University of Southern California Dissertations and Theses

The cancer stem-like phenotype: therapeutics, phenotypic plasticity and mechanistic studies

(USC Thesis Other)

The cancer stem-like phenotype: therapeutics, phenotypic plasticity and mechanistic studies

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THE CANCER STEM-LIKE PHENOTYPE:
THERAPEUTICS, PHENOTYPIC PLASTICITY AND
MECHANISTIC STUDIES

by
Kaijie He

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 2012

Copyright 2012 Kaijie He

ii
Acknowledgments
I would like to sincerely thank my PI Dr. Amir Goldkorn for his patience, kindness
and guidance during my graduate studies at USC. Your mentorship is paramount in
providing me not only knowledge and scientific skills but also essential life skills from
which I can benefit for a lifetime. I respect you as a great teacher and also a true friend
to me.
I also wish to thank my committee members, Dr. Michael Kahn, Dr. Gregor Adams,
Dr. Agnieszka Kobielak and Dr. Jacek Pinski for your invaluable suggestions on my
research. I am also grateful to my colleagues, Dr. Tong Xu, Dr. Yucheng Xu, Roy Lau
for sharing your knowledge on flow cytometry, molecular cloning and animal operations,
and teaching me many useful lab techniques; Ayesha Bhatia, Anisha Madhav and all
other lab members for your stimulating discussions.
My thanks also go to Drs. Elizabeth Blackburn, Peter Jones, Gerhard Coetzee,
Thomas Chen, Ite Laird-Offringa and Jacek Pinski for generously providing cancer cell
lines used in this study, Dr. Michael Kahn for providing ICG001, IQ-1 and CBP, P300
expression vectors, Drs. Daniel Weisenberger and Peter Laird (USC Epigenome Center)
for their generous help with the DNA methylation experiments.

iii
Table of Contents

Acknowledgments ii

List of Tables v

List of Figures vi

Abstract viii

Chapter 1: Introduction 1
Biology of telomeres and telomerase 2
Telomerase in cancer 4
Telomerase therapeutics 6
Telomerase and cancer stem cells 16
Phenotypic plasticity in Cancer stem cells 19
Signaling pathway regulating Cancer plasticity 22

Chapter 2: Prostate tumor cells with cancer stem-like properties have
high telomerase activity and are rapidly killed by telomerase interference 26
Abstract 26
Introduction 27
Materials and Methods 31
Results 36
Discussion 52

Chapter 3: Cancer cells cyclically lose and regain a drug-resistant highly
tumorigenic features characteristic of a cancer stem-like phenotype 57
Abstract 57
Introduction 58
Materials and Methods 61
Results 65
Discussion 89

Chapter 4: Plasticity between the cancer stem-like and non cancer stem
-like states is regulated by the PI3K/Akt/β-catenin/CBP pathway 96
Abstract 96
Introduction 98

iv
Materials and Methods 103
Results 107
Discussion 120

Chapter 5: Summary and future directions 125
Summary 125
Future directions 127

References 129

v
List of Tables

Table 1: GSTP-1 promoter methlylation measurements on prostate tumor
samples 41

Table 2: Sequences of PCR primers 42

Table 3: Oligo sequences and protocol for qPCR-TRAP assay 43

Table 4: Clinical and histological features of prostatectomy specimens 44

Table 5: Primer sequences used for qRT-PCR 75

vi
List of Figures

Figure 1: FACS isolation of cell subpopulations. 45

Figure 2: CSC phenotypes of cell subpopulations isolated from human
prostate tumors. 46

Figure 3: Telomerase and telomere characterization of CSC
and non-CSC cell subpopulations isolated from human prostate tumors. 47

Figure 4: Induction of telomerase interference (reprogramming of
telomerase) in tumor-derived CSC. 48

Figure 5: CSC phenotypes of cell subpopulations isolated from DU145
human prostate cancer cell line. 49

Figure 6: Effect of telomerase interference on in vivo tumor formation
by DU145-derived CSC. 51

Figure 7: Side population (SP) analysis in various cancer cell lines. 76

Figure 8: The side population (SP) is enriched for drug-resistant highly 77
-tumorigenic cells.

Figure 9: Sorting efficiency of FACSAria. 79

Figure 10: Relative gene expression. 80

Figure 11: FACS analysis for additional markers. 81

Figure 12: The NSP subpopulation generates SP and NSP subpopulations
in vitro and in vivo. 82

Figure 13: NSP single cell clones reconstitute SP and NSP subpopulations
in vitro and in vivo. 83

vii
Figure 14: SP subpopulation size fluctuates cyclically in the course of cell
culture. 84

Figure 15: SP cells arise through conversion of NSP cells. 85

Figure 16: Percentage of GFP-negative cells in the SP subpopulation. 87

Figure 17: Control experiments confirming that lentiviral-mediated
GFP expression is maintained in vivo throughout course of experiments. 88

Figure 18: Inhibition of PI3K/Akt pathway diminishes SP and inhibits
NSP to SP conversion. 115

Figure 19: Akt signaling affects side population. 116

Figure 20: β-catenin signaling regulates side population. 117

Figure 21: CBP functions downstream of β-catenin signaling to regulate
side population phenotype. 118

Figure 22: GSK3 β/ β-catenin/CBP signaling regulates regeneration of SP
from NSP. 119

viii
Abstract
Cancer claims over 500,000 lives in the U.S. annually, a mortality rate that is largely
attributable to solid tumors that have metastasized and become resistant to available
treatments. This type of disease progression may be mediated by cancer stem cells (CSC),
rare and unique cancer cells recently identified in many types of malignancies, and these
CSC are thought to play a central role in tumor formation, therapy-resistance, and
ultimately cancer progression and metastasis.
Unfortunately, however, to date there are very few effective CSC-targeting therapies.
To address this question, we isolated a putative CSC population from human prostate
tumors and cell lines, and we showed for the first time that these cells possess extremely
high telomerase activity relative to the bulk, unselected cancer cells. Strikingly,
telomerase interference – reprogramming telomerase to add incorrect “toxic” telomeres –
could induce rapid apoptosis and marked growth inhibition in prostate CSC and
abrogated their ability to form new tumors in SCID mice, which offers the first
tumor-derived and in vivo evidence that telomerase may ultimately form the basis for
more effective new CSC-targeting therapies.
The drug resistant and brisk tumor initiation abilities have been viewed as
pre-existing phenotypes only present in the small subpopulation of cancer stem cells, and
they have been intuitively conceptualized as self-renewing founder cells from which

ix
more differentiated cancer cells derive. However, recent work in cancer cell lines has
demonstrated that drug-resistant tumor initiating features can emerge de novo within
fractionated subpopulations of cells initially lacking these phenotypes. In our study, we
used a “side population” cancer stem cell model with GFP-labeling technique, and
demonstrated for the first time that adaptive, cancer-promoting traits like drug-resistance
and brisk tumor initiation arise not only as solitary events under selective pressures, but
also as highly orchestrated transitions occurring concurrently in large numbers of cells
even without specifically-induced drug selection, ectopic gene expression, or
fractionation into subpopulations. In addition, our follow-up mechanistic studies have
identified the PI3K/Akt/ β-catenin/CBP pathway to play a critical role in regulating this
dynamic equilibrium and regeneration of CSC. We hope our findings would contribute to
a better understanding of CSC, and can potentially offer new strategies for targeting these
drug-resistant, tumor-forming cells, ultimately leading to more effective treatments for
patients.

1
Chapter 1
Introduction

The phenomenon of therapy resistance and tumor re-initiation has always been a vital
challenge in cancer therapy (Dean et al., 2005; Kelland, 2007). In the past few years,
extensive studies have been done using varieties of animal models and clinical human
samples to investigate these properties of cancer cells (Diehn et al., 2009; Li et al.,
2008b; Olive et al., 2009; Reya et al., 2001; Visvader and Lindeman, 2008; Woodward
et al., 2007). However, the mechanisms of how cancer cells escape conventional therapy
and relapse in situ or further metastasize to other sites of the body still remain elusive.
Based on recent studies, a model has been proposed that anti-cancer therapy
characteristics may stem from a small subpopulation of putative “cancer stem cells”
(CSC), pluripotent cells with the capacity to differentiate and give rise to entire new
tumors, much the same as normal tissue stem cells are able to differentiate and
regenerate normal tissues. CSC are characterized by a long relative life span, activation
of pathways necessary for self-renewal (e.g. Wnt/ β-catenin, Notch, Shh, BMI1), relative
resistance to standard chemotherapy, and high tumorigenicity relative to unselected
tumor cells. CSC subpopulations were first isolated in acute myeloid leukemia, then
breast cancer and glioblastoma, and more recently reported in prostate, pancreatic, colon,

2
and bladder cancers (Akhtar et al., 2009; Gangemi et al., 2009; Gupta et al., 2009a;
O'Brien et al., 2009; Park et al., 2009; Visvader and Lindeman, 2008). Their therapy
resistance is attributed to high expression of ATP-binding cassette (ABC) drug
transporters (e.g. ABCG2, ABCB1) which can actively pump out many
chemotherapeutic agents (Chikazawa et al., 2010; Dean et al., 2005; Hu et al., 2010;
Singh et al., 2010; Steiniger et al., 2008). Given these properties, CSC have become an
important therapeutic target, and their biology and role in cancer progression are
subjects of intense investigation (Alix-Panabieres et al., 2007).

Biology of telomeres and telomerase
Telomere biology. The well-established canonical function of telomerase enzyme is
the maintenance and lengthening of telomeres, the tandem repetitive DNA sequences
located at the ends of human chromosomes (de Lange T, 2005). The 3’ telomeric strand
consists of G-rich tandem repeats (TTAGGG) terminating in a single stranded
3’-overhang with a lariat structure that often loops back and reinserts as a terminal T-loop
into the double-stranded telomeric region (Griffith et al., 1999). The two essential
functions of telomeres are protecting chromosome ends (the “capping" function of
telomeres), and facilitating their complete replication. Average telomere length in human
at birth is about 15-20 kb (Hastie et al., 1990; Verdun and Karlseder, 2007); however, as a

3
result of telomerase down-regulation in normal somatic cells, human chromosomes can
lose up to 50-200 nucleotides of telomeric sequence per cell division (Allsopp et al., 1992;
Smogorzewska and de Lange, 2004). Such shortening of telomeres is attributed to the
so-called “end replication problem” (Watson, 1972) wherein spaces left by RNA primers
during lagging strand replication lead to progressive shortening with each
division/replication cycle. The resulting telomeric shortening has been proposed to be a
mitotic clock that monitors cell divisions, and sufficiently short telomeres and the
absence of telomerase may signal replicative senescence at ~4-6Kb, known as mortality
stage 1 or M1 (Allsopp et al., 1992; Blackburn, 2000; Verdun and Karlseder, 2007). Some
cells may bypass the M1 via inactivation of p53 or RB1 and enter mortality stage 2 (M2
or crisis) manifested by genomic instability and fusion/breakage mutagenic events and
massive cell death. Activation of telomerase at M1 or M2 can stabilize telomere length
and immortalize cells, which – in the case of M2 – may potentiate cancer formation
(Verdun and Karlseder, 2007; Wright et al., 1989). In addition to telomerase, numerous
proteins have been shown to interact with telomeres, among them the six members of the
Shelterin complex (TRF1, TRF2, Pot1, Tin2, Rap1, TPP1) (de Lange, 2005) which
interact directly or indirectly with telomeric DNA to regulated telomere length and recruit
telomerase and additional proteins to single stranded or double stranded telomeric
regions.

4
Telomerase biology. The telomerase core ribonucleoprotein consists of two
components: a reverse transcriptase protein (TERT, 127 kD in humans) and an intrinsic
telomerase RNA molecule (Ter, 153kD and 451 bp in humans) (Cohen et al., 2007). Ter
contains a short template sequence used by TERT to reverse transcribe telomeric DNA
(Greider and Blackburn, 1989). The secondary and tertiary structures of TERT and Ter
and the elucidation of their functional domains are subject of ongoing investigation and
are beyond the scope of this review (Collins, 2006, 2008). Several additional proteins that
associate with the core RNP have been identified (Cohen et al., 2007; Egan and Collins,
2010), among them dyskerin, which plays a pivotal role in telomerase biogenesis and
function; mutations in dyskerin are implicated in the telomerase dysfunction disease
dyskeratosis congenita (Kirwan and Dokal, 2008; Mitchell et al., 1999). Whereas these
proteins play a critical role in telomerase holoenzyme biogenesis and function, their
potential as therapeutic targets has not been extensively explored to date.

Telomerase in cancer
Expression of telomerase protein (TERT) is tightly regulated at the transcriptional
level: with the exception of renewable progenitor compartments (hematopoietic,
epidermal, gastrointestinal), most benign, terminally differentiated tissues have extremely
low telomerase activity (Forsyth et al., 2002; Wright et al., 1996). In contrast, malignant

5
cells from as many as 90% of all human cancer – including prostate, melanoma, breast,
colon, sarcoma, and ovarian (Carey et al., 1999; Oishi et al., 1998; Okayasu et al., 1998;
Pirker et al., 2003; Poremba et al., 2002; Tomoda et al., 2002; Yoshida et al., 1999) –
have significant telomerase expression and telomerase activity levels that correlate
directly with malignant/metastatic potential by enabling continued proliferation and
telomere stabilization beyond M1 and M2/crisis. As a result of this sharp phenotypic
dichotomy between benign and malignant tissues, telomerase has been recognized as a
highly promising cancer therapeutic target – minimally toxic to host tissues while
potentially efficacious against a majority of malignancies. Indeed, early in vitro studies
demonstrated that activation of telomerase by ectopic expression of TERT, combined
with expression of SV40 antigen (inactivates pRB and p53) and H-ras, was sufficient to
transform benign cells in culture (Hahn, 2002; Hahn et al., 1999a). Conversely, attempts
to attenuate telomerase function in cell culture led to not only telomere shortening (Strahl
and Blackburn, 1996) but also cellular apoptosis, thus limiting cancer cell growth in vitro
(Hahn et al., 1999b; Zhang et al., 1999), thus providing additional compelling
mechanistic evidence that telomerase-dependent telomere maintenance is essential for
cancer cell immortalization, tumor progression, and disease metastasis.

6
Telomerase therapeutics
A variety of telomerase-based therapeutic strategies have been developed over the
past decade. Some have progressed into clinical trials, while others are still undergoing in
vitro study and pre-clinical development. For the purposes of this review, the main
therapeutic approaches will be discussed based on their general mechanism of action: 1.
Approaches that directly target the enzymatic function of telomerase; 2. Approaches that
target telomerase as a cancer specific marker; and 3. Approaches that target telomeres in
order to disrupt telomerase function.

1. Targeting the enzymatic function of telomerase
Telomerase inhibition. Perhaps the most straightforward therapeutic strategy seeks
to inhibit the enzymatic activity of telomerase, thus abolishing its telomere-lengthening
function, leaving telomeres to shorten with subsequent cell divisions, ultimately resulting
in senescence or apoptosis. Significant efforts to identify small molecule inhibitors of
telomerase reverse transcriptase function have failed to yield an agent with adequate
efficacy and specificity. However, an alternative tact undertaken by Geron, Corp. has
yielded an inhibitor which has been the single most clinically tested telomerase
therapeutics to date. GRN163L (Imetelstat) is an oligonucleotide with a sequence
TAGGGTTAGACAA that is complementary to the 11-nucleotide hTer template, the

7
highly conserved region of telomerase RNA used by the RNP to reverse transcribe
telomeric repeats. Binding of the hTer template region by GRN163L blocks the
biogenesis of an active telomerase RNP and results in progressive telomere shortening,
cellular senescence or apoptosis, and inhibition of cancer proliferation – either alone or in
combination with standard therapies – in a variety of in vitro and mouse cancer models
(Asai et al., 2003; Dikmen et al., 2005; Joseph et al., 2010; Tamakawa et al., 2010).
Modification of the GRN163L oligo backbone via an N3’ to P5’ thio-phosphoramidate
(NPS) transition stabilizes oligonucleotide-hTER duplex formation, and addition of a
lipid group at 5’ terminus of GRN163L facilitates cellular and tissue penetration.
Currently, GRN163L is undergoing extensive Phase I/II clinical testing in breast cancer,
lung cancer, multiple myeloma, and chronic myeloproliferative diseases. Preliminary
reports cite cytopenias, prolonged clotting, gastrointestinal side effects, fatigue, and
peripheral neuropathy as the most common toxicities (Fruh et al., 2008; M. Kozloff,
2010). As it proceeds towards additional Phase II and upcoming Phase III trials,
Imetelstat continues to be a highly promising agent and is among the most highly
developed of the telomerase therapeutics. One theoretical concern about this drug stems
from its mechanism of action: After Imetelstat inhibited telomerase in some preclinical
studies, multiple cell divisions with progressive telomere shortening had to occur over
several weeks before inhibition of cancer proliferation was observed (Herbert et al., 2002;

8
Shammas et al., 2008). This “phenotypic delay” raises the possibility that some cancer
cells might have the opportunity to develop resistance mechanisms such as upregulation
of TERT or alternative maintenance of telomeres via recombination (Cesare and Reddel,
2010). Whether such phenomena will play a clinical role and impact the efficacy of
Imetelstat will soon be addressed in additional Phase II and upcoming Phase III trials.
Telomerase interference. A different approach which directly targets telomerase
involves telomerase interference – altering the template region of hTer to reprogram the
RNP. This strategy, which was initially developed in the laboratory of Elizabeth
Blackburn as a tool to dissect telomerase reverse transcriptase function in ciliates (Yu et
al., 1990), was subsequently noted to exert an inhibitory effect on cancer cells (Goldkorn
and Blackburn, 2006; Guiducci et al., 2001; Kim et al., 2001; Li et al., 2004;
Marie-Egyptienne et al., 2009; Xu et al., 2010b). Specifically, endogenous wild-type hTer
is depleted using a short hairpin RNA knockdown, and at the same time, an hTer with a
mutated template region (MT-hTer) is ectopically introduced in its place. We and others
have shown that MT-hTer is incorporated into active telomerase in cancer cells, where it
essentially “reprograms” the enzyme to add incorrect telomeric tandem repeats. These
altered telomeric repeats are recognized as “uncapped” telomeres, eliciting a rapid DNA
damage response and apoptotic cascade, culminating in inhibition of proliferation
(Goldkorn and Blackburn, 2006; Li et al., 2004; Stohr and Blackburn, 2008). A potential

9
strength of telomerase interference is its immediate, dominant effects: Telomerase
reprogramming is not dependent on subsequent telomeric shortening and therefore has an
almost immediate effect on cancer cells by uncapping their telomeres within 1 or 2 cell
divisions, manifested by significant apoptosis and growth inhibition within 48 hours of
treatment. Moreover, cancer cells cannot upregulate TERT expression as a resistance
mechanism, because increased levels of TERT actually potentiate the pro-apoptotic
effects of MT-hTer by offering more enzyme to reprogram and thus even more dramatic
telomeric uncapping. On the other hand, the effects of telomerase reprogramming may be
so rapid and pervasive as to raise concerns about telomeric uncapping and toxicity in
normal stem cells that rely on telomerase activation to sustain progenitor tissue
compartments. A second, more practical obstacle is the challenge of effective systemic
delivery, as telomerase reprogramming currently is achieved via expression of the entire
451 nt MT-hTer from a DNA plasmid, making this an ineffective approach for systemic
treatment. The challenges of systemic delivery and possible stem cell toxicity are being
addressed in ongoing studies; our group has recently validated murine-targeting MT-mTer
and shRNA constructs (Xu et al., 2010b) that are being used to address these questions.

2. Targeting telomerase as a unique cancer marker
Telomerase expression and activity are high in most cancer types but low in benign,

10
differentiated cells, a specificity which has been exploited diagnostically and
prognostically by quantifying telomerase in primary tumor tissues and metastases, and
more recently in peripheral blood circulating tumor cells (Bravaccini et al., 2005; Hiyama
et al., 2000; Marchetti et al., 1999; Meeker, 2006; Poremba et al., 1999; Shay and
Bacchetti, 1997; Streutker et al., 2001; Wright et al., 1996; Xu et al., 2010a). Similarly,
the cancer-specificity of telomerase (TERT) expression has been exploited therapeutically
as a homing beacon for immune and virus mediated strategies.
Immunotherapy. Telomerase positive cancer cells display TERT peptide fragments
on their surface in association with MHC Class I molecules; hence TERT vaccines aim to
break immune tolerance and induce a TERT-specific cytotoxic T lymphocyte (CTL)
response. There are two main vaccine strategies: 1. direct inoculation with antigen or 2.
ex-vivo activation of autologous antigen presenting cells (APC, dendritic cells) or of B
lymphocytes.
1. Direct inoculation with antigen: Most clinically advanced in this group is GV1001
(GemVax, Denmark) a TERT-derived p611-626 16-mer with the sequence
EARPALLTSRLRFIPK that binds to and is subsequently presented by major
histocompatibility complex (MHC) class I. Multiple phase I/II studies have been
conducted in pancreatic cancer, malignant melanoma, and non-small-cell lung carcinoma
wherein patients received intradermal injections of GV1001. Toxicities were relatively

11
limited (local pain and inflammation at injection site, fevers, chills), and a majority of
patients developed quantifiable immune responses (CTL which recognize TERT) with a
suggestion of prolonged survival in immune responders (Brunsvig et al., 2006; S. Aamdal,
2006). However, to date, larger follow-up Phase II and III studies combining GV1001
with single agent chemotherapy in hepatocellular and pancreatic cancer, have
demonstrated no survival benefit (Greten et al., 2010; T. Buanes, 2009). Another Phase
III study of GV1001 in advanced pancreatic cancer (in combination with gemcitabine and
capecitabine) is still ongoing. A similar approach using a TERT-derived p540-548
ILAKFLHWL peptide also is currently in early phase clinical trials, where it has been
well-tolerated and was shown to generate hTERT-specific CD8
+
CTL, with tumor
infiltration and partial tumor regression observed in some cases (Brunsvig et al., 2006;
Domchek et al., 2007; Minev et al., 2000; Vonderheide et al., 2004; Vonderheide et al.,
1999).
2. Ex-vivo pulsing of antigen presenting cells (APC, Dendritic Cells) or of B
lymphocytes: Here, autologous immune cells are isolated from the cancer patient,
activated ex-vivo with TERT-derived peptides and reinfused into the patient. Several
Phase I/II trials using this approach have reported good tolerability, induction of immune
response (TERT-targeting CTL), and some instances of disease stabilization. A dendritic
cell approach using RNA based ex vivo activation (GRNVAC1, Geron Corp.), offers the

12
advantage of encoding multiple epitopes (compared to one epitope with peptide pulsing),
thus extending the scope of vaccination to strengthen the immune response (Su et al.,
2005). In a phase I/II trial in patients with prostate cancer, GRNVAC1 was well tolerated
and hTERT-specific CD8
+
cells were detected in 19 out 20 patients (Su et al., 2005). In a
similar approach using ex vivo DNA-pulsed autologous B lymphocytes, there were no
observed toxicities and vaccination induced TERT specific T cell responses (F. E. Millard,
2004).
In summary, TERT-targeting vaccine strategies have been aggressively developed in
the past decade, driven by enthusiasm for TERT’s specificity and ubiquity across a
majority of malignancies. Early Phase I/II trials have demonstrated that these vaccines
are able to break tolerance and activate a TERT-specific immune response, resulting in
tumor infiltration and a suggestion of clinical response in some cases. These promising
results are tempered by preliminary data from early phase III trials which failed to show
clinical benefit. There are two possible hurdles that may underlie these modest clinical
results: 1. Cancer patients are relatively immune suppressed, a condition attributed to the
cytokine milieu elaborated to varying degrees by their tumors; therefore, some patients
may have difficulty mounting a clinically significant response to the vaccine, a factor
which likely contributes to the modest success of most cancer vaccines to date. 2. The
absolute level of TERT even in telomerase positive cancer cells are quite low, on the

13
order of 100 molecules per cell (Cohen et al., 2007; Yi et al., 2001) ; therefore, even if
CTL are activated by the vaccine, the levels of TERT peptides displayed on tumor cells
may not constitute a sufficient “homing beacon” for a clinically significant immune
response. Newer vaccine strategies aimed at maximizing the immune response are
currently in preclinical development, and additional Phase III trials are ongoing which
hopefully will validate this approach.
Oncolytic viruses. A variety of preclinical viral and suicide gene strategies have
been developed to exploit the active hTERT promoter in cancer cells. Of these, the
furthest advanced towards clinical development is the telomerase specific oncolytic virus.
A conditionally replicative adenovirus is created by inserting the adenovirus E1A and
E1B genes downstream of the hTERT promoter, thus inducing adenoviral replication and
cellular lysis in a tumor specific manner (Irving et al., 2004). Telomelysin (OBP-301) is
the first telomerase specific oncolytic adenovirus to enter phase I study. Patients with
various solid tumors were administrated a single intratumoral injection of Telomelysin,
which was associated with grade 1 and 2 toxicities (pain at the injection site, fevers,
chills). Out of 16 enrolled subjects, 13 were shown to have viral DNA in plasma, and one
patient experienced partial response at the injected malignant lesion at day 56 after
injection (Nemunaitis et al., 2010). These early results are promising, and the presence of
viral DNA in plasma suggests a potential for therapeutic benefit beyond the local

14
intratumoral injection site; however, the clinical benefit of this approach in the metastatic
disease setting currently awaits further clinical testing.

3. Telomere-based therapeutic strategies
G-quadruplex stabilizers: The G-rich (TTAGGG) single stranded 3’ overhangs of
telomeres have been observed to form intramolecular four-stranded ribbon-like structures
termed G-quadruplexes (Smith and Feigon, 1992). Such structures, when stabilized by
small molecular compounds, prevent access of telomerase to telomeres, hence inhibiting
the canonical telomere lengthening and capping process. To exploit the potential
telomeric uncapping and shortening of the G-quadruplex phenomenon, several
G-quadruplex stabilizing agents were developed and shown to exert significant antitumor
efficacy both in vitro and in vivo, such as TMPyP4 (Mikami-Terao et al., 2008), RHPS4
(Leonetti et al., 2008), BRACO-19 (Burger et al., 2005; Gowan et al., 2002), and
Telomestatin (Doi et al., 2011; Waki et al., 2010). Studies of RHPS4 and Telomestatin
have demonstrated displacement of shelterin components (Pot1 and TRF2) from
telomeres associated with the uncapping DNA damage response (Gomez et al., 2006a;
Gomez et al., 2006b; Salvati et al., 2007; Tahara et al., 2006). However, to date, no
telomere-specific G-quadruplex agents have entered clinical trials. One general concern
about this class of agents is a potential lack of specificity, because the G-quadruplex

15
structure is not unique to telomeres and is shared by other genomic entities such as the
c-MYC promoter, the VEGF promoter, and guanine-rich genomic sequences (Eddy and
Maizels, 2006; Siddiqui-Jain et al., 2002; Sun et al., 2005). Accordingly, it is worth
noting that the only G-quadruplex targeting agent to have entered clinical testing,
CX-3543 (Quarfloxin, Cylene Pharmaceuticals), in fact does not target telomeric
G-quadruplexes but rather is designed to disrupt nucleolin/rDNA G-quadruplex
complexes. Preclinical studies of CX-3543 showed anticancer efficacy that was not
associated with altered telomere biology (Drygin et al., 2009), but rather with disruption
of nucleolin-rDNA interaction and inhibition of rRNA biosynethesis. This agent has
entered Phase I/II clinical trials, and the telomere-specific G-quadruplex stabilizers (i.e.
Telomestatin) are due to begin Phase I testing shortly.
Telomeric oligos: In an effort to mimic the telomeric uncapping and DNA damage
induced by telomerase interference (MT-hTer) and by G-quadruplex stabilizers,
investigators have attempted to treat cancer cells with DNA oligonucleotides homologous
to the uncapped TTAGGG telomeric repeats (Eller et al., 2002; Saretzki et al., 1999).
Introduction of so-called T-oligos directly into cancer cells has induced apoptosis,
autophagy, senescence, both in vitro and in vivo (Aoki et al., 2007; Puri et al., 2004;
Tsolou et al., 2008; Yaar et al., 2007). While these inhibitory effects were shown to be
more pronounced in cancer than in benign cells, their degree of cancer-specificity is the

16
subject of continuing investigation, and they have not as yet advanced into the clinical
trial arena.

