Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Functional role of chromatin remodeler proteins in cancer biology
(USC Thesis Other)
Functional role of chromatin remodeler proteins in cancer biology
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
FUNCTIONAL ROLE OF CHROMATIN REMODELER PROTEINS IN CANCER
BIOLOGY
By
Ranjani Lakshminarasimhan
A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirement for the Degree
DOCTOR OF PHILOSOPHY
(GENETIC, MOLECULAR AND CELLULAR BIOLOGY)
DECEMBER 2016
Copyright 2016 Ranjani Lakshminarasimhan
2
EPIGRAPH
“ கற் ற து ககமண் அள வு , கல்லாதது உலகள வு
Katrathu Kai Mann Alavu, Kallathathu Ulagalavu"
“What you have learnt is a mere handful; what you have yet to learn is the size of the
world."
-Avvaiyaar
3
DEDICATION
I dedicate this work to my father whose enthusiasm for science and the pursuit of
knowledge motivated me to become a scientist, to my mother whose unwavering
support pushed me to never give up, and to my grandfather whose brilliance and
curiosity remains a constant source of inspiration.
4
ACKNOWLEDGMENTS
I would like to acknowledge, with gratitude, the following people:
Dr. Peter A. Jones for his guidance, and support. Your attention to detail, and
passion for science are a great source of inspiration. Most of all, thank you for providing
me with the space and resources necessary to learn and do science.
Dr. Gangning Liang, for his mentorship through all of my projects. The work in
this thesis would not have been possible without your patience, and willingness to help,
especially this past year.
Drs. Jueng Soo You, and Claudia Andreu-Vieyra for guiding me through my first
year in the lab. Drs. Kamilla Mundberg, Jessica Charlet, Minmin Liu, Chris Duncan,
Xiaojing Yang, Elinne Becket, Sameer Chopra, and Hongtao Li for their helpful advice,
encouragement, and friendship throughout my time as a student.
My fellow graduate students of past and present, Dr. Fides Lay, Dr. Chris
Duymich, Dr. Sheng-Fang Su, Dr. Alexandra Soegaard, Yin-Wei Chang, and Weiya Ni
who enriched my life in and out of the Jones lab.
Dr. Yvonne Tsai, for your all your efforts in keep things organized in the lab as it
underwent so many changes. Thank you for all the support you have given me through
these years.
Drs. Simon Gayther, and Kate Lawrenson, for going above and beyond the call
of duty as collaborators and providing me with invaluable support and guidance. Dr.
Bernard Weissman for his collaboration. My committee members Drs. Gerry Coetzee,
and Peggy Farnham for their advice and feedback.
5
Finally, I would like to like to thank my close friends and family whose care and
encouragement have kept me sane these past 5 years. My brother for never failing to
be a source of entertainment and support. My husband for his incredible patience,
kindness, and love throughout this process.
6
TABLE OF CONTENTS
EPIGRAPH………………………………………………………………………………………2
DEDICATION……………………………………………………………………………………3
ACKNOWLEDGMENTS……………………………………………………………………….4
LIST OF TABLES……………………………………………………………………………....9
LIST OF FIGURES…………………………………………………………………………….10
ABSTRACT…………………………………………………………………………………….12
CHAPTER 1: AN OVERVIEW OF EPIGENETIC MECHANISMS IN CANCER
1.1 INTRODUCTION.................................................................................................................14
1.2 DNA METHYLATION .........................................................................................................15
1.2.1 DNA METHYLATION IN NORMAL MAMMALIAN TISSUE .............................................17
1.2.2 DNA METHYLATION IN CANCER ..................................................................................18
1.2.2.1 PROMOTER HYPERMETHYLATION ...........................................................................18
1.2.2.2 NON-CODING RNAs ....................................................................................................21
1.2.2.3 DNA HYPOMETHYLATION ..........................................................................................23
1.2.2.4 DNA METHYLATION AT INTERGENIC AND NON-PROMOTER INTRAGENIC
REGIONS IN CANCER .............................................................................................................24
1.2.2.4.1 DNA METHYLATION CHANGES IN TRANSCRIBED REGIONS ..............................25
1.2.2.4.2 DNA METHYLATION AT ENHANCERS ....................................................................26
1.2.2.7 CPG ISLAND METHYLATOR PHENOTYPE (CIMP) STRATIFY TUMOR SUBCLASS
.................................................................................................................................................27
1.3 NUCLEOSOME POSITIONING ..........................................................................................29
1.3.1 NUCLEOSOME POSITIONING IN CANCER ...................................................................31
1.4 HISTONE MODIFICATIONS ...............................................................................................32
1.4.1 HISTONE MODIFICATIONS IN CANCER .......................................................................33
1.5 THE EPIGENOME AS A THERAPEUTIC TARGET ...........................................................34
1.6 CONCLUSION ....................................................................................................................39
7
CHAPTER 2: INDUCTION OF SNF5 IN MALIGNANT RHABDOID TUMOR CELLS
RESCUES THE ABERRANT EXPRESSION IN PART THROUGH EPIGENETIC
MODULTATION
2.1 INTRODUCTION.................................................................................................................44
2.2 METHODS ..........................................................................................................................47
2.2.1 Cell culture .....................................................................................................................47
2.2.2 Protein extraction and western blot analysis ...............................................................47
2.2.3 RNA extraction and analysis .........................................................................................48
2.2.4 NOMe-seq and AcceSssIble assays .............................................................................48
2.2.5 Infinium HumanMethylation450 BeadChip data processing and AcceSssIble data
analysis....................................................................................................................................50
2.2.6 Computational analysis .................................................................................................51
2.3 RESULTS ...........................................................................................................................53
2.3.1 TCC642: MRT cell line with inducible SNF5 expression ..............................................53
2.3.2 Gene expression changes due to SNF5 induction reflects normal tissue expression
profile .......................................................................................................................................55
2.3.3 Gene expression alterations rectify the molecular pathways and regulatory
networks misfiring in MRTs ...................................................................................................59
2.3.4 Chromatin accessibility and DNA methylation characterized using AcceSssIble
finds accessibility increase at a limited number of regions. ...............................................62
2.3.5 A subset of genes presenting promoter accessibility changes are also significantly
altered in expression ..............................................................................................................66
2.3.6 TRIM2 expression positively correlates with prognosis in several cancer types .....69
2.3.7 TRIM2 is upregulated in response to SNF5 expression, independent of senescence,
and this is due to increased accessibility of the promoter. .................................................72
2.4 DISCUSSION ......................................................................................................................75
CHAPTER 3: DOWN-REGULATION OF ARID1A IS SUFFICIENT TO INDUCE
EPIGENETIC ALTERATIONS AND PHENOTYPIC TRANSFORMATION OF
ENDOMETRIOTIC CELLS
3.1 INTRODUCTION.................................................................................................................79
3.2 METHODS ..........................................................................................................................82
3.2.1 Cell culture .....................................................................................................................82
3.2.2 Lentiviral transfection ....................................................................................................82
3.2.3 Protein extraction and western blot analysis ...............................................................82
3.2.4 Anchorage-independent colony formation ..................................................................83
3.2.5 Cell adhesion to collagen ..............................................................................................83
8
3.2.6 Cell invasion through basement membrane extract ....................................................84
3.2.7 DNA content analysis by propidium iodide (PI) staining .............................................84
3.2.8 NOMe-seq and AcceSssIble assays .............................................................................84
3.2.9 Genome-wide NOMe-seq analysis ................................................................................86
3.2.9.1 Library preparation .....................................................................................................86
3.2.9.2 Sequence mapping and GC/CG methylation level extraction ..................................86
3.2.9.3 Nucleosome depleted region (NDR) detection ..........................................................87
3.2.10 Infinium HumanMethylation450 BeadChip data processing and AcceSssIble data
analysis....................................................................................................................................88
3.2.11 RNA extraction and analysis .......................................................................................88
3.2.12 ChIP-seq .......................................................................................................................89
3.3 RESULTS ...........................................................................................................................92
3.3.1 Establishing ARID1A KD cell line to model OCCC.......................................................92
3.3.2 Decreased ARID1A expression in an endometriosis cell line enhances colony
formation capacity, alters cell cycle distribution, promotes cell adhesiveness, and
invasiveness............................................................................................................................95
3.3.3 Alteration of gene expression by ARID1A knockdown mimics the gene expression
deregulation in OCCC. ............................................................................................................97
3.3.4 Global chromatin accessibility and DNA methylation profile is largely unperturbed
by down-regulation of ARID1A............................................................................................. 102
3.3.5 ARID1A loss contributes to the redistribution of H3K27ac and some redistribution
of H3K27me3 at promoters ................................................................................................... 108
3.3.6 Down-regulation of ARID1A also contributes to the redistribution of H3K27ac mark
at enhancers .......................................................................................................................... 114
3.3.7 A subset of differentially expressed genes are directly affected by epigenetic
changes due to loss of ARID1A ........................................................................................... 118
3.4 DISCUSSION .................................................................................................................... 124
CHAPTER 4: SUMMARY AND CONCLUSIONS
………………………………………………………. …………………………...…………..129
REFERENCES
…………………………………………...…………..………………....………….………….134
9
LIST OF TABLES
Table 2.1 Gene expression and NOMe-seq primer sequences……………………...52
Table 2.2 RNA-seq differential expression results of p21, p16, E2F gene family
members, and SNF5 in TCC642 with SNF5 induction…………………………58
Table 3.1 Primer list …………………………………………………………………………91
Table 3.2. Genes altered in iEEC16 ARID1A KD cell lines…………………………..101
10
LIST OF FIGURES
Figure 1.1 DNA methylation equilibrium between promoter and gene body
modulates gene expression…………………..…………………………………… 40
Figure 1.2 DNMTi exert anti-tumoral effect by eliciting immune response in cancer
cells …………………………………………….………………………………………41
Figure 2.1. SNF5 mRNA and protein induced in TCC642…………………………….. 54
Figure 2.2. Gene expression changes triggered by SNF5 induction in TCC642 are
consistent with deregulation observed in MRT primary tumors……………. 57
Figure 2.3. Most significant pathways and upstream regulators associated with
genes differentially expressed in TCC642 upon SNF5 induction………….. .61
Figure 2.4 Chromatin accessibility or DNA methylation of 4561 probes are
affected by expression of SNF5 in TCC642…………………………………….. 65
Figure 2.5. Genes changing in expression and promoter accessibility in TCC642
are correlated with SNF5 expression in primary samples……………………68
Figure 2.6. TRIM2 expression is prognostic in ccRCC, PRCC, LGG, and LUAD…71
Figure 2.7. TRIM2 expression is dependent on SNF5 expression and regulated by
promoter accessibility……..……………………………………………………….74
Figure 3.1 ARID1A knockdown lines established in iEEC16……...…………………. 93
Figure 3.2. Stable ARID1A knockdown lines generated in ES2 and iOSE4 cell lines
……………………………………………………………...………..………………….94
Figure 3.3. Loss of ARID1A contributes to phenotypic changes in iEEC16……… 96
Figure 3.4. Differentially expressed genes in ARID1A KD are deregulated in OCCC
…….……………………………………………………………………….…...............99
Figure 3.5. Pathway analysis of the differentially expressed genes………………100
Figure 3.6. Chromatin accessibility or DNA methylation of over 11,000 probes is
affected by decreased ARID1A expression in iEEC16 ..…………………….105
Figure 3.7. Minimal changes to chromatin accessibility and DNA methylation
occur in ES2 and iOSE4 ...………………………………………………………...106
11
Figure 3.8. NOMe-seq reveals majority of the genome unperturbed by loss of
ARID1A ..……………………………………………………………………………..107
Figure 3.9. Global H3K27ac and H3K27me3 remain largely unchanged………… 111
Figure 3.10. Down-regulation of ARD1A promotes re-distribution of H3K27ac and
H4K27me3……………………………………………………………………………112
Figure 3.11. Promoter accessibility changes accompany histone modification
alterations at a small subset of TSS…………………………………………….113
Figure 3.12. Loss of ARID1A expression affects enrichment of H3K27ac at
enhancers…….……………………………………………………………………... 116
Figure 3.13. Chromatin accessibility changes accompany histone modification
alterations at small subset of enhancers …..………………………………… 117
Figure 3.14. Subset of the differentially expressed genes are epigenetically
regulated by ARID1A………………………………………………………........... 122
Figure 3.15. H3K27ac redistribution affects both gain and loss of regulatory
elements…………………………………………………………………………….. 123
12
ABSTRACT
The complex interplay between epigenetic mechanisms such as DNA methylation,
histone modification, and nucleosome positioning regulate gene expression potential and
chromatin organization over cell generations. Mutations in mediators of epigenetic
processes can contribute to considerable alterations of gene expression and disrupt
cellular identity and function, resulting in the development of malignancies. Large-scale
sequencing efforts have uncovered germline and somatic mutations in many epigenetic
modulators including DNA methyltransferases, histone tail modifiers, and subunits of
chromatin remodeling complexes. Specifically, the components of the SWI/SNF
remodeling complexes are mutated in upwards of 20% of all tumors. However, our
understanding of the molecular changes that take place due to these mutations is still
incomplete. This dissertation aims to examine the role of SWI/SNF complex proteins in
shaping the epigenetic landscape of cancer and how the interplay between the various
epigenetic mechanisms contribute to functional alterations.
First, I demonstrated that induction of SNF5 in malignant rhabdoid tumor (MRT)
cell line TCC642 significantly altered the transcriptome and restores aberrant expression
of several genes closer to normal levels. Further, I surveyed the epigenome using the
DNA methylation array-based AcceSssIble assay and uncovered several potential direct
epigenetic targets of SNF5. Finally, I established TRIM2 as a downstream target of SNF5.
Thus, I found that re-expression of SNF5 in TCC642 cells regulated target gene
expression in part by remodeling chromatin.
13
Next, using an endometriosis cell line model, I comprehensively characterized
gene expression, histone modification, chromatin accessibility, and DNA methylation
changes that take place due to loss of ARID1A. Specifically, I showed that the loss of
ARID1A triggered a considerable re-distribution of the H3K27ac histone modification and
affected the expression of a number of genes critical to ovarian cancer development.
These molecular aberrations contributed to phenotypic changes including increased
anchorage-independent colony formation and increased invasion of basement membrane
extract in vitro.
Thus, this dissertation sheds light on how mutations in a single component of an
epigenetic complex can not only alter the epigenome but can also promote functional
changes and possibly contribute to oncogenesis.
14
CHAPTER 1
AN OVERVIEW OF EPIGENETIC MECHANISMS IN CANCER
Parts of the following introduction are adapted from a book chapter published in Springer
International Publishing Switzerland. 2017. A. Jeltsch, R.Z. Jurkowska (eds.), DNA
Methyltransferases - Role and Function, Advances in Experimental Medicine and Biology
945, DOI 10.1007/978-3-319-43624-1_7. Ranjani Lakshminarasimhan, and Gangning
Liang. Role of DNA methylation in Cancer.
I was responsible for writing the draft and have incorporated all suggested edits from
Gangning Liang. Peter A. Jones provided additional edits to this chapter.
1.1 INTRODUCTION
Epigenetic mechanisms include DNA methylation, nucleosome positioning, and
histone tail modifications, and coordinated management of these mechanisms allows
cells to heritably regulate gene expression potential over cellular generations. Proper
epigenetic control is critical to maintaining genomic integrity and enables genetically
identical cells to adopt distinct cellular phenotypes.
Technological advancements in next-generation sequencing have given
scientists an opportunity to methodically conduct large scale studies on cancer
genomes, transcriptomes, and epigenomes. As a result, it has become evident that
cancer genomes carry frequent and recurrent mutations in a wide variety of epigenetic
modulators including DNA methyltransferases, histone modifiers, and chromatin
remodelers (Shen and Laird, 2013; You and Jones, 2012). Loss of or altered activity of
15
these epigenetic players can contribute to genomic instability and aberrant gene
expression, thereby giving rise to diseases such as cancer (Hanahan and Weinberg,
2011; Sharma et al., 2010).
Classic hallmarks of cancer, as described by Hanahan and Weinberg, include
maintenance of cell proliferation, evasion of growth suppression and cell death,
promotion of angiogenesis, invasion, and metastasis (Hanahan and Weinberg, 2011). It
is now clear that both genetic and epigenetic alterations underlie these biological
processes. The impact of genetic mutations on tumorigenesis has long been recognized
and studied. Complementary to these genetic events, it is now accepted that oncogenic
traits also accumulate through epigenetic disruptions (Baylin and Jones, 2011;
Sandoval and Esteller, 2012). By investigating how these epigenetic mechanisms
influence oncogenesis, we can uncover previously unknown vulnerabilities and design
more effective therapeutic strategies to combat cancer progression.
1.2 DNA METHYLATION
Mammalian DNA methylation primarily occurs through the covalent addition of a
methyl group to the carbon-5 atom of cytosine in a cytosine-guanine (CpG) dinucleotide.
This enzymatic reaction can be catalyzed by three DNA methyltransferases (DNMTs).
DNMT3A and DNMT3B show equal preference to hemimethylated and unmethylated
DNA molecules and are essential for de novo DNA methylation (Okano et al., 1999).
DNMT3A and DNMT3B are highly expressed in embryonic stem (ES) cells and, though
downregulated, continue to be lowly expressed in somatic cells. (Sharma et al., 2011).
16
DNMT1 methylates CpGs in newly synthesized daughter strands during replication. While
DNMT1 carries out the majority of the DNA methylation in a dividing cell, DNMT3A/3B
strongly associate with nucleosomes to permit efficient propagation of DNA methylation
by methylating sites missed by DNMT1 (Jones and Liang, 2009; Liang et al., 2002; Okano
et al., 1999; Rhee et al., 2002; Sharma et al., 2011).
To complement DNA methylation by DNMTs, recently identified ten-eleven-
translocation (TET) family of dioxygenases provide a mechanistic basis for active DNA
demethylation. Through successive enzymatic reactions, these enzymes oxidize 5-
methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC), 5-carboxylcytosine (5caC),
and 5-formylcytosine (5fC) (Ko et al., 2010; Pastor et al., 2011, 2013). Finally, the oxidized
intermediates can be restored to cytosine by iterative oxidation followed by base excision
repair mediated by thymine DNA glycosylase (TDG) (Kohli and Zhang, 2013). Together
with DNMTs, TET enzymes contribute to a model for the dynamic regulation of DNA
methylation.
Methylated cytosine residues are more susceptible to spontaneous deamination
than unmethylated cytosine, resulting in cytosine-to-thymine (C→T) transitions
(Coulondre et al., 1978). A consequence of this phenomenon is a global reduction in CpG
dinucleotides in the human genome, with the exception of CpG islands (CGIs). CGIs
maintain a high density of CpG dinucleotides because they are generally unmethylated
and therefore “protected” from C→T transitions through deamination (Meissner et al.,
2008).
CGIs play an important role in gene expression regulation. They are often found in
promoters of ubiquitously expressed genes (Meissner et al., 2008; Mikkelsen et al.,
17
2007a). Methylation of CGIs at gene promoters can silence gene expression. Non-CGI
promoters, on the other hand, are more likely to be tissue-specifically expressed.
Therefore, only a small subset of non-CGI promoters remain unmethylated and
accessible to transcription factors in each tissue type (Eckhardt et al., 2006).
1.2.1 DNA METHYLATION IN NORMAL MAMMALIAN TISSUE
DNA methylation is vital to normal cellular physiology and has unique functions in
mammalian tissue. Throughout evolution, the mammalian genome has accumulated a
large number of long-terminal repeat (LTR) transposable retroviral elements. These
elements make up more than a third of the human genome (Cordaux and Batzer, 2009).
CpG methylation silences transposable elements and prevents their transcription. DNA
methylation of these repeat elements is central to the maintenance of genomic integrity
(Yoder et al., 1997).
Methylation is also necessary for genomic imprinting. This mechanism results in
the unequal contribution of genes on chromosomes inherited from each parent to
embryonic development. Imprinted genes are expressed in a parental-specific manner
rather than from both chromosomes. DNA methylation is a key mechanism by which
allele-specific expression is established. For example, if the maternal allele is imprinted
by DNA methylation, it becomes silenced, and only the gene inherited from the father is
expressed (Ferguson-Smith, 2011; Li et al., 1993).
Lastly, DNA methylation serves in X-chromosome inactivation, which is a
necessary event during development. In mammals, one copy of the X-chromosome is
18
stochastically inactivated in order to equalize the dosage of X-linked genes in females to
that in males (Pessia et al. 2012). The process is initiated and propagated by the
increased expression of the non-coding RNA XIST on the X-chromosome to be
inactivated (Xi). This then triggers a cascade of events that result in the exclusion of RNA
polymerase as well as recruitment of repressive histone marks to Xi (Pontier and Gribnau,
2011). Once inactivation of the X-chromosome has been established, DNA methylation
of CpG islands is necessary for the maintenance of the silenced state (Bestor et al., 2015;
Sharp et al., 2011).
1.2.2 DNA METHYLATION IN CANCER
Broad changes of the epigenome accompany cancer initiation and progression. It
has been known for decades that cancer cells display a global loss of CpG methylation,
including regions with low density of CpG sites, repeat elements, retrotransposons, and
laminin-associated domains (LADs). This phenomenon is juxtaposed to concomitant focal
hypermethylation at CpG islands and CpG island shores (Weisenberger and Liang,
2015).
1.2.2.1 PROMOTER HYPERMETHYLATION
In addition to genetic mutations, epigenetic processes such as DNA methylation
serve as an alternative mechanism for the inactivation of tumor suppressor genes (TSGs)
(Jones and Laird, 1999; You and Jones, 2012). The hypermethylation of CGI promoters
in cancer cells is inversely correlated with gene expression and results in the silencing of
19
many known tumor suppressor genes (Figure 1.1b) (Ehrlich and Lacey, 2013; Irizarry et
al., 2009; Jones and Baylin, 2007; Shen and Laird, 2013). Silencing of cell cycle
regulators and DNA repair genes through DNA methylation has been reported in many
different cancer types and is often mutually exclusive with mutational inactivation of the
same genes (Alvarez-Nuñez et al., 2006; Chiang et al., 2006; Costello et al., 1996; Sakai
et al., 1991). Sporadic breast and ovarian cancer display a loss of BRCA1 expression
due to promoter hypermethylation. Similarly, epigenetic silencing of tumor suppressor
VHL via promoter methylation predisposes individuals to several malignancies including
clear cell renal cell carcinoma (Chiang et al., 2006; Creighton et al., 2013; Esteller, 2001;
Esteller et al., 2000; Herman et al., 1994).
Silencing of DNA repair genes contributes to a greater burden of genomic
instability and genetic mutations. O6-Methylguanine-DNA-Methyltransferase (MGMT), a
DNA repair enzyme responsible for clearing out alkylation adducts on DNA, is frequently
hypermethylated in many cancers including gliomas and colorectal cancer (CRC).
Consequently, MGMT was one of the first tumorigenic DNA methylation biomarkers to be
discovered. The suppression of MGMT due to promoter hypermethylation results in
increased susceptibility to genetic mutations in other TSGs and oncogenes such as p53
and KRAS. Interestingly, loss of MGMT renders the cell more vulnerable to treatment by
the chemotherapeutic agent temozolomide (TMZ). Clinical studies in glioblastoma
multiforme (GBM) suggest treatment with TMZ is most beneficial in cases where the
tumor presents MGMT promoter hypermethylation (Donson et al., 2007; Silber et al.,
2012; Zarnett et al., 2015)
20
Similarly, promoter hypermethylation of the mismatch repair gene MLH1 leads to
decreased expression of MLH1 in many cancers (Lin et al., 2007). The acquired
hypermethylation of MLH1 promoter and the subsequent loss of mismatch repair activity
is considered to be a primary mechanism for microsatellite instability (MSI) (Shigeyasu et
al., 2015). MSI is responsible for the pathogenesis of many cancers including CRC and
endometrial carcinomas (Hinoue et al., 2012; Hitchins et al., 2007; Li et al., 2013;
Weisenberger et al., 2006).
Hypermethylation of CpG islands can also contribute to the loss of imprinting.
When the imprinted locus IGF2/H19 becomes aberrantly methylated, the expression of
H19 is suppressed and the expression of the growth factor IGF2 is increased (Kaneda
and Feinberg, 2005; Ravenel et al., 2001). Sustained overexpression of IGF2 has been
noted to contribute to the development and progression of colorectal and gastric cancers,
and the loss of imprinting at this locus is the most common alteration in Wilms’ tumor
(Bjornsson et al., 2007; Cui, 2007; Li et al., 1993; Taniguchi et al., 1995; Wu et al., 1997)
. DNA methylation aberration in cancers are widespread, and understanding which
of these changes are driving the tumor phenotype can of great help in guiding therapeutic
strategies. In a recent study, our group applied the concept of DNA methylation addiction
to identify epigenetic drivers of tumorigenesis. We hypothesized that cancer cells
depended on the methylation of a few vital regions for survival, and these regions would
be more likely to maintain DNA methylation when methylation levels were reduced
artificially. Because these regions contribute to the fitness of the cancer cell, they are
potential drivers of the tumor condition. To test this hypothesis, global DNA methylation
of the colorectal cell line HCT116 was compared to two of its derivatives lacking either
21
one or more DNMT (Rhee et al., 2000, 2002). Epigenetic drivers were determined by
identifying genomic regions that maintain methylation preferentially and in a cancer-
specific manner in the HCT116 derivative line. One of the candidate epigenetic drivers
identified by this approach was interleukin-1 receptor-associated kinase 3 (IRAK3). The
IRAK3 promoter is specifically hypermethylated in cancer and this correlates with its
reduced expression in tumors relative to normal tissue. Importantly, IRAK3 indirectly
inhibits multiple pathways essential for cancer survival, including STAT3, NF-kB, and
MAPK pathways. Therefore, downregulation of IRAK3 is beneficial for cancer
progression. Knocking down IRAK3 in a non-tumorigenic cell line was sufficient to
increase colony formation in vitro. IRAK3 is silenced in HCT116 by DNA methylation and
overexpression of IRAK3 in this line resulted in decreased cell viability (De Carvalho et
al., 2012).
1.2.2.2 NON-CODING RNAs
In addition to regulating mRNA expression, DNA methylation plays an important
role in regulating non-coding RNA (ncRNA), such as microRNA (miRNA), small nucleolar
RNA (snoRNA), vault RNA (vtRNA), and long non-coding RNA (lncRNA). These elements
are critical regulators of cellular processes including proliferation, differentiation, and
development (Esteller, 2011). Aberrant methylation can result in dysregulation of
microRNAs and contribute to cancer development. In bladder cancer cells, treatment with
the DNMT inhibitor (DNMTi) 5’-Aza-2’-deoxycytidine (5-Aza-CdR), upregulates mir-127
and subsequently downregulates the proto-oncogene BCL-6 (Ehrlich, 2010; Kulis et al.,
22
2013; Saito et al., 2006). Likewise, when the microRNA mir-124a is silenced due to
hypermethylation in acute lymphoid leukemia (ALL), it activates the CDK6-RB1 oncogene
pathway, contributing to poor patient survival (Agirre et al., 2009). It has also been
observed in ALL that CpG islands upstream of snoRNAs SNORD123, U70C, and
ACA59B undergo cancer-specific hypermethylation resulting in their transcriptional
silencing (Ferreira et al., 2012). Furthermore, ribonucleoprotein particles known as vaults
are in part comprised of small untranslated RNAs known as vtRNAs, and although their
function is not yet completely clear, studies have identified aberrant epigenetic regulation
of vtRNAs in diseases such as cancer. Gastric cancer and acute myeloid leukemia (AML)
patients with CpG hypermethylation of the ncRNA nc866, also known as vtRNA2-1, show
poor survival (Lee et al., 2014; Treppendahl et al., 2012). In vitro, knockdown of nc866 in
gastric cell lines leads to the induction of known oncogenes, and overexpression of the
ncRNA reduces cellular proliferation (Lee et al., 2014). In myelodysplastic syndrome
(MDS), both vtRNA1-2 and vtRNA1-3 have been found to be silenced by promoter
methylation, and the hypermethylation of the vtRNA1-3 promoter is associated with a
decreased survival in lower risk MDS patients (Helbo et al., 2015). Finally, a recent study
has detected epigenetic silencing of a partially annotated lncRNA MORT via DNA
hypermethylation to be highly significant for the immortalization of human mammary
epithelial cells. Deficient MORT expression is also common in most cancers and can be
reactivated by 5-Aza-CdR treatment, suggesting a role for this lncRNA in immortalization
during oncogenesis (Vrba et al., 2015). These findings and many others make it clear that
aberrant methylation of key ncRNAs is a fundamental feature of cancer and has a vital
role in disease progression.
23
1.2.2.3 DNA HYPOMETHYLATION
Although CpG hypomethylation was the first methylation change discovered in
cancer, the implication of this dysregulation in tumorigenesis has often been overlooked.
Over three decades ago, Feinberg & Vogelstein, and Gama-Sosa et al., identified a global
decrease in 5mC content across numerous cancer types (Feinberg and Vogelstein, 1983;
Gama-sosa et al., 1983). Hypomethylation can be an early event in tumorigenesis and is
frequently detected in benign hyperplasia. Loss of methylation is also prevalent in late
stage tumors with metastatic lesions possessing greater demethylation than primary
tumors (Li et al., 2014).
DNA methylation is key to suppressing repetitive and transposable elements and
hypomethylation can contribute to ectopic expression of these elements. Long
interspersed nuclear element 1 (LINE1) retrotransposons are mobile genetic elements
responsible for much of the endogenous mutagenesis in humans. The hypomethylation
of the CpG island at the LINE1 promoter stimulates the formation of a permissive
chromatin state at the alternate promoter of MET, thereby activating the transcription of
the oncogene (Wolff et al., 2010). LINE1 promoter hypomethylation has also been
recognized as an indicator of tumor progression and prognosis in several cancer types
including prostate, melanoma, bladder, and renal cancer (Andreotti et al., 2014; Ecsedi
et al., 2013; Karami et al., 2015; Su et al., 2014; Yegnasubramanian et al., 2008). Another
class of repeat elements known as short interspersed nuclear elements (SINEs) is also
similarly regulated by DNA methylation, and studies have observed loss of methylation at
these repeats in acute myeloid leukemia (AML) (Saied et al., 2012).
24
Although hypomethylation of non-CGI promoters is much less frequent than
hypermethylation of promoter CGIs, it can result in the upregulation of oncogenes and
proto-oncogenes (Feinberg and Vogelstein, 1983; Søes et al., 2014). In metastatic non-
small cell lung cancer tumors, for example, the putative oncogene engulfment and cell
motility 3 (ELMO3) gene is significantly overexpressed as a result of its promoter
hypomethylation (Søes et al., 2014). In osteosarcoma, Iroquois homeobox 1 (IRX1) is
upregulated and pro-metastatic. The increase of IRX1 gene expression is found in both
metastatic osteosarcoma cell lines as well as primary patient samples (Lu et al., 2015).
