University of Southern California Dissertations and Theses

Identification of novel epigenetic biomarkers and microRNAs for cancer therapeutics

(USC Thesis Other)

Identification of novel epigenetic biomarkers and microRNAs for cancer therapeutics

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content IDENTIFICATION OF NOVEL EPIGENETIC BIOMARKERS AND
MICRORNAS FOR CANCER THERAPEUTICS

by

Sheng-Fang Su

_______________________________________________

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

May 2013

Copyright 2013 Sheng-Fang Su

ii

EPIGRAPH
“ If A is a success in life, then A equals x plus y plus z. Work is x; y is play; and
z is keeping your mouth shut.”

-- Albert Einstein

iii

DEDICATION
I dedicate this work to both of my grandfathers and grandmothers who inspired me
to pursue a career in science, to my parents who always give me the biggest support
in everything I do and to my husband who is always by my side to encourage and
accompany me through all of the sadness and happiness.

iv

ACKNOWLEDGEMENTS
I would like to thank the following people who made the work in this thesis possible:
Dr. Peter A. Jones, my mentor, for his guidance and constant support. Your
passion for science and your “principles” have always inspired me to stay in the
scientific world and to enjoy science. You set an example of how to present our
works in a simple but clear and logic way. And your word: “one experiment, six
controls” have taught me to think thoroughly during every step of doing research.
Thank you for providing me an amazing working environment to learn and to do
science.
Dr. Gangning Liang, for his mentorship in all of my projects. It has been my
pleasure to work with you for the past four years. Your patience in handling
scientific works and helping everyone in the lab has truly been an inspiration. I have
learnt a lot from you and would like to thank you for all of your support which made
my research much better and deliberate.
Dr. Xiangning Qiu, for always being there when I needed her help during my first
year in the lab. Thank you for showing me the experimental procedure, supporting
me in any ways you can and being my friend. I really enjoyed the experience of
working with you.
Dr. Yvonne Tsai, for her hard work of collecting samples and processing urine
DNA extraction for my project, and for her well-organized lab management which
made doing researches in the lab much easier. Thank you for your assistance and
support in all the ways you provided.

v

My fellow graduate students Dr. Shikhar Sharma, Fides Lay, Yin-Wei Chang,
Chris Duymich, Ranjani Lakshminarasimhan, Kurinji Pandiyan, Ravi Agarwal and
Alexandra Soegaard, who were so much more than colleagues. Postdocs Dr. Terry
Kelly, Dr. Claudia Andreu-Vieyra, Dr. Jueng Soo You, Dr. Daniel De Carvalho, Dr.
Phillippa Oakford, Dr. Jessica Charlet, Dr. Kamilla Mundbjerg, Dr. Minmin Liu, Dr.
Xiaojing Yang and Dr. Elinne Becket, for all of their help. My committee members,
Dr. Gerhard Coetzee, Dr. Darryl Shibata, Dr. Michael Stallcup and Dr. Ite A. Laird-
Offringa, for all of the useful suggestion and feedback. My scientific collaborators,
Dr. Amy S. Lee, Dr. Bo Han, Dr. Kyle Pfaffenbach, Dr. Hyang-Min Byun and
Josephine Fang, my biostatistician collaborator, Dr. Kimberly
D Siegmund, my
clinical collaborators, Dr. André Luís De castro Abreu, Dr. Siamak Daneshmand and
Dr. Eila C. Skinner, my pathology collaborator, Moli Chen, for all of their support
and assistance.
My friends, Hui Shen, Suhn Kyong Rhie, and Wan-Ting Chen, for their friendship
and encouragement throughout this process.

vi

TABLE OF CONTENTS

Epigraph ii

Dedication iii

Acknowledgements iv

List of Tables ix

List of Figures x

Abstract xv

Chapter 1: Cancer epigenetics and its translational implications 1
Introduction 1
Overview of epigenetics in cancer 2
DNA methylation 3
Histone modifications 4
Nucleosome occupancy 6
MicroRNAs 7
Translational application of epigenetic strategies for cancer 9
Development of DNA methylation markers 10
Development of therapeutic microRNAs 13
Conclusion 15
Overview of thesis research 16

Chapter 2: A panel of three markers hyper and hypomethylated in
urine sediments accurately predicts bladder cancer
recurrence

18
Introduction 18
Materials and Methods 23
Results 27
DNA methylation analysis in urine sediments 27
Longitudinal study of DNA methylation changes
in urine sediments collected from TURBT patients at
the time of follow-up visits

29
A three-marker panel 36
Power of prediction of recurrence 37
Discussion 47
Conclusion 51

vii

Chapter 3: Functional organization of TJP2 (ZO2) at alterative
regulatory regions

52
Introduction 52
Materials and Methods 55
Results 59
Both CpG island promoters of TJP2 showed tissue-
specific expression patterns

59
Expression of both transcripts of TJP2 was down-
regulated in tumor samples from cancer patients

60
DNA methylation levels of TJP2 at P2 promoters
negatively correlated with mRNA expression in
bladder cancer cell lines

60
The TJP2 P2 at distal regulatory regions (R1) showed
nucleosome occupancy in T24 as well as in LD419
and UROtsa cells

62
Discussion 74

Chapter 4: MiR-30d, miR-181a and miR-199a-5p cooperatively
suppress the endoplasmic reticulum chaperone and
signaling regulator GRP78 in cancer

77
Introduction 77
Materials and Methods 81
Results 86
Inverse relationship between the expression of
miR-30d, miR-181a and miR-199a-5p and their
putative target GRP78

86
MiR-30d, miR-181a and miR-199a-5p directly target
the 3’UTR of GRP78 and significantly suppress
luciferase activity cooperatively

92
GRP78 down-regulation requires cooperation of
multiple miRNAs and leads to morphological
changes and apoptosis in C42B cells

95
MiR-30d, miR-181a, miR-199a-5p increase the
sensitivity of cancer cells to the HDAC inhibitor TSA

101
Lentiviral delivery of multiple co-transcribed
miRNAs decreases GRP78 protein levels and cell
viability and induces apoptosis in different cancer
cell lines

105
MiR-30d, miR-181a and miR-199a-5p inhibit tumor
growth in vivo

116
Discussion 118
Conclusion 120

viii

Chapter 5: Summary and Conclusions 121

References 126

ix

LIST OF TABLES

Table 2.1: Primer sequences used in pyrosequencing. 26

Table 2.2: The clinicopathological characteristics of 90 TURBT
patients.
31

Table 2.3: DNA methylation status in urine sediment samples of
TURBT patients at the time of first recurrence and the last
follow-up visit of no recurrence patients.
34

Table 2.4: The clinical characteristics of 34 recurrence bladder cancer
patients.
39

Table 2.5: Association between the score from the three-marker
signature and the bladder tumor pathological
characteristics of 90 TURBT patients.
40

Table 3.1: Primers sequence 58

Table 4.1: Primers sequence 85

Table 4.2: Ambion pre-miR
TM
miRNA ID 85

x

LIST OF FIGURES

Figure 2.1: DNA methylation alterations can be detected in urine
sediment samples from patients with bladder cancer.
28

Figure 2.2: Timeline of longitudinally collected urine sediment
samples in follow-up visits from bladder cancer patients
after tumor resection.
32

Figure 2.3: DNA methylation of HOXA9, SOX1, NPY, IRAK3, ZO2
(hypermethylated) and L1-MET (hypomethylated) in 368
urine sediment samples showed high sensitivity and
specificity in the detection of tumor recurrence.
33

Figure 2.4: A panel of six DNA methylation markers tested in urine
sediments from TURBT patients was positively correlated
with bladder tumor recurrence.
35

Figure 2.5: A three-marker signature showed high sensitivity and
specificity in detecting tumor recurrence.
38

Figure 2.6: The risk scores given by the combination of three DNA
methylation markers in the continuously monitored urine
showed distinct patterns in recurrence and no recurrence
patients.
42

Figure 2.7: Percentage of urine sediments that had positive scores
(DNA methylation score calculated to be higher than cut-
off values) at the time of recurrence (N=38) and the
comparison with cytology and cystoscopy reports at the
same visits to the clinic.
43

Figure 2.8: Three DNA methylation markers help predict the risk of
recurrence of bladder tumors in urine sediments.
45

xi

Figure 2.9: Three DNA methylation markers could detect the risk of
recurrence of bladder tumors in urine sediments.
46

Figure 3.1: Both TJP2-P1 and TJP2-P2 displayed tissue-specific
mRNA expression patterens.

65

Figure 3.2: Two alternative promoters of TJP2 (P1 and P2) showed
decreased mRNA expression in tumor tissues from
bladder, colon and prostate cancer patients.

66

Figure 3.3: The graphic map of CpG sites in two alternative promoters
of TJP2.

67

Figure 3.4: The relationship between DNA methylation levels and
mRNA expression of TJP2 at the two alternative
promoters (P1 and P2) in different cell lines.

68

Figure 3.5: Functional chromatin activity of the TJP2 gene at the
upstream promoter (P1).

69

Figure 3.6: Functional chromatin activity of the TJP2 gene at the
downstream promoter (P2).

70

Figure 3.7: mRNA expression of TJP2 P2 inversely correlated with
DNA methylation in LD419, UROtsa, and T24 cells.

71

Figure 3.8: Chromatin structure of GRP78 at promoter regions. 72

Figure 3.9: Chromatin structure of TJP2 at downstream promoter
region (P2).

73

Figure 4.1: Relative expression of miR-30d, miR-181a, and miR-199a-
5p, and their potential target, GRP78, in human cell lines.

87

xii

Figure 4.2: Endogenous expression of miR-30d, miR-181a, and miR-
199a-5p was determined by quantitative real-time PCR
analysis in two non-tumorigenic and four tumorigenic
human cell lines and normalized to U6 snRNA.

88

Figure 4.3: Relative expression of miR-30d, miR-181a, and miR-
199a-5p, and their potential target, GRP78, in human cell
lines.

90

Figure 4.4: GRP78 mRNA levels were assessed by quantitative real-
time PCR in untreated cell lines (top) and in cells treated
with the ER stressor Thapsigargin (Tg) (bottom).

91

Figure 4.5: MiR-30d, miR-181a, and miR-199a-5p directly target the
3’UTR of GRP78 with cooperative effects.

93

Figure 4.6: MiR-30d, miR-181a, and miR-199a-5p directly target the
3’UTR of GRP78 with cooperative effects.

94

Figure 4.7: Inhibition of GRP78 mRNA levels in C42B cells by co-
expression of multiple miRNAs.

96

Figure 4.8: Inhibition of GRP78 protein levels in C42B cells by co-
expression of multiple miRNAs.

97

Figure 4.9: Inhibition of GRP78 levels and induction of apoptosis in
C42B cells by co-expression of multiple miRNAs.

99

Figure 4.10: Inhibition of GRP78 levels and induction of apoptosis in
C42B cells by co-expression of multiple miRNAs.

100

Figure 4.11: Expression of miR-30d, miR-181a, and miR-199a-5p
increases TSA sensitivity in C42B cells.

102

xiii

Figure 4.12: Expression of miR-30d, miR-181a, and miR-199a-5p
increases TSA sensitivity, and reduces cell viability and
colony formation in C42B cells.

103

Figure 4.13: Expression of miR-30d, miR-181a, and miR-199a-5p
increases TSA sensitivity, reduces cell colony formation in
C42B cells.

104

Figure 4.14: Lentiviral vectors contain multiple co-transcribed
miRNAs.

107

Figure 4.15: Lentiviral delivery of multiple co-transcribed miRNAs
down- regulates GRP78 protein levels.

108

Figure 4.16: Lentiviral delivery of multiple co-transcribed miRNAs
down- regulates GRP78 protein levels, reduces cell
viability and induces apoptosis in different cancer cell
lines.

109

Figure 4.17: Stable expression of multiple miRNAs results in down-
regulation of GRP78 and up-regulation of PARP-1 protein
levels in different cancer cell lines.

110

Figure 4.18 Stable expression of multiple miRNAs results in inhibited
colony formation in different cells.

111

Figure 4.19: Inhibition of GRP78 mRNA levels in multiple miRNAs
transduced (LV miR30+181+199) C42B cells were
abrogated after GRP78 overexpression.

112

Figure 4.20: Inhibition of GRP78 protein levels in multiple miRNAs
transduced (LV miR30+181+199) C42B cells were
abrogated after GRP78 overexpression.

113

xiv

Figure 4.21: Cell viability in multiple miRNAs transduced (LV
miR30+181+199) C42B cells were abrogated after GRP78
overexpression.

114

Figure 4.22: Colony formation in multiple miRNAs transduced (LV
miR30+181+199) C42B cells were abrogated after GRP78
overexpression.

115

Figure 4.23: MiR-30d, miR-181a, and miR-199a-5p inhibit tumor
growth in athymic mice.

117

xv

ABSTRACT
Two of the most challenging issues clinics face in anti-cancer therapy are tumor
recurrence and drug-resistance. To overcome these two urgent problems, there is
certainly a need to develop promising markers to easily monitor cancer recurrence
and to develop novel strategies to overcome drug-resistance. Epigenetic alterations,
including DNA methylation, histone modifications, and nucleosome occupancy, as
well as non-coding RNAs, such as microRNAs, have been extensively observed in
virtually all types of human malignancies and may correlate with carcinogenesis,
tumor recurrence, and therapeutic outcomes. Therefore, these epigenetic alterations
and microRNAs serve as potential markers or drug targets, and thus harbour wide
applications in diagnostic and therapeutic intervention.

The high risk of recurrence in patients following transurethral resection of bladder
tumor (TURBT) for non-muscle invasive disease necessitates lifelong maintenance
treatment and surveillance. To aid in this process, I have identified a panel of DNA
methylation markers that can be analyzed in urine sediments to accurately predict
and detect bladder tumor recurrence in follow-up visits of TURBT patients.
Importantly, this panel, a combination of the hypermethylated markers SOX1 (a
transcription factor) and IRAK3 (a epigenetic driver), and one hypomethylated
marker L1-MET (a specific LINE1 element), showed a higher sensitivity than urine
cytology and cystoscopy in detecting tumor recurrence. Next, I showed the
functional roles and the regulatory mechanisms of one of these markers, ZO2 (TJP2),

xvi

at its two alterative CpG-rich promoters in the epigenetic levels. Specifically,
distinct mRNA expression and DNA methylation patterns were found at the two
promoters in various human tissues and tumors, indicating that tissue-specific
behaviours also correlated with CpG island promoters. In addition, to characterize a
putative enhancer of TJP2 around the downstream promoter, I analyzed endogenous
DNA methylation and nucleosome occupancy using Nucleosome occupancy and
methylome sequencing (NOMe-seq) in cells with active or repressed TJP2
expression. The results showed distinct nucleosome configurations at different DNA
regulatory regions. Finally, to address the challenge of drug-resistance in cancer
therapy, specifically concentrating on GRP78-mediated drug-resistance, I have
discovered multiple microRNAs (mir-30d, mir-181a, and mir-199a-5p) that act
cooperatively to suppress GRP78 levels and thus GRP78-mediated chemoresistance.
These results showed that the strategy of delivering co-transcribed microRNAs may
hold therapeutic potential.

In summary, I herein identified the confidential markers for the prediction and
detection of bladder cancer recurrence, as well as a microRNA-based approach for
the resensitization of tumor cells to chemotherapeutic drugs. My research in the field
of cancer epigenetics and its translational applications can potentially advance
cancer prevention and management.

1

CHAPTER 1

CANCER EPIGENETICS AND ITS TRANSLATIONAL
IMPLICATIONS

INTRODUCTION
Beyond the well-studied field of genetics and the profound impacts of its
disorganization on cancer, the field of epigenetics emerged to break the conventional
views of genome (Waddington, 2012). In addition to genetic mutations that drive
normal cells to disease states, destruction of epigenetic machineries (also termed
epimutations) serve as second hits for tumorigenesis (Jones and Laird, 1999). The
functional influences of epigenetics on genetics and their interaction have led
researchers to rethink the genetic code by studying the role of different layers of
epigenetics and their relevance to gene expression.

Epigenetics refers to the packaging and organizing of the chromatin and heritable
interpretation of the genome independent of the DNA sequence. Maintenance of the
epigenetic landscape is fundamental to mammalian embryonic development, and
disruption of the normal state causes several human diseases, including cancer. The
field of epigenetics research includes DNA cytosine methylation, modifications of
histone tails and histone variants, nucleosome occupancy at the transcriptional start
sites and microRNAs that cooperatively regulate gene expression (Force and

2

European Union, 2008). A gene can be activated usually by its promoter CpG islands
unmethylated, together with active histone modification marks and nucleosome
deplete regions (NDR) around the transcriptional start sites (TSS), leading to the
open chromatin structure that allows the entry of transcription factors (TF) and DNA
binding proteins (DBP) over the TSS and further activates gene transcription. A
silenced gene harbors methylation within the gene promoters when combining
repressive histone modification marks and nucleosome occupancy at the TSS,
resulting in compacted chromatin that the polycomb group (PcG) complex often
participates in and causes transcriptional repression (Baylin and Jones, 2011). In
addition, chromatin organization at the putative enhancer regions with signature
histone modification and nucleosome configuration has been extensively
documented as it plays a critical role in controlling chromatin architecture at
silenced, poised or active states (Andreu-Vieyra et al., 2011; Taberlay et al., 2011).
Each part of these epigenetic events occurs aberrantly in human disease and can
contribute to cancer. Perturbations in the balance of epigenetic mechanisms have
shown a remarkable correlation with the development of cancer. Therefore,
dissecting cancer epigenetics will expand the current concept of tumorigenesis and
help translate results to the clinic.

OVERVIEW OF EPIGENETICS IN CANCER
Epigenetic alterations are important events acquired in cancer initiation and
progression. Epigenetic abnormalities in any of these processes, including DNA

3

methylation, histone modification and nucleosome positioning characterize cancer
epigenetics. Deregulation of epigenetic settings leads to remodeling of chromatin
configuration and eventually to alteration of gene expression.

DNA METHYLATION
DNA methylation, the introduction of a methyl group to the 5-position of the
cytosine residue in CpG dinucleotides, plays an important role in embryonic
development and differentiation, X chromosome inactivation, and gene imprinting.
The covalent modification of cytosine within CpG in mammalian cells is mediated
by the de novo methyltransferase DNMT3A and DNMT3B, which target un- or
hemimethylated DNA, and DNMT1, which maintains the DNA methylation pattern
usually by targeting hemimethylated DNA during DNA replication (Sharma et al.,
2010). Cooperation of all three enzymes along with methyl-binding domain (MBD)
proteins and histone deacetylases or other co-modifiers can repress gene expression.
The CpG-rich regions (CpG islands) randomly distribute across the genome and
most of them are located at human promoters. CpG islands at specific promoter
regions usually stay unmethylated, thereby allowing gene expression, while
methylation occurs in some repetitive elements resulting in gene silencing (Jones,
2002, 2012).

Among epigenetic effects, abnormal DNA methylation of cancer-related genes at
their promoter CpG islands is one of the most well-characterized epigenetic changes

4

during the development of cancer. Hypermethylation of tumor suppressor genes
mainly on promoters containing CpG islands has been shown to be involved in gene
silencing and disease malignancies. On the other hand, global hypomethylation at
repetitive regions results in aberrant transcriptional activation and thus detection of
DNA methylation pattern changes is a suitable method for assessing the risk of
disease. In addition, DNA methylation alteration is chemically stable and detectable
in human body fluids, and available high throughput technology facilitates the
qualification of DNA methylation changes. For example, GSTP1 hypermethylation
has been described in both tumors and urine samples from patients with prostate
cancer (Cairns et al., 2001; Lee et al., 1994); a panel of candidate markers displayed
specific DNA methylation profiles in hepatocellular carcinoma (HCC), making them
useful in clinical application (Hernandez-Vargas et al., 2010); methylation changes
that occur in DNA-repair gene MGMT after chemotherapy act as predictors for
therapeutic outcome (Esteller et al., 2000); p14ARF promoter hypermethylation was
found in plasma DNA and the methylation status had a relationship with bladder
cancer recurrence (Domínguez et al., 2002).

HISTONE MODIFICATIONS
Post-translational modification of histone proteins on the N-terminal tails,
including methylation, acetylation, phosphorylation, ubiquitylation and sumoylation,
can modulate the stability and architecture of chromatin (Kouzarides, 2007).
Histone-modifying enzymes contain histone acetyltransferases (HAT) and

5

deacetylases (HDAC), which add or remove acetyl groups, respectively, and histone
methylatransferases (HMT) and demethylases (HDM), which mediate the presence
of methyl groups on the histone tails (Allis et al., 2007). A number of histone codes
followed by the activities of these enzymes have been identified to tightly control the
structure of the chromatin. Based on specific modifications on specific residues on
histone tails, histone modifications can result in either activation or silencing of the
gene - active histone modification marks, such as trimethylation of lysine 4 on
histone 3 (H3K4me3) or acetylation of lysine 27 on histone 3 (H3K27ac), usually
correlate with gene expression while repressive histone modification marks, like
trimethylation of lysine 27 or 9 on histone 3 (H3K27me3) (H3K9me3) participate in
the close state of chromatin (Liang et al., 2004). The possible combinations of these
marks indicate a complex control of histone modification in various biological
processes. One important example is that both H3K4me3 and H3K27me3 are co-
localized in embryonic stem cells (ESC). Such bivalent mark allows
developmentally-associated genes to flexibly switch their epigenetic states and
therefore determine the direction of cell fate with some key co-modifiers (Bernstein
et al., 2006).

Another type of epigenetic markers is histone modification whose anomalous
patterns are found in cancer (Seligson et al., 2005). Global abnormalities of histone
modification marks H4K20me3 and H4K16ac were reported in many types of cancer
(Fraga et al., 2005). Also, histone methylatransferases (HMT), for instance, can

6

methylate specific residues depended on different HMTs, including G9a, EZH2,
SUV and PRMT5, and further regulate various cellular pathways and function
(Scoumanne et al., 2009; Sharma et al., 2012). Overexpression of EZH2, which
methylates H3K27, is associated with metastasis of breast and prostate cancer (Valk-
Lingbeek et al., 2004). Generally speaking, mutations or dysregulation of the
corresponding histone-modifying enzymes that result in altered levels of histone
methylation or acetylation occur during tumorigenesis (Nakagawa et al., 2007; Shi,
2007).

NUCLEOSOME OCCUPANCY
Histone octamers (two copies of H2A, H2B, H3, and H4) wrapped by around 146
base pairs of DNA constitute the nucleosome core. Nucleosome occupancy around
the transcriptional start sites (TSS) plays a critical role in gene expression and is
dynamic in the human genome (Schones et al., 2008). The absence of nucleosomes
creates the accessibility for transcription factors (TF) to the specific regulatory loci
and thus acts as a considerable event in driving gene transcription. In contrast, lack
of nucleosome deplete regions (NDR) around the TSS shows a correlation with gene
repression. Nucleosome position has been described to be controlled by DNA
sequence, DNA methylation, TFs and chromatin remodelers which can move
nucleosome closer to or away from the promoters. Therefore, mutations of
nucleosome remodelers, such as SWI-SNF complexes, display tumorgenic
phenotype. For example, appropriate SNF5 expression is important for the survival

7

of patients with melanoma (Lin et al., 2009) while ARID1A mutations occur in
ovarian clear cell and endometrioid carcinomas (Jones et al., 2010; Wiegand et al.,
2010). In cancer, abnormal activities in subunits of nucleosome remodelers suggest
the importance of proper nucleosome organization (Reisman et al., 2009).

