Enhancers and super-enhancers in human disease and therapy
By
Heather Ashley Hoke
B.A. Integrated Science, Genetics and Molecular Biology
Northwestern University, 2009
Submitted to the Department of Biology in Partial Fulfillment
of the Requirements for the Degree of
fMASSACHUSETS INSTI1 WE
OF TECHNOLOGY
Doctor of Philosophy in Biology
SMAY28
at the
LIBRARIES
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
June 2014
©2014 Massachusetts Institute of Technology. All rights reserved.
Signature of Author:
Signature redacted
Department of Biology
Certified by:
Signature redacted
May22,2014
Richard A. Young
Professor of Biology
Accepted by:
2014
Signature redacted
Thesis Supervisor
Michael T. Hemann
Associate Professor of Biology
Chairperson, Biology Graduate Committee
Enhancers and super-enhancers in human disease and therapy
By
Heather Ashley Hoke
Submitted to the Department of Biology on May 22, 2014 in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy in Biology
Abstract
The human body is made up of a diverse array of cell types, each with specialized properties and
functions that support the organism as a whole. Despite this variability, with few exceptions,
these cells contain the same genetic information. The incredible diversity in cellular function
arises from differences in gene expression between cell types. Regulation of gene expression
involves complex interactions between transcription factors and cofactors and the transcriptional
machinery. These interactions occur both at the core promoter and at distal regulatory regions,
enhancers, which are looped into close physical proximity with the core promoter. Enhancer
regions are typically utilized in a cell-type-specific manner and help to drive the distinct gene
expression programs that define diverse cell identities. New technologies have allowed greater
ability to map enhancer regions in a variety of cell types in both healthy and diseased tissue. It is
becoming increasingly apparent that misregulation of enhancer regions is a major component of
many human diseases, including cancer. In addition, the availability of new classes of small
molecule inhibitors targeting enhancer-bound transcriptional cofactors suggests that disruption of
enhancer function may be an important therapeutic strategy in disease. Identification of superenhancers, clusters of enhancer regions that are occupied by exceptional levels of many
transcriptional coactivators, may further inform our understanding of human development and
disease. These super-enhancers are associated with genes that control and define cell state,
including many important disease-associated factors. In addition, super-enhancers are
particularly vulnerable to perturbation by small molecules targeting enhancer-associated factors.
This suggests that super-enhancers may be both biomarkers for disease-critical genes, and
Achilles heels, allowing these genes to be targeted by small molecule therapies.
Thesis supervisor: Dr. Richard A. Young
Title: Professor
Acknowledgements
First and foremost, I'd like to thank my advisor Rick Young for always challenging me to pursue
the most exciting and interesting projects. His confidence in me has motivated me to push
myself and become the scientist that I am today.
It has been my great pleasure to work with the many talented scientists in the Young lab. In
particular, I'd like to thank Tony Lee; I couldn't have asked for a better bay-mate, and will be
forever thankful for my bench assignment. His help and advice has been invaluable. I'd also
like to thank Jakob Loven and Charles Lin, who I worked closely with for the main part of my
thesis research and have been fantastic colleagues and mentors.
Finally, I'd like to thank my family and friends for their love and support throughout the past
five years. My parents have been a huge source of support during my graduate career, and I am
so thankful that they were always there for me, both to help me through tough times, and to
celebrate happy moments. I am especially proud of my mother, who as a non-scientist, took it
upon herself to learn basic biology so that she could better understand my work. Last, I'd like to
thank my fiance Sean for his immeasurable support, love, and patience as I finished my degree.
3
Table of Contents
Title page.....................................................................................................................................1
Abstract .......................................................................................................................................
2
Acknowledgem ents ............................................................................................................
3
Chapter 1: Introduction.........................................................................................................
5
Chapter 2: Selective inhibition of tumor oncogenes by disruption of super-enhancers .....
62
Chapter 3: Conclusions and future directions ............................................................................
111
Appendix A : Supplem entary m aterial for Chapter 2..................................................................124
Appendix B: Enhancer decommissioning by LSD 1 during embryonic stem cell
differentiation...............................................................................................................................157
4
Chapter 1
Introduction
Preface
Transcriptional enhancer regions play a critical role in driving the cell-type-specific gene
expression programs that underlie the incredible diversity in cell function within the human
body. These regions are often misregulated in disease, whether through the influence of human
sequence variants, somatic mutations, or misregulation of the transcription factors and
coactivators that act through enhancer regions. In chapter one of this thesis, I describe the basic
stages of the transcriptional cycle, and the transcription factors, cofactors and chromatin
regulators that control transcription through their actions at core promoters and enhancers. I then
discuss the basic properties of enhancer regions and the identification of super-enhancers, large
clusters of transcriptional enhancers that are associated with key cell-type-specific genes. Next, I
describe the involvement of enhancer and super-enhancer regions in human disease, and in
particular, cancer. Lastly, I discuss the recent development of many inhibitors targeting
enhancer-associated transcriptional regulators. Chapter two of this thesis describes the
association of super-enhancers with key oncogenes in cancer, and a mechanism by which
inhibition of a general transcriptional cofactor, BRD4, can lead to a selective effect on gene
expression through the disruption of super-enhancer regions. In chapter three, I offer concluding
thoughts on the use of super-enhancers as biomarkers, identifying key cell-state-defining or
disease-related genes, and the generality of super-enhancer disruption by transcriptional
inhibitors as a therapeutic strategy in human disease.