Telomerase and cancer stem cells
The discovery of CSC and the recognition of their potential role in cancer formation
and progression have prompted a broad search for therapies capable of targeting this
cancer phenotype (Gupta et al., 2009a; Maitland and Collins, 2008; Visvader and
Lindeman, 2008). Intriguingly, telomerase targeting may offer this therapeutic benefit.
Traditionally, telomerase activity has been considered a nearly universal characteristic of
cancer cells, an assumption that may exist because early surveys of telomerase activity
were conducted indiscriminantly from lysates of entire cancer populations (Shay and
Bacchetti, 1997), and because the oncogenic role of telomerase was demonstrated by
ectopically introducing the enzyme into unselected cell populations (Bodnar et al., 1998;
Hahn et al., 1999a). Contrary to this model of homogeneous telomerase activation,
studies of normal tissue stem cell compartments have demonstrated a unique role for
telomerase in stem cell activation (Lee et al., 1998; Sarin et al., 2005), raising the
possibility that perhaps telomerase plays a parallel unique role in CSC. In support of such
a role, ectopic overexpression of telomerase in cancer cell lines has indeed been shown to

17
enhance tumor initiation, perhaps reflecting a potentiation of the CSC phenotype (Gu et
al., 2007; Stewart et al., 2002).
Our group has investigated the relative telomerase activity and expression levels
within CSC and non-CSC subpopulations isolated (using flow cytometry) from freshly
resected human prostate tumors and from prostate cancer cell lines. Remarkably, both in
tumors and cell lines, CSC possessed markedly elevated levels of telomerase expression
and activity compared with non-CSC. Moreover, induction of telomerase interference via
ectopic expression of MT-hTer/siRNA effectively reprogrammed the active telomerase of
prostate CSC to induce rapid apoptosis and abrogate tumor initiation. Hence, these results
demonstrated that telomerase expression and activity may not be a uniform phenotype
common to all cancer cells, but rather may be concentrated in a subpopulation of cells
with CSC-like properties, which in turn renders these cells exceedingly susceptible to
telomerase interference. Our observations were consistent with findings from a handful of
other recent studies: Elevated telomerase activity was observed in CSC-like
subpopulations in breast and lung cancers (Ho et al., 2007; Lin et al., 2010), and the
telomerase inhibitor GRN163L (Imetelstat, Geron Corp.) was shown to effectively inhibit
the proliferation of CSC-like cells in prostate cancer and glioblastoma models (Marian et
al., 2010a; Marian et al., 2010b). Collectively, these studies suggest that targeting
telomerase may constitute a promising new therapeutic strategy for neutralizing CSC.

18
Recent findings have added an intriguing new twist to the CSC telomerase story.
Contrary to historical convention, it now appears that telomerase’s function in normal
stem cells and possibly in cancer is not limited to its known, canonical
telomere-maintenance catalytic function but may also involve a newly-discovered,
non-canonical direct signaling role. For example, activation of TERT in mouse hair
follicle stem cells induced active proliferation and differentiation even in the absence of
telomerase RNA (Ter) and even with use of a catalytically dead TERT; that is, without
telomere lengthening activity (Sarin et al., 2005). These effects were associated with
protein interactions and gene transcriptional events involving the Myc and Wnt/ β-catenin
pathways (Choi et al., 2008). Similarly in cancer, ectopic expression of a catalytically
dead TERT in fibroblasts which had been immortalized with SV40 and H-Ras rendered
the cells highly tumorigenic, even though this TERT variant lacked the capacity to
lengthen telomeres (Stewart et al., 2002). Our own preliminary data supports a
non-canonical role in cancer and possible in CSC: In our studies, ectopic expression of
TERT (wild-type or catalytically dead) rapidly and significantly expanded the size of the
CSC subpopulation (unpublished data), suggesting a telomere-independent signaling role
whereby telomerase potentiates the CSC phenotype. Thus, it may be reasonable to hope

19
that telomerase-targeting strategies may provide anti-CSC benefit not only by
interrupting the canonical, telomere-maintaining role of telomerase, but also by
disrupting its non-canonical, signaling function.

Phenotypic plasticity in Cancer stem cells
The resistance to chemotherapeutics is believed to be one of the most important
features of CSC, in which, highly expressed ATP-binding cassette (ABC) drug
transporters (such as ABCG2, ABCB1 etc.) can actively pump out the chemotherapeutic
agents and may potentially serve as an important mechanism to protect CSC (Dean et al.,
2005). Therefore, CSC can be identified as a “Side population” (SP) in flow cytometry
(Goodell et al., 1997; Patrawala et al., 2005), by efflux of Hoechst dye through an
ABCG2 mediated fashion (Zhou et al., 2001). SP cells isolated from various human
(Hirschmann-Jax et al., 2004; Wu et al., 2007) and murine (Bleau et al., 2009) tumors
have shown CSC properties and are more tumorigenic after inoculating into
immunodeficient mice compared to non-SP (NSP) cells. Moreover, SP cells isolated
from primary tumors and cancer cell lines show much higher resistance to chemotherapy
than their NSP counterparts (Chikazawa et al., 2010; Hu et al., 2010; Singh et al., 2010;
Steiniger et al., 2008).

20
One of the most fundamental questions in the rapidly evolving field of CSC —
“when and how do those drug resistant and tumorigenic CSC arise?” still remains
unanswered. Two general possibilities exist: The first postulates that CSC arise through
mutational events, perhaps in normal stem cells, and then function as a small seed
population capable of generating and replenishing tumor. The appeal of this model lies in
its intuitive analogy to normal stem cell function; indeed, CSC – like normal stem cells –
have been shown to proliferate slowly, express self-renewal genes, and possess efflux
pumps (Polyak and Hahn, 2006). The alternative possibility is that CSC are not at all a
static subpopulation, but rather a dynamic phenotype that can be displayed by any cancer
cell given the right stimulus (Gupta et al., 2009a; Polyak and Hahn, 2006; Reya et al.,
2001). In support of this hypothesis, it has been shown in non-cancer models that
differentiated cells can be reprogrammed into a so-called “induced progenitor state” (iPS)
by ectopic overexpression of several genes (Takahashi et al., 2007; Yu et al., 2007b).
Similarly, differentiated cancer cells induced with cytokines to undergo epithelial to
mesenchymal transformation (EMT) take on a phenotype with many similarities to CSC
(Mani et al., 2008).
In my study, I investigated when and how CSC arise by characterizing and tracking
CSC and non-CSC subpopulations in vitro and in vivo over time. For this purpose, we
analyzed cancer cells using flow cytometry with Hoechst dye exclusion, a

21
commonly-used method that yields a side population (SP) of cells with CSC properties.
Remarkably, when this approach was coupled with GFP labeling, we were able to
observe a dynamic two-way equilibrium between the SP and non-SP (NSP)
subpopulations in cell culture and in tumor xenografts. Specifically, the SP subpopulation
first became depleted by differentiation into NSP cells, and subsequently the SP
subpopulation was reconstituted by conversion of NSP cells back to the SP phenotype;
these transitions occurred spontaneously in the course of proliferation without exogenous
gene expression or cytokine application.
My study suggests that CSC may not comprise a static seed population from which
cancer emanates; but rather a transient phenotype, a two-way street through which a
subset of all cancer cells may pass at a given time. The proportion of cancer cells that
undertake the CSC phenotype is variable and likely governed by a combination of
environmental conditions and cellular signals that are as yet to be elucidated. Therefore,
current strategies to eradicate the CSC subpopulation may only have temporary effects,
because the CSC phenotype could be replenished from the non-CSC population. Hence,
recognizing this dynamic equilibrium and elucidating the mechanisms that govern it may
ultimately recast the CSC phenotype in a new light and lead to more effective cancer
therapies.

22
Signaling pathway regulating cancer plasticity
The PI3K/Akt pathway is a commonly activated pathway that controls the survival
and cell cycle regulation of cancer cells, and therefore the development of therapeutic
reagents targeting PI3K/Akt signaling as well as its downstream mediators appears to be
a hotspot in clinical trials for cancer treatment (Engelman, 2009; Liu et al., 2009). Several
recent reports have shown that pharmacological inhibition of this pathway could
effectively decrease a side population phenotype in cancer cells, which is consistently
associated with drug-resistant and tumorigenic abilities (Bleau et al., 2009; Zhou et al.,
2007). In addition, inhibition of PI3K/Akt activity sensitized cancer cells to
chemotherapeutic drugs in combination treatment (Dubrovska et al., 2010; Korkaya et al.,
2009), which are accompanied with significant decrease of drug-resistant, tumorigenic
cancer stem cell populations. These promising results have suggested the great potential
of directly targeting the major signaling pathways in elimination of drug resistant and
tumorigenic cancer populations. However, the important question of how these CSC
characteristics are regulated remains to be answered. Could the inhibition of PI3K/Akt
pathway specifically arrest the proliferation of cancer cells with particular traits (e.g.
drug-resistance, tumorigenicity)? Or, alternatively, could some major nodes in this
pathway function as switches, whose regulation may affect the transition of cancer cell
plasticity, which as a result, prevent cancer cells from reacquiring drug resistant and

23
tumorigenic properties, and sensitize them to drug response and decrease tumor
formation ability? More importantly, in a greater context of signaling network, what
would be the downstream mediators in PI3K/Akt pathway that may specifically control
the drug resistant and tumorigenic phenotypes?
To address these questions, we analyzed cancer cells in culture over time by using
flow cytometry to monitor the dynamic changes of the drug-resistant, highly
tumorigenic side population in response to the perturbation of PI3K/Akt pathway and its
downstream mediators. Interestingly, in our GFP-labeling experiment, we observed that
the previously reported phenomenon of spontaneous regeneration of side population
from non-side population was significantly prevented by the treatment of PI3K
inhibitor—LY294002. At the same time, the viability of non-labeled cells (initially SP
cells) was barely affected, indicating that LY294002 treatment indeed blocked
phenotypic plasticity rather than impacting the proliferation of side population cells.
The PI3K/Akt pathway has been reported to mediate the signaling of Wnt/ β-catenin
pathway through the shared downstream factor—GSK3 β, and regulate tumor
progression in various cancer types (Castellone et al., 2009; Korkaya et al., 2009; Song
et al., 2009). Therefore, we hypothesized that the downstream GSK3 β/ β-catenin axis
may play an important role in regulating drug-resistant, highly tumorigenic side
population phenotype. By inhibiting GSK3 β pharmacologically, we observed

24
accumulation of nuclear β-catenin, as well as significant increase of side population
cells, while specific knockdown of β-catenin by siRNA greatly diminished side
populations. Further downstream, cAMP response element-binding protein
(CREB)-binding protein (CBP), which is a co-activator recruited by β-catenin to form
active transcriptional complex, has been reported as critical factor in long-term
maintenance of embryonic stem cells (Miyabayashi et al., 2007; Takahashi-Yanaga and
Kahn, 2010). Moreover, CBP functions as a crucial co-activator for TCF/ β-catenin
mediated survivin transcription (Ma et al., 2005), whose expression is important for
drug resistance and cancer relapse (Park et al., 2011; Shoeneman et al., 2012). Using a
small molecular antagonist—ICG001 to specifically interrupt β-catenin/CBP interaction
(Emami et al., 2004), we showed that the drug resistant, highly tumorigenic side
population could be significantly decreased. Using another antagonist —IQ-1 to inhibit
the other related co-activator P300, which has been reported to play a differential role in
regulating differentiation and drug resistance (Ma et al., 2005), to potentiate
β-catenin/CBP signaling predictably increased side population. To determine whether
the newly generated side population cells are directly from conversion of non-side
population, we purified NSP cell by FACS, and treated them with pharmacological
reagents targeting different nodes in the PI3K/Akt/ β-catenin signaling hierarchy. We
demonstrate that activation of β-catenin by inhibiting GSK3 β significantly induced the

25
regeneration of side population cells, while interrupting the downstream β-catenin/CBP
signaling prevented SP conversion from NSP, even with the activation of nuclear
β-catenin. At the same time, proliferation assay results have shown that SP cells
exhibited little difference in growth inhibition from NSP cells under the treatment of
pharmacological inhibitors, indicating that that these inhibitors did not selectively target
side population cells. The proliferation experiments also helped to rule out the
possibility of SP cell contamination during NSP sorting. Taken together, my studies
have shown, for the first time, that PI3K/Akt/ β-catenin/CBP signaling regulates the
phenotypic plasticity of CSC, rather than solely affect their viability and maintenance;
therefore, targeting members of this pathway may potentially prevent the reacquisition
of drug-resistant and tumorigenic abilities in cancer cells, providing an opportunity to
sensitize cancer cells to conventional therapies, and enabling more effective cancer
eradication.

26
Chapter 2
Prostate tumor cells with cancer stem-like properties have
high telomerase activity and are rapidly killed by telomerase
interference

Abstract
Cancer Stem cells (CSC) have been postulated to promote treatment resistance and
disease progression in prostate and other malignancies. We investigated whether the
enzyme telomerase, which is active in cancer cells and in normal stem cells, plays an
important role in CSC which can be exploited to neutralize these cells.
We used flow cytometry and assays of gene expression, clonogenicity and
invasiveness to isolate and characterize a putative CSC subpopulation from
freshly-resected human prostatectomy specimens. Telomerase activity was measured by
qPCR-based Telomeric Repeat Amplification Protocol (TRAP). Telomerase interference
was achieved by ectopic expression of a mutated telomerase RNA construct which
reprograms telomerase to generate “toxic” uncapped telomeres. Treated cells were
assayed for apoptosis, proliferation in culture, and xenograft tumor formation.
CSC in prostate tumors expressed elevated levels of genes associated with a
progenitor phenotype and were highly clonogenic and invasive. Significantly, CSC

27
telomerase activity was 20 to 200-fold higher than in non-CSC from the same tumors,
and CSC were exquisitely sensitive to telomerase interference which induced rapid
apoptosis and growth inhibition. Similarly, induction of telomerase interference in
highly-tumorigenic CSC isolated from a prostate cancer cell line abrogated their ability to
form tumor xenografts.
Human prostate tumors contain a CSC subpopulation with markedly elevated
telomerase activity which renders them acutely susceptible to telomerase interference.
These findings offer the first tumor-derived and in vivo evidence that telomerase may
constitute a CSC “Achilles heel” which may ultimately form the basis for more effective
new CSC-targeting therapies.

Introduction
Cancer Stem cells (CSC) are self-renewing, highly tumorigenic cancer cells recently
identified in a broad spectrum of malignancies and implicated in tumor formation,
therapy resistance and disease dissemination (Gupta et al., 2009a; Visvader and
Lindeman, 2008). In prostate, as in other cancers, various strategies have been developed
to enrich and isolate CSC from tumors, including flow cytometry for specific cell surface
markers (Collins et al., 2005; Collins et al., 2001; Dubrovska et al., 2009; Hurt et al.,
2008; Klarmann et al., 2009; Miki et al., 2007; Patrawala et al., 2006; Vander Griend et

28
al., 2008), isolation of a side population (Brown et al., 2007), and formation of cell
spheres (Dubrovska et al., 2009) or holoclones (Li et al., 2008a; Marian et al., 2010b). In
addition, cancer cell lines, human xenograft tumors (Dubrovska et al., 2009; Hurt et al.,
2008; Klarmann et al., 2009; Miki et al., 2007; Patrawala et al., 2006) and mouse models
(Liao et al., 2010; Mulholland et al., 2009) have also been commonly used to characterize
prostate CSC. The mostly robust prostate CSC markers to emerge from this collective
work have been integrin α2β1, CD44 and CD133 (Collins et al., 2005; Collins et al.,
2001; Dubrovska et al., 2009; Hurt et al., 2008; Klarmann et al., 2009; Miki et al., 2007;
Patrawala et al., 2006; Vander Griend et al., 2008); these cell surface antigens enrich for a
subpopulation of prostate cancer cells with an embryonic stem-like gene expression
profile and with high clonogenic, metastatic, and tumorigenic potential relative to
marker-negative cells (Collins et al., 2005; Collins et al., 2001; Cunha et al., 1983;
Guzman-Ramirez et al., 2009; Hurt et al., 2008; Klarmann et al., 2009; Miki et al., 2007;
Patrawala et al., 2006; Vander Griend et al., 2008).
In light of CSC’s high tumorigenicity and therapy resistance, it has been suggested
that truly effective new cancer treatments should specifically aim to target this tumor
subpopulation (Gupta et al., 2009a; Maitland and Collins, 2008; Visvader and Lindeman,
2008); however, there are as yet no effective CSC-targeting treatments. Telomerase
activity is a recognized hallmark of malignancy which has been explored extensively by

29
our group and others for its therapeutic and biomarker potential (Goldkorn and Blackburn,
2006; Shay and Wright, 2006; Xu et al., 2010a; Xu et al., 2010b). Whereas benign,
terminally differentiated tissues have extremely low levels of telomerase (Wright et al.,
1996), malignant cells from a variety of cancers have significant telomerase expression
and activity that correlate with malignant potential and high tumor initiating ability (Gu
et al., 2007; Shay and Bacchetti, 1997; Stewart et al., 2002). In addition to its role in
cancer, telomerase recently has been found to play an equally important function in
normal stem cell function, inducing stem cell activation in human and mouse epidermal,
gastrointestinal, hematopoietic, neuronal, and reproductive niches (Mimeault and Batra,
2009). Given its dual roles in carcinogenesis and stem cell activation, we speculated that
perhaps telomerase activation plays an equally critical role in CSC and hence may
constitute an attractive therapeutic strategy for targeting these cells.
To address this question, we investigated the relative telomerase activity and
expression levels within cancer cell subpopulations isolated from freshly resected human
prostate tumors and from prostate cancer cell lines. Specifically, we used FACS to isolate
integrin α2β1
high
CD44
+
cells, previously reported to possess a cancer stem-like phenotype
(Collins et al., 2005; Collins et al., 2001; Garraway et al., 2010; Hurt et al., 2008; Patrawala
et al., 2007; Patrawala et al., 2006). When isolated from prostate tumors in our studies,
these cells expressed higher levels of self-renewal genes and were more clongenic and

30
invasive in vitro than α2 β1
high
CD44
-
cells from the same tumors. Similarly, when isolated
from the DU145 prostate cancer cell line, α2β1
high
CD44
+
cells expressed CSC-like
properties in vitro, as well as high tumor initiation in vivo compared to α2 β1
high
CD44
-
cells.
Remarkably, both in tumors and cell lines the putative CSC ( α2β1
high
CD44
+
) possessed
markedly elevated levels of telomerase expression and activity compared with bulk
non-CSC ( α2β1
high
CD44
-
). Therefore, we tested whether CSC were susceptible to
telomerase interference, a therapeutic strategy that specifically exploits the presence of
high telomerase activity (Goldkorn and Blackburn, 2006; Li et al., 2004;
Marie-Egyptienne et al., 2009; Xu et al., 2010b). Telomerase interference is a two-pronged
approach consisting of: 1. telomerase RNA with a mutated template region (MT-hTer), and
2. siRNA against endogenous wild-type telomerase RNA. Ectopic co-expression of these
two constructs from a single vector (MT-hTer/siRNA) effectively substitutes MT-hTer for
wild-type hTer, thus reprogramming telomerase to encode incorrect telomeres. The altered
“toxic” telomeres elicit a brisk DNA damage response and rapid apoptosis in cells with
high levels of active telomerase. In the present study, MT-hTer/siRNA effectively
reprogrammed the active telomerase of prostate CSC to induce rapid apoptosis and
abrogate tumor initiation, thus underscoring the therapeutic potential of targeting CSC
with telomerase interference.

31
Materials and Methods
Tissue collection, processing and cell culture. Prostate tumors freshly resected
from prostate cancer patients at USC Norris Comprehensive Cancer Center were
examined, inked, graded and staged by a pathologist in a de-identified, IRB-approved
protocol. Cells were obtained as described previously (Collins et al., 2005; Collins et al.,
2001) with minor modifications: Briefly, tissue was minced with scalpels and digested in
DMEM/F12 (50:50 mix) media supplemented with 8.75 µg/ml liberase blendzymes 3
(Roche Applied Science, Mannheim, Germany) and 1 µg/ml DNAse I (Invitrogen,
Carlsbad, CA) overnight at 37 °C in a shaker incubator. Epithelial organoids were
separated from the stromal fraction by unit gravity centrifugation and disaggregated into
single cell suspension by incubation with trypsin/EDTA (Mediatech, Manassas, VA) for 5
min at 37°C. A portion of the cells were stained and analyzed by flow cytometry. The rest
were plated on collagen coated plate (BD Pharmigen, San Diego, CA) and incubated at
37°C for 1 hr to enrich for an α
2
β
1
integrin
+
cell population as described previously
(Collins et al., 2001). Non-adherent cells were verified to be α
2
β
1
integrin
-
CD44
-
(Figure
1B). The adherent cells were collected by incubation for 5 min with accutase (Innovative
Cell Technologies, San Diego, CA) for FACS. Tumor cells were maintained on collagen
coated plate in CnT52 (PCT Prostate Epithelium Medium, Low BPE (Human), Millipore,
CA) at 37°C, 5% CO
2
. To further molecularly validate the cancer identity of CSC,

32
GSTP1 promoter methylation was confirmed in a subset of specimens and compared to
adjacent benign tissue (Table 1). Human prostate cancer cell line DU145 was cultured in
RPMI 1640 with fetal bovine serum (FBS, 10%, Omega) at 37°C, 5% CO
2
.
Flow cytometry. Fluorescence-activated cell sorting (FACS) analysis was used for
determination of marker expression by FACSAria (Becton Dickinson Immunocytometry
Systems). The following antibodies were used: integrin α2 (PE-CD49b) and FITC-CD44
(BD PharMingen, San Diego, CA); CD133/1 (allophycocyanin (APC) labeled; Miltenyi
Biotec, Inc., Auburn,CA). Single cell suspension was obtained from freshly resected and
digested prostate tumor samples as described above and resuspended in cell staining
buffer containing 2% FBS and 5mM EDTA. Cells were stained with the above antibodies
on ice for 20 min. Data were analyzed by FACSDiva.
Colony-forming assay and cell migration assay. To test clonogenicity, sorted cells
were seeded on collagen coated plates, and colonies were counted after 21 days. To test
invasiveness, cell migration assays were performed using Matrigel-coated 24-well inserts
(BD Pharmigen, San Diego, CA) per manufacturer’s instruction. Briefly, 10
4
cells were
placed in the upper chamber with CnT52 medium while the lower chamber was filled
with CnT52 containing 10 ng/ml SDF-1 (R&D systems, Minneapolis, MN).
Non-migrated cells in the upper chamber were removed following fixation and staining

33
of cells in the lower chamber. Migrated cells were analyzed using a Zeiss Imager.Z1
microscope with Axiovision software at 20X magnification.
RNA extraction, reverse transcription, and PCR. Total RNA was extracted by
RNAqueous-micro Kit (Applied Biosytems Inc, Foster City, CA). First strand cDNA was
synthesized using the RETROscript reverse transcription kit (Applied Biosytems Inc,
Foster City, CA). PCR primers were described previously (Patrawala et al., 2006; Wagner
et al., 2004) and in Table 2. GAPDH was used as internal control.
Quantitative PCR – Telomeric Repeat Amplification Protocol (qPCR-TRAP)
and telomere length assay. Telomerase activity from cell extracts was analyzed using
real-time PCR based telomeric repeat amplification protocol (TRAP) as described
(Herbert et al., 2006) and in Table 3. Briefly, cells were lysed in TRAPeze® 1X CHAPS
Lysis Buffer (Millipore, Temecula, CA). Cell lysate from 1000 cells was added for each
reaction for each sample. The iQ5 optical system software version 2.0 was used to
analyze the results. To determine telomere lengths, genomic DNA was extracted using
Qiagen DNeasy Blood & Tissue Kit (Qiagen), and relative telomere lengths were
analyzed in triplicate by real time PCR (Biorad MyIQ) as described previously using T
and S primers (Xu et al., 2010b).
Virus production and infection. Lentivirus was generated and cells were infected
as previously described (Goldkorn and Blackburn, 2006; Li et al., 2004). Briefly, 12 µg

34
of lentiviral vector, along with 3 µg of pMD.G and 9 µg of pCMVdR8.91 plasmids were
cotransfected into 293T cells by using the calcium phosphate co-precipitation method.
Virus-containing medium was harvested 48 and 72 hr after transfection and filtered
through 0.45 µm filter. Cells were seeded at 10
5
cells/well in 6-well plate overnight
before adding lentiviruses packaged from various constructs supplemented with 8 µg/ml
polybrene. After overnight infection, medium was changed and GFP signal was
confirmed at 48 hours post infection to ensure >90% transduction (>90% of cells GFP+).
Cell growth curve and apoptosis assay. Sorted cells were seeded at 10
5
cells/well
in 6 well plate and infected with control or MT-hTer/siRNA lentivirus. Cell proliferation
was measured by cell counting using a hemocytometer at subsequent time points after
infection. Apoptosis was analyzed at day 4 post infection with MT-hTer/siRNA by
TUNEL assays performed following the protocol described in In Situ Cell Death
Detection Kit, Fluorescein (Roche Applied Science, Indianapolis, IN) and analyzed on
BD LSR-II.
Subcutaneous tumor allografts. All experiments were approved and performed
following the rules of the Institutional Animal Care and Use Committees at University of
Southern California. 6-8 week old, male SCID mice were purchased from NIH. DU145
cells were infected overnight with control or MT-hTer/siRNA lentiviruses and cultured at
37°C, 5% CO
2
for 1 day after changing media. For each mouse, 5000 cells were

35
resuspended in media, mixed with 50 µl ice-cold matrigel (BD biosciences, San Jose, CA)
and placed on ice until inoculation. 1 ml insulin syringe was used for subcutaneous
inoculation onto the flank of each mouse (5 mice per treatment group). The growth of
tumors was observed and recorded as tumor volume by caliper measurement. Ninety days
after inoculation, mice were sacrificed, and tumors were resected and weighed.
Statistical analysis. Performed in collaboration with USC/Norris Biostatistics Core.
All experiments were conducted in triplicate with error bars representing standard
deviation around the mean. Student’s t-test was used to determine statistical significance
when comparing mean values at one point in time (e.g. gene expression, % apoptosis, cell
growth inhibition, relative telomerase activity in DU145 cells, tumor weights). Two-sided
Wilcoxon matched-pairs signed rank test was performed by using Graphpad Prism5.0
software to compare statistical significance of telomerase activity and telomerase (hTERT)
mRNA level for patients’ sample as well as tumor growth from DU145 cells over time.

