP due to suppression of OXPHOS
activity led to activation of the AMPK pathway (Hardie, 2011). Phosphorylation by AMPK
led to inhibition of the fatty acid synthesis enzyme, Acetyl-CoA carboxylase (ACC), and
activation of FAO (Hardie and Pan, 2002) (Fig. 2-8A). We found that the AMP/ATP ratio
was significantly increased compared to normal hepatocytes. Moreover, the AMP/ATP
ratio was decreased in Nanog-silenced TICs compared to control-scrambled TICs (Fig.
2-8B). Furthermore, the level of phospho-AMPKα was reduced in TICs following Nanog
silencing (Fig. 2-8C), which suggested that NANOG inhibition of OXPHOS promoted the
accumulation of AMP and led to activation of AMPKα via phosphorylation (Hawley et al.,
1996). Immunofluorescence staining consistently showed increased phosphorylated
AMPKα levels in the tumor (Fig. 2-8D).
20

Figure 2-8. Nanog metabolically reprogrammed TICs through AMPK pathway.
(A) Schematic diagram shows Nanog metabolically reprogrammed TIC through AMPK
pathway.
(B) TICs had a higher AMP/ATP ratio compared to normal hepatocytes and
Nanog-silenced TICs (n = 3, *p < 0.05).
(C) sh-Nanog treatment of TICs affected phosphorylation of AMPKa and AMPKb
associated with ACC phosphorylation.
(D) Phospho-AMPK level was increased in human tumor tissues.

21

2.2.6 NANOG prevents mitochondrial ROS production and maintains self-renewal
ability of TICs

We next evaluated the effect of ROS, a major by-product of the mitochondrial
respiratory chain, on TICs. Interestingly, we observed that more oxidatively active
mitochondria were present in sh-Nanog TICs (Fig. 2-9B) with no difference in total
mitochondria mass (Fig. 2-9A), indicating that NANOG suppressed mitochondrial ROS
production.
Moreover, treatment of TICs with Paraquat, an inducer of ROS, significantly reduced
tumor spheroid formation by TICs, which could be rescued by treatment with the ROS
scavenger, N-acetyl-L-cysteine (NAC) (Fig. 2-9C). These results suggested that NANOG
suppressed OXPHOS to prevent ROS production and maintain the self-renewal ability of
TICs.
To determine whether restoration of OXPHOS gene expression inhibited
self-renewal ability, we overexpressed Cox6a2 in TICs (Fig. 2-10A) and found that the
colony number was significantly reduced (Fig. 2-10B) and the level of ROS was
significantly increased (Fig. 2-10C).
In contrast, suppression of FAO function via silencing of the FAO gene (Fig. 2-10D),
Acadyl significantly reduced colony numbers (Fig. 2-10E) and increased ROS production
in TICs (Fig. 2-10F). These results indicate that NANOG suppresses OXPHOS and
alternatively activates FAO to prevent ROS production and maintain the self-renewal
ability of TIC.

22

Figure 2-9. Nanog prevented ROS production in TICs.
(A) FACS analysis and mitochondrial mass staining was unchanged in TICs compared to
sh-Nanog TICs. However, Nanog-silenced TICs had higher oxidized mitochondrial mass
and levels of ROS (B).
(C) The ROS inducer Paraquat (1mM Para), but not the ROS scavenger (1mM NAC),
inhibited tumor spheroid formation (n = 6, *p < 0.05).
23

Figure 2-10. Nanog regulated expression of OXPHOS and FAO genes to prevent
ROS production by TICs.
(A-C) Overexpression of the OXPHOS gene, Cox6a2, in TICs led to inhibition of
self-renewal ability (B) and ROS production (C).
(D-F) Silencing of the FAO gene, Acadvl, in TICs led to inhibition of self-renewal ability (E)
and ROS production (F).

24

2.2.7 NANOG orchestrates oncogenicity and therapeutic resistance via
mitochondrial metabolic reprogramming

Finally, to address the effects of alteration of OXPHOS/FAO gene expression on
oncogenicity and drug resistance, we tested the roles of the NANOG-repressed
OXPHOS gene (Cox6a2) or the FAO inhibitor, ETO, on the efficacy of the conventional
chemotherapeutic agent, sorafenib (Llovet and Bruix, 2008).
We restored the function of OXPHOS by overexpression of COX6A2 and/or employed
ETO as an inhibitor of FAO in TICs to assess their effects on cellular sorafenib sensitivity
in an orthotopic tumor transplantation model using both mouse and human TICs (Fig.
2-11A). As shown in Fig. 2-11A, restoration of OXPHOS function and/or inhibition of FAO
activation successfully rendered TICs susceptible to sorafenib treatment. Furthermore, to
test whether restoration of OXPHOS and/or inhibition of FAO promoted
sorafenib-mediated apoptosis through the mitochondrial-pathway, cytochrome c release
was examined in the mitochondria-enriched, heavy membrane fraction (HM) of total cell
extracts (Fig. 2-11B). Upon sorafenib treatment, cytochrome c was translocated from
mitochondria into the soluble fraction (cytoplasm) of hepatocytes within 1-3 hours post
treatment, while cytochrome c in TICs remained mostly in the HM fraction (mitochondria)
(Fig. 2-11B, left panel). In addition, silencing of Nanog (Fig. 2-11B, middle panel) or
overexpression of Cox6a2 and/or addition of the FAO inhibitor (ETO) (Fig. 2-11B, right
panel) enhanced cytochrome c release from mitochondria in response to sorafenib
treatment. These results demonstrated causal roles of NANOG-mediated repression of
OXPHOS and induction of FAO gene expression in chemoresistance.

25

Figure 2-11. Suppression of tumor growth and drug resistance through metabolic
regulation.
(A) Overexpression of Cox6a2 and ETO treatment abrogated drug resistance and
reduced tumor growth (*p < 0.05).
(B) Cytochrome c release from mitochondria was analyzed by immunoblotting of the
cytosol (soluble fraction: S) and mitochondria-rich (heavy membrane: HM) fractions of
the cell lysates. Mitochondrial cytochrome c release was increased by the silencing of
Nanog (middle panel) or combination of sorafenib and ETO treatment and
overexpression of COX6A2 in TICs (right panel). The fractions were then probed for
cytochrome c (Cyt c), VDAC1, and Cu/Zn SOD.
26

2.3 Discussion

During HCC development, environmental conditions such as hypoxia and
accumulation of lipid droplets (steatosis) may trigger cellular metabolic reprograming by
inhibition of mitochondrial respiration and activation of FAO. Because FAO-dependent
NADPH production promotes survival of leukemia cells (Caro et al., 2012; Samudio et al.,
2010), the concept of targeting FAO for intervention is of high therapeutic relevance
(Valent et al., 2012). In this study, we showed that Acadvl was repressed by NANOG in
TICs. Interestingly, it was shown that Acadvl knock-out mice have reduced FAO activity
and exhibit mitochondrial dysfunction, leading to hepatic steatosis, diacylglycerol
accumulation, and hepatic insulin resistance (Aoyama et al., 1995; Kurtz et al., 1998;
Zhang et al., 2007; Zhang et al., 2003). Moreover, the mitochondrial proteins BCL2 and
truncated-Bid (tBID) inhibit FAO via the fatty acid transporter CPT1 (Giordano et al., 2005;
Paumen et al., 1997). Inhibition of FAO facilitates oligomerization of proapoptotic
mitochondrial BCL2 family member molecules BAK and BAX, leading to apoptotic cell
death (Samudio et al., 2010). Moreover, it has been shown that leukemia-initiating cells
(LICs) rely on activation of FAO for their maintenance and function (Samudio et al., 2010).
Thus, it is possible that the fate of stem cells is metabolically switched by FAO (Ito et al.,
2012). Potential mechanisms by which elevation of FAO levels maintains self-renewal
ability include: (i) shunting of long-chain FA away from lipid and cell membrane synthesis;
(ii) down-regulation of ROS through production of NADPH to avoid loss of TICs; and (iii)
reduction of metabolic resistance to chemotherapy. By these criteria, NANOG function
could serve as a gatekeeper for FAO activity.
ROS play a critical role in maintaining of self-renewal ability and determining the fate
of stem cells. It has been shown that ROS induces differentiation in mouse and human
27

embryonic stem cells (Sauer et al., 2008 and Ji et al., 2010). Moreover, maintaining
certain levels of ROS level is critical for postnatal stem cell survival, such as
hematopoietic (Ludin et al., 2014) or lung stem cells (Paul et al., 2014). It has been shown
that an increase in endogenous ROS leads to DNA damage resulting in mutations, and
thus plays a vital role in tumorigenesis (Waris and Ahsan, 2006). Here we show that
similar to normal stem cells, the level of ROS plays a critical role in maintaining the
self-renewal ability of TICs.
In conclusion, we have elucidated several mechanisms of action of NANOG in the
maintenance and chemotherapy resistance of TICs involving not only the direct activation
of self-renewal via stemness genes. We also revealed the subsequent metabolic
reprogramming in these cells leading to amplification of TIC oncogenic activity and their
overall survival. Our data showed that NANOG reprogramming of mitochondrial
metabolism was indeed responsible for human TIC oncogenicity and chemoresistance.
The metabolic basis of altered cell functions and cell fate in TICs define potentially new
approaches for chemosensitization and elimination of TICs. Our findings have shifted the
paradigm in our understanding of the underlying basis of alcohol/HCV-associated cancer,
thus facilitating future development of new, personalized treatment strategies targeted
toward NANOG+ TICs arising from obesity, alcohol, or HCV-related HCC.

28

Chapter Three: Targeting the TIC population via epigenetic regulation

3.1 Introduction
Small molecule screening to identify agents that target TICs is performed worldwide
to select potential drug candidates. However, the drug development process is lengthy
and only a small fraction of drugs successfully gets FDA approval for clinical treatments
(Roses, 2008). In this study, we employed an FDA-approved drug library for screening
purposes to identify drug candidates that selectively target TICs. Successful repurposing
of FDA-approved drugs would greatly shorten the development cycle required for clinical
application compared to de novo drug development. If identified, such drug(s) could work
synergistically or in combination with current HCC treatment regimens. To find
compounds with minimum cytotoxicity and maximum anti-NANOG activity, we performed
the screen using three approaches: 1) a CD133 cell viability-based assay, 2) a NANOG
promoter-based activity assay, and 3) a combination screening method. We expect these
approaches will identify and characterize other NANOG-dependent mechanisms
underlying TIC chemoresistance compared to non-tumor cells.

3.2 Results
3.2.1 Identification of FDA-approved drug(s) that specifically target tumor-initiating
stem-like cells
To identify drugs that targeted the TIC population, we performed three different
screening methods: (1) a CD133 cell viability screening, (2) a NANOG-Green
Fluorescent Protein (GFP) reporter screen, and (3) a combination screening.
First, to select candidates that specifically inhibited the growth of CD133 (+) TICs,
we utilized Huh7, which is a human HCC cell line of which approximately 50–60% of the
cell population constitutively expresses CD133 (Fig. 3-1A). CD133 (+) and CD133 (-)
29

cells freshly sorted by either fluorescence-activated cell sorting (FACS) or magnetic
bead-associated cell sorting (MACS) were plated into individual wells of a 96-well plate
and compounds were assayed in duplicate for effect on cell viability after 48-hours
treatment. Among the 640 compounds tested, most exhibited a similar growth inhibitory
effect for both CD133 (+) and CD133 (-) cell populations (R
2
=0.80) (Fig. 3-1B). However,
two compounds showed a selective inhibitory effect on the CD133 (+) but not the CD133
(-) cell population. ATRA inhibited cell growth with resulting cell viabilities of 41.4% and
72%for CD133 (+) and CD133 (-) cells, respectively. Another drug, acitretin, a
second-generation retinoic acid derivative, resulted in cell viabilities of 12% and 70.4%
for CD133 (+) and CD133 (-) cells, respectively (Fig. 3-1C). Next, the IC50s for these two
compounds were determined to be 4.06 µg/ml (13.5 µM) for ATRA and 7.37 µg/ml (22.58
µM) for acitretin (Fig. 3-1D). In consideration of the lower pharmacologically deliverable
dose, we used ATRA for the subsequent study.