5
Transcription regulation
The transcriptioncycle
The transcription cycle progresses through several stages, each of which is subject to
regulation (Reviewed in (Berger, 2007; Kouzarides, 2007; Lee and Young, 2000; Li et al., 2007;
Saunders et al., 2006; Shandilya and Roberts, 2012)). Transcription begins by the recruitment of
factors to the core promoter, including the RNA Polymerase II holoenzyme (RNA Pol II),
general transcription factors (GTFs), and other cofactors such as the Mediator complex. This
group of factors is known as the pre-initiation complex (PIC). Following the formation of the
PIC at the transcription start site (TSS), transcription generally proceeds for about 25-60 bases
before becoming stalled. Release of polymerase from the paused state requires the recruitment
of additional factors, leading to the phosphorylation of several pause-factors as well as Pol II
itself. After pause release Pol II continues into productive elongation through the gene body.
RNA processing occurs concomitantly with transcriptional elongation, and many protein factors
are involved in facilitating RNA capping, splicing of introns, and finally cleavage and
polyadenylation at the 3' end of the transcript. Transcription usually continues several kilobases
after the polyadenylation signal before terminating, and releasing the Pol II from DNA.
Gene expression is influenced by both the core promoter region of a gene, and distal
enhancer regions. The core promoter is the region immediately flanking the transcription start
site of a gene, while enhancer elements can be located many megabases away from the gene that
they control. Both core promoters and enhancers contain binding sites for sequence specific
transcription factors that recruit transcriptional cofactors and the transcription machinery. The
6
many interactions between promoter- and enhancer- bound proteins also mediate looping
between the core promoter and enhancer, bringing the regions into close physical proximity,
despite their linear distance in the genome. Transcription factors at enhancers and promoters
influence transcription through the recruitment of both activating and repressive cofactors that
can influence several stages of the transcriptional cycle (Bulger and Groudine, 2011; Ong and
Corces, 2011; Shlyueva et al., 2014; Spitz and Furlong, 2012). The following section will
provide a brief description of the initiation and elongation stages of the transcriptional cycle, and
how these are regulated by the recruitment of key factors to core promoter and enhancer regions.
Initiation
The transcription cycle begins with the assembly of the PIC, which includes RNA Pol II,
associated general transcription factors (GTFs), and the Mediator complex (Lee and Young,
2000). Mediator is a large multisubunit complex that acts as a hub, bringing transcription
factors, transcriptional coactivators, and signaling pathway components into proximity with the
RNA Pol II holoenzyme (Borggrefe and Yue, 2011; Conaway and Conaway, 2011; Komberg,
2005; Malik and Roeder, 2010; Taatjes, 2010). The many subunits comprising the Mediator
complex allow for interactions with a wide variety of sequence-specific transcription factors.
Mediator can also interact with a variety of coactivator molecules, including GTFs. In addition,
Mediator binds directly to the c-terminal domain (CTD) of Pol II, providing a link between these
transcription factors, coactivators, and the RNA Pol II holoenzyme, contributing to the formation
of the PIC. Mediator also facilitates looping between enhancer regions and the core promoter
through its interaction with Cohesin, a protein complex also involved in the cohesion of sister
chromatids (Kagey et al., 2010). In addition, Mediator interacts with additional cofactors that
can influence later states of the transcriptional cycle, such as pause release.
7
Elongation
During transcriptional activation, the progression of RNA Pol II is typically stalled
shortly after initiation. This pausing of Pol II just downstream of the TSS appears to be a major
regulatory feature in gene expression, with a majority of mammalian genes occupied by paused
polymerases (Core and Lis, 2008; Core et al., 2008; Guenther et al., 2007; Rahl et al., 2010).
This pausing is mediated by several factors, including NELF (Negative elongation factor) and
DSIF (DRB-sensitivity-inducing factor) (Cheng and Price, 2007; Yamaguchi et al., 2013). The
transition from paused to elongating Pol II requires the action of the positive transcription
elongation factor, P-TEFb. P-TEFb consists of the cyclin-dependent kinase Cdk9, and its
partner, cyclin T 1, and exists in two forms, an active complex, and an inactive one bound by the
inhibitory factors HEXIM 1 and the 7SK snRNP. Active P-TEFb promotes elongation through
phosphorylation of the Pol II C-terminal domain (CTD) at serine 2, as well as phosphorylation of
the negative transcription elongation factors, DSIF and NELF (Peterlin and Price, 2006).
BRD4, a transcriptional coactivator that associates closely with the Mediator complex, is
involved in Pol II pause release through its interaction with P-TEFb (Dawson et al., 2011; Dey et
al., 2000; Jang et al., 2005; Jiang et al., 1998; Wu and Chiang, 2007; Yang et al., 2005).
Through its two bromodomains, BRD4 can bind to acetylated lysine residues on histones, as well
as other acetylated proteins, including transcription factors (Huang et al., 2009; LeRoy et al.,
2008; Rahman et al., 2011). In addition to these bromodomain-mediated interactions, BRD4 can
associate with P-TEFb through a separate C-terminal domain (Jang et al., 2005; Yang et al.,
2005). This binding event leads directly to the dissociation of the inhibitory factors HEXIM1
and the 7SK snRNP from P-TEFb, rendering it active and able to promote pause release
(Krueger et al., 2010). BRD4 is closely associated with Mediator, and was, in fact, first
8
identified as an interaction partner of the Mediator complex in murine cells (Jiang et al., 1998).