36
Results
Prostate tumor cells expressing integrin α2 β1
high
CD44
+
have a cancer stem cell
(CSC) phenotype. Freshly resected human prostate tumors (Table 4 and Figure 1) were
disaggregated, digested, and dissociated into single cell suspensions as described
previously (Collins et al., 2005; Collins et al., 2001) and in Materials and Methods. Cells
were analyzed by FACS for expression of integrin α
2
β
1
(CD49b), CD44 and CD133,
surface markers widely reported in prostate cancer to enrich for CSC properties such as
elevated expression of self-renewal genes, high in vitro clonogenicity and invasiveness,
and increased tumor initiation after selection from xenografted human cell lines (Collins
et al., 2005; Collins et al., 2001; Hurt et al., 2008; Patrawala et al., 2007; Patrawala et al.,
2006). In all tumor samples, we found α2β1
high
CD44
+
cells, which constituted 0.7% to
9.2% of all cancer cells (Table 4 and Figure 1). In contrast, minimal CD133 expression
was observed (Figure 1), consistent with previous reports (Garraway et al., 2010). In all
experiments, transient collagen adherence (Collins et al., 2001) was used to enrich for
integrin α2 β1
+
cells, followed by FACS to further select for an α2β1
high
CD44
+

subpopulation; purity (>98%) was confirmed by re-analysis with FACS (Figure 1).
To confirm the previously-reported CSC properties of the tumor-derived
FACS-sorted cells, we conducted phenotypic analyses which revealed that – compared
with the non-CSC majority of tumor cells ( α2β1
high
CD44
-
) – the α2β1
high
CD44
+

37
subpopulation expressed significantly higher mRNA levels of genes associated with
self-renewal, proliferation and invasiveness, such as β-catenin, NANOG, Oct3/4, SMO
and Bmi1; conversely, mRNA levels of genes associated with a differentiated prostate
phenotype, such as androgen receptor (AR) and prostate specific antigen (PSA), were
significantly lower (Figure 2A). Functionally, α2β1
high
CD44
+
subpopulations from all 8
tested tumors proliferated in culture, with three of these generating discrete colonies at an
average of 44 colonies per 1000 cells seeded after 21 days in culture; in contrast,
α2β1
high
CD44
-
cells from all 8 tumors tested did not proliferate in culture and produced
no colonies (Figure 2B). Moreover, α2β1
high
CD44
+
cells from 4 of 8 tumors migrated
across the membrane in a matrigel invasiveness assay at an average of 39 migrated cells
per 5 high powered fields after 24 hours (interestingly, the same subset of tumors yielded
α2β1
high
CD44
+
cells which were both clonogenic and invasive); in contrast, none of the
α2β1
high
CD44
-
cells from all 8 tumors tested were invasive (Figure 2C).
CSC have high telomerase expression and activity that can be targeted with
telomerase interference. We measured relative telomerase activity from the putative
tumor-derived CSC subpopulations ( α2β1
high
CD44
+
) and non-CSC subpopulations
( α2β1
+
CD44
-
) using qPCR-TRAP. Notably, we found that CSC had significantly higher
(approximately 20-fold to 200-fold) telomerase activity than non-CSC in 6 of 6 tumors
tested (Figure 3A, p=0.03). CSC also had significantly higher mRNA levels of hTERT,

38
the chief determinant of telomerase activity (Figure 3B, p=0.008), whereas their mean
bulk telomere lengths were not significantly different from those of non-CSC (Figure
3C).
Having observed the markedly elevated telomerase expression and activity of
tumor-derived CSC, we reasoned that this subpopulation may be particularly susceptible
to telomerase interference, ectopic introduction of telomerase RNA with a mutated
template region (MT-hTer/siRNA) that reprograms the telomerase enzyme to add
incorrect telomeres, resulting in telomeric uncapping, DNA damage, and apoptosis
(Goldkorn and Blackburn, 2006; Li et al., 2004; Xu et al., 2010b). We infected the
tumor-derived CSC with lentivirus expressing either MT-hTer/siRNA or vector control.
Two days after infection, MT-hTer expression by qPCR was ~6-fold compared to vector
control (Figure 4A), and a modified qPCR-TRAP assay to detect mutant telomeric
repeats showed a 5-fold increase in mutant telomerase activity (Figure 4B); together,
these data confirmed successful MT-hTer/siRNA expression and reprogramming of
telomerase activity in the CSC subpopulation. Importantly, telomerase interference
resulted in 80% apoptosis by day 4 of infection and ~95% cell growth inhibition by day 6
(Figure 4C-D). Hence, the high telomerase expression and activity of tumor-derived CSC
rendered them acutely susceptible to telomerase interference.

39
CSC derived from a prostate cancer cell line have high tumorigenicity that is
abrogated by telomerase interference. Having observed the marked efficacy of
telomerase interference in vitro, next we tested whether this approach could also inhibit
tumor formation in vivo. Currently there is no robust model for direct implantation and
growth of human tumor-derived prostate cancer cells; therefore, we investigated tumor
initiation in vivo using α2β1
high
CD44
+
cells isolated from the DU145 prostate cancer cell
line, a strategy previously shown to enrich for CSC properties (Hurt et al., 2008;
Patrawala et al., 2007; Patrawala et al., 2006). Consistent with previously published cell
line data (Hurt et al., 2008; Patrawala et al., 2007; Patrawala et al., 2006) and similar to
our primary tumor results, the DU145-derived putative CSC subpopulation
( α2β1
high
CD44
+
) had significantly higher self-renewal gene expression (Figure 5A) and
was significantly more clonogenic and invasive (Figure 5B) than the non-CSC
( α2β1
high
CD44
-
) subpopulation. Further mirroring our findings in primary tumors,
telomerase expression and activity of DU145-derived CSC were double those of
non-CSC (Figure 5C), and telomere lengths were not significantly different (Figure 5D).
Next, we tested the in vitro effects of telomerase interference (MT-hTer/siRNA) in the
CSC fraction relative to the non-CSC fraction of DU145 cells, a direct comparison which
had not been possible with primary tumor derived cells, because primary tumor-derived
non-CSC did not propagate in culture. Notably, ectopic expression of MT-hTer/siRNA in

40
DU145-derived CSC caused a significant 3-fold inhibition of proliferation relative to
vector control by day 6 after lentiviral infection (Figure 6A). In contrast, proliferation of
DU145-derived non-CSC

was not significantly inhibited by telomerase interference,
possibly because the lower telomerase expression and activity of these cells provided less
substrate to be reprogrammed by MT-hTer/siRNA (Figure 6A).
We investigated the in vivo impact of telomerase interference on the tumor initiating
ability of CSC and non-CSC subpopulations derived from the DU145 cell line.
MT-hTer/siRNA or vector control were ectopically overexpressed by lentiviral infection
in CSC and non-CSC, and 5000 cells were inoculated subcutaneously into SCID mice (4
groups, 5 mice per group). In the vector control groups, CSC generated measurable
tumors in 5 of 5 mice by day 45 with mean wet weight of 0.72 g at excision on day 90,
whereas non-CSC generated measurable tumors in only 3 of 5 mice by day 61 with mean
wet weight of only 0.32 g at excision on day 90 (i.e. fewer and smaller tumors with
greater lag time), thus confirming the greater baseline tumorigenicity of CSC relative to
non-CSC (Figure 6B, C). Notably, both subpopulations formed no tumors over 90 days of
follow-up when ectopically expressing MT-hTer/siRNA, suggesting that telomerase
interference effectively abrogated tumor initiation in vivo (Figure 6B, C).

41

Table 1. GSTP-1 promoter methlylation measurements on prostate tumor samples. GSTP-1
promoter methylation analysis in 3 primary CSC specimens derived from primary prostate tumors and
from 1 tumor-adjacent benign tissue specimen confirmed high methylation (PMR value) in the CSC
samples consistent with a prostate cancer molecular phenotype. Genomic DNA (500-1000 ng for each
sample) was treated with bisulfite using the Zymo EZ DNA methylation kit (Zymo Research, Orange,
CA) as described by the manufacturer. The DNA methylation status of the GSTP1 promoter region
was determined using MethyLight, a TaqMan-based, real-time PCR assay as previously described
(Eads et al., 2000; Eads et al., 1999; Weisenberger et al., 2006). The GSTP1 MethyLight oligomers
were described previously (Eads et al., 2001), except the probe contained a 5’-FAM fluorophore and a
3’-BHQ-1 quencher. In addition, we used a control reaction based on ALU repeats as a normalization
control reaction (Weisenberger et al., 2005). The GTSP1 DNA methylation score for each sample is
calculated as a Percent of Methylated Reference (PMR) as described (Weisenberger et al., 2006).

Sample ID GSTP1-M1 (HB-72)
Ct value
PMR
Benign adjacent None 0.0
1 29.8 105.5
2 29.6 89.5
3 29.3 124.2

42

Table 2. Sequences of PCR primers

Gene Forward primer (5’ – 3’) Reverse primer (5’ – 3’)
GAPDH ACCACAGTCCATGCCATCAC TCCACCACCCTGTTGCTGTA
CD44 TCCATCAAAGGCATTGGGCAG AACCTGCCGCTTTGCAGGTGT
AR GAAGCCATTGAGCCAGGTGT TCGTCCACGTGTAAGTTGCG
PSA GGTGACCAAGTTCATGCTGTG GTGTCCTTGATCCACTTCCG
β-catenin ACTGGCAGCAACAGTCTTACC TTTGAAGGCAGTCTGTCGTAAT
NANOG CAACTGGCCGAAGAATAGCA GCAGGAGAATTTGGCTGGAA
Oct3/4 ACACCTGGCTTCGGATTTCG GGCGATGTGGCTGATCTGCT
SMO ATCTCCACAGGAGAGACTGGTTCGG AAAGTGGGGCCTTGGGAACATG
Bmi1 GGAGACCAGCAAGTATTGTCCTTTTG CATTGCGCTGGGCATCGTAAG

43

Table 3. Oligo sequences and protocol for qPCR-TRAP assay
Cells were counted and lysed in TRAPeze® 1X CHAPS Lysis Buffer (Millipore, Temecula, CA) for
30 min on ice, centrifuged at 4 °C for 20 min to remove cell debris to collect cell lysates. Cell
lysates from equal numbers of cells were added for each reaction. For the first (extension) step, TS
oligonuleotides and dNTPS were added to the cell lysate, and the reaction was performed on
Bio-Rad MyiQ system for 30 min at 30 °C followed by inactivation of telomerase at 95 °C for 1 min.
For the second (amplification) step, one of two reverse primers was used: ACX primer was used to
detect wild-type telomeric repeats added by endogenous telomerase, whereas a MT-hTer-specific
reverse primer was used to detect mutant telomeric repeats added by ectopically-expressed MT-hTer.
SYBR green and Taq platinum polymerase (Invitrogen, Carlsbad, CA) were added to the same tube
in a master mix formula. The reaction was run on the same machine at 95
o
C, 0”; 50
o
C, 5"; 72
o
C,
10" for 40 cycles to amplify WT-hTer or MT-hTer respectively. Ct values reflecting telomerase
activity were measured and then analyzed further by iQ5 optical system software version 2.0.

44

Table 4. Clinical and histological features of prostatectomy specimens.

Tumor Pathologic TNM Stage Gleason Score PSA % CSC
1 pT2bN0MX 3+4 6.76 7.0
2 pT4N1MX 5+5 3.71 0.7
3 pT3bN1MX 4+3 11.1 5.3
4 pT3bN0MX 4+5

9.32 3.5
5 pT3aN0MX 4+4 0.26 5.2
6 pT3aN0MX 3+3 8.67 6.4
7 pT3bN0MX 3+5 9.44 9.2
8 pT2cN0MX 3+5 5.14 5.2
9 pT2bN0MX 3+4 10.2 0.8
10 pT3aN0MX 4+3 19.7 6.6
11 pT3bN0MX 3+4 7.72 3.9
12 pT4N1MX 4+5 7.29 4.3
13 pT2aN0MX 4+4 6.77 2.8

45

Figure 1. FACS isolation of cell subpopulations. (A) Scheme for identification and isolation of CSC
cells. (B) Cells that did not adhere to collagen coated plates were α
2
β
1
-
CD44
-
. (C) Sample FACS for
CD133: of the 13 tumors tested, none possessed any significant CD133
+
subpopulation. (D) Sorted
cells were reanalyzed by FACS for CD44 and CD49b to confirm >98% purity.

46

Figure 2. CSC phenotypes of cell subpopulations isolated from human prostate tumors. (A)
Left: Gene expression (by RT-PCR) of cell subpopulations from 3 individual tumors (“+” = CSC, and
“-” = non-CSC). Right: Semi-quantitative densitometric analysis of gel at left (mean values from 3
tumors). (B) Relative colony formation and (C) matrigel invasion of cell subpopulations from 8
individual tumors with sample micrographs (arrows on matrigel micrograph indicate cells that
migrated across the membrane).

47

Figure 3. Telomerase and telomere characterization of CSC and non-CSC cell subpopulations
isolated from human prostate tumors. (A) Telomerase activity of cell subpopulations from 6
individual tumors by qPCR-TRAP, normalized to a standard control telomerase activity from DU145
cancer cells (p=0.03). (B) Relative hTERT mRNA levels of cell subpopulations in 4 individual tumors
(p=0.008). (C) Telomere lengths of cell subpopulations in 3 individual tumors, respectively. In 2A-B,
“ND” = not detectable. In all panels, 1000 cells were used for each experiment, and assays were
conducted in triplicate and compared using a two-sided Wilcoxon matched-pairs signed rank test of
statistical significance.

48

Figure 4. Induction of telomerase interference (reprogramming of telomerase) in tumor-derived
CSC. (A) MT-hTer expression by qPCR and (B) MT-telomerase activity levels by MT-specific
qPCR-TRAP 48 hours after infection with lentivirus expressing MT-hTer/siRNA. (C) Percent
apoptosis by TUNEL assay at day 4 after treatment with MT-hTer/siRNA or vector control. (D)
Proliferation after treatment with MT-hTer/siRNA or vector control. Data in A-D reflects biological
triplicate experiments conducted using CSC cells isolated from 1 patient tumor. All experiments were
also repeated in triplicate using 2 additional patient tumors with similar results (data not shown).

49

Figure 5. CSC phenotypes of cell subpopulations isolated from DU145 human prostate cancer
cell line. (A) Relative gene expression and (B) Colony formation (left), and invasiveness (right) of
CSC and non-CSC subpopulations.

50

Figure 5 (continued). (C) Relative telomerase activity (left) and hTERT mRNA levels (right), and
bulk telomere lengths (D) of CSC

and non-CSC

subpopulations normalized to control standard
(LNCaP cell line).

51

Figure 6. Effect of telomerase interference on in vivo tumor formation by DU145-derived CSC.
(A) Relative growth inhibition of CSC

and non-CSC

subpopulations 6 days after treatment with
MT-hTer/siRNA. (B) CSC were more tumorigenic than non-CSC, and tumor formation by CSC was
abrogated by telomerase interference. (C) Mean tumor weights at excision on day 90 post inoculation
(NT = no tumors).

52
Discussion
Telomerase activity is considered a nearly universal characteristic of cancer cells, an
assumption that may exist because early surveys of telomerase activity were conducted
indiscriminantly from lysates of entire cancer populations (Shay and Bacchetti, 1997),
and because the oncogenic role of telomerase was demonstrated by ectopically
introducing the enzyme into unselected cell populations (Bodnar et al., 1998; Hahn et al.,
1999a). Contrary to this model of homogeneous telomerase activation, studies of normal
tissue stem cell compartments have demonstrated a unique role for telomerase in stem
cell activation (Lee et al., 1998; Sarin et al., 2005), raising the possibility that perhaps
telomerase plays a parallel unique role in so-called cancer stem cell (CSC) or tumor
initiating cells. In support of a unique telomerase role in CSC, ectopic overexpression of
telomerase in cancer cell lines has indeed been shown to enhance tumor initiation,
perhaps reflecting a potentiation of the CSC phenotype (Gu et al., 2007; Shay and
Bacchetti, 1997; Stewart et al., 2002). Hence, we sought to determine whether telomerase
activity in prostate tumors and cell lines is not uniformly distributed as previously
assumed, but rather is focused predominantly in the CSC subpopulation where it can be
used to neutralize these cells.
A variety of experimental systems have been described previously for the study of
prostate CSC, resulting in a wide array of observations. For instance, a luminal epithelial

53
stem cell was shown to be a cell of origin for prostate cancer using a mouse model in one
study (Wang et al., 2009), while in another report only basal cells from primary benign
human prostate tissue could initiate prostate cancer in immunodeficient mice (Goldstein
et al., 2010). Moreover, direct inoculation of primary prostate tumor subpopulations into
mice as proof of a CSC phenotype has not been achieved as is done routinely in several
other tumor types (Al-Hajj et al., 2003; O'Brien et al., 2009; O'Brien et al., 2007;
Quintana et al., 2008; Singh et al., 2004). Given this background of multiple models and
reports, we elected to conduct our study using primary human prostate tumors enriched
for α2β1
high
CD44
+
, the most robust and consistently cited markers for a CSC phenotype
in the prostate cancer literature (Collins et al., 2005; Collins et al., 2001; Hurt et al., 2008;
Klarmann et al., 2009; Patrawala et al., 2007; Patrawala et al., 2006). Functionally, CD44,
the primary receptor for hyaluronic acid (HA), plays critical roles in cancer cell adhesion,
migration and drug resistance, consistent with a CSC phenotype (Hao et al., 2010). This
choice of markers was further reaffirmed by a recent report wherein integrin α6
high
CD44
+
cell spheres generated in vitro from human prostate tumor tissues were successfully
implanted (along with rat urogenital sinus mesenchyme) into a new strain of highly
immunocompromised NOD-SCID/IL2r γNull mice, further substantiating the stem-like
phenotype of this population (Garraway et al., 2010).

54
In our present study, prostate tumor-derived α2 β1
high
CD44
+
cells expressed elevated
levels of genes associated with self-renewal and also were more clonogenic and invasive
than bulk unselected cells. Similarly, DU145-derived α2β1
high
CD44
+
cells also had
CSC-like in vitro properties as well as increased tumorigenicity in vivo relative to
α2β1
high
CD44
-
cells. Another study recently examined prostate cancer cell lines for
CSC-like cells and observed this subpopulation to have elevated telomerase activity
which was inhibited with an anti-telomerase oligonucleotide (Marian et al., 2010b). Our
present studies sought to further advance this line of investigation in several ways:
Comparing telomerase activity between CSC and non-CSC subpopulations from freshly
resected human tumors, testing the impact of telomerase reprogramming (via
MT-hTer/siRNA) on these tumor-derived subpopulations, and investigating the effects of
telomerase interference on tumor initiation in vivo using a similar cell line-derived CSC
subpopulation.
Strikingly, we found that a subpopulation of primary tumor cells with CSC properties
had markedly higher telomerase activity and hTERT expression than non-CSC from the
same prostate tumors. Moreover, the non-CSC fractions, which comprised the vast
majority (>90%) of tumor cells, did not propagate in vitro, a finding that further
highlighted the biologic dichotomy between the two subpopulations: The large non-CSC
subpopulation had very low telomerase activity and was unable to proliferate in culture,

55
whereas the small CSC subpopulation had very high telomerase expression and activity,
did proliferate in vitro, and underwent rapid apoptosis with exposure to telomerase
interference, a strategy which specifically reprograms and exploits telomerase to generate
toxic uncapped telomeres. This dichotomy in telomerase phenotype was further borne out
in the DU145 experiments, where telomerase interference significantly inhibited the
proliferation of CSC but exerted minimal effects on non-CSC. One possible explanation
for this differential effect was that CSC had much more telomerase available to be
reprogrammed by MT-hTer/siRNA, although contribution from other biological
differences between CSC and non-CSC cannot be ruled out. Lastly, in the in vivo
experiments, non-CSC derived from DU145 were unable to form tumors, whereas
DU145-derived CSC had brisk tumor formation that was efficiently abrogated by
telomerase interference with MT-hTer/siRNA.
Collectively, our experiments demonstrate that telomerase expression and activity are
not a uniform phenotype common to all cancer cells as generally assumed, but rather are
concentrated in a subpopulation of cells with CSC-like properties in prostate tumors and
cell lines. Their highly active telomerase phenotype rendered prostate CSC exceedingly
susceptible to telomerase interference, which induced apoptosis, inhibited proliferation,
and abrogated tumor formation. Moreover, the potent effects of MT-hTer/siRNA on
DU145-derived CSC (telomerase-high) versus its minimal effects on DU145-derived

56
non-CSC (telomerase-low) suggests that telomerase interference may have a degree of
therapeutic selectivity for CSC vis-à-vis their high telomerase. Therefore, targeting
telomerase in this manner may constitute a promising new therapeutic strategy for
neutralizing CSC in prostate and other cancers, ultimately leading to more effective
control of tumor recurrence and progression.

57
Chapter 3
Cancer cells cyclically lose and regain a drug-resistant highly
tumorigenic features characteristic of a cancer stem-like
phenotype

Abstract
Drug resistance and brisk tumor initiation have traditionally been viewed as
pre-existing phenotypes present in a small subpopulation of cancer cells which can
expand under selective pressures. However, recent work in cancer cell lines has
demonstrated that drug-resistant tumor initiating features can emerge de novo within
fractionated subpopulations of cells initially lacking these phenotypes. In the present
study, we asked whether such phenotypic plasticity exists broadly in unperturbed cancer
cell lines and tumor xenografts growing spontaneously without interventions like drug
selection or fractionation into subpopulations employed in prior studies. To address this
question, we used side population (SP) analysis combined with fluorescence labeling to
identify a drug-resistant highly-tumorigenic subpopulation and to track and analyze its
interaction with the larger phenotypically negative population over time. Remarkably, we
observed that SP size fluctuated in a cyclical manner: first contracting via differentiation
into the non-SP (NSP) population, and then re-expanding via simultaneous direct

58
conversion of numerous NSP cells back to the SP phenotype both in culture and in tumor
xenografts. These findings demonstrate for the first time that adaptive, cancer-promoting
traits like drug-resistance and brisk tumor initiation arise not only as solitary events under
selective pressures, but also as highly orchestrated transitions occurring concurrently in
large numbers of cells even without specifically-induced drug selection, ectopic gene
expression, or fractionation into subpopulations. This high level of coordinated
phenotypic plasticity bears consideration when using cancer cell lines as experimental
models and also may have significant implications for therapeutic efforts targeting the
drug-resistant tumor-initiating phenotype.

Introduction
The phenomena of drug resistance and tumor metastasis have long been recognized
as central challenges in cancer therapy (Dean et al., 2005; Nguyen et al., 2009).
Traditionally, these properties have been conceptualized as pre-existing molecular
phenotypes randomly present in a small subset of cancer cells, sometimes termed cancer
stem cells, that under the right environmental selection pressures (e.g. drug treatment or a
new tumor site) – would be favored to expand (Olive et al., 2009; Tsuchida et al., 2008).
To study these adaptive responses, a variety of immortalized and transformed cell lines
have been employed; while only offering an approximation of in vivo cancer behavior,

59
these models have allowed for serial labeling, tracking, and characterization of
phenotypically distinct subpopulations over time. In this manner, small subpopulations of
cells marked by drug resistance and vigorous tumor formation have been repeatedly
identified and expanded (Fukunaga-Kalabis et al., 2010; Hu et al., 2010; Singh et al.,
2010; Steiniger et al., 2008).
Recently, the established “unidirectional” model of a unique, pre-existing
population selected to expand has been called into question by a number of high-impact
studies: In several reports, cancer cells lacking the putative markers of drug resistance
and brisk tumorigenicity were nonetheless capable of tumor formation under certain
permissive conditions (Kelly et al., 2007; Quintana et al., 2008; Shmelkov et al., 2008).
In two recent reports, cells lacking drug-resistant and tumor initiating features were
clonally isolated and shown to be capable of reconstituting heterogeneous populations
comprised both of cells with drug-resistant and tumorigenic features as well as cells
lacking those phenotypes (Roesch et al., 2010; Sharma et al., 2010). These findings have
raised the intriguing possibility that a “pre-existing” seed population may not be
necessary for the emergence of tumor-forming or drug-resistant phenotypes; rather, a
subpopulation with these properties can arise de novo under certain conditions.
Given these observations, we wondered whether re-emergence of a drug-resistant
highly-tumorigenic phenotype could be achieved only under selective growth conditions

60
or through isolation from other subpopulations as done in previous reports, or was this a
more universal and spontaneous phenomenon. Specifically, could phenotypic plasticity
be readily observed even in cell lines propagated under standard conditions without
extrinsic selective pressures and without separation into constituent subpopulations?
Moreover, did such plasticity represent a clonal selection of one phenotype from the other
over time, or did it represent a real-time conversion occurring rapidly in many cells at
once?
To investigate this question, we analyzed cancer cell lines in vitro and in vivo over
time using flow cytometry with Hoechst dye exclusion, a commonly-used method which
yields a side population (SP) of cells with drug-resistant highly-tumorigenic properties in
tumors and cell lines (Bleau et al., 2009; Hirschmann-Jax et al., 2004; Ho et al., 2007;
Patrawala et al., 2005; Wu et al., 2007). Remarkably, when we coupled SP analysis with
GFP labeling, we observed a dynamic two-way equilibrium between the SP and non-SP
(NSP) subpopulations in cell culture and in tumor xenografts. Specifically, the SP
subpopulation first became depleted by differentiation into NSP cells, and subsequently
the SP subpopulation was reconstituted by direct conversion of numerous NSP cells
simultaneously back to the SP phenotype; these transitions occurred spontaneously in the
course of proliferation without exogenous selection pressures or separation into
constituent subpopulations. Our findings demonstrate for the first time that intact cancer

61
cell lines exhibit continuous, spontaneous plasticity whereby large numbers of cells lose
and subsequently regain a drug-resistant highly-tumorigenic phenotype in a cyclical
manner. These observations suggest that adaptive traits which confer a survival advantage
may be acquired by cancer cell populations not only through clonal selection of
pre-existing, solitary cells, but also through an ongoing, highly orchestrated process of
phenotypic interconversion occurring simultaneously in large numbers of cells to
regenerate a cancer stem-like population.

Materials and Methods
Cell culture and lentiviral infections. Cancer cell lines were obtained from
collaborators at the University of California, San Francisco and the University of
Southern California (see Acknowledgements) and were not re-authenticated prior to use
in these experiments. Human bladder cancer cells (J82, RT4, UM-UC3), human breast
cancer cells (MCF7) and rat glioma cells (C6) were maintained at 37°C, 5% CO
2
in
DMEM (Mediatech) supplemented with 10% of heat-inactivated fetal bovine serum
(Omega), penicillin (100 units/ml, Invitrogen), and streptomycin (100 µg/ml, Invitrogen).
Human prostate cancer cells (PC3, Du145, LNCap, LNCap-C4-2B) and human lung
cancer cell (H441) were maintained in RPMI medium (supplemented with 10% FBS, and
antibiotics 100 units/ml penicillin and 100 µg/ml streptomycin).