Recent studies used high-resolution nucleosome occupancy and methylome assay
(NOMe-seq) to simultaneously analyse endogenous DNA methylation and
nucleosome positioning at promoter and especially enhancer regions on single DNA
molecules, revealing the importance of nucleosome positioning in reprogramming
and stem cell pluripotency (Taberlay et al., 2011; You et al., 2011). Furthermore,
genome-wide nucleosome mapping uncovers the global nucleosome and DNA
methylation signature in CpG and non-CpG island promoters and demonstrates the
correlation of promoter nucleosome depletions with gene expression levels in a
direct manner (Kelly et al., 2012).

MICRORNAS
Small RNAs, such as microRNAs (miRNAs) are non-coding RNAs of
approximately 22 nucleotides in length that target the 3’untranslated region (UTR) of
mRNA, thereby causing translational repression or mRNA cleavage and degradation
(Bartel, 2009). Since the first miRNA, lin-4, was discovered in C. elegans in 1993
(Lee et al., 1993), hundreds of human miRNAs have been identified (Griffiths-Jones
et al., 2006) and have been shown to regulate several important biological processes,

8

including cell proliferation and differentiation, development and apoptosis (Ambros,
2004). The fact that a large number of human genes are targeted by miRNAs and that
many mRNAs contain multiple sites for miRNA binding indicate the potential
impact of miRNAs on the down-regulation of post-transcriptional events (Lewis et
al., 2005; Lim et al., 2005).

Aberrant miRNA expression is widely found in several human diseases, including
cancer (Croce, 2009). Genomic abnormalities (deletions, mutation, amplification, or
translocation), miRNA processing defects, and epigenetic alterations can lead to a
dysregulated miRNA signature in cancer (Fabbri et al., 2007; Friedman et al., 2009b;
Garzon et al., 2006; Garzon et al., 2010; Saito et al., 2006). Genome-wide miRNA
expression profiling of cancer patients has also revealed abnormal miRNA levels that
are tissue- and tumor-specific (Calin and Croce, 2006; Volinia et al., 2006).
Characterized by distinct functions, miRNAs can act either as oncogenes (oncomirs)
or as tumor suppressors by targeting molecules critically involved in carcinogenesis
(Calin et al., 2002; Croce, 2009; Esquela-Kerscher and Slack, 2006; Nicoloso et al.,
2009). For instance, miR-21 and miR-373 are involved in tumor metastasis whereas
miR-15a and miR-16-1 are associated with loss of function in most chronic
lymphocytic leukaemias (CLL) (Croce, 2009; Garzon et al., 2010). It was reported
that miR-504 could regulate p53-mediated effects, including apoptosis and cell-cycle
arrest by binding to the 3’UTR of p53, suggesting the involvement of miRNAs in the
tumor environment (Hu et al., 2010). Lack of tumor suppressor miR-101 expression

9

in bladder cancer up-regulated polycomb protein EZH2 and mediated epigenetic
events (Friedman et al., 2009) whereas overexpression of oncogenic miR-17-miR-92
cluster was extensively found in cancer (Mendell, 2008). All of these effects
demonstrate that miRNAs play a role in the regulation of gene expression and
improper miRNA expression will result in disease states.

TRANSLATIONAL APPLICATION OF EPIGENETIC STRATEGIES FOR
CANCER
Advances in cancer research move forward by connecting basic scientific
discoveries with clinical application and by translating results from bench work into
clinical practice. The fact that epigenetic aberrations present in cancer can be
reversed independently of the genome makes epigenetic therapy a useful and
important approach in cancer treatment. DNA methyltransferase (DNMT) inhibitors,
5-azacytidine (5-aza-CR; azacitidine) and 5-aza-2’-deoxycytidine (5-aza-CdR;
decitabine), which are FDA approved, have entered the clinical setting for the
treatment of myelodysplastic syndrome (MDS) patients. Inhibitors of histone
deacetylation (HDAC), such as suberoylanilide hydroxamic acid (SAHA,
vorinostat), can reactivate tumor suppressor genes that are epigenetically depressed
in cancer and cause inhibition of tumor growth. It is also known that the use of
DNMT inhibitors along with HDAC inhibitors has a combination effect to cure
cancer (Kelly et al., 2010; Soriano et al., 2007; Voso et al., 2009). In addition, the
use of miRNA-based therapy, either by restoring tumor suppressor miRNA

10

expression or by targeting oncogenic miRNAs, represents novel therapeutic
strategies (Saito et al., 2006). To date, a list of functional miRNAs have been
identified and utilized in the clinic. The emergence of such epigenetic drugs could
aid in sensitizing cancer cells to chemotherapy, avoiding drug resistance, and further
increasing efficacies through synergistic activities. Furthermore, epigenetic markers
with their advantages of stable and early appearance may offer an additional option
useful in the clinic that can detect, diagnose and prognose human disease and predict
therapeutic responses.

A better understanding of the mechanisms of epigenetic status disruption and a
deeper exploration of its correlation with disease progression will help apply our
findings in epigenetics to translational medicine, including biomarker development
for earlier detection of cancer and breakthrough strategies for effective cancer
treatment.

DEVELOPMENT OF DNA METHYLATION MARKERS
Aberrant gene silencing by DNA methylation at promoter regions of tumor
suppressor genes or oncogenic gene activation by DNA demethylation at repetitive
regions are both recognized as early and common epigenetic changes in cancer
development. DNA methylation analysis of cancer-related genes serves as promising
biomarkers for the early cancer detection and also for disease outcome prediction. In
addition to using DNA methylation as a marker for cancer diagnosis, close screening

11

of possible premalignancy can guide clinical decisions and benefit a patient’s quality
of life.

GSTP1 hypermethylation has been shown to be strongly correlated with prostate
cancer and its DNA methylation changes were detectable in patients’ urine (Cairns et
al., 2001), suggesting that GSTP1 may be a powerful DNA methylation marker for
prostate cancer. Results showing methylation of TFPI2 in stool DNA from colorectal
cancer patients indicated a novel biomarker for the detection of colorectal cancer
(Glöckner et al., 2009), and methylation of gene p16, CDH13, RASSF1A, and APC
in lung tumors were considered as markers for cancer recurrence (Brock et al.,
2008). In addition, hypermethylation of RUNX3 was highly associated with bladder
cancer development and progression (Kim WJ et al., 2005). Methylation of TIMP-3
(tissue inhibitor of metallinoproteinase 3) showed a significant association with
recurrence-free survival in patients with resection of superficial non-invasive bladder
carcinoma (Friedrich MG et al., 2005) while hypermethylation of DAPK (death-
associated protein kinase 1) indicated early recurrence of bladder cancers (Tada Y et
al., 2002). All of these results suggest that using DNA methylation as biomarkers for
clinical use is promising since there are frequent methylation alterations during
carcinogenesis and they are more stable and reliable than RNA or protein markers
(Laird, 2003).

12

Genome-wide high-throughput techniques have now been increasingly applied to
discover DNA methylation markers or to identify CpG island methylator phenotype
(CIMP) for the characterization of cancer (Noushmehr et al., 2010). Importantly,
DNA methylation can be measured and quantified on a high-throughput platform.
Revolution of powerful, next generation sequencing approaches enables researchers
to unmask cancer epigenomic profiles at a fast pace. Our group has previously
demonstrated the existence of hypermethylation of CpG island promoters in bladder
cancers relative to their normal counterparts using a variety of techniques including
bisulfite sequencing, MethyLight and methylation-sensitive single nucleotide primer
extension (Ms-SNuPE). This work has lead to the identification of several potential
DNA methylation markers in patients with bladder tumors (Salem et al., 2000; Wolff
et al., 2005). Furthermore, both hyper- and hypo-methylation markers specific to
bladder tumors were discovered in a study in which the Illumina GoldenGate
platform was used to analyze the methylation profiles of normal urothelial tissues
and different subtypes of bladder tumors (Wolff et al., 2010). This resulted in a panel
of candidate biomarkers, of which several were validated using bisulfite sequencing
and pyrosequencing in tumors and further, in paired urine samples. In addition to
altered methylation in tumors, changes in DNA methylation levels of certain
candidate markers can also be detected in the urine sediments of patients with tumors
when compared to cancer-free individuals (unpublished data). In chapter 2, I show
results using pyrosequencing for quantitative DNA methylation determinations in an
independent cohort of urine samples obtained from patients with bladder cancer and

13

from patients after tumor resection for longitudinal tracking of recurrence. All of the
results suggest good sensitivity and specificity of these markers, either alone or in
combination, in tumor detection as well as in tumor recurrence assessment, making
DNA methylation a potentially good epigenetic biomarker.

In summary, aberrant DNA methylation occurs early during tumorigenesis and is
present in premalignant lesions, and such alteration events can be detected in both
tissues and biological fluids, such as urine, blood and sputum on a genome-wide
platform. Given its non-invasive characteristics and ready accessibility, such samples
have been utilized in DNA methylation analysis to assist in the processes of tumor
diagnosis, detection and prognosis in routine clinical practice.

DEVELOPMENT OF THERAPEUTIC MICRORNAS
Another epigenetically controlled area of gene regulation is the field of non-
coding RNAs, such as microRNAs (miRNAs) (Kappelmann et al., 2012; Zhang et
al., 2012). Since miRNAs were identified as one class of the crucial regulators for
gene expression, scientists have taken efforts to explore the functions of miRNAs
during the past decade. Besides controlling normal development and cell function,
miRNAs play a critical role in human cancers, serving as oncomirs, or tumor
suppressor miRNAs (Esquela-Kerscher and Slack, 2006). Aberrant expression of
miRNAs is a common event in various cancers and has been considered as a novel
target for cancer therapy (Garzon et al., 2010). The down-regulation of

14

p21Cip1/Waf1 mRNA and protein levels was shown to be caused by the individual
binding of different p21-targeting miRNAs to p21 3’UTR in HEK 293 cells (Wu et
al., 2010). Multiple androgen receptor (AR)-inhibiting miRNAs decreased AR
protein levels and AR-induced proliferation when they were individually introduced
into prostate cancer cells (Ostling et al., 2011). Such functional studies indicate that
gene expression is tightly controlled and finely executed by miRNA networks and
therefore, allowing miRNAs to be used either as good targets for epigenetic therapy
or as potential epigenetic drugs themselves.

Moreover, a single gene sequence usually encodes for multiple predicted miRNA
binding sites (Krol et al., 2010; Thomas et al., 2010) and a given gene product could
be targeted by either single or multiple miRNAs (Kloosterman and Plasterk, 2006).
Given this concept, repression of a specific miRNA signature as a potential
therapeutic target or the use of multiple miRNAs to cooperately achieve better
inhibitory effects on oncogenic translation and drug resistance represents the basis
for miRNA-based therapy.

Epigenetic therapy using miRNAs, unlike conventional RNA interference, reaches
clinical value for their endogenous actions. Reactivation of tumor suppressor
miRNAs that are aberrantly repressed in cancer could be achieved via inducing
miRNA expression, like Aza treatment, or by introducing synthetic miRNAs to
restore their functions as tumor suppressors (Friedman et al., 2009a; Saito et al.,

15

2006). On the other hand, down-regulation of oncogenic miRNAs may also be useful
in suppressing tumor growth and sensitizing cancer cells to drug treatment. Anti-
miRNA therapy that can switch miRNA expression back to “normal” status assists
cancer treatment and provides an alterative approach beyond gene therapy.
Undoubtedly, increasing miRNA stability, enhancing drug efficacy and determining
the best method for miRNA delivery are important issues that are worthy of further
endeavors.

CONCLUSION
The field of cancer epigenetics and epigenetic implication in treating cancer have
rapidly evolved. Establishing epigenetic systems with the unique strengths of
quantitative high throughput techniques, endogenous regulators, and combination
activities will potentially benefit cancer prevention and management. Its usefulness
in premalignant detection and drug resistance reduction facilitates the translation of
basic biology into clinical applications.

16

OVERVIEW OF THESIS RESEARCH
Due to the high recurrence rate in patients with a history of non-invasive
superficial bladder tumor, I measured DNA methylation levels in serial urine
sediment samples from patients whom had undergone tumor resection by performing
pyrosequencing to monitor tumor recurrence. This reliable and quantitative
technology which can reproducibly quantify DNA methylation levels helps the
detection of reappearance of aberrant methylation in urine sediments. The
experimental design in chapter 2 has the distinct advantage of being non-invasive,
and of allowing multiple screenings for the same patient in long-term follow-up
visits after surgery. I identify significant indicators, including both hyper- and hypo-
DNA methylation markers for bladder tumor recurrence with a high degree of
sensitivity and specificity in urine sediments. Importantly, the combination of a
transcription factor, an epigenetic driver gene, and a specific LINE1 element could
accurately predict the risk of recurrence and could be used to routinely and easily
screen post-surgical patients during follow-up visits.

To determine epigenetic changes of specific genes as a functional mechanism
driving tumorigenesis, in chapter 3, I elucidated mRNA expression, DNA
methylation status and nucleosome positioning in gene of TJP2 at its two alternative
promoters. The distinct patterns were shown in various human normal tissues and
cancer cell lines, indicating different roles of alternative promoters in tissue

17

specificity and tumorigenesis. In addition, the putative enhancer regions require
further mapping of histone modifications and transcription factors.

It has consistently shown that GRP78 plays an important role in cancer
development and therapeutic resistance and that microRNAs, representing a part of
epigenetics, have a great potential as anti-cancer therapeutic agents. Thus, in chapter
4, I identified the specific microRNAs that target human GRP78, their cooperative
effect on cancer cells, and their combinatorial action with drug therapy by restoring
miR-30d, miR-181a, and miR-199a-5p into different types of cancer cell lines both
transiently and permanently. In addition, I analyzed normal-tumor paired samples
from patients with bladder, colon, and prostate cancer to highlight the clinical
relevance of GRP78 expression and the three microRNAs that target GRP78. The
effect of GRP78 overexpression on cell survival, colony formation, and PARP-1
levels was also shown to confirm the specificity of the microRNAs binding to the
3’UTR of GRP78. Furthermore, I showed the in vivo anti-tumor potential of these
microRNAs in the three-dimensional gelatin-TGase system. The use of multiple
microRNAs to target GPR78 could provide new avenues for cancer treatment to
inhibit the survival of tumor cells overexpressing GRP78.

Taken together, this work indicates that epigenetic events, including DNA
methylation and microRNAs strongly connect to cancer in both biological and
translational levels.

18

CHAPTER 2

A Panel of Three Markers Hyper and Hypomethylated in
Urine Sediments Accurately Predicts Bladder Cancer
Recurrence

INTRODUCTION
Bladder cancer was one of the ten most prevalent malignancies in males in 2011
ranking number fourth and number eighth in terms of deaths and new cases,
respectively (Siegel et al., 2011). The most common presenting symptom of bladder
cancer is hematuria, and general risk factors for the disease include smoking and
carcinogen exposure (Morgan and Clark, 2010). Non-muscle invasive bladder cancer
(NMIBC) accounts for 80% of all the cases, and can be further classified into
mucosa only (Ta), carcinoma in situ (Tis) and lamina propria invading (T1) lesions.
The rest of the cases present as muscle invasive bladder cancer (MIBC) (Babjuk et
al., 2011; Sobin et al., 2009). The primary treatment for NMIBC is transurethral
resection of bladder tumor (TURBT) with or without intravesical chemo or
immunotherapy; however, over 50% of patients will recur after the TURBT
procedure, with the highest rate of recurrence occurring in patients with high risk
disease (Millán-Rodríguez et al., 2000; Shelley et al., 2010). As a result, patients

19

with NMIBC require frequent and sometimes lifelong monitoring following TURBT,
making bladder cancer one of the most costly types of cancer to manage.

The current gold standard for monitoring of bladder cancer recurrence involves the
use of cystoscopy and cytology (Babjuk et al., 2011; Morgan and Clark, 2010).
Disease surveillance is cumbersome because of the invasive nature of cystoscopic
examination and because of the low sensitivity of urinary cytology in the detection of
low-grade tumors (Lintula and Hotakainen, 2010). In recent years, efforts have been
devoted to find better markers of disease diagnosis and prognosis in samples
collected by non-invasive methods, such as urine sediments (Sturgeon et al., 2010).
Bladder tumor cells have weaker cellular attachment than normal or benign bladder
urothelium and therefore they shed more and can be collected in urine (urine
sediments). The addition of NMP-22 (nuclear matrix protein 22), BTA (bladder
tumor antigen), or UroVysion FISH (fluorescence in situ), have shown to help with
the increased sensitivity of cytology (Parker and Spiess, 2011). However, mainly due
to the insufficient specificity (62%-85%), the markers proposed to date have not
been widely adopted in routine clinical practice (Reinert, 2012). Therefore, there is
an urgent need to find reliable markers to monitor recurrence in TURBT patients,
which in turn, may help facilitate and improve disease management.

Epigenetic changes, namely changes in chromatin structure that regulate gene
expression such as DNA methylation, occur during the process of tumorigenesis

20

(Jones, 2012). DNA methylation occurs via the introduction of a methyl group in the
5-position of the cytosine residue of CpG dinucleotides. In normal cells, CpG-rich
regions or CpG islands located at specific promoter regions are usually unmethylated
thereby allowing transcriptional activity, whereas methylation occurs in some
repetitive elements resulting in their silencing (Jones, 2012; Sharma et al., 2010).
Aberrant DNA methylation including increases and decreases at specific loci is both
one of the earliest and the most common epigenetic change that occurs during
tumorigenesis and it can be detected in premalignant lesions (Niwa et al., 2010;
Ushijima and Hattori, 2012; Wolff et al., 2010a; Wolff et al., 2010b). Changes in
DNA methylation are chemically stable and can be quantified, which makes them
potentially good tumor markers, including for bladder cancer (Laird, 2003; Wolff et
al., 2005). Inactivation of tumor suppressor genes by gain of DNA methylation
(hypermethylation) or global loss of DNA methylation (hypomethylation), which
activates genes that are normally not expressed have both been observed in bladder
tumors (Kim and Kim, 2009a; Reinert et al., 2011; Vallot et al., 2011; Wolff et al.,
2010a). Further studies have shown that methylation changes found in urine
sediments mirror those found in tumor tissues, indicating cancer-specific features
(Friedrich et al., 2004; Kim and Kim, 2009b; Reinert et al., 2011; Seifert et al.,
2007).

We previously identified both hyper- and hypomethylated regions in primary
bladder tumors and its premalignant lesions (Wolff et al., 2010b). Further studies

21

from our laboratory demonstrated that a specific LINE1 element, which is located
within the MET oncogene (L1-MET) and activates an alternate transcript of MET,
was hypomethylated, whereas the promoter of zona occludens 2 (ZO2; tight junction
protein 2) was hypermethylated in bladder tumors as well as in adjacent
histologically normal urothelium, suggesting that epigenetic changes precede
morphological changes, a phenomenon termed an “epigenetic field defect”, which
might be involved in malignant predisposition. (Wolff et al., 2010a; Wolff et al.,
2010b). We also found a group of genes that showed methylation changes both in
bladder tumors and urine sediments from bladder cancer patients (Friedrich et al.,
2004). Based on these previous studies, we hypothesize that DNA methylation
changes in urine sediments from TURBT patients can be used to detect early bladder
cancer recurrence. To test our hypothesis, we collected urine samples from TURBT
patients over a 7-year period at follow-up visits and assessed the methylation status
of a panel of markers, including cancer specific hypermethylated markers (HOXA9,
SOX, NPY), a epigenetic driver gene (IRAK3) (De Carvalho et al., 2012), and field
defect markers (ZO2 and L1-MET) (Wolff et al., 2010a; Wolff et al., 2010b). Our
results show that methylation changes of these markers individually display a
positive correlation with tumor recurrence with high sensitivity and specificity
(p<0·0001) . In addition, we show that the combination of SOX1, IRAK3 and L1-
MET markers provides better resolution than cytology and cystoscopy in the
detection of early recurrence changes. Overall, our results suggest a critical role of a
balance between hyper- and hypo-DNA methylation for bladder carcinogenesis, and

22

also provide a non-invasive and cost-effective way to assess patients post TURBT,
which if applied in the clinical setting, may help to detect unrecognized or
unanticipated tumor recurrence early, guide treatment direction and limit the use of
invasive procedures such as cystoscopies.

23

MATERIALS AND METHODS
Patients and sample collection
The study population includes patients under surveillance for tumor recurrence
following TURBT for non-invasive urothelial carcinoma (Tis, Ta, T1; grade1-3).
Urine samples were obtained from 90 such NMIBC patients at each available clinical
follow-up visit. Patient’s age ranged from 41 to 96 years old, with a median age of
69. Urine collection at follow-up visits was performed at the Department of Urology,
Keck School of Medicine of University of Southern California (USC) from 2004 to
2011 according to the institutional guidelines of the USC Norris Comprehensive
Cancer Center, in compliance with Institutional Review Board–approved protocols.
Patients at high risk of recurrence (carcinoma in situ, high grade Ta or T1 disease)
had received prior intravesical therapy with Bacillus Calmette-Guerin (BCG) or
mitomycin C as clinically indicated at the discretion of the treating physician. A total
of 368 samples were collected under patient informed consent at different follow-up
visits over a period ranging from 5 to 89 months. The timeline of urine sample
collection is presented in Figure 2.2. The baseline clinicopathological characteristics
of the patients are summarized in Table 2.2.

Tumor recurrence was defined as biopsy-proven bladder tumor, severe atypia or
papillary lesions or any concurrent suspicious criterion (neoplasia) in cytology or
cystoscopy after a previous surgery. Over the collection period, 34 patients had
tumor recurrence, while 56 patients were not diagnosed with recurrence through the

24

last follow-up visit. The clinical characteristics of 34 recurrent tumors are
summarized in Table 2.4. Out of the 34 patients with recurrence, 31 provided a urine
sample at the time of diagnosis. Approval for research on human subjects was
obtained from the USC Norris Comprehensive Cancer Center review boards. Tumors
were characterized according to the criteria of the American Joint Committee on
cancer (Edge et al., 2010) and staging and grading was based on the TNM
classification of the International Union Against Cancer (Sobin et al., 2009).

DNA extraction from urine sediments and DNA methylation analysis by
pyrosequencing
Urine samples (~50ml) were centrifuged for 10 min at 1500g and DNA extraction
from urine sediments was preformed as previously reported (Friedrich et al., 2004).
DNA was bisulfite-converted using EZ DNA Methylation Kit (Zymo Research,
Irvine, CA, USA) according to the manufacturer’s instructions. Six DNA
methylation markers were selected from our previous study (Wolff et al., 2010a;
Wolff et al., 2010b); the regions of interest were PCR amplified using biotin-labeled
primers (Table 2.1) and analyzed by pyrosequencing, a high throughput and
quantitative tool for DNA sequence detection. The percentage of methylated
cytosines divided by the sum of methylated and unmethylated cytosines was
measured using PSQ HS96 (Qiagen, Valencia, CA, USA) as previously described
(Wolff et al., 2010a).

25

Statistical analysis
ROC curves summarize the accuracy of our markers in DNA urine sediment from
87 independent samples, selected at the time of the last follow up visit for non-
recurrent patients (n=56), or at the time of first recurrence for the patients with
recurrence (n=31). A subset of 83 patients with complete data on all markers was
used to build a multivariable predictor model. We used stepwise logistic regression,
selecting variables to add or subtract based on the Akaike Information Criterion.
Sensitivity and specificity were estimated using 5-fold cross-validation, repeating the
model selection for each subdivision of the data. The final model was then evaluated
on the remaining samples from our data set. Control samples included visits prior to
the last follow-up visit where the patient was not diagnosed with bladder cancer; case
samples included recurrences occurring after the first recurrence and samples at the
initial clinic visit when the patient presented with bladder cancer.