This interaction has been further confirmed in human cells, where Mediator and BRD4 cooccupy both promoter and enhancer regions, together facilitating the recruitment of P-TEFb and
release of paused RNA Pol II into productive elongation (Dawson et al., 2011; Loven et al.,
2013; Wu and Chiang, 2007).
Additional factors involved in elongation control form a large complex, the super
elongation complex (SEC), which has also been implicated in the recruitment of P-TEFb to
active genes (Luo et al., 2012; Smith et al., 2011). The SEC is comprised of several elongation
factors, including eleven-nineteen lysine-rich leukemia (ELL) proteins, mixed lineage leukemia
translocation partners (MLLs), and the P-TEFb complex (Luo et al., 2012). The SEC component
ELL3 has been shown to occupy primarily enhancer regions in mouse embryonic stem cells
(mESCs), where it may play a role both in facilitating interactions between the enhancer and
promoter regions, and in recruiting other SEC components to promote gene activation in later
lineages (Lin et al., 2013). Lastly, some transcription factors appear to be involved in elongation
control. For example, the MYC transcription factor and proto-oncogene interacts directly with PTEFb and appears to be involved in mediating transcriptional pause release in a genome-wide
manner (Eberhardy and Farnham, 2001, 2002; Gargano et al., 2007; Kanazawa et al., 2003; Rahl
et al., 2010). Although MYC is primarily associated with promoter regions, overexpression of
MYC can lead to "enhancer invasion", wherein high levels of MYC lead its association with
lower-affinity binding sites at both enhancer and core promoter regions (Lin et al., 2012).
Divergent transcription
One emerging property of transcription is that it occurs in a bidirectional, divergent
manner at most mammalian genes (Core et al., 2008; Preker et al., 2008; Seila et al., 2008;
9
Sigova et al., 2013; Wu and Sharp, 2013). In addition to transcription in the sense, often proteincoding direction, a majority of genes also produce relatively short (50-2,000 nucleotides),
unstable antisense transcripts upstream of the promoter (Flynn et al., 2011; Ntini et al., 2013;
Preker et al., 2011; Preker et al., 2008). Antisense transcription shares many features with sense
transcription, including recruitment of the elongation factor P-TEFb (Flynn et al., 2011). The
factors that control promoter directionality are not well understood, but several mechanisms may
be at play. First, a minority of human promoter regions contain a TATA element, a motif bound
by the TATA-binding protein TBP, which is part of the TFIID GTF involved in the recruitment
of RNA Pol II. The sequence-specific binding of TBP can help to direct more precise
recruitment of RNA Pol II to the core promoter region, while at TATA-less genes (the majority
of human genes), less-specific protein-protein interactions with transcription factors recruit TBP
both upstream and downstream of these transcription factors binding sites, contributing to
bidirectional transcription (Core and Lis, 2008; Core et al., 2008; Wu and Sharp, 2013). Other
factors that may influence transcription directionality include chromatin remodeling, histone
deacetylation, and gene-loop formation between the promoter and 3' region of the gene
(Churchman and Weissman, 2011; Tan-Wong et al., 2012; Whitehouse et al., 2007). Control
may also occur by recognition of the antisense RNA species by the machinery that typically
controls the addition of poly A tracts. The poly A signal motif (PAS) or similar sequences are
enriched in the upstream antisense region, and recognition of this sequence can lead to cleavage
of the RNA and termination of transcription (Almada et al., 2013; Ntini et al., 2013; Proudfoot,
2011). Divergent transcription may help to increase expression of the sense transcript, for
example, by increasing negative supercoiling and DNA unwinding at promoters. Upstream
transcription can also lead to an increased propensity for mutagenesis, potentially contributing to
10
the evolution of new genes (Wu and Sharp, 2013). This divergent transcription also occurs at
enhancer regions, and may play an important role in enhancer function (discussed later) as well
as the evolution of new genes from enhancer regions (De Santa et al., 2010; Kim et al., 2010;
Sigova et al., 2013; Wu and Sharp, 2013).
Control of transcriptionby chromatin regulators
In addition to the mechanisms of regulation described above, transcription can also be
influenced by the chromatin state surrounding enhancers and promoters, and by the chromatin
modifying enzymes that act on histone proteins. Genomic DNA does not exist as a naked
molecule, but instead exists as chromatin, consisting of both DNA and closely associated
proteins, histones. Chromatin is comprised of units called nucleosomes: complexes containing
roughly 150 basepairs of DNA wrapped around positively charged histone proteins (Kornberg
and Lorch, 1999).
Histones can influence transcription both by their position, where they may create a
physical barrier to RNA Pol II initiation or elongation, and by posttranslational modifications
that can influence the recruitment of many transcriptional cofactors. To facilitate access of RNA
Pol II to histone-occupied DNA, transcription factors can recruit ATP-dependent chromatin
remodeling complexes, such as the SWI/SNF complex. Nucleosome remodeling complexes can
facilitate, or hinder, transcription, either by sliding histones along the DNA, ejecting them, or
mediating a switch between different histone variants (Hargreaves and Crabtree, 2011; Ho and
Crabtree, 2010).