In both cases the gain in expression is associated with hypomethylation of the gene
promoter.
1.2.2.4 DNA METHYLATION AT INTERGENIC AND NON-PROMOTER INTRAGENIC
REGIONS IN CANCER
For decades, much of the research efforts in cancer epigenetics had been
concentrated on the regulation of DNA methylation at gene promoters. Advances in next-
generation, and high-density array sequencing have allowed researchers to expand their
studies on DNA methylation to a genome-wide context. In doing so, it has become
increasingly evident that non-promoter intragenic and intergenic regions are also
dynamically regulated and contribute to physiological changes as well as to the
development of disease states.
25
1.2.2.4.1 DNA METHYLATION CHANGES IN TRANSCRIBED REGIONS
Unlike in promoters, where methylation contributes to a ‘closed’ chromatin architecture
resulting in transcriptional repression, the methylation level in transcribed regions (bodies)
of genes is often positively correlated with gene expression. A recent investigation of
glioblastoma samples revealed functional roles for gene body methylation in affecting
MGMT expression (Moen et al., 2014). The study found that tumors with unmethylated
MGMT promoter and high gene body methylation maintained a high MGMT expression.
As previously mentioned, MGMT expression confers resistance to TMZ therapy.
Consequently, pre-treating glioblastoma cell lines with DNMTi decitabine to reduce
MGMT body methylation significantly sensitized them to the temozolomide treatment
(Moen et al., 2014).
While DNA methylation inhibits initiation of transcription, studies have found it
enables transcription elongation (Kulis et al., 2013; Lou et al., 2014). Furthermore,
methylation in the gene body can also add to transcription efficiency by regulating the
usage of alternative start sites. Global methylome analysis of GBMs suggests a role for
gene body hypomethylation in stimulating the transcription from alternative promoters,
resulting in increased expression of alternative transcripts and some of which can code
for oncogenic protein isoforms (Nagarajan et al., 2014). Finally, a recent large scale
analysis of DNA methylation profiles chronic myeloid leukemia (CLL) subtypes revealed
widespread gene body hypomethylation was enriched in enhancer regions (Kulis et al.,
2012). Thus, loss of methylation in gene bodies can reveal distal regulatory elements
(enhancers), residing within gene bodies, that might have been muted tissue-specifically.
26
1.2.2.4.2 DNA METHYLATION AT ENHANCERS
Enhancers play a significant role in regulating the expression and activity of target
genes. Functional enhancers are decorated with active histone marks including H3K4me1
and H3K27ac. Enhancers serve as a platform for transcription factors (TFs), and through
long range interactions such as ‘looping’, these distal elements are able to deliver
necessary accessory proteins to promoters and stimulate robust transcription. Each
enhancer can regulate the activity of multiple promoters (Bulger and Groudine, 2011).
Although DNA methylation is generally absent in genomic regions with active histone
marks, such as those that delineate active enhancers (Jones, 2012; Kelly et al., 2012;
Lay et al.), expression-associated methylation sites co-localizing with enhancers have
been observed. Methylation at these sites is inversely correlated with gene expression
and can often be better predictors of expression levels than promoter methylation (Aran
and Hellman, 2013; Aran et al., 2013).
Subtle modulation of DNA methylation at enhancers can greatly affect gene
expression of multiple target genes. DNA methylation can thwart the binding of TFs to
DNA, and conversely, the presence of TFs can promote DNA hypomethylation by
preventing DNMTs from accessing DNA (Calo and Wysocka, 2013). Large scale analysis
of cancer and normal tissues have found widespread tumor-specific methylation changes
at enhancer regions (Yao et al., 2015a). Hypomethylation of intergenic and intragenic
enhancers can reveal binding motifs for TFs and induce downstream expression changes
(Aran et al., 2013; Kulis et al., 2013). On the other hand, DNA hypermethylation at
enhancers can decommission them, resulting in a loss of active histone marks and loss
27
of transcription factor binding. Such alterations can modulate gene transcription
independent of promoter methylation fluctuations (Aran et al., 2013; Kulis et al., 2013).
1.2.2.7 CPG ISLAND METHYLATOR PHENOTYPE (CIMP) STRATIFY TUMOR
SUBCLASS
In 1999, Toyota et al noted that a subset of colorectal cancers (CRC) showed
cancer-specific hypermethylation of specific CpG loci. Moreover, this subset of tumors
displayed a concordant hypermethylation of p16, THBS1, and hMLH1 promoters. The
group coined this phenomenon as CpG island methylator phenotype (CIMP). They further
postulated that CIMP contributes to tumorigenesis by concurrently incapacitating multiple
tumor suppressor genes through hypermethylation of their respective CGI promoters
(Toyota et al., 1999). In 2006, Weisenberger and colleagues utilized methylation data
from CRC samples to identify a panel of markers that identified the CIMP-positive tumors.
This subset of CRC tumors robustly correlated with the
v600E
BRAF mutation and
microsatellite instability (Weisenberger et al., 2006). While the molecular basis for onset
of CIMP in CRC is still unclear, several studies have now unequivocally proven its
existence.
One of the first mechanistic insights into CIMP generation in cancer came from
investigating promoter-associated hypermethylation in gliomas. Using data generated by
the cancer genome atlas (TCGA) consortium, Noushmehr et al comprehensively
characterized DNA methylation of GBM tumor and identified a CIMP type that defines a
subset of gliomas (G-CIMP). Interestingly, G-CIMP tumors were tightly associated with
28
high frequency of isocitrate dehydrogenase-1 (IDH1) somatic mutations (Brennan et al.,
2013; Noushmehr et al., 2010). Somatic mutations of IDH1 confer gain of function activity
in the mutant isoform allowing the mutated protein to produce the 2-hydroxygluterate (2-
HG). This onco-metabolite is an inhibitor of the TET family dioxygenases and Jumonji-C
–domain-containing histone lysine demethylases. Thus, production of 2-HG results in the
accumulation of DNA methylation along with aberrant histone methylation (Dang et al.,
2009). More recent studies have shown that the IDH1 mutation alone is sufficient to
establish a hypermethylator phenotype in gliomas and that this hypermethylator status is
retained in both early and late tumors of the same patient, suggesting that CIMP
phenotype is an early event that is likely driving the tumorigenesis (Hill et al., 2014; Turcan
et al., 2012).
Similar to GBMs, AML tumors bear IDH1 and IDH2 mutations as well as TET
mutations. IDH1, IDH2, and TET2 mutations are mutually exclusive, suggesting
redundant activity of the proteins. The TCGA consortium and others have shown that
AML tumors with mutations in IDH proteins or TET enzymes show substantial DNA
hypermethylation (Figueroa et al., 2010; Shih et al., 2012).
To date, several reports have described CIMPs in many additional cancers
including gastric, breast, bladder, melanoma, prostate, hepatocellular, and endometrial
cancer (Weisenberger, 2014). Stratifying cancers into subsets according to DNA
methylation can provide valuable prognostic, diagnostic, and therapeutic insights. In the
case of GBMs, G-CIMP patients tend to be younger in age and have better survival
outcomes than the non-G-CIMP patients. Similarly, Fang and colleagues found that B-
CIMP+ breast tumors were associated with estrogen receptor (ESR1) / progesterone
29
receptor (PGR) positive tumors and the CIMP status was a strong prognosis indicator. B-
CIMP+ patients had a lower risk of metastasis and better clinical survival (Fang et al.,
2011). Recognizing and understanding the onset of the methylator phenotype can thus
help researchers to better strategize therapeutic options.
1.3 NUCLEOSOME POSITIONING
The nucleosome is the basic unit of chromatin. It is formed by wrapping 146 bp of
DNA around a histone octamer comprised of two H2A-H2B dimers and a H3-H4 tetramer
(Luger et al., 1997). The compaction of DNA into chromatin provides organizational
structure, but also keeps the underlying DNA inaccessible to DNA binding proteins. Thus,
nucleosome positioning and occupancy can affect cellular phenotypes by promoting or
preventing processes such as gene transcription, DNA repair, and recombination. For
instance, promoters of actively expressed genes are characterized by a nucleosome
depleted region (NDR). This allows for transcription factors and RNA polymerase to bind
the region and promote gene expression. A nucleosome positioned at the TSS can
prevent the transcriptional machinery from binding the promoter and repress gene
expression (Kelly et al., 2012; Studitsky et al., 1995).
Nucleosome positioning is affected by numerous factors including DNA sequence,
transcription factor binding, histone modifications and variants, and chromatin
remodelers. Certain DNA sequences provide more favorable DNA interactions with
histone proteins, thus influencing nucleosome formation. For example, CG and GC
dinucleotides are preferred by nucleosomes while long A-tracts are disfavored (Struhl and
30
Segal, 2013; Wu and Li, 2010). However, the effect of DNA sequence can often be
overcome by other factors affecting nucleosome placement.
Nucleosome positioning, especially at the promoter, can be affected by
transcription factor binding. Here, TFs and the histone octamer compete for binding to
the DNA, and this competition is dependent on the affinities of the proteins to the DNA.
While the presence of nucleosomes can occlude TFs from interacting with DNA, so too
can a bound TF prevent nucleosome occupancy of the region. TFs can also recruit
chromatin remodeler proteins to actively mobilize nucleosomes (Li et al., 2015; Struhl and
Segal, 2013). Transcription factor binding can establish NDRs, such as in the case of
OCT4 binding to its distal enhancer and at the NANOG promoter to induce NDR formation
(You et al., 2011). Finally, modifications on the N-terminal of histone tails can sway
nucleosome positioning by affecting the DNA-nucleosome interactions, and by recruiting
non-histone proteins such as chromatin remodelers to the region.
The ISWI, CHD, and SWI/SNF families of chromatin remodelers use ATP to
mobilize nucleosomes, modulate chromatin compaction, and in turn affect the
accessibility of DNA. These remodeling complexes are critical to nucleosome sliding,
ejection, or insertion. Chromatin remodeling complexes can also exchange the canonical
histone proteins with histone variants that can enhance or repress transcription (Clapier
and Cairns, 2009). Furthermore, chromatin remodelers act in concert with transcription
factors and histone modifiers to alter chromatin and affect gene regulation (Muchardt and
Yaniv, 1999; Peterson and Workman, 2000; Yaniv, 2014). Thus, chromatin remodeler
proteins are critical to the regulation of gene expression and can affect biological
processes including DNA repair, cell cycle, and cell proliferation.
31
1.3.1 NUCLEOSOME POSITIONING IN CANCER
Recent investigations of whole genomes and exomes of numerous cancers have
revealed that many somatic mutations in tumors are in proteins that facilitate nucleosome
positioning and chromatin remodeling. For example, nearly 20% of all cancers show
mutation in one or more subunits of the SWI/SNF remodeling complex (Kadoch and
Crabtree, 2015). SNF5 is mutated in several cancers including in nearly 100% of all
malignant rhabdoid tumors, and SNF5 loss of function is also observed in metastatic
melanomas, and renal medullary carcinoma (Sansam and Roberts, 2006; Yaniv, 2014).
ARID1A, a subunit of SWI/SNF complex, is frequently altered in ovarian clear cell
carcinoma, gastric cancer, and hepatocellular carcinoma (Huang et al., 2012; Matsuzaki
et al., 2015; Skulte et al., 2014; 2012). Mutations in PBRM1, yet another SWI/SNF
complex member, occur in nearly 30% of clear cell renal cell carcinoma (Creighton et al.,
2013; Peña-Llopis et al., 2012).
Mutations and deregulated expression of subunits of the ISWI complex have been
detected in several cancers including prostate cancer, lymphomas, and serous epithelial
ovarian cancer (Skulte et al., 2014). Sequencing of gastric and colorectal cancer
genomes has uncovered mutations and loss of expression of the CHD4 ATPase as well
as mutations in CHD7. Further, CHD4 has been implicated as a tumor suppressor in
ovarian and prostate cancer (Andreu-Vieyra and Liang, 2013).
32
1.4 HISTONE MODIFICATIONS
A variety of post-translational modifications (PTMs) occur on the tails of histones and
studies have shown that these PTMs can affect chromosomal compaction. Condensed
heterochromatic regions and open euchromatic regions are decorated with distinct PTMs
(Murr, 2010). Unlike, DNA methylation, histone modifications are much more pliable and
dynamically regulated. Numerous enzymes responsible for placing, binding, or removing
histone modifications such as methylation, acetylation, phosphorylation, sumoylation,
glycosylation, and ubiquitination have been identified (Bannister and Kouzarides, 2011).
These include histone methyltransferases (HMTs), histone demethylases (HDMTs),
histone acetylases (HATs), and histone deacetylases (HDACs) (Gardner et al., 2011;
Sneppen and Dodd, 2012).
While the biological significance of many histone tail PTMs have yet to be
determined, combinations of PTMs at regulatory elements have been correlated with
functional consequences (Strahl and Allis, 2000). Histone modifications can influence
chromatin accessibility by either altering the interactions between nucleosomes and DNA
or by recruiting other proteins such as chromatin remodelers to modulate the
nucleosome-DNA interaction. For example, the addition of an acetyl group on the lysine
residues of histone tails removes the positive charge on the histone and thereby lowers
the affinity of the DNA to the histone (Struhl, 1998). In turn, this can make the DNA more
accessible to transcription factors. Therefore, acetylation is associated with accessible
DNA and active genes. Furthermore, proteins with bromodomain can recognize and bind
to the acetylated lysine and trigger recruitment of other factors affecting gene transcription
33
(Murr, 2010; Strahl and Allis, 2000). In contrast, methyl residues on histone tails do not
impart any change in charge, however, proteins with PHD domains can bind methylated
lysines (Jones and Liang, 2009; Ruthenburg et al., 2011). Methylated histones are found
in active and inactive regions of the genome. Some of these modifications include: tri-
methylation of H3 lysine 4 (correlated with active genes), tri-methylation of lysine 36 on
H3 (present in the body of actively transcribed genes), tri-methylation at H3 lysine 9
(closely associated with heterochromatin), and the polycomb repressive mark H3 lysine
27 tri-methylation (indicative of gene suppression) (Calo and Wysocka, 2013; Du et al.,
2015; Ruthenburg et al., 2007; Sawan, 2010).
1.4.1 HISTONE MODIFICATIONS IN CANCER
DNA replication, repair, and transcription hinges on precise regulation of DNA
structure. Dynamic control of histone modifications is critical to the stability of DNA
structure and regulation of biological processes. Thus, aberrations in this tightly controlled
epigenetic process can contribute to malignancies such as cancer.
As with DNA methylation mediators and nucleosome remodelers, histone
modifiers have also been identified as frequently mutated or abnormally expressed in
various cancers. Studies have noted overexpression of several HDACs in acute lymphoid
leukemia (Moreno et al., 2010; Sonnemann et al., 2012; Zhang et al., 2015a). Increased
expression of HDACs can decrease acetylation at promoters of tumor suppressors, and
thus affect the expression of these key genes. Similarly, upregulation of the polycomb
repressive complex members including the catalytic subunit EZH2 is seen in several
34
cancer types including OCCC, ALL, prostate, and breast (Chase and Cross, 2011; Han
Li and Chen, 2015; Kim et al., 2015a; McCabe and Creasy, 2014; Shih et al., 2012). The
histone demethylase LSD1, can demethylate both the active mark H3K4me2 and the
repressive mark H3K9me2 (Piao et al., 2015; Sharma et al., 2010). Thus, LSD1 mutations
can both contribute to activation and repression of target genes. Increased expression of
LSD1 has been observed in numerous cancers including neuroblastoma, colon, prostate,
and breast cancer (Amente et al., 2015; Ding et al., 2013; Hayami et al., 2011; Lim et al.,
2010; Metzger et al., 2005; Piao et al., 2015).
1.5 THE EPIGENOME AS A THERAPEUTIC TARGET
Epigenetic aberrations in cancers including differential DNA methylation can be
used to distinguish tumor subtypes, indicate treatment responsiveness, predict clinical
outcomes, and determine therapeutic strategies. Epigenetic profiles can reveal molecular
pathways most vulnerable to chemotherapeutic agents, and methylation changes can
often serve as a barometer for treatment efficacy (Kelly et al., 2010a).
Unlike genetic modifications, epigenetic modifications are both somatically
heritable and potentially reversible. Thus, finding pharmacological interventions aimed at
rescuing an altered epigenome is highly desirable and can have long-lasting impact.
Furthermore, because cancer cells can become addicted to their epigenetic aberrations,
targeting the epigenome can have the added benefit of increasing the sensitivity of the
tumor to standard chemotherapy regimen (Mair et al., 2014).
35
To this end, DNMTi have been successfully employed in preclinical and clinical
settings with the goal of eliminating aberrant methylation (Juo et al., 2015; Yamazaki and
Issa, 2013). DNMTi, such as the cytidine analogs 5’-Aza-2’-deoxycytidine (5-Aza-CdR)
and decitabine, incorporate into DNA during replication and are recognized as natural
substrates by DNMTs. DNMT initiates the methylation reaction by covalently binding
DNA. The resolution of this covalent bond is impeded by Aza-cytosine, resulting in the
sequestration of DNMTs to DNA. This in turn compromises the integrity of the DNA
molecule and triggers a DNA damage response. The subsequent proteomic degradation
of the bound DNMT results in the loss of methylation marks (Christman, 2002;
Stresemann and Lyko, 2008).
5’-Azacytidine (5-Aza-CR) is currently FDA approved to treat high-risk MDS
patients and has resulted in successful clinical outcomes (Fenaux et al., 2009). Pre-
clinical data are also available for other cytidine analogs, such as S110 which shows
better stability and activity relative to 5-Aza-CdR (Chuang et al., 2010; Yoo et al., 2007).
Treatment by DNMTi can sensitize cancers to other chemotherapeutic agents. For
example, in hepatocellular carcinoma cells, SGI-110 significantly synergized with
oxaliplatin resulting in greater cytotoxicity (Kuang et al., 2015). DNMTi can also prove
to be immunomodulatory. Hypomethylation induced in epithelial ovarian carcinoma cells
upon treatment with SGI-110 results in increased expression of cancer-testis antigens,
thereby, enhancing the recognition of EOC cells by antigen-specific CD8+ T-cells. This
contributes to restricted tumor growth and better survival in mouse xenograft models
(Srivastava et al., 2015).
36
Better understanding of the role of intergenic methylation in the recent years has
led researchers to realize that in addition to promoter methylation, gene body methylation
might also serve as a therapeutic target for demethylating chemotherapeutic agents such
as 5-Aza-CdR (Figure 1.1b, 1.1c). A study from our group on the effect of low-dose
transient treatment with 5-Aza-CdR from on gene body methylation confirmed that loss
of methylation from gene body correlated with loss of gene expression, and the rate of re-
methylation after drug withdrawal determined the strength of re-expression. By clustering
the genomic regions into groups according to the rates of re-methylation, the study noted
that rapidly re-methylating genes were enriched for oncogenic genes such as c-MYC
targets and metabolic pathway genes. Thus, a potential mechanism of action for DNA
methylation inhibitors could be through mitigating the effect of deregulated c-MYC
(Kasinathan and Henikoff, 2014; Yang et al., 2014).
The effects of DNMT inhibitors are diverse, and therapeutic responses have a slow
onset. Additionally, low-doses of DNMTi are sufficient for long-lasting loss of
tumorigenicity and self-renewal with minimal cytotoxic effect. All of this indicates that
supplementary to acute re-expression of tumor suppressor or downregulation of crucial
oncogene, other mechanism(s) must exist by which DNA methylation inhibitors can target
methylation (Licht, 2015; Oki et al., 2007; Tsai et al., 2012).
Recent investigations have shown that demethylating agents might be mediating
therapeutic response by rendering the cell more visible to the immune system
(Chiappinelli et al., 2015; Roulois et al., 2015). Specifically, demethylating agents are able
to trigger induction of an anti-viral immune response by permitting the expression of
endogenous retroviruses (ERVs) that had previously been silenced by DNA methylation
37
(Figure 2). Roulois et al determined that the majority of the late occurring (24 days past
initial exposure) expression changes upon transient low-dose 5-Aza-CdR treatment of
colorectal cell lines, were of interferon-responsive genes. These genes showed little
modulation of methylation at their promoters or coding region, and in fact, many of them
displayed low DNA methylation levels pre-treatment. Thus, it can be interpreted that the
change in gene expression upon treatment with 5-Aza-CdR is independent of the drug’s
capacity to demethylate the respective gene promoters. A series of genetic experiments
showed that the activation of these genes was through cytosolic pattern recognition
receptors which are primarily responsible for viral RNA detection. This then initiates a
signaling cascade dependent on the mitochondrial antiviral signaling (MAVS) adaptor
molecule, leading to the activation of downstream targets, such as IRF7, and culminating
in a strong anti-tumor response. The study found that 5-Aza-CdR induced a significant
increase of dsRNAs including a robust induction of endogenous retrovirus RNA
transcription (Roulois et al., 2015). Another group working with ovarian cancer cell lines
came to a similar conclusion that treatment with 5-Aza-CdR triggered the upregulation of
interferon signaling mediated by downstream activity of IRF7. Furthermore, the strength
of interferon response to the drug treatment was reflective of how well the tumor would
respond to the immune checkpoint therapy (Chiappinelli et al., 2015). Thus, a major mode
of action of DNMT inhibitors such as 5-Aza-CdR is that the loss of DNA methylation at
previously silenced elements, such as ERVs, and the subsequent induction of dsRNA
transcription trigger a strong anti-viral response. As a consequence, there is an overall
anti-tumoral affect including interferon induction, reduced cell proliferation, and loss of
self-renewal capacity upon treatment (Figure 1.2).
38
Epigenetic mechanisms often regulate each other, therefore pharmacological
targeting of one mechanism can confer an effect on the other epigenetic processes. The
chromatin remodeler protein CHD5 is considered a tumor suppressor in many cancer
types and is frequently silenced through multiple epigenetic mechanisms including
promoter hypermethylation (Fujita et al., 2008; Gorringe et al., 2008; Wang et al., 2011a).
A study performed in a colorectal cancer model found that treatment with 5-Aza-CdR
partially restored CHD5 protein expression (Fatemi et al., 2014).
Drugs targeting aberrant histone modifications have also been employed in the
fight against cancer. The histone deacetylation inhibitor (HDACi) suberoylanilide
hydroxamic acid (SAHA, vorinosat) can activate epigenetically silenced tumor
suppressors and has been used in treatment of cutaneous T-cell lymphoma (Olsen et al.,
2007). In vitro and murine studies have noted synthetic lethality to EZH2 inhibition in
cancers with mutations in DNA repair genes such as BRCA1 or chromatin remodeler
genes such as ARID1A (Bitler et al., 2015; Chase and Cross, 2011; Han Li and Chen,
2015; Mair et al., 2014). Finally, several in vitro studies have begun targeting multiple
epigenetic processes using combinatorial therapeutic strategies with promising results.
EZH2 inhibition along with HDACi in breast cancer cell lines and AML primary cells
displayed synergistic effects in reducing cell proliferation, and decreasing cell viability
(Chase and Cross, 2011). Similarly, a study on cancer cell lines using DNMTi in
combination with LSDi showed greater efficacy in restricting cell growth and colony
formation than either of the agents used separately (Han et al., 2013).
39
1.6 CONCLUSION
The intricate interactions between epigenetic mechanisms such as DNA methylation,
histone modifications, and nucleosome positioning facilitate gene expression. The proper
execution of cellular and biological processes including cell cycle, DNA repair, cell growth,
and proliferation hinges on the precise coordination of these mechanisms. Often during
tumorigenesis, aberrancies in these epigenetic processes can lead to inappropriate
silencing of tumor suppressors or expression of oncogenes. In contrast to genetic
mutations, epigenetic changes can potentially be reversed through pharmacological
intervention. Thorough understanding of the complexities governing the regulation of the
epigenome and the interplay between the genome and the epigenome will enhance our
abilities to effectively target cancer. Genome-wide screens can be efficiently used to
identify genes that are influenced by the pathways being affected by epigenetic
aberrations. Furthermore, with improved access to next generation sequencing, large-
scale multinational consortia have conducted research resulting in a wealth of genomic
and epigenomic data. Integrating this information with patient profiles will enable
researchers to validate putative therapeutic epigenetic targets, as well as stratify tumors
into clinically relevant subgroups according to their methylation status, thereby facilitating
the design of more effective therapeutic strategies.
40
Figure 1.1 DNA methylation equilibrium between promoter and gene body
modulates gene expression. In this diagram, methylated CpG sites are represented by
red circles, unmethylated CpG sites are represented by white circles, green arrows are
indicative of active expression, and red arrow marks the absence of expression. (a) In
normal mammalian tissue, genes that are actively transcribed have unmethylated
promoters and some methylation in the gene body. (b) With the onset of cancer, promoter
hypermethylation can turn off expression of genes and gene body hypermethylation can
permit robust expression of some genes. (c) Treatment with DNA methylation inhibitors
such as 5-Aza-CdR can restore gene expression by removing aberrant methylation.
41
Figure 1.2 DNMTi exert anti-tumoral effect by eliciting immune response in cancer
cells. Treatment with DNA methylation inhibitors induces transcription of endogenous
retroviral (ERV) elements. These double stranded RNAs are recognized by viral
recognition proteins such as RIG1 and MDA5 which in turn interact with the mitochondrial
anti-viral signaling (MAVs) proteins. MAVs mediated IRF7 activation leads to the
translocation of IRF7 from the cytoplasm to the nucleus where it initiates transcription of
interferons (IFNs), and interferon-stimulated genes (ISGs), which then contribute to
reduced proliferation (Modified from Chiappinelli et al. 2015; Licht 2015; Roulois et al.
2015).
42
OVERVIEW OF THE THESIS WORK
Advances in high-throughput sequencing techniques have made it possible for
large scale exploration of genomes, epigenomes, and transcriptomes of cancer and
normal tissues. Remarkably, it has become evident that a sizeable portion of all genetic
mutations occur in genes coding for epigenetic modifiers. Epigenetic mechanisms such
as DNA methylation, nucleosome positioning, and histone modifications are critical to
the proper maintenance of genomic integrity and cellular identity. These mechanisms
co-regulate each other and contribute to the precise control of gene expression
potential. It is now accepted that mutations in epigenetic modifiers can disrupt cellular
processes and affect cell identity, however, it remains unclear how mutation in one
mechanism can influence the overall epigenetic landscape genome-wide and cellular
phenotype. My graduate dissertation examines the role of non-catalytic protein
components of the SWI/SNF chromatin remodeler complex on the epigenome and its
consequence on cellular phenotype.
In chapter 2, I investigate the role of a core SWI/SNF member, SNF5, in the
maintenance of the oncogenic landscape of malignant rhabdoid tumors. Using an
inducible system that expresses SNF5 upon doxycycline administration, I identify genes
that are altered in MRT due to the absence of SNF5 and re-established upon SNF5
expression. Additionally, using the DNA methylation array-based footprinting assay,
AcceSssIble, I ascertain epigenetic networks driving these gene expression changes.
In chapter 3, I explore the capacity of ARID1A on effecting functional alterations
using an ARID1A-knockdown endometriosis cell line model. Using this model, I provide
43
evidence that ARID1A mutation can be an early-stage event in the transformation of
endometriosis cells and potentially giving rise to ovarian clear cell carcinoma (OCCC).
To this end, I employ cellular assays characterizing invasion, adhesion and colony
formation to understand the functional outcome of the down-regulation of ARID1A.
Using expression microarray, I identify gene expression deregulation due to loss of this
remodeler protein. I address the epigenetic consequence of ARID1A loss and its
relation to the phenotypic and expression changes observed using ChIP-seq to and
NOMe-seq, a DNA footprinting technique to simultaneously determine DNA methylation
and nucleosome occupancy. Finally, in chapter 4, I briefly summarize my work and
consider how my findings have contribute to the field of cancer epigenetics.
44
CHAPTER 2: INDUCTION OF SNF5 IN MALIGNANT RHABDOID TUMOR CELLS
RESCUES THE ABERRANT EXPRESSION IN PART THROUGH EPIGENETIC
MODULTATION
I performed all the experiments, and bioinformatics analyses unless specified below:
Cell lines transfected with inducible transgene were established and/or provided by the
lab of Bernard Weissman (BW) at UNC-Chapel Hill. The USC Epigenome Center
(Charlie Nicolet, Selene Tyndale, and Helen Troung) prepared and sequenced RNA-seq
libraries. USC NML Bioinformatics Services (Meng Li, and Yibu Chen) provided
assistance in mapping the RNA-seq data and identifying differentially expressed genes.
Dan Wiesenberger (DW) ran the samples on HM450K array and performed array
background normalization and beta-value extraction.
I am responsible for all the writing in this chapter and have incorporated edits provided
by Peter A Jones (PAJ) and Gangning Liang (GL). Experiments were conceived by me,
PAJ, GL, and BW.
2.1 INTRODUCTION
Malignant rhabdoid tumor (MRT) is a rare and highly aggressive pediatric cancer
and primarily affects the brain, kidney, and soft tissues (Biegel et al., 2014; Frühwald et
al., 2016; Haas et al., 1981). Even with advances in therapeutics, prognosis for MRT
patients remains dire. The overall survival rate for this cancer is between 20%-25% and
upon metastases the rate of survival falls to 11% (Reinhard et al., 2008). Thus, a better
45
understanding of MRT and the molecular underpinnings of the pathology is urgently
needed.
Remarkably, MRT has a stable diploid genome with the exception of mutations
and deletions at the SMARCB1/INI1/BAF47/hSNF5 locus, referred hereafter as SNF5
(Biegel et al., 1999; McKenna et al., 2008). Mouse models have confirmed that deletion
of SNF5 is necessary and sufficient for the rapid onset of MRT. Thus SNF5, a core
component of the SWI/SNF chromatin remodeling complex, has been validated as a bona
fide tumor suppressor in MRT (Biegel et al., 1999; Guidi et al., 2006; Kim and Roberts,
2014; McKenna et al., 2008).
The SWI/SNF complex uses ATP to mobilize nucleosome, modulate chromatin
compaction, and regulate accessibility of DNA (Clapier and Cairns, 2009; Euskirchen et
al., 2011; Peterson and Workman, 2000; Segal and Widom, 2009). By adjusting the
accessibility of DNA to transcription factors, DNA binding proteins, and the transcription
machinery, SWI/SNF complexes can influence numerous cellular activities critical to
development, differentiation, and the maintenance of genomic integrity (Euskirchen et al.,
2011; Muchardt and Yaniv, 1999; Yaniv, 2014).