30

Figure 3-1. Identification of selective inhibitors for CD133 (+) cells.
(A) For CD133 cell viability screening, the human HCC cell line, Huh7, was freshly sorted
into CD133 (+) and (-) cells for use in drug screening. The Huh7 cell line consistently has
50-60% CD133 (+) cells.
(B-D) For CD133 cell viability screening, most tested compounds showed similar effects
on CD133 (+) and CD133 (-) cells (R
2
=0.8), while two compounds (red dots) showed
specific growth inhibition of CD133 (+), but not CD133 (-) cells. One compound was
all-trans retinoic acid (ATRA) and the other was a second generation retinoic acid,
acitretin (C, D) (n=3, *p<0.05).
Next, to target the TIC population, we generated a NANOG-GFP reporter cell line
using TICs derived from mouse liver tumors. The lentivirus NANOG-GFP vector was
transduced into TICs and followed with antibiotic selection. To characterize this reporter
cell line, the GFP-high (top 20%) and -low (bottom 20%) populations were sorted by
FACS (Fig. 3-2A, left). qRT-PCR data revealed that Nanog expression in the GFP-high
population was 2-fold higher than in the GFP-low population (Fig. 3-2A, right).
31

Figure 3-2. Identification of selective Nanog inhibitors.
(A) For Nanog-GFP screening, we established a reporter cell line by transducing lentiviral
Nanog-GFP reporter into TICs. After antibiotic selection, the reporter cells were sorted
into high (top 20%) and low (bottom 20%). Right panel-The GFP
high
population had
higher levels of Nanog staining compared to the low population (n=3, *p<0.05).
(B) Nanog expression was further confirmed by immunofluorescence staining with a
Nanog-RFP antibody.
(C) Z-score distribution of drug library candidates. Candidates selected for repression of
Nanog had a z-score < -1.0.
(D) The hits selected from Nanog-GFP screening showed down-regulation of Nanog
gene expression (bottom panel).
32

In addition, immunofluorescence staining was conducted to confirm that the GFP
signal accurately represented Nanog expression (Fig. 3-2B). These data indicated that
the NANOG-GFP cell line was reliable for drug screening.
Freshly sorted NANOG-GFP-positive cells were plated into individual wells in a
96-well culture plate. Once the cells attached to the plate, each drug in the
aforementioned drug library was added to individual wells at a final concentration of 20
µg/ml in duplicate. The cells were fixed and stained with DAPI
(4',6-diamidino-2-phenylindole) after 12 h of incubation, and the GFP and DAPI signals
were read using a high-content screening reader. The criterion used for positive drug
candidates was a z-score less than -1.0 (the average z-score of vehicle control was
2.00±1.04) (Fig. 3-2C). Using this strategy, 56 hits were selected from the primary
screening. The subsequent confirmatory GFP screening was performed, and Nanog
gene expression was confirmed by qPCR. Ninety percent of the hit patterns matched
between the GFP screening and the qPCR results (Fig. 3-2D). Among these 56 hits, 14%
of the drugs were anti-neoplastic, 12.5% were anti-inflammatory, and 10.7% were
anti-hypertensive. More interestingly, based on the mechanisms of action of these drugs,
25% of the candidates were hormone-related (i.e., agonists or antagonists for adrenergic,
prostaglandin, or angiotensin receptors), 20% were neuron signal-related (i.e., positive or
negative modulators for the N-methyl-D-aspartate receptor, dopamine receptor, or
acetylcholine receptors), and 10% were regulators of ion-channels or proton pumps.
To increase the effectiveness of these drug candidates for elimination of the TIC
population, we combined ATRA with 56 candidate compounds from the NANOG
screening. Of the drugs combined with ATRA, one drug efficiently eliminated viability of
various HCC cell lines and mouse TICs. This hit was the HDAC inhibitor, suberoylanilide
hydroxamic acid (SAHA) (Fig. 3-3A).
33

Figure 3-3. Schematic diagram of drug combination screenings.
(A) The CD133 hit, ATRA, was further combined with 56 hits from Nanog screening
(Upper panel). Combination of ATRA and SAHA (red square) inhibited growth of various
34

HCC cell lines (bottom panel).
(B)The combination of ATRA and SAHA showed dose-dependent growth inhibition of
various HCC cells, such as Huh7, Hep3B, HepG2, human TICs and mouse TICs.
However, normal adult stem cells (mouse mesenchymal stem cells, Ms. MSCs) were less
sensitive to this drug combination.

We further tested this drug combination to determine the optimal combined
dosage. Various concentrations of SAHA and ATRA were tested in a block format.
Furthermore, the selected optimum drug combinations were tested to see if similar
inhibitory effects were observed with other cancer cell lines. The cell lines tested were
human and mouse HCC cell lines, which included HepG2, Hep3B, and mouse TICs.
The results showed that this combination had a similar dose-response effect on
these cancer cell lines (Fig. 3-3B). To test for specific killing activity toward TICs but not
normal stem cells, we assayed the viability of normal postnatal stem cells (mouse
mesenchymal stem cells) with the combination treatment and found that this combination
did not demonstrate any toxicity over the concentration ranges tested against TICs.
Cytotoxicity was observed only at very high combination dosages, which indicated that
this drug combination showed high specificity for the TIC population but spared the
normal stem cell population.

3.2.2 The ATRA-SAHA combination induces cell apoptosis pathways and reduces
the self-renewal ability of TICs in vitro.
The mechanism of cell killing exhibited by the ATRA-SAHA drug combination was
investigated. To determine whether this drug combination induced TIC apoptosis, we first
examined the occurrence of apoptosis by Annexin V-propidium iodide (PI) staining.
Indeed, the drug combination induced TIC apoptosis following treatment for 8 hours (Fig.
35

3-4A). The apoptotic process was further assayed for caspase activity originating from
the extrinsic (death receptor) and intrinsic (mitochondrial) pathways.

Figure 3-4. The ATRA-SAHA combination induces TIC apoptosis.
(A) The drug combination treatment (5 μg/ml of ATRA with 0.5 μg/ml of SAHA) induced
cell apoptosis as determined by Annexin V/PI double staining at each time point assayed
(0, 8, 16, and 24 hrs). We observed Annexin V-positive staining as early as 8 hrs after
treatment.
(B) To determine the drug combination treatment that best induced apoptosis, we
examined caspase activities of the extrinsic (death receptor; caspase-8) and intrinsic
36

(mitochondrial; caspase-9) pathways and found that the drug combination induced both
pathways and further activated caspase-3 (C).

The activities of caspase-3, -8, and -9 were tested at 6, 12, 16, and 24 hour time
points. Interestingly, we found the drug combination not only significantly activated the
extrinsic caspase-8 pathway, but also the intrinsic caspase-9 pathway at the early time
points (6 and 12 hours; p<0.05) (Fig. 3-4B and 3-4C). The combination with SAHA did
not have any other effect on cell apoptosis, suggesting that in this drug combination,
ATRA played the major role in the induction of cell apoptosis in the TIC population.
Figure 3-5. Combination treatment inhibited the self-renewal ability of TICs.
(A-B) To determine whether the drug combination treatment suppressed the self-renewal
ability of TIC, we performed a tumor spheroid assay (A) and anchorage-independent
colony formation assay (B) and found that the drug combination treatment reduced
colony number significantly (n=3, * p< 0.05). (C: Control, R: ATRA, S: SAHA, RS:
combination treatment).
37

Self-renewal and survival of TICs are the major issues regarding tumor recurrence.
We tested whether this drug combination affected the self-renewal ability of TICs, as
assessed by tumor spheroid formation assay. As Figure 3-5A shows, colony numbers in
the ATRA-only treated group were reduced by 83% in the CD133 (+) group, and in the
SAHA-only treated group, were reduced by 66%. In contrast, colony numbers in the
ATRA-SAHA group were reduced by 93.6% (p<0.05). Subsequently, we examined the
anchorage-independent growth of TICs in soft agar. The results of this assay showed that
the ATRA-only treated group had 50% fewer colonies, which was similar to the
SAHA-only treated group; however, the combination of ATRA and SAHA reduced colony
numbers by 95% (Fig. 3-5B). These results indicate that the combination of ATRA and
SAHA efficiently inhibits the self-renewal ability of TICs and their tumor forming ability.

3.2.3 Genome-wide transcriptome analysis revealed the mechanism for
ATRA-SAHA combination targeting of TICs.
To understand the mechanistic basis for the proapoptotic property of the
ATRA-SAHA drug combination, we conducted whole-transcriptome next-generation
sequencing (RNA-seq) following drug treatment (Fig. 3-6A). Principal Component
Analysis (PCA) of RNA-seq data showed that the RNA profile of ATRA-treated cells was
relatively similar to the control group while the gene expression pattern of SAHA-only
treated cells or combination-treated cells was different from the control group (Fig. 3-6B).
Only 189 genes were differentially expressed between treated cells and controls (Fig.
3-6C).
As expected, the subset of affected genes (66 genes) following ATRA treatment was
related to retinoid pathways (Fig. 3-7A and 3-7B, left panel). In contrast, the pattern of
gene expression of SAHA treatment was quite different from the control group (Fig. 3-7A
and 3-7B, right panel).
38

Figure 3-6. Genome-wide transcriptome analysis of drug treated TICs.
(A) To comprehensively illustrate the regulation network of drug treatment on TICs, we
performed RNA sequencing of TICs treated singly with ATRA or SAHA or in combination.
Gene expression heatmap (upper panel) and gene expression profile (lower panel).
(B) Principal component analysis (PCA) of RNA sequencing data showed that the gene
expression pattern of ATRA treatment only (red) was relatively similar to control (purple).
However, the gene expression pattern of SAHA-only treated cells (green) or combination
treatment (blue) was different from the control group.
(C) Differential gene expression in the three treatment groups presented as a Venn
diagram. This result revealed unique sets of genes affected by each drug combination,
which may play a critical role in regulating TICs.

39

Figure 3-7. Transcriptome analysis of individual drug treatment of TICs.
(A) Gene set enrichment analysis for ATRA and SAHA treatment only showed that each
drug alone activated the specific downstream target genes.
(B) Ingenuity Pathway Analysis for the unique set of genes in ATRA treatment-only group
and SAHA treatment-only group.

More interestingly, the gene set enrichment analysis (GSEA) showed that the stem cell
up-regulated gene set was highly enriched in the control group, but not in the drug
combination group (Fig. 3-8A), which was consistent with the IPA result (Fig. 3-8C). In
contrast, the apoptosis regulatory gene set, which includes caspase activation,
death-association protein kinase (DAPK), and protein ubiquitination pathways, was
highly enriched in the drug combination group (Fig. 3-8B). These results corroborated
the cell growth studies showing that this drug combination inhibited the self-renewal
ability and induced apoptosis of TICs.
40

Figure 3-8. The drug combination suppressed embryonic stem cell pluripotency
pathways and induced apoptosis pathway.
(A) GSEA analysis showed that the stem cell up-regulated gene set was highly enriched
in the control group, but not in the drug combination group, which indicated that the
combination treatment inhibited the stemness of TICs.
(B) GSEA analysis showed the regulation of apoptosis gene set was highly enriched in
the drug combination group, which indicated the combination treatment induced TIC
apoptosis.
41

(C) Ingenuity Pathway Analysis showed the combination treatment reduced embryonic
stem cell pluripotency pathways.

Figure 3-9. Ingenuity Pathway Analysis of the unique set of gene in the
combination treatment.
(A) Heatmap of gene expression patterns showed unique sets of genes activated in
response to individual drug treatments and dual drug combination.
42

(B) Ingenuity Pathway Analysis of the unique set of genes activated in the drug
combination group showed that these genes were highly associated with cancer
pathways and DNA repair pathways.

3.2.4 ATRA-SAHA combination treatment targeted the TIC population via
suppression of miR-22
Based on the RNA-seq data, a unique set of genes (595 genes) was differentially
expressed in the drug combination group versus the other three groups (Fig. 3-9A). IPA
indicated this unique subset of genes was associated with solid cancer pathways (such
as hereditary breast cancer) and with non-solid cancer pathways (such as acute myeloid
leukemia) (Fig. 3-9B). In addition, this unique subset of genes was related to DNA repair
signaling, such as nucleotide excision repair and DNA double strand break repair by
homologous recombination. These data suggest that the candidate genes in the unique
gene subset expressed upon drug combination treatment promoted TIC proliferation.
Notably, according to the RNA sequencing data, the unique set of affected genes in
the drug combination group included non-coding microRNA miR-22 and mirR-22 host
genes (miR-22hg) (Fig. 3-10A upper panel) and their corresponding target genes (Fig.
3-10A bottom panel). Recent evidence suggested that miRNAs regulate DNA damage
and repair (Tessitore et al., 2014). Therefore, we reasoned that any of the differentially
expressed miRNA transcripts in the pool of 595 genes could be a potential candidate to
explain the cause behind defective self-renewal in TICs post drug combination treatment.
To determine if miR-22 promoted self-renewal of TICs, we knocked down miR-22hg in
TICs. Silencing of miR-22hg reduced cell growth (Fig. 3-10B) and Nanog expression in
TICs (Fig. 3-10C). Furthermore, when we performed tumor spheroid and colony
formation assays with the miR-22hg-silenced cells, the colony numbers were significantly
reduced (Fig. 3-10D and 3-10E). These results indicate that down-regulation of
43

miR-22hg is important for suppressing TIC self-renewal.

Figure 3-10. Drug combination down-regulated miR-22 to suppress the
self-renewal of TICs.
(A) RNA-seq histograms showed that miR-22 and miR-22 host gene (miR-22hg) were
down-regulated (upper panel) along with their downstream target genes (bottom panel)
by the drug combination.
(B) Silencing by shmiR-22hg significantly reduced TIC growth (n=3, *p< 0.05).
(C) Silencing of miR-22hg reduced Nanog expression significantly (n=3, *p< 0.05).
(D) Silencing by shmiR-22hg significantly reduced the number of colonies appearing in
soft agar assay (n=4, ***p< 0.001).
(E) Silencing by shmiR-22hg reduced the number of sphere significantly (n=3, *p< 0.05).