Histone proteins are the targets of many posttranslational modifications, many of which
occur on the histone tail, 20-40 amino acid segments that extend beyond the main mass of the
nucleosome. Histone modifying enzymes can perform a wide variety of post-translational
11
modifications, including acetylation, methylation, phosphorylation, and ubiquitination, that are
involved in both gene activation and repression. In addition to the "writers" that deposit these
modifications, "erasers" such as histone deacetylases or demethylases can remove these marks.
"Reader" proteins can bind to specific histone marks, resulting in the recruitment of various
regulators of gene expression. Histone modifications are associated with different gene activity
states, depending both on the type of modification and the position of the modification on the
histone.
Lysine methylation can be associated with both activation and repression of transcription,
depending on the nature of the modification. Methylation of histone H3 lysine 4 (H3K4) is
generally considered to be a mark of active transcription, with trimethylation (H3K4me3)
occurring at promoter regions, while monomethylation (H3K4mel) occurs primarily at enhancer
regions. Interestingly, H3K4me 1 is present at both active enhancers as well as developmental
enhancers that appear to be primed for activation in later developmental stages (Creyghton et al.,
2010; Rada-Iglesias et al., 2011; Zentner et al., 2011). Several histone modifications are
associated with transcriptional elongation: H3K79me2, which occurs primarily in the 5' region
of elongating genes, and H3K36me3, which occurs more 3' in the gene body (Bernstein et al.,
2012).
Several histone methylation events are associated with inactive or repressed genes, with
different methylation patterns characteristic of different mechanisms of repression (Beisel and
Paro, 2011; Cedar and Bergman, 2012; Jones, 2012; Moazed, 2009; Reyes-Turcu and Grewal,
2012). Trimethylation of histone H3 lysine 27 (H3K27me3) is catalyzed by the EZH2 subunit of
the Polycomb complex. Genes marked by H3K27me3 are often silent, but poised for activation
in later developmental stages. Many of these Polycomb-repressed genes are also marked by
12
H3K4me3, and are referred to as "bivalent" genes, as they have marks associated with both gene
activity and repression (Orkin and Hochedlinger, 2011; Young, 2011). Genomic regions that are
more permanently repressed are often associated with trimethylation of H3K9 (H3K9me3),
which can be deposited by several histone methyltranferases, including SETDB 1, SUV39H 1,
SUV39H2, EHMT1 (GLP), and EHMT2 (G9A). These regions tend to be gene-poor and contain
repetitive elements or retrotransposons (Feng et al., 2010; Lejeune and Allshire, 2011;
Mikkelson et al., 2007).
Histone acetylation can occur at many different lysine residues in the histone tail and core
and is generally associated with activation of gene expression. These marks are deposited by
histone acetyltransferases (HATs) such as p300, GCN5, or TIP60, and removed by histone
deacetylases (HDACs). Acetylation of histone H3 lysine 27 (H3K27ac) is associated with active
enhancer regions (Bonn et al., 2012; Creyghton et al., 2010; Rada-Iglesias et al., 2011; Zentner et
al., 2011). In contrast to H3K4mel, which also occurs at inactive, primed enhancers, H3K27ac
appears to be exclusively a mark of active enhancers, making a distinction between these classes
of enhancer regions. In keeping with this model, during development, H3K27ac is acquired
almost exclusively at regions that are pre-marked by H3K4meI (Bonn et al., 2012). Unlike
lysine methylation, lysine acetylation removes the positive charge associated with the lysine
residue, potentially affecting the association of the histone protein with negatively charged DNA
in addition to creating a binding site for "reader" proteins.
Interestingly, many active genes are occupied by chromatin modifying enzymes with
opposing activities. For example, although histone deacetylation has long been associated with
repression of transcription, HDACs tend to be associated with actively transcribed genes
(Kurdistani et al., 2002; Wang et al., 2002; Wang et al., 2009). HDACs 1 and 2 exist in several
13
multisubunit complexes, including the NuRD complex, which also contains the histone
demethylase LSD 1, which is responsible for removing the active enhancer mark H3K4me 1.
Work from our lab (included as Appendix B) indicates that LSD 1 and HDACs 1 and 2 cooperate
at active enhancers to remove H3K4me 1 and histone acetyl marks upon differentiation, allowing
for deactivation of pluripotency genes (Whyte et al., 2012). LSD 1's demethylase activity is
hindered by the presence of acetylated histones (Forneris et al., 2005; Lee et al., 2006). We
therefore proposed a mechanism in which dynamic acetylation and deacetylation by HATs and
HDACs results in the maintenance of a level of acetylation that is inhibitory to LSD 1
demethylation. Upon differentiation, when HATs are no longer recruited to these enhancers, the
presence of HDACs results in deacetylation, followed by demethylation of the enhancers by
LSD1.
In addition to the writers and erasers that deposit and remove histone modifications,
"reader" proteins can recognize histone marks and lead to the recruitment of transcriptional
regulators to specific chromatin regions. As mentioned previously, BRD4 is an example of the
bromodomain class of acetyl-lysine binding proteins, and is involved in the recruitment of PTEFb to active enhancers and promoters marked by acetylated histones. Methylated histone
binding domains include the chromodomain, the WD40 repeat, and the PHD finger, which is
associated with methyl-lysine residues, and the Tudor domain, which can bind both methylated
lysines and arginine (Musselman et al., 2012). One example of a chromodomain protein is
heterochromatin protein 1 (HP 1), which recognizes H3K9me3 and facilitates chromatin
compaction at these repressed regions (Bannister et al., 2001; Lachner et al., 2001; Nakayama
and Takami, 2001). At enhancers, the TIP60 acetyltransferase is recruited in part through its
recognition of H3K4me I through its chromodomain (Jeong et al., 2011).