62
GFP labeling of cells. Lentivirus was generated as previously described to deliver
either GFP under a CMV promoter or control empty vector (Xu et al., 2010b). One day
prior to infection 2X10
5
J82 cells were seeded in 10cm plates, and on the next morning
media was replaced by 3 mL virus supernatant plus 7 mL media supplemented with 8
µg/ml polybrene. After 8 hr incubation at 37°C, the virus-containing media was replaced
with fresh media. Cells were observed for 48 hr to ensure >90% GFP expression prior to
fluorescence activated cell sorting (FACS) and side population studies.
Flow cytometry. Hoechst staining and FACS were conducted as described
previously (Goodell et al., 1996). Briefly, adherent cancer cells (1X10
6
/mL) were
trypsinized, counted, and resupended in prewarmed 10% FBS DMEM media. Hoechst
33342 (Sigma-Aldrich) was added at concentration of 5 µg/mL, incubated for 2 hrs in
37°C water bath and gently inverted several times during the course of incubation.
Parallel sample aliquots were prepared in the presence of 50 µM verapamil
(Sigma-Aldrich), an ATP-binding cassette transporter family inhibitor, at room
temperature for 10min before adding the Hoechst 33342 dye. Cells were centrifuged at
1000rpm for 5 min after incubation and resuspended in ice-cold DMEM media.
Propidium iodide (Sigma-Aldrich) was added to the cells at a final concentration of 2
µg/mL. Samples were incubated for at least 5 min on ice before FACS analysis
(FACSAria and FACSLSR-II, BD Biosciences, both equipped with UV lasers).

63
Drug resistance experiments. Cells were stained with Hoechst 33342, and FACS
sorted into SP and NSP cells. Fractionated SP or NSP or unsorted WP (whole population)
cells were seeded into 96 well plates at a concentration of 1x10
4
cells per well in the
presence of Cisplatin (25 µM) or Docetaxel (0.4 mM for J82 cells and 0.1 mM for MCF7
cells respectively). MTS assay was performed after 24 hours according to manufacture’s
protocol (Promega, Madison, WI).
Direct single cell isolation. Single cell suspension was made in DMEM media and
stained by Hoechst 33342 for side population as described above. Side population and
non-side population cells were sorted by FACSAria directly into 96-well plates at the
concentration of one single cell per well (FACS 96 well sorting program), and single
clone formation was confirmed visually on subsequent days.
Tumorigenicity assays. In vivo experiments were conducted under an approved
protocol in accordance with the institutional guidelines for the use of laboratory animals.
SP and NSP cells were suspended in 100 µL of media:matrigel=1:1 and inoculated
subcutaneously into the right flank of male SCID mice (6-8 weeks old, NCI-Frederick).
Time to onset of palpable tumor was recorded and tumor diameter was measured twice
weekly. Mice were sacrificed and tumors were excised for wet weight measurement when
the largest tumor diameters within each group exceeded 10 mm.

64
Analysis of excised tumors. In preparation for FACS, tumors were minced into
small pieces (about 1 mm
3
) and immersed in digestion solution. (DMEM/F12 (50:50)
supplemented with DNAse I (1 µg/mL ) and Liberase Blendzyme 3 (0.1 - 0.8 mg/mL)).
Cells were incubated at 37 °C overnight, followed by trypsin incubation at 37 °C for 5
min the next morning, then trypsin neutralization with DMEM 10% FBS media. Cell
suspension was filtered first by 100 µm strainer, then through 40 µm strainer (BD Falcon)
prior to FACS.
Colony formation assays. 1000 SP or NSP cells were seeded into 10 cm plates after
FACS sorting. As a control for the presence of Hoechst dye, 1000 SP cells were treated
with 50 µM verapamil for 10 min, and further stained by Hoechst33342 for another 2
hours, and then seeded into 10 cm plates. Cells were cultured at 37°C, 5% CO
2
. Plates
were washed 3x with PBS and fixed with 100% methanol (Room temperature, 15 min),
then stained for 1 hr at room temperature with 7 mL of Giemsa Stain (Sigma-Aldrich).
After staining, cells were washed with water and air-dried overnight. The number of
colonies with diameter of over 1mm was counted and photographed under bright-field
microscopy (Nikon Eclipse TS100). All experiments were done in triplicate, and repeated
at least twice.
Quantitative reverse transcription-PCR assays. Total RNA was isolated from
FACS-sorted SP and NSP cells using manufacture’s RNA isolation protocol (RNA-Bee,

65
Tel-Test). RNA was reverse transcribed using RETROscript kit (Ambion), and cDNA was
then subjected to real time PCR amplification using gene specific primers and Quanta
B-R Syber Green QPCR supermix (BioScience) with β-actin as a housekeeping gene
loading control (primer sequences and conditions in Table 5).
Statistical Analysis. Performed in collaboration with USC/Norris Biostatistics Core.
All experiments were conducted in triplicate with error bars representing standard
deviation around the mean. Student’s t-test was used to determine statistical significance
when comparing mean values at one point in time (e.g. gene expression, tumor weights,
SP%). Two-sided Wilcoxon matched-pairs signed rank test was performed by using
Graphpad Prism5.0 software to compare statistical significance of tumor growth over
time.

Results
The side population (SP) is enriched for drug-resistant highly-tumorigenic cells.
We used Hoechst efflux and FACS analysis to identify side populations of cells with
drug-resistant highly-tumorigenic properties as previously reported in tumors and cell
lines (Bleau et al., 2009; Hirschmann-Jax et al., 2004; Ho et al., 2007; Patrawala et al.,
2005; Wu et al., 2007). Using this technique, we found side populations in a variety of
cancer cell lines, including prostate, breast, lung, bladder and glioma (Figure 7 and

66
Figure 8A). Sorting of side population (SP) from non-side population (NSP) on a
FACSAria flow cytometer (BD) yielded population purities of >91% and >97%,
respectively as confirmed on LSR-II cell analyzer (BD) (Figure 9). The J82 and MCF7
cancer cell lines yielded SPs which constituted ~22.3% and 6.1%, respectively, of the
overall populations, and which diminished to 0.5% and 1.5%, respectively, after
treatment with verapamil, which blocks the ABC transporter and prevents Hoechst efflux
(Figure 8A). To confirm that SP cells were enriched for drug resistance, J82 bladder
cancer and MCF7 breast cancer cell lines were fractionated into SP and NSP and exposed
to cisplatin and docetaxel, two chemotherapeutic agents commonly used in bladder and
breast cancer (Figure 8B). For both cell lines, SP cells demonstrated significantly greater
survival than NSP or unfractionated whole population (WP); SP cells also had 2-fold
mRNA expression of the ABCG2 transporter responsible for Hoechst and drug efflux
(Figure 10).
Side populations derived from MCF7 cells have been extensively characterized and
previously shown to have high clonogenicity and tumorigenicity (Patrawala et al., 2005;
Zhou et al., 2007); therefore, we next confirmed the same to be true of the J82 cancer cell
line using plating assays and SCID mouse inoculations. As expected, SP cells were
significantly more clonogenic than NSP cells, generating ~3 times as many large colonies
(>1mm diameter) at 5 weeks after seeding (Figure 8C). To ensure that these clonogenic

67
differences were not spuriously caused by retention of Hoechst dye in NSP cells and
absence of dye in SP cells, we repeated the clonogenic assays in the presence of Hoechst
dye + verapamil. Even when SP cells contained Hoechst dye similarly to NSP cells, they
still were significantly more clonogenic than NSP cells (Figure 8C). In vivo, SP cells
formed tumors in all mice at all dilutions (5x10
4
, 1x10
4
, and 1x10
3
). In contrast, NSP
cells formed tumors in only 1 of 3 mice inoculated with 1x10
3
cells and in 2 of 3 mice
inoculated with 1x10
4
cells (Figure 8D). Moreover, NSP-derived tumors formed only
after a longer lag period and were smaller at excision than SP-derived tumors (Figure 8D).
Thus, in summary, SP cells – relative to NSP cells – were found to be more drug-resistant,
to form more colonies in vitro, and to form tumors more rapidly and efficiently in vivo,
thus validating that SP cells indeed were enriched for a drug-resistant highly-tumorigenic
phenotype. In a further analysis, J82-derived SP cells also possessed higher mRNA levels
of several genes associated with self-renewal (Figure 10), which was consistent with
previous reports (Harris et al., 2008; Patrawala et al., 2005; Zhou et al., 2007). At the
same time, J82-derived and MCF7-derived subpopulations expressed very low or
non-discriminatory levels of CD133, ALDH1, and CD44
+
24
low
(Figure 11), a finding that
was not surprising, because these markers are known to be highly variable in different
types of tumors and cell lines. In some cases they correlate with high tumorigenicity and

68
drug resistance, while in other cases they are uniformly expressed or totally absent
(Keysar and Jimeno, 2010; Stuelten et al., 2010).
The NSP subpopulation generates SP and NSP subpopulations in vitro and in
vivo. In the course of characterizing SP cells, we noted that NSP cells – though
significantly less clonogenic and tumorigenic than SP cells – were nonetheless capable of
forming some colonies in vitro and tumors in vivo (Figure 8C-D). We wondered whether
this delayed, inefficient capacity of NSP cells was predicated upon reconstitution of an
SP subpopulation, which in turn promoted colony and tumor formation. We tested this
possibility by sorting J82 cells and several other cancer cell lines into SP and NSP
subpopulations which were expanded and re-analyzed by FACS (Figure 12A). J82 NSP
cells re-seeded in vitro reconstituted SP subpopulations that increased from undetectable
at day 4, to ~6% at day 10, to ~12% at day 13 (Figure 12B). Notably, we observed a
similar phenomenon (Figure 12B) in 3 other cancer cell lines previously shown to
possess SP cells with drug-resistant highly-tumorigenic properties: MCF7 (breast cancer),
C6 (rat glioma), and H441 (lung cancer) (Ho et al., 2007; Kondo et al., 2004; Patrawala
et al., 2005; Zhou et al., 2007). These NSP-derived subpopulations recapitulated the
expected drug resistant phenotypes, with SP demonstrating significantly higher drug
resistance than NSP (Figure 12C). In vivo we inoculated SCID mice with equal
numbers of J82 SP, NSP, or unsorted whole population (WP), then excised the xenograft

69
tumors when they reached 10mm in largest diameter and analyzed the cells by FACS
LSR-II. Notably, although NSP cells required a longer lag period before generating tumor
xenografts (Figure 8D), all the excised tumors – regardless of SP, NSP, or WP origin –
contained similar (2-3%) SP subpopulations (Figure 12D). Thus, NSP subpopulations
from several different cancer cell types successfully reconstituted SP subpopulations over
time in vitro and in vivo.
Reconstitution of SP and NSP subpopulations from isolated NSP cells in vitro and in
vivo could have conceivably resulted from imperfect sorting (97% purity, Figure 9)
leading to unintentional inclusion of SP cells among the NSP population. To exclude this
possible cross-contamination, we used FACS to sort single NSP cells directly into
individual wells of 96-well plates and visually confirmed formation of a single clonal
population in each well (Figure 13A). Significantly, NSP-derived single cells generated
colonies in 78 of 600 wells (13%), a greater proportion than could have been caused by
SP cross-contamination during FACSAria sorting which has only 3% inaccuracy. A large
majority (~75%) of the NSP-derived colonies in turn generated a significant SP
subpopulation, defined as SP%>1% (Figure 13B), thus confirming that NSP cells indeed
gave rise to SP cells, and that this observation could not be explained by SP
cross-contamination of NSP during sorting.

70
We extended the single cell clone experiments in vivo to confirm that NSP-derived
SP cells (termed secondary SP, or SP2º) reacquired the characteristic highly-tumorigenic
phenotype displayed by “native” SP cells; that is, the ability to efficiently form new
tumors comprised of heterogeneous cell subpopulations. To test this, NSP-derived clones
were FACS sorted into secondary SP and NSP subpopulations (SP2º and NSP2º) and
inoculated into SCID mice. While both SP2º cells and NSP2º cells were capable of
forming tumors, SP2º cells were significantly more tumorigenic than NSP2º cells (Figure
13C). Subsequent xenograft excision and FACS analysis (Figure 13D) confirmed
reconstitution of heterogeneous cell subpopulations in all tumors; both SP2º and NSP2º
generated tumors with SP as well as NSP subpopulations, and there was no significant
difference in SP% between the two groups. Collectively the single-cell clonal
experiments confirmed both in vitro and in vivo that NSP cells were able to reconstitute
SP subpopulations with a characteristic highly-tumorigenic phenotype.
The side population expands and contracts cyclically in the course of cell culture.
Having observed that NSP cells were capable of reconstituting SP+NSP populations in
vitro and in vivo, we sought to characterize the day-by-day dynamics of the two
subpopulations growing in a “native” state (i.e. together in an intact cancer cell line). To
accomplish this, we propagated unsorted J82 cells in cell culture and monitored daily the
size of the SP, NSP, and whole population. Unexpectedly, we found that within one day

71
of seeding J82 cells into a new plate, the SP subpopulation became rapidly depleted from
a baseline of ~20% down to ~0.5% of the total population (Figure 14A). In subsequent
days, the SP subpopulation experienced a gradual increase, punctuated by a sudden spike
in SP subpopulation size on day 3 or 4 (Figure 14A, open arrows). The SP subpopulation
continued to increase, approaching its original baseline percentage (20%) as the overall
culture neared confluence. At that point, the cells were passaged into new plates and the
cycle repeated. Interestingly, the sudden increase in SP percentage on day 3 or 4 was
independent of the overall expansion rate of the cultured J82 cells; in fact, SP expansion
at those inflection points (Figure 14B, open arrow) was ~4- to 7-fold more rapid than the
expansion of the NSP subpopulation or of the entire population as a whole (WP).
SP cells arise through simultaneous direct conversion of numerous NSP cells. We
observed that the SP subpopulation experienced a sudden, marked expansion at
approximately mid-passage, a finding that could be explained one of two ways:
According to the traditional model, drug-resistant highly-tumorigenic cells could arise
only through expansion of pre-existing cells, implying that the observed sudden SP
expansion was caused by an explosive increase in SP cell proliferation. An alternative
explanation for the sudden SP expansion was that SP cells were arising by conversion of
cells from the NSP subpopulation rather than by proliferation of pre-existing SP cells
alone. To test this hypothesis, we generated a GFP-based model that enabled us to track

72
the fates of the SP and NSP subpopulations (Figure 15A). Briefly, J82 cells were infected
with lentivirus expressing either empty vector (EV) or GFP. J82
GFP
and J82
EV
were FACS
sorted into their respective SP and NSP subpopulations. These were re-combined to
create a J82 cell line consisting of 20% SP
EV
and 80% NSP
GFP
. This “hybrid” cell line
then could be analyzed daily by FACS for Hoechst dye efflux and GFP status.
When analyzed daily by FACS during in vitro culture, the reconstituted J82 cells
experienced the same initial (day 1-2) depletion of SP cells observed previously, from
~22% down to 0.7% (Figure 15B). During this early phase (day 1-2) the SP cells (red)
migrated over to the NSP gate. Then, as the SP percentage gradually increased on days 3
to 6, there was a concomitant gradual increase in the GFP+ percentage (green cells)
within the SP gate (Figure 15B). At the day 6 inflection point, the sudden sharp rise in SP
subpopulation was mirrored by an equally sharp increase in GFP+ percentage within the
SP gate. These findings confirmed that the rapid re-expansion of the SP subpopulation
was not caused by intrinsic self-renewal and proliferation of pre-existing SP cells, but
rather by conversion of NSP cells into the SP phenotype. Notably, this phenomenon was
confirmed (and any confounding role of GFP protein expression was ruled out) with a
reverse experiment wherein 20% SP
GFP
and 80% NSP
EV
were combined with similar
results (Figure 16). Hence, these experiments provide the first direct evidence that the
drug-resistant highly-tumorigenic phenotype within an established cancer cell line is

73
rapidly and simultaneously acquired by numerous cells previously lacking this
phenotype.
The principle of phenotypic plasticity was borne out further in vivo. The “hybrid” J82
cell line consisting of 20% SP
EV
+ 80% NSP
GFP
was inoculated into SCID mice, and the
xenograft tumors were resected at 3 time points and analyzed by FACS for SP percentage
and GFP percentage (tumor gating accuracy and persistence of GFP expression were
confirmed in separate experiments, Figure 17). Interestingly, the proliferating tumors
recapitulated the early rapid proliferative phase observed in vitro; that is, the SP
percentage declined from 20% in the “hybrid” J82 pre-inoculation down to ~1-3% in the
xenograft tumors at subsequent time points (Figure 15C). Consistent with this trend,
FACS plots of resected tumors showed that the original SP
EV
(red) cells migrated over to
the NSP gate by day 69. At the same time, the GFP+ percentage within the SP gate
increased from 30% GFP+ pre-inoculation to ~90% GFP+ in the xenograft-derived SP
subpopulations, suggesting a progressive replacement of the original SP
EV
subpopulation
(largely GFP-) by a new GFP+ SP subpopulation derived from the NSP
GFP
cells (Figure
15C). Hence, in tumor xenografts, as in vitro, drug-resistant highly-tumorigenic SP
subpopulation was replenished by conversion of cells from the NSP subpopulation. As
illustrated in the summary model (Figure 15D), the observed phenotypic
inter-conversions were spontaneous but also cyclical and occurring simultaneously in

74
large numbers of cells, a non-random pattern suggestive of a deliberate process, perhaps
in response to some environmental factor or factors.

75

Table 5: Primer sequences used for qRT-PCR

76

Figure 7. Side population (SP) analysis in various cancer cell lines. All lines were found to have
SP by Hoechst staining which were abrogated with Verapamil, a calcium channel blocker that
antagonizes ABCG2 transport.

77

Figure 8. The side population (SP) is enriched for drug-resistant highly-tumorigenic cells. (A)
FACS plots gated for a Hoechst-negative side population (SP) of J82 bladder cancer and MCF7 breast
cancer cells; treatment with Verapamil 50 µM drastically reduced the SPs. (B) Drug resistance of side
population (SP), non-side population (NSP) or unfractionated whole population (WP) 24 hours after
exposure to cisplatin (25 µM) or docetaxel (0.4 mM for J82, 0.1 mM for MCF7). (C) Colony forming
assays comparing NSP to SP and to SP + verapamil (“vera”) + Hoechst33342. Left: Sample images of
colony formation; Right: Colonies were quantified by counting those with radius >1mm 3 weeks after
plating.

78

Figure 8 (continued). (D) Xenograft tumor formation: Left: FACS-fractionated SP or NSP were
inoculated subcutaneously into SCID mice at three dilutions (5x10
4
, 1x10
4
and 1x10
3
) and tumor
volumes were measured twice weekly. Right: Wet weights of tumors: For each cell dilution, tumors
were excised when the largest tumors exceeded 10mm in greatest diameter. All results are means of
independent experiments conducted in triplicate.

79

Figure 9. Sorting efficiency of FACSAria. For each subpopulation, cells were first fractionated on
the BD FACSAria, then the sorted subpopulation was re-analyzed on a BD LSR-II analyzer. (A)
Sorting purity of J82 SP and NSP cells. (B) Sorting purity of empty vector (EV) labeled or GFP
labeled SP and NSP cells.

80

Figure 10. Relative gene expression. Quantitative RT-PCR measuring expression of genes associated
with self-renewal, normalized to β-actin (loading control).

81

Figure 11. FACS analysis for additional markers. (A) J82 and MCF7 cell lines were labeled with
CD44 antibody conjugated to FITC and CD24 antibody conjugated to PE and analyzed by FACS (left).
Percentage of CD44
+
CD24
-/low
present in SP, NSP and WP in J82 and MCF7 cells was tabulated
(right). (B) ALDH activity was analyzed in J82 and MCF7 cell lines by aldeflour assay. HT-29 cells
were used as a positive control; no significant aldeflour positive cells were detected in either J82 or
MCF7 cells. (C) FACS for CD133: Caco-2 cells were used as a positive control for CD133 staining;
neither J82 or MCF7 possessed any significant CD133
+
subpopulation.

82

Figure 12. The NSP subpopulation generates SP and NSP subpopulations in vitro and in vivo. (A)
Schematic of cancer cell subpopulation-based experiments. (B) SP composition over time of cultured
cells derived from FACS-fractionated NSP using 4 different cancer cell lines. (C) Relative drug
resistance of NSP-derived SP, NSP, and WP; (D) SP composition of mouse xenograft tumors derived
from FACS-fractionated J82 SP, NSP, and whole population (WP); tumors were excised and analyzed
when they reached 10mm in greatest diameter.

83

Figure 13. NSP single cell clones reconstitute SP and NSP subpopulations in vitro and in vivo. (A)
Schematic of single cell clone experiments: single NSP cells were sorted into 96-well plates, clonally
expanded, and subjected to FACS analysis for SP% as well as FACS sorting into secondary SP and
NSP (SP2
o
and NSP2
o
), which in turn were inoculated into SCID mice for in vivo analysis. (B)
Distribution of SP% within cell clones derived from single NSP cells. (C) Xenograft tumor formation:
1x10
4
SP2
o
or NSP2
o
cells FACS-fractionated from NSP-derived clones were inoculated
subcutaneously into SCID mice and tumor volumes were measured twice weekly. (D) SP% of
xenograft tumors derived from SP2
o
or NSP2
o
inoculation.

84

Figure 14. SP subpopulation size fluctuates cyclically in the course of cell culture. (A) SP
composition during in vitro culture. 1x10
5
cells were seeded into 6-well plates at day 0 (passage 1) in
triplicates and analyzed daily by Hoechst/FACS analysis. At day 6, 1x10
5
cells were reseeded into
6-well plates to regenerate the cycle (passage 2). (B) Relative daily change in size of SP, NSP and WP,
calculated as {cell # at day (N+1)} / {(cell # at day (N)}. Open arrows indicate inflection points of
sudden increase in SP size.

85

Figure 15. SP cells arise through conversion of NSP cells. (A) Schematic of GFP experiments:
FACS fractionated SP
EV
cells were mixed with NSP
GFP
cells at a 20:80 ratio at day 0; 1x10
5
of the
mixed cells were cultured in 6-well plates for in vitro analysis or inoculated into SCID mice for in vivo
xenograft tumor analysis. (B) In vitro analysis: Left: Daily FACS plots showing the relative
proportions of GFP+ and GFP- cells in the SP and NSP gates over time. Right: Graphs plotting the
relative size of SP and the GFP+ proportion within the SP over time.

86

Figure 15 (continued). (C) In vivo analysis: Left: FACS plots showing the relative proportions of
GFP+ and GFP- cells in the SP and NSP gates at 3 time points. Right: Graphs plotting the relative size
of SP and the GFP+ proportion within the SP in xenograft tumors derived from the 20:80
SP
EV
:NSP
GFP
J82 cancer cell line. (D) Illustrated model summarizing the direct interconversion
observed between the drug-resistant highly-tumorigenic phenotype (DRHT, SP) and the drug-sensitive
low-tumorigenic phenotype (DSLT, NSP). This plasticity occurred cyclically in many cells at once,
suggesting a possible adaptive response to environmental factors (italicized with question marks).

87

Figure 16. Percentage of GFP-negative cells in the SP subpopulation. This experiment was the
reverse of the one presented in Figure 15. Here, SP was combined with NSP at a 20:80 ratio. The
results paralleled and confirmed those in Figure 15, indicating that NSP cells (GFP-negative in this
experiment) were reconstituting the SP subpopulation over time.

88

Figure 17. Control experiments confirming that lentiviral-mediated GFP expression is
maintained in vivo throughout course of experiments. (A) FACS plot showing standard gate used
to isolate J82 cells in culture. (B) Same gate used to isolate J82 cells from mouse background cells and
debris in excised xenograft tumor. (C) GFP+ J82 cells inoculated into SCID mouse and excised after
60 days remain>99% GFP+. (D) Negative control (J82 cells not infected with lentiviral GFP).