44

3.2.5 ATRA-SAHA treatment induced TIC growth arrest and apoptosis via the
PTEN-FOXO pathway
Next, we examined the gene network(s) regulated by the drug combination
treatment. When we compared the candidate gene pathways between the three drug
treatment groups and untreated cells (Fig. 3-11A), TLR pathway was down-regulated in
drug combination group (Fig. 3-11B), which we previously showed to play critical roles in
oncogenesis and maintenance of the TIC population (Chen et al., 2013 and Lim et al.,
2007).
Figure 3-11. Drug combination regulated the key pathways of TICs.
(A) Heatmap of pathway comparison among ATRA, SAHA and combination (R+S)
groups (top ranking).
45

(B) IPA analysis showed that the drug combination treatment suppressed the Toll-like
receptor signaling; we previously showed that this pathway plays a vital role in TICs.
(C) IPA analysis showed that the combination treatment induced the PTEN pathway.

Specifically, we observed that the drug combination activated PTEN (Phosphatase
and tensin homolog deleted on chromosome 10) signaling (Fig. 3-11C). PTEN is a tumor
suppressor that regulates cell growth and apoptosis through the PTEN-FOXO pathway
(Song et al., 2012). In silico analysis by Oncomine showed that PTEN and its
downstream signals, FOXO1 and FOXO3, were down-regulated in two independent HCC
libraries (TCGA liver library and Guichard liver library) (Fig. 3-12A and Fig. 3-12B).

Figure 3-12. The drug combination activated PTEN, which is down-regulated in
HCC patients.
(A) Oncomine in silico analysis showed PTEN and its downstream signals (B) were
suppressed in HCC patients.
(C) PTEN promoter luciferase activity assay showed the combination treatment induced
PTEN prompter activity (n=3, ***p <0.01).
46

It was shown that miR-22 down-regulated PTEN via targeting the 3’-UTR (Bar and
Dikstein, 2010). The PTEN expression following drug combination treatment was
examined with a translation reporter for PTEN. We observed that PTEN 3’-UTR
luciferase reporter activity was increased in response to SAHA treatment, but not ATRA
(Fig. 3-12C). This result indicates that PTEN was subject to post-transcriptional
regulation.
Figure 3-13. The combination treatment activated PTEN-FOXO pathway.
(A) The drug combination induced PTEN-FOXOs through suppressed phosphorylation
(T308).
47

(B) Activation of FOXOs by the drug combination treatment induced cyclin-dependent
kinase inhibitors, p15
INK4b
, p19
INK4d
, p21
Cip1
, and p27
Kip1
, which led to suppression of
cyclins (Cyclin E and Cyclin D1) and cyclin-dependent kinases (CDK2).
(C) Activation of FOXO induced by the drug combination activates the BIM apoptosis
pathway.

The activation of PTEN by drug combination treatment reduced AKT
phosphorylation of Thr-308, which led to overexpression of FOXO1/3/4 (Fig. 3-13A). The
FOXO family not only regulates the cell cycle through CDK inhibitors (i.e. p15
INK4b
,
p19
INK4d
, p21
Cip1
and p27
Kip1
) (Katayama K et al., 2007), but also activates apoptosis
through transactivation of the BIM pathway (Fu Z and Tindall DJ, 2008). We observed
that the drug combination treatment induced expression of CDK inhibitors (p15
INK4b
,
p19
INK4d
, p21
Cip1
and p27
Kip1
), which led to reduction of cyclins (Cyclin D1 and Cyclin E)
and cyclin-dependent kinases (CDK2) (Fig. 3-13B). The drug combination treatment also
induced the expression of BIM, BAX, and cytochrome c (Fig. 3-13C). Thus, these results
demonstrated that the drug combination treatment induced TIC growth arrest and
apoptosis through the PTEN-FOXO pathway.

3.2.6 The ATRA+SAHA combination treatment alters the DNA methylation pattern
of Nanog via regulation of miR-22 and TET2
Up-regulation of microRNA 22 promoted tumor metastasis by directly
down-regulating members of the TET gene family, which are methylcytosine
dioxygenases (Song et al., 2013). In silico Oncomine analysis showed that TET2 was
down-regulated in two independent HCC libraries (TCGA liver library and Guichard liver
library) (Fig. 3-14A). We found that Tet2 gene expression was up-regulated after
treatment with the drug combination (Fig. 3-14B). This post-transcriptional regulation
48

was confirmed by employing luciferase reporter genes fused to the TET2a or TET2B
3’-UTR. As shown in Figure 3-14C, luciferase reporter activity was elevated after the
drug combination treatment. These data indicate that the dual drug combination
up-regulates TET2 by repressing miR-22 expression.
Figure 3-14. Suppression of miR-22 by drug combination activated TET2.
(A) In silico Oncomine analysis showed that TET2 was up-regulated in HCC patients.
(B) Drug combination induced Tet2 gene expression (n=3, *p< 0.05).
(C) Drug combination treatment activated the TET2 3’UTR luciferase reporter (n=3, *p<
0.05).
49

Figure 3-15. Drug combination activated epigenetic regulators, TET2 and DNMT3A.
(A) Western blot data showed that drug combination treatment induced TET2, DNMT3A,
and p53 expression. In contrast, the combination treatment reduced OCT4 and p53
suppressor, SIRT1 expression.
(B) Immuno-fluorescent staining shows drug combination reduced Nanog expression,
but induced p53, TET2, and DNMT3A expression.

Because members of the TET family are methylcytosine dioxygenases, we further
investigated whether or not the promoter pattern of DNA methylation is altered after drug
treatment by DNA bisulfite sequencing, especially on the Nanog promoter region.
Two important transcription factor binding sites, Oct4 (-285) and p53 (-790), are
located on the Nanog promoter region and regulate Nanog expression in contrasting
ways. OCT4 is recruited to the Nanog promoter to activate Nanog (Boyer et al., 2005,
Rodda et al., 2005, Lo, 2008, van den Berg et al., 2010), whereas p53 is recruited to the
Nanog promoter to suppress Nanog expression (Meletis et al., 2006, Pan et al., 2007 and
Han et al., 2008). We first examined TICs for changes in OCT4 and p53 levels following
50

drug treatment and found that the drug combination reduced mRNA levels of OCT4 and
SIRT1 (Fig. 3-15A), the latter is a suppressor of p53 (Li et al., 2012). By contrast, the
combination treatment induced TET2, DNMT3A, and p53 up-regulation (Fig. 3-15A and
3-15B).
Figure 3-16. Synergistic interaction of TET2 and DNMT3A induced by the drug
combination treatment altered the DNA methylation pattern of the Nanog promoter.
(A) Bisulfite sequencing of the Nanog promoter in the presence or absence of ATRA
and/or SAHA treatment. In TICs, the p53 binding site of Nanog promoter was highly
methylated; however, the OCT4 binding site was less methylated. Drug combination
treatment reduced the methylation of the p53 binding site of Nanog, but increased the
methylation of the OCT4 binding site of Nanog (* p< 0.05). C: Vehicle control, R: ATRA, S:
SAHA, RS: ATRA+SAHA combination treatment.
(B) ChIP-qPCR showed that TET2 and p53 were recruited to Nanog promoter but
DNMT3A was removed from Nanog promoter after drug combination treatment. In
51

contrast, DNMT3A was recruited to OCT4 binding site but TET2 and OCT4 were
removed from the Nanog promoter after drug combination treatment (n=3, * p< 0.05). (C:
Control, R: ATRA, S: SAHA, RS: combination treatment).

Next, we examined the methylation pattern of proximal promoter region of Nanog
and found that the p53 binding site in the Nanog promoter was highly methylated in TICs;
however, the OCT4 binding site of the Nanog promoter was less methylated in TICs
(58.3% and 90%, respectively) (Fig. 3-16A). Aafter the drug combination treatment
increased DNA methylation in the OCT4 binding site (the control and the combination
treatment were 58.3% and 79.2%, respectively), whereas methylation was decreased in
the p53 binding site (the control and the combination treatment were 90% and 58.3%,
respectively) (Fig. 3-16A).
As confirmation of the change in DNA binding activity of p53 and Oct4 to the Nanog
promoter, we performed ChIP-qPCR from TICs treated with the drug combination. Under
these conditions, we observed that p53 was recruited to the Nanog promoter region
whereas Oct4 was absent (Fig. 3-16B). Thus, the consequence of drug combination
treatment appeared to increase recruitment of two major DNA methylation regulators,
TETs and DNMTs to the Nanog promoter region with a subsequent effect on transcription
factor binding and a corresponding change in Nanog transcription. Indeed, we observed
that TET2 was detached from the Oct4 binding site and recruited to the p53 binding site
after combination treatment. On the other hand, DNMT3A was removed from the p53
binding site and recruited to the Oct4 binding site after the combination treatment. These
results demonstrated that the alteration of the DNA methylation pattern of the Nanog
promoter resulted in repression of Nanog expression.

52

3.2.7 Dual-drug combination treatment attenuated tumor growth in vivo
The efficacy of ATRA-SAHA on TIC viability in vitro prompted us to examine if the drug
combination inhibited tumor growth in vivo. For these studies, we subcutaneously
implanted 10
6
CD133 (+) Huh7 cells into NOD/Shi-scid/IL-2Rγ
null
(NOG) mice. In order to
specifically target the CD133 (+) population, we encapsulated ATRA into nanoparticles
conjugated with CD133 antibody using biodegradable poly(D,L-lactide-co-glycolide)
(PLGA) polymer.
Figure 3-17. The drug combination attenuated tumor growth.
(A) Treatment regimen designed to target CD133 (+) TICs. ATRA was encapsulated into
nanoparticles displaying a conjugated anti-CD133 antibody. Tumor growth was reduced
in the drug combination group (red), but not by single ATRA (blue) or SAHA (green)
treatment or control group (black) (C: Control, R: ATRA, S: SAHA, RS: combination
treatment).
(B) Histological analysis of tumors from each group showed that the drug combination
induced extensive cell death in the tumor.
(C) TUNEL staining of tumors from each group shows that the drug combination induced
cell apoptosis in the tumor.
53

Once a tumor size of 100 mm
3
was reached, the mice were treated with ATRA only
(5 µg/ml), SAHA only (0.5 µg/ml), or the combination with empty nanoparticles as a
control for all treatments. We observed that the single drug treatment groups did not
reduce or even promote tumor growth while the combination treatment significantly
inhibited tumor growth when compared to the single drug treatments and control groups
after 4 days of treatment (Fig. 3-17A). These data indicate that the drug combination
selectively inhibits tumor growth. The tumor morphology was examined in the control
group from hematoxylin and eosin stained tissue sections. Representative tumor tissues
sections from ATRA-treated mice were found to have necrotic regions (Fig. 3-17B). The
SAHA-treatment group did not induce cell death of tumor cells, but showed increased
vascularization, which may explain why SAHA-treated tumors had the largest tumor sizes.
The combination treatment induced extensive necrosis, indicating that the dual drug
regimen eliminates almost all tumor cells.
In order to understand the basis for cell death, TUNEL staining was performed for
each of the drug treatment groups. The tumors from the ATRA-treatment group had
TUNEL-positive tumor cells, but far less than that observed for tumors in the
SAHA-treatment group (Fig. 3-17C). The drug combination group showed a significant
increase of apoptosis, indicating that the combination of ATRA and SAHA effectively
inhibits tumor growth.
A comparison of gene expression patterns in liver cancers with overall survival was
performed using GSEA analysis. Both the liver cancer recurrence up-regulated gene set
(Fig. 3-18A) and liver cancer survival down-regulation gene set (Fig. 3-18B) are
enriched in control group, indicating that the ATRA-SAHA drug combination effectively
suppresses tumor growth and recurrence, improving the overall survival rate.
54

Figure 3-18. The combination treatment suppressed the tumor recurrence and
poor survival gene sets.
(A) GSEA analysis showed that tumor recurrence-associated gene set was highly
enriched in the control group, but not in the combination treatment group. This result
indicated the combination treatment suppressed the recrudesce-associated gene set. (B)
GSEA analysis showed that the gene set-associated with poor survival was highly
enriched in control group, but not in the combination group. These data indicate that this
drug combination treatment might improve the overall survival rate.

In conclusion, our results showed the drug combination suppressed miR-22
expression, which in turn inhibited the PTEN-regulated apoptosis pathway and
suppressed Nanog gene expression. The latter occurred through a change in the DNA
methylation pattern of the Nanog promoter itself, leading to a loss of self-renewal ability
and drug susceptibility. These results are summarized in the model shown in Fig. 3-19

55

Figure 3-19. Hypothetical Model: Combined drug treatment down-regulates miR-22,
leading to activation of PTEN-FOXO apoptosis pathway and TET-mediated
demethylation of p53-binding sites within the Nanog promoter. Specifically, TET2 is
recruited to p53-binding sites of the Nanog promoter while DNMT3A is recruited for
methylation of an OCT4 binding site within the Nanog promoter, leading to repression of
Nanog.