14
Conclusion
Cells can achieve specific gene expression programs by controlling transcription through
the influence of transcription factors, coactivators, and chromatin regulators. Many of these
transcriptional regulators are associated with distal enhancer regions which are looped into close
physical proximity with the core promoter region. The following sections describe the functional
properties of transcriptional enhancers in greater detail, and discuss how recent improvements in
our ability to map these enhancer regions have resulted in a greater understanding of enhancer
function in human development and disease.
Transcriptional enhancers
Although transcription requires only a core promoter, enhancers can greatly increase the
level of transcription from target genes, and, through binding of cell type-specific transcription
factors, can influence the timing and spatial distribution of gene expression within a cell or
organism. Early studies established these basic properties of enhancer function, and new
genome-wide technologies have extended our knowledge by allowing the mapping of enhancer
regions across many cell and tissue types. Estimates indicate that there may be over a million
enhancer regions encoded in the human genome (Bernstein et al., 2012). Investigating how these
enhancer regions mediate cell-type specific gene expression programs is critical to understanding
human development as well as disease. My thesis research led to the identification of a class of
enhancers, super-enhancers that consist of large clusters of individual enhancer elements. Superenhancers are associated with genes that control and define cell state, and thus may act as
biomarkers for identifying these key genes.
Historicalidentificationof enhancer elements
15
The first enhancer element identified was a 72 base pair sequence from the SV40 viral
genome. This viral sequence was capable of inducing a two hundred fold increase in the
expression of a reporter gene in human HeLa cells (Banerji et al., 1981). The enhancer region
retained its function when placed at different distances and orientations to the reporter gene; it
was able to enhance expression to a similar degree whether located proximal to the promoter or
several kilobases upstream or downstream of the gene (Banerji et al., 1981). Enhancers from
additional animal viruses were soon identified and described (de Villiers et al., 1982; Hansen and
Sharp, 1983; Schirm et al., 1985; Spandidos and Wilkie, 1983). Some of these enhancer
elements showed host-specificity, for example, an enhancer from the mouse polyoma virus was
most effective in mouse tissue, rather than primate tissue (de Villiers et al., 1982; Spandidos and
Wilkie, 1983).
Subsequently, enhancer elements were identified in metazoan genomes (Banerji et al.,
1983). One of the first endogenous mammalian enhancers described was a sequence located in
the immunoglobulin heavy chain locus (IGH) that acts as a transcriptional enhancer in reporter
assays. The enhancer activity of the IGH enhancer regions was found to be cell type specific; the
enhancer functioned in myeloma cells (a B-lymphocyte-derived tumor), but not HeLa cells
(derived from a cervical carcinoma) (Banerji et al., 1983; Davidson et al., 1986). In another
early example; several enhancer regions were found to control expression of the P-globin gene in
a cell-type, and developmental stage specific manner, first in chicken, and later in human cells
(Antoniou et al., 1988; Behringer et al., 1987; Choi and Engel, 1986; Hesse et al., 1986; Kollias
et al., 1987; Trudel and Costantini, 1987).
Further studies, using various techniques both in vitro and in vivo, indicated that
enhancer function depended on the binding of trans-acting protein factors (Augereau and
16
Chambon, 1986; Davidson et al., 1986; Lee et al., 1987; Mercola et al., 1985; Sassone-Corsi et
al., 1984; Scholer and Gruss, 1985; Sen and Baltimore, 1986; Sergeant et al., 1984; Wildeman et
al., 1984). These studies also suggested that these trans-factors bound to specific sequences in
the enhancer element, and that some of these factors were only present in specific cell types,
perhaps accounting for the cell-type-specificity of certain enhancers, such as the IGH or P-globin
enhancers (Lee et al., 1987; Mercola et al., 1985; Scholer and Gruss, 1985; Sen and Baltimore,
1986; Wildeman et al., 1984). For example, DNase footprinting experiments, in which
transcription factor binding sites are identified based on their protection from DNase I activity,
identified four erythroid-specific binding sites in an enhancer in the 3' region of the P-globin
gene (Wall et al., 1988). Gel-mobility shift assays revealed that these sites were bound by an
erythroid-specfic protein, Nuclear Factor Erythroid 1 (NF-E 1), later identified and characterized
as GATA- 1, a transcription factor that is essential for erythroid development (Ferreira et al.,
2005; Wall et al., 1988).