26

Table 2.1 Primer sequences used in pyrosequencing
Primer Name Sequence Amplicon size
HOXA9 91
sense 5’ ATGAAATTTGTAGTTTTATAATTTT
anti-sence 5’ Biotin-ATTACCCAAAACCCCAATAATAAC
sequencing 5’ GTTTTATAATTTT
SOX1 109
sense 5’ GGTATTTGGGATTAGTATATGTTTAG
anti-sence 5’ Biotin-CTATCTCCTTCCTCCTAC
sequencing 5’ TTAGTATATGTTTAG
NPY 106
sense 5’ GGGTTGTTTTTATTTTTGGTAGGATTAGA
anti-sence 5’ Biotin- CACCAAAACCCAAATATCTA
sequencing 5’ AGGAAAGTAGGGAT
IRAK3 136
sense 5’ GGAGTTTTGAGTTTTGGGTTTT
anti-sence 5’ Biotin- CCTAACCAAACCTAAAAATTACC
sequencing 5’ AGGTGTGAAGGGG
TJP2 84
sense 5’ GGTTTTTAGATAGGATTTAAAATTTTGAG
anti-sence 5’ Biotin-CAAAACCTCACACAAACAACTTC
sequencing 5’ AGGTTTTTTTAGTT
L1-MET 294
sense 5’ GTGTTTTTTAAGTGAGGTAATGTT
anti-sence 5’ Biotin- ATCCAACCACTACAAACTAC
sequencing 5’ GTTGGGAGTTGTAGAT

27

Results
DNA methylation analysis in urine sediments
We previously showed that HOXA9, SOX1, NPY, IRAK3, and ZO2 are
hypermethylated in bladder tumor samples (Wolff et al., 2010b) and the functional
role of IRAK3 as epigenetic driver gene in cancer (De Carvalho et al., 2012). In
addition, we demonstrated that hypermethylation of ZO2 or hypomethylation of L1-
MET in adjacent normal tissues may contribute to tumor recurrence (Wolff et al.,
2010a; Wolff et al., 2010b). To evaluate whether hypermethylation of HOXA9,
SOX1, NPY, IRAK3 and ZO2, and hypomethylation of L1-MET could also be
detected in urine sediments, we analyzed urine samples collected from patients with
bladder tumors (n=20) and from age-matched cancer-free controls (n=20) using
pyrosequencing. The results show that DNA methylation of HOXA9 (p<0·0001),
SOX1 (p=0·0017), NPY (p=0·005), IRAK3 (p<0·0001) and ZO2 (p<0·0001) was
significantly increased, while methylation of L1-MET (p<0·0001) was significantly
decreased in urine sediments from cancer patients compared to healthy donors,
indicating that the methylation status of these DNA methylation and epigenetic field
defect markers in urine sediments mirror that of the tumor (Figure 2.1).

28

Figure 2.1 DNA methylation alterations can be detected in urine sediment
samples from patients with bladder cancer. The DNA methylation status of
HOXA9, SOX1, NPY, IRAK3, ZO2 and L1-MET was analyzed by pyrosequencing
in urine sediments from bladder cancer patients (TU) and in control urine sediments
from age-matched cancer-free individuals (CU). Paired t-test was performed. ** :
p<0·01; *** : p< 0·001.

29

Longitudinal study of DNA methylation changes in urine sediments collected
from TURBT patients at the time of follow-up visits
To examine whether aberrant DNA methylation of HOXA9, SOX1, NPY, IRAK3,
ZO2 (hypermethylated) and L1-MET (hypomethylated) in urine sediments is
associated with tumor recurrence, we first analyzed their DNA methylation status in
368 urine sediments collected in follow-up visits from patients that had undergone
prior tumor resection (Figure 2.2) and calculated the Spearman correlation of DNA
methylation level for each marker (Figure 2.3A). Individual DNA methylation
marker success rates averaged 98·9% across all samples (range = 94·9-100%). Next,
the DNA methylation status of these markers in 31 urine sediments from patients
collected at the time of first recurrence was compared to that of 56 samples from the
last follow-up visit of patients who did not recur within the study period. The mean
DNA methylation levels of these markers are shown in Table 2.3. Our results show
that the six candidate markers individually showed high sensitivity and specificity in
recurrence detection as evidenced by the ROC curves and AUC values of 0·93
(HOXA9), 0·95 (SOX1), 0·94 (NPY), 0·9 (IRAK3), 0·93 (ZO2), and 0·95 (L1-
MET) (p<0·0001; Figure 2.3B). In the group of patients without bladder tumor
recurrence, urine sediment samples showed consistent DNA methylation levels
throughout the duration of surveillance follow-up visits; all the markers methylated
in bladder tumors displayed low methylation levels whereas the marker
hypomethylated in bladder tumors (L1-MET) maintained high methylation levels
after TURBT (Figure 2.4; patients 7873 and 9214). In contrast, the group of patients

30

who had bladder tumor recurrence displayed changes in the DNA methylation status
of all six markers at the time of clinically defined recurrence. For example, in patient
7258 DNA methylation levels of hypermethylated markers SOX1, NPY, IRAK3, and
ZO2 continued to increase until recurrence was confirmed with a positive cystoscopy
and biopsy 19 months after the first urine sample was obtained. Following resection
surgery, a decrease in previously elevated methylation levels can be seen (Figure
2.4). A similar pattern was observed in patient 7145; however, the overall
methylation levels measured at first recurrence still held at follow-up visits at six and
nine months after re-TURBT. This suggests incomplete removal of the recurrent
tumor, as suspicious cytology was recorded at these follow-up visits (Figure 2.4).
Our results demonstrate that hypermethylation of HOXA9, SOX1, NPY, IRAK3,
ZO2 and hypomethylation of L1-MET are consistent with disease recurrence.
Furthermore, the methylation levels of these markers displayed a clear trend in the
samples obtained at follow-up visits leading to the confirmation of recurrence:
hypermethylated markers continued to increase, while those of the hypomethylated
markers decreased (Figure 2.4). Taken together, the results demonstrate that the
methylation status of these markers in urine sediments not only shows a significant
correlation with recurrence (p<0·0001), but also has predictive value, as methylation
changes could be detected prior to clinical evidence of recurrence.

31

Table 2.2 The clinicopathological characteristics of 90 TURBT patients

No recurrence

Recurrence

Characteristic

N=56

N=34

Age - yr

Median

71

69

Range

42-96

41-87

Sex – no. (%)

Male

48 (86)

27 (79)

Female

8 (14)

7 (21)

Histology TCC – no. (%)

57 (100)

34 (100)

Number of tumors – no. (%)

Unifoci 19 (34)

18 (53)

Multifoci

16 (29)

12 (35)

Missing

21 (37)

4 (12)

T Stage – no. (%)

Tis

2 (4)

1 (3)

Ta

37 (66)

19 (56)

T1

17 (30)

14 (41)

Tumor grade
§
– no. (%)

Low

26 (46)

17 (50)

High

30 (54)

17 (50)

Concomitant CIS – no. (%)

11 (20)

7 (21)

Treatment – no. (%)

Adjuvant BCG

36 (64)

20 (59)

Adjuvant chemotherapy instillation

11 (19)

12 (35)

Follow-up time since TURBT -yr

5.5 (0.6-9.7)

4.7 (0.4-26)

Study follow-up time -yr

3.5 (0.4-7.1)

3.6 (0.5-7.4)

Total urines analyzed − no. 208 160
Urines analyzed/patient − no.
Mean (±SD) 3.7±1.8 4.7±2.1
Range 2-9 2-10
§ Grade1 and 2 are low grade. Grade 3 and more are high grade.
TCC: transitional cell carcinoma; TURBT: transurethral resection of bladder tumor;
CIS: carcinoma in situ; BCG: Bacillus Calmette-Guerin

32

Figure 2.2 Timeline of longitudinally collected urine sediment samples in follow-
up visits from bladder cancer patients after tumor resection. Each patient’s
starting point, denoted by time 0, refers to the first follow-up visit in the study when
a urine sample was collected. A follow-up visit marked in red indicates the time of
recurrence.

33

A B

Figure 2.3 DNA methylation of HOXA9, SOX1, NPY, IRAK3, ZO2
(hypermethylated) and L1-MET (hypomethylated) in 368 urine sediment
samples showed high sensitivity and specificity in the detection of tumor
recurrence. (A) Spearman correlation of DNA methylation levels of each marker in
urine sediment samples of TURBT patients. (B) Receiver operating characteristic
(ROC) curves of HOXA9, SOX1, NPY, IRAK3, ZO2 and L1-MET were created
using 31 urine sediments of TURBT patients at first recurrence and 56 urine
sediments from the last follow-up of recurrence-free patients. AUC: area under the
curve.

34

Table 2.3 DNA methylation status in urine sediment samples of TURBT
patients at the time of first recurrence and the last follow-up visit of no
recurrence patients.
#
n=55;
^
n=55;
*
n=29
P value was calculated by the paired t-test.

35

A

B

Figure 2.4 A panel of six DNA methylation markers tested in urine sediments
from TURBT patients was positively correlated with bladder tumor recurrence.
(A, B) Long-term DNA methylation analysis in TURBT patients and its relationship
with clinical status in patients who had no recurrence (A) and patients who had
recurrence (B). (-): negative; (*): suspicious; (+): positive (biopsy-proven bladder
tumor, severe atypia or papillary lesions in cytology or cystoscopy); R: recurrence;
BCG: Bacillus Calmette-Guerin.

36

A three-marker panel
To determine the combination of markers capable of detecting tumor recurrence in
urine sediments with the highest sensitivity and specificity, we built a model of
multiple markers by 5-fold cross-validation, using 29 samples taken at the time of
first recurrence after TURBT, and 54 samples from patients who were recurrence-
free at the last time of urine collection. From this model, three markers SOX1,
IRAK3, and L1-MET were found to be the best possible marker combination (risk
score = -0·37608 + 0·17095 × SOX1 + 0·21604 × IRAK3 – 0·09887 × L1-MET).
Scores above zero predict recurrence. Among the 54 samples from patients with no
recurrence, we found that 94% of patients showed negative scores (methylation score
lower than the cut-off), with three patients displaying positive scores (6%).
Importantly, in the 29 samples from patients with recurrence, 93% showed positive
scores for the presence of recurrence (p<0·0001; Figure 2.5A). The 5-fold cross-
validation analysis indicated that these markers can discriminate between recurrent
and non-recurrent patients in independent data with an estimated sensitivity of 86%
and specificity of 89% (Figure 2.5B). This three-gene model was then validated
using the remaining samples in our cohort: 23 samples taken at a visit where bladder
tumors were present (TU, nine recurrences after the first recurrence, and 14 at the
time of entry into the study), and 135 samples from patients who had not developed
cancer during a given follow-up time (CU). Notably, the three-marker model also
showed high sensitivity (83%) and specificity (97%) in the validation sample set:
131 urine sediment samples from patients with no recurrence displayed methylation

37

negative scores whereas 19 urine sediment samples taken from patients with known
urothelial carcinoma displayed a positive methylation score (Figure 2.5C; p<0·0001).
The DNA methylation status of our three-marker model showed no correlation with
any of the primary tumor characteristics, irrespective of tumor recurrence; however,
a positive correlation was found between DNA methylation status and tumor grade
of the recurrent tumor (Table 2.5). These results demonstrate that the combination of
tumor-specific hypermethylated marker SOX1, epigenetic driver IRAK3, and the
field defect hypomethylated marker L1-MET can detect disease recurrence with high
sensitivity and specificity. The results also suggest that a balance between hyper- and
hypo-DNA methylation is important for bladder carcinogenesis, further establishing
the epigenetic field defect as a factor involved in malignant predisposition.

Power of prediction of recurrence
To evaluate whether methylation of the three-marker model predicts recurrence in
our longitudinal study samples, we screened DNA methylation and calculated risk
scores (-0·37608 + 0·17095 × SOX1 + 0·21604 × IRAK3 – 0·09887 × L1-MET) in
every urine sample obtained at follow-up visits from 90 TURBT patients. DNA
methylation risk scores given by the combination of SOX1, IRAK3 and L1-MET in
the no recurrence group post-TURBT (CU samples reported in Figure 2.5C) were
lower than the cut-off value throughout months of continuous monitoring (Figure
2.6A; patients 8617 and 7789). In contrast, in the group of patients with tumor
recurrence, DNA methylation scores changed and showed higher than cut-off values

38

A C

B

Figure 2.5 A three-marker signature showed high sensitivity and specificity in
detecting tumor recurrence. (A) The risk score of -0·37608 + 0·17095 × SOX1 +
0·21604 × IRAK3 – 0·09887 × L1-MET was calculated in the urine sediments of
TURBT patients with no recurrence at the last follow-up and with recurrence. (B) 5-
fold cross-validation showed a sensitivity of 86% and specificity of 89%. (C) This
three-marker model was validated in a separate urine sediment samples that included
urine sediments from recurrence-free patients before the last follow-up visit (CU)
and urine sediments of patients with known urothelial carcinoma (TU), and the
sensitivity and specificity were determined. Risk scores above the cut-off value (red
dashed line) denote positive scores, while those below signify negative scores.

No Recurrence Recurrence
-10
0
10
20
30
Risk Score
***
Sensitivity= 93%
Specificity= 94%
(N=54)
(N=29)
CU TU
-10
0
10
20
30
Risk Score
***
Sensitivity= 83%
Specificity= 97%
(N=135)
(N=23)

39

Table 2.4 The clinical characteristics of 34 recurrence bladder cancer patients.

40

Table 2.5 Association between the score from the three-marker signature and
the bladder tumor pathological characteristics of 90 TURBT patients.

Paired t-test was performed. * : p<0·05; *** : p< 0·001.

41

(Figure 2.6B). Positive DNA methylation scores were also found in 90% of the
samples (34 out of 38) at the time of recurrence diagnosis. The sensitivity of these
markers is superior to that of both cytology (16%) and cystoscopy (8%) when
considering the same visits to the clinic (Figure 2.7). Furthermore, DNA methylation
scores were higher than the cut-off value in urine sediments collected before
recurrence, in some cases at least five months prior to the clinical diagnosis of
recurrence (Figure 2.6B; patients 8928 and 6804).

To quantify the prediction value of the three markers, we analyzed risk scores in
the period before recurrence in 189 samples from recurrence-free patients and 65
samples from patients who ultimately had recurrence. We found 46% (30 out of 65
urine samples) positive DNA methylation scores in the recurrence group prior to
recurrence and only 4% (7 out of 189 urine samples) positive scores in the no
recurrence group (Figure 2.8). When analyzing all samples (anytime visits) using our
three-marker panel, we found 62% (64 out of 103 urine samples) positive DNA
methylation scores in the recurrence group. This represents a 10- and 20-fold
increase in the number of positive samples detected at anytime visits compared to
urine cytology (6%) or cystoscopy (3%), respectively (Figure 2.9). Furthermore, our
results show that out of 71 samples with DNA methylation positive scores detected
anytime in the follow-up period, 64 were obtained from our 30 patients who
ultimately had recurrence (90%, positive prediction fraction, PPF), whereas 182 out
of 221 samples with negative scores correlated with no recurrence (82%, negative

42

A B

Figure 2.6 The risk scores given by the combination of three DNA methylation
markers in the continuously monitored urine showed distinct patterns in
recurrence and no recurrence patients. (A, B) DNA methylation levels of the
three-marker combination were used to calculate the risk score for recurrence in the
urine sediment samples from TURBT patients who had no recurrence (A) or had
recurrence (B) and of two individual patients. V: TURBT operation; R: recurrence;
risk score = - 0·37608 + 0·17095 × SOX1 + 0·21604 × IRAK3 – 0·09887 × L1-
MET. The red dashed line indicates the cut-off value. The orange arrow represents
positive scores before recurrence.

43

Figure 2.7 Percentage of urine sediments that had positive scores (DNA
methylation score calculated to be higher than cut-off values) at the time of
recurrence (N=38) and the comparison with cytology and cystoscopy reports at
the same visits to the clinic.

44

prediction fraction, NPF) (Figure 2.9). The results demonstrate that the three-marker
model can successfully detect current and predict subsequent recurrence in 90% of
the DNA methylation positive urine samples while these same samples showed 30%
suspicious and positive in cytology and 44% suspicious and positive in cystoscopy.

At the patient level, 80% of patients (16 out of 20 patients) whose urine samples
showed positive DNA methylation scores for the first time in the collection periods
developed recurrence later (Figure 2.8). Samples from five of these 16 patients
consistently displayed positive scores, samples from two showed subsequent
fluctuating results; and samples from the remaining nine had only one collection
point prior to recurrence. Out of the 70 patients who did not have a history of
positive DNA methylation scores, 52 (74%) did not recur (Figure 2.8). These data
highlight the importance of the history of positive DNA methylation scores in the
accurate interpretation of the results from the three-marker model. Our results
indicate that, unlike cytology, the three DNA methylation markers detected in urine
sediments collected in early follow-up visits can reliably predict recurrence in 80%
of patients having a history of DNA methylation positive tests compared to only 35%
by cytology and 15% by cystoscopy (Figure 2.8).

45

Figure 2.8 Three DNA methylation markers help predict the risk of recurrence
of bladder tumors in urine sediments. Percentage of urine sediments that had
positive scores (DNA methylation score calculated to be higher than cut-off values)
in the period before recurrence for recurrence-free patients (N=189) and patients who
ultimately had recurrence (N=65). Pie charts summarize all samples (left) or by
patient history (right) in the period before recurrence. A patient-level positive score
represents a history of positive DNA methylation scores at any eligible visits.
Sample-level charts report the percentage of samples from recurrence-free patients in
DNA methylation negative samples (Negative Predictive Fraction, NPF) and the
percentage of samples from patients with recurrence in DNA methylation positive
samples (Positive Predictive Fraction, PPF); Patient-level charts report the
percentage of recurrence-free patients in those without a history of positive samples
(Negative Predictive Value, NPV) and percentage of patients with recurrence in
those with a history of DNA methylation positive samples (Positive Predictive
Value, PPV). Also reported are the cytological and cystoscopic performance in these
same groups of samples/patients.

46

Figure 2.9 Three DNA methylation markers could detect the risk of recurrence
of bladder tumors in urine sediments. Percentage of urine sediments that had
positive scores at anytime for recurrence patients (N=103) and for recurrence-free
patients (N=189). Sample-level charts report the percentage of samples by DNA
methylation score (negative or positive), from patients without or with recurrence.
Patient-level charts report the percentage of patients by their history of a DNA
methylation positive score (negative/positive) having recurrence (no/yes), along with
the cytological and cystoscopic performance in these same groups of patients.

47

DISCUSSION
Markers that can be detected in urine sediments provide a non-invasive method to
test for the presence of bladder tumor cells and predisposed cell populations in the
urinary tract (Seifert et al., 2007). The Food and Drug Administration has approved
six urine markers for the detection of bladder cancer that are commercially available,
including NMP22, UroVysion, ImmunoCyt, and some new investigational urine
markers such as microsatellite alterations and gene mutations (ex: FGFR3), all of
which have shown higher sensitivity than cytology. However, to date, none of these
markers has been implemented as a routine screening method for recurrence in the
clinical setting mainly due to insufficient specificity (Kompier et al., 2010; Steiner et
al., 1997; Tilki et al., 2011).

Many studies have shown that aberrant DNA methylation of a single marker or a
combination of markers in urine sediments of patients carrying bladder cancer can
stably reflect their methylation status in bladder tumors independently of the
presence of other bladder benign conditions, thereby establishing DNA methylation
urine sediment screening as a promising non-invasive approach for bladder cancer
detection. (Chan et al., 2002; Chung et al., 2011; Costa et al., 2010; Dulaimi et al.,
2004; Friedrich et al., 2004; Lin et al., 2010; Reinert, 2012; Reinert et al., 2011).
Because of the high recurrence rate observed in TURBT patients, most studies have
focused on finding correlations between the methylation status of markers present in
the tumor or urine sediments at the time of diagnosis (prior to TURBT) and

48

recurrence (Oliveira et al., 2012). Although some of such markers showed positive
correlations with the number, size, grade and stage of primary tumors and prior
recurrence history, others did not, likely due to the variation of study population or
the sampling collection conditions (Friedrich et al., 2005; Negraes et al., 2008;
Reinert et al., 2012; Tada et al., 2002; Zhao et al., 2012). The variable correlation of
these markers with clinical outcome, the fact that only one sample was evaluated by
patient, the variety of methods used to detect methylation, and the reduced number of
control samples used in the different studies, have made it difficult to accurately
predict recurrence.

More recently, it has been proposed that longitudinal collection and testing of
urine sediments may help assess the prognostic, monitoring, and recurrence
predictive value of markers (Hoque et al., 2006; Reinert et al., 2012). Several studies
undertook this approach by using DNA methylation analysis, microsatellite markers
and a fibroblast growth factor receptor 3 (FGFR3) mutation assay (Rouprêt et al.,
2008; Zuiverloon et al., 2010). Although these markers were highly sensitive, they
displayed low specificity, in some cases comparable to that of cytology (Brems-
Eskildsen et al., 2010). A four DNA methylation marker panel provided better
specificity; however, it also displayed a high rate of false-positive results (33%)
(Zuiverloon et al., 2012). The three-marker model proposed in this study may
circumvent the specificity problem encountered in those early studies. Also, as far as
we know, we are the first group using multiple DNA methylation markers to directly

49

test risk value and monitor recurrence in serial urine samples from patients with a
history of non-invasive urothelial carcinoma. Although we found that the DNA
methylation status of HOXA9, SOX1, NPY, IRAK3, ZO2, and L1-MET was
significantly associated with recurrence with high sensitivity and specificity (all
p<0·0001), a three-marker signature that included SOX1, IRAK3, and L1-MET had
a recurrence predictive power far superior to that of cytology and cystoscopy (80%
vs. 35% vs. 15% accuracy), and therefore it could supplement visits that reveal
cytologically or cystoscopically atypical or suspicious results. In those cases in
which our markers failed to reveal early recurrence, it is possible that the low amount
of cancer cells shed into urine precluded detection, or that the quality of the DNA
required for the assay was lacking.

In addition, the three markers we identified here may also contribute to functional
changes during tumorigenesis. SRY (sex-determining region Y)-box 1 (SOX1), a
transcription factor, is involved in embryonic development and has been reported to
be hypermethylated in hepatocellular carcinoma, cervical, ovarian, and lung cancer
and thus functions as a tumor suppressor (Nelson et al., 2012; Tsao et al., 2012).
Interleukin-1 receptor-associated kinase 3 (IRAK3) shows significantly decreased
expression in various types of cancer in which it must be methylated when compared
to the normal. Our laboratory has identified IRAK3 as a key driver for cancer
survival through activating SURVIVIN (De Carvalho et al., 2012). A transcriptional
functional LINE1 element (L1-MET), a truncated transcript of MET, becomes

50

activated and expressed through hypomethylation of L1 in bladder tumor, and this
event is present across the entire bladders with cancer, suggesting L1-MET a
potentially specific marker for epigenetically altered urothelium which is thought as
premalignant (Wolff et al., 2010a). Consistent with our results, these three markers
whose DNA methylation changes - either increase or decrease - can contribute to
tumor malignancy.

The European Organization for Research and Treatment of Cancer (EORTC)
provides tables to calculate recurrence and progression scores for TURBT patients
based on the clinicopathological variables of primary tumors (Sylvester et al., 2006).
Our proposed risk score for recurrence using three-marker panel could also be used
in tandem with markers capable of predicting progression, thereby further increasing
their predictive accuracy (Kandimalla et al., 2012). However, this type of long-term
study may prove to be difficult to implement. Moreover, the fact that this study was
done with patients all of whom had a history of bladder cancer makes it uncertain
how the test would perform in a group with a very low incidence of cancer, such as
in the primary evaluation of hematuria. Further studies addressing the validation of
these urine sediment markers in a larger, multi-center patient cohort with appropriate
follow-up sample collection are required.