89
Discussion
We investigated whether emergence of drug-resistant highly-tumorigenic cells could
be observed in cancer cell lines propagated in culture and as tumor xenografts.
Specifically, we asked if phenotypic plasticity could be observed even in unfractionated,
intact cell lines grown without specific selective pressures such as chemotherapy
treatment. Moreover, could direct phenotypic conversion within groups of cells – rather
than clonal selection over time – be documented as an adaptive mechanism employed by
these cell lines. To study these questions, we used established cancer cell lines which
contained a Hoechst-effluxing side population (SP) characterized by drug resistance and
high tumorigenicity (Figure 8 and (Ho et al., 2007; Hu et al., 2010; Kondo et al., 2004;
Patrawala et al., 2005; Salcido et al., 2010; Singh et al., 2010; Steiniger et al., 2008; Zhou
et al., 2007)). We recognize that cancer cell lines may not perfectly represent the biology
of spontaneous in vivo tumors; however, tumor tissues and primary tumor-derived
cultures do not lend themselves to serial labeling and analyses of phenotypically distinct
subpopulations over time. In contrast, cell lines do allow this type of characterization and
therefore have been used extensively to explore how drug-resistant highly-tumorigenic
phenotypes emerge (Fukunaga-Kalabis et al., 2010; Lin et al., 2010; Roesch et al., 2010;
Sharma et al., 2010; Tsuchida et al., 2008). Cancer cell subpopulations possessing these
properties have sometimes been referred to as “cancer stem cells” or “tumor initiating

90
cells” (Visvader and Lindeman, 2008), because their typically high expression of
self-renewal genes (also noted in our studies, Figure 10) conjures a stem-like, progenitor
or hierarchical association. However, as reported previously (Stuelten et al., 2010) and
noted in our own studies (Figure 11), particular markers can vary widely in different
tumors and cell lines; therefore, for the purpose of these studies, we opted for
terminology based strictly on functional, therapeutically relevant properties that were
empirically demonstrated in our SP/NSP experiments: drug resistance and high
tumorigenicity.
As expected, FACS-sorted SPs gave rise to entire heterogeneous populations
(SP+NSP) in vitro and in vivo; however, NSPs from all of the cancer cell lines also were
capable of reconstituting heterogeneity (SP+NSP) albeit less efficiently (Figures 8 and
12), a finding that we validated with single cell clone experiments (Figure 13). These
findings were consistent with a recent study using embryonic stem cells where NSP was
shown to give rise to SP (Vieyra et al., 2009), as well as with recent observations in
cancer models where cells lacking the canonical surface markers for a highly-tumorigenic
phenotype were nonetheless capable of tumor formation (Bleau et al., 2009; Kelly et al.,
2007; Quintana et al., 2008; Shmelkov et al., 2008; Wu et al., 2007). A host of
explanations have been proposed to account for the observed tumorigenicity of
marker-negative cells: limited discrimination of cell surface markers, cross-contamination

91
during sorting, and variability of mouse models used to gauge tumorigenicity (Bleau et
al., 2009; Ho et al., 2007; Kelly et al., 2007; Quintana et al., 2008; Shmelkov et al., 2008).
However, another plausible explanation demonstrated in two recent papers (Roesch et al.,
2010; Sharma et al., 2010) and supported by our own findings is that drug-resistant
highly-tumorigenic subpopulations could arise directly from cells that initially lack these
traits.
Having confirmed that fractionated subpopulations expanded in isolation could
reconstitute phenotypic heterogeneity, next we wished to better-define the interaction of
these subpopulations in an intact unfractionated cell line (J82) by tracking the SP and
NSP cells in real time. Remarkably, we found that the SP subpopulation size fluctuated
cyclically with each passage in cell culture (Figure 14). Fluctuations in SP size have been
reported previously with exposure to radiation, hypoxia, or ectopic activation of specific
signaling pathways and were attributed to intrinsic changes in SP proliferation in
response to deliberate external stimuli (Bao et al., 2006; Das et al., 2008; Liang et al.,
2010; Mani et al., 2008; Zhou et al., 2007). In contrast, our experiments applied no
external stimuli to the cultured cells; nevertheless, we unexpectedly observed cyclical SP
fluctuations occurring spontaneously in two phases: First, the SP subpopulation
contracted drastically after seeding, diminishing from ~20% to less than 1% in the first
two days. There are three possible explanations for this: 1. SP cells died out in the first 2

92
days – this is unlikely as no such phenomenon was observed in our prior experiments
using fractionated SP cells. 2. SP cells did not proliferate and therefore were relatively
“swamped out” by the NSP cells – this also is unlikely because the overall population
doubling time (~24 hr) could not possibly reduce the SP percentage 20-fold in 2 days. 3.
SP cells rapidly “differentiated out” into NSP cells – this explanation is most consistent
with the classic drug-resistant highly-tumorigenic phenotype, and is in fact borne out by
the subsequent GFP labeling experiments, where GFP-negative cells were observed to
migrate from the SP gate to the NSP gate in the first 2 days (Figure 15).
The second phase in the cyclical fluctuation of SP size was even more striking: At
every passage, as the cell culture passed mid-passage and began to approach confluence,
the SP subpopulation underwent a rapid expansion (Figure 14). When we GFP-labeled
the NSP subpopulation (Figure 15), we observed that the rapid SP expansion derived not
from intrinsic proliferation of pre-existing SP cells, but rather from direct conversion of
NSP cells to the SP phenotype, a phenomenon that was recapitulated in vivo when labeled
cells were xenografted into SCID mice. Thus, both in vitro and in vivo, the SP
subpopulation was replenished by conversion of cells from the NSP subpopulation.
The results of our experiments suggest a higher order of population-wide plasticity
than previously suspected. Recent reports showed that drug-resistant tumor-initiating
traits could arise de novo within fractionated phenotypically negative subpopulations

93
expanded in isolation or under selective pressures, suggesting that solitary cells or small
numbers of cells were capable of altering their phenotypes under these ectopically
induced conditions (Das et al., 2008; Heddleston et al., 2009; Liang et al., 2010; Mani et
al., 2008; Roesch et al., 2010; Sharma et al., 2010). Our current findings demonstrate for
the first time that such plasticity is actually a robust phenomenon which can occur
spontaneously in intact cell lines without the need for fractionation or selective pressures.
Specifically, a drug-resistant highly-tumorigenic phenotype is lost and subsequently
regained by subpopulations of cells during passaging in culture, and similar phenotypic
interconversion is also observed in tumor xenografts.
Importantly, although the observed phenotypic inter-conversion was spontaneous (no
ectopic drugs or gene expression), it was not entirely random. Rather, it was a cyclical,
highly orchestrated process occurring simultaneously in large numbers of cells,
suggesting a real-time adaptive capability, perhaps in response to environmental factors
such availability of nutrients, oxygen or space (Cairns et al., 2011; Keith and Simon,
2007). This type of rapid, population-wide adaptation constitutes an alternative to the
traditional model wherein pre-existent traits that confer a survival advantage are
expanded in a “blind”, passive manner under selective pressures. Instead, entire
populations of cancer cells may possess innate, programmed adaptive responses that are
activated by real-time sensing and response to environmental conditions. According to

94
this hypothetical dynamic model (Figure 15D), when cancer cells are introduced into a
new, permissive growth environment (e.g. culture dish or subcutaneous inoculation), the
drug-resistant highly-tumorigenic subpopulation rapidly differentiates into the bulk
phenotypically negative population and thus effectively populates the new environment.
Conversely, as that environment becomes saturated and less hospitable – perhaps through
depletion of critical resources (nutrients, space, oxygen) or accumulation of toxins – a
portion of phenotypically negative cells respond by reverting to a drug-resistant
highly-tumorigenic phenotype, which is more robust and better-able to sustain the cancer
population until a new permissive environment (i.e. new culture plate or metastatic site)
is encountered and the cycle repeats. Indeed, this type of highly-orchestrated direct
conversion of phenotypically negative cells to a drug-resistant highly-tumorigenic
phenotype may underlie some of the changes in subpopulation size previously-observed
in response to deliberate manipulations (Bao et al., 2006; Das et al., 2008; Heddleston et
al., 2009; Liang et al., 2010; Mani et al., 2008; Zhou et al., 2007), as well as the observed
capacity of phenotypically negative cells to form tumors (perhaps by first reconstituting a
highly-tumorigenic subpopulation in vivo).
The environmental signals and cellular mechanisms governing these phenotypic
inter-conversions are the subject of intense investigation by our group and are as yet to be
fully elucidated. At the same time, the very presence of this phenomenon has important

95
experimental and clinical implications. Experimentally, cancer cell lines – so commonly
employed to study drug resistance and tumor formation – should be used with the
recognition that phenotypic heterogeneity exists in subpopulations which are engaged in
a continuous state of flux. Clinically, phenotypic plasticity – if confirmed in additional
models – would have significant implications for cancer therapeutics. Current efforts to
eradicate drug-resistant highly-tumorigenic subpopulations as distinct entities may meet
with limited success, because this phenotype may simply be replenished by the
phenotypically negative population. Hence, it may ultimately prove more productive to
conceptualize drug resistance and high tumorigenicity as adaptive phenotypes assumed
by a subset of the cancer population, rather than as discrete, static subpopulations of
cancer stem cells. Elucidating the signals and mechanisms that govern this dynamic,
ongoing plasticity may ultimately lead to more effective therapeutic strategies aimed at
disrupting the adaptive capabilities of cancer.

96
Chapter 4
Plasticity between the cancer stem-like and non cancer
stem-like states is regulated by the PI3K/Akt/ β-catenin/CBP
pathway

Abstract
Cancer stem-like cells (CSC) – a subpopulation of self-renewing, tumorigenic, drug
resistant tumor cells – are thought to promote cancer formation, therapy resistance and
disease progression. Recently, we and others reported that the CSC phenotype may not
represent a static pre-existing seed population; rather, CSC subpopulations can arise
through direct conversion of non-CSC to the CSC phenotype. Here we show that
PI3K/Akt/ β-catenin/CBP signaling plays a key role in mediating this important
phenotypic plasticity.
As previously reported, fluorescence activated cell sorting (FACS) and Hoechst dye
exclusion were applied to cancer cell lines in order to isolate side populations (SP) of
cells enriched for CSC properties (high tumorigenicity and drug resistance) versus
non-side-populations (NSP) of cells lacking these properties. GFP
+
NSP were
re-combined with GFP
–
SP to enable tracking of the two phenotypes in culture over time.
The cells were treated with pharmacological and siRNA inhibitors targeting members of

97
the PI3K/Akt/ β-catenin/CBP pathway, and their effects on SP size and NSP-to-SP
conversion were measured. We found that NSP-to-SP conversion was significantly
reduced by exposure to LY294002, a PI3K/Akt inhibitor, and the same effect was
recapitulated by direct siRNA knockdown of Akt. Further downstream, BIO
(6-bromoindirubin-3’-oxime) was used to inhibit GSK-3 β, a protein that normally
potentiates β-catenin degradation. Exposure to BIO increased both the overall levels and
the nuclear localization of β-catenin and significantly increased SP size. Conversely,
siRNA knockdown of β-catenin signficantly decreased SP size. Still further downstream,
two β-catenin transcription co-factors, CBP and P300, exerted opposite effects on the
CSC phenotype: Blocking β-catenin–CBP interaction with the specific inhibitor ICG001
significantly decreased SP size and sphere formation, whereas blocking β-catenin–P300
interaction with the specific inhibitor IQ-1 significantly increased SP size. Moreover,
ICG001 inhibition of β-catenin-CBP interaction abrogated the SP expansion caused by
upstream BIO inhibition of GSK-3 β. ICG001 did not affect the nuclear localization of
β-catenin, suggesting that CBP indeed cooperates with β-catenin further downstream to
transcriptionally impact NSP-to-SP conversion.
Cellular plasticity between the non-CSC and CSC phenotypes has considerable
therapeutic ramifications because it implies that the drug-resistant, aggressive CSC
subpopulation may be replenished from the larger non-CSC population. Our present

98
studies identify PI3K/Akt/ β-catenin/CBP signaling as playing a key role in mediating this
phenotypic plasticity. Therefore, therapeutic targeting of this pathway may be uniquely
effective at overcoming the therapy resistance and disease progression attributed to the
CSC phenotype.

Introduction:
Drug resistance and cancer progression constitute the central challenge in the
management of advanced malignancy (Dean et al., 2005; Kelland, 2007). While the
underlying biology of these phenomena has yet to be fully elucidated, a growing volume
of recent work has established that cancer cell populations are quite heterogeneous, and
that therapy resistance and brisk tumorigencity reside disproportionately in a relatively
small subpopulation of cancer cells sometimes termed “tumor initiating cells” or “cancer
stem-like” cells CSC (Reya et al., 2001; Visvader and Lindeman, 2008). As a result,
numerous ongoing preclinical and clinical studies aim to neutralize, differentiate or
eradicate cancer stem-like cells in order to avert drug resistance and halt disease
progression (Bar et al., 2007; Chan et al., 2009; Fan et al., 2006; Gupta et al., 2009b;
Hambardzumyan et al., 2008; Takahashi-Yanaga and Kahn, 2010; Yu et al., 2007a; Zhang
et al., 2008).

99
From a therapeutic perspective, a general consensus has emerged regarding the
desirability of eliminating cancer stem-like cells; however, from a biological standpoint,
the precise origin and function of these cells has generated significant uncertainty and
controversy (Clarke et al., 2006; Kim and Dirks, 2008), reflecting the great variability of
models, markers and assays devised to isolate and study them (Reya et al., 2001; Stuelten
et al., 2010; Visvader and Lindeman, 2008). Most early studies of the cancer stem-like
phenotype were conducted in models of hematopoietic malignancy and gave rise to the
presiding model that these cells – like their benign counterparts – were a
well-circumscribed, self-renewing seed population that arose hierarchically from some
common progenitor and in turn gave rise to cancer cell progeny (Bonnet and Dick, 1997;
Hope et al., 2004; Lapidot et al., 1994). However, several recent studies demonstrated
that even cancer cells lacking the putative cancer stem-like markers could nonetheless
demonstrate brisk tumorigenicity, a finding attributed to variability of markers and mouse
models or to contamination (Kelly et al., 2007; Quintana et al., 2008; Shmelkov et al.,
2008). Still other studies showed that under certain selective pressures (e.g. ectopic gene
expression), cell lacking cancer stem-like properties can acquire this phenotype de novo
(Gu et al., 2007; Mani et al., 2008; Sharma et al., 2010). Collectively, these reports
suggested that cancer stem-like cells might arise by more than one route, perhaps even
emerging de novo in an existing cancer cell population.

100
Recently, we reported that cancer stem-like cells could indeed arise by direct
conversion from non-cancer stem-like cells (He et al., 2011). Using fluorescence-labeled
cancer stem-like “side populations” (SP) isolated from cancer cell lines, we observed
significant plasticity whereby SP cells differentiated into “non-SP” (NSP) cells at the start
of each passage, followed by coordinated conversion of NSP cells back into the SP
phenotype towards late passage. This spontaneous, cyclical inter-conversion suggested a
high level of plasticity occurring without selection pressures such as ectopic gene
expression or drug treatment. Moreover, the rapid and coordinated nature of conversion
was not consistent with selection and amplification of rare stochastic events, but rather
was more consistent with a population-wide adaptive behavior, perhaps in response to
environmental signals. Indeed, a contemporaneous study reported within a week of our
own described a very similar observation of spontaneous plasticity (Chaffer et al., 2011).
These discoveries have cast the cancer stem-like phenotype in a new light: From a
biological perspective, a high degree of plasticity – if borne out in additional models –
would shift exiting models from a predominantly hierarchical framework to a more
functional definition whereby any population of cancer cells, given the right queues,
could shift to a cancer stem-like state. From a therapeutic standpoint, cancer stem-like
plasticity suggests that replenishment of the cancer stem-like population must be
contemplated and addressed by any new experimental agents. These considerations, both

101
biological and therapeutic, have motivated our group to identify signaling mechanisms
and pathways which mediate cancer stem-like plasticity.
The PI3K/Akt pathway is commonly activated and controls the survival and
proliferation of cancer cells; hence, it has been the target of intense developmental
therapeutic efforts (Engelman, 2009; Liu et al., 2009). Several recent reports have shown
that pharmacological inhibition of this pathway could effectively decrease the SP in
cancer cell populations (Bleau et al., 2009; Zhou et al., 2007), and sensitize cancer cells
to chemotherapeutic drugs in combination treatment, thus effectively attenuating the
drug-resistant, tumorigenic cancer stem-like phenotype (Dubrovska et al., 2010;
Knuefermann et al., 2003; Korkaya et al., 2009). These promising results have
highlighted the therapeutic potential of targeting the PI3K/Akt pathway, but they have left
the underlying mechanistic reasons unanswered: Did PI3K/Akt inhibition exert an
inhibitory effect specifically on cancer stem-like cells? Or, alternatively, did PI3K/Akt
and their downstream mediators function as switches that regulate the transition in and
out of a cancer stem-like phenotype observed in our previous plasticity experiments?
To investigate these questions, we tracked the drug-resistant, highly tumorigenic side
population of cancer cell lines in response to pharmacological and genetic perturbation of
the PI3K/Akt pathway and its downstream mediators. Using fluorescence activated cell
sorting (FACS) and GFP labeling, we observed that spontaneous NSP to SP conversion

102
was virtually abrogated by treatment with a pharmacological PI3K inhibitor. The
PI3K/Akt pathway mediates the signaling of wnt/ β-catenin pathway and regulates tumor
progression in various cancer types through a shared downstream factor—GSK3 β
(Castellone et al., 2009; Korkaya et al., 2009; Song et al., 2009). Here, we found that
GSK3 β inhibition increased β-catenin nuclear translocation with a concomitant rise in SP,
whereas β-catenin knock-down was associated with a drop in SP. Still further
downstream, β-catenin has been shown to interact with two transcriptional co-activators
in a switch-like manner: cAMP response element-binding protein (CREB)-binding
protein (CBP) is a critical factor in long-term maintenance of embryonic stem cells and
functions as a crucial co-activator for TCF/ β-catenin mediated survivin transcription (Ma
et al., 2005; Miyabayashi et al., 2007; Takahashi-Yanaga and Kahn, 2010), whose
expression is important for drug resistance and cancer relapse (Park et al., 2011;
Shoeneman et al., 2012). In contrast, P300 is a transcriptional co-activator whose
interaction with β-catenin has been shown to potentiate a differentiated state (Ma et al.,
2005; Miyabayashi et al., 2007). Here, we showed that potentiating the β-catenin-CBP
interaction (by inhibiting the competing p300- β-catenin binding) significantly increased
SP, while inhibiting the β-catenin-CBP interaction significantly decreased SP, even in the
setting of increased upstream nuclear translocation of β-catenin. All pharmacological and
molecular perturbations (both upstream and downstream) had little or no effect on SP

103
proliferation per se, consistent with a direct effect on phenotypic plasticity through
transcriptional regulation. Taken together, these findings define a novel role for
PI3K/Akt/ β-catenin-CBP/p300 pathway in the regulation of the cancer stem-like, highly
tumorigenic, drug resistant state. As such, members of this pathway play a direct
mechanistic role in cancer stem-like plasticity, warranting further study and therapeutic
targeting.

Materials and Methods
Cell culture and reagents. Cancer cell lines were obtained from collaborators at the
University of Southern California (see Acknowledgements) and were not re-authenticated
prior to use in these experiments. Human bladder cancer cells (J82) and human breast
cancer cells (MCF7) were maintained at 37°C, 5% CO
2
in DMEM (Mediatech)
supplemented with 10% of heat-inactivated fetal bovine serum (Omega), penicillin (100
units/ml, Invitrogen), and streptomycin (100 µg/ml, Invitrogen).
LY294002, Hoechst33342, verapamil and propidium iodide were purchased from
Sigma; BIO was obtained from Millipore; ICG001 and IQ-1 were provided by Dr.
Michael Kahn. Antibodies to pan-Akt, phosphor-Akt (Ser473) were purchased from Cell
Signaling Technology; GAPDH was from Millipore cooperation; β-catenin antibody was
from BD Bioscience; CBP and P300 antibodies were from Santa Cruz Biotechnology.

104
siRNA to Akt and β-catenin were purchased from Cell Signaling Technology;
Negative siRNA control and transfection reagent were from QIAGEN.
Lenti-virus infection: Lentivirus was generated as previously described to deliver
either GFP under a CMV promoter or control empty vector (Xu et al., 2010b). One day
prior to infection 2X10
5
J82 cells were seeded in 10cm plates, and on the next morning
media was replaced by 3 mL virus supernatant plus 7 mL media supplemented with 8
µg/ml polybrene. After 8 hr incubation at 37°C, the virus-containing media was replaced
with fresh media. Cells were observed for 48 hr to ensure >90% GFP expression prior to
fluorescence activated cell sorting (FACS) and side population studies.
Flow cytometry. Hoechst staining and FACS were conducted as described
previously (Goodell et al., 1996). Briefly, adherent cancer cells (1X10
6
/mL) were
trypsinized, counted, and resupended in prewarmed 10% FBS DMEM media. Hoechst
33342 (Sigma-Aldrich) was added at concentration of 5 µg/mL, incubated for 2 hrs in
37°C water bath and gently inverted several times during the course of incubation.
Parallel sample aliquots were prepared in the presence of 50 µM verapamil
(Sigma-Aldrich), an ATP-binding cassette transporter family inhibitor, at room
temperature for 10min before adding the Hoechst 33342 dye. Cells were centrifuged at
1000rpm for 5 min after incubation and resuspended in ice-cold DMEM media.
Propidium iodide (Sigma-Aldrich) was added to the cells at a final concentration of 2

105
µg/mL. Samples were incubated for at least 5 min on ice before FACS analysis
(FACSAria and FACSLSR-II, BD Biosciences, both equipped with UV lasers).
Immunofluorescence microscopy. J82 and MCF7 (2x10
4
) were plated on coverslips
in 12-well plates, the next day, media were changed and drugs or DMSO control were
added. After 72 hrs treatment, cells were fixed by 10% formalin and stained as previously
described (Xu et al., 2010b) by primary antibody against β-catenin, secondary antibody
conjugated with Alexor 647 (Invitrogen) and DAPI for nuclear visualization. Coverslips
were mounted on slides, and β-catenin localization was analyzed using Zeiss Imager.Z1
microscope with Axiovision software at x63 magnification.
siRNA transfection. siRNAs were transfected using QIAGEN HiPerFect
transfection reagent following manufacture’s fast-forward transfection protocol. Briefly,
2x10
5
cells in 6-well plate shortly before transfection. siRNA or control siRNA were
diluted in serum free medium, and incubated with 10 uL HiPerFect reagent for 10 min at
room temperature. Complexes were added drop-wise onto cells. Media were changed the
next morning, and cells were harvested 72 hrs posttransfection. Knockdown of
endogenous protein levels were determined by western blot.
Western blot. 4%-12% Bis/Tris precast gels (Invitrogen) were used for pan-Akt,
pAkt, GSK3 β, pGSK3 β and β-catenin. 3%-8% Tris/acetate gels (Invitrogen) were used
for CBP and P300. Proteins were transferred using Invitrogen iBlot Dry Blotting system

106
(8 min transfer). The membranes were blocked with Odyssey blocking buffer (Li-COR),
and incubated with primary antibodies, secondary antibodies labeled with IRDye infrared
dyes and then detected with Infrared Imaging system (Li-COR).
Sphere formation assay. Sphere culture was done as previously described (PNAS
paper). 10
4
cells were plated in 6-well low attachment plate (with ultra-low attachment
surface, Corning Incorporated). Cells were grown in MEBM media (Lonza Inc.)
supplemented with B27 (Stem Cell Technologies), 20 ng/mL EGF, and 20 ng/mL basic
FGF (Invitrogen), 4 ug/mL insulin, and 1% penicillin/streptomycin. After 6 days, spheres
formed were counted under microscope visually.
Protein co-immunoprecipitation. For detection the association of CBP and P300
with β-catenin, J82 and MCF7 cells were treated with or without ICG001 or IQ-1for 24
hrs respectively. Cells were washed with PBS, and extracted for nuclear proteins by
Nulcear extraction kit (Millipore). Protein concentrations were determined by using
Coomassie (Bradford) protein assay kit (Thermo Scientific). 100~200 ug of total nuclear
proteins were diluted in Co-IP buffer (20 nM Tris-HCl (pH 8.0), 150 mM NaCl, 1.5 mM
MgCl
2
, 1 mM EDTA, 5 mM DTT, 0.5% NP-40, 10% glycerol and protease inhibitor
cocktail), and then incubated with 2 ug primary antibodies or rabbit IgG control at 4
o
C
overnight. 30 uL of a 50% slurry of protein-A agarose was added and incubated in co-IP
buffer for 1 hour at 4
o
C on a rotating platform. Beads were centrifuged at 1000xg for 1

107
min, and washed 3 times with co-IP buffer and twice with PBS, and then boiled with 30
uL co-IP buffer. After centrifuge at 6000xg for 1 min, supernatant was collected and
subjected to SDS-PAGE and western blot using a mouse monoclonal antibody for
β-catenin.
Statistical Analysis. All experiments were conducted in triplicate with error bars
representing standard deviation around the mean. Student’s t-test was used to determine
statistical significance when comparing mean values at one point in time.

Results
The simultaneous direct conversion of numerous NSP cells to SP cells can be
blocked by PI3K/Akt pathway inhibition. We have shown previously that
drug-resistant highly-tumorigenic SP cells can arise direct from the conversion of NSP
cells (He et al., 2011). Recent studies have also suggested that the PI3K/Akt signaling
pathway may be required for the maintenance of side population phenotype and
tumorigenicity (Bleau et al., 2009; Zhou et al., 2007), but how does this pathway regulate
side population phenotype—is it regulated through cell viability/proliferation or by
shifting the dynamic equilibrium between SP and NSP conversion? To answer this
question, we applied our GFP-based model to dynamically track the behaviors of SP and
NSP in the culturing environment. Briefly, J82 cells were infected with lentivirus

108
expressing either empty vector (EV) or GFP. J82
EV
and J82
GFP
were FACS sorted into
their respective SP and NSP subpopulations. These were re-combined to create a J82 cell
line consisting of ~20% SP
EV
and 80% NSP
GFP
. These cells were treated with or without
PI3K inhibitor—LY294002, and then analyzed by FACS for Hoechst dye efflux and GFP
status.
When analyzed (every 2 days) by FACS during in vitro culture, we found that in
DMSO control group, the hybrid J82 cells exhibited a dramatic drop of SP percentage in
the first 2 days, and then SP% started to increase gradually until day 6, a time point after
which SP started an explosive expansion, and this trend continued until the cells reached
confluence (day 10) (Figure 18A-B). However, in the LY294002 treated group, SP% only
had a minimum increase after day2, and remained at a very low level until day 10
compared to DMSO control. As shown by FACS plots in Figure 18A, in the first 2 days,
original SP cells (red) disappeared from SP gate, but reappeared in NSP population,
indicating that they differentiated into NSP cells rather than dying out from the major
population. After LY294002 was added, we found that GFP% in the whole population
remained relatively constant (~80%) , and showed no difference from DMSO group,
suggesting that original SP cells (GFP negative) were not selectively targeted by
LY294002 (Figure 18C). In fact, whole population GFP% decreased a little (from 88% to
81%), meaning original SP cells showed higher proliferation rate than original NSP cells

109
(GFP positive), which is very reasonable, considering SP cells are much more clonogenic
and tumorigenic than NSP cells.
The most significant difference was observed after day6 between DMSO control and
LY294002 treated group. We found in DMSO control, SP% dramatically expanded (from
1.5% at day6 to 25.2% at day10) after this time point, accompanied with a significant
increase of GFP% cells in the side population gate (27.3% at day6 to 62.4% at day10),
indicated that the SP expansion was greatly contributed by the conversion of GFP
positive NSP cells to SP cells (as we had shown previously, REF). However, in the
LY294002 treated group, side population percentage remained very low at day10
(~0.75%), and GFP positive cells in SP gate was only 30.45%, much lower than the
62.4% in DMSO group (Figure 18B), suggesting that few GFP positive NSP cells had
converted into SP cells after PI3K/Akt pathway was inhibited.
Activation of Akt signaling to potentiate conversion. Given the limited specificity
of pharmacological reagents, we also applied siRNA to specifically knockdown Akt
proteins. After 3 days of 50nM Akt siRNA treatment, western blot results clearly showed
knockdown of total Akt protein level, as well as marked decrease of phosphor-Akt
signaling. More importantly, inhibition of akt activation by siRNA significantly decreased
SP% in both J82 and MCF7 cells compared to treatment with scrambled sequence
(nonspecific) siRNA control (p=0.02 and p=0.03 respectively) (Figure 19A)

110
To determine whether activation of PI3K/Akt pathway can also regulate side
population, we used a pharmacological inhibitor bpV(pic) to inhibit PTEN, which results
in further activation of Akt signaling (Zhou et al., 2007). We found that bpV treatment
significantly increased SP in PTEN positive MCF7 cells (from ~3.5% to 5%, p=0.01),
while bpV exhibited no effects on J82 cells, a cell line which does not express PTEN
(Gildea et al., 2004) (Figure 19B).
GSK3 β/ β-catenin signaling functions downstream of PI3K/Akt pathway in
regulation of conversion. GSK3 β/ β-catenin signaling is a downstream signaling
mediator of PI3K/Akt pathway which has been shown to impact mammary tumor stem
cell properties in mouse models and in human cancer cells (Korkaya et al., 2009; Zhang
et al., 2010). To determine whether GSK3 β/ β-catenin signaling also regulates the
conversion from NSP to SP, we first applied a pharmacological inhibitor of GSK3 β —
BIO (6-bromoindirubin-3’-oxime) to inhibit this signaling. We found that BIO treated
cells showed nuclear accumulation of β-catenin in both J82 and MCF7 cell lines, a major
event in β-catenin signaling activation (Moon et al., 2004; Sato et al., 2004). However,
DMSO control exhibited mostly membrane and cytoplasm localization of β-catenin
(Figure 20A). In addition, induction of β-catenin nuclear translocation by BIO treatment
resulted in the significantly increase of SP% (from 6.9% to 16.3% in J82 and from 4.2%
to 12.2% in MCF7) (Figure 20B), consistent with previous findings that activation of