Until very recently, the causative role of epigenetic mechanisms such as chromatin
remodeling in oncogenesis was largely unrecognized. The advent of next-generation
sequencing technologies has made it possible to analyze entire genomes and exomes of
numerous cancers. These efforts have revealed that genes coding for epigenetic
modifiers often carry somatic mutations in tumors and the subunits of the SWI/SNF
complexes are some of the most commonly mutated (Andreu-Vieyra and Liang, 2013;
46
Biegel et al., 2014; Masliah-Planchon et al., 2015; Sharma et al., 2010; You and Jones,
2012). In fact, studies have shown that up to 20% of all cancers display mutations in at
least one of the SWI/SNF remodeler complex members (Kadoch and Crabtree, 2015).
Understanding how these epigenetic modulators can shape the epigenome and permit
oncogenesis is critical to designing effective therapies.
Considering the paucity of somatic mutations and the near-universal loss of SNF5
in MRT, I hypothesized that modulating the expression of a core SWI/SNF component
would disrupt chromatin organization. Therefore, in this study I aim to identify epigenetic
driver genes and gene networks mediating the effects of SNF5 in MRT. Using an inducible
SNF5 system in TCC642, a cell line derived from MRT, I first characterized alterations to
the transcriptome upon SNF5 expression. Next, I successfully correlated these findings
with changes seen in primary MRT. I then surveyed the epigenome for DNA methylation
and chromatin accessibility changes using the DNA methylation array based AcceSssIble
assay and uncovered several potential direct epigenetic targets of SNF5. Finally, I
established TRIM2 as a down-stream target of SNF5 and showed that it is positively
associated with patient survival. Together, these findings confirmed that induction of
SNF5 expression in TCC642 cells regulated some target gene expression by remodeling
chromatin architecture.
47
2.2 METHODS
2.2.1 Cell culture
Flag tagged SNF5 was amplified from an existing plasmid, pcDNA3-fSNF5, by
PCR (T7 promoter F primer, hSNF5 ORF R primer – TTA CCA GGC CGG CGT GTT)
(Chai et al., 2005) and the resulting PCR product was TOPO-cloned into a Gateway
vector using the pCR®8/GW/TOPO® TA cloning kit (45-0642, Invitrogen). The fSNF5
was transferred into the pINDUCER20 vector that provided a c-terminal HA tag using
Gateway® LR Clonase ™ II Plus Enzyme Mix (12538-120, Invitrogen) (Meerbrey et al.,
2011). The identity of the fSNF5HA insert was confirmed by DNA sequencing. The cell
lines were cultured in RPMI1640 media supplemented with 10% FBS, at 37°C and 5%
CO
2
(Kuwahara et al., 2013; Wei et al., 2014). 1 μg/ml doxycycline was used to induce
maximal SNF5 expression and experiments were performed on cells isolated at the 24-
hour mark.
2.2.2 Protein extraction and western blot analysis
Cells were rinsed with PBS, trypsinized, and resuspended in RIPA buffer (50 mM
Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% DOC, 0.1% SDS) supplemented with
protease inhibitors (Roche, 04693132001). The cells were then briefly sonicated on ice
and cellular debris was removed by centrifugation. Bradford reagent (Bio-Rad) as used
to quantify protein concentration and 40ug of protein was mixed with SDS/β-
mercaptoethanol loading buffer. Proteins were resolved on a 4-15% gradient
48
SDS/PAGE gel (Bio-Rad). Antibodies against HA (Sigma, H3663, 1:1000) and β-ACTIN
(Sigma, A2228, 1:2,000) were used.
2.2.3 RNA extraction and analysis
Total RNA was extracted from two independent cell cultures in Trizol and purified
using Direct-zol RNA MiniPrep (Zymo Research). Libraries were constructed using the
poly-A selected method and sequences were generated at USC Epigenome Center.
Partek Genomic Suite licensed through the USC NML Bioinformatic Services was used
to map sequences to the genome, filter out non-unique/PCR duplicates, and determine
differentially expressed genes (Partek).
For quantitative PCR (qPCR) analysis, total RNA was treated with DNaseI
(Invitrogen), followed by EDTA and heat inactivation. DNase treated RNA was reverse-
transcribed using iScript Reverse Transcription Supermix (Bio-Rad). QPCR reactions
were performed using KAPA SYBR FAST Universal master mix. Genes of interest were
normalized to PCNA and TBP (Table 2.1).
2.2.4 NOMe-seq and AcceSssIble assays
Nucleosome-occupancy and DNA methylome sequencing (NOMe-seq) was
performed as previously described in (Lay et al.; Kelly et al. 2012) and AcceSssIble assay
was performed as previously described in (Pandiyan et al. 2013).
49
Briefly, 250,000 exponentially growing cells were washed with PBS, trypsinized
and lysed 5 minutes on ice in 1mL of ice-cold nuclei-extraction buffer (10 mM Tris, pH7.4,
10mM NaCl, 3 mM MgCl2, 0.1 mM EDTA and 0.25% NP-40). The extracted nuclei were
washed with 500 μL of ice-cold wash buffer (10 mM Tris, pH7.4, 10 mM NaCl, 3 mM
MgCl2, and 0.1 mM EDTA).
For NOMe-seq, the nuclei were then resuspended in 1X GC reaction buffer (New
England Biolabs). The nuclei were supplemented with sucrose, and 1.5uL of S-
adenosylmethionine (SAM) (New England Biolabs) and treated with 200 units of
M.CviPI enzyme (New England Biolabs) for 7.5 minutes at 37°C followed by a boost of
100 unites of M.CviPI enzyme and 0.75 μL of SAM for 7.5 additional minutes. Reactions
were terminated by adding stop solution (10 mM Tris-HCl (7.9), 600 mM NaCl, 1% SDS,
and 0.1 mM EDTA) and incubated with proteinase K at 55⁰C overnight. Genomic DNA
was then isolated by phenol-chloroform extraction and ethanol precipitation. Zymo’s EZ
DNA methylation kit was used to bisulfite convert 1 μg of genomic DNA. To analyze the
methylation and accessibility status of individual DNA molecules at regions of interest,
bisulfite converted DNA was PCR amplified with primers listed in Table 2.1 and cloned
into the pCR4.1 vector using the TOPO-TA cloning kit (Invitrogen, Carlsbad, CA).
Individual colonies were screened for the insert and the region of interest was
sequenced using M13 primers (Genewiz).
For AcceSssIble, the nuclei were resuspended in 1X NEB 2 buffer (New England
Biolabs). Isolated nuclei were supplemented with sucrose, 1μL of SAM (New England
Biolabs) and treated with either ddH2O (no enzyme control) or with 50 units of M.SssI
50
enzyme (New England Biolabs). Reactions were terminated by adding equal volume of
stop solution (10 mM Tris-HCl pH7.9, 600 mM NaCl, 1% SDS, 0.1 mM EDTA) and
incubated with Proteinase-K at 55°C overnight. Genomic DNA (gDNA) was isolated by
phenol-chloroform extraction and ethanol precipitation. EZ DNA methylation (Zymo
Research) kit was used to bisulfite-convert 1 μg of gDNA. Previously described
Methylight-based quality control tests were used to determine conversion completeness
(Campan et al., 2009). DNA methylation was assayed using Infinium
HumanMethylation450 (HM450) BeadChip array at the USC Epigenome Center
according to manufacturer’s protocol.
2.2.5 Infinium HumanMethylation450 BeadChip data processing and AcceSssIble
data analysis
HM450 data processing and accessible data analysis were performed as
described in previously published work (Becket et al., 2016; Pandiyan et al., 2013).
Probes with a detection p-value >0.05, located within 15 bp of a SNP, overlapping
repetitive elements, mapping to multiple locations, or located on the X and Y-
chromosomes were excluded from the analysis. After filtering, 385,826 probes remained
for downstream analysis. CpG sites with a change of 0.2 (20%) in β-value were
considered to be altered in DNA methylation.
CpG accessibility was determined by subtracting the endogenous DNA
methylation of CpG sites (DNA methylation of the No Enzyme control sample) from the
51
DNA methylation of CpG site from the M.SssI treated sample. CpG sites with 0.2 (20%)
change in CpG accessibility were considered to be altered in chromatin accessibility.
2.2.6 Computational analysis
Data analysis and visualization was performed using the statistical language R,
and the package ggplot. OncoLnc was used to access TCGA level 3 RNA-seq expression
data, clinical data, and generate survival curves (www.oncolnc.com). Normal kidney
tissue and MRT primary tissue expression microarray data were obtained through GEO
(GSE15641, and GSE11482). The microarray data files from previously published data
were normalized using gcrma. Pathway analysis was performed using Ingenuity Pathway
Analysis (IPA®, QIAGEN Redwood City, www.qiagen.com/ingenuity).
52
Primer Use Sequence
SMARCB1-FWD Expression 5’- GACGACGGCGAGTTCTACAT
SMARCB1-REV Expression 5’- TAGTCGCCTCCAGAGTGAGG
TRIM2-FWD Expression 5‘- TTTGCAGGTCCCCATTTTGC
TRIM2-REV Expression 5‘- TCCACTGCTACACCTGTTGG
PCNA-FWD Expression 5’- GCAGATGTACCCCTTGTTGTAGAGT
PCNA-REV Expression 5’- TCTTCATCCTCGATCTTGGGA
TBP-FWD Expression 5’- GAGGTGGGGATTTTGTT
TBP-REV Expression 5’- CTCCTCCCTCCAATAA
TRIM2-F NOMe-seq 5’- GATATTTAGAAATTTAGTAGAAA
TRIM2-R NOMe-seq 5’- CTATAATTAAAAAAACAAAAAAAA
Table 2.1 Gene expression and NOMe-seq primer sequences
53
2.3 RESULTS
2.3.1 TCC642: MRT cell line with inducible SNF5 expression
The MRT cell line TCC642 lacks endogenous SNF5 expression and has been
stably transfected with a transgene expressing SNF5 under the control of a tetracycline
regulatory element (TRE) (Figure 2.1A) (Wei et al., 2014). Strong induction of RNA and
protein expression of SNF5 were observed 24hrs after the administration of doxycycline
(Figure 2.1B). As noted by previous studies, expression of SNF5 triggered cellular
senescence in the MRT cells (Kuwahara et al., 2013; Reincke et al., 2003; Wei et al.,
2014). Canonical senescence associated chromatin architecture alterations are observed
3-9 days after initial growth arrest (Aird and Zhang, 2013). Therefore, all the experiments
in this study have been performed at the 24hrs mark after SNF5 induction to minimize
capturing senescence-specific changes.
54
Figure 2.1. SNF5 mRNA and protein induced in TCC642. (A) SNF5 expression is under
the control of a tetracycline regulatory element (TRE). At 0hr there is no endogenous
expression of SNF5 and upon doxycycline (dox) administration, the HA-tagged SNF5
transgene is transcribed. (B) Treatment with doxycycline for 24hrs is sufficient for strong
induction of SNF5 mRNA and protein.
55
2.3.2 Gene expression changes due to SNF5 induction reflects normal tissue
expression profile
To understand the functional role of SNF5 in MRT, I compared the transcriptome
of two biological replicates harvested before doxycycline treatment (0hr, no SNF5) and
24hr post-doxycycline treatment (with SNF5). By performing RNA-seq experiments on
these samples, I found 662 genes to be differentially expressed (fold change > 2, FDR <
0.1) (Figure 2.2A). The majority of the genes altered in expression were up-regulated
(569) while 92 genes were down-regulated upon induction of SNF5 (Figure 2.2A). Earlier
studies have noted that SNF5 re-expression induces cell cycle arrest through the
activation of p21 and p16 (Kuwahara et al., 2013; Vries et al., 2005; Wang et al., 2014).
Although the upregulation of p21 and p16 did not meet the FDR cut-off set for this study,
I confirmed a similar trend of increased expression (Table 2.2).
SNF5-positive cells upregulated expression of several key regulators of cellular
phenotypes. These included tumor suppressors such as APC, RASSF8, and FOXO3, cell
fate and differentiation factors such as HEY1, KLF4, TGFB1 & TGFB3, nuclear receptors
such as EGFR. On the other hand, genes downregulated in SNF5 expressing cells were
often amplified or upregulated in tumors such as CCNE1, and TERT.
E2F family members, along with RB, are the primary regulators of cell-cycle. E2F
1-3 are transcriptional activators (promote cell cycle progression) while E2F 4-8 are
transcriptional inhibitors (antagonize cell cycle progression) (Johnson et al., 2012; Nevins,
2001). Interestingly, I found that gene expression of both activators such as E2F2, and
E2F3 as well as inhibitory E2F members such as E2F4, E2F5, and E2F8, tended to be
56
down regulated in the SNF5-induced cells (Table 2.2). However only the down-regulation
of the transcriptional activator E2F2 met the 2-fold change cut-off set for this experiment.
The down-regulation of activating E2Fs, especially E2F2, could be mediating the growth
arrest triggered by SNF5 induction. On the other hand, the down-regulation of inhibitory
E2Fs could be a reflection of heterogeneity of SNF5 expression in the cell population.
Next, to ensure that I am identifying expression alterations relevant to those
observed during MRT oncogenesis, I validated the gene expression findings in a cohort
of normal kidney and primary MRT samples. Using previously published normal and MRT
expression microarray data (Gadd et al., 2010; Jones et al., 2005), I confirmed that most
of the genes up-regulated upon SNF5 induction in TCC642 were more expressed in
normal tissue than in MRTs and genes down-regulated upon SNF5 induction were less
expressed in normal than in MRTs (Figure 2.2B). Thus, expressing SNF5 in the MRT cell
line TCC642, re-calibrated the mRNA levels of many genes that are aberrantly expressed
in primary MRTs.
57
Figure 2.2. Gene expression changes triggered by SNF5 induction in TCC642 are
consistent with deregulation observed in MRT primary tumors. Heat-maps plotting
row-normalized expression z-scores. The color scale ranges from green to red (-4 to 4)
where green indicates lower expression and red indicates higher expression (A) RNA
expression data of TCC642 cells at 0hr (no SNF5) and 24hr (with SNF5) after dox
administration. Only genes with minimum 2-fold change in expression and a FDR < 0.1
are shown. (B) Expression data of primary normal and MRT tissue were obtained from
previously published Affymetrix microarray experiments. Heatmap displays row-
normalized expression of genes differentially expressed in TCC642 (as described in
2.2A). The top panel displays genes that were upregulated in TCC642 with SNF5
expression and the bottom panel shows genes that were downregulated in TCC642 with
SNF5 expression.
58
Gene Fold Change p-Value FDR
CDKN1A (p21) 1.25803 0.05 0.15
CDKN2A (p16) 1.35201 0.05 0.14
E2F1 -1.5768 0.0288785 0.109381
E2F2 -2.79904 0.0035916 0.060748
E2F3 -1.06097 0.0099083 0.074666
E2F4 -1.25331 0.0103209 0.075276
E2F5 -1.4002 0.0121657 0.078359
E2F6 -1.32619 0.0374976 0.123216
E2F7 1.13159 0.360495 0.49692
E2F8 -1.64976 0.0141639 0.083381
SMARCB1 (SNF5) 8.70529 0.0001 0.05
Table 2.2 RNA-seq differential expression results of p21, p16, E2F gene family
members, and SNF5 in TCC642 with SNF5 induction.
59
2.3.3 Gene expression alterations rectify the molecular pathways and regulatory
networks misfiring in MRTs
To gain further insight into the gene networks hijacked by cancer and potentially
restored by the induction of SNF5, I conducted pathway analyses of the differentially
expressed genes. Many of the pathways significantly associated (FDR < 0.05) with genes
up-regulated upon SNF5 expression were involved in cellular assembly, organization, and
signaling. These included integrin, paxillin, FAK, and MAPK/ERK signaling pathways
(Figure 2.3A). These pathways often contribute to cellular functions critical in oncogenesis
such as cell adhesion, cell proliferation, and invasion (Behmoaram et al., 2008;
Hildebrand et al., 1995; Macagno et al., 2014). In addition, peroxisome proliferator-
activated receptor (PPAR) pathway which regulates differentiation, metabolism and
immunity as well as several cytokine-activated pathways such as tissue factor,
macrophage pathways were also affected. Interestingly, the majority of the altered
pathways were upstream of the MAPK/ERK with the exception of PPAR which can be
suppressed by MAPK activity (Behmoaram et al., 2008; Burgermeister and Seger, 2007;
Hildebrand et al., 1995; Papageorgiou et al., 2007). Additionally, many of the upstream
regulators identified are known to be critical effectors of oncogenesis and some have
been previously recognized to be modulated by SNF5 activity. These included TGFB1,
ERRB2, HNF1B, and the SWI/SNF ATPase SMARCA4 (Brg1) (Amano et al., 2015;
Brenca et al., 2013; Sun et al., 2016a; Wei et al., 2014) (Figure 2.3B).
On the other hand, pathways associated (FDR <0.05) with genes down-regulated
upon SNF5 expression were critical to cell cycle regulation. These pathways include
60
estrogen-dependent S-phase entry and G1/S checkpoint regulation (Figure 2.3C). The
upstream regulators of these down-regulated genes were largely RB, RB-like
transcription factors, and the E2F family of transcription factors (Figure 2.3D). E2F2 is
one of the upstream regulator identified and this gene was significantly down-regulated
in TCC642 upon SNF5 induction (Table 2.2). Furthermore, these findings are highly
consistent with previous studies that have found inactivation of SNF5 results in
upregulation of RB-E2F related genes (Guidi et al., 2006; Isakoff et al., 2005).
Thus, pathway analyses showed that the introduction of the epigenetic regulator
SNF5 in TCC642 resulted in differential expression of genes closely associated with
cellular processes such as cell proliferation, metabolism, differentiation, and
oncogenesis. Notably, the MAPK/ERK pathway could be a key mediator of these
perceived alterations. These findings are in line with the biological function of SNF5 in
cell cycle, growth, and differentiation (Euskirchen et al., 2011; Frühwald et al., 2016;
Isakoff et al., 2005; Reincke et al., 2003; Wang et al., 2014).
61
Figure 2.3. Most significant pathways and upstream regulators associated with
genes differentially expressed in TCC642 upon SNF5 induction. Pathways identified
with a FDR cut-off of 0.05 and upstream regulator with a p-value of overlap < 0.001 are
shown. (A) Pathways associated with upregulated genes (B) Top 10 most significant
upstream regulators associated with the upregulated genes (C) Pathways associated with
downregulated genes (D) Top 10 most significant upstream regulators associated with
downregulated genes.
62
2.3.4 Chromatin accessibility and DNA methylation characterized using
AcceSssIble finds accessibility increase at a limited number of regions.
Studies have shown that the loss of SNF5 does not compromise the integrity of
the SWI/SNF complex, and it is also accepted that the target genes of the complex can
vary depending on the complex composition (Kadoch et al., 2013; McKenna et al., 2008).
Additionally, recent work by our collaborators has confirmed that re-expression of SNF5
in MRT cells indeed alters SWI/SNF complex composition (Wei et al., 2014). Thus, I
investigated how the change in SWI/SNF complex composition might be affecting
chromatin accessibility or DNA methylation of the regions targeted by the complex.
To this end, I performed AcceSssIble assay, an HM450 DNA methylation array-
based foot-printing technique developed in our lab (Becket et al., 2016; Pandiyan et al.,
2013). Nuclei isolated from TCC642 cell line at 0hr and 24hr post-doxycycline treatment
were either treated with the CpG methyltransferase M.SssI or untreated (no-enzyme,
control). Methylation levels of both enzyme-treated and untreated samples were analyzed
on the HM450 methylation array. Chromatin accessibility was inferred by subtracting the
endogenous methylation level (β-value of the CpG site in the no-enzyme control) from
the methylation β-value of the M.SssI enzyme-treated sample. A difference of 20% in
chromatin accessibility (Δ Accessibility, 24hr - 0hr) or DNA methylation (Δ Methylation,
24hr - 0hr) was considered to be a change.
Epigenetic changes upon SNF5 induction as characterized by AcceSssIble assay
were sorted into six categories (A-F, Figure 2.4A, and 2.4B). Groups A and D contained
loci altered in DNA methylation and chromatin accessibility. Specifically, group A loci
63
increased in methylation and decreased in accessibility, while group D regions decreased
in DNA methylation and increased in accessibility. In contrast, groups E and F, displayed
alteration of accessibility independent of DNA methylation. Upon induction of SNF5,
Group E regions decreased in accessibility and group F increased in accessibility. Finally,
loci in groups B and C altered in DNA methylation but this was not accompanied by any
accessibility change.
Surprisingly, I found that the effect of SNF5 induction on chromatin accessibility or
DNA methylation was limited to approximately 1% of the CpG sites probed by the array
(4561 loci). Of these, the most frequent alterations were chromatin accessibility changes
independent of DNA methylation (group E – 1016 loci, group F – 2320 loci). The least
frequent alterations were DNA methylation changes independent of chromatin
accessibility (group B – 108 loci, group C – 76 loci) (Figure 2.4A, and 2.4B). It should be
noted that DNA methylation in somatic cells is maintained in a cell-cycle dependent
manner and changes in DNA methylation patterns will occur over multiple cell generation
(Baylin et al., 1997; Bird, 2002; Robertson and Jones, 1997). Since these experiments
were designed to identify changes that take place within 24hrs of SNF5 induction, I do
not expect to capture all the DNA methylation alterations that might be taking place due
to SNF5 induction. Thus, the limited number of DNA methylation dependent changes
identified could be a reflection of our study constraint. Further experiments are required
to confirm that induction of SNF5 truly has minimal effect on DNA methylation profile of
MRT cells. TSS200 and TSS1500 locations were considered to be TSS loci while all
others were classified as non-TSS loci. In line with the probe distribution of the array, I
64
found that the CpG sites altered in accessibility or methylation were mostly of the non-
TSS group (Figure 2.4B).
Given that nucleosome occupancy in regulatory elements such as gene promoters
or enhancers can modulate target gene expression (Kelly et al., 2012; Taberlay et al.,
2014; You et al., 2011), I took a closer look at the regions altered in chromatin accessibility
(with or without DNA methylation changes). These include groups A and E (losing
chromatin accessibility) or groups D and F (gaining chromatin accessibility) (Figure 2.4A,
marked by red boxes). Specifically, I focused on genes with accessibility alterations in the
promoter upon SNF5 induction. Pathway analyses using these genes revealed that
several of the most significant pathways and upstream regulators were similar to those
identified by the mRNA expression analysis. These networks which include pathways
such as MAPK signaling, Tissue Factor (TF) signaling, and G1/S checkpoint regulation,
and regulators such as F2, Beta-estradiol, and tumor necrosis factor (TNF) could be
epigenetically driving the effects of SNF5 in MRT (Figure 2.4A, and 2.4D). Interestingly,
one of the upstream regulators of the epigenetically altered genes, ATF3, has TSS loci
gaining accessibility in SNF5 positive cells and is also upregulated in expression upon
SNF5 induction (Fold-change: 3.73, FDR: 0.056, Figure 2.4D). Consequently, ATF3 is an
example of a gene where the epigenetic reconfiguration of the promoter, mediated by
SNF5, could be directly influencing the upregulation of ATF3 in SNF5-positive cells. ATF3
is a stress-inducible transcription factor and studies have noted ATF3 over-expression
under stress conditions can induce growth arrest (Fan et al., 2002; Kim et al., 2015b).
Thus, ATF3 could be a putative mediator of the tumor-suppressor activity of SNF5.
65
Figure 2.4 Chromatin accessibility or DNA methylation of 4561 probes are affected
by expression of SNF5 in TCC642. (A) Density smooth scatterplot summarizing the results
of the AcceSssIble assay. The difference in chromatin accessibility (Δ Accessibility) between
cell line with induced SNF5 expression and control is on the x-axis. The change in DNA
methylation (Δ DNA methylation) between cells line with induced SNF5 and control is on the
y-axis. The dotted lines at 0.2 and -0.2 indicate the minimum threshold of difference required
to be considered as an alteration. (B) Bar-graphs summarizing the probe count in groups A-
F. (C) Pathways and upstream regulators of genes with TSS loci increasing in accessibility.
(D) Pathways and upstream regulators of genes with TSS loci decreasing in accessibility.
66
2.3.5 A subset of genes presenting promoter accessibility changes are also
significantly altered in expression
Using the AcceSssIble assay I identified several epigenetic alterations that were
concordant with the networks associated with gene expression changes. This suggested
that some of the genes altered in expression by SNF5 induction could be due to the
promoter remodeling activity of the SWI/SNF complex. To confirm this hypothesis and to
isolate putative direct epigenetic targets of SNF5, I cross-referenced the accessibility of
the TSS probes present in the HM450 array with the mRNA expression fold-change for
the corresponding gene (Figure 2.5A). I found most genes with at least a 2-fold change
in expression (Figure 2.5A, above the solid green line or below the solid red line) were
unperturbed in promoter accessibility. However, as summarized by the blue regression
line, there was a trend towards increased chromatin accessibility with increased
expression. I found that 78 TSS loci corresponding to 52 genes were altered in chromatin
accessibility by 20% and increased in gene expression by 2-fold (Figure 2.5A, beyond the
dotted green line) while none of the down-regulated genes showed loss of promoter
accessibility.
Next, I confirmed these findings with expression data from primary tumor samples.
Of the 52 genes altered in chromatin accessibility and expression in TCC642, 42 genes
were present in the primary sample expression dataset (Figure 2.5B). Twenty-seven of
these 42 genes were down-regulated in MRT primary samples relative to normal tissue.
I performed a Pearson correlation test of SNF5 expression with the expression of each of
the 27 genes and found 16 genes significantly correlated (r > 0.5 or <-0.5, p-value < 0.05)
67
with SNF5 mRNA expression (Figure 2.5C). I posited these 16 genes to be potential direct
epigenetic targets of SNF5.
68
Figure 2.5. Genes changing in expression and promoter accessibility in TCC642
are correlated with SNF5 expression in primary samples. (A) Scatter plot of all TSS
loci with their respective gene expression fold-change. Change in accessibility between
cells with SNF5 induced and cells with no SNF5 present on the x-axis and fold-change of
RNA expression with and without SNF5 is on the y-axis. The dotted red and green lines
mark minimum cut-off of accessibility change -0.2 (20% loss) and 0.2 (20% gain)
respectively. The solid red and green lines mark 2-fold down-regulation and 2-fold
upregulation respectively. (B) Gene expression heat-map with primary (normal, tumor)
samples of 42 genes gaining in expression in TCC642 with SNF5 induction and having a
20% gain in accessibility. Of these, 27 genes showed an expression pattern were
consistent with the cell line TCC642 cell line mRNA data. (C) Pearson correlation of each
of the 27 genes to SNF5 expression was performed. The bars-graph plots the correlation-
coefficient (r) and the line-plot denotes the significance of the correlation.
69
2.3.6 TRIM2 expression positively correlates with prognosis in several cancer types
Of the 16 genes significantly associated with SNF5 expression, TRIM2 had the
highest correlation coefficient (r = 0.85, p-value = 1.5E-8). Furthermore, two TSS probes
at the TRIM2 locus showed an increase in accessibility (cg03854543 Δ accessibility: 0.31,
cg27442931 Δ accessibility: 0.33) upon SNF5 introduction.
The E3 ubiquitin ligase protein TRIM2 has primarily been recognized for its
neuroprotective role. In neural cells, TRIM2 mediates the ubiquitination of the pro-
apoptotic BCL-2 family-member BIM through the ERK1/2 pathway (Balastik et al., 2008;
Khazaei et al., 2011; Thompson et al., 2011).
Although its role in cancer is largely unknown, a recent study reported TRIM2 to
be down-regulated in epithelial ovarian cancers (EOC) relative to normal ovarian tissue
and identified TRIM2 as a target of the tumor suppressor miR-145 in EOC (Chen et al.,
2015). In addition, using the OncoLnc data portal (Anaya, 2016), I mined publicly available
data from TCGA (Tomczak et al., 2015) to correlate TRIM2 expression with patient
survival. Using the multivariate Cox regression analysis function in OncoLnc, I found
TRIM2 expression was significantly (FDR < 0.1) and negatively correlated with death in
patients with clear cell renal cell carcinoma (ccRCC), papillary renal cell carcinoma
(PRCC), low grade glioma (LGG), and lung adenocarcinoma (LUAD) (Figure 2.6A).
Further, sorting the patients by TRIM2 expression and performing Kaplan-Maier survival
analysis comparing the top quartile with the bottom quartile (ccRCC, PRCC, LUAD) or
the top third with the bottom third (LGG) revealed that the higher expressing cohort had
better survival (Log-rank p-value < 0.05) (Figure 2.6B). Together, these results suggested
70
that TRIM2 could potentially be a critical player in oncogenesis. Moreover, since TRIM2
is strongly associated with SNF5 expression in MRT, I anticipated that it could be a direct
target of SNF5. Therefore, I chose to further investigate TRIM2 in my MRT cell line model
system.
71
Figure 2.6. TRIM2 expression is prognostic in ccRCC, PRCC, LGG, and LUAD. (A)
Multivariate regression analysis correlating TRIM2 expression with death (FDR < 0.1)
performed using OncoLnc. (B) Survival curves of patients with high TRIM2 expression
and low TRIM2 expression generated in OncoLnc. Patients in the highest and lowest
TRIM2 expression quartiles were compared for ccRCC, PRCC, and LUAD. Upper third
and lower third expression quartiles were compared for LGG.
72
2.3.7 TRIM2 is upregulated in response to SNF5 expression, independent of
senescence, and this is due to increased accessibility of the promoter.
As previously mentioned, MRT cells undergo senescence upon SNF5 induction.
Hence, I wanted ensure that the TRIM2 promoter chromatin accessibility and mRNA
changes I found in my cell line model were directly dependent on SNF5 induction and not
a side-effect of cellular senescence. To do this, I confirmed TRIM2 expression in two
additional MRT cell lines (TCC549 and G401) that have been also been transfected with
inducible SNF5 (Figure 2.7A). In addition, I also assayed for TRIM2 expression in a
fibroblast cell line (NHF-1) with inducible BRAF
V600E
. This cell line undergoes senescence
without any upregulation of SNF5 (Figure 2.7A) and serves as a negative control. I found
that the expression of TRIM2 is significantly increased only in the cell lines with inducible
SNF5 and not in the negative control (Figure 2.7B). These results further corroborated
my hypothesis that TRIM2 is a direct target of SNF5 and is upregulated due to SNF5
expression.
It is well known that gene expression is strongly correlated with an open promoter
and the presence of a nucleosome depleted region (NDR) upstream of the TSS (Kelly et
al., 2012; Lay et al.; Lin et al., 2007; Weber et al., 2007). Although the AcceSssIble assay
result indicated that the TRIM2 promoter gains accessibility upon SNF5 induction, the
assay is dependent on probe distribution of the array, thus limited in resolution. Therefore,
I performed additional characterization of the TRIM2 promoter architecture of the using
nucleosome-occupancy and methylome sequencing (NOMe-Seq), a high-resolution,
single-molecule, nucleosome-positioning assay. This technique utilizes the
73
methyltransferase M.CviPI to methylate open cytosines in a GpC dinucleotide context.