51

CONCLUSION
Our study provides new insights into the value of a combination of
hypermethylated and hypomethylated markers specific for tumor and epigenetic field
defect to screen urine sediments from patients that underwent bladder tumor
resections. To our knowledge, this is the first study to follow and monitor multiple
urine sediment samples over the course of many years and to incorporate both hyper-
and hypo-DNA methylation profiles during the process of bladder tumor recurrence.
This study provides evidence that a marker panel such as the one investigated here
may offer the means to minimize the frequency of cystoscopy for patients with a
negative score. We suggest that patients with a positive urinary methylation risk
score but no clinical evidence of bladder cancer disease should still be closely
monitored since they carry a high risk of recurrence.

52

CHAPTER 3

FUNCTIONAL ORGANIZATION OF TJP2 (ZO2) AT
ALTERATIVE REGULATORY REGIONS

INTRODUCTION
Tight junction protein 2 (TJP2), also named zona occludens 2 (ZO2), locates at the
cell-cell contact sites and acts as a barrier in epithelial and endothelial cells (Matter
and Balda, 2007). TJP2 proteins bind to claudins, occludins, and actin filaments; at
specific cellular regions, they are known as scaffolds or adaptors capable of
assembling molecular signaling complexes, receptors, ion channels, transporters, and
cell adhesion molecules through the PDZ domain which is a protein-protein
interaction motif (Zahraoui et al., 2000). PDZ proteins can interact with various PDZ
binding partners and then mediate diverse cellular and biological processes. Binding
to disease-associated proteins, on the other hand, can promote cancer progression.
Having the PDZ domain, TJP2 can enforce organization of multi-protein complexes
at the membrane thereby, playing an important role in cell-cell adhesion, polarization
and development (Oka et al., 2010). TJP2 also controls the nuclear localization
machinery and thus regulates different cellular signaling pathways, such as cell
proliferation and differentiation (Balda and Matter, 2009; Craven and Bredt, 1998).
In addition, it has been found that TJP2 functions as a tumor suppressor gene and can

53

target and suppress expression of tumorigenic proteins, such as adenovirus type 9
E4-ORF1 (Glaunsinger et al., 2001). As the result, deregulation of TJP2 or abnormal
interaction between its PDZ domain and other oncogenic proteins could contribute to
several human diseases, including cancer (Walsh et al., 2010).

TJP2 gene contains two alternative promoters, resulting in the presence of two
isoforms: TJP2-P1 and TJP2-P2. The upstream promoter (P1) gives rise to TJP2-P1
protein which is 23 amino acids at the N-terminus shorter than TJP2-P2. Both
isoforms were reported to be present in normal human pancreatic duct cells, but only
TJP2-P2 showed decreased expression in pancreatic adenocarcinoma (Chlenski et
al., 1999). Although such organization of TJP2 was identified in 1999, few studies
had paid attention to the functional roles of the two alternative isoform proteins,
especially the shorter isoform TJP2-P1 which is given by the upstream promoter
(P1). Whether the two alternative proteins play distinct functions in cancer, whether
they have exclusive behavior in the signaling transduction, or whether they feature
tissue specificity remain unclear.

Expression of TJP2 has shown significantly decreased in breast cancer (Chlenski
et al., 2000). Moreover, hypermethylation of the downstream promoter P2 and
inactivation of TJP2-P2 expression were found in pancreatic cancer cells. In 2010,
our group found that TJP2 at the downstream promoter region (P2) was
hypermethylated in tumors as well as across the entire tumor-bearing urothelium

54

when compared to the normal tissues from cancer-free individuals, indicating the
presence of epigenetic field defect (Wolff et al., 2010c). The epigenetically altered
normal-appearing urothelium might correlate with tumor malignancy, but the
underlying mechanisms require further studied. Epigenetic mechanisms that regulate
the activation of the two alternative TJP2 isoforms also need a deeper exploration.

55

MATERIALS AND METHODS

Cell lines and primary tumors
Normal fibroblast LD419, non-tumorigenic human urothelial UROtsa and
NK2464 cells were obtained and cultured as described previously (Wolff et al.,
2010b). T24, RT4, TCCSUP, HT1376, SCa-BER, J82 and UMUC3 cells were
obtained from the ATCC (American Type Culture Collection, Manassas, VA) and
cultured according to recommended protocols. Patient samples were obtained
through the University of Southern California/Norris Tissue Procurement Core
Resource after informed consent and Institutional Review Board approval at the
University of Southern California/Norris Comprehensive Cancer Center.

Reverse transcription and quantitative Real-Time PCR analysis
Total RNA was extracted by Trizol (Invitrogen, Carlsbad, CA) and cDNA was
prepared by M-MLV reversed transcriptase and random hexamers (Promega,
Madison, WI). The mRNA levels were measured by real-time PCR as described
(Friedman et al., 2009). The mRNA expression was normalized to human PCNA.
The RNAs of 20 human tissues were obtained from the Clontech total RNA Master
Panel (Clontech, Mountain View, CA).

DNA extraction and methylation analysis by pyrosequencing
DNA was extracted with a procedure using lysis buffer along with proteinase K

56

and phenol/chloroform. DNA was precipitated with isopropanol using glycogen as a
carrier, washed with 70% EtOH, dissolved in TE-4 buffer, and then stored at 4°C.
DNA extraction from frozen tissues has been described previously (Byun et al.,
2007). DNA was bisulfite-converted using the EZ DNA Methylation Kit (Zymo
Research, Irvine, CA, USA). The regions of interest were PCR amplified using
biotin-labeled primers (Table 3.1) and analyzed by pyrosequencing. The percentage
of methylated cytosines divided by the sum of methylated and unmethylated
cytosines was measured using PSQ HS96 Pyrosequencing System (Qiagen,
Valencia, CA, USA) as previously described (Wolff et al., 2010a).

Nucleosome Occupancy and Methylome Sequencing (NOMe-seq)

Analyses of nucleosome occupancy and endogenous methylation at regulatory
regions were determined as previously described with minor modifications (Andreu-
Vieyra et al., 2011; You et al., 2011). After nuclei extraction from 100,000 cells,
treatment of a GpC methyltransferase, M.CviPI (New England Biolabs), which
methylates GpC sites, was performed for 7.5 minutes with 200 U of the enzyme,
followed by another 7.5 minutes incubation with 100 U of M. CviPI at 37 °C.
Reactions were stopped by adding an equal volume of stop solution [20nM Tris·HCl
(pH 7.9), 600 mM NaCl, 1% SDS, 10 mM EDTA, and 200ug/mL Proteinase K] and
incubated at 55 °C for 16 hours. Then DNA was extracted by phenol/chloroform,
precipitated by ethanol and bisulfite converted by the EZ DNA Methylation Kit
(Zymo Research, Irvine, CA, USA). Regulatory regions of interest were PCR

57

amplified and cloned using the TOPO TA-cloning Kit (Invitrogen, Carlsbad, CA,
USA) according to the manufacturers’ instructions. Primers used were shown in
Table 3.1. Colonies were screened by PCR and at least 15 positive clones per
amplicon were sequenced. Sequencing results were analyzed by the BiQ analyzer
software (Bock et al., 2005).

58

Table 3.1 Primer sequences
Primer Name Sequence
Pyrosequencing
TJP2 P1
sense 5’ GGGTTTTAGTGATAAAGGTTTTTTAGG
anti-sence 5’ Biotin-ACTATCACCTACTTCCTTAAAACC
sequencing 5’
TJP2 P2
sense 5’ GGTTTTTAGATAGGATTTAAAATTTTGAG
anti-sence 5’ Biotin-CAAAACCTCACACAAACAACTTC
sequencing 5’ AGGTTTTTTTAGTT
NOMe-seq
TJP2 P2 R1
sense 5’ GGATTTAGTTATTTTAAAGGAGTT
anti-sence 5’ AAAAAAACCTCAAAATTTTAAATCC
TJP2 P2 R2
sense 5’ GGATTTAAAATTTTGAGGTTTTTTT
anti-sence 5’ AAAACCCACCCCTTTATCC
Real-time qPCR
TJP2 P1
sense 5’ TCTAGGGTGCGAAGTACCACATT
anti-sence 5’ CTTGCTCCTGGAGCTGCAG
TJP2 P2
sense 5’ AGCAGGAGCAGAAGCAGAAG
anti-sence 5’ TGGAATCCTTTTGTAGGGTCA

59

RESULTS
Both CpG island promoters of TJP2 showed tissue-specific expression patterns.
TJP2 is located at the chromosome 9q21. Two alternative promoters of TJP2 are
CpG-rich as shown in the top of figure 3.1 (obtained from the UCSC genome
browser). In order to distinguish functions of upstream promoter (P1) from
downstream promoter (P2) of TJP2, we designed real-time PCR primers for the
specific detection of mRNA expression of two forms of TJP2 (Table 3.1). The length
of transcripts for P1 is 78bp, and 183bp for P2. First, the expression levels of mRNA
were analyzed using primer sets specific to P1 and P2 in various human tissues,
including stomach, small intestine, lung, heart, placenta, spleen, uterus, kidney,
prostate, brain, brain cerebellum, fetal brain, liver, fetal liver, trachea, skeletal
muscle, thymus, and thyroid. Differential mRNA expressions of the P1 and P2 were
observed in human normal tissues (Figure 3.1). In general, P1 showed low
transcription activities in most of human tissues, whereas P2 had relatively higher
mRNA expression, with the highest in brain, liver, heart, and lung. Expression of P2
was higher than P1 in the majority of these tissues whereas expression of P1 was
higher than P2 in the small intestine, prostate, trachea and thyroid. Together, we
found that both TJP2-P1 and TJP2-P2 expressed in normal tissues at mRNA levels,
but TJP2-P2 showed the dominant activity. The results demonstrated that expression
patterns of TJP2-P1 and TJP2-P2 might be tissue-specific.

60

Expression of both transcripts of TJP2 was down-regulated in tumor samples
from cancer patients.
Next, to study whether the mRNA expression patterns of TJP2-P1 and TJP2-P2
could have a clinical relevance, I measured the expression of their mRNA levels in
normal-tumor paired samples from patients with bladder, colon and prostate cancer
(10 bladder cancer, 10 colon cancer and 4 prostate cancer patients). We found that
both TJP2 P1 and P2 mRNA levels were down-regulated in tumors in all three types
of cancer when compared to their normal counterparts (Figure 3.2). Although some
patients had opposite levels of P1 and P2 expression (for example, bladder cancer
patient 3670 showed high P1 and low P2 expression in normal counterparts, and
patient 4728 displayed low P1 and high P2 expression), it was unlikely that a
correlation appeared between the mRNA expression of P1 and P2 in the same
patients (for example, bladder cancer patient 4541 showed low P1 and low P2
expression). Overall, the results indicated that both TJP2-P1 and TJP2-P2 transcripts
were down-regulated in tumors.

DNA methylation levels of TJP2 at P2 promoters negatively correlated with
mRNA expression in bladder cancer cell lines.
For in vitro studies, we examined the mRNA expression levels of TJP2-P1 and
TJP2-P2 in three non-tumorigenic (LD419, NK2464 and UROtsa) and seven bladder
cancer cell lines (T24, RT4, TCCSUP, HT1376, Sca-BER, J82 and UMUC3).
Moreover, patterns of DNA methylation are possibly responsible to gene expression

61

levels. Therefore, we also analyzed DNA methylation status at the two alternative
promoter regions (P1 and P2) using pyrosequencing. The CpG sites (blue arrows)
detected around the upstream promoter (P1) in this study were 247bp from the
transcription start sit (TSS; red arrows) while those at the downstream promoter (P2)
were 444bp from the TSS (Figure 3.3).

The TJP2-P1 showed differential mRNA expression levels in both non-
tumorigenic (LD419, NK2464 and UROtsa) and bladder tumorigenic cell lines with
the highest expression in RT4 and HT1376, which are the bladder cancer cell lines
(Figure 3.4 bottom). In contrast to clinical samples that showed a clear decrease in
TJP2-P1 mRNA expression in tumors compared to normal paired controls (Figure
3.2), bladder cancer cell lines RT4, HT1376, T24 and Sca-BER expressed higher
TJP2-P1 than non-tumorigenic LD419, NK2464 and UROtsa cell lines. This might
be due to culturing issues or distinct modifying behaviors between various bladder
cancer cells. In the DNA methylation levels, P1 promoters were unmethylated in all
of the cell lines analyzed, except in the UMUC3 cells (Figure 3.4 top).
Hypermethylation of P1 and low mRNA expression of TJP2-P1 were observed in the
UMUC3 cells. On the other hand, the mRNA expression of TJP2-P2 was decreased
in bladder cancer cell lines compared to non-tumorigenic cell lines (Figure 3.4
bottom). And P2 promoters were fully methylated in cancer (Figure 3.4 top).
However, highly methylated of P2 was also found in non-tumorigenic NK2464.
Together, these results indicate that mRNA expression of TJP2-P2 was silenced in

62

cancer along with P2 promoters highly methylated. TJP2-P2, in general, stably
showed a negative correlation between gene expression and DNA methylation status
in both cell lines and clinical samples whereas TJP2-P1 demonstrated differential
patterns in gene expression and DNA methylation, even though in the same types of
cancer.

The TJP2 P2 at distal regulatory regions (R1) showed nucleosome occupancy in
T24 as well as in LD419 and UROtsa cells.
To investigate the functional roles of epigenetic mechanisms in gene expression at
distinct regulatory regions, we mapped histone modifications for TJP2 at P1 and P2
using the UCSC genome browser (Figure 3.5 and 3.6). It showed a H3K4me1-
enriched region accompanied with many transcription factors binding sites at non-
CpG islands upstream of P1, and a CpG islands promoter region marked with
H3K27ac, partial H3K4me1 and H3K4me3 as well as few transcription factors at P1
(Figure 3.5). For P2, the enrichments of H3K4me1 and H3K27ac that indicated
features of enhancers (Taberlay et al., 2011) were found at a region upstream of the
CpG-rich P2 promoter of TJP2 (Figure 3.6). Some putative binding sites for
transcription factors were also shown in this region. P2, especially, the CpG sites
analyzed above using pyrosequencing for the DNA methylation analysis (Figure 3.3)
are likely around the boundary of active promoter marks and active enhancer marks
(red arrow in Figure 3.6). Therefore, to determine the possibility of the presence of
putative enhancers and whether the nucleosome positioning patterns at such

63

enhancers play an important role in gene expression, we selected a TJP2-P2-
repressed bladder cancer cell line, T24, and two non-tumorigenic cell lines, LD419
and UROtsa, which showed TJP2-P2 expression for the analysis of nucleosome
occupancy and endogenous DNA methylation on TJP2 P2. The DNA methylation
patterns of the three cell lines at P2 were described by pyrosequencing (Figure 3.7).

Nucleosome occupancy and methylome sequencing (NOMe-seq), is an approach
capable of mapping nucleosomes positioning and detecting DNA methylation status
at gene regulatory regions. Using methyl-transferases to methylate nucleosome-
depleted GpCs allows for the detection of both endogenous DNA methylation and
nucleosome occupancy simultaneously in a single DNA molecule. It can provide
digital results at both CpG-rich and CpG-poor regions. So first, we used NOMe-seq
to analyze the chromatin structure at the promoter of GRP78 as a control gene for the
constitutive gene expression in T24, LD419 and UROtsa. The results confirmed that
all of the three cells displayed the nucleosome-depleted regions (NDR) at GRP78
promoter regions (Figure 3.8). To characterized whether a putative enhancer exists
upstream of the promoter of TJP2 P2 where we had analyzed the DNA methylation
by pyrosequencing (Figure 3.7), we then examined the status of nucleosome
occupancy as well as endogenous DNA methylation across the TJP2 P2 at R1, which
is located ~ 1kb and R2, which is located ~ 0.5kb upstream of the TSS of P2 in both
TJP2-P2-active and -repressed cells (Figure 3.9).

64

The results showed the presence of nucleosome occupancy at both R1 and R2
where DNA was fully methylated in TJP2-P2-repressed cells, T24, whereas an NDR
was detected around R2 in UROtsa and LD419 cells that express TJP2-P2. However,
the open chromatin structure only exhibited in 20% of R2 upstream of the TSS in
LD419 cells (Figure 3.9). Nearly 100% of R1 was occupied by nucleosomes in
UROtsa and LD419 cells, indicating that the lack of an NDR was inconsistent with
TJP2 P2 transcription, in this case of LD419 and possibly of UROtsa. In addition,
hypomethylation was observed in TJP2-P2-active cells, UROtsa and LD419, but
DNA tended to have increased methylation along the increased distance from the
TSS (Figure 3.10). Taken together, results of NOMe-seq analysis suggest
nucleosome-depleted regions (NDR) around promoter regions of TJP2 gene in cells
with TJP2-P2 transcriptional active state. However, we failed to identify clear NDRs
at a potential enhancer region because the nucleosome occupancy was shown at R1
in UROtsa and LD419 cells.

65

TJP2 (ZO2)

Figure 3.1 Both TJP2-P1 and TJP2-P2 displayed tissue-specific mRNA
expression patterens. TJP2 mRNA expression levels in different human tissue
samples were determined by quantitative real-time PCR analysis using specific
primer sets and normalized to human PCNA.

66

Figure 3.2 Two alternative promoters of TJP2 (P1 and P2) showed decreased
mRNA expression in tumor tissues from bladder, colon and prostate cancer
patients. TJP2 mRNA expression levels in normal-tumor paired samples from
patients with prostate, colon and bladder cancer were determined by quantitative
real-time PCR analysis using specific primer sets and normalized to human PCNA.

67

Figure 3.3 The graphic map of CpG sites in two alternative promoters of TJP2.
The green bars (CpG: 99 and CpG: 82) represent CpG islands at the TJP2 upstream
promoter (P1) and the downstream promoter (P2). Each vertical black line represents
CpG dinucleotides. Blue arrows represent CpG sites measured in the pyrosequencing
assay. Red arrows denote the transcriptional start sites.

68

Figure 3.4 The relationship between DNA methylation levels and mRNA
expression of TJP2 at the two alternative promoters (P1 and P2) in different cell
lines. DNA methylation and mRNA expression levels of TJP2 P1 and P2 in non-
tumorigenic (LD419, UROtsa, and NK2464) and bladder cancer cell lines (T24,
RT4, TCCSUP, HT1376, Sca-BER, J82 and UMUC3) were analyzed using specific
primer sets. The percentage of DNA methylation was measured by pyrosequencing
(top). TJP2 mRNA levels were assessed by quantitative real-time PCR and
normalized to human PCNA (bottom).

69

Figure 3.5 Functional chromatin activity of the TJP2 gene at the upstream
promoter (P1). The UCSC human genome browser maps active histone
modification marks- H3K4me1, H3K4me3, and H3K27Ac, and the binding of the
transcription factors around the regulatory regions of TJP2 P1. Red arrow indicates
the CpG sites measured in the pyrosequencing (PSQ).

70

Figure 3.6 Functional chromatin activity of the TJP2 gene at the downstream
promoter (P2). The UCSC human genome browser maps active histone
modification marks- H3K4me1, H3K4me3, and H3K27Ac, and the binding of the
transcription factors around the regulatory regions of TJP2 P2. Red arrow indicates
the CpG sites measured in the pyrosequencing (PSQ).

71

Figure 3.7 mRNA expression of TJP2 P2 inversely correlated with DNA
methylation in LD419, UROtsa, and T24 cells. DNA methylation and mRNA
expression levels of TJP2 P2 were measured in two non-tumorigenic (LD419 and
UROtsa) and one bladder cancer cell line (T24). TJP2 mRNA levels were assessed
by quantitative real-time PCR and normalized to human PCNA (left). The
percentage of DNA methylation was measured by pyrosequencing (right).

72

GRP78

Figure 3.8 Chromatin structure of GRP78 at promoter regions. NOMe-seq was
performed in T24, UROtsa and LD419 cells. Each vertical black line represents CpG
or GpC dinucleotides. Black arrows denote the transcriptional start sites. Methylation
status in each CpG and GpC dinucleotides were plotted as bubble charts. White
bubble represents unmethylation and black represents methylation (on the top of the
graphic map). Nucleosome accessible regions were shown in green and inaccessible
regions were shown in white bubbles, scaled with ~150bp pink bar (on the bottom of
the graphic map).

73

Figure 3.9 Chromatin structure of TJP2 at downstream promoter region (P2).
NOMe-seq was performed at R1 (left) and R2 (right) of TJP2 P2 in T24, UROtsa
and LD419 cells. Each vertical black line represents CpG or GpC dinucleotides.
Black arrows denote the transcriptional start sites. Methylation status in each CpG
and GpC dinucleotides were plotted as bubble charts. White bubble represents
unmethylation and black represents methylation (on the top of the graphic map).
Nucleosome accessible regions were shown in green and inaccessible regions were
shown in white bubbles, scaled with ~150bp pink bar (on the bottom of the graphic
map).

74

DISCUSSION
Nucleosome depleted regions (NDR), lacking nucleosomes around the TSSs of
active genes, have been reported as key drivers for gene transcription (You and
Jones, 2012). However, patterns of nucleosome occupancy at some regulatory
regions, such as enhancers remain unclear. Recent studies have indicated that
nucleosome positioning at DNA regulatory regions plays a critical role in controlling
the expression of pluripotent genes (You et al., 2011); without androgen treatment,
androgen receptors (AR) showed NDRs at enhancers that lead to gene activation
(Andreu-Vieyra et al., 2011); the presence of NDRs determined permissive
enhancers in polycomb-repressed genes and this is important in ESCs
reprogramming (Taberlay et al., 2011); using genome-wide mapping, DNA
methylation profiles were found not correlated with nucleosome occupancy at CTCF
regions (Kelly et al., 2012). DNA regulatory regions, like enhancers or promoters,
have differential nucleosome configurations that characterize functional activity of
genes and correlate with transcriptional activities.

It has been described that TJP2-P1 isoform is shorter than TJP2-P2 isoform with a
difference of 23 peptides and they are resulted from two alternative promoters.
Moreover, hypermethylation of P2 has been documented in most of cancer with
silenced TJP2-P2 expression. Characterization of epigenetic alterations on various
gene regulatory regions in both normal and cancer cells may reveal the functional
importance of different regulation on gene expression. Therefore, I examined mRNA

75

expression, DNA methylation and nucleosome positioning profiles especially at P2
to explore potential enhancers that could regulate gene transcription. The advantages
of NOMe-seq analysis used in our study over the MNase I digestion to map the
location of nucleosomes at gene regulatory regions are that it provides information of
endogenous DNA methylation and nucleosome position simultaneously in a single
DNA molecule and it provides digital results at both CpG-rich and CpG-poor
regions. Although we couldn’t find a putative enhancer upstream of TJP2 P2 that
might correlate with gene activity, our results showed the chromatin configurations
upstream of the TSS of TJP2 P2 in cells with or without TJP2-P2 expression. The
range we analyzed in the NOMe-seq assay was just within 1kb upstream of the TSS
of TJP2 P2. Based on studies which suggested that regions for enhancers could be 2-
4 kb in range (Taberlay et al., 2011), potential key elements for TJP2 P2 might locate
at somewhere much further.

Histone modifications, such as methylation and acylation to the tails of histone
proteins represent an important part of epigenetic features that correlate with gene
expression. A specific modification of a specific histone protein on a specific residue
(also called histone mark), alone or in combination, can regulate transcriptional
activity. To investigate specific histone modification patterns, chromatin
immunoprecipitation (ChIP) or ChIP-seq is required in selected cell lines that
characterized as active or repressed contexts of gene TJP2. Further studies, such as
genome-wide NOMe-seq or global DNA methylation analysis will be useful to

76

provide information regarding whether some undefined regulatory elements
containing specific epigenetic features could connect with their cognate promoters
through certain transcription factors, histone modifiers or nucleosome remodelers
and ultimately cooperatively regulate gene expression. It will help us answer the
questions that specific chromatin configurations at alterative regulatory regions could
affect gene transcription in a tissue–specific manner through changes in chromatin
accessibility to transcription factors, and in this case, further analysis at P1 is
necessary for the future study.