111
β-catenin by BIO can help to maintain pluripotency in embryonic stem cells as well as in
caner cells (Sato et al., 2004).
To exclude the possibility that pharmacological inhibitors may have non-specific
targets, and to further substantiate the role of β-catenin as a direct mediator of plasticity
in and out of the CSC phenotype, we used siRNAs to specifically knock down β-catenin
in cancer cells (Figure 20C). Western blots confirmed effective knockdown of β-catenin
protein (~70% in J82 and ~80% in MCF7), and cells transfected with siRNA after 72 hrs
were subject to FACS analysis for side population phenotype. As expected, side
populations were significantly diminished after β-catenin knockdown in both cell lines
(from 5% to ~1% in J82 and 5% to 0.5% in MCF7), while cells treated with control
scrambled siRNA were not significantly affected (p>0.1).
β-catenin/CBP signaling is required for NSP to SP conversion. Recent studies
have shown the β-catenin transcription co-activator -- CBP (CREB-binding protein) is
required for TCF/ β-catenin-mediated survivin gene expression, which is critical for the
inhibition of apoptosis pathway as well as cancer cell proliferation (Kumar et al., 2009;
Ma et al., 2005). Side population phenotype, which is recognized as being associated
with high drug resistance and tumorigenicity, might be regulated through the
β-catenin/CBP/survivin signaling. To test this hypothesis, we used a pharmacological
inhibitor – ICG001 to specifically interfere with β-catenin/CBP interaction (Emami et al.,

112
2004). We found that ICG001 treatment significantly decreased side population, from
10% to 2% in J82 (p=0.02) and ~4.5% to 1.5% in MCF7 (p=0.01) respectively (Figure
21A), and also ICG001 treatment significantly inhibited sphere formation (a
characteristic cancer stem-like property) in both cancer cell lines (p=0.03) (Figure 21B).
It is reported that another β-catenin coactivator – P300, as opposed to CBP, functions
to decrease the expression of survivin and other (eg. Oct4 and Sox2)(Ma et al., 2005).
In addition, treatment of small molecule drug IQ-1 decreased the P300 and β-catenin
interaction, and potentiated CBP signaling by preventing the switch of β-catenin
coactivator from CBP to P300 (Miyabayashi et al., 2007). To determine if activating CBP
signaling could affect side population phenotype, we used IQ-1 to increase the
CBP/ β-catenin interaction. We found that both cell lines showed significant increase of
SP after treatment with IQ-1 (1uM) for 3 days (Figure 21C), thus further confirming that
β-catenin/CBP signaling is required for regulating SP phenotype.
To confirm that ICG001 affects SP only through disrupting β-catenin/CBP
interaction, and to exclude the possibility that ICG001 may non-specifically target
β-catenin itself, or even some upstream pathway that may inhibit β-catenin activation, we
treated cells with ICG001, and immunocytochemistry results showed no difference of
β-catenin localization compared to DMSO control (Figure 21D). In addition, the
activation of β- catenin by BIO was not affected by ICG001 co-treatment (Figure 21D),

113
confirming that ICG001 may only affect the downstream activity of β-catenin. This
hypothesis was further supported by FACS results which showed that ICG001 treatment
significantly abrogated the BIO effects in potentiating SP, and dramatically reduced the
side population (from 25% in J82 and 12% in MCF7 after BIO induction to around 4% in
J82 and 5% in MCF7 respectively) (Figure 21E).
β-catenin/CBP signaling regulates regeneration of SP from NSP. We have already
demonstrated that the β-catenin/CBP signaling could affect the drug resistant and
tumorigenic side population phenotype, but whether this regulation is mediated through
targeting the SP cell proliferation and survival, or by directly controlling the conversion
of NSP to SP cells are still not very clear. To answer this question, we tested whether
perturbing the GSK3 β/ β-catenin/CBP signaling would impact the regeneration of SP
from FACS purified NSP cells. Sorted NSP cells were cultured with or without
pharmacological inhibitors (or DMSO control) to either activate or inhibit this pathway,
and the regeneration of SP from NSP was monitored by FACS on subsequent days. As
shown in Figure 22, BIO significantly induced SP regeneration after 7 days (p<0.01),
while ICG001 significantly prevented SP regeneration even after 10 days in culture
(p<0.01). Furthermore, results of concurrent treatment with both drugs demonstrated that
ICG001 abrogated BIO induced SP regeneration from NSP (SP decrease from 32.4% to
5.4% in J82 at day10, and from 18.2% to 2.1% in MCF7 at day12). To determine whether

114
the GSK3 β/ β-catenin/CBP signaling selectively targets the proliferation and survival of
SP cells, we examined the effects of these inhibitors on the proliferation and viability of
SP and NSP cells. As the MTS proliferation assay results show in Figure 22, we observed
that BIO exhibited little effects on SP and NSP growth in J82 cells. Although BIO
inhibited MCF7 cells growth after longer treatment time (after day 4), there is no
difference in growth inhibition between SP and NSP cells. Similarly, ICG001 alone or
co-treatment with BIO showed inhibition effects on both J82 and MCF7 cells, but
ICG001 did not selectively target either SP or NSP group. Taken together, these results
suggested that GSK3 β/ β-catenin/cBP pathway directly regulates NSP to SP conversion
rather than differentially impacting the proliferation rates of these subpopulations; at the
same time, this set of experiments eliminated the possibility that newly regenerated SP
cells originated from contamination with SP cells during sorting.

115

Figure 18. Inhibition of PI3K/Akt pathway diminishes SP and inhibits NSP to SP conversion. (A)
GFP-labeling experiment shows significant decrease in conversion from GFP
+
NSP to SP with
LY294002 treatment. (B) Decrease of SP percentage (top graph) and GFP
+
cells in SP (bottom table)
after treatment with 10 µM LY294002. (C) GFP% in whole population shows no significant difference
with or without LY294002 treatment.

116

Figure 19. Akt signaling affects side population. (A) siRNA knockdown of total Akt and
phosphor-Akt protein reduces SP in J82 and MCF7 cells. (B) PTEN inhibition with bpV(pic)
increases SP in MCF7 cells (PTEN wt), but does not affect SP in J82 (PTEN null).

117

Figure 20. β-catenin signaling regulates side population. (A) GSK3 β inhibition with BIO induces
β-catenin nuclear localization. (B) Induction of β-catenin nuclear translocation results in SP
expansion. (C) siRNA knockdown of β-catenin protein decreases side population.

118

Figure 21. CBP functions downstream of β-catenin signaling to regulate side population
phenotype. (A) Inhibition of CBP/ β-catenin interaction ICG001 significantly decreases SP. (B)
ICG001 prevents sphere formation, a cancer stem-like phenotype. (C) Promoting CBP/ β-catenin
interaction by IQ-1 increases SP. (D) ICG001 treatment does not affect β-catenin nuclear
translocation. (E) ICG001 abrogates BIO-induced SP expansion. ( ∗: p<0.05; ∗∗: p<0.01)

119

Figure 22. GSK3 β/ β-catenin/CBP signaling regulates regeneration of SP from NSP. (A)
Percentage of SP regenerated from pre-sorted NSP cells with or without pharmacological
perturbations of GSK3 β/ β-catenin/CBP signaling. (B) Growth inhibition of SP compared to NSP
with BIO, ICG001 or co-treatment of BIO and ICG001. ( ∗: p<0.05; ∗∗: p<0.01)

120
Discussion:
Recently we demonstrated the drug resistant and highly tumorigenic side population
is not a discrete population but actually can arise from the differentiated non-side
population of cancer cells in a highly-orchestrated dynamic equilibrium (He et al., 2011).
The replenishment of SP from the bulk NSP suggests a phenotypic plasticity between
drug resistant, highly tumorigenic state and the drug sensitive, low tumorigenic state, and
that heterogeneity in cancer cells could spontaneously arise from a homogeneous
population or even single cell of either state through dynamic conversion. Therefore,
efforts focusing on eradicating the drug resistant highly tumorigenic population may only
have a temporary effect (Scheel and Weinberg, 2011), as more differentiated cancer cells
that originally lack these properties may emerge as new “CSC” de novo with
drug-tolerant and tumorigenic state (Roesch et al., 2010; Sharma et al., 2010). In this
study, we sought to elucidate the mechanisms which may govern this conversion process,
and to provide evidence for pursuing therapeutic strategies to prevent regeneration of
drug resistant and tumorigenic cells.
As one of the most commonly activated pathway in cancer (Liu et al., 2009),
PI3K/Akt signaling has previously been reported to control the tumorigenicity and
therapy resistance in different types of cancer cells, through regulating cell proliferation
and protection machinery to prevent apoptosis (Dubrovska et al., 2009; Hambardzumyan

121
et al., 2008; Knuefermann et al., 2003). In addition, activation of PI3K/Akt pathway can
induce an expansion of cancer stem-like population (determined by CD133+, CD44+,
SP), while inhibition of this signaling results in differentiation and decease of CSC
population (Dubrovska et al., 2009; Vermeulen et al., 2008). According to the
unidirectional CSC model, it could be postulated that PI3K/Akt signaling plays a critical
role in CSC, but not in non-CSC, to regulate their self-renewal and maintenance of CSC
population. The reason behind this divergence of signaling response in different cancer
subpopulations may be due to differential activation of endogenous PI3K/Akt signaling,
as studies have shown that cancer cells with high self-renewal and tumorigenic abilities
are usually associated with elevated PI3K/Akt activation (Bleau et al., 2009; Dubrovska
et al., 2009; Hambardzumyan et al., 2008; Zhou et al., 2007), and this is generally
considered as a result of highly organized hierarchy in cancer cells. However, the
findings of spontaneous dedifferentiation in cancer, and recent studies in iPS cells and
EMT have provided a new perspective that the activation of signaling network in cancer
subpopulations may not be a result of highly organized hierarchy, but on the contrary,
reactivation of these pathways and some pluripotency factors could be the driving force
to generate phenotypic heterogeneity in cancer. This suggests the possibility that
reactivation of certain signaling pathways (like PI3K/Akt and further downstream
mediators) in non-CSC could potentially increase the phenotypic plasticity and generate

122
drug resistant, highly tumorigenic population de novo, which serves as an alternative
route to replenish CSC in addition to CSC self-renewal.
To test this possibility, we used a previously described GFP-labeling experiment to
track the behavior of drug-resistant, tumorigenic side population with or without the
inhibition of PI3K/Akt signaling, and we found that SP% in LY294002 treatment group
remained at a very low level, and few SP cells were regenerated compared to DMSO
control. This finding can be explained in 2 different ways: inhibition of PI3K may
preferentially prevent the self-renewal and growth of SP cells, while the NSP cells are
less affected. But if this was true, we would expect to see a significant decrease of
original GFP
-
SP cells (red) and largely increase of GFP
+
cells (originally NSP) in the
whole population after LY treatment. However, the total GFP% shows little difference
between the two groups (Figure 18C), indicating LY294002 does not selectively target SP
cells. The alternative explanation is that blocking PI3K/Akt pathway decreases the
phenotypic plasticity and inhibits the regeneration of SP cells from NSP cells, which is
well supported by the results showing marked reduction of GFP
+
NSP cells to convert
into SP cells. Even though we still see a minimum regeneration of SP cells (~1%) at day
10, the majority of those new SP cells are GFP
-
, indicating their origin as initial GFP
-
SP
cells.

123
Wnt/ β-catenin pathway is another important pathway that has been reported to
promote therapy resistance and tumorigenic capacity in various types of cancers
(Vermeulen et al., 2010; Woodward et al., 2007; Zhang et al., 2010). Recent studies have
shown that PI3K/Akt and Wnt signaling can converge at the common downstream
mediator GSK3 β to further regulate β-catenin signaling (Castellone et al., 2009; Song et
al., 2009). Our results have demonstrated that β-catenin signaling is required to maintain
the SP phenotype; however, overexpression of β-catenin protein alone is not sufficient to
induce SP expansion (data not shown), indicating the possibility that other factors may
also be recruited in combination with β-catenin to regulate SP. CBP and P300 are 2
co-activators on TCF/ β-catenin mediated signal transduction, but play differential roles in
controlling differentiation in embryonic stem cells and tumorigenic abilities in cancer
cells (Emami et al., 2004; Miyabayashi et al., 2007). By using ICG001 to interrupt
β-catenin/CBP signaling, we found significant inhibition of self-renewal ability in cancer
cells (sphere formation ability) and dramatic decrease of SP, indicating a promoted
differentiation process. More importantly, inhibition of CBP signaling by ICG001
prevents direct regeneration of SP from NSP counterpart without affecting upstream
β-catenin activation, and abrogates BIO induced SP expansion, indicating CBP functions
downstream of β-catenin to regulate phenotypic plasticity in cancer cells.

124
Our findings suggest that PI3K/Akt/GSK3 β/ β-catenin/CBP signaling plays an
important role in regulating the plasticity of drug resistant/highly tumorigenic phenotype,
and suggest the possibility that tumorigenic potential in cancer may not limit to a static
population of cancer cells, and cancer population do not necessarily need to follow a
strict hierarchy of differentiation pattern. Instead, activation of this signaling pathway
could reverse the differentiation process and promote the regeneration of drug resistant
and tumorigenic properties.

125
Chapter 5
Summary and future directions

Summary
Cancer stem-like cells (CSC) – a subpopulation of self-renewing, tumorigenic, drug
resistant tumor cells – are thought to promote cancer formation, therapy resistance and
disease progression. Therefore, CSC are being intensely studied, and new therapeutic
strategies are being developed to specifically target CSC. My graduate research work has
aimed to target and better understand the CSC phenotype.
The first part of my research (in collaboration with Dr. Tong Xu in the Goldkorn Lab)
aimed to use telomerase interference to specifically target CSC in prostate tumor.
Telomerase interference is a two-pronged approach consisting of: 1. siRNA to
knockdown wild type human telomerase RNA (hTer) template, and 2. replacing wild type
telomerase RNA by a new mutant-template (MT-hTer) RNA. Active telomerase
incorporates with MT-hTer and adds wrong sequence to telomere ends, inducing telomere
uncapping and rapid apoptosis. In this project, we purified α2β1
high
CD44
+
cells from
both primary prostate tumor and cancer cell lines, and characterized them as highly
clonogenic, invasive and tumorigenic cancer stem-like phenotype compared to CD44
-

cells. Significantly, those α2β1
high
CD44
+
CSC exhibited 20 to 200-fold higher telomerase
activity than in non-CPC from the same tumors, and CSC were exquisitely sensitive to

126
telomerase interference which induced rapid apoptosis and growth inhibition both in vitro
and in vivo.
The second part of my work focused on the phenotypic plasticity of CSC. Until very
recently, CSC have been widely conceptualized as a small population of self-renewing
founder cells from which all other cancer cells arise, known as the unidirectional model.
However, in my recent studies we used flow cytometry and GFP labeling to carefully
track CSC-like side populations (SP) and non-side populations (NSP) of cancer cells, and
we observed evidence to challenge the unidirectional model. Based on this evidence, we
proposed a novel alternative bidirectional model based on our discovery of a dynamic
two-way equilibrium: CSCs indeed generate differentiated cancer cells, but the
differentiated cancer cells in turn de-differentiate and replenish the CSC population. Our
studies demonstrated for the first time that CSCs are not a small and static group of
founder cells; rather, they are continuously depleted and repleted in an ongoing state of
dynamic equilibrium with differentiated cancer cells.
My third project has focused on exploring the possible mechanisms that may regulate
this phenotypic plasticity in cancer, Beginning with a pathway already implicated in the
CSC phenotype, we showed that PI3K/Akt/ β-catenin/CBP signaling plays a key role in
mediating CSC phenotypic plasticity. NSP-to-SP conversion was significantly reduced by
exposure to pharmacological manipulation as well as siRNA knock-down of key

127
members of this pathway, from PI3K to AKT, to GSK-3 β to β-catenin and finally to its
transcriptional partners CBP and p300. Conversely, activating this pathway significantly
induced regeneration of CSC from non-CSC population.

Future directions
We realize that signaling pathways do not function in isolation, and cannot be
presented or considered in a simple linear fashion. A more accurate picture is that
signaling pathways are linked together in a large protein network that is subjected to
multiple stimulatory and inhibitory inputs, as well as complex feedback mechanisms.
Therefore, we have initiated more comprehensive transcriptome and epigenome
comparisons of CSC and non-CSC to identify still more pathways and gene targets that
are involved in CSC plasticity. We have already completed the whole genome DNA
methylation analysis for SP vs NSP cells (J82) in collaboration with Dr. Peter Laird at the
USC Epigenome Center. We found that SP cells generally exhibited a more
hypomethylated status; interestingly, the differentially hypomethylated loci were found
mostly at non-CpG island regions (Figure 23). In a preliminary analysis, out of 450,000
probes analyzed, we filtered out 92 probes/genes that have shown significant difference
in DNA methylation (absolute ∆ β value > 0.2) for further study (Table 6). At the same
time, we have initiated ChIP-seq experiments to study the differential modifications

128
between SP and NSP on histone marks (H3K4me, H3K9Ac, H3K27Ac), as some reports
have shown that histone acetyltransferases CBP and P300 can mediate histone
acetylation and gene activation (Crump et al., 2011; Rada-Iglesias et al., 2011; Wang et
al., 2010). Furthermore, we are also interested in comparing the whole transcriptome by
using RNA-seq for SP vs. NSP cells, and correlate the RNA-seq results with DNA
methylation and ChIP-seq results. These high throughput discovery platforms will yield
interesting new correlations and candidates that may help to identify new mediators and
pathways that regulate the phenotypic plasticity of cancer cells.
In summary, my studies have demonstrated that cancer stem cells have high levels of
telomerase expression and activity that can be effectively targeted by telomerase
interference therapy. Moreover, we have provided evidence that cancer cells may have
the capacity to spontaneously transition in and out of the cancer stem cell phenotype, and
such phenotypic changes may be regulated by PI3K/Akt/ β-catenin/CBP signaling. These
and future studies will help us to catalog and better-understand the signaling networks
which underlie CSC plasticity, ultimately leading to more effective control of cancer
progression.

129
References

Akhtar, K., Bussen, W., and Scott, S.P. (2009). Cancer stem cells - from initiation to elimination,
how far have we reached? (Review). Int J Oncol 34, 1491-1503.

Al-Hajj, M., Wicha, M.S., Benito-Hernandez, A., Morrison, S.J., and Clarke, M.F. (2003).
Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci U S A 100,
3983-3988.

Alix-Panabieres, C., Muller, V., and Pantel, K. (2007). Current status in human breast cancer
micrometastasis. Curr Opin Oncol 19, 558-563.

Allsopp, R.C., Vaziri, H., Patterson, C., Goldstein, S., Younglai, E.V., Futcher, A.B., Greider,
C.W., and Harley, C.B. (1992). Telomere length predicts replicative capacity of human fibroblasts.
Proc Natl Acad Sci U S A 89, 10114-10118.

Aoki, H., Iwado, E., Eller, M.S., Kondo, Y., Fujiwara, K., Li, G.Z., Hess, K.R., Siwak, D.R.,
Sawaya, R., Mills, G.B., et al. (2007). Telomere 3' overhang-specific DNA oligonucleotides
induce autophagy in malignant glioma cells. FASEB J 21, 2918-2930.

Asai, A., Oshima, Y., Yamamoto, Y., Uochi, T.A., Kusaka, H., Akinaga, S., Yamashita, Y.,
Pongracz, K., Pruzan, R., Wunder, E., et al. (2003). A novel telomerase template antagonist
(GRN163) as a potential anticancer agent. Cancer Res 63, 3931-3939.

Bao, S., Wu, Q., McLendon, R.E., Hao, Y., Shi, Q., Hjelmeland, A.B., Dewhirst, M.W., Bigner,
D.D., and Rich, J.N. (2006). Glioma stem cells promote radioresistance by preferential activation
of the DNA damage response. Nature 444, 756-760.

Bar, E.E., Chaudhry, A., Lin, A., Fan, X., Schreck, K., Matsui, W., Piccirillo, S., Vescovi, A.L.,
DiMeco, F., Olivi, A., et al. (2007). Cyclopamine-mediated hedgehog pathway inhibition depletes
stem-like cancer cells in glioblastoma. Stem Cells 25, 2524-2533.

Blackburn, E.H. (2000). Telomere states and cell fates. Nature 408, 53-56.

Bleau, A.M., Hambardzumyan, D., Ozawa, T., Fomchenko, E.I., Huse, J.T., Brennan, C.W., and
Holland, E.C. (2009). PTEN/PI3K/Akt pathway regulates the side population phenotype and
ABCG2 activity in glioma tumor stem-like cells. Cell Stem Cell 4, 226-235.

130
Bodnar, A.G., Ouellette, M., Frolkis, M., Holt, S.E., Chiu, C.P., Morin, G.B., Harley, C.B., Shay,
J.W., Lichtsteiner, S., and Wright, W.E. (1998). Extension of life-span by introduction of
telomerase into normal human cells. Science 279, 349-352.

Bonnet, D., and Dick, J.E. (1997). Human acute myeloid leukemia is organized as a hierarchy
that originates from a primitive hematopoietic cell. Nat Med 3, 730-737.

Bravaccini, S., Sanchini, M.A., Amadori, A., Medri, L., Saragoni, L., Calistri, D., Monti, F., V olpi,
A., and Amadori, D. (2005). Potential of telomerase expression and activity in cervical specimens
as a diagnostic tool. J Clin Pathol 58, 911-914.

Brown, M.D., Gilmore, P.E., Hart, C.A., Samuel, J.D., Ramani, V.A., George, N.J., and Clarke,
N.W. (2007). Characterization of benign and malignant prostate epithelial Hoechst 33342 side
populations. Prostate 67, 1384-1396.

Brunsvig, P.F., Aamdal, S., Gjertsen, M.K., Kvalheim, G., Markowski-Grimsrud, C.J., Sve, I.,
Dyrhaug, M., Trachsel, S., Moller, M., Eriksen, J.A., et al. (2006). Telomerase peptide
vaccination: a phase I/II study in patients with non-small cell lung cancer. Cancer Immunol
Immunother 55, 1553-1564.

Burger, A.M., Dai, F., Schultes, C.M., Reszka, A.P., Moore, M.J., Double, J.A., and Neidle, S.
(2005). The G-quadruplex-interactive molecule BRACO-19 inhibits tumor growth, consistent
with telomere targeting and interference with telomerase function. Cancer Res 65, 1489-1496.

Cairns, R.A., Harris, I.S., and Mak, T.W. (2011). Regulation of cancer cell metabolism. Nat Rev
Cancer 11, 85-95.

Carey, L.A., Kim, N.W., Goodman, S., Marks, J., Henderson, G., Umbricht, C.B., Dome, J.S.,
Dooley, W., Amshey, S.R., and Sukumar, S. (1999). Telomerase activity and prognosis in primary
breast cancers. J Clin Oncol 17, 3075-3081.

Castellone, M.D., De Falco, V ., Rao, D.M., Bellelli, R., Muthu, M., Basolo, F., Fusco, A., Gutkind,
J.S., and Santoro, M. (2009). The beta-catenin axis integrates multiple signals downstream from
RET/papillary thyroid carcinoma leading to cell proliferation. Cancer Res 69, 1867-1876.

Cesare, A.J., and Reddel, R.R. (2010). Alternative lengthening of telomeres: models, mechanisms
and implications. Nat Rev Genet 11, 319-330.

131
Chaffer, C.L., Brueckmann, I., Scheel, C., Kaestli, A.J., Wiggins, P.A., Rodrigues, L.O., Brooks,
M., Reinhardt, F., Su, Y., Polyak, K., et al. (2011). Normal and neoplastic nonstem cells can
spontaneously convert to a stem-like state. Proc Natl Acad Sci U S A 108, 7950-7955.

Chan, K.S., Espinosa, I., Chao, M., Wong, D., Ailles, L., Diehn, M., Gill, H., Presti, J., Jr., Chang,
H.Y., van de Rijn, M., et al. (2009). Identification, molecular characterization, clinical prognosis,
and therapeutic targeting of human bladder tumor-initiating cells. Proc Natl Acad Sci U S A 106,
14016-14021.

Chikazawa, N., Tanaka, H., Tasaka, T., Nakamura, M., Tanaka, M., Onishi, H., and Katano, M.
(2010). Inhibition of Wnt signaling pathway decreases chemotherapy-resistant side-population
colon cancer cells. Anticancer Res 30, 2041-2048.

Choi, J., Southworth, L.K., Sarin, K.Y., Venteicher, A.S., Ma, W., Chang, W., Cheung, P., Jun, S.,
Artandi, M.K., Shah, N., et al. (2008). TERT promotes epithelial proliferation through
transcriptional control of a Myc- and Wnt-related developmental program. PLoS Genet 4, e10.

Clarke, M.F., Dick, J.E., Dirks, P.B., Eaves, C.J., Jamieson, C.H., Jones, D.L., Visvader, J.,
Weissman, I.L., and Wahl, G.M. (2006). Cancer stem cells--perspectives on current status and
future directions: AACR Workshop on cancer stem cells. Cancer Res 66, 9339-9344.

Cohen, S.B., Graham, M.E., Lovrecz, G.O., Bache, N., Robinson, P.J., and Reddel, R.R. (2007).
Protein composition of catalytically active human telomerase from immortal cells. Science 315,
1850-1853.

Collins, A.T., Berry, P.A., Hyde, C., Stower, M.J., and Maitland, N.J. (2005). Prospective
identification of tumorigenic prostate cancer stem cells. Cancer Res 65, 10946-10951.

Collins, A.T., Habib, F.K., Maitland, N.J., and Neal, D.E. (2001). Identification and isolation of
human prostate epithelial stem cells based on alpha(2)beta(1)-integrin expression. J Cell Sci 114,
3865-3872.

Collins, K. (2006). The biogenesis and regulation of telomerase holoenzymes. Nat Rev Mol Cell
Biol 7, 484-494.

Collins, K. (2008). Physiological assembly and activity of human telomerase complexes. Mech
Ageing Dev 129, 91-98.

132
Crump, N.T., Hazzalin, C.A., Bowers, E.M., Alani, R.M., Cole, P.A., and Mahadevan, L.C.
(2011). Dynamic acetylation of all lysine-4 trimethylated histone H3 is evolutionarily conserved
and mediated by p300/CBP. Proc Natl Acad Sci U S A 108, 7814-7819.

Cunha, G.R., Sekkingstad, M., and Meloy, B.A. (1983). Heterospecific induction of prostatic
development in tissue recombinants prepared with mouse, rat, rabbit and human tissues.
Differentiation 24, 174-180.

Das, B., Tsuchida, R., Malkin, D., Koren, G., Baruchel, S., and Yeger, H. (2008). Hypoxia
enhances tumor stemness by increasing the invasive and tumorigenic side population fraction.
Stem Cells 26, 1818-1830.

de Lange, T. (2005). Shelterin: the protein complex that shapes and safeguards human telomeres.
Genes Dev 19, 2100-2110.

de Lange T, I.V., Blackburn, E. H. (2005). Telomeres (New York, Cold spring harbor Laboratory
Press).