Transcriptionfactor binding
This binding of sequence-specific transcription factors is a key feature of enhancer
regions (reviewed in (Levine, 2010)). As described above, the binding of transcription factors to
enhancer regions and subsequent recruitment of the RNA Pol II transcription apparatus and
cofactors is a critical step in the activation of gene expression. Enhancers typically contain
binding motifs for many distinct classes of sequence-specific transcription factors. Proteinprotein interactions between these factors can act in a cooperative manner, increasing their
apparent affinity for the enhancer element. Binding of one transcription factor may also
indirectly facilitate the binding of another, by altering DNA conformation, or by recruiting
cofactors, for example a nucleosome remodeling complex, that mediates the binding of a second
17
factor. In addition, these transcription factors can act synergistically to recruit transcriptional
coactivators or RNA Pol II. These synergistic interactions can result in greater transcriptional
activation than could be achieved by independently acting factors (Carey et al., 1990; Giese et
al., 1995; Kim and Maniatis, 1997; Thanos and Maniatis, 1995). In addition, the convergence of
many transcription factors to an enhancer region contributes to the control of the precise
temporal and spatial regulation of gene expression, by requiring a specific combination of
transcription factors for gene activation (Ptashne and Gann, 1998). One particularly well-studied
enhancer is the J-interferon "enhanceosome", which is activated by a combination of several
transcription factors, c-Jun, ATF2, IRF3, IRF7, and NF-KB (Agalioti et al., 2000; Merika and
Thanos, 2001; Thanos and Maniatis, 1995). Crystal structures of this complex reveal the many
protein-DNA and protein-protein interactions that contribute to the function of this enhancer
(Panne et al., 2007).
Enhancer promoter looping
Communication between distal enhancer regions and the core promoter through looping
is another critical feature of enhancer function. Looping interactions between different DNA
regions were first observed in bacterial systems (Adhya, 1989; Bulger and Groudine, 1999;
Matthews, 1992; Ptashne, 1986; Saiz and Vilar, 2006; Schleif, 1992). For example, nitrogen
regulatory protein C (NtrC) bound at upstream enhancer regions in the glnA operon can interact
with RNA polymerases at the promoter regions, forming a DNA loop (Su et al., 1990). DNA
looping interactions can be detected by chromosome conformation capture assays, which rely on
proximity-based ligation of DNA fragments, identifying genomic regions that are physically
close, regardless of linear distance in the genome (Dekker et al., 2002). These experiments have
demonstrated looping interactions between many individual enhancers and promoters in human
18
cells, including the P-globin enhancers and promoter (Deng et al., 2012; Tolhuis et al., 2002;
Vakoc et al., 2005). These looping interactions are likely critical for gene activation; at the
p-
globin locus, forced looping by tethering of a looping factor was sufficient to activate gene
expression (Deng et al., 2012).
Loop formation is stabilized by interactions between the Mediator and cohesin
complexes. The cohesin complex, which is also involved in sister chromatin cohesion, is a ringshaped complex. Although it is unclear precisely how cohesin facilitates looping, the ring
structure has the required dimensions to encircle two strands of nucleosome-occupied DNA and
it is possible cohesin stabilizes looping interactions in this manner. Cohesin is loaded on to
DNA by NIPBL, a protein that is associated with the Mediator complex. Both Mediator and
cohesin globally co-occupy cell-type-specific enhancer and promoter regions, likely helping to
drive the cell-type-specific expression program by promoting enhancer-promoter loops (Kagey et
al., 2010). Global mapping of enhancer-promoter looping interactions is possible through the
ChIA-PET technique (chromatin interaction analysis by paired end tag sequencing). ChIA-PET
combines chromatin immunoprecipitation and proximity-based ligation to identify loops between
regions bound by a particular protein factor. ChIA-PET has been carried out in several systems,
identifying enhancer-promoter loops that are highly cell-type-specific, and correlated with
transcription levels (Chepelev et al., 2012; Fullwood et al., 2009; Li et al., 2012; Zhang et al.,
2013). Given that enhancers can act at large genomic distances from their target genes,
identifying the target gene of an enhancer is often difficult, and greater ability to map the
physical interactions between promoters and enhancers will improve our ability to understand
how enhancers influence gene expression in a given cell type.
EnhancerRNA
19
One emerging property of enhancers is the production of RNA from these regions
(reviewed in (Lai and Shiekhattar, 2014; Lam et al., 2014; Orom and Shiekhattar, 2013)).
Recent genome-wide sequencing of RNA species has revealed that enhancers are transcribed,
producing non-coding RNAs (De Santa et al., 2010; Kim et al., 2010). The level of RNAs
produced from enhancer regions is correlated with the expression of nearby genes, and in several
systems, induction of protein coding gene expression was preceded by an increase in the
production of enhancer RNAs, suggesting that the transcription of RNA from enhancers may
play a role in regulating the expression of protein coding genes (Hah et al., 2013; Kim et al.,
2010; Koch and Andrau, 2011). There are several hypotheses as to the functional significance of
the transcription of enhancer regions. This transcription may be only noise, simply a
consequence of the localization of the Pol II transcription apparatus to the looped
promoter/enhancer region. Alternatively, the act of transcription itself may facilitate the
maintenance of open chromatin at enhancer regions. Recently, evidence has suggested that the
RNA produced by enhancer transcription may have a functional role itself. Several studies have
indicated that depletion of enhancer RNAs can lead to reduced expression of nearby genes (Lam
et al., 2013; Melo et al., 2013; Orom et al., 2010).
The mechanism by which enhancer RNA contributes to enhancer function remains
unclear, but many recent studies suggest that RNA may interact with many enhancer-bound
proteins, stabilizing their association with the enhancer, and looping to the promoter region. In
these studies, depletion of enhancer RNA led to decreased looping interactions between
enhancers and promoters, and decreased occupancy of the Mediator coactivator complex (Lai et
al., 2013; Li et al., 2012). Evidence suggests that the Mediator complex may interact directly
with RNA, and given that many DNA-binding transcription factors also have RNA-binding
20
capabilities, these physical interactions between enhancer RNA and enhancer-associated protein
factors may play a key role in enhancer function (Burdach et al., 2012; Lai et al., 2013; Ng et al.,
2013).