77

CHAPTER 4

MiR-30d, miR-181a, and miR-199a-5p cooperatively
suppress the endoplasmic reticulum chaperone and
signaling regulator GRP78 in cancer

INTRODUCTION
The glucose regulated protein GRP78, also referred to as BiP or HSPA5, is a
major endoplasmic reticulum (ER) chaperone and a master regulator of the unfolded
protein response (UPR) (Kahali et al., 2012; Lee, 2007; Pfaffenbach and Lee, 2011;
Ron and Walter, 2007). GRP78 can also be detected on the surface of cancer cells,
where it mediates oncogenic signals (Ni et al., 2011). The stress-mediated up-
regulation of GRP78 represents a key adaptive response for cancer cell survival (Lee,
2007). However, under severe stress conditions, apoptosis is triggered through the
induction of BAX/BAK and CHOP (Hetz et al., 2006; Ron and Walter, 2007).
Tumors are subjected to endoplasmic reticulum stress due to intrinsic metabolic
alterations and extrinsic factors in the tumor microenvironment, such as acidosis and
hypoxia, which leads to UPR activation and GRP78 induction (Biquan and Lee,
2012; Dong et al., 2011). GRP78 is up-regulated in a variety of tumors, and plays
roles in anti-apoptosis, tumor progression, angiogenesis and metastasis (Biquan and

78

Lee, 2012; Grkovic et al., 2012; Lee, 2007; Li et al., 2012; Yeung et al., 2008).
Increased GRP78 levels are associated with poor outcome and early recurrence in
prostate cancer (Pootrakul et al., 2006). Conversely, GRP78 haploinsufficiency in
mouse cancer models slows the progression of both solid tumors and hematopoietic
malignancies, and inhibits tumor neo-angiogenesis (Dong et al., 2011; Fu et al.,
2008; Wey et al., 2011). Increased GRP78 expression in cancer cells facilitates drug
resistance by suppressing apoptosis (Lee, 2007; Roue et al., 2011), whereas
knockdown of GRP78 sensitizes human cancer cells to the histone deacetylase
inhibitor trichostatin A (TSA), malignant gliomas to temozolomide, and tumor
associated endothelial cells to chemotherapeutic agents (Baumeister et al., 2009;
Pyrko et al., 2007; Virrey et al., 2008). Collectively, these studies suggest that
GRP78 is a potential therapeutic target in cancer (Backer et al., 2011; Lee, 2007).

Small regulatory RNAs, such as microRNAs (miRNAs), endogenously suppress
gene expression (Lee et al., 1993), and hundreds of miRNAs have been described
(Griffiths-Jones et al., 2006). Human miRNAs regulate several important biological
processes, including cell proliferation and differentiation, development, apoptosis
and epigenetic changes (Ambros, 2004; Krol et al., 2010), and are implicated in
diseases, such as cancer (Calin and Croce, 2006; Croce, 2009; Kappelmann et al.,
2012). Aberrant expression of miRNAs is common in various cancers and can be
caused by genomic abnormalities, miRNA processing defects, and epigenetic
alterations (Calin et al., 2004; Friedman et al., 2009; Garzon et al., 2010; Lopez-

79

Serra and Esteller, 2012; Saito et al., 2006). MiRNAs can act as oncogenes or as
tumor suppressors by targeting molecules critically involved in carcinogenesis (Calin
et al., 2002; Croce, 2009; Hu et al., 2010) and thus, they are good candidate targets
for cancer therapy (Garzon et al., 2010; Zhang et al., 2012). In addition, miRNAs
have been implicated in the stress response (Duan et al., 2012; Leung and Sharp,
2010). Since increased GRP78 expression is involved in tumor progression and
chemo-resistance, we hypothesized that miRNAs that target GRP78, which is highly
expressed in cancer, may act as tumor suppressors and may be clinically relevant. To
date, however, the specific miRNAs involved in the regulation of GRP78 have yet to
be characterized.

Here, we show that miR-30d, miR-181a, and miR-199a-5p, three miRNAs
predicted to target GRP78 in silico, are down-regulated in tumors of various origins
and in human cancer cell lines. Further, we demonstrate an inverse correlation
between miRNA and GRP78 expression levels, suggesting that these miRNAs may
regulate GRP78 and are clinically relevant. Our results indicate that the three
miRNAs directly bind to and down-regulate GRP78 levels in vitro. In cancer cells,
miRNAs showed combinatorial effects on GRP78 repression, and apoptosis
induction as well as on the reduction of cell survival and colony formation.
Importantly, the cooperation among these miRNAs increased the sensitivity of
cancer cells to drug treatment. Altogether, our results identify three specific miRNAs

80

that regulate GPR78 and provide evidence that the delivery of multiple miRNAs
holds promise for the development of novel cancer therapies.

81

MATERIALS AND METHODS
Cell lines, primary tumors and drugs
Normal fibroblast LD419, non-tumorigenic human urothelial UROtsa and
NK2464 cells were obtained and cultured as described previously (Wolff et al.,
2010). C42B cells were a gift from Dr. M Stallcup (USC, CA, USA) and were
maintained in RPMI supplemented with 10% fetal bovine serum. UM-UC-3, J82,
HCT116, HL60 cells were obtained from ATCC (Manassas, VA, USA) and cultured
according to recommended protocols. Patient samples were obtained through
University of Southern California/Norris Tissue Procurement Core Resource after
informed consent and Institutional Review Board approval at the University of
Southern California/Norris Comprehensive Cancer Center. Thapsigargin (Tg) and
trichostatin A (TSA) (Sigma-Aldrich, St Louis, MO, USA) were prepared as
previously described (Baumeister et al., 2009).

Reverse transcription and quantitative Real-Time PCR analysis
MiRNA Taqman assays (Applied Biosystems, Foster City, CA, USA) were
performed following the manufacturer’s instructions. Total RNA (15ng) was reverse
transcribed into miR- and U6-specific cDNA and miRNA expression was
normalized to U6 snRNA. The mRNA levels were measured by real-time PCR as
described (Friedman et al., 2009). Total RNA was extracted by Trizol (Invitrogen,
Carlsbad, CA, USA) and cDNA was prepared by M-MLV reversed transcriptase and

82

random hexamers (Promega, Madison, WI, USA). The miRNA expression was
normalized to human GAPDH.

Western blot assay
The same amount of total cell lysates were prepared for Western blot analysis as
described (Friedman et al., 2009). Antibodies against GRP78 (BD Pharmingen, San
Jose, CA, USA); CHOP protein, PARP-1 (Santa Cruz Biotechnology, Santa Cruz,
CA, USA); β-actin (Sigma, St Louis, MO, USA) were used.

Transfection
MiRNA precursors (Ambion, Foster City, CA, USA) were transfected into cells at
the final concentration of 50nM using Oligofectamine (Invitrogen, Carlsbad, CA,
USA) following the manufacturer’s protocol.

Generation of GRP78 3’UTR luciferase constructs
The human GRP78 3’UTR (387bp) was cloned into the XbaI site of pGL3-control
vector (Promega, Madison, WI, USA). Mutant vectors were generated using
designed mutagenic oligonucleotide primers by the QuikChange II XL site-directed
mutagenesis kit (Stratagene, La Jolla, CA, USA). Each miRNA binding site mutation
was carried out by one complete procedure of mutant synthesis.

83

Luciferase reporter assay
Luciferase reporter vectors, either wild-type or mutant (500ng), together with
pRL-SV40 (5ng) (Promega, Madison, WI, USA) were co-transfected with miRNA
precursor at the final concentration of 150nM using Lipofectamine 2000 (Invitrogen,
Carlsbad, CA, USA). Luciferase activity was measured using the Dual Luciferase
assay kit (Promega, Madison, WI, USA) according to the manufacturer’s
instructions. Firefly luciferase activity was normalized to the internal control Renilla.

Apoptosis assay
Annexin V-FITC apoptosis detection kit (Medical & Biological Laboratories,
Woburn, MA, USA) was used according to the manufacturer’s instructions. Briefly,
both attached and floating cells were collected and resuspended in binding buffer
before adding the Annexin V-FITC antibody and propidium iodide (PI). Stained cells
were analyzed by flow cytometry (Beckman Coulter, Brea, CA, USA).

Cell viability assay and colony formation assay
Cell viability assay and colony formation were described previously (Friedman et
al., 2009). Total cell numbers were counted at the indicated time points. 1000 of cells
were seeded and incubated at 37
o
C, 5% CO2 for 14 days to allow colonies to form.
Colonies were fixed in methanol, stained with 10% of Giemsa solution (Sigma, St
Louis, MO, USA) and counted.

84

Expression vectors and virus transduction
Expression vectors were made by sequentially cloning each of the precursor
miRNA into pcDNA3.1(+) (Invitrogen, Carlsbad, CA, USA). MiRNAs were then
cloned into the lentivirus vector using In-Fusion
TM
Advantage PCR cloning kit
(Clontech, Mountain View, CA, USA). Cells were infected with RNA virus stock in
the presence of polybrene (70 µg/ml). After 24h, the medium was substituted with
fresh medium containing puromycin (1.25µg/ml). After 7 days of selection, the
puromycin was removed from the medium.

In vivo cell injection with 3D gelatin-TGase
Athymic nude male mice (25-35g) were used in the study according to an
approved protocol by the Institutional Animal Care and Committee, University of
Southern California. Animals were anesthetized with ketamine/xylazine (10:1,w/w)
before the injection. Gelatin-TGase was prepared as previously described (Kuwahara
et al., 2010). About 200µl of gelatin-TGase cocktail mixed with 10
6
cells was
injected through a 27-gauge needle on the shank of the mouse. Tumors were allowed
to grow for 45 days, after which time, mice were necropsied and tumor size was
measured.

85

Table 4.1 Primers sequence
Inserts for pGL3
5'-GRP78 3’UTR-XbaI TTAATCTAGATGATCTGCTAGTGCTGTAATATTG
3'-GRP78 3’UTR-XbaI GAATTCTAGAAGTAATTGATCTAATTAGAAGCTTCTC
GRP78 3’UTR mutant
5’ miR30d CTATAGCCTAAGCGGCTCTTTAGTGCTTTTCATTAGC
AGTTGC
3’ miR30d GCAACTGCTAATGAAAAGCACTAAAGAGCCGCTTAG
GCTATAG
5’ miR181a GGAAAAAATTGAAAGAACTTAAGTCTCCAATCTAAT
TGGAATCTTCACCTC
3’ miR181a GAGGTGAAGATTCCAATTAGATTGGAGACTTAAGTT
CTTTCAATTTTTTCC
5’ miR199a CCAATAAATGTTTGTTATTTACAGTCGTCTAATGTTT
GTGAGAAGCTTC
3’ miR199a GAAGCTTCTCACAAACATTAGACGACTGTAAATAAC
AAACATTTATTGG
Expression vector
5' miR30d-kpnl TATAGGTACCATTAGCTGAAGATGATGACTG
3' miR30d-EcoRl TATAGAATTCCACATTTTATAGCCTCCTCAAC
5' miR181a-EcoRI TATAGAATTCTCGACTTGAAACCCAGAG
3' miR181a-EcoRV TATAGATATCAAAATTCACTGGACCACATTTGG
5' miR199a-EcoRV TATAGATATCGTTTCCTTGGCTGCTCAG
3' miR199a-XhoI TATACTCGAGCTCGAATCTTCTATGCGAG
Real-time PCR
5’ GRP78 AACCATACATTCAAGTTGATATTGGAGGTG
3’ GRP78 TCCCAAATAAGCCTCAGCGGTTTC

Table 4.2 Ambion pre-miR
TM
miRNA ID
Mature ID Stem-loop ID
hsa-miR-30d hsa-mir-30d
hsa-miR-181a hsa-mir-181a-2
hsa-miR-199a-5p hsa-mir-199a-1

86

RESULTS
Inverse relationship between the expression of miR-30d, miR-181a, and miR-
199a-5p and their putative target GRP78
In order to identify the specific miRNAs that target the 3’UTR of GRP78, we used
the TargetScan database (http://www.targetscan.org). Of all predicted miRNAs, we
selected three - miR-30d, miR-181a, and miR-199a-5p - that have highly conserved
seed sequences at the 3’UTR of GRP78 (Figure 4.1A). We first explore whether the
correlation of these three miRNAs and GRP78 could be clinical relevant by
measuring the expression of these three miRNAs and GRP78 mRNA levels in
normal-tumor paired samples from patients with prostate, colon and bladder cancer.
We found an inverse correlation between miRNAs and GRP78 in all three types of
tumors (p<0.05) (Figure 4.1B). In order to find optimal subjects for in vitro studies,
we next examined the endogenous expression levels of these miRNAs in non-
tumorigenic (LD419, UROtsa, and NK2464) and tumorigenic cell lines (lymphoma,
lung, cervical, bladder, colon, and prostate). Differential expression of the three
miRNAs was observed in cell lines (Figure 4.2). In general, miR-30d and miR-181a
were down-regulated in most cancer cell lines and tumor samples, whereas miR-
199a-5p maintained low expression levels in most normal and cancer cell lines, with
the exception of LD419. In contrast, clinical samples showed a clear decrease in
miR-199a-5p expression compared to normal paired controls (Figure 4.1B).

87

Figure 4.1 Relative expression of miR-30d, miR-181a, and miR-199a-5p, and
their potential target, GRP78, in human normal-tumor paired samples from
cancer patients. (A) Prediction of putative miRNAs targeting human GRP78 using
TargetScan (http://www.targetscan.org). The positions of miR-30d, miR-181a, and
miR-199a-5p base paring with the 3’UTR of GRP78 are indicated. (B) miR-30d,
miR-181a, and miR-199a-5p and GRP78 expression in normal-tumor paired samples
from patients with prostate, colon and bladder cancer were determined by
quantitative real-time PCR analysis. miRNA expression was normalized to U6
snRNA and GRP78 mRNA expression was normalized to human GAPDH. Paired t-
test was performed. *=p<0.05.

88

Figure 4.2 Endogenous expression of miR-30d, miR-181a, and miR-199a-5p in
both non-tumorigenic and tumorigenic cell lines. miRNA expression was
determined by quantitative real-time PCR analysis and normalized to U6 snRNA.
Two non-tumorigenic and four tumorigenic human cell lines used for further study
were indicated in dashed lines.

HL60

89

Next, we examined the correlation between the expression levels of miR-30d,
miR-181a, and miR-199a-5p and GRP78 protein levels. We selected a panel of two
non-tumorigenic cell lines (LD419 and UROtsa), which have high expression levels
of all three miRNAs (except miR-199a-5p in UROtsa), and four cancer cell lines,
which displayed both high (UM-UC-3, J82) and low (HCT116 and C42B) miRNA
levels. Based on the miRNA expression profile, we found that GRP78 expression
was higher in HCT116 and C42B, and quite low in other cell lines (Figure 4.3A),
suggesting a regulatory role of these miRNAs on GRP78 expression.

GRP78 expression is induced in cells following exposure to the endoplasmic
reticulum stress inducer, thapsigargin (Tg) (Baumeister et al., 2009; Pyrko et al.,
2007). To determine how the endogenous expression levels of these miRNAs
affected the cell response to Tg, we examined GRP78 protein levels 20 hours after
Tg treatment. Treatment with Tg caused a 4 fold increase in GRP78 protein levels
compared to untreated cells. We found that GRP78 induction was the highest in
HCT116 and C42B cells, which lack the expression of all three miRNAs (Figure
4.3B). In addition, GRP78 levels were significantly different between cells with up-
regulated expression of the three miRNAs (LD419, UROtsa, UMUC3, and J82), and
cells with low expression of the three miRNAs (HCT116 and C42B), and these
differences persisted even under conditions known to induce GRP78, such as Tg
treatment (p<0.05, n=3). Our results also demonstrate a similar pattern of GRP78
expression at the mRNA level (Figure 4.4). Collectively, these studies show that both

90

Figure 4.3 The relationship between expression of miR-30d, miR-181a, and
miR-199a-5p, and protein levels of GRP78 in six human cell lines. (A) GRP78
protein levels were assessed by Western blot in untreated cell lines (B) and in cells
treated with the endoplasmic reticulum stressor Thapsigargin (Tg), quantitated using
Quantity One (Bio-Rad), and plotted as a ratio to Actin levels. *=p<0.05; **=p<0.01;
***=p<0.001.

91

Figure 4.4 The relationship between expression of miR-30d, miR-181a, and
miR-199a-5p, and mRNA levels of GRP78 in six human cell lines. GRP78
mRNA levels were assessed by quantitative real-time PCR in untreated cell lines
(top) and in cells treated with the ER stressor Thapsigargin (Tg) (bottom). GRP78
mRNA expression was normalized to human GAPDH. Data are shown as the mean ±
SD (n=3). * = p<0.05.

92

basal and Tg-induced GRP78 levels inversely associate with the endogenous levels
of the three miRNAs analyzed.

MiR-30d, miR-181a, and miR-199a-5p directly target the 3’UTR of GRP78 and
significantly suppress luciferase activity cooperatively
To investigate whether GRP78 is a direct target of miR-30d, miR-181a, and miR-
199a-5p, we generated a firefly luciferase reporter vector containing the GRP78
3’UTR (Figure 4.5). The vector was transfected along with miRNA precursors into
C42B cells, since these cells lack the expression of all three miRNAs. The lysates
were analyzed for luciferase activity 24 hours post-transfection. We found mild
repression of luciferase activity by individual miRNAs, which did not reach
statistical significance, whereas the transfection with all three miRNAs resulted in
approximately a 60% decrease in activity, suggesting that they act cooperatively
(p<0.05, n=3) (Figure 4.5). However, the combination of any given two miRNAs
may also have a similar effect. To assess this possibility, we generated mutant
GRP78 luciferase constructs, carrying two base pair changes in each miRNA
putative binding site located at the GRP78 3’UTR (Figure 4.6). All three miRNA
precursors were transfected along with wild-type or mutated GRP78 vectors into
C42B cells to test the binding ability of different miRNA combinations. Mutation in
any of the three miRNA binding sequences reduced the ability of all three miRNAs
to inhibit luciferase activity, suggesting that they directly bind to the 3’UTR of
GRP78 and all three are required to achieve efficient inhibition.

93

Figure 4.5 MiR-30d, miR-181a, and miR-199a-5p directly target the 3’UTR of
GRP78 with cooperative effects. The luciferase reporter vector (pGL3) containing
the human GRP78 3’UTR was co-transfected with miR-30d, miR-181a, or miR-
199a-5p precursors into C42B cells at a final concentration of 150nM. Luciferase
activity was measured 24h after transfection and normalized to Renilla (pRL-SV40).
Data are shown as the mean ± SD of three individual experiments. * = p<0.05.
miRNC: negative control.

*

94

Figure 4.6 MiR-30d, miR-181a, and miR-199a-5p directly target the 3’UTR of
GRP78 with cooperative effects. Individual miRNA binding sites (Luc-m1, Luc-
m2, and Luc-m3) or all three miRNA binding sites (Luc-m123) were mutated in the
GRP78 3’UTR vector. Luciferase reporter vectors, either wild-type (WT) or mutant
(m1, m2, m3, and m123) were co-transfected with either miRNC or all three miRNA
precursors into C42B cells at a final concentration of 150nM. Luciferase activity was
measured 24h after transfection and normalized to Renilla (pRL-SV40). Data are
shown as the mean ± SD of three individual experiments. * = p<0.05. miRNC:
negative control.

*

mutant WT

95

GRP78 down-regulation requires cooperation of multiple miRNAs and leads to
morphological changes and apoptosis in C42B cells
GRP78 is a stable protein present in cancer cell lines and miRNAs target mRNAs
at their 3’UTR region. To further verify the ability of miR-30d, miR-181a and miR-
199a-5p to down-regulate GRP78, C42B cells were transfected with miRNA
precursors and treated 24 hours after transfection with Tg to induce GRP78
expression. MiRNA expression and GRP78 mRNA levels were confirmed by
quantitative PCR analysis 20 hours post-treatment, an optimal time-point for
endoplasmic reticulum chaperon induction, including GRP78, in C42B cells (data
not shown). As expected, GRP78 mRNA and protein levels significantly increased
after Tg treatment in the control samples (miRNC) (p<0.05, n=3) (Figure 4.7 and
4.8). A small, non-significant decrease in GRP78 mRNA levels was found in cells
treated with Tg and transfected with each individual miRNAs, compared to controls
(miRNC). In contrast, when the three miRNAs were transfected together, we
observed a significant decrease in GRP78 mRNA levels (p<0.05, n=3) (Figure 4.7),
which is consistent with the findings of the luciferase experiments described above
(Figure 4.5). Similarly, a 50% reduction in GRP78 protein levels was observed only
when the three miRNAs were transfected together, suggesting again that they act
cooperatively (Figure 4.8).

96

Figure 4.7 Inhibition of GRP78 mRNA levels in C42B cells by co-expression of
multiple miRNAs. GRP78 mRNA levels were determined by quantitative real-time
PCR in cells transfected with miRNA precursors at a final concentration of 50nM.
GRP78 mRNA expression was normalized to human GAPDH. Data are shown as the
mean ± SD (n=3). * = p<0.05. miRNC: negative control.

97

Figure 4.8 Inhibition of GRP78 protein levels in C42B cells by co-expression of
multiple miRNAs. GRP78 protein levels were determined by Western blot,
respectively, in cells transfected with miRNA precursors at a final concentration of
50nM. GRP78 protein level was normalized to human Actin. Data are shown as the
mean ± SD (n=3). * = p<0.05. miRNC: negative control.

*
Ctrl
Tg

98

Since over-expression of GRP78 in cancer cells can inhibit apoptosis (Cook et al.,
2012; Lee, 2007; Yeung et al., 2008), we next evaluated whether the modulation of
GRP78 levels by miRNAs affects apoptosis in C42B cells treated with Tg. In control
samples, which show high GRP78 levels, the protein levels of the apoptotic marker
PARP-1, and CHOP, a UPR target that is induced in response to Tg, were low. In
contrast, when GRP78 levels were decreased by transfection of miR-30d, miR-181a
and miR-199a-5p, PARP-1 and CHOP levels were significantly up-regulated
(p<0.05, n=3) (Figure 4.9). In addition, C42B cells transfected with miRNAs
underwent morphological changes and became rounded. Treatment with Tg resulted
in the formation of small vesicles that resemble apoptotic bodies and a decrease in
cell numbers (Figure 4.10A and B). We further confirmed the induction of apoptosis
by FACS, using AnnexinV and propidium iodide staining. The results showed a
significant increase (p<0.05, n=3) in apoptosis in cells transfected with all three
miRNAs, but not with miRNC at 24, 48 and 72 hours post-transfection. Treatment
with Tg significantly increased the apoptotic response in cells transfected with all
three miRNAs, compared to miRNC (38.2% ± 2.1 vs. 12.8% ± 1.6 at 48h; 60.8% ±
1.3 vs. 33.9% ± 2.8 at 72h) (p<0.05, n=3) (Figure 4.10C). Altogether, the results
show that a combination of miRNAs efficiently sensitizes C42B cells to apoptosis by
reducing GRP78 levels.

99

Figure 4.9 Inhibition of GRP78 levels and induction of apoptosis in C42B cells
by co-expression of multiple miRNAs. Multiple miRNA expression induced a
decrease in GRP78 protein levels and increases in the levels of CHOP, a UPR
indicator, and the apoptosis indicator PARP-1. Protein levels were normalized to
Actin. Data are shown as the mean ± SD (n=3). * = p<0.05.