Dean, M., Fojo, T., and Bates, S. (2005). Tumour stem cells and drug resistance. Nat Rev Cancer
5, 275-284.

Diehn, M., Cho, R.W., Lobo, N.A., Kalisky, T., Dorie, M.J., Kulp, A.N., Qian, D., Lam, J.S.,
Ailles, L.E., Wong, M., et al. (2009). Association of reactive oxygen species levels and
radioresistance in cancer stem cells. Nature 458, 780-783.

Dikmen, Z.G., Gellert, G.C., Jackson, S., Gryaznov, S., Tressler, R., Dogan, P., Wright, W.E., and
Shay, J.W. (2005). In vivo inhibition of lung cancer by GRN163L: a novel human telomerase
inhibitor. Cancer Res 65, 7866-7873.

Doi, T., Shibata, K., Yoshida, M., Takagi, M., Tera, M., Nagasawa, K., Shin-Ya, K., and
Takahashi, T. (2011). (S)-Stereoisomer of telomestatin as a potent G-quadruplex binder and
telomerase inhibitor. Org Biomol Chem 9, 387-393.

Domchek, S.M., Recio, A., Mick, R., Clark, C.E., Carpenter, E.L., Fox, K.R., DeMichele, A.,
Schuchter, L.M., Leibowitz, M.S., Wexler, M.H., et al. (2007). Telomerase-specific T-cell
immunity in breast cancer: effect of vaccination on tumor immunosurveillance. Cancer Res 67,
10546-10555.

133
Drygin, D., Siddiqui-Jain, A., O'Brien, S., Schwaebe, M., Lin, A., Bliesath, J., Ho, C.B., Proffitt,
C., Trent, K., Whitten, J.P., et al. (2009). Anticancer activity of CX-3543: a direct inhibitor of
rRNA biogenesis. Cancer Res 69, 7653-7661.

Dubrovska, A., Elliott, J., Salamone, R.J., Kim, S., Aimone, L.J., Walker, J.R., Watson, J.,
Sauveur-Michel, M., Garcia-Echeverria, C., Cho, C.Y., et al. (2010). Combination therapy
targeting both tumor-initiating and differentiated cell populations in prostate carcinoma. Clin
Cancer Res 16, 5692-5702.

Dubrovska, A., Kim, S., Salamone, R.J., Walker, J.R., Maira, S.M., Garcia-Echeverria, C.,
Schultz, P.G., and Reddy, V.A. (2009). The role of PTEN/Akt/PI3K signaling in the maintenance
and viability of prostate cancer stem-like cell populations. Proc Natl Acad Sci U S A 106,
268-273.

Eads, C.A., Danenberg, K.D., Kawakami, K., Saltz, L.B., Blake, C., Shibata, D., Danenberg, P.V .,
and Laird, P.W. (2000). MethyLight: a high-throughput assay to measure DNA methylation.
Nucleic Acids Res 28, E32.

Eads, C.A., Danenberg, K.D., Kawakami, K., Saltz, L.B., Danenberg, P.V ., and Laird, P.W. (1999).
CpG island hypermethylation in human colorectal tumors is not associated with DNA
methyltransferase overexpression. Cancer Res 59, 2302-2306.

Eads, C.A., Lord, R.V., Wickramasinghe, K., Long, T.I., Kurumboor, S.K., Bernstein, L., Peters,
J.H., DeMeester, S.R., DeMeester, T.R., Skinner, K.A., et al. (2001). Epigenetic patterns in the
progression of esophageal adenocarcinoma. Cancer Res 61, 3410-3418.

Eddy, J., and Maizels, N. (2006). Gene function correlates with potential for G4 DNA formation
in the human genome. Nucleic Acids Res 34, 3887-3896.

Egan, E.D., and Collins, K. (2010). Specificity and stoichiometry of subunit interactions in the
human telomerase holoenzyme assembled in vivo. Mol Cell Biol 30, 2775-2786.

Eller, M.S., Puri, N., Hadshiew, I.M., Venna, S.S., and Gilchrest, B.A. (2002). Induction of
apoptosis by telomere 3' overhang-specific DNA. Exp Cell Res 276, 185-193.

Emami, K.H., Nguyen, C., Ma, H., Kim, D.H., Jeong, K.W., Eguchi, M., Moon, R.T., Teo, J.L.,
Kim, H.Y., Moon, S.H., et al. (2004). A small molecule inhibitor of beta-catenin/CREB-binding
protein transcription [corrected]. Proc Natl Acad Sci U S A 101, 12682-12687.

134
Engelman, J.A. (2009). Targeting PI3K signalling in cancer: opportunities, challenges and
limitations. Nat Rev Cancer 9, 550-562.

F. E. Millard, M.G., D. Darrah, P. Farness, M. Zanetti (2004). Phase I study of transgenic B
lymphocyte immunization (TLI) against telomerase in androgen-independent prostate cancer
(PC). Paper presented at: ASCO.

Fan, X., Matsui, W., Khaki, L., Stearns, D., Chun, J., Li, Y.M., and Eberhart, C.G. (2006). Notch
pathway inhibition depletes stem-like cells and blocks engraftment in embryonal brain tumors.
Cancer Res 66, 7445-7452.

Forsyth, N.R., Wright, W.E., and Shay, J.W. (2002). Telomerase and differentiation in
multicellular organisms: turn it off, turn it on, and turn it off again. Differentiation 69, 188-197.

Fruh, M., Rolland, E., Pignon, J.P., Seymour, L., Ding, K., Tribodet, H., Winton, T., Le Chevalier,
T., Scagliotti, G.V., Douillard, J.Y., et al. (2008). Pooled analysis of the effect of age on adjuvant
cisplatin-based chemotherapy for completely resected non-small-cell lung cancer. J Clin Oncol 26,
3573-3581.

Fukunaga-Kalabis, M., Martinez, G., Nguyen, T.K., Kim, D., Santiago-Walker, A., Roesch, A.,
and Herlyn, M. (2010). Tenascin-C promotes melanoma progression by maintaining the
ABCB5-positive side population. Oncogene.

Gangemi, R., Paleari, L., Orengo, A.M., Cesario, A., Chessa, L., Ferrini, S., and Russo, P. (2009).
Cancer stem cells: a new paradigm for understanding tumor growth and progression and drug
resistance. Curr Med Chem 16, 1688-1703.

Garraway, I.P., Sun, W., Tran, C.P., Perner, S., Zhang, B., Goldstein, A.S., Hahm, S.A., Haider, M.,
Head, C.S., Reiter, R.E., et al. (2010). Human prostate sphere-forming cells represent a subset of
basal epithelial cells capable of glandular regeneration in vivo. Prostate 70, 491-501.

Gildea, J.J., Herlevsen, M., Harding, M.A., Gulding, K.M., Moskaluk, C.A., Frierson, H.F., and
Theodorescu, D. (2004). PTEN can inhibit in vitro organotypic and in vivo orthotopic invasion of
human bladder cancer cells even in the absence of its lipid phosphatase activity. Oncogene 23,
6788-6797.

135
Goldkorn, A., and Blackburn, E.H. (2006). Assembly of mutant-template telomerase RNA into
catalytically active telomerase ribonucleoprotein that can act on telomeres is required for
apoptosis and cell cycle arrest in human cancer cells. Cancer Res 66, 5763-5771.

Goldstein, A.S., Huang, J., Guo, C., Garraway, I.P., and Witte, O.N. (2010). Identification of a
cell of origin for human prostate cancer. Science 329, 568-571.

Gomez, D., O'Donohue, M.F., Wenner, T., Douarre, C., Macadre, J., Koebel, P., Giraud-Panis,
M.J., Kaplan, H., Kolkes, A., Shin-ya, K., et al. (2006a). The G-quadruplex ligand telomestatin
inhibits POT1 binding to telomeric sequences in vitro and induces GFP-POT1 dissociation from
telomeres in human cells. Cancer Res 66, 6908-6912.

Gomez, D., Wenner, T., Brassart, B., Douarre, C., O'Donohue, M.F., El Khoury, V., Shin-Ya, K.,
Morjani, H., Trentesaux, C., and Riou, J.F. (2006b). Telomestatin-induced telomere uncapping is
modulated by POT1 through G-overhang extension in HT1080 human tumor cells. J Biol Chem
281, 38721-38729.

Goodell, M.A., Brose, K., Paradis, G., Conner, A.S., and Mulligan, R.C. (1996). Isolation and
functional properties of murine hematopoietic stem cells that are replicating in vivo. J Exp Med
183, 1797-1806.

Goodell, M.A., Rosenzweig, M., Kim, H., Marks, D.F., DeMaria, M., Paradis, G., Grupp, S.A.,
Sieff, C.A., Mulligan, R.C., and Johnson, R.P. (1997). Dye efflux studies suggest that
hematopoietic stem cells expressing low or undetectable levels of CD34 antigen exist in multiple
species. Nat Med 3, 1337-1345.

Gowan, S.M., Harrison, J.R., Patterson, L., Valenti, M., Read, M.A., Neidle, S., and Kelland, L.R.
(2002). A G-quadruplex-interactive potent small-molecule inhibitor of telomerase exhibiting in
vitro and in vivo antitumor activity. Mol Pharmacol 61, 1154-1162.

Greider, C.W., and Blackburn, E.H. (1989). A telomeric sequence in the RNA of Tetrahymena
telomerase required for telomere repeat synthesis. Nature 337, 331-337.

Greten, T.F., Forner, A., Korangy, F., N'Kontchou, G., Barget, N., Ayuso, C., Ormandy, L.A.,
Manns, M.P., Beaugrand, M., and Bruix, J. (2010). A phase II open label trial evaluating safety
and efficacy of a telomerase peptide vaccination in patients with advanced hepatocellular
carcinoma. BMC Cancer 10, 209.

136
Griffith, J.D., Comeau, L., Rosenfield, S., Stansel, R.M., Bianchi, A., Moss, H., and de Lange, T.
(1999). Mammalian telomeres end in a large duplex loop. Cell 97, 503-514.

Gu, G., Yuan, J., Wills, M., and Kasper, S. (2007). Prostate cancer cells with stem cell
characteristics reconstitute the original human tumor in vivo. Cancer Res 67, 4807-4815.

Guiducci, C., Cerone, M.A., and Bacchetti, S. (2001). Expression of mutant telomerase in
immortal telomerase-negative human cells results in cell cycle deregulation, nuclear and
chromosomal abnormalities and rapid loss of viability. Oncogene 20, 714-725.

Gupta, P.B., Chaffer, C.L., and Weinberg, R.A. (2009a). Cancer stem cells: mirage or reality? Nat
Med 15, 1010-1012.

Gupta, P.B., Onder, T.T., Jiang, G., Tao, K., Kuperwasser, C., Weinberg, R.A., and Lander, E.S.
(2009b). Identification of selective inhibitors of cancer stem cells by high-throughput screening.
Cell 138, 645-659.

Guzman-Ramirez, N., Voller, M., Wetterwald, A., Germann, M., Cross, N.A., Rentsch, C.A.,
Schalken, J., Thalmann, G.N., and Cecchini, M.G. (2009). In vitro propagation and
characterization of neoplastic stem/progenitor-like cells from human prostate cancer tissue.
Prostate 69, 1683-1693.

Hahn, W.C. (2002). Immortalization and transformation of human cells. Mol Cells 13, 351-361.

Hahn, W.C., Counter, C.M., Lundberg, A.S., Beijersbergen, R.L., Brooks, M.W., and Weinberg,
R.A. (1999a). Creation of human tumour cells with defined genetic elements. Nature 400,
464-468.

Hahn, W.C., Stewart, S.A., Brooks, M.W., York, S.G ., Eaton, E., Kurachi, A., Beijersbergen, R.L.,
Knoll, J.H., Meyerson, M., and Weinberg, R.A. (1999b). Inhibition of telomerase limits the
growth of human cancer cells. Nat Med 5, 1164-1170.

Hambardzumyan, D., Becher, O.J., Rosenblum, M.K., Pandolfi, P.P., Manova-Todorova, K., and
Holland, E.C. (2008). PI3K pathway regulates survival of cancer stem cells residing in the
perivascular niche following radiation in medulloblastoma in vivo. Genes Dev 22, 436-448.

Hao, J.L., Cozzi, P.J., Khatri, A., Power, C.A., and Li, Y. (2010). CD147/EMMPRIN and CD44
are potential therapeutic targets for metastatic prostate cancer. Curr Cancer Drug Targets 10,
287-306.

137
Harris, M.A., Yang, H., Low, B.E., Mukherjee, J., Guha, A., Bronson, R.T., Shultz, L.D., Israel,
M.A., and Yun, K. (2008). Cancer stem cells are enriched in the side population cells in a mouse
model of glioma. Cancer Res 68, 10051-10059.

Hastie, N.D., Dempster, M., Dunlop, M.G., Thompson, A.M., Green, D.K., and Allshire, R.C.
(1990). Telomere reduction in human colorectal carcinoma and with ageing. Nature 346, 866-868.
He, K., Xu, T., and Goldkorn, A. (2011). Cancer cells cyclically lose and regain drug-resistant
highly tumorigenic features characteristic of a cancer stem-like phenotype. Mol Cancer Ther 10,
938-948.

Heddleston, J.M., Li, Z., McLendon, R.E., Hjelmeland, A.B., and Rich, J.N. (2009). The hypoxic
microenvironment maintains glioblastoma stem cells and promotes reprogramming towards a
cancer stem cell phenotype. Cell Cycle 8, 3274-3284.

Herbert, B.S., Hochreiter, A.E., Wright, W.E., and Shay, J.W. (2006). Nonradioactive detection of
telomerase activity using the telomeric repeat amplification protocol. Nat Protoc 1, 1583-1590.

Herbert, B.S., Pongracz, K., Shay, J.W., and Gryaznov, S.M. (2002). Oligonucleotide N3'-->P5'
phosphoramidates as efficient telomerase inhibitors. Oncogene 21, 638-642.

Hirschmann-Jax, C., Foster, A.E., Wulf, G.G., Nuchtern, J.G., Jax, T.W., Gobel, U., Goodell, M.A.,
and Brenner, M.K. (2004). A distinct "side population" of cells with high drug efflux capacity in
human tumor cells. Proc Natl Acad Sci U S A 101, 14228-14233.

Hiyama, E., Saeki, T., Hiyama, K., Takashima, S., Shay, J.W., Matsuura, Y., and Yokoyama, T.
(2000). Telomerase activity as a marker of breast carcinoma in fine-needle aspirated samples.
Cancer 90, 235-238.

Ho, M.M., Ng, A.V., Lam, S., and Hung, J.Y. (2007). Side population in human lung cancer cell
lines and tumors is enriched with stem-like cancer cells. Cancer Res 67, 4827-4833.

Hope, K.J., Jin, L., and Dick, J.E. (2004). Acute myeloid leukemia originates from a hierarchy of
leukemic stem cell classes that differ in self-renewal capacity. Nat Immunol 5, 738-743.

Hu, L., McArthur, C., and Jaffe, R.B. (2010). Ovarian cancer stem-like side-population cells are
tumourigenic and chemoresistant. Br J Cancer 102, 1276-1283.

138
Hurt, E.M., Kawasaki, B.T., Klarmann, G.J., Thomas, S.B., and Farrar, W.L. (2008). CD44+
CD24(-) prostate cells are early cancer progenitor/stem cells that provide a model for patients
with poor prognosis. Br J Cancer 98, 756-765.

Irving, J., Wang, Z., Powell, S., O'Sullivan, C., Mok, M., Murphy, B., Cardoza, L., Lebkowski,
J.S., and Majumdar, A.S. (2004). Conditionally replicative adenovirus driven by the human
telomerase promoter provides broad-spectrum antitumor activity without liver toxicity. Cancer
Gene Ther 11, 174-185.

Joseph, I., Tressler, R., Bassett, E., Harley, C., Buseman, C.M., Pattamatta, P., Wright, W.E., Shay,
J.W., and Go, N.F. (2010). The telomerase inhibitor imetelstat depletes cancer stem cells in breast
and pancreatic cancer cell lines. Cancer Res 70, 9494-9504.

Keith, B., and Simon, M.C. (2007). Hypoxia-inducible factors, stem cells, and cancer. Cell 129,
465-472.

Kelland, L. (2007). The resurgence of platinum-based cancer chemotherapy. Nat Rev Cancer 7,
573-584.

Kelly, P.N., Dakic, A., Adams, J.M., Nutt, S.L., and Strasser, A. (2007). Tumor growth need not
be driven by rare cancer stem cells. Science 317, 337.

Keysar, S.B., and Jimeno, A. (2010). More than markers: biological significance of cancer stem
cell-defining molecules. Mol Cancer Ther 9, 2450-2457.

Kim, C.F., and Dirks, P.B. (2008). Cancer and stem cell biology: how tightly intertwined? Cell
Stem Cell 3, 147-150.

Kim, M.M., Rivera, M.A., Botchkina, I.L., Shalaby, R., Thor, A.D., and Blackburn, E.H. (2001).
A low threshold level of expression of mutant-template telomerase RNA inhibits human tumor
cell proliferation. Proc Natl Acad Sci U S A 98, 7982-7987.

Kirwan, M., and Dokal, I. (2008). Dyskeratosis congenita: a genetic disorder of many faces. Clin
Genet 73, 103-112.

Klarmann, G.J., Hurt, E.M., Mathews, L.A., Zhang, X., Duhagon, M.A., Mistree, T., Thomas,
S.B., and Farrar, W.L. (2009). Invasive prostate cancer cells are tumor initiating cells that have a
stem cell-like genomic signature. Clin Exp Metastasis 26, 433-446.

139
Knuefermann, C., Lu, Y., Liu, B., Jin, W., Liang, K., Wu, L., Schmidt, M., Mills, G.B.,
Mendelsohn, J., and Fan, Z. (2003). HER2/PI-3K/Akt activation leads to a multidrug resistance in
human breast adenocarcinoma cells. Oncogene 22, 3205-3212.

Kondo, T., Setoguchi, T., and Taga, T. (2004). Persistence of a small subpopulation of cancer
stem-like cells in the C6 glioma cell line. Proc Natl Acad Sci U S A 101, 781-786.

Korkaya, H., Paulson, A., Charafe-Jauffret, E., Ginestier, C., Brown, M., Dutcher, J., Clouthier,
S.G., and Wicha, M.S. (2009). Regulation of mammary stem/progenitor cells by
PTEN/Akt/beta-catenin signaling. PLoS Biol 7, e1000121.

Kumar, S.R., Scehnet, J.S., Ley, E.J., Singh, J., Krasnoperov, V., Liu, R., Manchanda, P.K.,
Ladner, R.D., Hawes, D., Weaver, F.A., et al. (2009). Preferential induction of EphB4 over
EphB2 and its implication in colorectal cancer progression. Cancer Res 69, 3736-3745.

Lapidot, T., Sirard, C., Vormoor, J., Murdoch, B., Hoang, T., Caceres-Cortes, J., Minden, M.,
Paterson, B., Caligiuri, M.A., and Dick, J.E. (1994). A cell initiating human acute myeloid
leukaemia after transplantation into SCID mice. Nature 367, 645-648.

Lee, H.W., Blasco, M.A., Gottlieb, G.J., Horner, J.W., 2nd, Greider, C.W., and DePinho, R.A.
(1998). Essential role of mouse telomerase in highly proliferative organs. Nature 392, 569-574.

Leonetti, C., Scarsella, M., Riggio, G., Rizzo, A., Salvati, E., D'Incalci, M., Staszewsky, L.,
Frapolli, R., Stevens, M.F., Stoppacciaro, A., et al. (2008). G-quadruplex ligand RHPS4
potentiates the antitumor activity of camptothecins in preclinical models of solid tumors. Clin
Cancer Res 14, 7284-7291.

Li, H., Chen, X., Calhoun-Davis, T., Claypool, K., and Tang, D.G. (2008a). PC3 human prostate
carcinoma cell holoclones contain self-renewing tumor-initiating cells. Cancer Res 68,
1820-1825.

Li, S., Rosenberg, J.E., Donjacour, A.A., Botchkina, I.L., Hom, Y .K., Cunha, G.R., and Blackburn,
E.H. (2004). Rapid inhibition of cancer cell growth induced by lentiviral delivery and expression
of mutant-template telomerase RNA and anti-telomerase short-interfering RNA. Cancer Res 64,
4833-4840.

Li, X., Lewis, M.T., Huang, J., Gutierrez, C., Osborne, C.K., Wu, M.F., Hilsenbeck, S.G., Pavlick,
A., Zhang, X., Chamness, G.C., et al. (2008b). Intrinsic resistance of tumorigenic breast cancer
cells to chemotherapy. J Natl Cancer Inst 100, 672-679.

140
Liang, Y., Zhong, Z., Huang, Y., Deng, W., Cao, J., Tsao, G., Liu, Q., Pei, D., Kang, T., and Zeng,
Y .X. (2010). Stem-like cancer cells are inducible by increasing genomic instability in cancer cells.
J Biol Chem 285, 4931-4940.

Liao, C.P., Adisetiyo, H., Liang, M., and Roy-Burman, P. (2010). Cancer-associated fibroblasts
enhance the gland-forming capability of prostate cancer stem cells. Cancer Res 70, 7294-7303.

Lin, T., Meng, L., Li, Y., and Tsai, R.Y. (2010). Tumor-initiating function of
nucleostemin-enriched mammary tumor cells. Cancer Res 70, 9444-9452.

Liu, P., Cheng, H., Roberts, T.M., and Zhao, J.J. (2009). Targeting the phosphoinositide 3-kinase
pathway in cancer. Nat Rev Drug Discov 8, 627-644.

M. Kozloff, G.W.S., F. M. Benedetti, A. Starr, J. A. Wallace, M. J. Stuart, D. Gruver,and K. Miller
(2010). Phase I study of imetelstat (GRN163L) in combination with paclitaxel (P) and
bevacizumab (B) in patients (pts) with locally recurrent or metastatic breast cancer (MBC). Paper
presented at: ASCO.

Ma, H., Nguyen, C., Lee, K.S., and Kahn, M. (2005). Differential roles for the coactivators CBP
and p300 on TCF/beta-catenin-mediated survivin gene expression. Oncogene 24, 3619-3631.

Maitland, N.J., and Collins, A.T. (2008). Prostate cancer stem cells: a new target for therapy. J
Clin Oncol 26, 2862-2870.

Mani, S.A., Guo, W., Liao, M.J., Eaton, E.N., Ayyanan, A., Zhou, A.Y ., Brooks, M., Reinhard, F.,
Zhang, C.C., Shipitsin, M., et al. (2008). The epithelial-mesenchymal transition generates cells
with properties of stem cells. Cell 133, 704-715.

Marchetti, A., Bertacca, G., Buttitta, F., Chella, A., Quattrocolo, G., Angeletti, C.A., and
Bevilacqua, G. (1999). Telomerase activity as a prognostic indicator in stage I non-small cell lung
cancer. Clin Cancer Res 5, 2077-2081.

Marian, C.O., Cho, S.K., McEllin, B.M., Maher, E.A., Hatanpaa, K.J., Madden, C.J., Mickey,
B.E., Wright, W.E., Shay, J.W., and Bachoo, R.M. (2010a). The telomerase antagonist, imetelstat,
efficiently targets glioblastoma tumor-initiating cells leading to decreased proliferation and tumor
growth. Clin Cancer Res 16, 154-163.

Marian, C.O., Wright, W.E., and Shay, J.W. (2010b). The effects of telomerase inhibition on
prostate tumor-initiating cells. Int J Cancer 127, 321-331.

141
Marie-Egyptienne, D.T., Brault, M.E., Nimmo, G.A., Londono-Vallejo, J.A., and Autexier, C.
(2009). Growth defects in mouse telomerase RNA-deficient cells expressing a template-mutated
mouse telomerase RNA. Cancer Lett 275, 266-276.

Meeker, A.K. (2006). Telomeres and telomerase in prostatic intraepithelial neoplasia and prostate
cancer biology. Urol Oncol 24, 122-130.

Mikami-Terao, Y., Akiyama, M., Yuza, Y., Yanagisawa, T., Yamada, O., and Yamada, H. (2008).
Antitumor activity of G-quadruplex-interactive agent TMPyP4 in K562 leukemic cells. Cancer
Lett 261, 226-234.

Miki, J., Furusato, B., Li, H., Gu, Y., Takahashi, H., Egawa, S., Sesterhenn, I.A., McLeod, D.G.,
Srivastava, S., and Rhim, J.S. (2007). Identification of putative stem cell markers, CD133 and
CXCR4, in hTERT-immortalized primary nonmalignant and malignant tumor-derived human
prostate epithelial cell lines and in prostate cancer specimens. Cancer Res 67, 3153-3161.

Mimeault, M., and Batra, S.K. (2009). Recent insights into the molecular mechanisms involved in
aging and the malignant transformation of adult stem/progenitor cells and their therapeutic
implications. Ageing Res Rev 8, 94-112.

Minev, B., Hipp, J., Firat, H., Schmidt, J.D., Langlade-Demoyen, P., and Zanetti, M. (2000).
Cytotoxic T cell immunity against telomerase reverse transcriptase in humans. Proc Natl Acad Sci
U S A 97, 4796-4801.

Mitchell, J.R., Wood, E., and Collins, K. (1999). A telomerase component is defective in the
human disease dyskeratosis congenita. Nature 402, 551-555.

Miyabayashi, T., Teo, J.L., Yamamoto, M., McMillan, M., Nguyen, C., and Kahn, M. (2007).
Wnt/beta-catenin/CBP signaling maintains long-term murine embryonic stem cell pluripotency.
Proc Natl Acad Sci U S A 104, 5668-5673.

Moon, R.T., Kohn, A.D., De Ferrari, G.V., and Kaykas, A. (2004). WNT and beta-catenin
signalling: diseases and therapies. Nat Rev Genet 5, 691-701.

Mulholland, D.J., Xin, L., Morim, A., Lawson, D., Witte, O., and Wu, H. (2009).
Lin-Sca-1+CD49fhigh stem/progenitors are tumor-initiating cells in the Pten-null prostate cancer
model. Cancer Res 69, 8555-8562.

142
Nemunaitis, J., Tong, A.W., Nemunaitis, M., Senzer, N., Phadke, A.P., Bedell, C., Adams, N.,
Zhang, Y.A., Maples, P.B., Chen, S., et al. (2010). A phase I study of telomerase-specific
replication competent oncolytic adenovirus (telomelysin) for various solid tumors. Mol Ther 18,
429-434.

Nguyen, D.X., Bos, P.D., and Massague, J. (2009). Metastasis: from dissemination to
organ-specific colonization. Nat Rev Cancer 9, 274-284.

O'Brien, C.A., Kreso, A., and Dick, J.E. (2009). Cancer stem cells in solid tumors: an overview.
Semin Radiat Oncol 19, 71-77.

O'Brien, C.A., Pollett, A., Gallinger, S., and Dick, J.E. (2007). A human colon cancer cell capable
of initiating tumour growth in immunodeficient mice. Nature 445, 106-110.