Because GpC sites are more frequently present in the genome than CpG sites, assaying
for GpC accessibility can provide better resolution (Schlesinger et al., 2007; Taberlay et
al., 2011; Widschwendter et al., 2007; Wolff et al., 2010; You et al., 2011). I performed
NOMe-seq followed by bisulfite conversion and locus-specific DNA sequencing to create
a digital footprint of nucleosome positioning using the methylation readout at the GpC
sites, as well as obtain the endogenous DNA methylation status of the CpG sites.
In the cells expressing SNF5, I find fewer DNA molecules with regions of
inaccessibility large enough to accommodate nucleosomes (>146bp) directly upstream of
the TSS (Figure 2.7C). These regions of inaccessibility are denoted by pink bars in Figure
2.7C. Additionally, the TRIM2 promoter region is unmethylated in both the control cells
and the SNF5 expressing cells, signifying that the gain in expression is independent of
DNA methylation (Figure 2.7C). Together these findings indicate that a greater
percentage of cells maintain an NDR upstream of the TSS upon SNF5 induction resulting
in the upregulation of TRIM2.
74
Figure 2.7. TRIM2 expression is dependent on SNF5 expression and regulated by
promoter accessibility. (A) SNF5 expression in MRT cell lines (TCC642, TCC549,
G401) with dox-inducible SNF5 and fibroblast NHF-1 with dox-inducible BRAF-V600E.
(B) TRIM2 expression MRT cell lines is upregulated (p-value <0.05) and not in NHF-1.
(C) NOMe-seq map of the TRIM2 promoter in TCC642. Unfilled circles are unmethylated
CpG or GpC sites, black filled circles are methylated CpG and teal filled circles denote
methylated (accessible) GpC sites. Nucleosome is indicated with a pink bar representing
region of inaccessibility greater than 146bp. TSS is marked by black arrow.
75
2.4 DISCUSSION
Aberrant activity of several tumorigenic pathways including FGFR, Hedgehog
signaling, WNT and AKT signaling have been identified in MRTs (Darr et al., 2014;
Frühwald et al., 2016; Wöhrle et al., 2013). The prevalence of SNF5 loss in MRTs and
the lack of other concomitant genomic alterations, suggest “epigenetic instability” could
be driving the deregulation of the oncogenic pathways. In this study I aimed to identify
these epigenetic alterations driving MRT oncogenesis. To this end, I used a
tetracycline-driven inducible SNF5 system to understand how expressing SNF5 in a
MRT cell line lacking endogenous SNF5 expression could promote epigenetic
alterations and reverse tumor specific gene expression changes.
By doing so, I found that SNF5 re-expression in malignant rhabdoid tumor is
sufficient to rescue aberrant gene expression of hundreds of genes. However, these
expression changes were mostly independent of promoter accessibility alterations. The
small overlap in gene expression and accessibility changes found in my investigation is
consistent with prior studies examining the inactivation of chromatin remodelers in yeast
and mouse fibroblasts that have also noted only a modest association of nucleosome
occupancy perturbations with expression changes (Gkikopoulos et al., 2011;
Tolstorukov et al., 2013).
Although, I found that the induction of SNF5 in TCC642 cells triggered epigenetic
alterations at a surprisingly limited number of genomic loci, this study is limited by the
resolution of the HM450K array. It is possible that there are many additional chromatin
accessibility changes taking place outside the regions probed. An additional limitation is
76
the heterogeneous nature of the cell population being assayed. The level of SNF5
expression is not homogenous and not all the cells begin growth arrest at the same
time. These factors could be diluting the magnitude of the resultant epigenetic changes
making it difficult to identify the accessibility changes. Finally, the majority of the
epigenetic changes I found were actually outside the promoters. This suggests, that the
SWI/SNF complex plays an important role in organizing chromatin in regions outside of
the promoters. These areas of the genome encompass elements such as gene bodies
and enhancers, both of which are critical in the regulation of gene expression (Aran and
Hellman, 2013; Calo and Wysocka, 2013; Yang et al., 2014). Therefore, investigating
the functionality of these non-promoter epigenetic changes mediated through SNF5 will
be crucial for a more complete understanding of the role of SNF5 in MRT and should
explored in future studies.
Focusing on TSSs, I found genes with alterations in promoter accessibility were
often associated with ERK/MAPK signaling, cell cycle signaling, and cell adhesion
signaling. Similarly, I found that genes upregulated in expression were significantly
associated with several cell signaling pathways that are upstream of the ERK/MAPK
pathway. Downregulated genes, on the other hand, were mostly associated with cell
cycle pathways. Thus, in this study, I identified ERK/MAPK signaling pathway to be a
putative epigenetic target of SNF5.
Aside from remodeling chromatin, SWI/SNF complex can modulate gene
expression through its interactions with nuclear receptors (Wang et al., 2014). For
example, the SWI/SNF complex can bind the retinoblastoma protein and aid in
77
suppression of retinoblastoma (Rb) targets (Trouche et al., 1997). The complex can
also interact with hormone receptors such as glucocorticoid and estrogen receptors to
initiate gene expression changes (Muratcioglu et al., 2015; Romero and Sanchez-
Cespedes, 2014). I identified many of these interactor proteins as upstream regulators
of the deregulated genes including PGR, ERBB2, Rb, and E2Fs. This suggests that in
this study I am capturing both gene expression changes due to epigenetic alterations of
the chromatin along with expression changes mediated through protein-protein
interactions.
Correlating the genes altering in both promoter accessibility and gene expression
in TTC642 cells with expression profiles of normal kidney and primary MRTs, I identified
sixteen potentially direct epigenetic targets of SNF5. Using this strategy, I have
established TRIM2 as a novel SNF5 target. SNF5 induction remodeled the TRIM2
promoter and triggered the upregulation of TRIM2 mRNA.
In summary, I surveyed the transcriptome and the epigenome of TCC642 cells with
and without the presence of SNF5 to elucidate epigenetically regulated genes and gene
networks. Significant alterations of several hundred genes were noted and importantly,
these alterations accurately reflected the expression changes observed between primary
normal and tumor samples. Further, a subset of the gene expression changes was
accompanied by chromatin accessibility aberrations. Finally, I successfully identified
novel SNF5 target genes and demonstrated that TRIM2 is an epigenetic target of SNF5.
Together, these findings show that SNF5 induction triggers gene expression changes
through modulation of promoter accessibility of target genes as well as protein-protein
78
interactions with nuclear receptors. These expression changes in turn alter oncogenic
pathways and effectively prevent the unlimited proliferation of the MRT cell line TCC642.
79
CHAPTER 3
DOWN-REGULATION OF ARID1A IS SUFFICIENT TO INDUCE EPIGENETIC
ALTERATIONS AND PHENOTYPIC TRANSFORMATION OF ENDOMETRIOTIC
CELLS
The work presented in this chapter is being prepared as a manuscript for submission.
I performed all experiments, and bioinformatics analyses unless specified below:
ARID1A-shRNA and SCR-shRNA lentiviral constructs were generated by Christopher
Duymich and Claudia Andreu-Vieyra (CAV). IEEC16 cell line was established by Kate
Lawrenson (KL) and Simon Gayther. Soft-agar assays were performed by KL and me.
CAV established knockdown and control lines in ES2, and iEEC16. The USC
Epigenome Center (Charlie Nicolet, Selene Tyndale, and Helen Troung) prepared and
sequenced the NOMe-seq libraries. RNA samples were run on expression microarray at
the Sanford-Burnham Medical Research facility (Kang Liu). Dan Wiesenberger (DW)
ran the AcceSssIble samples on HM450K array and performed array background
normalization and beta-value extraction. I wrote the draft for this chapter and
incorporated edits provided by all the co-authors. Experiments were conceived by me,
CAV, Gangning Liang, and Peter A. Jones.
3.1 INTRODUCTION
Epithelial ovarian cancer (EOC) has the highest mortality of all female reproductive
malignancies in the United States (U.S. Cancer Statistics Working Group, 2015). Ovarian
80
clear cell carcinoma (OCCC), is a subtype of EOC characterized by glycogen-containing
clear cells and “hob-nail” cells that secrete glycogen. OCCC accounts for 5% - 25% of
EOCs, depending on the population group, and advanced stages of OCCC have the worst
patient outcomes of all EOC subtypes (Helleman et al., 2006; Itamochi et al., 2008;
Nezhat et al., 2015). Lack of effective therapies and drug resistance, particularly to
conventional platinum-based chemotherapeutic agents, contribute to the poor prognosis
of late-stage OCCC (Anglesio et al., 2011; Matsuzaki et al., 2015). Thus, gaining a clearer
understanding of the biological and functional processes governing the pathogenesis of
OCCC is crucial to the development of more effective early detection tools and treatment
strategies.
Several studies have found that the non-catalytic protein AT-rich Interactive
Domain 1A (ARID1A), which is part of the SWI/SNF chromatin remodeling complex, is
mutated in numerous cancer types including OCCC, endometrioid carcinoma, and gastric
and breast cancers (Brueggmann et al., 2014; Huang et al., 2012; Wang et al., 2011b).
ARID1A is the most commonly mutated gene in OCCC (46%-57%) and endometrioid
carcinoma (26%-40%) (Guan et al., 2011a; Jones et al., 2010; Wiegand et al., 2010,
2011). Compared to other ovarian subtypes, OCCC shows the strongest association with
pre-existing endometriosis (Ayhan et al., 2012; Pearce et al., 2012; Wiegand et al., 2010).
Moreover, mutation or loss of ARID1A expression has been noted in 15%-44% of
endometriosis (Wiegand et al., 2010; Xiao et al., 2012; Yamamoto et al., 2012).
Consequently, it has been proposed that ARID1A mutation might be an early stage event
in the oncogenic transformation of endometriotic cells giving rise to OCCC (Ayhan et al.,
81
2012; Guan et al., 2011a; Wiegand et al., 2011). Though ARID1A has been implicated as
a tumor suppressor, the contribution of epigenetic deregulation due to ARID1A loss
remains unclear (Guan et al., 2011b; Mamo et al., 2012).
Here, I investigate the potential functional role and the epigenetic impact of
decreased ARID1A expression on oncogenesis by studying an immortalized
endometriosis cell line iEEC16 (Brueggmann et al., 2014; Lawrenson et al., 2014) for
alterations of DNA methylation, nucleosome positioning, and histone modifications.
82
3.2 METHODS
3.2.1 Cell culture
iEEC16 and iOSE4 cells were cultured in normal ovarian surface epithelial
complete medium at 37°C and 5% CO2
(Brueggmann et al., 2014; Lawrenson et al.,
2014). ES2 cells were cultured in McCoy’s-5A supplemented with 10% FBS. The ES2
cell line was obtained from ATCC; iEEC16 and iOSE4 cell lines were generated by Drs.
Kate Lawrenson and Simon Gayther.
3.2.2 Lentiviral transfection
Lentiviral vector pLKO.1 was used to express shRNA targeting ARID1A or a
scrambled control sequence. To generate lentiviral particles, HEK293T cells were
transfected with shRNA vectors along with packaging plasmids pCML1 and pMDG1 using
Lipofectamine LTX Reagent with Plus (Life Technologies) according to manufacturer’s
protocol. Virus supernatant was added to iEEC16, iOSE4, and ES2 cells in culture media
supplemented with 7ug/mL polybrene (Sigma). Cells were selected with 1 mg/ml
puromycin for 1 week.
3.2.3 Protein extraction and western blot analysis
Cells were rinsed with PBS, trypsinized, and resuspended in RIPA buffer (50 mM
Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40, 0.5% DOC, 0.1% SDS) supplemented with
83
protease inhibitors (Roche, 04693132001). The cells were then briefly sonicated on ice
and cellular debris was removed by centrifugation. Bradford reagent (Bio-Rad) was used
to quantify protein concentration and 40ug of protein was mixed with SDS/β-
mercaptoethanol loading buffer. Proteins were resolved on a 4-15% gradient SDS/PAGE
gel (Bio-Rad). Antibodies against ARID1A (Santa Cruz Biotechnology, 1:500) and β-
ACTIN (Sigma, A2228, 1:2,000) were used.
3.2.4 Anchorage-independent colony formation
Colony formation on soft-agar was performed as described in [Lawrenson et al.,
2014]. Briefly, 2 x 10
4
cells of iEEC16, SCR, and KD cells were plated in triplicates on
0.3% Noble Agar (Sigma) containing culture medium. A base layer of 0.6% Noble Agar
containing medium was used as a base layer. Cells were fixed and stained with 1% p-
iodonitrotetrazolium violet (Sigma) in methanol (VWR).
3.2.5 Cell adhesion to collagen
Cell adhesion to collagen was assayed by seeding 20,000 cells to 96-well plates
with uncoated wells or coated with collagen I (Gibco A10483-01). After 30 minutes of
incubation at 37°C, wells were washed three times with PBS and fixed with 4%
paraformaldehyde for 10 minutes at room temperature. Wells were washed again with
PBS and stained with 5mg/ml crystal violet in 2% ethanol for 10 minutes at room
temperature. The stain was washed off with water and plates were allowed to dry
84
completely overnight. The stained cells were lysed in 2% SDS and absorbance was read
at 590 um using a microplate reader (Microwin).
3.2.6 Cell invasion through basement membrane extract
Cell invasion assay performed using Cultrex® BME cell invasion assay (R&D
Systems) cell invasion assay was used according to manufacturer’s protocol. Data was
acquired using microplate reader (Microwin).
3.2.7 DNA content analysis by propidium iodide (PI) staining
Cells were trypsinized, washed with cold PBS (without Ca
2+
or Mg
2+
), and fixed in
70% ethanol overnight at 4°C. Fixed cells were washed twice with cold PBS and
resuspended in staining solution (200 µg PI, 2 mg DNAse-free RNAse A, 10mL 0.1%
TritonX-100 in PBS). Cells in staining solution were incubated at 37°C for 15 minutes and
20°C for 30 minutes. Data was acquired on LSR II flow cytometer (BD Bioscience).
3.2.8 NOMe-seq and AcceSssIble assays
Nucleosome-occupancy and DNA methylome sequencing (NOMe-seq) was
performed as previously described in (Kelly et al., 2012; Lay et al., 2015) and AcceSssIble
assay was performed as previously described in (Pandiyan et al. 2013).
85
Briefly, 250,000 exponentially growing cells were washed with PBS, trypsinized
and lysed 5 minutes on ice in 1mL of ice-cold nuclei-extraction buffer (10 mM Tris, pH7.4,
10mM NaCl, 3 mM MgCl2, 0.1 mM EDTA and 0.25% NP-40). The extracted nuclei were
washed with 500 µL of ice-cold wash buffer (10 mM Tris, pH7.4, 10 mM NaCl, 3 mM
MgCl2, and 0.1 mM EDTA).
For NOMe-seq, the nuclei were then resuspended in 1X GC Reaction buffer
(New England Biolabs). The nuclei were supplemented with sucrose, and 1.5uL of S-
adenosylmethionine (SAM) (New England Biolabs) and treated with 200 units of
M.CviPI enzyme (New England Biolabs) for 7.5 minutes at 37°C followed by a boost of
100 unites of M.CviPI enzyme and 0.75 µL of SAM for 7.5 additional minutes. Reactions
were terminated by adding stop solution (10 mM Tris-HCl (7.9), 600 mM NaCl, 1% SDS,
and 0.1 mM EDTA) and incubated with proteinase K at 55°C overnight. Genomic DNA
was then isolated by phenol-chloroform extraction and ethanol precipitation. Zymo’s EZ
DNA methylation (overnight protocol) kit was used to bisulfite convert 1µg of genomic
DNA. To analyze the methylation and accessibility status of individual DNA molecules at
regions of interest, bisulfite-converted DNA was PCR amplified with primers listed in
Table 3.1 and cloned into the pCR4.1 vector using the TOPO-TA cloning kit (Invitrogen,
Carlsbad, CA). Individual colonies were screened for the insert and the region of
interest was sequenced using M13 primers (Genewiz).
For AcceSssIble, the nuclei were resuspended in 1X NEB 2 buffer (New England
Biolabs). Isolated nuclei were supplemented with sucrose, 1µL of SAM (New England
Biolabs) and treated with either ddH2O (no enzyme control) or with 50 units of M.SssI
86
enzyme (New England Biolabs). Reactions were terminated by adding equal volume of
stop solution (10 mM Tris-HCl (7.9), 600 mM NaCl, 1% SDS, 0.1 mM EDTA) and
incubated with Proteinase-K at 55°C overnight. Genomic DNA (gDNA) was isolated by
phenol-chloroform extraction and ethanol precipitation. EZ DNA methylation (Zymo
Research) kit was used to bisulfite-convert 1µg of gDNA. Previously described
Methylight-based quality control tests were used to determine conversion completeness
(Campan et al., 2009). Bisulfite DNA from single AcceSssIble experiment for each cell
line was used to assay for DNA methylation levels by Infinium HumanMethylation450
(HM450) BeadChip array at the USC Epigenome Center according to manufacturer’s
protocol.
3.2.9 Genome-wide NOMe-seq analysis
3.2.9.1 Library preparation
NOMe-seq libraries were prepared from 5µg of M.CviPI treated DNA by the USC
Epigenome center as previously described in (Kelly et al., 2012) and elsewhere. DNA
was sonicated into 200-bp pieces, END-repaired (Epicenter), methylated adaptors were
ligated (Illumina), bisulfite-converted (Zymo EZ DNA Methylation), and PCR amplified
and sequenced on Illumina Hi-seq 2000 using 76-bp single-end protocol.
3.2.9.2 Sequence mapping and GC/CG methylation level extraction
Genome-wide sequencing reads from NOMe-seq experiments were mapped to the
hg19 genome, and CG and GC methylation levels were extracted using the previously
87
described pipeline (Berman et al., 2012; Kelly et al., 2012; Liu et al., 2012). Briefly, bad
quality reads were removed or trimmed. Fastq-mcf (Aronesty, 2013) was used trim
adapters and a program established in previously published work (Kelly et al., 2012)
was used to detect inverted duplication reads. Aligned BAM files were inputted to Bis-
SNP software (Liu et al., 2012) and SNP and methylation information were obtained in
standard VCF formats and bed/bedGraph/wig formats, respectively. Previously
published Perl script AlignWig2Loc along with several R scripts was used to align
sequence reads to a given set of coordinates and produce heatmaps (Lay et al., 2015).
3.2.9.3 Nucleosome depleted region (NDR) detection
A hidden Markov model (HMM) based approach previously established in (Lay et al.,
2015) was used to detect nucleosome-depleted regions (NDRs). Briefly, using GCH
methylation levels and a two-state beta-binomial HMM were used to segment regions
into methyltransferase accessible regions (MARs) and methyltransferase protected
regions (MPRs). Segments with a minimum of three contiguous GCHs in the same state
were required for MAR and MPR calling. Significance of MAR was calculated in relation
to all MPRs present in the surrounding (+/- 100kb) region using one-tailed binomial test.
MARs with FDR-corrected p-value <0.01 and larger than 140bp were considered as
NDRs.
88
3.2.10 Infinium HumanMethylation450 BeadChip data processing and AcceSssIble
data analysis
HM450 data processing and accessible data analysis were performed as
described in previously published work (Becket et al., 2016; Pandiyan et al., 2013).
Probes with a detection p-value >0.05, located within 15 bp of a SNP, overlapping
repetitive elements, mapping to multiple locations, or located on the X and Y-
chromosomes were excluded from the analysis. After filtering, 385,826 probes remained
for downstream analysis. CpG sites with a change of 0.2 (20%) in β-value were
considered to be altered in DNA methylation.
CpG accessibility was determined by subtracting the endogenous DNA
methylation of CpG sites (DNA methylation of the No Enzyme control sample) from the
DNA methylation of CpG site from the M.SssI-treated sample. CpG sites with 0.2 (20%)
change in CpG accessibility were considered to be altered in chromatin accessibility.
3.2.11 RNA extraction and analysis
Trypsinized cells were rinsed with PBS and lysed with Trizol reagent. Direct-zol
RNA MiniPrep (Zymo Research) was used to isolate total RNA. Total RNA from biological
replicates of each cell line was sent to Sanford-Burnham Medical Institute (La Jolla) for
expression analysis using Illumina expression BeadChIP HumanHT-12_V4 (Illumina).
Expression microarray data was processed using lumi package in R. Log2-transformed
and quantile-normalized data were analyzed for differential expression using limma
89
package in R. Ingenuity Pathway Analysis (Qiagen) was used to identify pathways,
processes, and diseases affected due to loss of ARID1A. Oncomine (Compendia
Bioscience) was used to mine publicly available gene expression data.
For quantitative PCR (qPCR) analysis, total RNA was treated with DNaseI
(Invitrogen), followed by EDTA and heat inactivation. DNase-treated RNA was reverse-
transcribed using iScript Reverse Transcription Supermix (Bio-Rad). Quantitative PCR
reactions were performed using KAPA SYBR FAST Universal master mix. Genes of
interest were normalized to GAPDH. Primers used in qPCR analysis are listed in Table
3.1.
3.2.12 ChIP-seq
ChIP was performed using 50 µg of chromatin as previously described (Kelly et
al., 2010b). ChIP assays on scrambled control samples were performed in biological
replicates. ChIP assays performed with chromatin from ARID1A-knockdown1 (KD1) and
ARID1A-knockdown2 (KD2) were treated as biological replicates. The following
antibodies were used: H3K4me3 (Active Motif, 39160), H3K4me1 (Active Motif, 39298),
H3K27ac (Active Motif, 39297), H3K27me3 (Active Motif, 39155). Genome-wide libraries
were generated from purified ChIP and input DNA, barcoded, and sequenced for 50
single-end reads according to [Kelly et al., 2012].
All ChIP-seq reads were extended to 200bp, mapped to hg19, non-unique and
PCR duplicates were removed using SeqMonk
90
(http://www.bioinformatics.babraham.ac.uk/projects/seqmonk). Peaks were called on
biological replicates using the model-based analysis of ChIP-seq (MACS)(Zhang et al.,
2008) algorithm with a p-value cutoff of 10
-10
against input and only peaks common
among the replicates were retained. Read count quantification, correcting for data store
size and peak width, followed by log2 transformation was used to quantify peaks.
91
Primer Use Sequence
ARID1A-FWD Expression 5’- TGGGCATTAGATACCATCAACATCC
ARID1A-REV Expression 5’- TGGACTAGACACCTTGCTGAACCTC
GAPDH-FWD Expression 5’- GATAAGGATTAAATTTTGTTTTF
GAPDH-REV Expression 5’- TTCAAAAAAAAAAACATCATAA
ZNF583-F NOMe-Seq 5’- GAGGTGGGGATTTTGTT
ZNF583-R NOMe-Seq 5’- CTCCTCCCTCCAATAA
NTF3-f NOMe-Seq 5’- GATAAGGATTAAATTTTGTTTTF
NTF3-R NOMe-Seq 5’- TTCAAAAAAAAAAACATCATAA
Table 3.1. Gene expression and NOMe-seq primer sequences
92
3.3 RESULTS
3.3.1 Establishing ARID1A KD cell line to model OCCC
As previously mentioned, up to 44% of endometriosis cases harbor either
mutations in the ARID1A gene or loss of ARID1A protein expression (Ayhan et al., 2012;
Xiao et al., 2012). Thus, ARID1A deficiency could be an early event in the transformation
of endometriotic lesions to OCCC (Yamamoto et al., 2012). To test this hypothesis, I used
stable knockdowns of ARID1A established in the immortalized endometriosis cell line
iEEC16. I confirmed down-regulation of ARID1A at the mRNA and protein level in two
biological repeat experiments performed with independent short hairpin RNAs (shRNAs)
(KD1 and KD2) (Figure 3.1). In addition, I also generated ARID1A knockdowns in iOSE4,
a cell line derived from normal ovarian surface epithelium, and in ES2, an OCCC cell line
with wildtype ARID1A expression (Figure 3.2).
93
Figure 3.1 ARID1A knockdown lines established in iEEC16. Two ARID1A
knockdown (KD1 and KD2) and one scrambled control (SCR) lines were established in
iEEC16. (A) Bar-graph of ARID1A mRNA normalized to GAPDH shows statistically
significant mRNA knockdown (n=3, *p-value < 0.05). (B) Representative western blot
image of ARID1A protein and β-Actin loading control confirm marked reduction of
ARID1A protein.
94
Figure 3.2. Stable ARID1A knockdown lines generated in ES2 and iOSE4 cell
lines. ARID1A knockdown (KD1) and scrambled control (SCR) lines were established in
ES2 and iOSE4. Bar-graph of ARID1A mRNA normalized to GAPDH shows statistically
significant (p-value < 0.05) mRNA knockdown. Western blot image of ARID1A protein
and β-Actin loading control confirm marked reduction of ARID1A protein. (A) ES2 (B)
iOSE4.
95
3.3.2 Decreased ARID1A expression in an endometriosis cell line enhances colony
formation capacity, alters cell cycle distribution, promotes cell adhesiveness, and
invasiveness.
Endometriosis cells transduced with ARID1A-shRNA to knockdown (KD) ARID1A
gene expression showed a statistically significant increase in anchorage-independent
colony formation in soft agar relative to the cells with scrambled control (SCR) shRNA (p-
value < 0.05) (Figure 3.3A). A previous study has described increased accumulation of
cells in S-phase in the absence of ARID1A (Ayhan et al., 2012). While my data does not
replicate this finding, I did find an increase in the aneuploid compartment (Figure 3.3 B,
C). Finally, ARID1A knockdown cells displayed a tendency towards increased adhesion
to collagen I and greater invasion through a basement membrane extract (p < 0.05)
(Figure 3.3 D, E). Taken together, this cell line model suggests a potential functional role
for ARID1A loss that can lead to acquisition of phenotypes associated with transformation
of endometriotic lesions.
96
Figure 3.3. Loss of ARID1A contributes to phenotypic changes in iEEC16. (a) Bar-
graph quantifying the number of colonies grown in soft agar along with representative
phase-contrast microscopy images show KD cells form more colonies independent of
anchorage than SCR. (b) Propidium iodide staining for DNA content summarized in bar
chart shows a smaller percentage of cells in G1, and S phase as well as a strong
increase in aneuploidy population. (c.) KD lines also display a trend of greater adhesion
to collagen relative to SCR and (d) are more invasive through basement membrane
extract (BME) in-vitro. For all of the above experiments N=3, *p-value <0.05.
97
3.3.3 Alteration of gene expression by ARID1A knockdown mimics the gene
expression deregulation in OCCC.
To understand how the loss of ARID1A could contribute to direct functional
changes, I performed gene expression microarray analysis of the control and KD cell
lines. I found 99 genes that were commonly affected in both the KD lines relative to the
control. Twenty-seven genes were upregulated, and 72 were down-regulated (Fold
Change (KD/SCR) > 1.4, p-value < 0.05) (Figure 3.4A, Table 3.2). I compared these
expression data with publicly-available data on primary OCCC tumors obtained from the
Oncomine
TM
data portal. For many of the deregulated genes, I observed remarkably
similar expression pattern in primary OCCC tissues. The top 20 genes with differential
expression changes in iEEC16 KD cell line that correlate with either up- or down-
regulation observed in OCCC are summarized in Figure 3.4B.
Next, I conducted pathway analyses of the differentially expressed genes in the
iEEC16 KD cell line to gain further understanding of the networks affected by the gene
expression alterations. Of the pathways most significantly altered (p-value < 0.05), many
were either related to cancer or immune response (Figure 3.4C). Notable affected
pathways involved in cell proliferation and motility include the integrin signaling pathway,
which contributes to several cellular functions including cell adhesion and invasion, and
the paxillin pathway, which is critical in mediating cell integrin signaling with p38/MAPK,
and JNK pathways (Cavallaro and Christofori, 2004; Hildebrand et al., 1995; Kumar,
1998). Gene expression alterations in these pathways might be responsible for the
increased cell adhesion to collagen and invasion observed in ARID1A-KD cell line (Figure
98
3.3C, D). In addition, ‘cancer’, ‘reproductive system disease’, and ‘embryonic
development’ were the top diseases and functions predicted to be influenced by the
differential gene expression (Figure 3.5A); I also identified genes that are involved in
several gynecological malignancies including endometrial and ovarian cancer (Figure
3.5B). Finally, many of the upstream regulators of the differentially expressed genes,
including chorionic gonadotrophin (Cg), TNF, PDGF-BB, VEGF, and RASSF1 have been
previously implicated in ovarian cancer (Fu and Zhang, 2014; Gupta et al., 2016; Matei
et al., 2007; Tashiro et al., 2003) (Figure 3.5C). Taken together, these data show that
genes exhibiting expression changes due to down-regulation of ARID1A in immortalized
endometriosis cells function in pathways and networks critical to tumorigenesis and are
frequently altered in primary ovarian tumors.
99
Figure 3.4. Differentially expressed genes in ARID1A KD are deregulated in
OCCC. Expression changes were assayed from two biological replicates of each of the
two KD lines and the SCR line using Illumina BeadArray. (A) The expression of 108
transcripts (99 genes) were significantly (Log2 (Fold Change) >0.5; p-value <0.05)
affected by ARID1A downregulation. (B) Oncomine
TM
heatmap analysis of the publicly
available Hendrix dataset (Hendrix et al., 2006) comparing normal ovary (Normal) to
OCCC (Tumor) for expression of genes altered by ARID1A downregulation. Genes are
ordered by p-value of the differential expression between normal/tumor. Twenty genes
either down-regulated (top) or up-regulated (bottom) in KD cell line are shown. (C)
Canonical pathways affected by differentially expressed genes as identified by Ingenuity
Pathway Analysis (IPA). The length of the bar connotes significance. (p-value < 0.05).
Pathways involved in cancer are highlighted in blue and those involved in immune
response are highlighted in pink.
100
Figure 3.5. Pathway analysis of the differentially expressed genes. (A) Ingenuity
Pathway Analysis (IPA) was used to identify diseases and functions affected by the
genes that were differentially expressed in the KD cell line. (B) IPA analysis revealed
differentially expressed genes contributed to numerous cancers including several
gynecological malignancies (highlighted in green). The length of the bar connotes
significance (p-value < 0.01). (C) IPA analysis also identified many upstream regulators
known to be involved in ovarian cancer (highlighted in grey).
101
Table 3.2. Genes altered in iEEC16 ARID1A KD cell lines. All genes commonly altered
in the ARID1A KD cell lines by (log2FC >0.5, p-value < 0.05).