*

*

*

100

Figure 4.10 Inhibition of GRP78 levels and induction of apoptosis in C42B cells
by co-expression of multiple miRNAs. (A) Representative micrographs showing
apoptotic cells (arrowhead) in cells transfected with all three miRNAs with or
without Tg treatment. (B) Total cell numbers were counted 4 days post-transfection.
Data represents mean ± SD (n=3). *=p<0.05. (C) Attached and floating cells were
both collected at the indicated time-points and stained with Annexin V-FITC and
propidium iodide (PI). The percentage of Annexin V-FITC positive (early apoptotic)
cells was determined by FACS. Data are shown as the mean ± SD (n=3). *=p<0.05;
**=p<0.001; ***=p<0.00001.

A
B C

101

MiR-30d, miR-181a, miR-199a-5p increase the sensitivity of cancer cells to the
HDAC inhibitor TSA
It has been demonstrated that increased GRP78 expression confers resistance to
the HDAC inhibitor Trichostatin A (TSA), thereby decreasing its therapeutic
efficacy (Baumeister et al., 2009). Therefore, we next examined whether down-
regulation of GRP78 by miRNAs in C24B cells modifies their response to TSA. We
found that TSA treatment induced morphological changes in C24B cells transfected
with all three miRNAs, but not with control (miRNC) (Figure 4.11, upper panel).
TSA induced GRP78 protein levels in cells transfected with miRNC, and this
increase was suppressed by transfection of miR-30d, miR-181a, and miR-199a-5p,
which also resulted in increased expression of the apoptotic marker PARP-1 (Figure
4.11 lower panel).

As shown above, transfection of all three miRNAs caused a significant decrease in
cell numbers (Figure 4.10B), which can also be observed in cells transfected with
siRNAs against GRP78 (p<0.05, n=3). TSA treatment of cells transfected with
miRNC caused a significant decrease in cell numbers (p<0.05, n=3). Importantly,
TSA caused further decreases in cell numbers in cells transfected with the three
miRNAs (p<0.05, n=3) (Figure 4.12). In addition, colony formation assays showed
that transfection with the three miRNAs significantly inhibited colony formation in
the presence or absence of TSA, in levels comparable to those of cells transfected
with siGRP78 (p<0.05, n=3) (Figure 4.13). The results indicate that restoring the

102

Figure 4.11 Expression of miR-30d, miR-181a, and miR-199a-5p increases TSA
sensitivity in C42B cells. C42B cells were transfected with miR-30d, miR-181a and
miR-199a-5p precursors and with a negative (miRNC) and positive (siRNA against
GRP78, siGRP78) controls. 24 hours post-transfection, C42B cells were treated with
vehicle (Ctrl), Thapsigargin (Tg), or the histone deacetylase inhibitor (TSA).
Representative micrographs showing changes in cell morphology (top). GRP78 and
PARP-1 protein levels were assessed by Western blot analysis. Actin levels were
used as loading control.

103

Figure 4.12 Expression of miR-30d, miR-181a, and miR-199a-5p increases TSA
sensitivity and reduces cell viability in C42B cells. C42B cells were transfected
with miR-30d, miR-181a and miR-199a-5p precursors and with a negative (miRNC)
and positive (siRNA against GRP78, siGRP78) controls. 24 hours post-transfection,
C42B cells were treated with vehicle (Ctrl), Thapsigargin (Tg), or the histone
deacetylase inhibitor (TSA). Total cell numbers were counted 4 days post-
transfection. Data represents mean ± SD (n=3). *=p<0.05.

Ctrl
TSA

(10
5
)
*
*
*

104

Figure 4.13 Expression of miR-30d, miR-181a, and miR-199a-5p increases TSA
sensitivity and reduces cell colony formation in C42B cells. C42B cells were
transfected with miR-30d, miR-181a and miR-199a-5p precursors and with a
negative (miRNC) and positive (siRNA against GRP78, siGRP78) controls. 24 hours
post-transfection, C42B cells were treated with vehicle (Ctrl), Thapsigargin (Tg), or
the histone deacetylase inhibitor (TSA). Colony formation was assessed 14 days
after transfection. Data are shown as the mean ± SD (n=3). *= p<0.05.

Ctrl
TSA

*
*
*

105

expression of multiple miRNAs that target GRP78 can sensitize cancer cells to
therapeutic epigenetic agents.

Lentiviral delivery of multiple co-transcribed miRNAs decreases GRP78
protein levels and cell viability and induces apoptosis in different cancer cell
lines
In order to improve the delivery efficiency of multiple miRNAs and to evaluate
the long-term effects of their expression, we generated a lentiviral expression vector
containing the combination of miR-30d, miR-181a, and miR-199a-5p (Figure
4.14A), transduced C42B, HCT116 and HL60 cells, and confirmed expression of
each miRNA (Figure 4.14B) and GRP78 protein levels (Figure 4.15) in transduced
cells. Transduction of the lentivector carrying multiple miRNAs caused a significant
decrease in GRP78 protein levels and cell viability, and an increase in PARP-1
protein levels in C42B cells treated with Tg and TSA (p<0.05, n=3) (Figure 4.16A,
B, and C, and Figure 4.17), as shown in transient transfections. Tg-treated HCT116
and HL60 transduced cells showed a decrease and an increase, but not significant in
GRP78 and PARP-1 protein levels, respectively, whereas a significant decrease in
GRP78 protein levels was observed in TSA-treated HCT116 cells (p<0.05, n=3)
(Figure 4.16A and Figure 4.17). Cell viability was significantly reduced in Tg-
treated HL60 cells and in both HCT116 and HL60 cells treated with TSA (p<0.05,
n=3) (Figure 4.16B). Finally, Tg-treated HL60 and HCT116 cells transduced with

106

three miRNAs showed no changes in PARP-1 levels, whereas a non-significant
increase in PARP-1 was observed in TSA-treated cells (Figure 4.16C and Figure
4.17). In addition, decreased colony formation was observed in cells transduced with
the three miRNAs (Figure 4.18).

To further confirm the specificity of the three miRNAs, we transfected the coding
sequence of GRP78 lacking its 3’UTR into C42B cells stably expressing miR-30d,
miR-181a, and miR-199a-5p. Our data showed that, after transfection, GRP78
expression was restored at both mRNA and protein levels, and that the inhibitory
effects of the three miRNAs on cell survival and colony formation were abrogated,
suggesting that miR-30d, miR-181a, and miR-199a-5p require the 3’UTR region of
GRP78 to exert their actions (Figure 4.19-4.22). Their binding is lost upon
overexpressing GRP78 lacking the 3’UTR sequence which is consistence with
mutation of their binding sites in the luciferase assay (Figure 4.6). Taken together,
these results indicate that the cooperative effect of multiple miRNAs that target
GRP78 is specific and maintained in a stable expression system. The results also
suggest that the magnitude of the overall effect of decreasing GRP78 levels is cell-
type-specific.

107

Figure 4.14 Lentiviral vectors contain multiple co-transcribed miRNAs. (A)
Expression vectors were generated by cloning miR-30d, miR-181a, and miR-199a-
5p into lentiviral vectors. (B) Expression of miR-30d, miR-181a, and miR-199a-5p
was measured by quantitative real-time PCR and normalized to U6 snRNA. Data are
shown as the mean ± SD.

LV LV miR 30+181+199
A
B

108

Figure 4.15 Lentiviral delivery of multiple co-transcribed miRNAs down-
regulates GRP78 protein levels. GRP78 and PARP-1 protein levels were
determined by Western blot in C42B cells infected with lentivirus vectors only (LV)
or containing multiple miRNAs (LV miR30+181+199). Protein levels were
normalized to Actin.

109

Figure 4.16 Lentiviral delivery of multiple co-transcribed miRNAs down-
regulates GRP78 protein levels, reduces cell viability and induces apoptosis in
different cancer cell lines. C42B, HCT116 and HL60 cells infected with LV or LV
miR30+181+199 were treated with Tg, or the histone deacetylase inhibitor (TSA).
(A) GRP78 and (C) PARP-1 protein levels were determined by Western blot. Protein
levels were normalized to Actin. (B) Total cell numbers were counted 4 days after
the treatment. Data represents mean ± SD (n=3). *=p<0.05.

110

Figure 4.17 Stable expression of multiple miRNAs results in down-regulation
of GRP78 and up-regulation of PARP-1 protein levels in different cancer cell
lines. GRP78 and PARP-1 protein levels were assessed by Western blot in C42B,
HCT116, and HL60 cells infected with LV only or LV miR30+181+199, and treated
with Tg or the histone deacetylase inhibitor (TSA). Protein levels were normalized to
Actin.

111

Figure 4.18 Stable expression of multiple miRNAs results in inhibited colony
formation in different cells. Colony formation was assessed in cells infected with
lentivirus vectors only (LV), or containing multiple miRNAs (LV miR30+181+199)
14 days after the treatment of vehicle, or the histone deacetylase inhibitor (TSA).
Data are shown as the mean ± SD (n=3). * = p<0.05.

112

Figure 4.19 Inhibition of GRP78 mRNA levels in multiple miRNAs transduced
(LV miR30+181+199) C42B cells were abrogated after GRP78 overexpression.
GRP78 coding region lacking its 3’UTR was cloned into pcDNA3.1 and transfected
into cells stably expressing miR30d, miR181a, and miR199a-5p. GRP78 mRNA
levels was determined 48h after GRP78 overexpression by quantitative real-time
PCR, respectively, in cells treated with the histone deacetylase inhibitor (TSA).
GRP78 mRNA expression was normalized to human GAPDH. Data are shown as the
mean ± SD (n=3). * = p<0.05.

113

Figure 4.20 Inhibition of GRP78 protein levels in multiple miRNAs transduced
(LV miR30+181+199) C42B cells were abrogated after GRP78 overexpression.
GRP78 coding region lacking its 3’UTR was cloned into pcDNA3.1 and transfected
into cells stably expressing miR30d, miR181a, and miR199a-5p. GRP78 protein
levels were determined 48h after GRP78 overexpression by Western blot,
respectively, in cells treated with Tg or the histone deacetylase inhibitor (TSA).
GRP78 protein levels were normalized to Actin. Data are shown as the mean ± SD
(n=3). * = p<0.05.

114

Figure 4.21 Cell viability in multiple miRNAs transduced (LV miR30+181+199)
C42B cells were abrogated after GRP78 overexpression. Total cell numbers
were counted 4 days after TSA treatment. Data represents mean ± SD (n=3).
*=p<0.05.

115

Figure 4.22 Colony formation in multiple miRNAs transduced (LV
miR30+181+199) C42B cells were abrogated after GRP78 overexpression.
GRP78 coding region lacking its 3’UTR was cloned into pcDNA3.1 and transfected
into cells stably expressing miR30d, miR181a, and miR199a-5p. Colony formation
was assessed 14 days after GRP78 overexpression. Data are shown as the mean ± SD
(n=3). *= p<0.05.

116

MiR-30d, miR-181a, and miR-199a-5p inhibit tumor growth in vivo
To investigate the potential anti-tumor activity of these three miRNAs in vivo,
cells stably expressing miR-30d, miR-181a and miR-199a-5p were cultured in a
three-dimensional matrix (transglutaminase-gelatin gel) to mimic the tumor
environment and then injected subcutaneously into athymic nude male mice (n=6 in
each control and experimental group). Tumors were allowed to grow for 45 days.
HCT116 successfully formed tumors, whereas HCT116 expressing miR-30d, miR-
181a and miR-199a-5p showed tumors that were two to three fold smaller in size
than those of control cells, suggesting that the three miRNAs inhibit tumorigenesis in
vivo (Figure 4.23 A and B). In mice injected with C42B cells, angiogenesis was
observed but no tumor formation was detected. After necropsy, we found that some
cells were still trapped in the gel matrix (data not shown). A possible explanation for
these findings is that C42B cells may require longer time for tumor formation when
injected subcutaneously (Zhang et al., 2001). C42B cells stably expressing the three
miRNAs did not form tumors. Small nodules, and cells trapped in the gel matrix
were observed in mice injected with HL-60 cells, whereas no nodules were observed
in HL-60 cells stably expressing miR-30d, miR-181a and miR-199a-5p (data not
shown). Since HL-60 was derived from a liquid tumor, we cannot rule out the
possibility that cell may have migrated to the lymph nodes. Further studies will help
clarify this point.

117

Figure 4.23 MiR-30d, miR-181a, and miR-199a-5p inhibit tumor growth in
athymic mice. HCT116 cells were cultured in the three-dimensional
transglutaminase-gelatin matrix and injected subcutaneously (1X10
6
cells)

into
athymic nude male mice (n=6 in each control and experimental group). Tumors were
allowed to grow for 45 days after which time, mice were necropsied. Tumors were
photographed (A) and tumor sized was evaluated (B). Data are shown as the mean ±
SD (n=6).

A
B
HCT116

118

DISCUSSION
GRP78 up-regulation in various tumor types and its induction after drug treatment
has been shown to be a major contributor to tumorigenesis and therapeutic resistance
(Booth et al., 2012; Lee, 2007). Despite advances in our understanding of GRP78
actions and its induction by endoplasmic reticulum stress, little is known about
endogenous inhibitors controlling its expression other than repression of its basal
expression by histone deacetylases (Baumeister et al., 2009). Thus, discovery of such
regulatory moieties will be important for designing more efficacious therapeutic
approaches. In our study, we identify miR-30d, miR-181a, and miR-199a-5p as
regulatory small RNAs that act cooperatively to control GRP78 levels. This effect
cannot be attributed to the miRNA dose, since the same amount of total miRNA was
used irrespective of whether they were transfected alone or in combination. Because
the reduction in GRP78 levels occurs at both mRNA and protein levels, it is likely
that the three miRNAs used in our study act through GRP78 mRNA destabilization
and not translational repression (Guo et al., 2010). The fact that miR-30d, miR-181a,
and miR-199a-5p are down-regulated and GRP78 is up-regulated in samples of
prostate, colon, and bladder cancer patients, suggests that these miRNAs are
clinically relevant and that all three miRNAs are required to suppress GRP78
(Daneshmand et al., 2007). Roue et al. demonstrated that inhibition of GRP78 using
siRNA can overcome drug resistance in mantle cell lymphoma (Roue et al., 2011). In
our study, we show that decreased levels of GRP78 achieved either by transient
transfection or transduction of multiple miRNAs can also increase the sensitivity of

119

cancer cells to TSA treatment, resulting in the induction of apoptosis, and inhibition
of cell growth and colony formation. Therefore, the use of multiple specific miRNAs
to target key genes involved in tumorigenesis could provide an exciting avenue for
the development of new cancer therapies. Our results also suggest that co-expression
of multiple miRNAs from a single lentiviral vector platform (Qiu et al., 2011) is an
efficient method to deliver and test the putative combinatorial actions of miRNAs on
a single target gene. Moreover, in vivo data show that the tumor size of the HCT116
stably expressing miR-30d, miR-181a and miR-199a-5p is significantly smaller than
that of control cells, consistent with the notion that these miRNAs suppress GRP78
expression, leading to inhibition of tumor growth. However, the effect of the
miRNAs on their targets within the tumor microenvironment requires further
investigation.

The critical role of miRNAs in cancer and their involvement in common cellular
pathways make them valuable and comprehensive targets (Garzon et al., 2010). In
addition to targeting GRP78, miR-30d can act as a negative regulator for p53 and
regulate cell cycle arrest and apoptosis (Kumar et al., 2011), while down-regulation
of miR-199a-5p in human hepatocellular carcinoma (HCC) is highly associated with
cell invasion (Shen et al., 2010). On the other hand, it is well-established that a
single mRNA molecule can be targeted by multiple miRNAs, and studies focusing
on the combinatorial actions of miRNAs have begun to emerge (Lewis et al., 2005;
Lim et al., 2005; Marasa et al., 2009; Mavrakis et al., 2011b). For instance, multiple

120

miRNAs have been recently shown to regulate PTEN expression in T-cell
lymphoblastic leukemia cells (Mavrakis et al., 2011b). Such studies indicate that
gene expression is tightly controlled by miRNA networks (Mavrakis et al., 2011a).
The results presented here suggest that the combined action of multiple miRNAs
might be essential to achieve efficient down-regulation of GRP78. If this holds true
for other genes, it may help explain why the validation of the majority of miRNA
targets found using prediction algorithms, which is based on single miRNA-based
experiments, has been unsuccessful (Bartel, 2009; Selbach et al., 2008).

CONCLUSION
We report that miR-30d, miR-181a, and miR-199a-5p regulate GRP78 and that
their decreased expression in tumor cells results in increased GRP78 levels, which in
turn promotes tumorigenesis and therapeutic resistance. To the best of our
knowledge, this is the first report identifying the specific miRNAs that repress
GRP78 and their combinatorial regulatory action. Notably, our results suggest that
the use of miRNAs and TSA or other therapeutic agents in combination therapy may
provide a powerful approach in the treatment of GRP78-overexpressing and drug
resistant tumors.

121

CHAPTER 5

SUMMARY AND CONCLUSIONS

Epigenetic therapy has emerged as one of the hottest fields in cancer research
today, due in part for its possible reversal of epigenetic abnormalities and role as a
hallmark in cancer assessment. Among many promising areas and applications of
epigenetics research, DNA methylation markers and tumor suppressor microRNAs
have been widely studied because of their potential role in serving as biomarkers and
providing therapeutic approaches for cancer (Grkovic et al., 2012; Jones, 2002; Li et
al., 2012; Luo and Lee, 2012). DNA methylation and microRNAs as markers to
detect, diagnose, and monitor the aggressive or malignant behavior of cancer, as well
as anti-cancer tools to overcome therapeutic resistance, are critical and worth further
researches and discussion (Baylin and Jones, 2011; Cook et al., 2012; Kahali et al.,
2012; Laird, 2003)

Non-muscle invasive bladder cancer, characterized by a high rate of recurrence, is
a relatively high-cost disease in cancer management. Urine markers are useful and
hold potential for recurrence surveillance due to their easy accessibility. Given their
stable, reliable and early appearance, DNA methylation changes have shown their
importance in bladder carcinogenesis. While most studies of DNA methylation

122

primarily address the potential utility of markers in urine of patients carrying bladder
tumors, we provide new insights into the value of a combination of hypermethylated
and hypomethylated markers, including a transcription factor (SOX1), a specific
LINE1 element (L1-MET), and a driver gene (IRAK3) to screen urine sediments for
recurrence from patients that underwent bladder tumor resections. Our group
identified IRAK3 as a key epigenetic driver for cancer survival and published it in
Cancer Cell last year (De Carvalho et al., 2012). Of six markers I examined in
chapter 2, IRAK3, importantly, was involved in our three-marker model and showed
not only an ability to detect recurrence, but also predictive value in the return of
tumor malignancy. We showed that, for the first time, the translational implication of
using a functional epigenetic driver could represent a novel approach in the
surveillance of bladder tumor recurrence in patients’ urine with high sensitivity and
specificity.

Despite advances in our understanding of markers for the detection of tumor
recurrence, little is know about personalized following and monitoring of multiple
urine sediment samples over the course of many years. The study in chapter 2 is
longitudinal; I analyzed DNA methylation changes in urine sediments serially
collected from TURBT patients at the time of follow-up visits. Discovery of DNA
methylation markers for the long-term surveillance will be important for the early
detection of unrecognized or unanticipated tumor recurrence. To our knowledge, this
is the first study to incorporate both hyper- and hypo-DNA methylation profiles for

123

monitoring bladder tumor recurrence. These markers I showed in chapter 2 may be
useful in guiding treatment direction and in avoiding unnecessarily invasive exams
such as cystoscopies.

In addition to clinical study, the underlying mechanisms of these markers in
epigenetic regulation are also important to explore. In chapter 3, the functional roles
of TJP2 (ZO2) at the alterative promoters were described at both gene expression
and epigenetic levels. In order to functionally distinguish two promoters from one
another, mRNA expression and DNA methylation levels were analyzed in both
human normal and cancer cell lines as well as in clinical samples from patients with
bladder, prostate and colon cancer. TJP2- P2 displayed down-regulated in tumors
compared to the normal counterparts and exhibited negative correlation between
levels of mRNA expression and DNA methylation whereas TJP2-P1 showed
differential patterns among various cell lines. Moreover, endogenous DNA
methylation and nucleosome positioning profiles at different regulatory regions
around P2 were studied in the TJP2-P2-active and TJP2-P2-repressed cell lines using
NOMe-seq analysis. Our results present here showed distinct nucleosome
configurations at different DNA regulatory regions in cells with the presence or
absence of TJP2-P2 activity. However, genome-wide studies to map epigenetic
signatures of TJP2 across two alterative promoter regions (P1 and P2) are required.

GRP78, an ER stress inducible chaperone and a master regulator, is well

124

established to be critical for the maintenance of ER function and homeostasis.
GRP78 up-regulation in various tumor types and its induction after drug treatment
has been shown to be a major contributor to tumorigenesis and therapeutic
resistance. Thus, GRP78 has been increasingly recognized as a novel target for anti-
cancer therapy (Fu et al., 2008; Lee, 2007; Yeung et al., 2008). Despite advances in
our understanding of GRP78 actions and its induction by ER stress, little is known
about endogenous inhibitors controlling its expression. Thus, discovery of such
regulatory moieties and their mechanisms of action will be important for designing
more efficacious therapeutic approaches. Many laboratories, including ours, have
shown that microRNAs endogenously regulate gene expression and, hence, gene
function and signaling pathways (Friedman et al., 2009; Saito et al., 2006). Most
studies on microRNA, however, primarily address the function and the biological
effects of a single microRNA. In chapter 4, I found that the expression of GRP78 at
the mRNA and protein levels can be directly and potently modulated by the
coordinate expression of multiple GRP78-targeting microRNAs. We showed that the
cooperative effect of multiple microRNAs might be essential for the repression of
some genes, in this case, GRP78. To our knowledge, this is the first report on the
identification of specific microRNAs that target human GRP78.

Furthermore, our observation that these microRNAs levels are inversely
associated with GRP78 expression in normal-tumor paired samples from patients
with prostate, colon and bladder tumors directs clinical relevance. The translational

125

implication of this work is that the use of multiple microRNAs to suppress GPR78
could represent a novel anti-cancer approach in inhibiting the survival of tumor cells
overexpressing GRP78 that are resistant to therapeutic treatment. In chapter 4, we
showed that microRNAs can be used in combination with other targeted
chemotherapy to enhance treatment efficacy.

In summary, my study presented in this thesis would be interesting and exciting
findings to basic, translational and clinical researchers working on DNA methylation
markers discovery, prediction markers for cancer recurrence, the mammalian stress
response, molecular chaperones, functional microRNAs, tumor malignancy,
experimental therapeutics and combination drug therapy. Epigenetic study has
important applications in tumor biology and cancer detection and is likely to change
clinical practice and thinking about cancer therapy.

126

REFERENCES
Allis, C.D., Berger, S.L., Cote, J., Dent, S., Jenuwien, T., Kouzarides, T., Pillus, L.,
Reinberg, D., Shi, Y., Shiekhattar, R., et al. (2007). New nomenclature for
chromatin-modifying enzymes. Cell 131, 633-636.

Ambros, V. (2004). The functions of animal microRNAs. Nature 431, 350-355.

Andreu-Vieyra, C., Lai, J., Berman, B.P., Frenkel, B., Jia, L., Jones, P.A., and
Coetzee, G.A. (2011). Dynamic nucleosome-depleted regions at androgen receptor
enhancers in the absence of ligand in prostate cancer cells. Mol Cell Biol 31, 4648-
4662.

Babjuk, M., Oosterlinck, W., Sylvester, R., Kaasinen, E., Böhle, A., Palou-Redorta,
J., Rouprêt, M., and (EAU), E.A.o.U. (2011). EAU guidelines on non-muscle-
invasive urothelial carcinoma of the bladder, the 2011 update. Eur Urol 59, 997-
1008.