56

3.3 Discussion

The goal of this work was to identify drugs and drug combinations that would
specifically target the TIC population in tumor. As such, most molecular screens focus
only on one marker when assaying a large molecule library (Gupta et al., 2009); however,
the marker may not be efficient in eliminating the target population of malignant cells. In
this study, we conducted three different kinds of screens for the TIC population, including
a CD133 cell viability screen, a NANOG-GFP high-content screen, and a combination
screen.
By employing a high-throughput, TIC viability screen tested against a library of
FDA-approved drugs, we found that a retinoic acid derivative, ATRA, displayed the best
inhibition ability of cell growth. This drug was included in a refined screening approach
targeting Nanog expression in a secondary dual drug regimen with the FDA-approved
drug library. From this secondary drug screening, we found that the combination of ATRA
and SAHA demonstrated the best efficacy for inhibition of TIC growth in vitro and in vivo.
Our analysis of the mechanism by which these drugs killed tumor cells showed a bipartite
process leading to cell death. Increased TET2 suppressed Nanog expression, leading to
a change in promoter methylation and subsequent repression of Nanog transcription.
Repression of miR-22 expression initiates this pathway. As a consequence of the latter,
PTEN activity increased with a corresponding induction of the apoptosis pathway in TICs
leading to cell death. Interestingly, knockdown of miR-22 expression also sensitized cells
to killing by other chemotherapeutic agents (e.g., rapamycin). Although our favored drug
combination is ATRA + SAHA, this general strategy of repressing miR-22 may be useful
for sensitizing cancer cells to other therapeutic drugs.
Regulation of cell growth via retinoic acid signaling has been widely used to treat
57

various types of cancer, such as breast cancer (Garattini et al., 2007), lung cancer (Dahl
et al., 2000), ovarian cancer (Harant et al., 1993), prostate cancer (Zhao et al., 1999),
neuroblastoma (Reynolds et al., 2003), renal cell carcinoma (Motzer et al., 2000),
pancreatic cancer (Weiss et al., 2009), liver cancer (Meyskens et al., 1998), head and
neck cancer (Rubin Grandis et al., 1996), and acute promyelocytic leukemia (Huang et
al., 1988). Retinoic acid is also an inducer of embryonic stem cell and hematopoietic
stem cell differentiation (Simandi et al., 2010, Rochette-Egly, 2015, Chanda et al., 2013).
In HCC, induction and intracellular localization of the nerve growth factor IB (NGFIB, aka
Nur77) via Fenretinide, a structural analogue of retinoic acid, could induce cell apoptosis
through activation of caspase-3/7 (Yang et al., 2010). In our study, we demonstrated that
retinoic acid not only activated the extrinsic caspase-8 pathway, but the intrinsic
caspase-9 pathway as well.
The HDAC inhibitors are widely used in treatment of various cancers, such as
leukemia (Rosato et al., 2003), pancreatic cancer (Kumagai et al., 2007), lung cancer
(Komatsu et al., 2006), breast and colon tumors (Butler et al., 2002), ovarian cancer
(Strait et al., 2005), and cervical cancer (Li and Wu, 2004). These inhibitors have broad
effects on the regulation of the cell cycle, apoptosis, cell differentiation, and autophagy,
and they are anti-angiogenic (Khan & La Thangue, 2012). In addition, the HDAC
inhibitors can induce cell cycle arrest through the induction of p21 and down-regulation of
cyclins (Sabdor et al., 2000). Furthermore, HDAC inhibitor treatments induce
accumulation of ROS, which results in DNA damage and subsequent apoptosis
(Petruccelli et al., 2011). In this study, we showed that treatment with the HDAC inhibitor
(SAHA) alone failed to reduce cell growth in vitro or to reduce tumor growth in vivo,
strongly suggesting that the single treatment for conventional cancer therapy was not
sufficient. We showed that only the combination of the HDAC inhibitor with ATRA was
able to successfully reprogram the TIC population for cell apoptosis and suppress tumor
58

growth.
MicroRNAs (miRNAs) are small non-coding RNAs (17–22 nucleotides) = involved
in RNA silencing via translation inhibition or mRNA degradation. Increasing evidence has
revealed that miRNAs play a critical role in tumorigenicity. For example, in HCC,
miR-130b is up-regulated in CD133 (+) TICs, leading to the down-regulation of tumor
protein 53-inducible protein 1 (TP53INP1) and enhanced self-renewal (Ma et al., 2010).
Similarly, miR-155 targets TP53INP1 to regulate the self-renewal ability of liver TICs
(Chiou et al., 2015). Overexpression of miR-150 in CD133 (+) TICs led to inhibition of
self-renewal and tumor growth via interaction with the 3’UTR of c-Myb (Zhang et al.,
2012). miR-22 promotes Hepatitis B virus-related HCC development through
down-regulation of estrogen receptor alpha (ERα) transcription (Jiang et al., 2011). More
interestingly, these miRNAs often function with epigenetic regulators to alter target gene
expression. For instance, miR-22 promotes genes associated with
epithelial-to-mesenchymal transition (EMT) by directly down-regulating members of the
TET family (Song et al., 2013). The miR-29 gene family is down-regulated in lung cancer,
which directly regulates the de novo DNA methyltransferases (DNMTs) DNMT3A and
DNMT3B, and leads to aberrant DNA methylation (Fabbri et al., 2007). Recently,
miR-34b was shown to regulate DNMTs and histone deacetylases (HDACs) in prostate
cancer (Majid et al., 2013).
We found that miR-22 was down-regulated following drug combination treatment.
Moreover, TET2, the target of miR22, was up-regulated, which indicated that epigenetic
modification, especially DNA methylation, was a response to the drug combination
therapy. It is well known that this combination is widely used in acute myeloid leukemia
patients to induce leukemia cell differentiation (Salomini and Pandolfi, 2000). In human
malignant melanoma, the combination of 13-cis-retinoic acid with the HDAC inhibitor
LAQ824 induced cell growth arrest and apoptosis (Kato et al., 2007). Additionally, the
59

HDAC inhibitor DWP0016 suppressed miR-22 via p53-independent PTEN activation and
inhibited neuroblastoma cell growth (Jin et al., 2013). In cervical cancer, the combination
of retinoic acid with the HDAC inhibitor BML-210 could induce HeLa cell apoptosis
through the p53 pathway (Borutinskaite et al., 2006). In HCC, the combination of
Fenretinide with TSA, another general HDAC inhibitor, could further induce cell apoptosis
via up-regulation of Nur77 (Yang et al., 2010). However, few of these results provided any
detailed epigenetic mechanism dependent upon the combination treatment. Our data
indicated that this drug combination not only induced cell apoptosis but also inhibited the
ability of self-renewal via epigenetic regulation.
MicroRNA analogues or antagonist therapies are an emerging anti-cancer
strategy; however, these miRNA-based therapies are still in the clinical trial phase, and
the therapeutic concerns regarding dosage, stability, and safety remain unclear. In this
study, we demonstrated that the combination of the FDA-approved drugs ATRA and
SAHA could manipulate microRNA expression with improved safety control. In future
studies, we will investigate if an examination of a patient’s gene expression pattern in
TICs, especially if positive for higher miR-22/Nanog/CD133 levels, would render these
tumors more susceptible to this drug combination for treatment.

60

Chapter Four: Experimental Procedures

4.1 Experimental procedures related to Chapter 2

4.1.1 ChIP-seq sample preparation and bioinformatics analysis
Four pairs of CD133 (+) TICs and CD133 (–) control cells (~1x10
5
per mouse) were
isolated from four independent mouse liver tumors. ChIP was performed with NANOG
antibody using CD 133 (+) as well as CD133 (–) cell lines following a standard protocol
as suggested by the manufacturer (Millipore). To generate sequencing library constructs,
ChIP DNA fragments (1–10 ng) were used for adapter ligation, gel purification and PCR,
followed by ligation. ChIP-seq library constructions and high-throughput DNA sequencing
was performed using Illumina HiSeq 2000 (Illumina, San Diego, CA, USA) using a 50 bp
single end reads at the USC Genomic Core.
Approximately, 20 million reads were aligned with the mm9 reference genome using
Bowtie 2 (version 0.12.7) to generate around 18 million aligned reads with mapping
quality ≥20, allowing only two mismatches per alignment (Li and Durbin, 2009). Only
uniquely mapped reads were retained and redundant reads were filtered out. Further,
each read was extended in the sequencing orientation to a total of 200 bases to infer the
coverage at each genomic position. The genome was divided into non-overlapping
windows 200 bp, and aligned reads were considered to be within a window of the
midpoint of its estimated fragment. Mid-points in each window were counted, and
empirical distributions of windows counts were created. The genomic bins, which
contained statistically significant ChIP-seq enrichment, were identified by comparison to
a Poisson background model, assuming that background reads are spread randomly
throughout the genome.
In addition, fold enrichment was calculated in CD133 (+) cells over CD133 (–) cells.
61

The mapping output files were also converted to browser-extensible data (BED) files. For
visualization, wiggle tracks and TDF file were generated by computing mean read density
over 25 bp bins of mouse genome with aligned and filtered reads from ChIP-seq data.
Wiggle tracks were visualized in the IGV (Integrated Genomic Viewer) as well as
Seqmonk (Seqmonk v0.26.9). To assign ChIP-seq enriched regions to genes, a complete
set of Refseq genes was downloaded from the UCSC genome dataset and, genes with
enriched regions within 5 kb of their TSSs were called bound. Accession number: GSE
68237.

4.1.2 XF24 extracellular flux analyzer for measurement of cellular OCR and ECAR
To measure cellular bioenergetics using extracellular flux, a Seahorse XF24 Extracellular
Flux Analyzer (Seahorse Biosciences) was used. Cells were plated in XF 24-well cell
culture microplates at 5 × 10
4
cells per well (Seahorse Bioscience) and incubated in
unbuffered DMEM medium for overnight. Next day, cells was replaced with unbuffered
DMEM assay medium supplemented with 1 mM pyruvate and 25 mM glucose for one
hour equilibration. The oxygen consumption rate (OCR) and extracellular acidification
rate (ECAR) were measured over time at 10 min intervals. The first three measurements
were conducted to establish a baseline rate, followed by three measurements after the
addition of 2 μM Oligomycin, to determine ATP turnover and the degree of proton leakage.
Next, the maximal respiratory capacity was measured after the addition of the electron
transport chain decoupler (FCCP, 0.5 μM). Finally, we administered 5 μM
Antimycin/Rotenon to inhibit the flux of electrons through complex III and prevent oxygen
consumption by the cytochrome c oxidase in the mitochondria. Both OCR and ECAR
were determined by plotting the oxygen tension and acidification of the medium in the
chamber as a function of time and normalized to protein concentration. Absolute values of
OCR and ECAR were expressed as pmol per minute and mpH per minute, respectively.
62

4.1.3 Stable-isotope carbon labeling is traced for flux analysis
Cells were cultured in DMEM/F12 medium (17.5 mM unlabeled glucose) supplemented
with 7.5 mM [U
13
C6]-glucose (Cambridge Isotope Laboratories) for 48 hr and total ion
chromatography of fatty acids was performed by stable isotope tracing using
[U
13
C6]-glucose for 48 hr. Three independent replicates of 2 × 10
6
cells for each cell line
were collected, and the cell pellets were suspended in 0.5 ml of water and lysed by
sonication. Cell debris was separated by centrifugation and proteins precipitated by
treating the clarified supernatant with 1 ml of cold acetone. The final supernatant was
air-dried and the free glutamic acid was converted to its trifluoroacetamide butyl ester for
GC-MS analysis. Rate of fatty acid synthesis is represented by Oleate
C18:1/Palmitoleate C16:1 ratio.

4.1.4 Fatty acid β-oxidation assay
Rates of fatty acid β-oxidation were determined, in which the rate of carbon dioxide
production from the oxidation of [
14
C] palmitate was measured in Metabolomic Core
facility in MMR building. Cells were cultured in the presence of [
14
C] palmitate–BSA
complex and the released [
14
C] carbon dioxide trapped for 1 h at 37 °C onto filter paper
soaked in 100 mM sodium hydroxide. The rate of β-oxidation was calculated as the
amount of trapped [
14
C] carbon dioxide in relative units produced per mg protein per hour.