SYMBOL log2FC_KD1 adj.P.Val_KD1 log2FC_KD2 adj.P.Val_KD2 SYMBOL log2FC_KD1 adj.P.Val_KD1 log2FC_KD2 adj.P.Val_KD2
ACTC1 -1.7 0.016192331 -4.8 2.16E-07 PMP22 -1.0 0.003265305 -1.6 3.11E-06
ACTG2 -0.8 0.009524004 -0.8 0.000323989 POTEE -0.7 0.036228101 -1.6 5.49E-06
ADARB1 -1.1 0.00058339 -2.1 7.28E-08 PPP4R4 -0.9 0.000880981 -1.4 9.29E-07
ADARB1 -0.8 0.002974711 -1.5 1.00E-06 PPP4R4 -0.8 0.012310651 -1.1 2.85E-05
ALDH1B1 -0.6 0.029434826 -1.0 7.38E-05 PRUNE2 -0.7 0.009201434 -1.0 3.25E-05
AMPH -0.9 0.014683005 -1.4 2.66E-05 PRUNE2 -0.6 0.034747318 -0.9 9.02E-05
ARSJ -0.5 0.041182793 -0.8 7.71E-05 PTGFRN -0.9 0.005857752 -1.6 3.20E-06
BCAR3 -0.5 0.041364091 -1.2 3.81E-06 PTGFRN -0.8 0.017817948 -1.3 1.07E-05
BDNF -0.7 0.014683005 -2.4 2.16E-08 RHOD -0.6 0.036228101 -0.7 0.000502819
C11orf87 -0.8 0.003283976 -0.7 0.000307626 RNF150 -0.9 0.002964519 -1.2 7.41E-06
C1QTNF5 -1.4 7.32E-05 -2.9 7.67E-09 RTN1 -0.9 0.001353014 -1.2 3.87E-06
C5orf13 -0.6 0.014683005 -1.6 3.65E-07 SCN3A -0.5 0.041851749 -0.6 0.001019878
CA2 -0.9 0.019588502 -0.6 0.007351008 SLC38A1 -0.6 0.029824199 -0.5 0.001596532
CABLES1 -0.8 0.025248644 -1.7 2.92E-06 SON -0.5 0.043415346 -0.8 0.000162886
CCL13 -2.0 2.58E-06 -1.2 5.76E-06 ST6GALNAC5 -0.6 0.041364091 -1.9 3.65E-07
CCL2 -1.5 0.000440745 -6.1 4.04E-10 SULF1 -1.2 0.000883614 -1.5 5.44E-06
CDH11 -0.6 0.029434826 -3.5 1.71E-09 SYNC -0.5 0.041669584 -1.8 2.67E-07
CGNL1 -0.9 0.001807741 -2.0 8.51E-08 SYNC1 -0.5 0.041851749 -2.8 3.27E-09
COL3A1 -1.0 0.001238962 -1.4 3.39E-06 SYNM -0.6 0.039796736 -1.5 4.59E-06
CPE -0.6 0.025248644 -0.7 0.000472778 TAGLN -0.6 0.047556531 -1.2 9.32E-06
CYP2S1 -0.7 0.014683005 -1.5 1.67E-06 TLE4 -0.6 0.040854508 -1.6 8.74E-07
DAAM1 -0.5 0.042651121 -1.3 2.06E-06 TMEM171 -0.6 0.016229988 -0.7 0.0001943
DSG2 -0.6 0.031416562 -0.9 3.57E-05 TMEM43 -0.5 0.04223168 -0.6 0.000720263
DSP -0.7 0.017446416 -1.3 2.82E-06 TNC -1.0 0.001525642 -1.0 4.65E-05
EFEMP1 -1.0 0.041364091 -3.1 3.27E-07 TNC -0.8 0.017456872 -1.3 1.61E-05
EFEMP1 -0.6 0.017817948 -0.8 0.000221926 TRIM58 -1.0 0.000682662 -1.2 3.09E-06
EFNB2 -0.8 0.041364091 -0.7 0.005188759 C13orf15 2.7 1.05E-05 1.8 7.12E-06
EIF4E3 -0.5 0.038713439 -0.7 0.000143404 ELFN2 0.5 0.034504685 0.8 0.000124084
FABP4 -0.8 0.036370986 -1.4 2.13E-05 GPR177 0.9 0.003181513 1.8 5.20E-07
FAR2 -0.8 0.006301962 -1.1 2.21E-05 LGR6 0.6 0.012342875 0.9 5.58E-05
FBLN5 -0.7 0.041364091 -1.8 1.51E-06 SVIL 0.6 0.028468441 0.8 9.02E-05
FBN2 -0.6 0.029824199 -1.1 3.25E-05 PHLDA1 0.7 0.017817948 1.1 3.85E-05
FEM1B -0.5 0.041769236 -0.5 0.003830678 G0S2 0.8 0.028468441 1.1 0.000102369
FNDC1 -1.0 0.000682662 -1.5 6.67E-07 HS3ST3A1 1.9 0.000148198 1.3 9.09E-05
FNDC1 -0.9 0.045233974 -1.5 4.38E-05 ANXA10 0.8 0.036370986 1.6 1.62E-05
GAS1 -1.4 5.50E-05 -1.3 3.47E-06 NRP1 0.5 0.042651121 1.5 5.13E-07
GAS1 -0.9 0.000923012 -0.7 0.000319566 ITGA10 0.6 0.021403306 1.1 6.32E-06
GPR37 -0.6 0.041851749 -0.6 0.00218138 TGM2 0.9 0.029824199 3.2 5.82E-08
IGFBP3 -0.5 0.035162204 -2.9 3.07E-09 FAM43A 0.7 0.012310651 0.5 0.00522737
IL11 -1.1 0.008552431 -1.0 0.000572517 MT1G 0.9 0.004322831 2.2 1.05E-07
IL6 -0.8 0.041364091 -1.3 4.19E-05 CLDN1 0.6 0.012310651 0.6 0.000410032
IL7R -0.7 0.022569976 -0.6 0.001897272 DHRS3 0.9 0.006271523 0.7 0.000713196
KCTD20 -0.5 0.036370986 -1.0 2.34E-05 JAM3 0.8 0.004322831 1.1 1.52E-05
KIAA0367 -1.1 0.001238962 -2.6 5.70E-08 TMEM158 1.1 0.025248644 2.2 5.27E-06
LBH -1.3 0.000529937 -2.5 8.51E-08 C1orf97 0.6 0.026690706 1.0 3.83E-05
LDB2 -0.8 0.003722183 -1.1 1.23E-05 NPTX1 2.7 9.25E-07 0.6 0.002807135
LIMCH1 -1.3 0.000346366 -3.1 1.08E-08 CCDC56 0.6 0.035162204 0.9 0.000114865
LMO4 -0.5 0.04223168 -1.1 1.35E-05 SCG5 1.3 9.23E-05 1.3 2.39E-06
LMOD1 -0.6 0.027688268 -1.3 1.66E-06 CCND2 0.8 0.032488962 3.3 2.28E-08
MBP -0.6 0.041769236 -1.3 4.63E-06 LRP3 0.7 0.011290311 0.8 0.000118243
NRXN3 -0.7 0.011582601 -1.2 1.09E-05 OSR1 0.9 0.029434826 0.7 0.004866906
NTF3 -1.9 1.05E-05 -2.6 3.54E-08 GPR177 0.7 0.029434826 2.5 7.76E-08
PAPSS2 -0.6 0.016317082 -1.2 3.92E-06 FAM89A 0.5 0.032361114 0.7 0.000137551
PLAC9 -0.6 0.032265319 -2.0 1.34E-07 ZNF583 0.5 0.042078397 0.7 0.000486973
PMP22 -0.8 0.003048736 -1.5 9.83E-07
102
3.3.4 Global chromatin accessibility and DNA methylation profile is largely
unperturbed by down-regulation of ARID1A
Due to its potential role as a chromatin remodeler protein, I expected that the
decreased ARID1A levels may affect chromatin architecture and contribute to the
observed phenotypic and gene expression alterations. I used two whole genome
accessibility assays developed in-house to interrogate the chromatin and achieve a
readout for both DNA methylation and chromatin accessibility.
Firstly, single AcceSssIble experiments were performed on each of the SCR and
KD cell lines. The Infinium HumanMethylation450 array (HM450)-based AcceSssIble
assay compares DNA extracted from nuclei treated with M.SssI CpG methyltransferase
with an untreated control to infer the accessibility of unmethylated CpGs and
simultaneously determine the methylation levels of CpG sites that are endogenously
methylated.
The vast majority (>95%) of the CpG sites probed did not show a robust
difference in methylation or accessibility upon decrease of ARID1A expression.
However, a 20% change in accessibility or methylation was observed in nearly 11,000
loci in both the ARID1A-KD iEEC16 cell lines. In contrast, only 6000 probes in ES2 and
1084 loci in iOSE4 displayed a methylation or accessibility change (Figure 3.6A-B,
Figure 3.7 A-B). I cross-referenced the genes with accessibility or methylation
alterations in the promoter regions (TSS loci) to investigate whether comparable genes
were being affected by ARID1A down-regulation in all three model systems.
Surprisingly, I found limited overlap amongst the genes that are affected by ARID1A
103
loss in the three cell line models. This suggested that the role of ARID1A may be cell
type specific (Figure 3.7C). However, experiments on each of the cell line models were
performed at different times, thus, it cannot be excluded that the minimal common
changes could be an experimental artifact.
Pathway analyses of the genes with altered promoter accessibility in iEEC16,
revealed epigenetic networks perturbed by the downregulation of ARID1A. Gene
promoters with increased accessibility affected several pathways previously
characterized as significant in ovarian clear cell subtype including IGF-1, and endothelin-
1 (Rosanò et al., 2009; Teoh et al., 2014; Ukaji et al., 2015) (Figure 3.6C). I also observed
greater accessibility in promoters of genes that are part of metabolic pathways such as
glycolysis I, and gluconeogenesis. This is of interest because studies have shown that
glycolysis is affected in ovarian cancer. For example, HNF1b, a metabolic gene often
overexpressed in OCCC, mediates glycolysis and promotes tumorigenesis in OCCC
(Amano et al., 2015; Okamoto et al., 2015). Additionally, among the pathways affected
by the gene TSSs with decreased promoter accessibility, I identified several key signaling
pathways including CDK5, ERK/MAPK and ERK5 (Figure 3.6D). Of note, CDK5 signaling
has been suggested to mediate sensitivity of ovarian cancer to paclitaxil treatment (Zhang
et al., 2015b). Taken together, the epigenetic alterations I observed in iEEC16 were
relevant to the putative changes taking place during ovarian carcinogenesis. Therefore, I
opted to conduct a high-resolution whole genome assessment of chromatin accessibility
and DNA methylation in iEEC16 using nucleosome occupancy and methylome
sequencing (NOMe-seq) (Kelly et al., 2012; Lay et al., 2015).
104
NOMe-seq uses the methyltransferase M.CviPI to methylate GpC sites present in
DNA regions that are free of nucleosomes or other DNA binding proteins (Schlesinger et
al., 2007; Taberlay et al., 2011; Widschwendter et al., 2007; Wolff et al., 2010; You et al.,
2011). Upon bisulfite conversion and sequencing, NOMe-seq delivers a precise digital
readout of both endogenous DNA methylation at the CpG sites and nucleosome
occupancy or chromatin accessibility from the GpC sites.
Firstly, I interrogated accessibility and DNA methylation levels at aligned
transcription start sties (TSSs) and CTCF insulator binding sites to attain a global
overview of the effect of ARID1A down-regulation. Studies have shown that active gene
promoters harbor a nucleosome-depleted region (NDR) upstream of TSS (Kelly et al.,
2012; Lay et al., 2015; Weber et al., 2007). In line with this, the majority of CpG Island
(CGI) promoters in iEEC16 (both KD and SCR cell lines) were devoid of DNA methylation
and maintained a strong NDR upstream of the TSS (Kelly et al., 2012; Lay et al.; Lin et
al., 2007; Weber et al., 2007) (Figure 3.8, top panel). In contrast, most of the non-CGI
promoters were methylated and NDRs were present in only a small subset of
unmethylated non-CGI promoters (Figure 3.8, middle panel). Finally, the well-positioned
and highly regular anti-correlative phasing of nucleosomes and DNA methylation
surrounding CTCF binding sites (Kelly et al., 2012; Lay et al., 2015) was also sustained
in ARID1A knockdown cell line (Figure 3.8, bottom panel). Therefore, reduced expression
of ARID1A in iEEC16 did not result in a noticeable impact on global chromatin
accessibility or DNA methylation at transcription start sites or CTCF insulator sites (Figure
3.8). This was consistent with the results from AcceSssIble assay.
105
Figure 3.6. Chromatin accessibility or DNA methylation of over 11,000 probes is
affected by decreased ARID1A expression in iEEC16. (A, B) Density smooth
scatterplot summarizing the results of the AcceSssIble assay. The difference in chromatin
accessibility (Delta Accessibility) between ARID1A-KD cell line and SCR is on the x-axis.
The change in DNA methylation (Delta Methylation) between ARID1A-KD cells and SCR
is on the y-axis. The dotted lines at 0.2 and -0.2 indicate the minimum threshold of change
used for it to be considered as an alteration. (A) Scatter plot of accessibility and
methylation changes in KD1, (B) scatter plot of accessibility and methylation changes in
KD2. (C) Genes with TSS probes increasing in accessibility were used in a pathway
analysis by Ingenuity Pathway Analysis (IPA). Signaling pathways of interest to
oncogenesis and/or ovarian cancer are marked by the red boxes. (D) Genes with TSS
probes decreasing in accessibility were used in a pathway analysis by IPA. Signaling
pathways of interest to oncogenesis and/or ovarian cancer are marked by the red boxes.
A B
D C
106
Figure 3.7. Minimal changes to chromatin accessibility and DNA methylation occur
in ES2 and iOSE4. (A, B) Density scatterplot summarizing AcceSssIble assay results in
(A) ES2 and (B) iOSE4 with change in accessibility between KD and SCR on thee x-axis
and change in DNA methylation on the y-axis. The dotted lines at 0.2 and -0.2 indicate
the minimum threshold of change used for it to be considered as an alteration. (C, D)
Venn-diagram displaying the overlap of gene promoters with accessibility or methylation
changes in each cell line investigated.
B A
C
D
107
Figure 3.8. NOMe-seq reveals majority of the genome unperturbed by loss of
ARID1A. Heatmaps of NOMe-seq methylation levels (CG methylation for “DNA
methylation” and GC methylation for “Accessibility”) were aligned to the center of CGI
TSS (N=20145), non-CGI TSS (N=23594) or CTCF binding sites (N=13506), extended
by +/- 1kb. CpG methylation is plotted on a scale of 0 (blue) to 100 (yellow) and GpC
methylation is plotted on a scale of 0 (white) to 100 (green). TSS or CTCF sites were
hierarchically clustered based on the accessibility within +/- 250 bp of the site.
108
3.3.5 ARID1A loss contributes to the redistribution of H3K27ac and some
redistribution of H3K27me3 at promoters
Since decreased ARID1A expression did not dramatically alter accessibility or
methylation at gene promoters and CTCF sites. I coupled the NOMe-seq data with
chromatin immunoprecipitation and sequencing (ChIP-seq) data of various histone
modifications to segment the genome into functional regions. I primarily focused on the
variations observed in H3K27ac and H3K27me3 modifications as previous reports have
suggested that these marks might be involved in mediating ARID1A activity (Bitler et al.,
2015; Kim et al., 2015a; Sun et al., 2016b). H3K4me3 and H3K4me1 data were used to
delineate potential promoter and enhancers respectively. ChIP-seq peaks were called on
biological replicates (two biological replicates of SCR, and one replicate each of KD1 and
KD2) using the model-based analysis of ChIP-seq (MACS) algorithm with cutoff of p-value
< 10
-10
. Active promoters were identified by H3K27ac and H3K4me3 peaks within 2kb of
an annotated TSS, and inactive promoters were identified by H3K27me3 peaks
overlapping TSS. H3K27ac peaks in regions >2kb away from TSS marked by the
presence of H3K4me1 and absence of H3K4me3 were designated as enhancers (Calo
and Wysocka, 2013; Kim and Shiekhattar, 2015; Mikkelsen et al., 2007b). Peaks were
called over biological replicates and read count quantification, correcting for data store
size and peak width, was performed to quantify the peaks in SCR and KD cell lines
(Charlet et al., 2016). A fold-difference of 1.5 between peak values in SCR and KD cells
was used as a cut-off of change.
109
The global levels of H3K27ac and H3K27me3 were only marginally changed in
response to ARID1A KD in iEEC16 cells (Figure 3.9). However, when the H3K27ac and
H3K27me3 peaks were segmented into regulatory elements such as promoter and
enhancers, an interesting pattern emerged. Over 60% of all the H3K27ac peaks at gene
promoters showed greater enrichment in ARID1A KD cells, while less than 0.5% of the
promoter peaks were depleted in H3K27ac (Figure 3.10A). The peaks with increased read
counts of H3K27ac at promoters were on average 2-fold more enriched in the KD, and
the peaks with decrease read counts at promoters had an average of 1.88-fold decrease
in the KD (p-value < 2E-5) (Figure 3.10A, box-plots). To understand the functional
relevance of the altered H3K27ac distribution on the rest of the epigenetic landscape of
iEEC16, I examined the top 10% most altered H3K27ac promoter peaks, corresponding
to 711 TSSs, by overlapping them with the NOMe-seq data. All of the promoters were
unmethylated in both SCR and KD samples (Figure 3.11A, left 2 panels) and the majority
of the promoters retained an open chromatin architecture in both SCR and KD cells
(Figure 3.11A, right 2 panels). Finally, most of these promoters maintained a highly
accessible region upstream of the TSS indicating the presence of an NDR (Figure 3.11A,
right 2 panels) and only a subset of promoters showed more accessibility in KD (Figure
3.11A, marked by orange box).
Unlike the large number of H3K27ac promoter peaks altered, I found that 85% of
the H3K27me3 promoter peaks were unchanged in read count enrichment due to the
decrease in ARID1A. Only 8.5% (257) of the promoter peaks were more enriched for the
repressive mark in the KD cells while 6.5% (196) of the promoter peaks were showed
110
decreased enrichment in the KD cell line (Figure 4b). On average both the “increased”
and “decreased” peaks showed a 1.8-fold difference in enrichment between the SCR and
the KD cells (p-value < 2E-5) (Figure 3.11b, box-plots). The restrained effect on
H3K27me3 distribution corroborated the idea that the dependency of SWI/SNF complex
members on PRC2 complex may not be limited to its catalytic capacity (Kim et al., 2015a).
As with H3K27ac peaks, NOMe-seq data mapping the underlying DNA methylation and
accessibility of the H3K27me3 peaks showed no visible effect on the DNA methylation
levels (Figure 3.11b; left two panels). Visually, a subset of the gene promoters losing
H3K27me3 (Figure 3.11b; right two panels, orange box) showed a modest gain in
chromatin accessibility, while the promoters gaining H3K27me3 in the KD sample
displayed a general decrease in accessibility in these regions (Figure 3.11b; right two
panels, below the dotted line).
These findings indicated that the downregulation of ARID1A can stimulate a
general increase of H3K27ac enrichment at gene promoters, and a context-dependent
alteration of H3K27me3 enrichment. In both cases, the change in histone mark
distribution had minimal effect on the underlying chromatin accessibility and DNA
methylation.
111
Figure 3.9. Global H3K27ac and H3K27me3 remain largely unchanged. H3K27ac and
H3K27me3 data from ARID1A KD and SCR cell lines were used. Box plots summarizing
read count quantifications of (A) H3K27ac-peaks and (B) H3K27me3 peaks in biological
replicates of SCR and KD cell lines (log2 of read count).
A B
112
Figure 3.10. Down-regulation of ARD1A promotes re-distribution of H3K27ac and
H4K27me3. (A, B) H3K27ac and H3K27me3 data from ARID1A KD and SCR cell lines
were used. Pie-charts summarizing histone peaks with increased read count enrichment
by 1.5-fold (“increased”), decreased read count enrichment by 1.5-fold (“decreased”), or
unchanged in read count enrichment (“unchanged”) in the KD cells relative to SCR cells
and box-plots quantifying log2 of read count enrichment in the “increased” or “decreased”
peaks are shown. (A) H3K27ac peaks at gene promoters (n=6910) (B) H3K27me3 peaks
at gene promoters (n=3019).
A
B
*p-value < 2E-5
113
Figure 3.11. Promoter accessibility changes accompany histone modification
alterations at a small subset of TSS. H3K27ac or H3K27me3 peaks overlapped with
NOMe-seq methylation levels (CG methylation levels as “DNA Methylation” and GC
methylation levels as “Accessibility”) were aligned based on TSS center, extended by +/- 1kb
and plotted as heatmaps. Genomic loci were hierarchically clustered based on the
accessibility within +/- 250 bp of TSS centers. (a) Top 10% most altered H3K27ac promoter
peaks overlapped with NOMe-seq data and centered at known TSS. Visually, a subset of
promoters (highlighted by the orange box) show some gain in accessibility in the KD cell line.
(b) Top 10% most altered H3K27me3 promoter peaks overlapped with NOMe-seq data and
centered at TSS. Promoter peaks losing H3K27me3 are above the dotted line and peaks
gaining H3K27me3 are plotted below the dotted line. Dotted-black line separates promoters
with decreased H3K27me3 enrichment in KD cell line (above the dotted-line) from promoter
regions with increased H3K27me3 in KD cell line. Visually, a subset of promoters (highlighted
by the orange box) show some gain in accessibility in the KD.
114
3.3.6 Down-regulation of ARID1A also contributes to the redistribution of H3K27ac
mark at enhancers
In contrast to the increased enrichment of H3K27ac at most promoters, the
majority of the altered enhancer peaks were reduced in H3K27ac upon the down-
regulation of ARID1A (Figure 3.12). Close to 24% of the enhancer peaks were decreased
in H3K27ac enrichment while fewer than 5% of the peaks in the ARID1A KD cells showed
increase of H3K27ac enrichment (Figure 3.12). On average, the peaks with increased
H3K27ac were 1.9-fold more enriched for the mark while the enhancers losing acetylation
were 1.85-fold less enriched for the mark in the KD (p-value < 2E-5) (Figure 3.12, box-
plots).
Unlike promoters, enhancers do not have a defined start site. Therefore, to identify
a functional change I aligned the enhancers to NDRs called in the control and KD line
using a Hidden Markov Model (HMM) approach previously described by our lab (Kelly et
al., 2012; Lay et al., 2015) and briefly outlined in the methods (Chapter 3, Section 3.2.9.3).
I then intersected the top 10% of most altered H3K27ac enhancer peaks with NOMe-seq
data to generate chromatin accessibility and methylation maps of the enhancers (Figure
3.13). Nearly all of the enhancer peaks contained NDRs in SCR sample and strikingly, a
small portion of these enhancers lose NDRs in the KD (Figure 3.13, left 2 panels, marked
by orange box). Additionally, similar to the promoters (Figure 3.11a), DNA methylation in
the 2kb window around the NDRs remained unperturbed between the SCR and KD
samples (Figure 3.13, left 2 panels).
115
Together, I found that, in contrast to promoters in which loss of ARID1A resulted
in nearly 60% of H3K27ac peaks gaining acetylation, enhancers lose H3K27ac at about
24% of peaks in the KD. However, these changes in histone mark distribution correlated
with modulation of chromatin accessibility at only a limited number of genomic regions.
These findings suggest that ARID1A plays an important role in the distribution of H3K27ac
histone marks and a potentially targeted role in driving changes to nucleosome
positioning and DNA methylation in these regions.
116
Figure 3.12. Loss of ARID1A expression affects enrichment of H3K27ac at
enhancers. H3K27ac and H3K27me3 data from ARID1A KD and SCR cell lines were
used. Pie-chart summarizes histone peaks with increased read count enrichment by 1.5-
fold (“increased”), decreased read count enrichment by 1.5-fold (“decreased”), or
unchanged in read count enrichment (“unchanged”) in the KD cells relative to SCR cells.
Box-plots quantify log2 of read count enrichment in the “increased” or “decreased” peaks.
Majority of the H3K27ac enhancer peaks (n=33627) are unchanged in read count
enrichment between SCR and KD cell lines.
*p-value < 2E-5
117
Figure 3.13. Chromatin accessibility changes accompany histone modification
alterations at small subset of enhancers. Hidden-Markov model identified
methyltransferase accessible regions (MARS), and MARS greater than 140p were
considered nucleosome-depleted regions (NDRs). H3K27ac peaks overlapped with
NOMe-seq methylation levels (CG methylation levels as “DNA Methylation” and GC
methylation levels as “Accessibility”) were aligned based on either center of the NDRs,
extended by +/- 1kb and plotted as heatmaps. Genomic loci were hierarchically clustered
based on the accessibility within +/- 250 bp of NDR centers. Top 10% of most altered
H3K27ac enhancer peaks overlapped with NOMe-seq methylation levels aligned to NDR
centers and extended to +/- 1kb are plotted as heatmaps. Dotted-black line separates
enhancers with decreased enrichment of H3K27ac in KD cell line (above the dotted-line)
from genomic regions with increased H3K27ac in KD cell line. Orange box denotes subset
of enhancers that, upon visual inspection, show some loss in accessibility in the KD.
118
3.3.7 A subset of differentially expressed genes are directly affected by epigenetic
changes due to loss of ARID1A
Next, I integrated the epigenomic data with the gene expression changes observed
in iEEC16 KD cell line to understand how the differentially expressed genes may be
epigenetically regulated by the decreased expression of ARID1A. While the genes
regulated by a particular promoter can be easily identified, associating an enhancer with
its target gene can be much more challenging. Enhancers often regulate multiple genes
over considerable distances and in doing so they habitually skip over the closest gene.
However, studies have shown that enhancers affect the gene corresponding to the
nearest TSS with the highest frequency of any other TSS in the genome (Li et al., 2012;
Yao et al., 2015b). Therefore, I annotated the putative enhancer peaks with the nearest
gene. Among the down-regulated genes, 9 gene promoters displayed some loss of
H3K27ac, 9 gene promoters displayed some gain of H3K27me3, and 39 putative
enhancers showed loss of H3K27ac (Figure 3.14a). Conversely among upregulated
genes, 16 promoters saw gain of H3K27ac, 9 promoters saw a loss of H3K27me3, and 7
putative enhancers gained acetylation (Figure 3.14b).
Thus, epigenetic alteration could be involved in up to 43 out of the 72 down-
regulated genes and 16 out of the 27 up-regulated genes. For example, the SVIL and
ZNF583 promoters showed a gain of the H3K27ac mark and both of these genes were
increased in expression in the KD line (mean expression fold-change = 1.6 and 1.5
respectively). In addition, the gain of H3K27ac mark at SVIL promoter is complemented
with decreased H3K27me3 (Figure 15a). On the other hand, SULF1, and CCL2 were two
119
of the few promoters to lose H3K27ac in the KD line and these genes were downregulated
in the KD cell line (mean expression fold-change = -2.6 and -13.3 respectively).
Additionally, SULF1 promoter also increased in enrichment of H3K27me3 mark (Figure
3.15B).
I selected ZNF583, a gene that is up-regulated in KD cells (Table 3.2), to further
investigate the effect of ARID1A knockdown on promoter accessibility at a region gaining
active marks (Figure 3.15A). I performed NOMe-seq followed by locus-specific DNA
sequencing to create a digital footprint of nucleosome positioning at this gene promoter
in the control and KD samples. The ZNF583 CpG island promoter has 3 TSSs all located
within the H3K27ac peak found in the SCR. In the KD cells, fewer DNA molecules with
regions of inaccessibility large enough to accommodate nucleosomes (>146bp) were
present directly upstream of the TSSs (Figure 2.15C, highlighted with blue boxes). These
regions of inaccessibility are represented by pink bars in Figure 2.15C. Furthermore, the
nucleosomes were less consistently positioned in the KD indicating weaker positioning
relative to the SCR. Together, this suggests ARID1A may be required for the down-
regulation of ZNF583 expression by the maintenance of strongly positioned nucleosomes
at its promoter.
The enhancer peaks located upstream of LRP3 and upstream of JAM3, displayed
an increase in the enrichment of H3K27ac in the KD (Figure 3.15D). These genes were
upregulated in expression in the KD cell line (mean expression fold-change = 1.70 and
1.98, respectively). On the other hand, the enhancers in the intron of TLE and NTF3 were
strongly depleted in the KD (Figure 3.15E) and these genes were downregulated in the
120
KD cell line (mean expression fold-change = -2.34 and -5.59, respectively). None of these
locations see any enrichment of H3K27me3 in either the SCR or the KD line.
I selected the putative enhancer located upstream of NTF3 to investigate the effect
of ARID1A knockdown on chromatin accessibility of a region losing active histone marks
(Figure 2.15D). The neurotrophic tyrosine kinase receptor, NTF3, was downregulated in
both the iEEC16 cell line expression data (Table 2.2) and in primary OCCC expression
data from Oncomine
TM
(Figure 2.15B top panel). Locus-specific NOMe analysis of the
NTF3 enhancer revealed an accessible region downstream of the enhancer peak center
in the SCR, marked by a green line in Figure 2.15F, which is nearly completely replaced
with regions of inaccessibility large enough to accommodate nucleosomes (represented
by pink bars) in the KD (Figure 2.15F). This decreased accessibility at the enhancer could
explain the loss of NTF3 gene expression, thus suggesting that ARID1A could be required
to keep the chromatin open and accessible in the enhancer for proper transcription and
expression of NTF3.
Although only a limited number of genes were differentially expressed upon loss
expression of ARID1A, the deregulation of nearly 58% of the downregulated genes and
56% of the upregulated genes was accompanied with some epigenomic variation (Figure
3.14a, and 3.14b). Closer inspection of some of these regulatory elements showed
chromatin accessibility alterations with complementary histone modification changes
(Figure 3.156c, and 3.15f). Thus, alteration of gene expression due to loss of ARID1A
could be directly controlled through the modulation of histone modifications and the
accessibility of the chromatin. However, conclusive case for epigenetic regulation of
121
chromatin accessibility and gene expression can be made only after more thorough
mechanistic investigations.
122
Figure 3.14. Subset of the differentially expressed genes are epigenetically
regulated by ARID1A. (A) Venn-diagrams of down-regulated genes showing any loss
in H3K27ac (at promoters or enhancers) or gain in H3K27me3 at gene promoters. (B)
Venn-diagrams of up-regulated genes showing any gain in H3K27ac (at promoters or
enhancers) or loss of H3K27me3 at gene promoters.