Backer, M.V., Backer, J.M., and Chinnaiyan, P. (2011). Targeting the unfolded
protein response in cancer therapy. Methods Enzymol 491, 37-56.

Bartel, D.P. (2009). MicroRNAs: target recognition and regulatory functions. Cell
136, 215-233.

Balda, M.S., and Matter, K. (2009). Tight junctions and the regulation of gene
expression. Biochim Biophys Acta 1788, 761-767.

Bartel, D.P. (2009). MicroRNAs: target recognition and regulatory functions. Cell
136, 215-233.

127

Baumeister, P., Dong, D., Fu, Y., and Lee, A.S. (2009). Transcriptional induction of
GRP78/BiP by histone deacetylase inhibitors and resistance to histone deacetylase
inhibitor-induced apoptosis. Mol Cancer Ther 8, 1086-1094.

Baylin, S.B., and Jones, P.A. (2011). A decade of exploring the cancer epigenome -
biological and translational implications. Nat Rev Cancer 11, 726-734.

Bernstein, B.E., Mikkelsen, T.S., Xie, X., Kamal, M., Huebert, D.J., Cuff, J., Fry, B.,
Meissner, A., Wernig, M., Plath, K., et al. (2006). A bivalent chromatin structure
marks key developmental genes in embryonic stem cells. Cell 125, 315-326.

Biquan, L., and Lee, A.S. (2012). The critical roles of endoplasmic reticulum
chaperones and unfolded protein response in tumorigenesis and anti-cancer therapies
(Oncogene).

Bock, C., Reither, S., Mikeska, T., Paulsen, M., Walter, J., and Lengauer, T. (2005).
BiQ Analyzer: visualization and quality control for DNA methylation data from
bisulfite sequencing. Bioinformatics 21, 4067-4068.

Booth, L., Cazanave, S.C., Hamed, H.A., Yacoub, A., Ogretmen, B., Chen, C.S.,
Grant, S., and Dent, P. (2012). OSU-03012 suppresses GRP78/BiP expression that
causes PERK-dependent increases in tumor cell killing. Cancer Biol Ther 13, 224-
236.

Brems-Eskildsen, A.S., Zieger, K., Toldbod, H., Holcomb, C., Higuchi, R., Mansilla,
F., Munksgaard, P.P., Borre, M., Ørntoft, T.F., and Dyrskjøt, L. (2010). Prediction
and diagnosis of bladder cancer recurrence based on urinary content of hTERT,
SENP1, PPP1CA, and MCM5 transcripts. BMC Cancer 10, 646.

128

Brock, M.V., Hooker, C.M., Ota-Machida, E., Han, Y., Guo, M., Ames, S.,
Glöckner, S., Piantadosi, S., Gabrielson, E., Pridham, G., et al. (2008). DNA
methylation markers and early recurrence in stage I lung cancer. N Engl J Med 358,
1118-1128.

Calin, G.A., Sevignani, C., Dumitru, C.D., Hyslop, T., Noch, E., Yendamuri, S.,
Shimizu, M., Rattan, S., Bullrich, F., Negrini, M., et al. (2004). Human microRNA
genes are frequently located at fragile sites and genomic regions involved in cancers.
Proc Natl Acad Sci U S A 101, 2999-3004.

Cairns, P., Esteller, M., Herman, J.G., Schoenberg, M., Jeronimo, C., Sanchez-
Cespedes, M., Chow, N.H., Grasso, M., Wu, L., Westra, W.B., et al. (2001).
Molecular detection of prostate cancer in urine by GSTP1 hypermethylation. Clin
Cancer Res 7, 2727-2730.

Calin, G.A., and Croce, C.M. (2006). MicroRNA signatures in human cancers. Nat
Rev Cancer 6, 857-866.

Calin, G.A., Dumitru, C.D., Shimizu, M., Bichi, R., Zupo, S., Noch, E., Aldler, H.,
Rattan, S., Keating, M., Rai, K., et al. (2002). Frequent deletions and down-
regulation of micro- RNA genes miR15 and miR16 at 13q14 in chronic lymphocytic
leukemia. Proc Natl Acad Sci U S A 99, 15524-15529.

Chan, M.W., Chan, L.W., Tang, N.L., Tong, J.H., Lo, K.W., Lee, T.L., Cheung,
H.Y., Wong, W.S., Chan, P.S., Lai, F.M., et al. (2002). Hypermethylation of
multiple genes in tumor tissues and voided urine in urinary bladder cancer patients.
Clin Cancer Res 8, 464-470.

129

Chlenski, A., Ketels, K.V., Engeriser, J.L., Talamonti, M.S., Tsao, M.S.,
Koutnikova, H., Oyasu, R., and Scarpelli, D.G. (1999). zo-2 gene alternative
promoters in normal and neoplastic human pancreatic duct cells. Int J Cancer 83,
349-358.

Chlenski, A., Ketels, K.V., Korovaitseva, G.I., Talamonti, M.S., Oyasu, R., and
Scarpelli, D.G. (2000). Organization and expression of the human zo-2 gene (tjp-2)
in normal and neoplastic tissues. Biochim Biophys Acta 1493, 319-324.

Cook, K.L., Shajahan, A.N., Wärri, A., Jin, L., Hilakivi-Clarke, L.A., and Clarke, R.
(2012). Glucose-Regulated Protein 78 Controls Cross-talk between Apoptosis and
Autophagy to Determine Antiestrogen Responsiveness. Cancer Res 72, 3337-3349.

Craven, S.E., and Bredt, D.S. (1998). PDZ proteins organize synaptic signaling
pathways. Cell 93, 495-498.

Chung, W., Bondaruk, J., Jelinek, J., Lotan, Y., Liang, S., Czerniak, B., and Issa, J.P.
(2011). Detection of bladder cancer using novel DNA methylation biomarkers in
urine sediments. Cancer Epidemiol Biomarkers Prev 20, 1483-1491.

Costa, V.L., Henrique, R., Danielsen, S.A., Duarte-Pereira, S., Eknaes, M.,
Skotheim, R.I., Rodrigues, A., Magalhães, J.S., Oliveira, J., Lothe, R.A., et al.
(2010). Three epigenetic biomarkers, GDF15, TMEFF2, and VIM, accurately predict
bladder cancer from DNA-based analyses of urine samples. Clin Cancer Res 16,
5842-5851.

Croce, C.M. (2009). Causes and consequences of microRNA dysregulation in
cancer. Nat Rev Genet 10, 704-714.

130

Daneshmand, S., Quek, M.L., Lin, E., Lee, C., Cote, R.J., Hawes, D., Cai, J.,
Groshen, S., Lieskovsky, G., Skinner, D.G., et al. (2007). Glucose-regulated protein
GRP78 is up-regulated in prostate cancer and correlates with recurrence and
survival. Hum Pathol 38, 1547-1552.

De Carvalho, D.D., Sharma, S., You, J.S., Su, S.F., Taberlay, P.C., Kelly, T.K.,
Yang, X., Liang, G., and Jones, P.A. (2012). DNA methylation screening identifies
driver epigenetic events of cancer cell survival. Cancer Cell 21, 655-667.

Domínguez, G., Carballido, J., Silva, J., Silva, J.M., García, J.M., Menéndez, J.,
Provencio, M., España, P., and Bonilla, F. (2002). p14ARF promoter
hypermethylation in plasma DNA as an indicator of disease recurrence in bladder
cancer patients. Clin Cancer Res 8, 980-985.

Dong, D., Stapleton, C., Luo, B., Xiong, S., Ye, W., Zhang, Y., Jhaveri, N., Zhu, G.,
Ye, R., Liu, Z., et al. (2011). A critical role for GRP78/BiP in the tumor
microenvironment for neovascularization during tumor growth and metastasis.
Cancer Res 71, 2848-2857.

Duan, Q., Wang, X., Gong, W., Ni, L., Chen, C., He, X., Chen, F., Yang, L., Wang,
P., and Wang, D.W. (2012). ER Stress Negatively Modulates the Expression of the
miR-199a/214 Cluster to Regulates Tumor Survival and Progression in Human
Hepatocellular Cancer. PLoS One 7, e31518.

Dulaimi, E., Uzzo, R.G., Greenberg, R.E., Al-Saleem, T., and Cairns, P. (2004).
Detection of bladder cancer in urine by a tumor suppressor gene hypermethylation
panel. Clin Cancer Res 10, 1887-1893.

131

Edge, S.B., Byrd, D.R., Compton, C.C., Fritz, A.G., Greene, F.L., Trotti, A., Edge,
S., Byrd, D., Compton, C., Fritz, A., et al. (2010). AJCC Cancer Staging Manual, 7
th

edition, seven edn (New York: Springer-Verlag).

Esquela-Kerscher, A., and Slack, F.J. (2006). Oncomirs - microRNAs with a role in
cancer. Nat Rev Cancer 6, 259-269.

Esteller, M., Garcia-Foncillas, J., Andion, E., Goodman, S.N., Hidalgo, O.F.,
Vanaclocha, V., Baylin, S.B., and Herman, J.G. (2000). Inactivation of the DNA-
repair gene MGMT and the clinical response of gliomas to alkylating agents. N Engl
J Med 343, 1350-1354.

Fabbri, M., Garzon, R., Cimmino, A., Liu, Z., Zanesi, N., Callegari, E., Liu, S.,
Alder, H., Costinean, S., Fernandez-Cymering, C., et al. (2007). MicroRNA-29
family reverts aberrant methylation in lung cancer by targeting DNA
methyltransferases 3A and 3B. Proc Natl Acad Sci U S A 104, 15805-15810.

Force, A.A.f.C.R.H.E.T., and European Union, N.o.E., S.ientific Advisory Board
(2008). Moving AHEAD with an international human epigenome project. Nature
454, 711-715.

Fraga, M.F., Ballestar, E., Villar-Garea, A., Boix-Chornet, M., Espada, J., Schotta,
G., Bonaldi, T., Haydon, C., Ropero, S., Petrie, K., et al. (2005). Loss of acetylation
at Lys16 and trimethylation at Lys20 of histone H4 is a common hallmark of human
cancer. Nat Genet 37, 391-400.

Friedrich, M.G., Chandrasoma, S., Siegmund, K.D., Weisenberger, D.J., Cheng, J.C.,
Toma, M.I., Huland, H., Jones, P.A., and Liang, G. (2005). Prognostic relevance of

132

methylation markers in patients with non-muscle invasive bladder carcinoma. Eur J
Cancer 41, 2769-2778.

Friedrich, M.G., Weisenberger, D.J., Cheng, J.C., Chandrasoma, S., Siegmund, K.D.,
Gonzalgo, M.L., Toma, M.I., Huland, H., Yoo, C., Tsai, Y.C., et al. (2004).
Detection of methylated apoptosis-associated genes in urine sediments of bladder
cancer patients. Clin Cancer Res 10, 7457-7465.

Friedman, J.M., Jones, P.A., and Liang, G. (2009). The tumor suppressor
microRNA-101 becomes an epigenetic player by targeting the polycomb group
protein EZH2 in cancer. Cell Cycle 8, 2313-2314.

Friedman, J.M., Liang, G., Liu, C.C., Wolff, E.M., Tsai, Y.C., Ye, W., Zhou, X., and
Jones, P.A. (2009). The putative tumor suppressor microRNA-101 modulates the
cancer epigenome by repressing the polycomb group protein EZH2. Cancer Res 69,
2623-2629.

Fu, Y., Wey, S., Wang, M., Ye, R., Liao, C.P., Roy-Burman, P., and Lee, A.S.
(2008). Pten null prostate tumorigenesis and AKT activation are blocked by targeted
knockout of ER chaperone GRP78/BiP in prostate epithelium. Proc Natl Acad Sci U
S A 105, 19444-19449.

Garzon, R., Fabbri, M., Cimmino, A., Calin, G.A., and Croce, C.M. (2006).
MicroRNA expression and function in cancer. Trends Mol Med 12, 580-587.

Garzon, R., Marcucci, G., and Croce, C.M. (2010). Targeting microRNAs in cancer:
rationale, strategies and challenges. Nat Rev Drug Discov 9, 775-789.

133

Glaunsinger, B.A., Weiss, R.S., Lee, S.S., and Javier, R. (2001). Link of the unique
oncogenic properties of adenovirus type 9 E4-ORF1 to a select interaction with the
candidate tumor suppressor protein ZO-2. EMBO J 20, 5578-5586.

Glöckner, S.C., Dhir, M., Yi, J.M., McGarvey, K.E., Van Neste, L., Louwagie, J.,
Chan, T.A., Kleeberger, W., de Bruïne, A.P., Smits, K.M., et al. (2009). Methylation
of TFPI2 in stool DNA: a potential novel biomarker for the detection of colorectal
cancer. Cancer Res 69, 4691-4699.

Grkovic, S., O'Reilly, V.C., Han, S., Hong, M., Baxter, R.C., and Firth, S.M. (2012).
IGFBP-3 binds GRP78, stimulates autophagy and promotes the survival of breast
cancer cells exposed to adverse microenvironments. Oncogene.

Griffiths-Jones, S., Grocock, R.J., van Dongen, S., Bateman, A., and Enright, A.J.
(2006). miRBase: microRNA sequences, targets and gene nomenclature. Nucleic
Acids Res 34, D140-144.

Guo, H., Ingolia, N.T., Weissman, J.S., and Bartel, D.P. (2010). Mammalian
microRNAs predominantly act to decrease target mRNA levels. Nature 466, 835-
840.

Hernandez-Vargas, H., Lambert, M.P., Le Calvez-Kelm, F., Gouysse, G., McKay-
Chopin, S., Tavtigian, S.V., Scoazec, J.Y., and Herceg, Z. (2010). Hepatocellular
carcinoma displays distinct DNA methylation signatures with potential as clinical
predictors. PLoS One 5, e9749.

Hetz, C., Bernasconi, P., Fisher, J., Lee, A.H., Bassik, M.C., Antonsson, B., Brandt,
G.S., Iwakoshi, N.N., Schinzel, A., Glimcher, L.H., et al. (2006). Proapoptotic BAX

134

and BAK modulate the unfolded protein response by a direct interaction with
IRE1alpha. Science 312, 572-576.

Hoque, M.O., Begum, S., Topaloglu, O., Chatterjee, A., Rosenbaum, E., Van
Criekinge, W., Westra, W.H., Schoenberg, M., Zahurak, M., Goodman, S.N., et al.
(2006). Quantitation of promoter methylation of multiple genes in urine DNA and
bladder cancer detection. J Natl Cancer Inst 98, 996-1004.

Hu, W., Chan, C.S., Wu, R., Zhang, C., Sun, Y., Song, J.S., Tang, L.H., Levine, A.J.,
and Feng, Z. (2010). Negative regulation of tumor suppressor p53 by microRNA
miR-504. Mol Cell 38, 689-699.

Jones, P.A. (2002). DNA methylation and cancer. Oncogene 21, 5358-5360.

Jones, P.A. (2012). Functions of DNA methylation: islands, start sites, gene bodies
and beyond. Nat Rev Genet 13, 484-492.

Jones, P.A., and Laird, P.W. (1999). Cancer epigenetics comes of age. Nat Genet 21,
163-167.

Jones, S., Wang, T.L., Shih, I.M., Mao, T.L., Nakayama, K., Roden, R., Glas, R.,
Slamon, D., Diaz, L.A., Vogelstein, B., et al. (2010). Frequent mutations of
chromatin remodeling gene ARID1A in ovarian clear cell carcinoma. Science 330,
228-231.

Kandimalla, R., van Tilborg, A.A., Kompier, L.C., Stumpel, D.J., Stam, R.W.,
Bangma, C.H., and Zwarthoff, E.C. (2012). Genome-wide analysis of CpG island
methylation in bladder cancer identified TBX2, TBX3, GATA2, and ZIC4 as pTa-
specific prognostic markers. Eur Urol 61, 1245-1256.

135

Kahali, S., Sarcar, B., Prabhu, A., Seto, E., and Chinnaiyan, P. (2012). Class I
histone deacetylases localize to the endoplasmic reticulum and modulate the
unfolded protein response. FASEB J 26, 2437-2445.

Kappelmann, M., Kuphal, S., Meister, G., Vardimon, L., and Bosserhoff, A.K.
(2012). MicroRNA miR-125b controls melanoma progression by direct regulation of
c-Jun protein expression. Oncogene.

Kelly, T.K., De Carvalho, D.D., and Jones, P.A. (2010). Epigenetic modifications as
therapeutic targets. Nat Biotechnol 28, 1069-1078.

Kelly, T.K., Liu, Y., Lay, F.D., Liang, G., Berman, B.P., and Jones, P.A. (2012).
Genome-wide mapping of nucleosome positioning and DNA methylation within
individual DNA molecules. Genome Res 22, 2497-2506.

Kim, W.J., and Kim, Y.J. (2009). Epigenetic biomarkers in urothelial bladder cancer.
Expert Rev Mol Diagn 9, 259-269.

Kim, Y.K., and Kim, W.J. (2009). Epigenetic markers as promising prognosticators
for bladder cancer. Int J Urol 16, 17-22.

Kloosterman, W.P., and Plasterk, R.H. (2006). The diverse functions of microRNAs
in animal development and disease. Dev Cell 11, 441-450.

Kompier, L.C., Lurkin, I., van der Aa, M.N., van Rhijn, B.W., van der Kwast, T.H.,
and Zwarthoff, E.C. (2010). FGFR3, HRAS, KRAS, NRAS and PIK3CA mutations
in bladder cancer and their potential as biomarkers for surveillance and therapy.
PLoS One 5, e13821.

136

Kouzarides, T. (2007). Chromatin modifications and their function. Cell 128, 693-
705.

Krol, J., Loedige, I., and Filipowicz, W. (2010). The widespread regulation of
microRNA biogenesis, function and decay. Nat Rev Genet 11, 597-610.

Laird, P.W. (2003). The power and the promise of DNA methylation markers. Nat
Rev Cancer 3, 253-266.

Kumar, M., Lu, Z., Takwi, A.A., Chen, W., Callander, N.S., Ramos, K.S., Young,
K.H., and Li, Y. (2011). Negative regulation of the tumor suppressor p53 gene by
microRNAs. Oncogene 30, 843-853.

Kuwahara, K., Yang, Z., Slack, G.C., Nimni, M.E., and Han, B. (2010). Cell delivery
using an injectable and adhesive transglutaminase-gelatin gel. Tissue Eng Part C
Methods 16, 609-618.

Lee, A.S. (2007). GRP78 induction in cancer: therapeutic and prognostic
implications. Cancer Res 67, 3496-3499.

Lee, R.C., Feinbaum, R.L., and Ambros, V. (1993). The C. elegans heterochronic
gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 75,
843-854.

Lee, W.H., Morton, R.A., Epstein, J.I., Brooks, J.D., Campbell, P.A., Bova, G.S.,
Hsieh, W.S., Isaacs, W.B., and Nelson, W.G. (1994). Cytidine methylation of
regulatory sequences near the pi-class glutathione S-transferase gene accompanies
human prostatic carcinogenesis. Proc Natl Acad Sci U S A 91, 11733-11737.

137

Lewis, B.P., Burge, C.B., and Bartel, D.P. (2005). Conserved seed pairing, often
flanked by adenosines, indicates that thousands of human genes are microRNA
targets. Cell 120, 15-20.

Leung, A.K., and Sharp, P.A. (2010). MicroRNA functions in stress responses. Mol
Cell 40, 205-215.

Li, N., Zoubeidi, A., Beraldi, E., and Gleave, M.E. (2012). GRP78 regulates clusterin
stability, retrotranslocation and mitochondrial localization under ER stress in
prostate cancer. Oncogene.

Liang, G., Lin, J.C., Wei, V., Yoo, C., Cheng, J.C., Nguyen, C.T., Weisenberger,
D.J., Egger, G., Takai, D., Gonzales, F.A., et al. (2004). Distinct localization of
histone H3 acetylation and H3-K4 methylation to the transcription start sites in the
human genome. Proc Natl Acad Sci U S A 101, 7357-7362.

Lim, L.P., Lau, N.C., Garrett-Engele, P., Grimson, A., Schelter, J.M., Castle, J.,
Bartel, D.P., Linsley, P.S., and Johnson, J.M. (2005). Microarray analysis shows that
some microRNAs downregulate large numbers of target mRNAs. Nature 433, 769-
773.

Lin, H., Wong, R.P., Martinka, M., and Li, G. (2009). Loss of SNF5 expression
correlates with poor patient survival in melanoma. Clin Cancer Res 15, 6404-6411.

Lin, H.H., Ke, H.L., Huang, S.P., Wu, W.J., Chen, Y.K., and Chang, L.L. (2010).
Increase sensitivity in detecting superficial, low grade bladder cancer by combination
analysis of hypermethylation of E-cadherin, p16, p14, RASSF1A genes in urine.
Urol Oncol 28, 597-602.

138

Lintula, S., and Hotakainen, K. (2010). Developing biomarkers for improved
diagnosis and treatment outcome monitoring of bladder cancer. Expert Opin Biol
Ther 10, 1169-1180.

Lopez-Serra, P., and Esteller, M. (2012). DNA methylation-associated silencing of
tumor-suppressor microRNAs in cancer. Oncogene 31, 1609-1622.

Marasa, B.S., Srikantan, S., Masuda, K., Abdelmohsen, K., Kuwano, Y., Yang, X.,
Martindale, J.L., Rinker-Schaeffer, C.W., and Gorospe, M. (2009). Increased MKK4
abundance with replicative senescence is linked to the joint reduction of multiple
microRNAs. Sci Signal 2, ra69.

Matter, K., and Balda, M.S. (2007). Epithelial tight junctions, gene expression and
nucleo-junctional interplay. J Cell Sci 120, 1505-1511.

Mavrakis, K.J., Leslie, C.S., and Wendel, H.G. (2011). Cooperative control of tumor
suppressor genes by a network of oncogenic microRNAs. Cell Cycle 10, 2845-2849.

Mavrakis, K.J., Van Der Meulen, J., Wolfe, A.L., Liu, X., Mets, E., Taghon, T.,
Khan, A.A., Setty, M., Setti, M., Rondou, P., et al. (2011). A cooperative
microRNA-tumor suppressor gene network in acute T-cell lymphoblastic leukemia
(T-ALL). Nat Genet 43, 673-678.

Mendell, J.T. (2008). miRiad roles for the miR-17-92 cluster in development and
disease. Cell 133, 217-222.

139

Millán-Rodríguez, F., Chéchile-Toniolo, G., Salvador-Bayarri, J., Palou, J., Algaba,
F., and Vicente-Rodríguez, J. (2000). Primary superficial bladder cancer risk groups
according to progression, mortality and recurrence. J Urol 164, 680-684.

Morgan, T.M., and Clark, P.E. (2010). Bladder cancer. Curr Opin Oncol 22, 242-
249.

Nakagawa, M., Oda, Y., Eguchi, T., Aishima, S., Yao, T., Hosoi, F., Basaki, Y.,
Ono, M., Kuwano, M., Tanaka, M., et al. (2007). Expression profile of class I
histone deacetylases in human cancer tissues. Oncol Rep 18, 769-774.

Negraes, P.D., Favaro, F.P., Camargo, J.L., Oliveira, M.L., Goldberg, J., Rainho,
C.A., and Salvadori, D.M. (2008). DNA methylation patterns in bladder cancer and
washing cell sediments: a perspective for tumor recurrence detection. BMC Cancer
8, 238.

Nelson, H.H., Marsit, C.J., Christensen, B.C., Houseman, E.A., Kontic, M.,
Wiemels, J.L., Karagas, M.R., Wrensch, M.R., Zheng, S., Wiencke, J.K., et al.
(2012). Key epigenetic changes associated with lung cancer development: results
from dense methylation array profiling. Epigenetics 7, 559-566.