4.1.5 Mitochondria labeling and measurement of ROS levels
To evaluate the status of mitochondria in TICs, the MitoTracker® Mitochondrion-selective
probes for total mitochondrial mass (MitoTracker® Deep Red FM, Invitrogen) and for
oxidized state mitochondria (MitoSOX™ Red mitochondrial superoxide indicator,
Invitrogen) were added to the media, respectively, and cells subjected to FACS analyses.
ROS labeling was performed as per the instructions for CellROX® Oxidative stress
63

reagent Probes (CellROX® green reagent, Invitrogen C10444). In brief, the cells were
incubated with staining solution (100 nM) in culture media at 37
o
C for 30 minutes. After
staining was complete, cells were washed with PBS and

4.1.6 ATP production measurements
Relative ATP/cell assays were performed in 96-well plates. After cells were treated with
inhibitors for 4 hr, culture media was removed. Cell Titer-Glo (100 μl: Promega) and
CyQUANT(Invitrogen) were immediately added to each well. Luminescence and
fluorescence readings were consecutively measured after room-temperature incubation
for 10 min.

4.1.7 Cox6a2 and Acadvl promoter luciferase assay
The promoter regions of Cox6a2 and Acadvl were inserted into a pGL3 Firefly luciferase
reporter vector as different truncation forms. The luciferase assay was performed as per
vendor instructions (Promega). Briefly, 1 μg of pGL3 luciferase plasmid was transfected
with Fugene. A 100 ng of Renilla plasmid was co-transfected as an internal control. Cells
were harvested 24 hr after transfection, and cell-free lysates were assayed for luciferase
activity of cell lysate was measured with the dual-luciferase reporter assay kit (Promega)
using a luminometer.

4.1.8 Quantitative real-time PCR (qPCR)
Total RNA was extracted from the cells by RNeasy Mini kit (Qiagen). 1 µg of RNA was
treated with DNase I (Invitrogen) and used for reverse-transcription (Omniscript RT kit,
Qiagen). Quantitative real-time PCR was performed with Taqman Fast Advanced master
mix (Invitrogen) using ABI 7900 system (Applied BioSystems). Taqman primers and
probes for beta Actin (assay ID: Mm00607939_s1), Nanog (assay ID: Mm02384862_g1),
64

Atp6v1g2 (assay ID: Mm01159330_g1), Atp5d (assay ID: Mm00502864_m1), Atp5h
(assay ID: Mm02392026_g1), Atp8b2 (assay ID: Mm01220121_m1), Acaa2 (assay ID:
Mm00624282_m1), Cox15 (assay ID: Mm00523096_m1), Cox6a2 (assay ID:
Mm00438295_g1), Ndufs2 (assay ID: Mm00467603_g1), Ndufv2 (assay ID:
Mm01239727_m1), and Uqcrfs1 (assay ID: Mm00481849_m1) were obtained from
Applied Biosystems.

4.1.9 In vivo rescue experiments of OXPHOS gene inhibition and FAO by
implantation of TICs into immunocompromised mice
The effect of restoration of an OXPHOS gene and/or inhibition of FAO for effect on
tumorigenicity of TICs in a xenograft model was examined. Cryopreserved human TICs
obtained from liver tumors were tested for tumorigenicity in NOG mice. Prior to
implantation, these cells were expanded through several passages and infected with the
lentiviral vector expressing Cox6a2 cDNA and dsRed (as a fluorescence tracing marker
for in vivo imaging) (1 x 10
5
TU/ml: MOI 10). Ten days post-lentivirus infection, TICs (1 x
10
4
) were subcutaneously injected into 6-8-week-old NOG. Tumor growth was monitored
and palpable tumors were measured by caliper every 4 days for 44 days.

65

4.2 Experimental procedures related to Chapter 3

4.2.1 TICs isolation
The mouse or human TICs were isolated as previously described (Chen et al., 2013). In
brief, the TICs were isolated from liver tumors of HCV transgenic mice that were fed with
alcohol for 12 months. Human TICs were isolated from alcoholic patients with or without
HCV infection, as previously described (Chen et al., 2013). Fresh liver cancer tissues
were collected from the USC transplant surgery unit in collaboration with Dr. Linda Sher.
The minced tumor was digested by collagenase (Roche) with DNase I (Roche) to obtain
cell suspensions, which were washed and adjusted to a concentration of 1×10
7
cells/ml.
These cells were incubated with antibodies against CD133, CD49f, and CD45 (BD
Biosciences) and sorted by FACS to isolate CD133+CD49f+CD45- (short as CD133(+))
vs. CD133-CD49f+CD45- CD133(-) in short) populations as previously described (Parent
et al., 2004).

4.2.2 Cell culture
Huh7, HepG2, and Hep3B human HCC cell lines were cultured in DMEM (high-glucose)
medium supplemented with 10% fetal bovine serum (FBS), non-essential amino acids
(NEAA, Invitrogen), and Glutamine/Penicillin/Streptomycin (Invitrogen). The mouse TICs
grown in DMEM/F12 medium (Sigma-Aldrich) supplemented with 10% FBS,
non-essential amino acids (NEAA, Invitrogen), Glutamine/Penicillin/Streptomycin
(Invitrogen), nucleosides (Sigma), 20ng/ml mEGF (Invitrogen), and 100nM
dexamethasone (Sigma-Aldrich). Both cell lines were grown at 37°C and 5% CO2.

4.2.3 Chemical screening and analysis
The FDA approved drug library (ENZO Life BML-2841-0100) containing 640 FDA
66

approved drugs was selected to maximize chemical and pharmacological diversity. The
library included 44 different drug categories including analgesics, COX2 inhibitors, and
cholinergics.

4.2.4 Cell viability assay
CD133(+) and CD133(-) Huh7 cells were freshly sorted using the MACS CD133 micro
bead kit (Miltenyi Biotec, 130-050-801) and were seeded in 100 µl medium containing
5000 cells per well in a 96-well plate. Once the cells attached, the FDA approved drug
library was added to each well to a final concentration of 20 µg/ml in duplicate. After
incubation for 48 hours, the cell viability was determined by a luminescence assay. The
selected drug candidates had to show a significant cell growth inhibition effect on the
CD133 (+) population (percentage of cell viability less than 30%), but with no or only a
minor effect on the CD133 (–) population (percentage of cell viability greater than 70%)
compared to the vehicle control group (1% DMSO). After 16 hours, 1 µl of a 2 µg/ml drug
solution was individually added to the 96 wells, resulting in a 20 µg/ml final concentration
for most compounds. After 48 hours, CellTiter-Glo® Reagent (G8233, Promega) was
added and the luminescence signal was measured with an automated plate reader. The
raw data for each well was background-corrected by DMSO control wells on the same
plate. The selected hit compounds exhibited a marked effect on CD133 (+) cells (cell
viability <30%) and low/no effect on CD133 (–) cells (cell viability >70%).

4.2.5 Nanog-GFP screening
The Nanog-GFP liver cancer stem cell line was transduced with pGreenZeo-Nanog
transcriptional reporter lentivirus vector (System Biosciences SR10031VA-1), containing
1200-bp of Nanog promoter region. The transduced cells were positively selected with
zeomycin (10 µg/ml) and further sorted for the GFP-high population (~20% of total
67

population) for drug screening. Nanog-GFP liver cancer stem cells were then seeded in
100 µl medium containing 5000 cells per well in a 96-well plate. After 16 hours, 0.5 µl of a
2 µg/ml compound was added to each well, resulting in a 10 µg/ml final concentration for
most compounds. After 12 hours, the cells were fixed with 1% Paraformadehyde and
stained with DAPI; compounds were screened in duplicate. The GFP and DAPI images
were acquired using a BD Pathway Bioimaging Systems instrument. Z-score was
calculated from the data using the formula z = (X-u)/s.d., where u is the mean, s.d. is the
standard deviation of the whole population and X is the sample value calculated based
on the ratio of GFP intensity to DAPI intensity. The z-score of selected hits must be less
than -1.0. The average of z-score of vehicle control is 2.0±1.04.

4.2.6 Determination of combination dose
Freshly sorted CD133 (+) Huh7 or mouse TICs were plated in 100 µl of medium
containing 5000 cells per well in a 96-well plate. After 16 hours, 0.5 µl of a 2 µg/ml, 1
µg/ml, 0.2 µg/ml, 0.1 µg/ml, 0.02 µg/ml, or 0.01 µg/ml compound was added to each well
in triplicate, resulting in a 10 µg/ml, 5 µg/ml, 1 µg/ml, 0.5 µg/ml, 0.1 µg/ml, and 0.05 µg/ml
final concentration for most compounds, respectively. After 48 hours, the cells were either
measured for cell viability by Cell-Glo® Reagent (Promega), or fixed with 1% PFA
(paraformaldehyde) and stained with DAPI for high-throughput screening.

4.2.7 Annexin V staining
Annexin V staining was performed according to the manufacturer's instructions (A35110,
Invitrogen). In brief, after drug treatment, the cells were washed twice with ice-cold
phosphate-buffered saline (PBS) and detached by trypsin/EDTA. The cells were then
incubated with 5 µl of Annexin V-APC in a 100 µl of cell suspension at room temperature
for 15 minutes. After incubation, the cells were mixed with propidium iodide solution and
68

analyzed by flow cytometry.

4.2.8 Caspase activity analysis
The caspase activity assay was performed according to the manufacturer’s instructions
(G8090, G8200, and G8210, Promega). In brief, 10,000 cells were plated into each well
of a 96-well plate. After the cells attached, the drugs (5 µg/ml of ATRA and 0.5 µg/ml of
SAHA) were added to each well and incubated for 6, 12, 16, and 24 hours. After
incubation, the Caspase-Glo® substrate reagent was added to each well followed by
incubation for 30 minutes. After incubation, the luminescence signal was measured with
a luminometer.

4.2.9 TUNEL staining assay
The TUNEL staining was performed according to the manufacturer’s instructions
(4810-30-K, TREVIGEN). In brief, paraffin-embedded tumor sections from each group
were de-paraffinized, re-hydrated, and washed twice in PBS. Samples were covered with
Proteinase K solution for 30 minutes at room temperature and then washed two times in
deionized water. Slides were immersed in quenching solution for 5 minutes at room
temperature and then washed in PBS. Slides were incubated in TdT labeling buffer for 5
minutes, immersed with labeling reaction mix, and incubated at 37°C for 1 hour in a
humidity chamber. Samples were immersed in TdT stop buffer for 5 minutes and then
washed twice in deionized water for 5 minutes each at room temperature. Samples were
covered with Strep-HRP solution and incubated for 10 minutes at 37°C and washed twice
in PBS. Samples were incubated in DAB solution for 5 minutes and then washed in
deionized water several times. The samples were counterstained with Methyl Green and
mounted on slides for observation.

69

4.2.10 Tumor spheroid formation assay
Freshly sorted CD133(+)/(–) Huh7 cells were plated in a low binding culture plate (NUNC
145397) containing 100 cells per well in 100 µl of culture medium with ATRA (5 µg/ml),
SAHA (0.5 µg/ml), or a combination of both. After 2 weeks, colony numbers were
counted.

4.2.11 Anchorage-independent growth assay
Freshly sorted CD133 (+)/(–) Huh7 cells were mixed with 0.35% agarose containing 1000
cells per well in culture medium with retinoic acid (5 µg/ml), SAHA (0.5 µg/ml), or a
combination of both. After 2 weeks, the colony numbers were counted.

4.2.12 All-trans Retinoic acid nanoparticles conjugated with CD133
CD133 was conjugated to the terminal amine functionality on a polyethylene glycol block
of polylactide-polyethylene glycol (PLA-PEG) as previously described (Swaminathan et
al., 2013). In brief, 30mg of PLGA (lactide to glycolide ratio 50:50, average Mw 40KDa,
Absorbable polymers) and 6mg of all-trans retinoic acid (Enzo Life Sciences) were
dissolved in 1ml of chloroform. An oil-in-water emulsion was formed by emulsifying the
polymer-drug solution in 8ml of 2.5% (w/v) aqueous PVA (polyvinyl alcohol, Mw
30-70KDa, Sigma) solution by probe sonication for 7 minutes on ice. The 8mg of diblock
copolymer polyactide-polyethylene glycol with terminal maleimide functionalization
(PLA-PEG-MAL, Laysan Bio Inc.,) was dissolved in 200ul of chloroform and added
drop-wise to the above emulsion with stirring. The emulsion was stirred overnight
followed by 2 hours vacuum to remove the residual chloroform. Nanoparticles were then
washed three times by repeat ultracentrifugation at 35,000rpm and 4
o
C for 35 minutes
and reconstitution in deionized water. Nanoparticle dispersion was then lyophilized. The
size of nanoparticle was determined by Delsa Nano C (Beckman Coulter Inc.).
70

Before CD133 antibody conjugation, 20ug of human CD133 antibody or the IgG isotype
control antibody (Miltenyi Biotec) was diluted with 100mM sodium phosphate buffer (pH 8)
containing 150mM sodium chloride to a final volume 1ml. The antibody solution was then
mixed with 4ul of 1mM solution of 2-iminothiolane and incubate at 4
o
C for 2 hours. The
iminothiolated antibody solution was de-salted and exchanged with HEPES buffer
mixture (50mM HEPES, pH 7.4, 150mM sodium chloride, 2mM EDTA) using Zeba
desalting column (Pierce). Conjugation of PLA-PEG-MAL on nanoparticles surface was
determine by Amplite Fluorimetric maleimide assay (AAT Bioquest).
For CD133 antibody conjugation, the iminothiolated antibody solution was then
further concentrated by using ultra-4 centrifugal filter unit (Amicon) by centrifugation at
4000 rpm and 4
o
C for 30 minutes. 20ug of the iminothiolated antibody (anti-CD133 or
control IgG) was mixed with 15mg of maleimide functionalized PLGA nanoparticles
dispersed in HEPES buffer mixture and incubated at room temperature overnight in a
rotating shaker. The particles were then pelleted by ultracentrifugation at 14,000 rpm and
4
o
C for 30 minutes and washed twice with de-ionized water. Nanoparticle suspension
then lyophilized. Conjugation of CD133 on nanoparticle surface was determined by
Western Blot. Before injection, the lyophilized nanoparticles were re-dissolved in PBS
and filtered with a 0.22 micron filter.