Oishi, T., Kigawa, J., Minagawa, Y., Shimada, M., Takahashi, M., and Terakawa, N. (1998).
Alteration of telomerase activity associated with development and extension of epithelial ovarian
cancer. Obstet Gynecol 91, 568-571.

Okayasu, I., Mitomi, H., Yamashita, K., Mikami, T., Fujiwara, M., Kato, M., and Oshimura, M.
(1998). Telomerase activity significantly correlates with cell differentiation, proliferation and
lymph node metastasis in colorectal carcinomas. J Cancer Res Clin Oncol 124, 444-449.

Olive, K.P., Jacobetz, M.A., Davidson, C.J., Gopinathan, A., McIntyre, D., Honess, D., Madhu,
B., Goldgraben, M.A., Caldwell, M.E., Allard, D., et al. (2009). Inhibition of Hedgehog signaling
enhances delivery of chemotherapy in a mouse model of pancreatic cancer. Science 324,
1457-1461.

Park, C.Y., Tseng, D., and Weissman, I.L. (2009). Cancer stem cell-directed therapies: recent data
from the laboratory and clinic. Mol Ther 17, 219-230.

Park, E., Gang, E.J., Hsieh, Y .T., Schaefer, P., Chae, S., Klemm, L., Huantes, S., Loh, M., Conway,
E.M., Kang, E.S., et al. (2011). Targeting survivin overcomes drug resistance in acute
lymphoblastic leukemia. Blood 118, 2191-2199.

Patrawala, L., Calhoun-Davis, T., Schneider-Broussard, R., and Tang, D.G. (2007). Hierarchical
organization of prostate cancer cells in xenograft tumors: the CD44+alpha2beta1+ cell population
is enriched in tumor-initiating cells. Cancer Res 67, 6796-6805.

143
Patrawala, L., Calhoun, T., Schneider-Broussard, R., Li, H., Bhatia, B., Tang, S., Reilly, J.G.,
Chandra, D., Zhou, J., Claypool, K., et al. (2006). Highly purified CD44+ prostate cancer cells
from xenograft human tumors are enriched in tumorigenic and metastatic progenitor cells.
Oncogene 25, 1696-1708.

Patrawala, L., Calhoun, T., Schneider-Broussard, R., Zhou, J., Claypool, K., and Tang, D.G.
(2005). Side population is enriched in tumorigenic, stem-like cancer cells, whereas ABCG2+ and
ABCG2- cancer cells are similarly tumorigenic. Cancer Res 65, 6207-6219.

Pirker, C., Holzmann, K., Spiegl-Kreinecker, S., Elbling, L., Thallinger, C., Pehamberger, H.,
Micksche, M., and Berger, W. (2003). Chromosomal imbalances in primary and metastatic
melanomas: over-representation of essential telomerase genes. Melanoma Res 13, 483-492.

Polyak, K., and Hahn, W.C. (2006). Roots and stems: stem cells in cancer. Nat Med 12, 296-300.

Poremba, C., Heine, B., Diallo, R., Heinecke, A., Wai, D., Schaefer, K.L., Braun, Y., Schuck, A.,
Lanvers, C., Bankfalvi, A., et al. (2002). Telomerase as a prognostic marker in breast cancer:
high-throughput tissue microarray analysis of hTERT and hTR. J Pathol 198, 181-189.

Poremba, C., Willenbring, H., Hero, B., Christiansen, H., Schafer, K.L., Brinkschmidt, C.,
Jurgens, H., Bocker, W., and Dockhorn-Dworniczak, B. (1999). Telomerase activity distinguishes
between neuroblastomas with good and poor prognosis. Ann Oncol 10, 715-721.

Puri, N., Eller, M.S., Byers, H.R., Dykstra, S., Kubera, J., and Gilchrest, B.A. (2004).
Telomere-based DNA damage responses: a new approach to melanoma. FASEB J 18, 1373-1381.

Quintana, E., Shackleton, M., Sabel, M.S., Fullen, D.R., Johnson, T.M., and Morrison, S.J. (2008).
Efficient tumour formation by single human melanoma cells. Nature 456, 593-598.

Rada-Iglesias, A., Bajpai, R., Swigut, T., Brugmann, S.A., Flynn, R.A., and Wysocka, J. (2011). A
unique chromatin signature uncovers early developmental enhancers in humans. Nature 470,
279-283.

Reya, T., Morrison, S.J., Clarke, M.F., and Weissman, I.L. (2001). Stem cells, cancer, and cancer
stem cells. Nature 414, 105-111.

Roesch, A., Fukunaga-Kalabis, M., Schmidt, E.C., Zabierowski, S.E., Brafford, P.A., Vultur, A.,
Basu, D., Gimotty, P., Vogt, T., and Herlyn, M. (2010). A temporarily distinct subpopulation of
slow-cycling melanoma cells is required for continuous tumor growth. Cell 141, 583-594.

144
S. Aamdal, S.D., O. Engebraaten, K. Owre, M. Dyrhaug, S. Trachsel, G. Gaudernack (2006). A
phase I/II study of telomerase peptide vaccination in combination with chemotherapy in patients
with stage IV malignant melanoma. Paper presented at: ASCO.

Salcido, C.D., Larochelle, A., Taylor, B.J., Dunbar, C.E., and Varticovski, L. (2010). Molecular
characterisation of side population cells with cancer stem cell-like characteristics in small-cell
lung cancer. Br J Cancer 102, 1636-1644.

Salvati, E., Leonetti, C., Rizzo, A., Scarsella, M., Mottolese, M., Galati, R., Sperduti, I., Stevens,
M.F., D'Incalci, M., Blasco, M., et al. (2007). Telomere damage induced by the G-quadruplex
ligand RHPS4 has an antitumor effect. J Clin Invest 117, 3236-3247.

Saretzki, G., Sitte, N., Merkel, U., Wurm, R.E., and von Zglinicki, T. (1999). Telomere shortening
triggers a p53-dependent cell cycle arrest via accumulation of G-rich single stranded DNA
fragments. Oncogene 18, 5148-5158.

Sarin, K.Y., Cheung, P., Gilison, D., Lee, E., Tennen, R.I., Wang, E., Artandi, M.K., Oro, A.E.,
and Artandi, S.E. (2005). Conditional telomerase induction causes proliferation of hair follicle
stem cells. Nature 436, 1048-1052.

Sato, N., Meijer, L., Skaltsounis, L., Greengard, P., and Brivanlou, A.H. (2004). Maintenance of
pluripotency in human and mouse embryonic stem cells through activation of Wnt signaling by a
pharmacological GSK-3-specific inhibitor. Nat Med 10, 55-63.

Scheel, C., and Weinberg, R.A. (2011). Phenotypic plasticity and epithelial-mesenchymal
transitions in cancer and normal stem cells? Int J Cancer 129, 2310-2314.

Shammas, M.A., Koley, H., Bertheau, R.C., Neri, P., Fulciniti, M., Tassone, P., Blotta, S.,
Protopopov, A., Mitsiades, C., Batchu, R.B., et al. (2008). Telomerase inhibitor GRN163L
inhibits myeloma cell growth in vitro and in vivo. Leukemia 22, 1410-1418.

Sharma, S.V., Lee, D.Y., Li, B., Quinlan, M.P., Takahashi, F., Maheswaran, S., McDermott, U.,
Azizian, N., Zou, L., Fischbach, M.A., et al. (2010). A chromatin-mediated reversible
drug-tolerant state in cancer cell subpopulations. Cell 141, 69-80.

Shay, J.W., and Bacchetti, S. (1997). A survey of telomerase activity in human cancer. Eur J
Cancer 33, 787-791.

145
Shay, J.W., and Wright, W.E. (2006). Telomerase therapeutics for cancer: challenges and new
directions. Nat Rev Drug Discov 5, 577-584.

Shmelkov, S.V., Butler, J.M., Hooper, A.T., Hormigo, A., Kushner, J., Milde, T., St Clair, R.,
Baljevic, M., White, I., Jin, D.K., et al. (2008). CD133 expression is not restricted to stem cells,
and both CD133+ and CD133- metastatic colon cancer cells initiate tumors. J Clin Invest 118,
2111-2120.

Shoeneman, J.K., Ehrhart, E.J., 3rd, Eickhoff, J.C., Charles, J.B., Powers, B.E., and Thamm, D.H.
(2012). Expression and function of survivin in canine osteosarcoma. Cancer Res 72, 249-259.

Siddiqui-Jain, A., Grand, C.L., Bearss, D.J., and Hurley, L.H. (2002). Direct evidence for a
G-quadruplex in a promoter region and its targeting with a small molecule to repress c-MYC
transcription. Proc Natl Acad Sci U S A 99, 11593-11598.

Singh, A., Wu, H., Zhang, P., Happel, C., Ma, J., and Biswal, S. (2010). Expression of ABCG2
(BCRP) is regulated by Nrf2 in cancer cells that confers side population and chemoresistance
phenotype. Mol Cancer Ther 9, 2365-2376.

Singh, S.K., Hawkins, C., Clarke, I.D., Squire, J.A., Bayani, J., Hide, T., Henkelman, R.M.,
Cusimano, M.D., and Dirks, P.B. (2004). Identification of human brain tumour initiating cells.
Nature 432, 396-401.

Smith, F.W., and Feigon, J. (1992). Quadruplex structure of Oxytricha telomeric DNA
oligonucleotides. Nature 356, 164-168.

Smogorzewska, A., and de Lange, T. (2004). Regulation of telomerase by telomeric proteins.
Annu Rev Biochem 73, 177-208.

Song, S., Mazurek, N., Liu, C., Sun, Y., Ding, Q.Q., Liu, K., Hung, M.C., and Bresalier, R.S.
(2009). Galectin-3 mediates nuclear beta-catenin accumulation and Wnt signaling in human colon
cancer cells by regulation of glycogen synthase kinase-3beta activity. Cancer Res 69, 1343-1349.

Steiniger, S.C., Coppinger, J.A., Kruger, J.A., Yates, J., 3rd, and Janda, K.D. (2008). Quantitative
mass spectrometry identifies drug targets in cancer stem cell-containing side population. Stem
Cells 26, 3037-3046.

146
Stewart, S.A., Hahn, W.C., O'Connor, B.F., Banner, E.N., Lundberg, A.S., Modha, P., Mizuno, H.,
Brooks, M.W., Fleming, M., Zimonjic, D.B., et al. (2002). Telomerase contributes to
tumorigenesis by a telomere length-independent mechanism. Proc Natl Acad Sci U S A 99,
12606-12611.

Stohr, B.A., and Blackburn, E.H. (2008). ATM mediates cytotoxicity of a mutant telomerase RNA
in human cancer cells. Cancer Res 68, 5309-5317.

Strahl, C., and Blackburn, E.H. (1996). Effects of reverse transcriptase inhibitors on telomere
length and telomerase activity in two immortalized human cell lines. Mol Cell Biol 16, 53-65.

Streutker, C.J., Thorner, P., Fabricius, N., Weitzman, S., and Zielenska, M. (2001). Telomerase
activity as a prognostic factor in neuroblastomas. Pediatr Dev Pathol 4, 62-67.

Stuelten, C.H., Mertins, S.D., Busch, J.I., Gowens, M., Scudiero, D.A., Burkett, M.W., Hite, K.M.,
Alley, M., Hollingshead, M., Shoemaker, R.H., et al. (2010). Complex display of putative tumor
stem cell markers in the NCI60 tumor cell line panel. Stem Cells 28, 649-660.

Su, Z., Dannull, J., Yang, B.K., Dahm, P., Coleman, D., Yancey, D., Sichi, S., Niedzwiecki, D.,
Boczkowski, D., Gilboa, E., et al. (2005). Telomerase mRNA-transfected dendritic cells stimulate
antigen-specific CD8+ and CD4+ T cell responses in patients with metastatic prostate cancer. J
Immunol 174, 3798-3807.

Sun, D., Guo, K., Rusche, J.J., and Hurley, L.H. (2005). Facilitation of a structural transition in
the polypurine/polypyrimidine tract within the proximal promoter region of the human VEGF
gene by the presence of potassium and G-quadruplex-interactive agents. Nucleic Acids Res 33,
6070-6080.

T. Buanes, J.M., W. Liauw, M. Hebbar, and J. Nemunaitis (2009). A randomized phase III study
of gemcitabine (G) versus GV1001 in sequential combination with G in patients with
unresectable and metastatic pancreatic cancer (PC) Paper presented at: ASCO.

Tahara, H., Shin-Ya, K., Seimiya, H., Yamada, H., Tsuruo, T., and Ide, T. (2006). G-Quadruplex
stabilization by telomestatin induces TRF2 protein dissociation from telomeres and anaphase
bridge formation accompanied by loss of the 3' telomeric overhang in cancer cells. Oncogene 25,
1955-1966.

Takahashi-Yanaga, F., and Kahn, M. (2010). Targeting Wnt signaling: can we safely eradicate
cancer stem cells? Clin Cancer Res 16, 3153-3162.

147
Takahashi, K., Tanabe, K., Ohnuki, M., Narita, M., Ichisaka, T., Tomoda, K., and Yamanaka, S.
(2007). Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell
131, 861-872.

Tamakawa, R.A., Fleisig, H.B., and Wong, J.M. (2010). Telomerase inhibition potentiates the
effects of genotoxic agents in breast and colorectal cancer cells in a cell cycle-specific manner.
Cancer Res 70, 8684-8694.

Tomoda, R., Seto, M., Tsumuki, H., Iida, K., Yamazaki, T., Sonoda, J., Matsumine, A., and
Uchida, A. (2002). Telomerase activity and human telomerase reverse transcriptase mRNA
expression are correlated with clinical aggressiveness in soft tissue tumors. Cancer 95,
1127-1133.

Tsolou, A., Passos, J.F., Nelson, G., Arai, Y., and Zglinicki, T. (2008). ssDNA fragments induce
cell senescence by telomere uncapping. Exp Gerontol 43, 892-899.

Tsuchida, R., Das, B., Yeger, H., Koren, G., Shibuya, M., Thorner, P.S., Baruchel, S., and Malkin,
D. (2008). Cisplatin treatment increases survival and expansion of a highly tumorigenic
side-population fraction by upregulating VEGF/Flt1 autocrine signaling. Oncogene 27,
3923-3934.

Vander Griend, D.J., Karthaus, W.L., Dalrymple, S., Meeker, A., DeMarzo, A.M., and Isaacs, J.T.
(2008). The role of CD133 in normal human prostate stem cells and malignant cancer-initiating
cells. Cancer Res 68, 9703-9711.

Verdun, R.E., and Karlseder, J. (2007). Replication and protection of telomeres. Nature 447,
924-931.

Vermeulen, L., De Sousa, E.M.F., van der Heijden, M., Cameron, K., de Jong, J.H., Borovski, T.,
Tuynman, J.B., Todaro, M., Merz, C., Rodermond, H., et al. (2010). Wnt activity defines colon
cancer stem cells and is regulated by the microenvironment. Nat Cell Biol 12, 468-476.

Vermeulen, L., Todaro, M., de Sousa Mello, F., Sprick, M.R., Kemper, K., Perez Alea, M., Richel,
D.J., Stassi, G., and Medema, J.P. (2008). Single-cell cloning of colon cancer stem cells reveals a
multi-lineage differentiation capacity. Proc Natl Acad Sci U S A 105, 13427-13432.

Vieyra, D.S., Rosen, A., and Goodell, M.A. (2009). Identification and characterization of side
population cells in embryonic stem cell cultures. Stem Cells Dev 18, 1155-1166.

148
Visvader, J.E., and Lindeman, G.J. (2008). Cancer stem cells in solid tumours: accumulating
evidence and unresolved questions. Nat Rev Cancer 8, 755-768.

Vonderheide, R.H., Domchek, S.M., Schultze, J.L., George, D.J., Hoar, K.M., Chen, D.Y.,
Stephans, K.F., Masutomi, K., Loda, M., Xia, Z., et al. (2004). Vaccination of cancer patients
against telomerase induces functional antitumor CD8+ T lymphocytes. Clin Cancer Res 10,
828-839.

V onderheide, R.H., Hahn, W.C., Schultze, J.L., and Nadler, L.M. (1999). The telomerase catalytic
subunit is a widely expressed tumor-associated antigen recognized by cytotoxic T lymphocytes.
Immunity 10, 673-679.

Wagner, W., Ansorge, A., Wirkner, U., Eckstein, V ., Schwager, C., Blake, J., Miesala, K., Selig, J.,
Saffrich, R., Ansorge, W., et al. (2004). Molecular evidence for stem cell function of the
slow-dividing fraction among human hematopoietic progenitor cells by genome-wide analysis.
Blood 104, 675-686.

Waki, K., Anno, K., Ono, T., Ide, T., Chayama, K., and Tahara, H. (2010). Establishment of
functional telomerase immortalized human hepatocytes and a hepatic stellate cell line for
telomere-targeting anticancer drug development. Cancer Sci 101, 1678-1685.

Wang, J., Weaver, I.C., Gauthier-Fisher, A., Wang, H., He, L., Yeomans, J., Wondisford, F.,
Kaplan, D.R., and Miller, F.D. (2010). CBP histone acetyltransferase activity regulates embryonic
neural differentiation in the normal and Rubinstein-Taybi syndrome brain. Dev Cell 18, 114-125.

Wang, X., Kruithof-de Julio, M., Economides, K.D., Walker, D., Yu, H., Halili, M.V., Hu, Y.P.,
Price, S.M., Abate-Shen, C., and Shen, M.M. (2009). A luminal epithelial stem cell that is a cell
of origin for prostate cancer. Nature 461, 495-500.

Watson, J.D. (1972). Origin of concatemeric T7 DNA. Nat New Biol 239, 197-201.

Weisenberger, D.J., Campan, M., Long, T.I., Kim, M., Woods, C., Fiala, E., Ehrlich, M., and
Laird, P.W. (2005). Analysis of repetitive element DNA methylation by MethyLight. Nucleic
Acids Res 33, 6823-6836.

Weisenberger, D.J., Siegmund, K.D., Campan, M., Young, J., Long, T.I., Faasse, M.A., Kang,
G.H., Widschwendter, M., Weener, D., Buchanan, D., et al. (2006). CpG island methylator
phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF
mutation in colorectal cancer. Nat Genet 38, 787-793.

149
Woodward, W.A., Chen, M.S., Behbod, F., Alfaro, M.P., Buchholz, T.A., and Rosen, J.M. (2007).
WNT/beta-catenin mediates radiation resistance of mouse mammary progenitor cells. Proc Natl
Acad Sci U S A 104, 618-623.

Wright, W.E., Pereira-Smith, O.M., and Shay, J.W. (1989). Reversible cellular senescence:
implications for immortalization of normal human diploid fibroblasts. Mol Cell Biol 9,
3088-3092.

Wright, W.E., Piatyszek, M.A., Rainey, W.E., Byrd, W., and Shay, J.W. (1996). Telomerase
activity in human germline and embryonic tissues and cells. Dev Genet 18, 173-179.

Wu, C., Wei, Q., Utomo, V., Nadesan, P., Whetstone, H., Kandel, R., Wunder, J.S., and Alman,
B.A. (2007). Side population cells isolated from mesenchymal neoplasms have tumor initiating
potential. Cancer Res 67, 8216-8222.

Xu, T., Lu, B., Tai, Y.C., and Goldkorn, A. (2010a). A cancer detection platform which measures
telomerase activity from live circulating tumor cells captured on a microfilter. Cancer Res 70,
6420-6426.

Xu, T., Xu, Y ., Liao, C.P., Lau, R., and Goldkorn, A. (2010b). Reprogramming murine telomerase
rapidly inhibits the growth of mouse cancer cells in vitro and in vivo. Mol Cancer Ther 9,
438-449.

Yaar, M., Eller, M.S., Panova, I., Kubera, J., Wee, L.H., Cowan, K.H., and Gilchrest, B.A. (2007).
Telomeric DNA induces apoptosis and senescence of human breast carcinoma cells. Breast
Cancer Res 9, R13.

Yi, X., Shay, J.W., and Wright, W.E. (2001). Quantitation of telomerase components and hTERT
mRNA splicing patterns in immortal human cells. Nucleic Acids Res 29, 4818-4825.

Yoshida, R., Kiyozuka, Y., Ichiyoshi, H., Senzaki, H., Takada, H., Hioki, K., and Tsubura, A.
(1999). Change in telomerase activity during human colorectal carcinogenesis. Anticancer Res 19,
2167-2172.

Yu, F., Yao, H., Zhu, P., Zhang, X., Pan, Q., Gong, C., Huang, Y., Hu, X., Su, F., Lieberman, J., et
al. (2007a). let-7 regulates self renewal and tumorigenicity of breast cancer cells. Cell 131,
1109-1123.

150
Yu, G.L., Bradley, J.D., Attardi, L.D., and Blackburn, E.H. (1990). In vivo alteration of telomere
sequences and senescence caused by mutated Tetrahymena telomerase RNAs. Nature 344,
126-132.

Yu, J., Vodyanik, M.A., Smuga-Otto, K., Antosiewicz-Bourget, J., Frane, J.L., Tian, S., Nie, J.,
Jonsdottir, G.A., Ruotti, V., Stewart, R., et al. (2007b). Induced pluripotent stem cell lines derived
from human somatic cells. Science 318, 1917-1920.

Zhang, M., Atkinson, R.L., and Rosen, J.M. (2010). Selective targeting of radiation-resistant
tumor-initiating cells. Proc Natl Acad Sci U S A 107, 3522-3527.

Zhang, X., Komaki, R., Wang, L., Fang, B., and Chang, J.Y. (2008). Treatment of radioresistant
stem-like esophageal cancer cells by an apoptotic gene-armed, telomerase-specific oncolytic
adenovirus. Clin Cancer Res 14, 2813-2823.

Zhang, X., Mar, V., Zhou, W., Harrington, L., and Robinson, M.O. (1999). Telomere shortening
and apoptosis in telomerase-inhibited human tumor cells. Genes Dev 13, 2388-2399.

Zhou, J., Wulfkuhle, J., Zhang, H., Gu, P., Yang, Y., Deng, J., Margolick, J.B., Liotta, L.A.,
Petricoin, E., 3rd, and Zhang, Y . (2007). Activation of the PTEN/mTOR/STAT3 pathway in breast
cancer stem-like cells is required for viability and maintenance. Proc Natl Acad Sci U S A 104,
16158-16163.

Zhou, S., Schuetz, J.D., Bunting, K.D., Colapietro, A.M., Sampath, J., Morris, J.J., Lagutina, I.,
Grosveld, G.C., Osawa, M., Nakauchi, H., et al. (2001). The ABC transporter Bcrp1/ABCG2 is
expressed in a wide variety of stem cells and is a molecular determinant of the side-population
phenotype. Nat Med 7, 1028-1034.

Abstract (if available)

Abstract Cancer claims over 500,000 lives in the U.S. annually, a mortality rate that is largely attributable to solid tumors that have metastasized and become resistant to available treatments. This type of disease progression may be mediated by cancer stem cells (CSC), rare and unique cancer cells recently identified in many types of malignancies, and these CSC are thought to play a central role in tumor formation, therapy-resistance, and ultimately cancer progression and metastasis. ❧ Unfortunately, however, to date there are very few effective CSC-targeting therapies. To address this question, we isolated a putative CSC population from human prostate tumors and cell lines, and we showed for the first time that these cells possess extremely high telomerase activity relative to the bulk, unselected cancer cells. Strikingly, telomerase interference – reprogramming telomerase to add incorrect “toxic” telomeres – could induce rapid apoptosis and marked growth inhibition in prostate CSC and abrogated their ability to form new tumors in SCID mice, which offers the first tumor-derived and in vivo evidence that telomerase may ultimately form the basis for more effective new CSC-targeting therapies. ❧ The drug resistant and brisk tumor initiation abilities have been viewed as pre-existing phenotypes only present in the small subpopulation of cancer stem cells, and they have been intuitively conceptualized as self-renewing founder cells from which more differentiated cancer cells derive. However, recent work in cancer cell lines has demonstrated that drug-resistant tumor initiating features can emerge de novo within fractionated subpopulations of cells initially lacking these phenotypes. In our study, we used a “side population” cancer stem cell model with GFP-labeling technique, and demonstrated for the first time that adaptive, cancer-promoting traits like drug-resistance and brisk tumor initiation arise not only as solitary events under selective pressures, but also as highly orchestrated transitions occurring concurrently in large numbers of cells even without specifically-induced drug selection, ectopic gene expression, or fractionation into subpopulations. In addition, our follow-up mechanistic studies have identified the PI3K/Akt/b-catenin/CBP pathway to play a critical role in regulating this dynamic equilibrium and regeneration of CSC. We hope our findings would contribute to a better understanding of CSC, and can potentially offer new strategies for targeting these drug-resistant, tumor-forming cells, ultimately leading to more effective treatments for patients.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Exploration of the roles of cancer stem cells and survivin in the pathogenesis and progression of prostate cancer

PDF

Identification and characterization of adult stem cells in the oral cavity

PDF

Investigating the effects of polycomb repressive complex inhibitors on cancer cell phenotypic plasticity

PDF

Studies of murine prostate cancer stem / progenitor cells

PDF

Role of cancer-associated fibroblast secreted annexin A1 in generation and maintenance of prostate cancer stem cells

PDF

Identification of CBP/FOXM1 as a molecular target in triple negative breast cancer

PDF

Targeting glioma cancer stem cells for the treatment of glioblastoma multiforme

PDF

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

PDF

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

PDF

The mechanism by which extracellular Hsp90α promotes cell migration: implications in wound healing and cancer

PDF

RNA methylation in cancer plasticity and drug resistance

PDF

Characterization of a fragment in secreted Hsp90α with potential therapeutic benefits in wound healing and cancer

PDF

The effect of tumor-mediated immune suppression on prostate cancer immunotherapy

PDF

The calcium-sensing receptor in the specification of normal and malignant hematopoietic cell localization in the bone marrow microenvironment

PDF

Role of the bone marrow niche components in B cell malignancies

PDF

Cisplatin activates mitochondrial oxphos leading to acute treatment resistance in bladder cancer

PDF

Molecular mechanisms of chemoresistance in breast cancer

PDF

Co-expression of monoamine oxidase A and prostate cancer stem cell markers in Pten knockout mice

PDF

Ancestral inference and cancer stem cell dynamics in colorectal tumors

PDF

The role of LGR5 in the pathogenesis of Ewing sarcoma: a marker of aggressive disease and a contributor to the malignant phenotype

Asset Metadata

Creator He, Kaijie (author)

Core Title The cancer stem-like phenotype: therapeutics, phenotypic plasticity and mechanistic studies

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 07/24/2012

Defense Date 05/17/2012

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

Tag cancer stem cell,OAI-PMH Harvest,plasticity,side population,signaling pathway,telomerase

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Pinski, Jacek K. (committee chair), Adams, Gregor B. (committee member), Goldkorn, Amir (committee member), Kahn, Michael (committee member), Kobielak, Agnieszka (committee member)

Creator Email hekaijie@gmail.com,khe@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-61631

Unique identifier UC11290452

Identifier usctheses-c3-61631 (legacy record id)

Legacy Identifier etd-HeKaijie-973.pdf

Dmrecord 61631

Document Type Dissertation

Rights He, Kaijie

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