123
Figure 3.15. H3K27ac redistribution affects both gain and loss of regulatory
elements. IGV browser view of gene promoters and enhancers (A, B, D, E). The first
track shows the gene or genomic region, which is followed by a track of CGI. Next, wiggle
plots of H3K27ac ChIP-seq enrichment in SCR cells and ARID1A KD cells are shown,
and finally wiggle plots of H3K27me3 ChIP-seq enrichment data are displayed. (C, F)
Unfilled circles are unmethylated (inaccessible) GpC sites, teal filled circles denote
methylated (accessible) GpC sites. Nucleosome is indicated with a pink bar representing
region of inaccessibility greater than 146bp. TSS is marked by black arrow and center of
enhancer peak is denoted by a green line. (A) SVIL and ZNF583, are gaining acetylation
peak at their respective promoters, and (B) SULF1 and CCL2, are genes showing loss of
the active regulatory element at the promoter. (C) The gain in active regulatory element
at the promoter of ZNF583 is accompanied by increase in chromatin accessibility
upstream of each of the 3 TSSs. (D) LRP3 and JAM3, gain the acetylation peak at their
respective gene body enhancer, and (E) TLE4 and NTF3 showing loss of the active
enhancer element in the gene body. (F) The loss in active regulatory element at the
enhancer of NTF3 is accompanied by the loss of chromatin accessibility downstream of
the center of the enhancer peak.
124
3.4 DISCUSSION
In this study I found that down-regulation of ARID1A in a non-malignant
endometriosis cell line was sufficient to induce higher efficiency of anchorage-
independent growth, increase the cells’ propensity to adhere to collagen, and promote
invasion of basement membrane extract in vitro. The ability for adherent cells to grow in
an anchorage-independent fashion in vitro has been shown to correlate with neoplastic
growth in animal models and therefore considered to be a barometer of tumorigenicity
(Freedman et al., 1974; Mori et al., 2009). In addition, cell adhesion to extracellular matrix
is critical for the proper growth and proliferation of cells. Type I collagen, the most
abundant ECM protein, can induce potent chemotactic responses, adhesion to collagen
can be crucial for invasion, metastasis as well as signaling in cancer cells (Cavallaro and
Christofori, 2004; Chen et al., 2009). Thus, the phenotypic alterations due to ARID1A
down-regulation in the non-tumorigenic iEEC16 cell line suggest that ARID1A mutation
could be a critical step in the transformation of endometriotic lesions into carcinoma.
A set of around 100 genes showed altered expression after down-regulation of
ARID1A. These genes overlapped to a considerable extent with previously noted
changes such as gain of GRPr expression and loss of SULF1 (Sulfatase I) expression.
GRPr (gastrin-releasing peptide receptor) upregulation is present in several solid tumors
including breast, ovary, and prostate (Cornelio et al., 2013; Elshafae et al., 2016). In
contrast, SULF1 is often down-regulated in ovarian cancer especially in the clear cell
subtype. Loss of SULF1 promotes cellular proliferation through growth factor signaling
(He et al., 2014). In addition, I identified several genes whose expression were highly
125
concordant with OCCC and have yet to be studied in the context of ovarian cancer
development, and could be worth exploring in the future.
Pathway analyses of the expression changes revealed that many of the
differentially expressed genes affect networks significant to oncogenesis and contribute
to disease phenotype that include ovarian and endometrial cancers. Several studies have
shown a epistatic relationship between ARID1A and the PI3/AKT signaling pathway in
OCCC, where loss of ARID1A is often accompanied by gain in PI3K/AKT signaling and
ARID1A mutant OCCC are sensitized to PI3K/AKT inhibitors (Alvarez-Nuñez et al., 2006;
Chandler et al., 2015; Guan et al., 2014; Samartzis et al., 2013; Shinya Matsuzaki* et al.,
2015; Wiegand et al., 2014). However, my analysis did not detect any expression
modulation of genes within the PI3K/AKT pathway, suggesting that in the iEEC16 cell line
model, gain of PI3K/AKT activity may not be necessary to trigger phenotypic
transformation or PI3K might not be downstream of ARID1A.
The limited number of expression changes observed is consistent with the
literature and is not surprising considering the molecular redundancies present in
chromatin remodeler complexes as well as the experimental limitations of this study
(Guan et al., 2011b; Raab et al., 2015). For example, a recent study showed that up to
41% ARID1A targets genes are also targeted by the ARID1A homologs ARID2 or ARID1B
(Raab et al., 2015). Further, this study was conducted using a knockdown system and
therefore a low level mRNA and protein expression persists; this residual ARID1A
expression could also be contributing to the small number of expression changes
observed.
126
A surprising outcome of my work was that down-regulation of ARID1A impacted
global H3K27ac distribution but had targeted effects on the rest of the epigenome. I found
nearly 60% of promoters and 25% of enhancers were altered in H3K27ac enrichment
upon loss of ARID1A yet only a few of these histone changes translated to robust
alterations in chromatin accessibility, or gene expression. Thus, the quantitative level of
H3K27ac marking might not correlate highly with the degree of accessibility and the level
of histone modification alone might not be sufficient to induce strong changes to the
chromatin accessibility.
Studies attempting to validate putative enhancers have noted that only a subset of
enhancers have functional role in mediating gene expression changes (Kwasnieski et al.,
2014; Nord et al., 2013). My findings are consistent with these studies; though nearly a
quarter of the enhancer peaks are losing H3K27ac, the majority of these enhancer
changes are not associated with alteration of chromatin accessibility or gene expression.
Furthermore, a recent study examining role of ARID protein homologs on hepatocellular
epigenome found an enrichment of HDAC targets amongst ARID1A ChIP-seq peaks
(Raab et al., 2015). While I do not find statistically significant differential expression of
HATs or HDACs, it is possible that changes in the interaction dynamics of the SWI/SNF
complex with these histone modifiers could be contributing to the H3K27ac changes
observed. Finally, in a locus-specific manner, I showed that a strong change in histone
modification, when co-occurring with gene expression alteration, can be accompanied by
chromatin accessibility modulation, suggesting ARID1A has targeted rather than global
effects on the epigenome and transcriptional regulation in OCCC.
127
Gene transcription is controlled by the availability of regulatory factors and
transcriptional machinery, along with the permissiveness of the regulatory elements
(marked by active histone marks). A recent study has suggested the loss of ARID1A can
restrict promoter accessibility to transcription factors, thereby affecting the histone
modifications, and in turn effecting gene expression and phenotypic changes (Sun et al.,
2016b). As previously mentioned, SWI/SNF complexes can compensate for the loss of
ARID1A by incorporating homolog such as ARID1B (Raab et al., 2015; Sun et al., 2016b).
Thus, the redistribution of histone modification and gene expression could be a
combination of direct and indirect effects mediated through the interactions of ARID1A,
or other compensatory homologs, with histone modifiers, transcription factors, and other
DNA binding proteins. ChIP-seq experiments assaying for ARID1A may help us to better
elucidate the potential mechanism. Unfortunately, the lack of DNA-sequence specificity
of ARID1A binding, in concert with poor quality of antibodies currently available in the
market, prevent us from performing a successful genome-wide assay of ARID1A binding
sites. Whether direct or indirect, these findings imply that down-regulation of ARID1A can
alter the distribution of histone modifications, modulate the permissiveness of some
regulatory elements, affect gene expression, and finally contribute to cellular
transformation.
Lastly, it has been long speculated that endometriosis could be a precursor to
OCCC (Ayhan et al., 2012; King et al., 2016; Xiao et al., 2012). In this study, I provided
evidence that loss of ARID1A in non-tumorigenic endometriosis cells may be sufficient
128
for the initiation of phenotypic and molecular alterations that could potentially contribute
to oncogenic transformation.
129
CHAPTER 4
SUMMARY AND CONCLUSIONS
As reviewed in Chapter 1, epigenetic regulation has emerged as a hallmark of
cancer development. Mechanisms such as DNA methylation, histone modification, and
nucleosome positioning act in concert with each other to maintain gene expression
potential and chromatin organization over cell generations. Thus, mutations in
mediators of epigenetic processes can potentially contribute to considerable alterations
of gene expression, disrupt cell identity and contribute to the development of
malignancies such as cancer (Baylin et al., 1997; Sharma et al., 2010). In line with this,
large-scale sequencing efforts have uncovered germline and somatic mutations in many
epigenetic modulators including DNA methyltransferases, histone tail modifiers, and
chromatin remodeling complex members (Shen and Laird, 2013; You and Jones, 2012).
Consequently, it is now apparent that aberrant chromatin regulation is a key mechanism
in pathogenesis of tumors.
First discovered in yeast, the SWI/SNF chromatin remodeler complex, named for
yeast mating types switching (SWI) and sucrose non-fermenting (SNF), uses ATP to
drive its chromatin remodeling activity. Inactivating mutations in several components of
this complex have been identified with high frequencies in a wide array of tumor types
(Halliday et al., 2009; Oike et al., 2013). Furthermore, many components of the
SWI/SNF complex are capable of establishing protein-protein interactions with nuclear
receptors to further facilitate changes in gene expression potential. Thus, the goals of
130
this thesis were: to elucidate the effect of SWI/SNF on the epigenetic landscape, gene
expression, and cellular phenotype, as well as to identify the epigenetic driver genes
and gene networks that are altered in cancers due to the loss of SWI/SNF protein
members.
To understand how SWI/SNF complexes could epigenetically alter gene networks
and promote oncogenic phenotype, I modulated the expression of SWI/SNF subunits in
tumor and non-tumor cell line models. Firstly, in Chapter 2, I explored the role of SNF5 in
malignant rhabdoid tumor (MRT). I characterized the epigenetic changes that occur upon
re-expression of SNF5 in MRT, and identified epigenetic targets of SNF5 that might
contribute to disease progression. I used MRT cell lines stably transfected inducible SNF5
transgene. Using this model system, I first detected gene expression changes upon
induction of SNF5. Next, I validated these findings in primary tumor samples and found a
remarkable consistency of the expression profiles between the cell line model and primary
MRT samples. Surveying the epigenome before and after SNF5 induction in vitro using
chromatin accessibility assay AcceSssIble, I identified gene networks that are
epigenetically regulated and established TRIM2 as an epigenetic target of SNF5 in MRT.
In Chapter 3, I demonstrated that knocking down ARID1A in an immortalized
endometriosis cell line was sufficient to promote phenotypic transformation as evidenced
by higher efficiency of anchorage-independent growth, increased propensity to adhere to
collagen and greater capacity to invade basement membrane extract in vitro. Using the
NOMe-seq and AcceSssIble accessibility assays, along with ChIP-seq and expression
microarray, I explored the interplay between the various epigenetic processes and how
131
they mediated functional changes. First, I compared the transcription profiles of
endometriosis cells with and without ARID1A knockdown and identified significant
expression changes in the gene expression due to ARID1A knockdown. Remarkably,
there was a strong overlap between the alterations due to down-regulation of ARID1A
and expression deregulation observed in primary OCCC. Further, pathway analyses
showed that these genes impacted networks highly relevant to tumorigenesis. Next, I
demonstrated that reduced expression of ARID1A had targeted effects on chromatin
accessibility, yet, correlated with a strong increase in the active H3K27ac mark at majority
of promoter regions and a decrease at a large number of potential enhancers. These
findings provided evidence that ARID1A mutation can be an early stage event in the
oncogenic transformation of endometrial cells in giving rise to OCCC and this is partly
mediated through epigenetic alterations.
Together, my findings demonstrated that though ARID1A and SNF5 are
components of a chromatin remodeling complex, their roles in mediating functional
changes go beyond direct organization of chromatin. Both in the case of SNF5 induction
in MRT and ARD1A knockdown in endometriosis cells, I found that chromatin accessibility
and DNA methylation changes were largely limited. This is in line with other studies that
have shown downregulation of a SWI/SNF component does not necessarily compromise
the complex integrity (Kadoch et al., 2013; Wei et al., 2014). Further, the minimal overlap
of gene expression changes and chromatin accessibility changes are also supported by
earlier findings in which deregulated chromatin is not always accompanied by expression
changes (Gkikopoulos et al., 2011; Tolstorukov et al., 2013). This provides a model for
132
nucleosome positioning to be a mechanism of fine tuning expression and transcription
can be largely maintained even with aberrant positioning. Secondly, gene expression is
modulated by the availability of transcriptional activators and repressors, and SWI/SNF
subunits are known to interact with nuclear proteins such as transcription factors (Wang
et al., 2014). Thus, many of the expression changes observed could be due to novel
interactions of the complex with such factors and independent of the chromatin
remodeling capacity of the SWI/SNF complex.
Due to their role as chromatin remodeler complex members and because neither
SNF5 nor ARID1A bind DNA with sequence specificity, it could be imagined that
modulation of their expression would have a global effect on nucleosome positioning and
occupancy. Indeed, there have been somewhat differing reports of large scale occupancy
and positioning changes due to loss of SWI/SNF subunits. One study, performed in
mouse embryonic stem cells, showed that loss of ARID1A resulted in increased
nucleosome occupancy at TSS (Lei et al., 2015). Another study in mouse epidermal
fibroblasts found that loss of SNF5 decreased nucleosome occupancy at the peri-TSS
regions (Tolstorukov et al., 2013). Finally, loss of SWI/SNF in yeast showed no change
in nucleosome occupancy but a significant alteration of nucleosome positioning (Yen et
al., 2012). However, my investigations found modulation of SWI/SNF complex members
had a restrained effect on global chromatin accessibility levels and a much larger effect
at individual targets. Together, these findings highlight the highly context-dependent role
of SWI/SNF complexes.
133
While my work provides direct evidence that loss of epigenetic modulators, such
as members of the SWI/SNF complex, can initiate phenotypic changes and contribute to
potential oncogenic transformation. It also underscores the need for more thorough
mechanistic studies integrating epigenomic information along with complex composition
in the context of oncogenesis.
134
REFERENCES
Agirre, X., Vilas-Zornoza, A., Jiménez-Velasco, A., Martin-Subero, J.I., Cordeu, L.,
Gárate, L., San José-Eneriz, E., Abizanda, G., Rodríguez-Otero, P., Fortes, P., et al.
(2009). Epigenetic silencing of the tumor suppressor microRNA Hsa-miR-124a
regulates CDK6 expression and confers a poor prognosis in acute lymphoblastic
leukemia. Cancer Res. 69, 4443–4453.
Aird, K.M., and Zhang, R. (2013). Detection of senescence-associated heterochromatin
foci (SAHF). Methods Mol. Biol. 965, 185–196.
Alvarez-Nuñez, F., Bussaglia, E., Mauricio, D., Ybarra, J., Vilar, M., Lerma, E., de Leiva,
A., and Matias-Guiu, X. (2006). PTEN promoter methylation in sporadic thyroid
carcinomas. Thyroid 16, 17–23.
Amano, Y., Mandai, M., Yamaguchi, K., Matsumura, N., Kharma, B., Baba, T., Abiko,
K., Hamanishi, J., Yoshioka, Y., and Konishi, I. (2015). Metabolic alterations caused by
HNF1β expression in ovarian clear cell carcinoma contribute to cell survival. Oncotarget
6, 26002–26017.
Amente, S., Milazzo, G., Sorrentino, M.C., Ambrosio, S., Di Palo, G., Lania, L., Perini,
G., and Majello, B. (2015). Lysine-specific demethylase (LSD1/KDM1A) and MYCN
cooperatively repress tumor suppressor genes in neuroblastoma. Oncotarget 6, 14572–
14583.
Anaya, J. (2016). OncoLnc: linking TCGA survival data to mRNAs, miRNAs, and
lncRNAs. PeerJ Comput. Sci. 2, e67.
135
Andreotti, G., Karami, S., Pfeiffer, R.M., Hurwitz, L., Liao, L.M., Weinstein, S.J.,
Albanes, D., Virtamo, J., Silverman, D.T., Rothman, N., et al. (2014). LINE1 methylation
levels associated with increased bladder cancer risk in pre-diagnostic blood DNA
among US (PLCO) and European (ATBC) cohort study participants. Epigenetics 9,
404–415.
Andreu-Vieyra, C. V, and Liang, G. (2013). Nucleosome occupancy and gene regulation
during tumorigenesis. Adv. Exp. Med. Biol. 754, 109–134.
Anglesio, M.S., Carey, M.S., Köbel, M., MacKay, H., and Huntsman, D.G. (2011). Clear
cell carcinoma of the ovary: A report from the first Ovarian Clear Cell Symposium, June
24th, 2010. Gynecol. Oncol. 121, 407–415.
Aran, D., and Hellman, A. (2013). DNA methylation of transcriptional enhancers and
cancer predisposition. Cell 154, 11–13.
Aran, D., Sabato, S., and Hellman, A. (2013). DNA methylation of distal regulatory sites
characterizes dysregulation of cancer genes. Genome Biol. 14, R21.
Aronesty, E. (2013). Comparison of Sequencing Utility Programs. Open Bioinforma. J.
7, 1–8.
Ayhan, A., Mao, T.-L., Seckin, T., Wu, C.-H., Guan, B., Ogawa, H., Futagami, M.,
Mizukami, H., Yokoyama, Y., Kurman, R.J., et al. (2012). Loss of ARID1A expression is
an early molecular event in tumor progression from ovarian endometriotic cyst to clear
cell and endometrioid carcinoma. Int. J. Gynecol. Cancer 22, 1310–1315.
Balastik, M., Ferraguti, F., Pires-da Silva, A., Lee, T.H., Alvarez-Bolado, G., Lu, K.P.,
136
and Gruss, P. (2008). Deficiency in ubiquitin ligase TRIM2 causes accumulation of
neurofilament light chain and neurodegeneration. Proc. Natl. Acad. Sci. U. S. A. 105,
12016–12021.
Bannister, A.J., and Kouzarides, T. (2011). Regulation of chromatin by histone
modifications. Cell Res. 21, 381–395.
Baylin, S.B., and Jones, P.A. (2011). A decade of exploring the cancer epigenome —
biological and translational implications. Nucleus.
Baylin, S.B., Bird, A., Chuang, L.S.H., Antequera, F., Bird, A., Antequera, F., Turker,
M.S., Bestor, T.H., Bestor, T.H., Tycko, B., et al. (1997). Tying it all together:
epigenetics, genetics, cell cycle, and cancer. Science 277, 1948–1949.
Becket, E., Chopra, S., Duymich, C.E., Lin, J.J., You, J.S., Pandiyan, K., Nichols, P.W.,
Siegmund, K.D., Charlet, J., Weisenberger, D.J., et al. (2016). Identification of DNA
Methylation-Independent Epigenetic Events Underlying Clear Cell Renal Cell
Carcinoma. Cancer Res. 76, 1954–1964.
Behmoaram, E., Bijian, K., Jie, S., Xu, Y., Darnel, A., Bismar, T.A., and Alaoui-Jamali,
M.A. (2008). Focal adhesion kinase-related proline-rich tyrosine kinase 2 and focal
adhesion kinase are co-overexpressed in early-stage and invasive ErbB-2-positive
breast cancer and cooperate for breast cancer cell tumorigenesis and invasiveness.
Am. J. Pathol. 173, 1540–1550.
Berman, B.P., Weisenberger, D.J., Aman, J.F., Hinoue, T., Ramjan, Z., Liu, Y.,
Noushmehr, H., Lange, C.P.E., van Dijk, C.M., Tollenaar, R.A.E.M., et al. (2012).
137
Regions of focal DNA hypermethylation and long-range hypomethylation in colorectal
cancer coincide with nuclear lamina-associated domains. Nat. Genet. 44, 40–46.
Bestor, T.H., Edwards, J.R., and Boulard, M. (2015). Notes on the role of dynamic DNA
methylation in mammalian development. Proc. Natl. Acad. Sci. 112, 6796–6799.
Biegel, J.A., Zhou, J.Y., Rorke, L.B., Stenstrom, C., Wainwright, L.M., and Fogelgren, B.
(1999). Germ-line and acquired mutations of INI1 in atypical teratoid and rhabdoid
tumors. Cancer Res. 59, 74–79.
Biegel, J.A., Busse, T.M., and Weissman, B.E. (2014). SWI/SNF chromatin remodeling
complexes and cancer. Am. J. Med. Genet. C. Semin. Med. Genet. 166C, 350–366.
Bird, A. (2002). DNA methylation patterns and epigenetic memory. Genes Dev. 16, 6–
21.
Bitler, B.G., Aird, K.M., Garipov, A., Li, H., Amatangelo, M., Kossenkov, A. V, Schultz,
D.C., Liu, Q., Shih, I.-M., Conejo-Garcia, J.R., et al. (2015). Synthetic lethality by
targeting EZH2 methyltransferase activity in ARID1A-mutated cancers. Nat. Med. 21,
231–238.
Bjornsson, H.T., Brown, L.J., Fallin, M.D., Rongione, M.A., Bibikova, M., Wickham, E.,
Fan, J.-B., and Feinberg, A.P. (2007). Epigenetic specificity of loss of imprinting of the
IGF2 gene in Wilms tumors. J. Natl. Cancer Inst. 99, 1270–1273.
Brenca, M., Rossi, S., Lorenzetto, E., Piccinin, E., Piccinin, S., Rossi, F.M., Giuliano, A.,
Dei Tos, A.P., Maestro, R., and Modena, P. (2013). SMARCB1/INI1 genetic inactivation
is responsible for tumorigenic properties of epithelioid sarcoma cell line VAESBJ. Mol.
138
Cancer Ther. 12, 1060–1072.
Brennan, C.W., Verhaak, R.G.W., McKenna, A., Campos, B., Noushmehr, H., Salama,
S.R., Zheng, S., Chakravarty, D., Sanborn, J.Z., Berman, S.H., et al. (2013). The
somatic genomic landscape of glioblastoma. Cell 155, 462–477.
Brueggmann, D., Templeman, C., Starzinski-Powitz, A., Rao, N.P., Gayther, S.A.,
Lawrenson, K., Eskenazi, B., Warner, M., Bulun, S., Pearce, C., et al. (2014). Novel
three-dimensional in vitro models of ovarian endometriosis. J. Ovarian Res. 7, 17.
Bulger, M., and Groudine, M. (2011). Functional and mechanistic diversity of distal
transcription enhancers. Cell 144, 327–339.
Burgermeister, E., and Seger, R. (2007). MAPK- Kinases as Nucleo-Cytoplasmic
Shuttles for PPARγ. Cell Cycle 6, 1539–1548.
Calo, E., and Wysocka, J. (2013). Modification of enhancer chromatin: what, how, and
why? Mol. Cell 49, 825–837.
Campan, M., Weisenberger, D.J., Trinh, B., and Laird, P.W. (2009). MethyLight. pp.
325–337.
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.
Cavallaro, U., and Christofori, G. (2004). Cell adhesion and signalling by cadherins and
Ig-CAMs in cancer. Nat. Rev. Cancer 4, 118–132.
139
Chai, J., Charboneau, A.L., Betz, B.L., and Weissman, B.E. (2005). Loss of the hSNF5
gene concomitantly inactivates p21CIP/WAF1 and p16INK4a activity associated with
replicative senescence in A204 rhabdoid tumor cells. Cancer Res. 65, 10192–10198.
Chandler, R.L., Damrauer, J.S., Raab, J.R., Schisler, J.C., Wilkerson, M.D., Didion,
J.P., Starmer, J., Serber, D., Yee, D., Xiong, J., et al. (2015). Coexistent ARID1A-
PIK3CA mutations promote ovarian clear-cell tumorigenesis through pro-tumorigenic
inflammatory cytokine signalling. Nat. Commun. 6, 6118.
Charlet, J., Duymich, C.E., Lay, F.D., Mundbjerg, K., Dalsgaard Sørensen, K., Liang,
G., and Jones, P.A. (2016). Bivalent Regions of Cytosine Methylation and H3K27
Acetylation Suggest an Active Role for DNA Methylation at Enhancers. Mol. Cell 62,
422–431.
Chase, A., and Cross, N.C.P. (2011). Aberrations of EZH2 in cancer. Clin. Cancer Res.
17, 2613–2618.
Chen, X., Dong, C., Law, P.T.Y., Chan, M.T. V, Su, Z., Wang, S., Wu, W.K.K., Xu, H.,
Jemal, A., Siegel, R., et al. (2015). MicroRNA-145 targets TRIM2 and exerts tumor-
suppressing functions in epithelial ovarian cancer. Gynecol. Oncol. 139, 513–519.
Chen, Y., Lu, B., Yang, Q., Fearns, C., Yates, J.R., and Lee, J.-D. (2009). Combined
integrin phosphoproteomic analyses and small interfering RNA--based functional
screening identify key regulators for cancer cell adhesion and migration. Cancer Res.
69, 3713–3720.
Chiang, J.W., Karlan, B.Y., Cass, L., and Baldwin, R.L. (2006). BRCA1 promoter
140
methylation predicts adverse ovarian cancer prognosis. Gynecol. Oncol. 101, 403–410.
Chiappinelli, K.B., Strissel, P.L., Desrichard, A., Li, H., Henke, C., Akman, B., Hein, A.,
Rote, N.S., Cope, L.M., Snyder, A., et al. (2015). Inhibiting DNA Methylation Causes an
Interferon Response in Cancer via dsRNA Including Endogenous Retroviruses. Cell
162, 974–986.
Christman, J.K. (2002). 5-Azacytidine and 5-aza-2’-deoxycytidine as inhibitors of DNA
methylation: mechanistic studies and their implications for cancer therapy. Oncogene
21, 5483–5495.
Chuang, J.C., Warner, S.L., Vollmer, D., Vankayalapati, H., Redkar, S., Bearss, D.J.,
Qiu, X., Yoo, C.B., and Jones, P.A. (2010). S110, a 5-Aza-2’-deoxycytidine-containing
dinucleotide, is an effective DNA methylation inhibitor in vivo and can reduce tumor
growth. Mol. Cancer Ther. 9, 1443–1450.
Clapier, C.R., and Cairns, B.R. (2009). The biology of chromatin remodeling complexes.
Annu. Rev. Biochem. 78, 273–304.
Cordaux, R., and Batzer, M.A. (2009). The impact of retrotransposons on human
genome evolution. Nat. Rev. Genet. 10, 691–703.
Cornelio, D.B., DE Farias, C.B., Prusch, D.S., Heinen, T.E., Dos Santos, R.P.,
Abujamra, A.L., Schwartsmann, G., and Roesler, R. (2013). Influence of GRPR and
BDNF/TrkB signaling on the viability of breast and gynecologic cancer cells. Mol. Clin.
Oncol. 1, 148–152.
Costello, J.F., Berger, M.S., Huang, H.S., and Cavenee, W.K. (1996). Silencing of
141
p16/CDKN2 expression in human gliomas by methylation and chromatin condensation.
Cancer Res. 56, 2405–2410.
Coulondre, C., Miller, J.H., Farabaugh, P.J., and Gilbert, W. (1978). Molecular basis of
base substitution hotspots in Escherichia coli. Nature 274, 775–780.
Creighton, C.J., Morgan, M., Gunaratne, P.H., Wheeler, D.A., Gibbs, R.A., Gordon
Robertson, A., Chu, A., Beroukhim, R., Cibulskis, K., Signoretti, S., et al. (2013).
Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature
499, 43–49.
Cui, H. (2007). Loss of imprinting of IGF2 as an epigenetic marker for the risk of human
cancer. Dis. Markers 23, 105–112.
Dang, L., White, D.W., Gross, S., Bennett, B.D., Bittinger, M.A., Driggers, E.M., Fantin,
V.R., Jang, H.G., Jin, S., Keenan, M.C., et al. (2009). Cancer-associated IDH1
mutations produce 2-hydroxyglutarate. Nature 462, 739–744.
Darr, J., Klochendler, A., Isaac, S., and Eden, A. (2014). Loss of IGFBP7 expression
and persistent AKT activation contribute to SMARCB1/Snf5-mediated tumorigenesis.
Oncogene 33, 3024–3032.
Ding, J., Zhang, Z.-M., Xia, Y., Liao, G.-Q., Pan, Y., Liu, S., Zhang, Y., and Yan, Z.-S.
(2013). LSD1-mediated epigenetic modification contributes to proliferation and
metastasis of colon cancer. Br. J. Cancer 109, 994–1003.
Donson, A.M., Addo-Yobo, S.O., Handler, M.H., Gore, L., and Foreman, N.K. (2007).
MGMT promoter methylation correlates with survival benefit and sensitivity to
142
temozolomide in pediatric glioblastoma. Pediatr. Blood Cancer 48, 403–407.
Du, J., Johnson, L.M., Jacobsen, S.E., and Patel, D.J. (2015). DNA methylation
pathways and their crosstalk with histone methylation. Nat. Publ. Gr. 16, 519–532.
Eckhardt, F., Lewin, J., Cortese, R., Rakyan, V.K., Attwood, J., Burger, M., Burton, J.,
Cox, T. V, Davies, R., Down, T.A., et al. (2006). DNA methylation profiling of human
chromosomes 6, 20 and 22. Nat. Genet. 38, 1378–1385.
Ecsedi, S.I., Hernandez-Vargas, H., Lima, S.C., Herceg, Z., Adany, R., and Balazs, M.
(2013). Transposable hypomethylation is associated with metastatic capacity of primary
melanomas. Int. J. Clin. Exp. Pathol. 6, 2943–2948.
Ehrlich, M. (2010). DNA hypomethylation in cancer cells. Epigenomics 1, 239–259.
Ehrlich, M., and Lacey, M. (2013). Epigenetic Alterations in Oncogenesis. Epigenetic
Alterations Oncog. Adv. Exp. 31 Med. Biol. 754, 31–56.
Elshafae, S.M., Hassan, B.B., Supsavhad, W., Dirksen, W.P., Camiener, R.Y., Ding, H.,
Tweedle, M.F., and Rosol, T.J. (2016). Gastrin-releasing peptide receptor (GRPr)
promotes EMT, growth, and invasion in canine prostate cancer. Prostate 76, 796–809.
Esteller, M. (2001). DNA methylation patterns in hereditary human cancers mimic
sporadic tumorigenesis. Hum. Mol. Genet. 10, 3001–3007.
Esteller, M. (2011). Non-coding RNAs in human disease. Nat. Rev. Genet. 12, 861–874.
Esteller, M., Silva, J.M., Dominguez, G., Bonilla, F., Matias-Guiu, X., Lerma, E.,
Bussaglia, E., Prat, J., Harkes, I.C., Repasky, E.A., et al. (2000). Promoter
143
hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors. J.
Natl. Cancer Inst. 92, 564–569.
Euskirchen, G.M., Auerbach, R.K., Davidov, E., Gianoulis, T. a, Zhong, G., Rozowsky,
J., Bhardwaj, N., Gerstein, M.B., and Snyder, M. (2011). Diverse roles and interactions
of the SWI/SNF chromatin remodeling complex revealed using global approaches.
PLoS Genet. 7, e1002008.
Fan, F., Jin, S., Amundson, S.A., Tong, T., Fan, W., Zhao, H., Zhu, X., Mazzacurati, L.,
Li, X., Petrik, K.L., et al. (2002). ATF3 induction following DNA damage is regulated by
distinct signaling pathways and over-expression of ATF3 protein suppresses cells
growth. Oncogene 21, 7488–7496.
Fang, F., Turcan, S., Rimner, A., Kaufman, A., Giri, D., Morris, L.G.T., Shen, R.,
Seshan, V., Mo, Q., Heguy, A., et al. (2011). Breast cancer methylomes establish an
epigenomic foundation for metastasis. Sci. Transl. Med. 3, 75ra25.