Genome-wide identification of enhancerelements
With the completion of the human genome project, and ever lowering costs of genomic
sequencing, we have been increasingly able to obtain global maps of enhancer regions in many
cell types. The work of individual laboratories, as well as large consortiums, such as the
ENCODE project, have contributed to the identification of genomic regulatory regions across a
wide variety of cell types (Bernstein et al., 2012). Several strategies exist to identify enhancer
regions. As described above, enhancers are occupied by a transcription factors, as well as many
transcriptional coactivators and the transcriptional apparatus. By determining the genomic
localization of these enhancer-bound factors, we can generate a genome-wide map of enhancer
regions. One strategy for determining genomic binding sites is chromatin immunoprecipitation
followed by massively parallel sequencing (ChIP-seq). In this method, chemical crosslinking is
used to link protein factors to their DNA binding sites. Chromatin is then sheared, and the
protein factors and their accompanying DNA fragments are immunoprecipitated using an
antibody to the desired DNA-associated protein factor. The immunoprecipitated DNA is then
purified and sequenced, giving the genome-wide binding sites for the desired factor. Several
alternate strategies can be used, including ChIP-exo, in which exonuclease digestion is used to
generate DNA fragments centered around the protein bound, protected regions (Rhee and Pugh,
2011). In DNA adenine methyltransferase identification (DamID), the protein of interest is fused
to DNA adenine methyltransferase, and these methylated regions are amplified and sequenced
(Luo et al., 2011; Wu and Yao, 2013).
21
When the transcription factors for a particular cell type are known, they can be used to
map enhancer regions. For example, in mouse embryonic stem cells (mESC), Oct4, Sox2 and
Nanog have been described as master transcription factors (Ng and Surani, 2011; Orkin and
Hochedlinger, 2011; Young, 2011). ChIP-seq of Oct4, Sox2 and Nanog can identify putative
enhancer regions that are bound by these three factors. Function testing of these regions reveals
that binding of the three transcription factors is predictive of enhancer function in a reporter
assay (Chen et al., 2008). In addition to transcription factors, enhancers are occupied by many
transcriptional cofactors, including the Mediator complex, the bromodomain containing protein
BRD4, and the histone acetyltransferase (HAT) p300. ChIP-Seq of each of these factors has
been used to identify enhancer regions, and may be particularly useful in cases where the celltype-specific transcription factors are unknown (Chapuy et al., 2013; Heintzman et al., 2007;
Kagey et al., 2010; Rada-Iglesias et al., 2011; Visel et al., 2009).
The binding of transcription factors and associated cofactors creates areas of open
chromatin at enhancers, a feature that can be exploited in enhancer identification. One such
technique, DNase hypersensitive site sequencing (DNase-Seq), involves digestion of accessible
DNA regions by DNase I, followed by sequencing of these hypersensitive regions (Crawford et
al., 2006; John et al., 2013; Song and Crawford, 2010). In addition, several newer techniques
have been developed based on chromatin accessibility. FAIRE-Seq (Formaldehyde-assisted
identification of regulatory elements) uses the fact that formaldehyde crosslinking is more
efficient at nucleosome occupied regions to separate these from nucleosome free regions, which
are then sequenced (Giresi et al., 2007; Simon et al., 2012). ATAC-seq, (assay for transposase
accessible chromatin) utilizes the transposase Tn5 to insert sequencing adapters in to open
chromatin regions (Buenrostro et al., 2013).
22
Although enhancer regions bound by transcription factors are typically depleted of
nucleosomes, the nucleosomes immediately surrounding transcription factor binding sites in
enhancers often carry stereotypical modification that can be used to identify these regions. As
described previously, promoters tend to be trimethylated at histone H3 lysine 4 (H3K3me3);
enhancer regions are often monomethylated at this residue (H3K4mel). This H3K4mel mark
occurs both at active enhancers and at enhancer regions that appear to be primed, and are utilized
in later lineages. Acetylation of H3K27 appears to mark primarily active enhancers (Creyghton
et al., 2010; Rada-Iglesias et al., 2011). Identification of enhancers using histone modifications
has been widely used, including by the ENCODE project, to map enhancer regions in a wide
variety of organisms and cell types (Bernstein et al., 2012; Bonn et al., 2012; Ernst et al., 2011;
Kharchenko et al., 2011; Shen et al., 2012; Wamstad et al., 2012).
In addition to these enhancer-mapping strategies, several new methods for functional
identification of enhancers have emerged in recent years. Traditionally, as with the discovery of
the SV40 enhancer, enhancer function is confirmed using reporter assays testing the ability of a
candidate regulatory sequence to increase the transcription of a reporter gene. Recently, this
strategy has been adapted to be used in a high throughput, genome-wide manner, allowing global
identification of functional enhancer elements. One such strategy, STARR-seq (self-transcribing
active regulatory region sequencing) relies on an enhancer's ability to drive expression in a
position- and orientation-independent manner by cloning putative enhancers downstream of a
minimal promoter. In this assay, regulatory regions that enhance their own transcription will
produce higher levels of corresponding RNA, which is measured by sequencing (Arnold et al.,
2013).
Super-enhancers
23
My thesis research, described in chapter two, led to the identification of super-enhancers.