4.2.13 In vivo tumorigenicity experiments
A half million freshly sorted CD133 (+) TICs were suspended in 100 µl of Matrigel™ (BD)
and injected subcutaneously into NOD/Shi-scid/IL-2Rγ
null
(NOG) mice, six mice per group.
After the tumor volume reached 100 mm
3
, animals received one intravenous dose of
CD133-conjugated RA nanoparticle (5 µg/ml) and/or SAHA (0.5 µg/ml) daily. The animals
were monitored regularly for tumor growth and survival every day by tail vein injection. All
animals work was performed according to national and international guidelines. Tumor
71

volumes were calculated by measuring two perpendicular diameter and using the formula
(L x W
2
) / 2, where L is the longest diameter and W is perpendicular to L. Animal studies
were based on a protocol approved by the Institutional Animal Care and Use Committee
at University of Southern California.

4.2.14 RNA sequencing
RNA sequencing samples were collected at 16 hours of treatment with ATRA (5 µg/ml),
SAHA (0.5 µg/ml) or combination treatment. Total RNA for RNA sequencing was
extracted using RNeasy Plus Mini Kit (Qiagen), which includes a DNA depletion column.
DNase I treatment and rRNA depletion with Tibozero technology were performed before
RNA sequencing. Sample quantity and quality was verified by spectrophotometry
(NanoDrop 1000), fluorimetry (Qubit), and the Aglient Bioanalyzer 2100 profiler. RNA
Integrity Number (RIN) values of >7.0 and OD260/280 =2.0-2.2 were used for RNA-seq
library preparation. Extracted RNA (1 µg) was used for RNA sequencing (Illumina
HiSeq2500 system), using a 100 bp single end reads. Sequenced one-million reads were
cleaned according to a rigorous pre-processing workflow (Trimmomatic-0.32) before
mapping them to the mouse genome (mm10) using SHRiMP2.2.3
(http://compbio.cs.toronto.edu/shrimp/ ). Cufflinks2.0.2 (cuffdiff2 - Running Cuffdiff) was
then used to perform differential expression analysis with a FDR cutoff of 0.05 (95%
confidence interval). A Perl script was used after differential expression analysis to
improve the readability of the results files. Quality control information was generated via
Fastqc: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/. The log2 (fold
change) seen in these files was such that fold change = Sample2_fpkmValue /
Sample1_fpkmValue. All work was performed by the University of Rochester Genomics
Research Center (URGRC). All gene expression profiles were analyzed by Partek Flow,
QIAGEN’s Ingenuity Pathway Analysis (IPA®,QIAGEN Redwood City,
72

www.qiagen.com/ingenuity) and Gene Set Enrichment Analysis. Accession Number:
GSE78139

4.2.15 Bisulfite sequencing
Bisulfite sequencing was performed according to the manufacturer's instructions (D5005,
Zymo Research). In brief, 2 µg of genomic DNA from each group was treated with CT
conversion reagent in the following thermal cycle: 98°C for 10 minutes, 64°C for 2.5
hours, and 4°C for storage for up to 20 hours. Converted DNA was treated with
M-Desulphonation Buffer for 20 minutes at room temperature. After desulfonation, DNA
was washed and eluted. Bisulfite-treated DNA (150 ng) was used for PCR.
Bisulfite PCR primers:
Oct4 binding site Forward (-483): 5’-TTTAATGTGAAGAGTAAGTAAGAAA-3’
Reverse (-185): 5’-ATAAAATAACCCAAACTAAAAAAAA-3’
p53 binding site Forward (-952): 5’-GTTTTTTGTAGAATAAAATTTAGGAAGA-3’
Reverse (-790): 5’-CAAACTTATCTACCACCATACCCAA-3’

4.2.16 Chromatin immunoprecipitation assays (ChIP-qPCR)
The cells were fixed in 1% formaldehyde for 10 minutes at room temperature and the
reaction was quenched by 0.125M glycine. The cells were washed twice with ice-cold
PBS, resuspended in lysis buffer [1% SDS, 10 mM EDTA, 50 mM Tris-HCl pH 8.0, 1 mM
phenylmethylsulphonyl fluoride (PMSF), 1ml per 10
6
cells] and incubated on ice for 10
minutes. The cell suspension was sonicated 5 times for 1 minute each. The sonicated
samples were centrifuged at 14,000rpm at 4°C for 15 minutes and the supernatant (input)
was stored at -80°C. The supernatants (50 µl) were immunoprecipitated with 5 µg of
relevant antibodies in RIPA buffer (1% Triton X-100, 0.1% deoxycholate, 140 mM NaCl, 1
mM PMSF) overnight at 4°C under rotation. Protein G beads were incubated with 100
73

µg/ml sonicated salmon sperm DNA and 1 µg/ml bovine serum albumin in RIPA buffer
under the same conditions. Blocked beads and immunoprecipitated samples were
combined the next day and were incubated under rotation for 3 hours at 4°C. The
immunoprecipitates were washed 7 times with RIPA wash buffer (1% Triton X-100, 0.1%
DOC, 0.1% SDS, 500 mM NaCl, 1 mM PMSF). Input samples (10 µl) and beads were
resuspended in 100 µl of 100 mM Tris-SDS and proteinase K to a final concentration of
200 µg/ml and incubated for 4 hours at 55°C and then overnight at 65°C. The next day,
samples were phenol-chloroform extracted and ethanol immunoprecipitated with NaOAc
and 20 mg of glycogen as a carrier. DNAs from input and immunoprecipitated pellets
were resuspended in 50 µl and 250 µl of TE buffer, respectively. The DNA content was
analyzed using qPCR (5 µl per 20 µl reaction)
ChIP-qPCR primers:
Oct4 binding site Forward (-285): 5’-AGTGAAATGAGGTAAAGCCTCT-3’
Reverse (-80): 5’-TATTCTCCCAGGCACCCA-3’
p53 binding site Forward (-741): 5’-TACAGTGAGAACTTGTCTCAAA-3’
Reverse (-541): 5’-GAGCCTGTGTCCTGCTTA-3’

74

Bibliography
American Cancer Society. Cancer Facts and Figures 2016. Atlanta, GA: American
Cancer Society; 2016.
Aoyama, T., Souri, M., Ueno, I., Kamijo, T., Yamaguchi, S., Rhead, W.J., Tanaka, K., and
Hashimoto, T. (1995). Cloning of human very-long-chain acyl-coenzyme A
dehydrogenase and molecular characterization of its deficiency in two patients. Am J
Hum Genet 57, 273-283.
Bar N, Dikstein R. miR-22 forms a regulatory loop in PTEN/AKT pathway and modulates
signaling kinetics. PLoS One. 2010 May 27;5(5):e10859.
Borutinskaite VV, Navakauskiene R, Magnusson KE. Retinoic acid and histone
deacetylase inhibitor BML-210 inhibit proliferation of human cervical cancer HeLa cells.
Ann N Y Acad Sci. 2006 Dec;1091:346-55.
Boyer LA, Lee TI, Cole MF, Johnstone SE, Levine SS, Zucker JP, Guenther MG, Kumar
RM, Murray HL, Jenner RG, Gifford DK, Melton DA, Jaenisch R, Young RA. Core
transcriptional regulatory circuitry in human embryonic stem cells. Cell. 2005 Sep
23;122(6):947-56.
Bussolati B, Bruno S, Grange C, Ferrando U, Camussi G. Identification of a
tumor-initiating stem cell population in human renal carcinomas. FASEB J. 2008
Oct;22(10):3696-705.
Butler LM, Zhou X, Xu WS, Scher HI, Rifkind RA, Marks PA, Richon VM. The histone
deacetylase inhibitor SAHA arrests cancer cell growth, up-regulates thioredoxin-binding
protein-2, and down-regulates thioredoxin. Proc Natl Acad Sci U S A. 2002 Sep
3;99(18):11700-5.
Caro, P., Kishan, A.U., Norberg, E., Stanley, I.A., Chapuy, B., Ficarro, S.B., Polak, K.,
Tondera, D., Gounarides, J., Yin, H., et al. (2012). Metabolic signatures uncover distinct
targets in molecular subsets of diffuse large B cell lymphoma. Cancer Cell 22, 547-560.
75

Chanda B, Ditadi A, Iscove NN, Keller G. Retinoic acid signaling is essential for
embryonic hematopoietic stem cell development. Cell. 2013 Sep 26;155(1):215-27.
Chen CL, Tsukamoto H, Liu JC, Kashiwabara C, Feldman D, Sher L, Dooley S, French
SW, Mishra L, Petrovic L, Jeong JH, Machida K. Reciprocal regulation by TLR4 and
TGF-β in tumor-initiating stem-like cells. J Clin Invest. 2013 Jul;123(7):2832-49.
Chiou SH, Yu CC, Huang CY, Lin SC, Liu CJ, Tsai TH, Chou SH, Chien CS, Ku HH, Lo
JF. Positive correlations of Oct-4 and Nanog in oral cancer stem-like cells and high-grade
oral squamous cell carcinoma. Clin Cancer Res. 2008 Jul 1;14(13):4085-95.
Chiou SH, Yu CC, Huang CY, Lin SC, Liu CJ, Tsai TH, Chou SH, Chien CS, Ku HH, Liu F,
Kong X, Lv L, Gao J. MiR-155 targets TP53INP1 to regulate liver cancer stem cell
acquisition and self-renewal. FEBS Lett. 2015 Feb 13;589(4):500-6.
Dahl AR, Grossi IM, Houchens DP, Scovell LJ, Placke ME, Imondi AR, Stoner GD, De
Luca LM, Wang D, Mulshine JL. Inhaled isotretinoin (13-cis retinoic acid) is an effective
lung cancer chemopreventive agent in A/J mice at low doses: a pilot study. Clin Cancer
Res. 2000 Aug;6(8):3015-24.
Ezeh UI, Turek PJ, Reijo RA, Clark AT. Human embryonic stem cell genes OCT4,
NANOG, STELLAR, and GDF3 are expressed in both seminoma and breast carcinoma.
Cancer. 2005 Nov 15;104(10):2255-65.
Fabbri M, Garzon R, Cimmino A, Liu Z, Zanesi N, Callegari E, Liu S, Alder H, Costinean
S, Fernandez-Cymering C, Volinia S, Guler G, Morrison CD, Chan KK, Marcucci G, Calin
GA, Huebner K, Croce CM. MicroRNA-29 family reverts aberrant methylation in lung
cancer by targeting DNA methyltransferases 3A and 3B. Proc Natl Acad Sci U S A. 2007
Oct 2;104(40):15805-10.
Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, Parkin DM, Forman
D, Bray F. Cancer incidence and mortality worldwide: sources, methods and major
patterns in GLOBOCAN 2012. Int J Cancer. 2015 Mar 1;136(5):E359-86.
76

Fu Z, Tindall DJ. FOXOs, cancer and regulation of apoptosis. Oncogene. 2008 Apr
7;27(16):2312-9.
Garattini E, Gianni M, Terao M. Retinoids as differentiating agents in oncology: a
network of interactions with intracellular pathways as the basis for rational therapeutic
combinations. Curr Pharm Des. 2007;13(13):1375-400.
Giordano, A., Calvani, M., Petillo, O., Grippo, P., Tuccillo, F., Melone, M.A., Bonelli, P.,
Calarco, A., and Peluso, G. (2005). tBid induces alterations of mitochondrial fatty acid
oxidation flux by malonyl-CoA-independent inhibition of carnitine palmitoyltransferase-1.
Cell Death Differ 12, 603-613.
Gupta PB, Onder TT, Jiang G, Tao K, Kuperwasser C, Weinberg RA, Lander ES.
Identification of selective inhibitors of cancer stem cells by high-throughput screening.
Cell. 2009 Aug 21;138(4):645-59.
Han MK, Song EK, Guo Y, Ou X, Mantel C, Broxmeyer HE. SIRT1 regulates apoptosis
and Nanog expression in mouse embryonic stem cells by controlling p53 subcellular
localization. Cell Stem Cell. 2008 Mar 6;2(3):241-51.
Hardie DG, Pan DA. Regulation of fatty acid synthesis and oxidation by the
AMP-activated protein kinase. Biochem Soc Trans. 2002 Nov;30(Pt 6):1064-70.
Hardie DG. AMP-activated protein kinase: an energy sensor that regulates all aspects of
cell function. Genes Dev. 2011 Sep 15;25(18):1895-908.
Harant H, Korschineck I, Krupitza G, Fazeny B, Dittrich C, Grunt TW. Retinoic acid
receptors in retinoid responsive ovarian cancer cell lines detected by polymerase chain
reaction following reverse transcription. Br J Cancer. 1993 Sep;68(3):530-6.
Hassan MM, Hwang LY, Hatten CJ, Swaim M, Li D, Abbruzzese JL, Beasley P, Patt YZ.
Risk factors for hepatocellular carcinoma: synergism of alcohol with viral hepatitis and
diabetes mellitus. Hepatology. 2002 Nov;36(5):1206-13.
Hawley, S.A., Davison, M., Woods, A., Davies, S.P., Beri, R.K., Carling, D., and Hardie,
77