Fatemi, M., Paul, T.A., Brodeur, G.M., Shokrani, B., Brim, H., and Ashktorab, H. (2014).
Epigenetic silencing of CHD5, a novel tumor-suppressor gene, occurs in early colorectal
cancer stages. Cancer 120, 172–180.
Feinberg, A.P., and Vogelstein, B. (1983). Hypomethylation of ras oncogenes in primary
human cancers. Biochem. Biophys. Res. Commun. 111, 47–54.
Fenaux, P., Mufti, G.J., Hellstrom-Lindberg, E., Santini, V., Finelli, C., Giagounidis, A.,
Schoch, R., Gattermann, N., Sanz, G., List, A., et al. (2009). Efficacy of azacitidine
compared with that of conventional care regimens in the treatment of higher-risk
144
myelodysplastic syndromes: a randomised, open-label, phase III study. Lancet. Oncol.
10, 223–232.
Ferguson-Smith, A.C. (2011). Genomic imprinting: the emergence of an epigenetic
paradigm. Nat. Rev. Genet. 12, 565–575.
Ferreira, H.J., Heyn, H., Moutinho, C., and Esteller, M. (2012). CpG island
hypermethylation-associated silencing of small nucleolar RNAs in human cancer. RNA
Biol. 9, 881–890.
Figueroa, M.E., Abdel-Wahab, O., Lu, C., Ward, P.S., Patel, J., Shih, A., Li, Y.,
Bhagwat, N., Vasanthakumar, A., Fernandez, H.F., et al. (2010). Leukemic IDH1 and
IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and
impair hematopoietic differentiation. Cancer Cell 18, 553–567.
Freedman, V., Shin, S., Aaronson, S.A., Todaro, G.J., Berman, L., Stulberg, C.S., Chen,
T.T., Heidelberger, C., Dulbecco, R., Freedman, V.H., et al. (1974). Cellular
tumorigenicity in nude mice: Correlation with cell growth in semi-solid medium. Cell 3,
355–359.
Frühwald, M.C., Biegel, J.A., Bourdeaut, F., Roberts, C.W.M., and Chi, S.N. (2016).
Atypical teratoid/rhabdoid tumors-current concepts, advances in biology, and potential
future therapies. Neuro. Oncol. 18, 764–778.
Fu, L., and Zhang, S. (2014). RASSF1A promotes apoptosis and suppresses the
proliferation of ovarian cancer cells. Int. J. Mol. Med. 33, 1153–1160.
Fujita, T., Igarashi, J., Okawa, E.R., Gotoh, T., Manne, J., Kolla, V., Kim, J., Zhao, H.,
145
Pawel, B.R., London, W.B., et al. (2008). CHD5, a tumor suppressor gene deleted from
1p36.31 in neuroblastomas. J. Natl. Cancer Inst. 100, 940–949.
Gadd, S., Sredni, S.T., Huang, C.-C., Perlman, E.J., and Renal Tumor Committee of the
Children’s Oncology Group (2010). Rhabdoid tumor: gene expression clues to
pathogenesis and potential therapeutic targets. Lab. Invest. 90, 724–738.
Gama-sosa, M. a., Slagel, V. a., Trewyn, R.W., Oxenhandler, R., Kuo, K.C., Gehrke,
C.W., and Ehrlich, M. (1983). The 5-methylcytosine content of DNA from human tumors.
Nucleic Acids Res. 11, 6883–6894.
Gardner, K.E., Allis, C.D., and Strahl, B.D. (2011). OPERating ON Chromatin, a Colorful
Language where Context Matters. J. Mol. Biol. 409, 36–46.
Gkikopoulos, T., Schofield, P., Singh, V., Pinskaya, M., Mellor, J., Smolle, M.,
Workman, J.L., Barton, G.J., and Owen-Hughes, T. (2011). A role for Snf2-related
nucleosome-spacing enzymes in genome-wide nucleosome organization. Science 333,
1758–1760.
Gorringe, K.L., Choong, D.Y., Williams, L.H., Ramakrishna, M., Sridhar, A., Qiu, W.,
Bearfoot, J.L., and Campbell, I.G. (2008). Mutation and methylation analysis of the
chromodomain-helicase-DNA binding 5 gene in ovarian cancer. Neoplasia 10, 1253–
1258.
Guan, B., Mao, T.-L., Panuganti, P.K., Kuhn, E., Kurman, R.J., Maeda, D., Chen, E.,
Jeng, Y.-M., Wang, T.-L., and Shih, I.-M. (2011a). Mutation and loss of expression of
ARID1A in uterine low-grade endometrioid carcinoma. Am. J. Surg. Pathol. 35, 625–
146
632.
Guan, B., Wang, T.-L., and Shih, I.-M. (2011b). ARID1A, a factor that promotes
formation of SWI/SNF-mediated chromatin remodeling, is a tumor suppressor in
gynecologic cancers. Cancer Res. 71, 6718–6727.
Guan, B., Rahmanto, Y.S., Wu, R.-C., Wang, Y., Wang, Z., Wang, T.-L., and Shih, I.-M.
(2014). Roles of deletion of Arid1a, a tumor suppressor, in mouse ovarian
tumorigenesis. J. Natl. Cancer Inst. 106.
Guidi, C.J., Mudhasani, R., Hoover, K., Koff, A., Leav, I., Imbalzano, A.N., and Jones,
S.N. (2006). Functional interaction of the retinoblastoma and Ini1/Snf5 tumor
suppressors in cell growth and pituitary tumorigenesis. Cancer Res. 66, 8076–8082.
Gupta, M., Babic, A., Beck, A.H., and Terry, K. (2016). TNF-α expression, risk factors,
and inflammatory exposures in ovarian cancer: evidence for an inflammatory pathway of
ovarian carcinogenesis? Hum. Pathol. 54, 82–91.
Haas, J.E., Palmer, N.F., Weinberg, A.G., and Beckwith, J.B. (1981). Ultrastructure of
malignant rhabdoid tumor of the kidney. A distinctive renal tumor of children. Hum.
Pathol. 12, 646–657.
Han, H., Yang, X., Pandiyan, K., and Liang, G. (2013). Synergistic re-activation of
epigenetically silenced genes by combinatorial inhibition of DNMTs and LSD1 in cancer
cells. PLoS One 8, e75136.
Han Li, C., and Chen, Y. (2015). Targeting EZH2 for cancer therapy: progress and
perspective. Curr. Protein Pept. Sci. 16, 559–570.
147
Hanahan, D., and Weinberg, R.A. (2011). Hallmarks of cancer: the next generation. Cell
144, 646–674.
Hayami, S., Kelly, J.D., Cho, H.-S., Yoshimatsu, M., Unoki, M., Tsunoda, T., Field, H.I.,
Neal, D.E., Yamaue, H., Ponder, B.A.J., et al. (2011). Overexpression of LSD1
contributes to human carcinogenesis through chromatin regulation in various cancers.
Int. J. Cancer 128, 574–586.
He, X., Khurana, A., Roy, D., Kaufmann, S., and Shridhar, V. (2014). Loss of HSulf-1
expression enhances tumorigenicity by inhibiting Bim expression in ovarian cancer. Int.
J. Cancer 135, 1783–1789.
Helbo, A., Treppendahl, M., Aslan, D., Dimopoulos, K., Nandrup-Bus, C., Holm, M.,
Andersen, M., Liang, G., Kristensen, L., and Grønbæk, K. (2015). Hypermethylation of
the VTRNA1-3 Promoter is Associated with Poor Outcome in Lower Risk
Myelodysplastic Syndrome Patients. Genes (Basel). 6, 977–990.
Helleman, J., Jansen, M.P.H.M., Span, P.N., van Staveren, I.L., Massuger, L.F.A.G.,
Meijer-van Gelder, M.E., Sweep, F.C.G.J., Ewing, P.C., van der Burg, M.E.L., Stoter,
G., et al. (2006). Molecular profiling of platinum resistant ovarian cancer. Int. J. Cancer
118, 1963–1971.
Hendrix, N.D., Wu, R., Kuick, R., Schwartz, D.R., Fearon, E.R., and Cho, K.R. (2006).
Fibroblast growth factor 9 has oncogenic activity and is a downstream target of Wnt
signaling in ovarian endometrioid adenocarcinomas. Cancer Res. 66, 1354–1362.
Herman, J.G., Latif, F., Weng, Y., Lerman, M.I., Zbar, B., Liu, S., Samid, D., Duan, D.S.,
148
Gnarra, J.R., and Linehan, W.M. (1994). Silencing of the VHL tumor-suppressor gene
by DNA methylation in renal carcinoma. Proc. Natl. Acad. Sci. U. S. A. 91, 9700–9704.
Hildebrand, J.D., Schaller, M.D., and Parsons, J.T. (1995). Paxillin, a tyrosine
phosphorylated focal adhesion-associated protein binds to the carboxyl terminal domain
of focal adhesion kinase. Mol. Biol. Cell 6, 637–647.
Hill, V.K., Shinawi, T., Ricketts, C.J., Krex, D., Schackert, G., Bauer, J., Wei, W.,
Cruickshank, G., Maher, E.R., and Latif, F. (2014). Stability of the CpG island
methylator phenotype during glioma progression and identification of methylated loci in
secondary glioblastomas. BMC Cancer 14, 506.
Hinoue, T., Weisenberger, D.J., Lange, C.P.E., Shen, H., Byun, H.-M., Van Den Berg,
D., Malik, S., Pan, F., Noushmehr, H., van Dijk, C.M., et al. (2012). Genome-scale
analysis of aberrant DNA methylation in colorectal cancer. Genome Res. 22, 271–282.
Hitchins, M.P., Wong, J.J.L., Suthers, G., Suter, C.M., Martin, D.I.K., Hawkins, N.J., and
Ward, R.L. (2007). Inheritance of a cancer-associated MLH1 germ-line epimutation. N.
Engl. J. Med. 356, 697–705.
Huang, J., Deng, Q., Wang, Q., Li, K.-Y., Dai, J.-H., Li, N., Zhu, Z.-D., Zhou, B., Liu, X.-
Y., Liu, R.-F., et al. (2012). Exome sequencing of hepatitis B virus-associated
hepatocellular carcinoma. Nat. Genet. 44, 1117–1121.
Irizarry, R.A., Ladd-Acosta, C., Wen, B., Wu, Z., Montano, C., Onyango, P., Cui, H.,
Gabo, K., Rongione, M., Webster, M., et al. (2009). The human colon cancer
methylome shows similar hypo- and hypermethylation at conserved tissue-specific CpG
149
island shores. Nat. Genet. 41, 178–186.
Isakoff, M.S., Sansam, C.G., Tamayo, P., Subramanian, A., Evans, J.A., Fillmore, C.M.,
Wang, X., Biegel, J.A., Pomeroy, S.L., Mesirov, J.P., et al. (2005). Inactivation of the
Snf5 tumor suppressor stimulates cell cycle progression and cooperates with p53 loss
in oncogenic transformation. Proc. Natl. Acad. Sci. U. S. A. 102, 17745–17750.
Itamochi, H., Kigawa, J., and Terakawa, N. (2008). Mechanisms of chemoresistance
and poor prognosis in ovarian clear cell carcinoma. Cancer Sci. 99, 653–658.
Johnson, J.L., Pillai, S., Pernazza, D., Sebti, S.M., Lawrence, N.J., and Chellappan,
S.P. (2012). Regulation of matrix metalloproteinase genes by E2F transcription factors:
Rb-Raf-1 interaction as a novel target for metastatic disease. Cancer Res. 72, 516–526.
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 Baylin, S.B. (2007). The Epigenomics of Cancer. Cell 128, 683–692.
Jones, P.A., and Laird, P.W. (1999). Cancer epigenetics comes of age. Nat. Genet. 21,
163–167.
Jones, P.A., and Liang, G. (2009). Rethinking how DNA methylation patterns are
maintained. Nat. Rev. Genet. 10, 805–811.
Jones, J., Otu, H., Spentzos, D., Kolia, S., Inan, M., Beecken, W.D., Fellbaum, C., Gu,
X., Joseph, M., Pantuck, A.J., et al. (2005). Gene signatures of progression and
metastasis in renal cell cancer. Clin. Cancer Res. 11, 5730–5739.
150
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.
Juo, Y.-Y., Gong, X.-J., Mishra, A., Cui, X., Baylin, S.B., Azad, N.S., and Ahuja, N.
(2015). Epigenetic therapy for solid tumors: from bench science to clinical trials.
Epigenomics 7, 215–235.
Kadoch, C., and Crabtree, G.R. (2015). Mammalian SWI/SNF chromatin remodeling
complexes and cancer: Mechanistic insights gained from human genomics. Sci. Adv. 1,
e1500447.
Kadoch, C., Hargreaves, D.C., Hodges, C., Elias, L., Ho, L., Ranish, J., and Crabtree,
G.R. (2013). Proteomic and bioinformatic analysis of mammalian SWI/SNF complexes
identifies extensive roles in human malignancy. Nat. Genet. 45, 592–601.
Kaneda, A., and Feinberg, A.P. (2005). Loss of imprinting of IGF2: a common
epigenetic modifier of intestinal tumor risk. Cancer Res. 65, 11236–11240.
Karami, S., Andreotti, G., Liao, L.M., Pfeiffer, R.M., Weinstein, S.J., Purdue, M.P.,
Hofmann, J.N., Albanes, D., Mannisto, S., and Moore, L.E. (2015). LINE1 methylation
levels in pre-diagnostic leukocyte DNA and future renal cell carcinoma risk. Epigenetics
10, 282–292.
Kasinathan, S., and Henikoff, S. (2014). 5-Aza-CdR Delivers a Gene Body Blow.
Cancer Cell 26, 449–451.
Kelly, T.K., De Carvalho, D.D., and Jones, P.A. (2010a). Epigenetic modifications as
151
therapeutic targets. Nat. Biotechnol. 28, 1069–1078.
Kelly, T.K., Miranda, T.B., Liang, G., Berman, B.P., Lin, J.C., Tanay, A., and Jones, P.A.
(2010b). H2A.Z maintenance during mitosis reveals nucleosome shifting on mitotically
silenced genes. Mol. Cell 39, 901–911.
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.
Khazaei, M.R., Bunk, E.C., Hillje, A.-L., Jahn, H.M., Riegler, E.M., Knoblich, J.A.,
Young, P., and Schwamborn, J.C. (2011). The E3-ubiquitin ligase TRIM2 regulates
neuronal polarization. J. Neurochem. 117, 29–37.
Kim, K.H., and Roberts, C.W.M. (2014). Mechanisms by which SMARCB1 loss drives
rhabdoid tumor growth. Cancer Genet. 207, 365–372.
Kim, T.-K., and Shiekhattar, R. (2015). Architectural and Functional Commonalities
between Enhancers and Promoters. Cell 162, 948–959.
Kim, K.H., Kim, W., Howard, T.P., Vazquez, F., Tsherniak, A., Wu, J.N., Wang, W.,
Haswell, J.R., Walensky, L.D., Hahn, W.C., et al. (2015a). SWI/SNF-mutant cancers
depend on catalytic and non-catalytic activity of EZH2. Nat. Med. 21, 1491–1496.
Kim, K.-H., Park, B., Rhee, D.-K., and Pyo, S. (2015b). Acrylamide Induces Senescence
in Macrophages through a Process Involving ATF3, ROS, p38/JNK, and a Telomerase-
Independent Pathway. Chem. Res. Toxicol. 28, 71–86.
152
King, C.M., Barbara, C., Prentice, A., Brenton, J.D., and Charnock-Jones, D.S. (2016).
Models of endometriosis and their utility in studying progression to ovarian clear cell
carcinoma. J. Pathol. 238, 185–196.
Ko, M., Huang, Y., Jankowska, A.M., Pape, U.J., Tahiliani, M., Bandukwala, H.S., An,
J., Lamperti, E.D., Koh, K.P., Ganetzky, R., et al. (2010). Impaired hydroxylation of 5-
methylcytosine in myeloid cancers with mutant TET2. Nature 468, 839–843.
Kohli, R.M., and Zhang, Y. (2013). TET enzymes, TDG and the dynamics of DNA
demethylation. Nature 502, 472–479.
Kuang, Y., El-Khoueiry, A., Taverna, P., Ljungman, M., and Neamati, N. (2015).
Guadecitabine (SGI-110) priming sensitizes hepatocellular carcinoma cells to
oxaliplatin. Mol. Oncol. 9, 1799–1814.
Kulis, M., Heath, S., Bibikova, M., Queirós, A.C., Navarro, A., Clot, G., Martínez-Trillos,
A., Castellano, G., Brun-Heath, I., Pinyol, M., et al. (2012). Epigenomic analysis detects
widespread gene-body DNA hypomethylation in chronic lymphocytic leukemia. Nat.
Genet. 44, 1236–1242.
Kulis, M., Queirós, A.C., Beekman, R., and Martín-Subero, J.I. (2013). Intragenic DNA
methylation in transcriptional regulation, normal differentiation and cancer. Biochim.
Biophys. Acta - Gene Regul. Mech. 1829, 1161–1174.
Kumar, C.C. (1998). Signaling by integrin receptors. Oncogene 17, 1365–1373.
Kuwahara, Y., Wei, D., Durand, J., and Weissman, B.E. (2013). SNF5 reexpression in
malignant rhabdoid tumors regulates transcription of target genes by recruitment of
153
SWI/SNF complexes and RNAPII to the transcription start site of their promoters. Mol.
Cancer Res. 11, 251–260.
Kwasnieski, J.C., Fiore, C., Chaudhari, H.G., and Cohen, B.A. (2014). High-throughput
functional testing of ENCODE segmentation predictions. Genome Res. 24, 1595–1602.
Lawrenson, K., Lee, N., Torres, H.A.M., Lee, J.M., Brueggmann, D., Rao, P.N.,
Noushmehr, H., and Gayther, S.A. (2014). Src as a novel therapeutic target for
endometriosis. Gynecol. Oncol. 135, 100–107.
Lay, F.D., Liu, Y., Kelly, T.K., Witt, H., Farnham, P.J., Jones, P. a, and Berman, B.P.
The role of DNA methylation in directing the functional organization of the cancer
epigenome. Genome Res. 1–11.
Lay, F.D., Liu, Y., Kelly, T.K., Witt, H., Farnham, P.J., Jones, P.A., and Berman, B.P.
(2015). The role of DNA methylation in directing the functional organization of the
cancer epigenome. Genome Res. 25, 467–477.
Lee, K.-S., Park, J.-L., Lee, K., Richardson, L.E., Johnson, B.H., Lee, H.-S., Lee, J.-S.,
Kim, S.-B., Kwon, O.-H., Song, K.S., et al. (2014). nc886, a non-coding RNA of anti-
proliferative role, is suppressed by CpG DNA methylation in human gastric cancer.
Oncotarget 5, 3944–3955.
Lei, I., West, J., Yan, Z., Gao, X., Fang, P., Dennis, J.H., Gnatovskiy, L., Wang, W.,
Kingston, R.E., and Wang, Z. (2015). BAF250a Protein Regulates Nucleosome
Occupancy and Histone Modifications in Priming Embryonic Stem Cell Differentiation. J.
Biol. Chem. 290, 19343–19352.
154
Li, E., Beard, C., and Jaenisch, R. (1993). Role for DNA methylation in genomic
imprinting. Nature 366, 362–365.
Li, G., Ruan, X., Auerbach, R.K., Sandhu, K.S., Zheng, M., Wang, P., Poh, H.M., Goh,
Y., Lim, J., Zhang, J., et al. (2012). Extensive promoter-centered chromatin interactions
provide a topological basis for transcription regulation. Cell 148, 84–98.
Li, J., Huang, Q., Zeng, F., Li, W., He, Z., Chen, W., Zhu, W., and Zhang, B. (2014). The
prognostic value of global DNA hypomethylation in cancer: a meta-analysis. PLoS One
9, e106290.
Li, M., Hada, A., Sen, P., Olufemi, L., Hall, M.A., Smith, B.Y., Forth, S., McKnight, J.N.,
Patel, A., Bowman, G.D., et al. (2015). Dynamic regulation of transcription factors by
nucleosome remodeling. Elife 4, e06249.
Li, X., Yao, X., Wang, Y., Hu, F., Wang, F., Jiang, L., Liu, Y., Wang, D., Sun, G., and
Zhao, Y. (2013). MLH1 promoter methylation frequency in colorectal cancer patients
and related clinicopathological and molecular features. PLoS One 8, e59064.
Liang, G., Chan, M.F., Tomigahara, Y., Tsai, Y.C., Gonzales, F.A., Li, E., Laird, P.W.,
and Jones, P.A. (2002). Cooperativity between DNA methyltransferases in the
maintenance methylation of repetitive elements. Mol. Cell. Biol. 22, 480–491.
Licht, J.D. (2015). DNA Methylation Inhibitors in Cancer Therapy: The Immunity
Dimension. Cell 162, 938–939.
Lim, S., Janzer, A., Becker, A., Zimmer, A., Schüle, R., Buettner, R., and Kirfel, J.
(2010). Lysine-specific demethylase 1 (LSD1) is highly expressed in ER-negative breast
155
cancers and a biomarker predicting aggressive biology. Carcinogenesis 31, 512–520.
Lin, J.C., Jeong, S., Liang, G., Takai, D., Fatemi, M., Tsai, Y.C., Egger, G., Gal-Yam,
E.N., and Jones, P.A. (2007). Role of nucleosomal occupancy in the epigenetic
silencing of the MLH1 CpG island. Cancer Cell 12, 432–444.
Liu, Y., Siegmund, K.D., Laird, P.W., and Berman, B.P. (2012). Bis-SNP: combined
DNA methylation and SNP calling for Bisulfite-seq data. Genome Biol. 13, R61.
Lou, S., Lee, H.-M., Qin, H., Li, J.-W., Gao, Z., Liu, X., Chan, L.L., Kl Lam, V., So, W.-
Y., Wang, Y., et al. (2014). Whole-genome bisulfite sequencing of multiple individuals
reveals complementary roles of promoter and gene body methylation in transcriptional
regulation. Genome Biol. 15, 408.
Lu, J., Song, G., Tang, Q., Zou, C., Han, F., Zhao, Z., Yong, B., Yin, J., Xu, H., Xie, X.,
et al. (2015). IRX1 hypomethylation promotes osteosarcoma metastasis via induction of
CXCL14 / NF- κ B signaling. J. Clin. Invest. 125, 1–18.
Luger, K., Mäder, A.W., Richmond, R.K., Sargent, D.F., and Richmond, T.J. (1997).
Crystal structure of the nucleosome core particle at 2.8 A resolution. Nature 389, 251–
260.
Macagno, J.P., Diaz Vera, J., Yu, Y., MacPherson, I., Sandilands, E., Palmer, R.,
Norman, J.C., Frame, M., and Vidal, M. (2014). FAK acts as a suppressor of RTK-MAP
kinase signalling in Drosophila melanogaster epithelia and human cancer cells. PLoS
Genet. 10, e1004262.
Mair, B., Kubicek, S., and Nijman, S.M.B. (2014). Exploiting epigenetic vulnerabilities for
156
cancer therapeutics. Trends Pharmacol. Sci. 35, 136–145.
Mamo, A., Cavallone, L., Tuzmen, S., Chabot, C., Ferrario, C., Hassan, S., Edgren, H.,
Kallioniemi, O., Aleynikova, O., Przybytkowski, E., et al. (2012). An integrated genomic
approach identifies ARID1A as a candidate tumor-suppressor gene in breast cancer.
Oncogene 31, 2090–2100.
Masliah-Planchon, J., Bièche, I., Guinebretière, J.-M., Bourdeaut, F., and Delattre, O.
(2015). SWI/SNF chromatin remodeling and human malignancies. Annu. Rev. Pathol.
10, 145–171.
Matei, D., Kelich, S., Cao, L., Menning, N., Emerson, R.E., Rao, J., Jeng, M.H., and
Sledge, G.W. (2007). PDGF BB induces VEGF secretion in ovarian cancer. Cancer
Biol. Ther. 6, 1951–1959.
Matsuzaki, S., Yoshino, K., Ueda, Y., Matsuzaki, S., Kakuda, M., Okazawa, A., Egawa-
Takata, T., Kobayashi, E., Kimura, T., Siegel, R., et al. (2015). Potential targets for
ovarian clear cell carcinoma: a review of updates and future perspectives. Cancer Cell
Int. 15, 117.
McCabe, M.T., and Creasy, C.L. (2014). EZH2 as a potential target in cancer therapy.
Epigenomics 6, 341–351.
McKenna, E.S., Sansam, C.G., Cho, Y.-J., Greulich, H., Evans, J.A., Thom, C.S.,
Moreau, L.A., Biegel, J.A., Pomeroy, S.L., and Roberts, C.W.M. (2008). Loss of the
epigenetic tumor suppressor SNF5 leads to cancer without genomic instability. Mol.
Cell. Biol. 28, 6223–6233.
157
Meerbrey, K.L., Hu, G., Kessler, J.D., Roarty, K., Li, M.Z., Fang, J.E., Herschkowitz, J.I.,
Burrows, A.E., Ciccia, A., Sun, T., et al. (2011). The pINDUCER lentiviral toolkit for
inducible RNA interference in vitro and in vivo. Proc. Natl. Acad. Sci. U. S. A. 108,
3665–3670.
Meissner, A., Mikkelsen, T.S., Gu, H., Wernig, M., Hanna, J., Sivachenko, A., Zhang,
X., Bernstein, B.E., Nusbaum, C., Jaffe, D.B., et al. (2008). Genome-scale DNA
methylation maps of pluripotent and differentiated cells. Nature 454, 766–770.
Metzger, E., Wissmann, M., Yin, N., Müller, J.M., Schneider, R., Peters, A.H.F.M.,
Günther, T., Buettner, R., and Schüle, R. (2005). LSD1 demethylates repressive histone
marks to promote androgen-receptor-dependent transcription. Nature 437, 436.
Mikkelsen, T.S., Ku, M., Jaffe, D.B., Issac, B., Lieberman, E., Giannoukos, G., Alvarez,
P., Brockman, W., Kim, T.-K., Koche, R.P., et al. (2007a). Genome-wide maps of
chromatin state in pluripotent and lineage-committed cells. Nature 448, 553–560.
Mikkelsen, T.S., Ku, M., Jaffe, D.B., Issac, B., Lieberman, E., Giannoukos, G., Alvarez,
P., Brockman, W., Kim, T.-K., Koche, R.P., et al. (2007b). Genome-wide maps of
chromatin state in pluripotent and lineage-committed cells. Nature 448, 553–560.
Moen, E.L., Stark, A.L., Zhang, W., Dolan, M.E., and Godley, L.A. (2014). The Role of
Gene Body Cytosine Modifications in MGMT Expression and Sensitivity to
Temozolomide. Mol. Cancer Ther. 13, 1334–1344.
Moreno, D.A., Scrideli, C.A., Cortez, M.A.A., de Paula Queiroz, R., Valera, E.T., da
Silva Silveira, V., Yunes, J.A., Brandalise, S.R., and Tone, L.G. (2010). Differential
158
expression of HDAC3, HDAC7 and HDAC9 is associated with prognosis and survival in
childhood acute lymphoblastic leukaemia. Br. J. Haematol. 150, 665–673.
Mori, S., Chang, J.T., Andrechek, E.R., Matsumura, N., Baba, T., Yao, G., Kim, J.W.,
Gatza, M., Murphy, S., and Nevins, J.R. (2009). Anchorage-independent cell growth
signature identifies tumors with metastatic potential. Oncogene 28, 2796–2805.
Muchardt, C., and Yaniv, M. (1999). ATP-dependent chromatin remodelling: SWI/SNF
and Co. are on the job. J. Mol. Biol. 293, 187–198.
Muratcioglu, S., Presman, D.M., Pooley, J.R., Grøntved, L., Hager, G.L., Nussinov, R.,
Keskin, O., Gursoy, A., Mangelsdorf, D.J., Thummel, C., et al. (2015). Structural
Modeling of GR Interactions with the SWI/SNF Chromatin Remodeling Complex and
C/EBP. Biophys. J. 109, 1227–1239.
Murr, R. (2010). Interplay between different epigenetic modifications and mechanisms.
(Elsevier Inc.).
Nagarajan, R.P., Zhang, B., Bell, R.J.A., Johnson, B.E., Olshen, A.B., Sundaram, V., Li,
D., Graham, A.E., Diaz, A., Fouse, S.D., et al. (2014). Recurrent epimutations activate
gene body promoters in primary glioblastoma. Genome Res. 24, 761–774.
Nevins, J.R. (2001). The Rb/E2F pathway and cancer. Hum. Mol. Genet. 10, 699–703.
Nezhat, F.R., Apostol, R., Nezhat, C., and Pejovic, T. (2015). New insights in the
pathophysiology of ovarian cancer and implications for screening and prevention. Am. J.
Obstet. Gynecol. 213, 262–267.
159
Nord, A.S., Blow, M.J., Attanasio, C., Akiyama, J.A., Holt, A., Hosseini, R.,
Phouanenavong, S., Plajzer-Frick, I., Shoukry, M., Afzal, V., et al. (2013). Rapid and
pervasive changes in genome-wide enhancer usage during mammalian development.
Cell 155, 1521–1531.
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.
Okamoto, T., Mandai, M., Matsumura, N., Yamaguchi, K., Kondoh, H., Amano, Y.,
Baba, T., Hamanishi, J., Abiko, K., Kosaka, K., et al. (2015). Hepatocyte nuclear factor-
1β (HNF-1β) promotes glucose uptake and glycolytic activity in ovarian clear cell
carcinoma. Mol. Carcinog. 54, 35–49.
Okano, M., Bell, D.W., Haber, D.A., and Li, E. (1999). DNA methyltransferases Dnmt3a
and Dnmt3b are essential for de novo methylation and mammalian development. Cell
99, 247–257.
Oki, Y., Aoki, E., and Issa, J.-P.J. (2007). Decitabine--bedside to bench. Crit. Rev.
Oncol. Hematol. 61, 140–152.
Olsen, E.A., Kim, Y.H., Kuzel, T.M., Pacheco, T.R., Foss, F.M., Parker, S., Frankel,
S.R., Chen, C., Ricker, J.L., Arduino, J.M., et al. (2007). Phase IIB Multicenter Trial of
Vorinostat in Patients With Persistent, Progressive, or Treatment Refractory Cutaneous
T-Cell Lymphoma. J. Clin. Oncol. 25, 3109–3115.
160
Pandiyan, K., You, J.S., Yang, X., Dai, C., Zhou, X.J., Baylin, S.B., Jones, P.A., and
Liang, G. (2013). Functional DNA demethylation is accompanied by chromatin
accessibility. Nucleic Acids Res. 41, 3973–3985.
Papageorgiou, E., Pitulis, N., Msaouel, P., Lembessis, P., and Koutsilieris, M. (2007).