Nicoloso, M.S., Spizzo, R., Shimizu, M., Rossi, S., and Calin, G.A. (2009).
MicroRNAs--the micro steering wheel of tumour metastases. Nat Rev Cancer 9, 293-
302.

Ni, M., Zhang, Y., and Lee, A.S. (2011). Beyond the endoplasmic reticulum:
atypical GRP78 in cell viability, signalling and therapeutic targeting. Biochem J 434,
181-188.

140

Niwa, T., Tsukamoto, T., Toyoda, T., Mori, A., Tanaka, H., Maekita, T., Ichinose,
M., Tatematsu, M., and Ushijima, T. (2010). Inflammatory processes triggered by
Helicobacter pylori infection cause aberrant DNA methylation in gastric epithelial
cells. Cancer Res 70, 1430-1440.

Noushmehr, H., Weisenberger, D.J., Diefes, K., Phillips, H.S., Pujara, K., Berman,
B.P., Pan, F., Pelloski, C.E., Sulman, E.P., Bhat, K.P., et al. (2010). Identification of
a CpG island methylator phenotype that defines a distinct subgroup of glioma.
Cancer Cell 17, 510-522.

Oka, T., Remue, E., Meerschaert, K., Vanloo, B., Boucherie, C., Gfeller, D., Bader,
G.D., Sidhu, S.S., Vandekerckhove, J., Gettemans, J., et al. (2010). Functional
complexes between YAP2 and ZO-2 are PDZ domain-dependent, and regulate YAP2
nuclear localization and signalling. Biochem J 432, 461-472.

Oliveira, A.I., Jerónimo, C., and Henrique, R. (2012). Moving forward in bladder
cancer detection and diagnosis: the role of epigenetic biomarkers. Expert Rev Mol
Diagn 12, 871-878.

Ostling, P., Leivonen, S.K., Aakula, A., Kohonen, P., Makela, R., Hagman, Z.,
Edsjo, A., Kangaspeska, S., Edgren, H., Nicorici, D., et al. (2011). Systematic
analysis of microRNAs targeting the androgen receptor in prostate cancer cells.
Cancer Res 71, 1956-1967.

Parker, J., and Spiess, P.E. (2011). Current and emerging bladder cancer urinary
biomarkers. ScientificWorldJournal 11, 1103-1112.

Pfaffenbach, K.T., and Lee, A.S. (2011). The critical role of GRP78 in physiologic
and pathologic stress. Curr Opin Cell Biol 23, 150-156.

141

Pootrakul, L., Datar, R.H., Shi, S.R., Cai, J., Hawes, D., Groshen, S.G., Lee, A.S.,
and Cote, R.J. (2006). Expression of stress response protein Grp78 is associated with
the development of castration-resistant prostate cancer. Clin Cancer Res 12, 5987-
5993.

Pyrko, P., Schönthal, A.H., Hofman, F.M., Chen, T.C., and Lee, A.S. (2007). The
unfolded protein response regulator GRP78/BiP as a novel target for increasing
chemosensitivity in malignant gliomas. Cancer Res 67, 9809-9816.

Qiu, X., Friedman, J.M., and Liang, G. (2011). Creating a flexible multiple
microRNA expression vector by linking precursor microRNAs. Biochem Biophys
Res Commun.

Reinert, T. (2012). Methylation Markers for Urine-Based Detection of Bladder
Cancer: The Next Generation of Urinary Markers for Diagnosis and Surveillance of
Bladder Cancer. Adv Urol 2012, 503271.

Reinert, T., Borre, M., Christiansen, A., Hermann, G.G., Orntoft, T.F., and Dyrskjøt,
L. (2012). Diagnosis of Bladder Cancer Recurrence Based on Urinary Levels of
EOMES, HOXA9, POU4F2, TWIST1, VIM, and ZNF154 Hypermethylation. PLoS
One 7, e46297.

Reinert, T., Modin, C., Castano, F.M., Lamy, P., Wojdacz, T.K., Hansen, L.L., Wiuf,
C., Borre, M., Dyrskjøt, L., and Orntoft, T.F. (2011). Comprehensive genome
methylation analysis in bladder cancer: identification and validation of novel
methylated genes and application of these as urinary tumor markers. Clin Cancer Res
17, 5582-5592.

142

Reisman, D., Glaros, S., and Thompson, E.A. (2009). The SWI/SNF complex and
cancer. Oncogene 28, 1653-1668.

Ron, D., and Walter, P. (2007). Signal integration in the endoplasmic reticulum
unfolded protein response. Nat Rev Mol Cell Biol 8, 519-529.

Roue, G., Perez-Galan, P., Mozos, A., Lopez-Guerra, M., Xargay-Torrent, S.,
Rosich, L., Saborit-Villarroya, I., Normant, E., Campo, E., and Colomer, D. (2011).
The Hsp90 inhibitor IPI-504 overcomes bortezomib resistance in mantle cell
lymphoma in vitro and in vivo by down-regulation of the prosurvival ER chaperone
BiP/Grp78. Blood 117, 1270-1279.

Rouprêt, M., Hupertan, V., Yates, D.R., Comperat, E., Catto, J.W., Meuth, M.,
Lackmichi, A., Ricci, S., Lacave, R., Gattegno, B., et al. (2008). A comparison of the
performance of microsatellite and methylation urine analysis for predicting the
recurrence of urothelial cell carcinoma, and definition of a set of markers by
Bayesian network analysis. BJU Int 101, 1448-1453.

Saito, Y., Liang, G., Egger, G., Friedman, J.M., Chuang, J.C., Coetzee, G.A., and
Jones, P.A. (2006). Specific activation of microRNA-127 with downregulation of the
proto-oncogene BCL6 by chromatin-modifying drugs in human cancer cells. Cancer
Cell 9, 435-443.

Salem, C., Liang, G., Tsai, Y.C., Coulter, J., Knowles, M.A., Feng, A.C., Groshen,
S., Nichols, P.W., and Jones, P.A. (2000). Progressive increases in de novo
methylation of CpG islands in bladder cancer. Cancer Res 60, 2473-2476.

143

Schones, D.E., Cui, K., Cuddapah, S., Roh, T.Y., Barski, A., Wang, Z., Wei, G., and
Zhao, K. (2008). Dynamic regulation of nucleosome positioning in the human
genome. Cell 132, 887-898.

Scoumanne, A., Zhang, J., and Chen, X. (2009). PRMT5 is required for cell-cycle
progression and p53 tumor suppressor function. Nucleic Acids Res 37, 4965-4976.

Seligson, D.B., Horvath, S., Shi, T., Yu, H., Tze, S., Grunstein, M., and Kurdistani,
S.K. (2005). Global histone modification patterns predict risk of prostate cancer
recurrence. Nature 435, 1262-1266.

Selbach, M., Schwanhäusser, B., Thierfelder, N., Fang, Z., Khanin, R., and
Rajewsky, N. (2008). Widespread changes in protein synthesis induced by
microRNAs. Nature 455, 58-63.

Seifert, H.H., Schmiemann, V., Mueller, M., Kazimirek, M., Onofre, F., Neuhausen,
A., Florl, A.R., Ackermann, R., Boecking, A., Schulz, W.A., et al. (2007). In situ
detection of global DNA hypomethylation in exfoliative urine cytology of patients
with suspected bladder cancer. Exp Mol Pathol 82, 292-297.

Sharma, S., Gerke, D.S., Han, H.F., Jeong, S., Stallcup, M.R., Jones, P.A., and
Liang, G. (2012). Lysine methyltransferase G9a is not required for DNMT3A/3B
anchoring to methylated nucleosomes and maintenance of DNA methylation in
somatic cells. Epigenetics Chromatin 5, 3.

Sharma, S., Kelly, T.K., and Jones, P.A. (2010). Epigenetics in cancer.
Carcinogenesis 31, 27-36.

144

Shelley, M.D., Mason, M.D., and Kynaston, H. (2010). Intravesical therapy for
superficial bladder cancer: a systematic review of randomised trials and meta-
analyses. Cancer Treat Rev 36, 195-205.

Shen, Q., Cicinnati, V.R., Zhang, X., Iacob, S., Weber, F., Sotiropoulos, G.C.,
Radtke, A., Lu, M., Paul, A., Gerken, G., et al. (2010). Role of microRNA-199a-5p
and discoidin domain receptor 1 in human hepatocellular carcinoma invasion. Mol
Cancer 9, 227.

Shi, Y. (2007). Histone lysine demethylases: emerging roles in development,
physiology and disease. Nat Rev Genet 8, 829-833.

Siegel, R., Ward, E., Brawley, O., and Jemal, A. (2011). Cancer statistics, 2011: the
impact of eliminating socioeconomic and racial disparities on premature cancer
deaths. CA Cancer J Clin 61, 212-236.

Sobin, L.H., Gospodarowicz, M.K., and Christian, W. (2009). TNM Classification of
Malignant Tumours, 7th Edition (Wiley-Blackwell).

Soriano, A.O., Yang, H., Faderl, S., Estrov, Z., Giles, F., Ravandi, F., Cortes, J.,
Wierda, W.G., Ouzounian, S., Quezada, A., et al. (2007). Safety and clinical activity
of the combination of 5-azacytidine, valproic acid, and all-trans retinoic acid in acute
myeloid leukemia and myelodysplastic syndrome. Blood 110, 2302-2308.

Steiner, G., Schoenberg, M.P., Linn, J.F., Mao, L., and Sidransky, D. (1997).
Detection of bladder cancer recurrence by microsatellite analysis of urine. Nat Med
3, 621-624.

145

Sturgeon, C.M., Duffy, M.J., Hofmann, B.R., Lamerz, R., Fritsche, H.A.,
Gaarenstroom, K., Bonfrer, J., Ecke, T.H., Grossman, H.B., Hayes, P., et al. (2010).
National Academy of Clinical Biochemistry Laboratory Medicine Practice
Guidelines for use of tumor markers in liver, bladder, cervical, and gastric cancers.
Clin Chem 56, e1-48.

Sylvester, R.J., van der Meijden, A.P., Oosterlinck, W., Witjes, J.A., Bouffioux, C.,
Denis, L., Newling, D.W., and Kurth, K. (2006). Predicting recurrence and
progression in individual patients with stage Ta T1 bladder cancer using EORTC risk
tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol 49,
466-465; discussion 475-467.

Taberlay, P.C., Kelly, T.K., Liu, C.C., You, J.S., De Carvalho, D.D., Miranda, T.B.,
Zhou, X.J., Liang, G., and Jones, P.A. (2011). Polycomb-repressed genes have
permissive enhancers that initiate reprogramming. Cell 147, 1283-1294.

Tada, Y., Wada, M., Taguchi, K., Mochida, Y., Kinugawa, N., Tsuneyoshi, M.,
Naito, S., and Kuwano, M. (2002). The association of death-associated protein
kinase hypermethylation with early recurrence in superficial bladder cancers. Cancer
Res 62, 4048-4053.

Thomas, M., Lieberman, J., and Lal, A. (2010). Desperately seeking microRNA
targets. Nat Struct Mol Biol 17, 1169-1174.

Tilki, D., Burger, M., Dalbagni, G., Grossman, H.B., Hakenberg, O.W., Palou, J.,
Reich, O., Rouprêt, M., Shariat, S.F., and Zlotta, A.R. (2011). Urine markers for
detection and surveillance of non-muscle-invasive bladder cancer. Eur Urol 60, 484-
492.

146

Tsao, C.M., Yan, M.D., Shih, Y.L., Yu, P.N., Kuo, C.C., Lin, W.C., Li, H.J., and
Lin, Y.W. (2012). SOX1 functions as a tumor suppressor by antagonizing the
WNT/β-catenin signaling pathway in hepatocellular carcinoma. Hepatology 56,
2277-2287.

Ushijima, T., and Hattori, N. (2012). Molecular pathways: involvement of
Helicobacter pylori-triggered inflammation in the formation of an epigenetic field
defect, and its usefulness as cancer risk and exposure markers. Clin Cancer Res 18,
923-929.

Vallot, C., Stransky, N., Bernard-Pierrot, I., Hérault, A., Zucman-Rossi, J.,
Chapeaublanc, E., Vordos, D., Laplanche, A., Benhamou, S., Lebret, T., et al.
(2011). A novel epigenetic phenotype associated with the most aggressive pathway
of bladder tumor progression. J Natl Cancer Inst 103, 47-60.

Valk-Lingbeek, M.E., Bruggeman, S.W., and van Lohuizen, M. (2004). Stem cells
and cancer; the polycomb connection. Cell 118, 409-418.

Virrey, J.J., Dong, D., Stiles, C., Patterson, J.B., Pen, L., Ni, M., Schonthal, A.H.,
Chen, T.C., Hofman, F.M., and Lee, A.S. (2008). Stress chaperone GRP78/BiP
confers chemoresistance to tumor-associated endothelial cells. Mol Cancer Res 6,
1268-1275.

Volinia, S., Calin, G.A., Liu, C.G., Ambs, S., Cimmino, A., Petrocca, F., Visone, R.,
Iorio, M., Roldo, C., Ferracin, M., et al. (2006). A microRNA expression signature
of human solid tumors defines cancer gene targets. Proc Natl Acad Sci U S A 103,
2257-2261.

147

Voso, M.T., Santini, V., Finelli, C., Musto, P., Pogliani, E., Angelucci, E., Fioritoni,
G., Alimena, G., Maurillo, L., Cortelezzi, A., et al. (2009). Valproic acid at
therapeutic plasma levels may increase 5-azacytidine efficacy in higher risk
myelodysplastic syndromes. Clin Cancer Res 15, 5002-5007.

Waddington, C.H. (2012). The epigenotype. 1942. Int J Epidemiol 41, 10-13.

Walsh, T., Pierce, S.B., Lenz, D.R., Brownstein, Z., Dagan-Rosenfeld, O., Shahin,
H., Roeb, W., McCarthy, S., Nord, A.S., Gordon, C.R., et al. (2010). Genomic
duplication and overexpression of TJP2/ZO-2 leads to altered expression of
apoptosis genes in progressive nonsyndromic hearing loss DFNA51. Am J Hum
Genet 87, 101-109.

Wey, S., Luo, B., Tseng, C.C., Ni, M., Zhou, H., Fu, Y., Bhojwani, D., Carroll,
W.L., and Lee, A.S. (2011). Inducible knockout of GRP78/BiP in the hematopoietic
system suppresses Pten-null leukemogenesis and AKT oncogenic signaling. Blood.

Wiegand, K.C., Shah, S.P., Al-Agha, O.M., Zhao, Y., Tse, K., Zeng, T., Senz, J.,
McConechy, M.K., Anglesio, M.S., Kalloger, S.E., et al. (2010). ARID1A mutations
in endometriosis-associated ovarian carcinomas. N Engl J Med 363, 1532-1543.

Wolff, E.M., Chihara, Y., Pan, F., Weisenberger, D.J., Siegmund, K.D., Sugano, K.,
Kawashima, K., Laird, P.W., Jones, P.A., and Liang, G. (2010). Unique DNA
methylation patterns distinguish noninvasive and invasive urothelial cancers and
establish an epigenetic field defect in premalignant tissue. Cancer Res 70, 8169-
8178.

Wolff, E.M., Liang, G., and Jones, P.A. (2005). Mechanisms of Disease: genetic and
epigenetic alterations that drive bladder cancer. Nat Clin Pract Urol 2, 502-510.

148

Wolff, E.M., Byun, H.M., Han, H.F., Sharma, S., Nichols, P.W., Siegmund, K.D.,
Yang, A.S., Jones, P.A., and Liang, G. (2010). Hypomethylation of a LINE-1
promoter activates an alternate transcript of the MET oncogene in bladders with
cancer. PLoS Genet 6, e1000917.

Wu, S., Huang, S., Ding, J., Zhao, Y., Liang, L., Liu, T., Zhan, R., and He, X.
(2010). Multiple microRNAs modulate p21Cip1/Waf1 expression by directly
targeting its 3' untranslated region. Oncogene 29, 2302-2308.

Yeung, B.H., Kwan, B.W., He, Q.Y., Lee, A.S., Liu, J., and Wong, A.S. (2008).
Glucose-regulated protein 78 as a novel effector of BRCA1 for inhibiting stress-
induced apoptosis. Oncogene 27, 6782-6789.

You, J.S., Kelly, T.K., De Carvalho, D.D., Taberlay, P.C., Liang, G., and Jones, P.A.
(2011). OCT4 establishes and maintains nucleosome-depleted regions that provide
additional layers of epigenetic regulation of its target genes. Proc Natl Acad Sci U S
A 108, 14497-14502.

You, J.S., and Jones, P.A. (2012). Cancer genetics and epigenetics: two sides of the
same coin? Cancer Cell 22, 9-20.

Zahraoui, A., Louvard, D., and Galli, T. (2000). Tight junction, a platform for
trafficking and signaling protein complexes. J Cell Biol 151, F31-36.

Zhang, J., Dai, J., Qi, Y., Lin, D.L., Smith, P., Strayhorn, C., Mizokami, A., Fu, Z.,
Westman, J., and Keller, E.T. (2001). Osteoprotegerin inhibits prostate cancer-
induced osteoclastogenesis and prevents prostate tumor growth in the bone. J Clin
Invest 107, 1235-1244.

149

Zhang, Y., Roccaro, A.M., Rombaoa, C., Flores, L., Obad, S., Fernandes, S.M.,
Sacco, A., Liu, Y., Ngo, H., Quang, P., et al. (2012). LNA-mediated anti-
microRNA-155 silencing in low-grade B cell lymphomas. Blood.

Zhao, Y., Guo, S., Sun, J., Huang, Z., Zhu, T., Zhang, H., Gu, J., He, Y., Wang, W.,
Ma, K., et al. (2012). Methylcap-seq reveals novel DNA methylation markers for the
diagnosis and recurrence prediction of bladder cancer in a Chinese population. PLoS
One 7, e35175.

Zuiverloon, T.C., Beukers, W., van der Keur, K.A., Munoz, J.R., Bangma, C.H.,
Lingsma, H.F., Eijkemans, M.J., Schouten, J.P., and Zwarthoff, E.C. (2012). A
methylation assay for the detection of non-muscle-invasive bladder cancer (NMIBC)
recurrences in voided urine. BJU Int 109, 941-948.

Zuiverloon, T.C., van der Aa, M.N., van der Kwast, T.H., Steyerberg, E.W.,
Lingsma, H.F., Bangma, C.H., and Zwarthoff, E.C. (2010). Fibroblast growth factor
receptor 3 mutation analysis on voided urine for surveillance of patients with low-
grade non-muscle-invasive bladder cancer. Clin Cancer Res 16, 3011-3018.

Abstract (if available)

Abstract Two of the most challenging issues clinics face in anti-cancer therapy are tumor recurrence and drug-resistance. To overcome these two urgent problems, there is certainly a need to develop promising markers to easily monitor cancer recurrence and to develop novel strategies to overcome drug-resistance. Epigenetic alterations, including DNA methylation, histone modifications, and nucleosome occupancy, as well as non-coding RNAs, such as microRNAs, have been extensively observed in virtually all types of human malignancies and may correlate with carcinogenesis, tumor recurrence, and therapeutic outcomes. Therefore, these epigenetic alterations and microRNAs serve as potential markers or drug targets, and thus harbour wide applications in diagnostic and therapeutic intervention. ❧ The high risk of recurrence in patients following transurethral resection of bladder tumor (TURBT) for non-muscle invasive disease necessitates lifelong maintenance treatment and surveillance. To aid in this process, I have identified a panel of DNA methylation markers that can be analyzed in urine sediments to accurately predict and detect bladder tumor recurrence in follow-up visits of TURBT patients. Importantly, this panel, a combination of the hypermethylated markers SOX1 (a transcription factor) and IRAK3 (a epigenetic driver), and one hypomethylated marker L1-MET (a specific LINE1 element), showed a higher sensitivity than urine cytology and cystoscopy in detecting tumor recurrence. Next, I showed the functional roles and the regulatory mechanisms of one of these markers, ZO2 (TJP2), at its two alterative CpG-rich promoters in the epigenetic levels. Specifically, distinct mRNA expression and DNA methylation patterns were found at the two promoters in various human tissues and tumors, indicating that tissue-specific behaviours also correlated with CpG island promoters. In addition, to characterize a putative enhancer of TJP2 around the downstream promoter, I analyzed endogenous DNA methylation and nucleosome occupancy using Nucleosome occupancy and methylome sequencing (NOMe-seq) in cells with active or repressed TJP2 expression. The results showed distinct nucleosome configurations at different DNA regulatory regions. Finally, to address the challenge of drug-resistance in cancer therapy, specifically concentrating on GRP78-mediated drug-resistance, I have discovered multiple microRNAs (mir-30d, mir-181a, and mir-199a-5p) that act cooperatively to suppress GRP78 levels and thus GRP78-mediated chemoresistance. These results showed that the strategy of delivering co-transcribed microRNAs may hold therapeutic potential. ❧ In summary, I herein identified the confidential markers for the prediction and detection of bladder cancer recurrence, as well as a microRNA-based approach for the resensitization of tumor cells to chemotherapeutic drugs. My research in the field of cancer epigenetics and its translational applications can potentially advance cancer prevention and management.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

DNA methylation inhibitors and epigenetic regulation of microRNA expression

PDF

Effects of chromatin regulators during carcinogenesis

PDF

The role of microRNAs in cancer

PDF

Functional DNA methylation changes in normal and cancer cells

PDF

Creating a multiple micrornia expression vector to target GRP78, an ER chaperone and signaling regulator in cancer

PDF

DNA methylation markers for blood-based detection of small cell lung cancer in mouse models

PDF

Functional role of chromatin remodeler proteins in cancer biology

PDF

Understanding DNA methylation and nucleosome organization in cancer cells using single molecule sequencing

PDF

Integrated genomic & epigenomic analyses of glioblastoma multiforme: Methods development and application

PDF

Identification and characterization of cancer-associated enhancers

PDF

Discovery of mature microRNA sequences within the protein- coding regions of global HIV-1 genomes: Predictions of novel mechanisms for viral infection and pathogenicity

PDF

RNA methylation in cancer plasticity and drug resistance

PDF

Exploring serum and tear micro-RNA as biomarkers for early diagnosis of Sjögren’s Syndrome

PDF

CpG poor promoter SULT1C2 regulated by DNA methylation and is induced by cigarette smoke condensate in lung cell lines

PDF

Air pollution, mitochondrial function, and growth in children

PDF

The role of endoplasmic reticulum chaperones in adipogenesis, liver cancer and mammary gland development

PDF

The role of endoplasmic reticulum chaperone protein GRP78 in breast cancer

PDF

TLR8-transferred miR-192 acts as a tumor suppressor in neuroblastoma by inhibiting CTCF

PDF

Targeting glioma cancer stem cells for the treatment of glioblastoma multiforme

PDF

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

Asset Metadata

Creator Su, Sheng-Fang (author)

Core Title Identification of novel epigenetic biomarkers and microRNAs for cancer therapeutics

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 04/23/2014

Defense Date 03/07/2013

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

Tag cancer therapeutics,DNA methylation,epigenetic biomarkers,microRNA,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Jones, Peter Anthony (committee chair), Coetzee, Gerhard (Gerry) A. (committee member), Shibata, Darryl K. (committee member)

Creator Email shengfas@usc.edu,shwchen@ntu.edu.tw

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-241461

Unique identifier UC11294457

Identifier etd-SuShengFan-1579.pdf (filename),usctheses-c3-241461 (legacy record id)

Legacy Identifier etd-SuShengFan-1579.pdf

Dmrecord 241461

Document Type Dissertation

Format application/pdf (imt)

Rights Su, Sheng-Fang

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