D.G. (1996). Characterization of the AMP-activated protein kinase kinase from rat liver
and identification of threonine 172 as the major site at which it phosphorylates
AMP-activated protein kinase. J Biol Chem 271, 27879-27887.
Hoei-Hansen CE. Application of stem cell markers in search for neoplastic germ cells in
dysgenetic gonads, extragonadal tumours, and in semen of infertile men. Cancer Treat
Rev. 2008 Jun;34(4):348-67.
Huang ME, Ye YC, Chen SR, Chai JR, Lu JX, Zhoa L, Gu LJ, Wang ZY. Use of all-trans
retinoic acid in the treatment of acute promyelocytic leukemia. Blood. 1988
Aug;72(2):567-72.
Ito, K., Carracedo, A., Weiss, D., Arai, F., Ala, U., Avigan, D.E., Schafer, Z.T., Evans,
R.M., Suda, T., Lee, C.H., et al. (2012). A PML-PPAR-delta pathway for fatty acid
oxidation regulates hematopoietic stem cell maintenance. Nat Med 18, 1350-1358.
Jeter CR, Badeaux M, Choy G, Chandra D, Patrawala L, Liu C, Calhoun-Davis T,
Zaehres H, Daley GQ, Tang DG. Functional evidence that the self-renewal gene NANOG
regulates human tumor development. Stem Cells. 2009 May;27(5):993-1005.
Jeter CR, Liu B, Liu X, Chen X, Liu C, Calhoun-Davis T, Repass J, Zaehres H, Shen JJ,
Tang DG. NANOG promotes cancer stem cell characteristics and prostate cancer
resistance to androgen deprivation. Oncogene. 2011 Sep 8;30(36):3833-45.
Ji AR, Ku SY, Cho MS, Kim YY, Kim YJ, Oh SK, Kim SH, Moon SY, Choi YM. Reactive
oxygen species enhance differentiation of human embryonic stem cells into
mesendodermal lineage. Exp Mol Med. 2010 Mar 31;42(3):175-86.
Jiang R, Deng L, Zhao L, Li X, Zhang F, Xia Y, Gao Y, Wang X, Sun B. miR-22 promotes
HBV-related hepatocellular carcinoma development in males. Clin Cancer Res. 2011 Sep
1;17(17):5593-603.
Jin H, Liang L, Liu L, Deng W, Liu J. HDAC inhibitor DWP0016 activates p53
transcription and acetylation to inhibit cell growth in U251 glioblastoma cells. J Cell
78

Biochem. 2013 Jul;114(7):1498-509.
Katayama K, Nakamura A, Sugimoto Y, Tsuruo T, Fujita N. FOXO transcription
factor-dependent p15(INK4b) and p19(INK4d) expression. Oncogene. 2008 Mar
13;27(12):1677-86.
Kato Y, Salumbides BC, Wang XF, Qian DZ, Williams S, Wei Y, Sanni TB, Atadja P, Pili R.
Antitumor effect of the histone deacetylase inhibitor LAQ824 in combination with
13-cis-retinoic acid in human malignant melanoma. Mol Cancer Ther. 2007
Jan;6(1):70-81.
Khan O, La Thangue NB. HDAC inhibitors in cancer biology: emerging mechanisms and
clinical applications. Immunol Cell Biol. 2012 Jan;90(1):85-94.
Kim SY, Jeong S, Jung E, Baik KH, Chang MH, Kim SA, Shim JH, Chun E, Lee KY.
AMP-activated protein kinase-α1 as an activating kinase of TGF-β-activated kinase1 has
a key role in inflammatory signals. Cell Death Dis. 2012 Jul 26;3:e357.
Komatsu N, Kawamata N, Takeuchi S, Yin D, Chien W, Miller CW, Koeffler HP. SAHA, a
HDAC inhibitor, has profound anti-growth activity against non-small cell lung cancer cells.
Oncol Rep. 2006 Jan;15(1):187-91.
Kumagai T, Wakimoto N, Yin D, Gery S, Kawamata N, Takai N, Komatsu N, Chumakov A,
Imai Y, Koeffler HP. Histone deacetylase inhibitor, suberoylanilide hydroxamic acid
(Vorinostat, SAHA) profoundly inhibits the growth of human pancreatic cancer cells. Int J
Cancer. 2007 Aug 1;121(3):656-65.
Kurtz, D.M., Rinaldo, P., Rhead, W.J., Tian, L., Millington, D.S., Vockley, J., Hamm, D.A.,
Brix, A.E., Lindsey, J.R., Pinkert, C.A., et al. (1998). Targeted disruption of mouse
long-chain acyl-CoA dehydrogenase gene reveals crucial roles for fatty acid oxidation.
Proceedings of the National Academy of Sciences of the United States of America 95,
15592-15597.
Lee TK, Castilho A, Cheung VC, Tang KH, Ma S, Ng IO. CD24(+) liver tumor-initiating
79

cells drive self-renewal and tumor initiation through STAT3-mediated NANOG regulation.
Cell Stem Cell. 2011 Jul 8;9(1):50-63.
Li H, Wu X. Histone deacetylase inhibitor, Trichostatin A, activates p21WAF1/CIP1
expression through downregulation of c-myc and release of the repression of c-myc from
the promoter in human cervical cancer cells. Biochem Biophys Res Commun. 2004 Nov
12;324(2):860-7.
Li L, Wang L, Li L, Wang Z, Ho Y, McDonald T, Holyoake TL, Chen W, Bhatia R.
Activation of p53 by SIRT1 inhibition enhances elimination of CML leukemia stem cells in
combination with imatinib. Cancer Cell. 2012 Feb 14;21(2):266-81.
Lim CA, Yao F, Wong JJ, George J, Xu H, Chiu KP, Sung WK, Lipovich L, Vega VB.
Genome-wide mapping of RELA(p65) binding identifies E2F1 as a transcriptional
activator recruited by NF-kappaB upon TLR4 activation. Mol Cell. 2007 Aug
17;27(4):622-35.
Lo JF. Positive correlations of Oct-4 and Nanog in oral cancer stem-like cells and
high-grade oral squamous cell carcinoma. Clin Cancer Res. 2008 Jul 1;14(13):4085-95.
Ludin A, Gur-Cohen S, Golan K, Kaufmann KB, Itkin T, Medaglia C, Lu XJ,
Ledergor G, Kollet O, Lapidot T. Reactive oxygen species regulate hematopoietic stem
cell self-renewal, migration and development, as well as their bone marrow
microenvironment. Antioxid Redox Signal. 2014 Oct 10;21(11):1605-19. doi:
10.1089/ars.2014.5941.Epub 2014 Jun 26.
Ma S, Lee TK, Zheng BJ, Chan KW, Guan XY. CD133+ HCC cancer stem cells confer
chemoresistance by preferential expression of the Akt/PKB survival pathway. Oncogene.
2008 Mar 13;27(12):1749-58.
Machida K, Cheng KT, Sung VM, Lee KJ, Levine AM, Lai MM. Hepatitis C virus infection
activates the immunologic (type II) isoform of nitric
oxide synthase and thereby enhances DNA damage and mutations of cellular genes. J
80

Virol. 2004 Aug;78(16):8835-43.
Machida K, Cheng KT, Lai CK, Jeng KS, Sung VM, Lai MM. Hepatitis C virus triggers
mitochondrial permeability transition with production of reactive oxygen species, leading
to DNA damage and STAT3 activation. J Virol. 2006 Jul;80(14):7199-207.
Machida K, Cheng KT, Sung VM, Levine AM, Foung S, Lai MM. Hepatitis C virus induces
toll-like receptor 4 expression, leading to enhanced production of beta interferon and
interleukin-6. J Virol. 2006 Jan;80(2):866-74.
Machida K, Tsukamoto H, Mkrtchyan H, Duan L, Dynnyk A, Liu HM, Asahina K,
Govindarajan S, Ray R, Ou JH, Seki E, Deshaies R, Miyake K, Lai MM. Toll-like receptor
4 mediates synergism between alcohol and HCV in hepatic oncogenesis involving stem
cell marker Nanog. Proc Natl Acad Sci U S A. 2009 Feb 3;106(5):1548-53.
Ma S, Chan KW, Hu L, Lee TK, Wo JY, Ng IO, Zheng BJ, Guan XY. Identification and
characterization of tumorigenic liver cancer stem/progenitor cells. Gastroenterology.
2007 Jun;132(7):2542-56.
Ma S, Tang KH, Chan YP, Lee TK, Kwan PS, Castilho A, Ng I, Man K, Wong N, To KF,
Zheng BJ, Lai PB, Lo CM, Chan KW, Guan XY. miR-130b Promotes CD133(+) liver
tumor-initiating cell growth and self-renewal via tumor protein 53-induced nuclear protein
1. Cell Stem Cell. 2010 Dec 3;7(6):694-707.
Majid S, Dar AA, Saini S, Shahryari V, Arora S, Zaman MS, Chang I, Yamamura S,
Tanaka Y, Chiyomaru T, Deng G, Dahiya R. miRNA-34b inhibits prostate cancer through
demethylation, active chromatin modifications, and AKT pathways. Clin Cancer Res.
2013 Jan 1;19(1):73-84.
Meletis K, Wirta V, Hede SM, Nistér M, Lundeberg J, Frisén J. p53 suppresses the
self-renewal of adult neural stem cells. Development. 2006 Jan;133(2):363-9.
Meyskens FL Jr, Jacobson J, Nguyen B, Weiss GR, Gandara DR, MacDonald JS.
Phase II trial of oral beta-all trans-retinoic acid in hepatocellular carcinoma (SWOG 9157).
81

Invest New Drugs. 1998;16(2):171-3.
Motzer RJ, Murphy BA, Bacik J, Schwartz LH, Nanus DM, Mariani T, Loehrer P, Wilding
G, Fairclough DL, Cella D, Mazumdar M. Phase III trial of interferon alfa-2a with or
without 13-cis-retinoic acid for patients with advanced renal cell carcinoma. J Clin Oncol.
2000 Aug;18(16):2972-80.
Noh KH, Kim BW, Song KH, Cho H, Lee YH, Kim JH, Chung JY, Kim JH, Hewitt SM,
Seong SY, Mao CP, Wu TC, Kim TW. Nanog signaling in cancer promotes stem-like
phenotype and immune evasion. J Clin Invest. 2012 Nov;122(11):4077-93.
Ofengeim D, Yuan J. Regulation of RIP1 kinase signalling at the crossroads of
inflammation and cell death. Nat Rev Mol Cell Biol. 2013 Nov;14(11):727-36.
Paul MK, Bisht B, Darmawan DO, Chiou R, Ha VL, Wallace WD, Chon AT, Hegab AE,
Grogan T, Elashoff DA, Alva-Ornelas JA, Gomperts BN(6). Dynamic changes in
intracellular ROS levels regulate airway basal stem cell homeostasis through
Nrf2-dependent Notch signaling. Cell Stem Cell. 2014 Aug 7;15(2):199-214.
Pan G, Thomson JA. Nanog and transcriptional networks in embryonic stem cell
pluripotency. Cell Res. 2007 Jan;17(1):42-9.
Paumen, M.B., Ishida, Y., Han, H., Muramatsu, M., Eguchi, Y., Tsujimoto, Y., and Honjo,
T. (1997). Direct interaction of the mitochondrial membrane protein carnitine
palmitoyltransferase I with Bcl-2. Biochem Biophys Res Commun 231, 523-525.
Petruccelli LA, Dupéré-Richer D, Pettersson F, Retrouvey H, Skoulikas S, Miller WH Jr.
Vorinostat induces reactive oxygen species and DNA damage in acute myeloid leukemia
cells. PLoS One. 2011;6(6):e20987.
Reynolds CP, Matthay KK, Villablanca JG, Maurer BJ. Retinoid therapy of high-risk
neuroblastoma. Cancer Lett. 2003 Jul 18;197(1-2):185-92.
Rochette-Egly C. Retinoic acid signaling and mouse embryonic stem cell differentiation:
Cross talk between genomic and non-genomic effects of RA. Biochim Biophys Acta.
82