The non-genomic crosstalk between PPAR-γ ligands and ERK1/2 in cancer cell lines.
Expert Opin. Ther. Targets 11, 1071–1085.
Pastor, W.A., Pape, U.J., Huang, Y., Henderson, H.R., Lister, R., Ko, M., McLoughlin,
E.M., Brudno, Y., Mahapatra, S., Kapranov, P., et al. (2011). Genome-wide mapping of
5-hydroxymethylcytosine in embryonic stem cells. Nature 473, 394–397.
Pastor, W.A., Aravind, L., and Rao, A. (2013). TETonic shift: biological roles of TET
proteins in DNA demethylation and transcription. Nat. Rev. Mol. Cell Biol. 14, 341–356.
Pearce, C.L., Templeman, C., Rossing, M.A., Lee, A., Near, A.M., Webb, P.M., Nagle,
C.M., Doherty, J.A., Cushing-Haugen, K.L., Wicklund, K.G., et al. (2012). Association
between endometriosis and risk of histological subtypes of ovarian cancer: a pooled
analysis of case-control studies. Lancet. Oncol. 13, 385–394.
Peña-Llopis, S., Vega-Rubín-de-Celis, S., Liao, A., Leng, N., Pavía-Jiménez, A., Wang,
S., Yamasaki, T., Zhrebker, L., Sivanand, S., Spence, P., et al. (2012). BAP1 loss
defines a new class of renal cell carcinoma. Nat. Genet. 44, 751–759.
Peterson, C.L., and Workman, J.L. (2000). Promoter targeting and chromatin
remodeling by the SWI/SNF complex. Curr. Opin. Genet. Dev. 10, 187–192.
Piao, L., Suzuki, T., Dohmae, N., Nakamura, Y., and Hamamoto, R. (2015). SUV39H2
161
methylates and stabilizes LSD1 by inhibiting polyubiquitination in human cancer cells.
Oncotarget 6, 16939–16950.
Pontier, D.B., and Gribnau, J. (2011). Xist regulation and function explored. Hum.
Genet. 130, 223–236.
Raab, J.R., Resnick, S., Magnuson, T., Clapier, C., Cairns, B., Wang, W., Xue, Y.,
Zhou, S., Kuo, A., Cairns, B., et al. (2015). Genome-Wide Transcriptional Regulation
Mediated by Biochemically Distinct SWI/SNF Complexes. PLOS Genet. 11, e1005748.
Ravenel, J.D., Broman, K.W., Perlman, E.J., Niemitz, E.L., Jayawardena, T.M., Bell,
D.W., Haber, D.A., Uejima, H., and Feinberg, A.P. (2001). Loss of Imprinting of Insulin-
Like Growth Factor-II (IGF2) Gene in Distinguishing Specific Biologic Subtypes of Wilms
Tumor. JNCI J. Natl. Cancer Inst. 93, 1698–1703.
Reincke, B.S., Rosson, G.B., Oswald, B.W., and Wright, C.F. (2003). INI1 expression
induces cell cycle arrest and markers of senescence in malignant rhabdoid tumor cells.
J. Cell. Physiol. 194, 303–313.
Reinhard, H., Reinert, J., Beier, R., Furtwängler, R., Alkasser, M., Rutkowski, S.,
Frühwald, M., Koscielniak, E., Leuschner, I., Kaatsch, P., et al. (2008). Rhabdoid tumors
in children: prognostic factors in 70 patients diagnosed in Germany. Oncol. Rep. 19,
819–823.
Rhee, I., Jair, K.W., Yen, R.W., Lengauer, C., Herman, J.G., Kinzler, K.W., Vogelstein,
B., Baylin, S.B., and Schuebel, K.E. (2000). CpG methylation is maintained in human
cancer cells lacking DNMT1. Nature 404, 1003–1007.
162
Rhee, I., Bachman, K.E., Park, B.H., Jair, K.-W., Yen, R.-W.C., Schuebel, K.E., Cui, H.,
Feinberg, A.P., Lengauer, C., Kinzler, K.W., et al. (2002). DNMT1 and DNMT3b
cooperate to silence genes in human cancer cells. Nature 416, 552–556.
Robertson, K.D., and Jones, P.A. (1997). Dynamic interrelationships between DNA
replication, methylation, and repair. Am. J. Hum. Genet. 61, 1220–1224.
Romero, O.A., and Sanchez-Cespedes, M. (2014). The SWI/SNF genetic blockade:
effects in cell differentiation, cancer and developmental diseases. Oncogene 33, 2681–
2689.
Rosanò, L., Cianfrocca, R., Masi, S., Spinella, F., Di Castro, V., Biroccio, A., Salvati, E.,
Nicotra, M.R., Natali, P.G., and Bagnato, A. (2009). Beta-arrestin links endothelin A
receptor to beta-catenin signaling to induce ovarian cancer cell invasion and metastasis.
Proc. Natl. Acad. Sci. U. S. A. 106, 2806–2811.
Roulois, D., Loo Yau, H., Singhania, R., Wang, Y., Danesh, A., Shen, S.Y., Han, H.,
Liang, G., Jones, P.A., Pugh, T.J., et al. (2015). DNA-Demethylating Agents Target
Colorectal Cancer Cells by Inducing Viral Mimicry by Endogenous Transcripts. Cell 162,
961–973.
Ruthenburg, A.J., Allis, C.D., and Wysocka, J. (2007). Methylation of Lysine 4 on
Histone H3: Intricacy of Writing and Reading a Single Epigenetic Mark. Mol. Cell 25,
15–30.
Ruthenburg, A.J., Li, H., Milne, T. a, Dewell, S., McGinty, R.K., Yuen, M., Ueberheide,
B., Dou, Y., Muir, T.W., Patel, D.J., et al. (2011). Recognition of a mononucleosomal
163
histone modification pattern by BPTF via multivalent interactions. Cell 145, 692–706.
Saied, M.H., Marzec, J., Khalid, S., Smith, P., Down, T.A., Rakyan, V.K., Molloy, G.,
Raghavan, M., Debernardi, S., and Young, B.D. (2012). Genome wide analysis of acute
myeloid leukemia reveal leukemia specific methylome and subtype specific
hypomethylation of repeats. PLoS One 7, e33213.
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.
Sakai, T., Toguchida, J., Ohtani, N., Yandell, D.W., Rapaport, J.M., and Dryja, T.P.
(1991). Allele-specific hypermethylation of the retinoblastoma tumor-suppressor gene.
Am. J. Hum. Genet. 48, 880–888.
Samartzis, E.P., Noske, A., Dedes, K.J., Fink, D., and Imesch, P. (2013). ARID1A
mutations and PI3K/AKT pathway alterations in endometriosis and endometriosis-
associated ovarian carcinomas. Int. J. Mol. Sci. 14, 18824–18849.
Sandoval, J., and Esteller, M. (2012). Cancer epigenomics: beyond genomics. Curr.
Opin. Genet. Dev. 22, 50–55.
Sansam, C.G., and Roberts, C.W.M. (2006). Epigenetics and cancer: altered chromatin
remodeling via Snf5 loss leads to aberrant cell cycle regulation. Cell Cycle 5, 621–624.
Sawan, C. (2010). 3 – Histone Modifications and Cancer. In Advances in Genetics, pp.
57–85.
164
Schlesinger, Y., Straussman, R., Keshet, I., Farkash, S., Hecht, M., Zimmerman, J.,
Eden, E., Yakhini, Z., Ben-Shushan, E., Reubinoff, B.E., et al. (2007). Polycomb-
mediated methylation on Lys27 of histone H3 pre-marks genes for de novo methylation
in cancer. Nat. Genet. 39, 232–236.
Segal, E., and Widom, J. (2009). What controls nucleosome positions? Trends Genet.
25, 335–343.
Sharma, S., Kelly, T.K., and Jones, P.A. (2010). Epigenetics in cancer. Carcinogenesis
31, 27–36.
Sharma, S., De Carvalho, D.D., Jeong, S., Jones, P.A., and Liang, G. (2011).
Nucleosomes containing methylated DNA stabilize DNA methyltransferases 3A/3B and
ensure faithful epigenetic inheritance. PLoS Genet. 7, e1001286.
Sharp, A.J., Stathaki, E., Migliavacca, E., Brahmachary, M., Montgomery, S.B., Dupre,
Y., and Antonarakis, S.E. (2011). DNA methylation profiles of human active and inactive
X chromosomes. Genome Res. 21, 1592–1600.
Shen, H., and Laird, P.W. (2013). Interplay between the Cancer Genome and
Epigenome. Cell 153, 38–55.
Shigeyasu, K., Nagasaka, T., Mori, Y., Yokomichi, N., Kawai, T., Fuji, T., Kimura, K.,
Umeda, Y., Kagawa, S., Goel, A., et al. (2015). Clinical Significance of MLH1
Methylation and CpG Island Methylator Phenotype as Prognostic Markers in Patients
with Gastric Cancer. PLoS One 10, e0130409.
Shih, A.H., Abdel-Wahab, O., Patel, J.P., and Levine, R.L. (2012). The role of mutations
165
in epigenetic regulators in myeloid malignancies. Nat. Rev. Cancer 12, 599–612.
Shinya Matsuzaki*, Kiyoshi Yoshino, Yutaka Ueda, Satoko Matsuzaki, Mamoru Kakuda,
A.O., and Tomomi Egawa‑Takata, E.K. and T.K. (2015). Potential targets for ovarian
clear cell carcinoma: a review of updates and future perspectives. Cancer Cell Int. 15.
Silber, J.R., Bobola, M.S., Blank, A., and Chamberlain, M.C. (2012). O(6)-
methylguanine-DNA methyltransferase in glioma therapy: promise and problems.
Biochim. Biophys. Acta 1826, 71–82.
Skulte, K.A., Phan, L., Clark, S.J., and Taberlay, P.C. (2014). Chromatin remodeler
mutations in human cancers: epigenetic implications. Epigenomics 6, 397–414.
Sneppen, K., and Dodd, I.B. (2012). A simple histone code opens many paths to
epigenetics. PLoS Comput. Biol. 8, e1002643.
Søes, S., Daugaard, I.L., Sørensen, B.S., Carus, A., Mattheisen, M., Alsner, J.,
Overgaard, J., Hager, H., Hansen, L.L., and Kristensen, L.S. (2014). Hypomethylation
and increased expression of the putative oncogene ELMO3 are associated with lung
cancer development and metastases formation. Oncoscience 1, 367–374.
Sonnemann, J., Gruhn, B., Wittig, S., Becker, S., and Beck, J.F. (2012). Increased
activity of histone deacetylases in childhood acute lymphoblastic leukaemia and acute
myeloid leukaemia: support for histone deacetylase inhibitors as antileukaemic agents.
Br. J. Haematol. 158, 664–666.
Srivastava, P., Paluch, B.E., Matsuzaki, J., James, S.R., Collamat-Lai, G., Taverna, P.,
Karpf, A.R., and Griffiths, E.A. (2015). Immunomodulatory action of the DNA
166
methyltransferase inhibitor SGI-110 in epithelial ovarian cancer cells and xenografts.
Epigenetics 10, 237–246.
Strahl, B.D., and Allis, C.D. (2000). The language of covalent histone modi ® cations.
403, 41–45.
Stresemann, C., and Lyko, F. (2008). Modes of action of the DNA methyltransferase
inhibitors azacytidine and decitabine. Int. J. Cancer 123, 8–13.
Struhl, K. (1998). Histone acetylation and transcriptional regulatory mechanisms. Genes
Dev. 12, 599–606.
Struhl, K., and Segal, E. (2013). Determinants of nucleosome positioning. Nat. Struct.
Mol. Biol. 20, 267–273.
Studitsky, V.M., Clark, D.J., and Felsenfeld, G. (1995). Overcoming a nucleosomal
barrier to transcription. Cell 83, 19–27.
Su, S.-F., de Castro Abreu, A.L., Chihara, Y., Tsai, Y., Andreu-Vieyra, C., Daneshmand,
S., Skinner, E.C., Jones, P.A., Siegmund, K.D., and Liang, G. (2014). A panel of three
markers hyper- and hypomethylated in urine sediments accurately predicts bladder
cancer recurrence. Clin. Cancer Res. 20, 1978–1989.
Sun, H., Zhong, X., Wang, C., Wang, S., Lin, L., Zou, R., Wu, Y., Sun, N., Sun, G.,
Wen, T., et al. (2016a). SNF5 is Involved in Suppression of Hepatocellular Carcinoma
Progression via TGF-Beta 1 Signaling. Anat. Rec. (Hoboken). 299, 869–877.
Sun, X., Chuang, J.-C., Kanchwala, M., Wu, L., Celen, C., Li, L., Liang, H., Zhang, S.,
167
Maples, T., Nguyen, L.H., et al. (2016b). Suppression of the SWI/SNF Component
Arid1a Promotes Mammalian Regeneration. Cell Stem Cell 18, 456–466.
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.
Taberlay, P.C., Statham, A.L., Kelly, T.K., Clark, S.J., and Jones, P.A. (2014).
Reconfiguration of nucleosome-depleted regions at distal regulatory elements
accompanies DNA methylation of enhancers and insulators in cancer. Genome Res. 24,
1421–1432.
Taniguchi, T., Sullivan, M.J., Ogawa, O., and Reeve, A.E. (1995). Epigenetic changes
encompassing the IGF2/H19 locus associated with relaxation of IGF2 imprinting and
silencing of H19 in Wilms tumor. Proc. Natl. Acad. Sci. U. S. A. 92, 2159–2163.
Tashiro, H., Katabuchi, H., Begum, M., Li, X., Nitta, M., Ohtake, H., and Okamura, H.
(2003). Roles of luteinizing hormone/chorionic gonadotropin receptor in anchorage-
dependent and -independent growth in human ovarian surface epithelial cell lines.
Cancer Sci. 94, 953–959.
Teoh, J., Park, K., Wang, Y., Hu, Q., Kim, S., Wu, G., Huang, S., Maihle, N., and Kim, I.
(2014). Endothelin-1/Endothelin A receptor-mediated biased signaling is a new player in
modulating human ovarian cancer cell tumorigenesis. Cell. Signal. 26, 2885–2895.
Thompson, S., Pearson, A.N., Ashley, M.D., Jessick, V., Murphy, B.M., Gafken, P.,
Henshall, D.C., Morris, K.T., Simon, R.P., and Meller, R. (2011). Identification of a novel
168
Bcl-2-interacting mediator of cell death (Bim) E3 ligase, tripartite motif-containing protein
2 (TRIM2), and its role in rapid ischemic tolerance-induced neuroprotection. J. Biol.
Chem. 286, 19331–19339.
Tolstorukov, M.Y., Sansam, C.G., Lu, P., Koellhoffer, E.C., Helming, K.C., Alver, B.H.,
Tillman, E.J., Evans, J.A., Wilson, B.G., Park, P.J., et al. (2013). Swi/Snf chromatin
remodeling/tumor suppressor complex establishes nucleosome occupancy at target
promoters. Proc. Natl. Acad. Sci. U. S. A. 110, 10165–10170.
Tomczak, K., Czerwińska, P., and Wiznerowicz, M. (2015). The Cancer Genome Atlas
(TCGA): an immeasurable source of knowledge. Contemp. Oncol. (Poznań, Poland) 19,
A68–A77.
Toyota, M., Ahuja, N., Ohe-Toyota, M., Herman, J.G., Baylin, S.B., and Issa, J.P.
(1999). CpG island methylator phenotype in colorectal cancer. Proc. Natl. Acad. Sci. U.
S. A. 96, 8681–8686.
Treppendahl, M.B., Qiu, X., Søgaard, A., Yang, X., Nandrup-Bus, C., Hother, C.,
Andersen, M.K., Kjeldsen, L., Möllgård, L., Möllgaard, L., et al. (2012). Allelic
methylation levels of the noncoding VTRNA2-1 located on chromosome 5q31.1 predict
outcome in AML. Blood 119, 206–216.
Trouche, D., Le Chalony, C., Muchardt, C., Yaniv, M., and Kouzarides, T. (1997). RB
and hbrm cooperate to repress the activation functions of E2F1. Proc. Natl. Acad. Sci.
U. S. A. 94, 11268–11273.
Tsai, H.-C., Li, H., Van Neste, L., Cai, Y., Robert, C., Rassool, F. V, Shin, J.J., Harbom,
169
K.M., Beaty, R., Pappou, E., et al. (2012). Transient low doses of DNA-demethylating
agents exert durable antitumor effects on hematological and epithelial tumor cells.
Cancer Cell 21, 430–446.
Turcan, S., Rohle, D., Goenka, A., Walsh, L. a., Fang, F., Yilmaz, E., Campos, C.,
Fabius, A.W.M., Lu, C., Ward, P.S., et al. (2012). IDH1 mutation is sufficient to establish
the glioma hypermethylator phenotype. Nature 483, 479–483.
U.S. Cancer Statistics Working Group (2015). United States Cancer Statistics: 1999–
2012 Incidence and Mortality Web-based Report.
Ukaji, T., Lin, Y., Banno, K., Okada, S., and Umezawa, K. (2015). Inhibition of IGF-1-
Mediated Cellular Migration and Invasion by Migracin A in Ovarian Clear Cell
Carcinoma Cells. PLoS One 10, e0137663.
Vrba, L., Garbe, J.C., Stampfer, M.R., and Futscher, B.W. (2015). A lincRNA connected
to cell mortality and epigenetically-silenced in most common human cancers.
Epigenetics 10, 1074–1083.
Vries, R.G.J., Bezrookove, V., Zuijderduijn, L.M.P., Kia, S.K., Houweling, A.,
Oruetxebarria, I., Raap, A.K., and Verrijzer, C.P. (2005). Cancer-associated mutations
in chromatin remodeler hSNF5 promote chromosomal instability by compromising the
mitotic checkpoint. Genes Dev. 19, 665–670.
Wang, J., Chen, H., Fu, S., Xu, Z.-M., Sun, K.-L., and Fu, W.-N. (2011a). The
involvement of CHD5 hypermethylation in laryngeal squamous cell carcinoma. Oral
Oncol. 47, 601–608.
170
Wang, K., Kan, J., Yuen, S.T., Shi, S.T., Chu, K.M., Law, S., Chan, T.L., Kan, Z., Chan,
A.S.Y., Tsui, W.Y., et al. (2011b). Exome sequencing identifies frequent mutation of
ARID1A in molecular subtypes of gastric cancer. Nat. Genet. 43, 1219–1223.
Wang, X., Haswell, J.R., and Roberts, C.W.M. (2014). Molecular pathways: SWI/SNF
(BAF) complexes are frequently mutated in cancer--mechanisms and potential
therapeutic insights. Clin. Cancer Res. 20, 21–27.
Weber, M., Hellmann, I., Stadler, M.B., Ramos, L., Pääbo, S., Rebhan, M., and
Schübeler, D. (2007). Distribution, silencing potential and evolutionary impact of
promoter DNA methylation in the human genome. Nat. Genet. 39, 457–466.
Wei, D., Goldfarb, D., Song, S., Cannon, C., Yan, F., Sakellariou-Thompson, D.,
Emanuele, M., Major, M.B., Weissman, B.E., and Kuwahara, Y. (2014). SNF5/INI1
deficiency redefines chromatin remodeling complex composition during tumor
development. Mol. Cancer Res. 12, 1574–1585.
Weisenberger, D.J. (2014). Characterizing DNA methylation alterations from the cancer
genome atlas. J. Clin. Invest. 124, 17–23.
Weisenberger, D.J., and Liang, G. (2015). Contributions of DNA methylation
aberrancies in shaping the cancer epigenome. Transl. Cancer Res. 4, 219–234.
Weisenberger, D.J., Siegmund, K.D., Campan, M., Young, J., Long, T.I., Faasse, M.A.,
Kang, G.H., Widschwendter, M., Weener, D., Buchanan, D., et al. (2006). CpG island
methylator phenotype underlies sporadic microsatellite instability and is tightly
associated with BRAF mutation in colorectal cancer. Nat. Genet. 38, 787–793.
171
Widschwendter, M., Fiegl, H., Egle, D., Mueller-Holzner, E., Spizzo, G., Marth, C.,
Weisenberger, D.J., Campan, M., Young, J., Jacobs, I., et al. (2007). Epigenetic stem
cell signature in cancer. Nat. Genet. 39, 157–158.
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.
Wiegand, K.C., Lee, A.F., Al-Agha, O.M., Chow, C., Kalloger, S.E., Scott, D.W., Steidl,
C., Wiseman, S.M., Gascoyne, R.D., Gilks, B., et al. (2011). Loss of BAF250a (ARID1A)
is frequent in high-grade endometrial carcinomas. J. Pathol. 224, 328–333.
Wiegand, K.C., Hennessy, B.T., Leung, S., Wang, Y., Ju, Z., McGahren, M., Kalloger,
S.E., Finlayson, S., Stemke-Hale, K., Lu, Y., et al. (2014). A functional proteogenomic
analysis of endometrioid and clear cell carcinomas using reverse phase protein array
and mutation analysis: protein expression is histotype-specific and loss of
ARID1A/BAF250a is associated with AKT phosphorylation. BMC Cancer 14, 120.
Wöhrle, S., Weiss, A., Ito, M., Kauffmann, A., Murakami, M., Jagani, Z., Thuery, A.,
Bauer-Probst, B., Reimann, F., Stamm, C., et al. (2013). Fibroblast growth factor
receptors as novel therapeutic targets in SNF5-deleted malignant rhabdoid tumors.
PLoS One 8, e77652.
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
172
Genet. 6, e1000917.
Wu, R., and Li, H. (2010). Positioned and G/C-capped poly(dA:dT) tracts associate with
the centers of nucleosome-free regions in yeast promoters. Genome Res. 20, 473–484.
Wu, M.S., Wang, H.P., Lin, C.C., Sheu, J.C., Shun, C.T., Lee, W.J., and Lin, J.T.
(1997). Loss of imprinting and overexpression of IGF2 gene in gastric adenocarcinoma.
Cancer Lett. 120, 9–14.
Xiao, W., Awadallah, A., and Xin, W. (2012). Loss of ARID1A/BAF250a expression in
ovarian endometriosis and clear cell carcinoma. Int. J. Clin. Exp. Pathol. 5, 642–650.
Yamamoto, S., Tsuda, H., Takano, M., Tamai, S., and Matsubara, O. (2012). PIK3CA
mutations and loss of ARID1A protein expression are early events in the development
of cystic ovarian clear cell adenocarcinoma. Virchows Arch. 460, 77–87.
Yamazaki, J., and Issa, J.-P.J. (2013). Epigenetic aspects of MDS and its molecular
targeted therapy. Int. J. Hematol. 97, 175–182.
Yang, X., Han, H., De Carvalho, D.D., Lay, F.D., Jones, P.A., and Liang, G. (2014).
Gene Body Methylation Can Alter Gene Expression and Is a Therapeutic Target in
Cancer. Cancer Cell 26, 577–590.
Yaniv, M. (2014). Chromatin remodeling: from transcription to cancer. Cancer Genet.
207, 352–357.
Yao, L., Shen, H., Laird, P.W., Farnham, P.J., and Berman, B.P. (2015a). Inferring
regulatory element landscapes and transcription factor networks from cancer
173
methylomes. Genome Biol. 16, 105.
Yao, L., Shen, H., Laird, P.W., Farnham, P.J., and Berman, B.P. (2015b). Inferring
regulatory element landscapes and transcription factor networks from cancer
methylomes. Genome Biol. 16, 105.
Yegnasubramanian, S., Haffner, M.C., Zhang, Y., Gurel, B., Cornish, T.C., Wu, Z.,
Irizarry, R.A., Morgan, J., Hicks, J., DeWeese, T.L., et al. (2008). DNA hypomethylation
arises later in prostate cancer progression than CpG island hypermethylation and
contributes to metastatic tumor heterogeneity. Cancer Res. 68, 8954–8967.
Yen, K., Vinayachandran, V., Batta, K., Koerber, R.T., and Pugh, B.F. (2012). Genome-
wide nucleosome specificity and directionality of chromatin remodelers. Cell 149, 1461–
1473.
Yoder, J.A., Walsh, C.P., and Bestor, T.H. (1997). Cytosine methylation and the ecology
of intragenomic parasites. Trends Genet. 13, 335–340.
Yoo, C.B., Jeong, S., Egger, G., Liang, G., Phiasivongsa, P., Tang, C., Redkar, S., and
Jones, P.A. (2007). Delivery of 5-aza-2’-deoxycytidine to cells using
oligodeoxynucleotides. Cancer Res. 67, 6400–6408.
You, J.S., and Jones, P.A. (2012). Cancer Genetics and Epigenetics: Two Sides of the
Same Coin? Cancer Cell 22, 9–20.
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.
174
A. 108, 14497–14502.
Zarnett, O.J., Sahgal, A., Gosio, J., Perry, J., Berger, M.S., Chang, S., and Das, S.
(2015). Treatment of elderly patients with glioblastoma: a systematic evidence-based
analysis. JAMA Neurol. 72, 589–596.
Zhang, C., Zhong, J.F., Stucky, A., Chen, X.-L., Press, M.F., and Zhang, X. (2015a).
Histone acetylation: novel target for the treatment of acute lymphoblastic leukemia. Clin.
Epigenetics 7, 117.
Zhang, S., Lu, Z., Mao, W., Ahmed, A.A., Yang, H., Zhou, J., Jennings, N., Rodriguez-
Aguayo, C., Lopez-Berestein, G., Miranda, R., et al. (2015b). CDK5 Regulates
Paclitaxel Sensitivity in Ovarian Cancer Cells by Modulating AKT Activation, p21Cip1-
and p27Kip1-Mediated G1 Cell Cycle Arrest and Apoptosis. PLoS One 10, e0131833.
Zhang, Y., Liu, T., Meyer, C.A., Eeckhoute, J., Johnson, D.S., Bernstein, B.E.,
Nusbaum, C., Myers, R.M., Brown, M., Li, W., et al. (2008). Model-based analysis of
ChIP-Seq (MACS). Genome Biol. 9, R137.
(2012). Comprehensive molecular characterization of human colon and rectal cancer.
Nature 487, 330–337.
Abstract (if available)
Abstract
The complex interplay between epigenetic mechanisms such as DNA methylation, histone modification, and nucleosome positioning regulate gene expression potential and chromatin organization over cell generations. Mutations in mediators of epigenetic processes can contribute to considerable alterations of gene expression and disrupt cellular identity and function, resulting in the development of malignancies. Large-scale sequencing efforts have uncovered germline and somatic mutations in many epigenetic modulators including DNA methyltransferases, histone tail modifiers, and subunits of chromatin remodeling complexes. Specifically, the components of the SWI/SNF remodeling complexes are mutated in upwards of 20% of all tumors. However, our understanding of the molecular changes that take place due to these mutations is still incomplete. This dissertation aims to examine the role of SWI/SNF complex proteins in shaping the epigenetic landscape of cancer and how the interplay between the various epigenetic mechanisms contribute to functional alterations. ❧ First, I demonstrated that induction of SNF5 in malignant rhabdoid tumor (MRT) cell line TCC642 significantly altered the transcriptome and restores aberrant expression of several genes closer to normal levels. Further, I surveyed the epigenome using the DNA methylation array-based AcceSssIble assay and uncovered several potential direct epigenetic targets of SNF5. Finally, I established TRIM2 as a downstream target of SNF5. Thus, I found that re-expression of SNF5 in TCC642 cells regulated target gene expression in part by remodeling chromatin. ❧ Next, using an endometriosis cell line model, I comprehensively characterized gene expression, histone modification, chromatin accessibility, and DNA methylation changes that take place due to loss of ARID1A. Specifically, I showed that the loss of ARID1A triggered a considerable re-distribution of the H3K27ac histone modification and affected the expression of a number of genes critical to ovarian cancer development. These molecular aberrations contributed to phenotypic changes including increased anchorage-independent colony formation and increased invasion of basement membrane extract in vitro. ❧ Thus, this dissertation sheds light on how mutations in a single component of an epigenetic complex can not only alter the epigenome but can also promote functional changes and possibly contribute to oncogenesis.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Effects of chromatin regulators during carcinogenesis
PDF
Molecular mechanism of the recruitment of SWI/SNF chromatin remodeling complex and histone acetyltransferase to estrogen-responsive promoters
PDF
Functional DNA methylation changes in normal and cancer cells
PDF
Understanding DNA methylation and nucleosome organization in cancer cells using single molecule sequencing
PDF
Identification and characterization of cancer-associated enhancers
PDF
Identification of novel epigenetic biomarkers and microRNAs for cancer therapeutics
PDF
Epigenetic plasticity of cultured female human embryonic stem cells and regulation of gene expression and chromatin by PR-SET7 mediated H4K20me1
PDF
The relationship between DNA methylation and transcription factor binding in colon cancer cells
PDF
DNA methylation inhibitors and epigenetic regulation of microRNA expression
PDF
Breast epithelial cell type specific enhancers and functional annotation of breast cancer risk loci
PDF
Using epigenetic toggle switches to repress tumor-promoting gene expression
PDF
Identification and characterization of PR-Set7 and histone H4 lysine 20 methylation-associated proteins
PDF
Epigenetic mechanisms driving bladder cancer
PDF
Functional characterization of colon cancer risk-associated enhancers: connecting risk loci to risk genes
PDF
Transcriptional regulation by epigenetic mechanisms
PDF
The role of Hic-5 in glucocorticoid receptor binding to chromatin
PDF
Multiple functions of the PR-Set7 histone methyltransferase: from transcription to the cell cycle
PDF
Integrated genomic & epigenomic analyses of glioblastoma multiforme: Methods development and application
PDF
Genetics and the environment: evaluating the role of noncoding RNA in autism spectrum disorder
PDF
DNA methylation changes in the development of lung adenocarcinoma
Asset Metadata
Creator
Lakshminarasimhan, Ranjani
(author)
Core Title
Functional role of chromatin remodeler proteins in cancer biology
School
Keck School of Medicine
Degree
Doctor of Philosophy
Degree Program
Genetic, Molecular and Cellular Biology
Publication Date
11/17/2016
Defense Date
09/12/2016
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Arid1A,chromatin remodeler,epigenetics,OAI-PMH Harvest,SNF5,SWI/SNF
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Farnham, Peggy (
committee chair
), Coetzee, Gerhard (
committee member
), Jones, Peter A. (
committee member
), Liang, Gangning (
committee member
)
Creator Email
ranjanil@usc.edu,rlaks86@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-321589
Unique identifier
UC11214221
Identifier
etd-Lakshminar-4928.pdf (filename),usctheses-c40-321589 (legacy record id)
Legacy Identifier
etd-Lakshminar-4928.pdf
Dmrecord
321589
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Lakshminarasimhan, Ranjani
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
Tags
Arid1A
chromatin remodeler
epigenetics
SNF5
SWI/SNF