2015 Jan;1851(1):66-75.
Rodda DJ, Chew JL, Lim LH, Loh YH, Wang B, Ng HH, Robson P. Transcriptional
regulation of nanog by OCT4 and SOX2. J Biol Chem. 2005 Jul 1;280(26):24731-7.
Rosato RR, Almenara JA, Dai Y, Grant S. Simultaneous activation of the intrinsic and
extrinsic pathways by histone deacetylase (HDAC) inhibitors and tumor necrosis
factor-related apoptosis-inducing ligand (TRAIL) synergistically induces mitochondrial
damage and apoptosis in human leukemia cells. Mol Cancer Ther. 2003
Dec;2(12):1273-84.
Roses AD. Pharmacogenetics in drug discovery and development: a translational
perspective. Nat Rev Drug Discov. 2008 Oct;7(10):807-17.
Rountree CB, Senadheera S, Mato JM, Crooks GM, Lu SC. Expansion of liver cancer
stem cells during aging in methionine adenosyltransferase 1A-deficient mice. Hepatology.
2008 Apr;47(4):1288-97.
Rubin Grandis J, Zeng Q, Tweardy DJ. Retinoic acid normalizes the increased gene
transcription rate of TGF-alpha and EGFR in head and neck cancer cell lines. Nat Med.
1996 Feb;2(2):237-40.
Salomoni P, Pandolfi PP. Transcriptional regulation of cellular transformation. Nat Med.
2000 Jul;6(7):742-4.
Samudio, I., Harmancey, R., Fiegl, M., Kantarjian, H., Konopleva, M., Korchin, B.,
Kaluarachchi, K., Bornmann, W., Duvvuri, S., Taegtmeyer, H., et al. (2010).
Pharmacologic inhibition of fatty acid oxidation sensitizes human leukemia cells to
apoptosis induction. The Journal of clinical investigation 120, 142-156.
Sandor V, Senderowicz A, Mertins S, Sackett D, Sausville E, Blagosklonny MV, Bates
SE. P21-dependent g(1)arrest with downregulation of cyclin D1 and upregulation of
cyclin E by the histone deacetylase inhibitor FR901228. Br J Cancer. 2000
Sep;83(6):817-25.
83

Sauer H, Ruhe C, Müller JP, Schmelter M, D'Souza R, Wartenberg M. Reactive oxygen
species and upregulation of NADPH oxidases in mechanotransduction of embryonic
stem cells. Methods Mol Biol. 2008;477:397-418.
Shen JJ, Tang DG. NANOG promotes cancer stem cell characteristics and prostate
cancer resistance to androgen deprivation. Oncogene. 2011 Sep 8;30(36):3833-45.
Simandi Z, Balint BL, Poliska S, Ruhl R, Nagy L. Activation of retinoic acid receptor
signaling coordinates lineage commitment of spontaneously differentiating mouse
embryonic stem cells in embryoid bodies. FEBS Lett. 2010 Jul 16;584(14):3123-30.
Strait KA, Warnick CT, Ford CD, Dabbas B, Hammond EH, Ilstrup SJ. Histone
deacetylase inhibitors induce G2-checkpoint arrest and apoptosis in cisplatinum-resistant
ovarian cancer cells associated with overexpression of the Bcl-2-related protein Bad. Mol
Cancer Ther. 2005 Apr;4(4):603-11.
Song MS, Salmena L, Pandolfi PP. The functions and regulation of the PTEN tumour
suppressor. Nat Rev Mol Cell Biol. 2012 Apr 4;13(5):283-96.
Song SJ, Poliseno L, Song MS, Ala U, Webster K, Ng C, Beringer G, Brikbak NJ, Yuan X,
Cantley LC, Richardson AL, Pandolfi PP. MicroRNA-antagonism regulates breast cancer
stemness and metastasis via TET-family-dependent chromatin remodeling. Cell. 2013
Jul 18;154(2):311-24.
Song SJ, Ito K, Ala U, Kats L, Webster K, Sun SM, Jongen-Lavrencic M,
Manova-Todorova K, Teruya-Feldstein J, Avigan DE, Delwel R, Pandolfi PP. The
oncogenic microRNA miR-22 targets the TET2 tumor suppressor to promote
hematopoietic stem cell self-renewal and transformation. Cell Stem Cell. 2013 Jul
3;13(1):87-101.
Stickel F. Alcoholic cirrhosis and hepatocellular carcinoma. Adv Exp Med Biol.
2015;815:113-30.
Suetsugu A, Nagaki M, Aoki H, Motohashi T, Kunisada T, Moriwaki H. Characterization
84

of CD133+ hepatocellular carcinoma cells as cancer stem/progenitor cells. Biochem
Biophys Res Commun. 2006 Dec 29;351(4):820-4.
Sun AX, Liu CJ, Sun ZQ, Wei Z. NANOG: a promising target for digestive malignant
tumors. World J Gastroenterol. 2014 Sep 28;20(36):13071-8.
Swaminathan SK, Roger E, Toti U, Niu L, Ohlfest JR, Panyam J. CD133-targeted
paclitaxel delivery inhibits local tumor recurrence in a mouse model of breast cancer. J
Control Release. 2013 Nov 10;171(3):280-7.
Tessitore A, Cicciarelli G, Del Vecchio F, Gaggiano A, Verzella D, Fischietti M, Vecchiotti
D, Capece D, Zazzeroni F, Alesse E. MicroRNAs in the DNA Damage/Repair Network
and Cancer. Int J Genomics. 2014;2014:820248.
Testro AG, Visvanathan K. Toll-like receptors and their role in gastrointestinal disease. J
Gastroenterol Hepatol. 2009 Jun;24(6):943-54.
Torres, J., and Watt, F.M. (2008). Nanog maintains pluripotency of mouse embryonic
stem cells by inhibiting NFkappaB and cooperating with Stat3. Nat Cell Biol 10, 194-201.
Uthaya Kumar DB, Chen CL, Liu JC, Feldman DE, Sher LS, FrenchS, DiNorcia J,
French SW, Naini BV, Junrungsee S, Agopian VG, Zarrinpar A, Machida K. TLR4
Signaling via NANOG Cooperates With STAT3 to Activate Twist1 and Promote Formation
of Tumor-Initiating Stem-Like Cells in Livers of Mice. Gastroenterology. 2016 Mar;
150(3):707-19.
van den Berg DL, Snoek T, Mullin NP, Yates A, Bezstarosti K, Demmers J, Chambers I,
Poot RA. An Oct4-centered protein interaction network in embryonic stem cells. Cell
Stem Cell. 2010 Apr 2;6(4):369-81.
Wang XQ, Ng RK, Ming X, Zhang W, Chen L, Chu AC, Pang R, Lo CM, Tsao SW, Liu X,
Poon RT, Fan ST. Epigenetic regulation of pluripotent genes mediates stem cell features
in human hepatocellular carcinoma and cancer cell lines. PLoS One. 2013 Sep
4;8(9):e72435.
85

Waris G, Ahsan H. Reactive oxygen species: role in the development of cancer and
various chronic conditions. J Carcinog. 2006 May 11;5:14.
Weiss FU, Marques IJ, Woltering JM, Vlecken DH, Aghdassi A, Partecke LI, Heidecke
CD, Lerch MM, Bagowski CP. Retinoic acid receptor antagonists inhibit miR-10a
expression and block metastatic behavior of pancreatic cancer. Gastroenterology. 2009
Dec;137(6):2136-45.e1-7.
Xu F, Dai C, Zhang R, Zhao Y, Peng S, Jia C. Nanog: a potential biomarker for liver
metastasis of colorectal cancer. Dig Dis Sci. 2012 Sep;57(9):2340-6.
Yang H, Bushue N, Bu P, Wan YJ. Induction and intracellular localization of Nur77 dictate
fenretinide-induced apoptosis of human liver cancer cells. Biochem Pharmacol. 2010 Apr
1;79(7):948-54.
Ye F, Zhou C, Cheng Q, Shen J, Chen H. Stem-cell-abundant proteins Nanog,
Nucleostemin and Musashi1 are highly expressed in malignant cervical epithelial cells.
BMC Cancer. 2008 Apr 18;8:108.
Zhang, D., Liu, Z.X., Choi, C.S., Tian, L., Kibbey, R., Dong, J., Cline, G.W., Wood, P.A.,
and Shulman, G.I. (2007). Mitochondrial dysfunction due to long-chain Acyl-CoA
dehydrogenase deficiency causes hepatic steatosis and hepatic insulin resistance.
Proceedings of the National Academy of Sciences of the United States of America 104,
17075-17080.
Zhang J, Luo N, Luo Y, Peng Z, Zhang T, Li S. microRNA-150 inhibits human
CD133-positive liver cancer stem cells through negative regulation of the transcription
factor c-Myb. Int J Oncol. 2012 Mar;40(3):747-56.
Zhang, L.F., Ding, J.H., Yang, B.Z., He, G.C., and Roe, C. (2003). Characterization of the
bidirectional promoter region between the human genes encoding VLCAD and PSD-95.
Genomics 82, 660-668.
Zhang S, Balch C, Chan MW, Lai HC, Matei D, Schilder JM, Yan PS, Huang TH, Nephew
86

KP. Identification and characterization of ovarian cancer-initiating cells from primary
human tumors. Cancer Res. 2008 Jun 1;68(11):4311-20.
Zhao XY, Ly LH, Peehl DM, Feldman D. Induction of androgen receptor by
1alpha,25-dihydroxyvitamin D3 and 9-cis retinoic acid in LNCaP human prostate cancer
cells. Endocrinology. 1999 Mar;140(3):1205-12.
Zhou BB, Zhang H, Damelin M, Geles KG, Grindley JC, Dirks PB. Tumour-initiating cells:
challenges and opportunities for anticancer drug discovery. at Rev Drug Discov. 2009
Oct;8(10):806-23.

Abstract (if available)

Abstract Hepatocellular carcinoma (HCC) is one of the most common cancers and is a leading cause of deaths worldwide. Conventional chemotherapy and surgery are the current treatments for HCC patients

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

Mechanisms of nuclear translocation of pyruvate dehydrogenase and its effects on tumorigenesis of hepatocellular carcinoma…

PDF

Hepatitis B virus X protein regulation of β-catenin and NANOG and co-regulatory role with YAP1 in HCC malignancy

PDF

Targeting mitochondrion-nucleus PDH1 transfer to suppress self-renewal and epigenetic NANOG reprogramming of tumor-initiating cells

PDF

Lnc-RNA and WNT signal synergistically promotes self-renewal of tumor initiating stem-like cells via epigenetic regulation

PDF

Tri-specific T cell engager immunotherapy targeting tumor initiating cells

PDF

The roles of endoplasmic reticulum chaperones in regulating liver homeostasis and tumorigenesis

PDF

Musashi-2 promotes c-MYC expression through IRES-dependent translation and self-renewal ability in hepatocellular carcinoma

PDF

Steatosis induces Wnt/β-catenin pathway to stimulate proliferation of hepatic tumor initiating cells and promote liver cancer development

PDF

Cell fate reprogramming of liver tumor-initiating stem-like cells via phosphorylated NUMB and TBC1D15

PDF

Tri-specific T cell engager immunotherapy targeting tumor initiating cells

PDF

MSI2 level alters histone transcription rate in HepG2 cells

PDF

Regulation of miR21, miR221, CXCL5 and OPN in liver cancer

PDF

Targeting glioma cancer stem cells for the treatment of glioblastoma multiforme

PDF

The essential role of histone H2A deubiquitinase MYSM1 in natural killer cell maturation and HSC homeostasis

PDF

Regulation of T cell HLA-DR by CD3 ζ signaling

PDF

PTEN deletion induced tumor initiating cells: Strategies to accelerate the disease progression of liver cancer

PDF

AKT exhibits isoform-specific function in hepatocyte transformation and response to cellular stress

PDF

Novel long non-coding RNA, LINC00824, contributes to stemness in hepatocytes

Asset Metadata

Creator Chen, Chia-Lin (author)

Core Title Targeting tumor-initiating stem-like cells through metabolic and epigenetic regulations

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 07/26/2018

Defense Date 06/08/2016

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

Tag epigenetic regulation,hepatocellular carcinoma,metabolic reprogramming,NANOG,OAI-PMH Harvest,tumor-initiating stem-like cells

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Ou, J.-H. James (committee chair), Machida, Keigo (committee member), Stiles, Bangyan L. (committee member)

Creator Email chen104@usc.edu,truip75@yahoo.com.tw

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-285027

Unique identifier UC11280383

Identifier etd-ChenChiaLi-4664.pdf (filename),usctheses-c40-285027 (legacy record id)

Legacy Identifier etd-ChenChiaLi-4664.pdf

Dmrecord 285027

Document Type Dissertation

Format application/pdf (imt)

Rights Chen, Chia-Lin

Type texts

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

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

Repository Name University of Southern California Digital Library

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