Super-enhancers are clusters of transcriptional enhancers, that span large regions of the genome
and are occupied by high levels of transcription factors and transcriptional coactivators, as well
as the Pol II transcription apparatus itself. Work by others in the Young lab described superenhancers in mouse embryonic stem cells (mESCs). Here, enhancers were identified as regions
occupied by the mESC master transcription factors Oct4, Sox2, and Nanog, determined by ChIPseq. While most mESC enhancer regions were a few hundred base pairs in size, a small subset
of regions were made up of clusters of enhancers that spanned large regions of the genome, up to
50kb. These super-enhancers could be separated from smaller typical enhancers based on their
occupancy by exceptional amounts of the Mediator coactivator (Whyte et al., 2013). Out of over
8,000 enhancer regions in mESCs, 231 were identified as super-enhancers, regions whose
median length is over an order of magnitude greater than typical enhancers, and which are
occupied by an order of magnitude greater level of Mediator. The high level of occupancy of
RNA Pol II and coactivators at super-enhancers is concomitant with higher expression levels of
genes associated with super-enhancers when compared to genes associated with typical
enhancers.
Examination of the genes associated with super-enhancers revealed that many of these
genes are known to be important in mESC biology, including Oct4, Sox2 and Nanog themselves,
as well as other well-known factors, such as Klf4, Esrrb, and the mir290 cluster. Extending this
analysis to other mouse tissues, super-enhancers were again found to be associated with celltype-specific genes, many which are known to play key roles in defining a particular tissue type.
For example, the muscle master transcription factor, MyoD, is associated with a super-enhancer
in mouse skeletal muscle myoblasts, while the B-cell-specific transcription factor Pu. 1 is
24
associated with a super-enhancer in mouse B-cells. Therefore, we believe that super-enhancers
play a role in sustaining high levels of transcription of key genes that function in defining cell
state, across a variety of cell types.
As described above, enhancers function through the cooperative interactions between
many transcription factors and cofactors (Carey, 1998; Carey et al., 1990; Giese et al., 1995; Kim
and Maniatis, 1997; Thanos and Maniatis, 1995). Enhancers bound by many cooperatively
interacting factors can lose activity more rapidly than enhancers with fewer binding sites when
the levels of enhancer-bound factors are reduced (Giniger and Ptashne, 1988; Griggs and
Johnston, 1991). Super-enhancers, comprising of clusters of enhancer regions highly occupied
by transcriptional coactivators, are indeed more sensitive to the loss of the enhancer-associated
factors Oct4 and Mediator. Knockdown of these proteins resulted in a greater loss in expression
of super-enhancer-associated genes compared to genes associated with typical enhancers. This
sensitivity to perturbation may facilitate gene expression changes during development, when a
cell must decommission cell-type-specific enhancers to make way for a new gene expression
program.
Furthering our study of super-enhancers, we examined these regulatory regions in a panel
of 86 human cell types, as well as several human tumor cell lines (Hnisz et al., 2013). In this
study, the widely profiled enhancer-associated histone modification H3K27ac was used to
identify enhancer regions, and to rank these regions based on ChIP-seq signal to differentiate
super- and typical enhancers. Confirming the previous results in mouse tissue, super-enhancers
in human tissue were associated with cell type-specific genes linked to biological processes that
were important determinants of the cell type and function (Hnisz et al., 2013). Many key celltype-specific transcription factors are associated with super-enhancers, possibly forming an
25
interconnected autoregulatory loop. For example, in mESC, the Oct4, Sox2 and Nanog genes are
associated with super-enhancers, which are bound by each of these three transcription factors
(Whyte et al., 2013). Therefore, identifying super-enhancers may help to identify the master
transcription factors, and the regulatory network driving cell state in a variety of cell types.
Some super-enhancers overlap with other historically characterized examples of large
enhancer regions, for example, the locus control regions (LCRs), corresponding to the IGH
enhancer and
p-globin enhancer (Forrester et al.,
1990; Grosveld et al., 1987; Li et al., 2002;
Madisen and Groudine, 1994). More recent global analyses have identified similar large
enhancer regions. Examples include transcription initiation platforms (TIPs), characterized by
binding of RNA Pol II and GTFs (Koch and Andrau, 2011); stretch enhancers, enhancer regions
defined by bioinformatics analysis of chromatin state (Parker et al., 2013); and DNA methylation
valleys, large regions of hypomethylated DNA (Xie et al., 2013). The identification, by different
strategies, of these large cell-type-specific enhancer regions supports the idea that large, superenhancer regions are critical for regulating genes that define and control cell state.
Conclusion
Enhancers are now well recognized as key drivers of cell-type-specific expression,
helping to specify expression programs through the integration of many cooperatively acting
sequence-specific transcription factors and coactivators. Other key properties of enhancers
include their ability to act at great genomic distances, and without regard to the orientation of
their target genes. This function is facilitated by looping interactions that bring enhancers and
promoters into close physical proximity. In addition, transcription of enhancers themselves, and
the RNA produced, may prove to be an important aspect of enhancer function. Recent advances
allowing the genome-wide mapping of enhancer regions underscores the importance of these
26
elements in controlling cell-type-specific gene expression programs. Although each cell type
typically contains thousands of active enhancer regions, the identification of super-enhancers
may help to further define the critical genes and regulatory regions that control and define cell
state.
Dysregulation of enhancer function in human disease
Early examples of enhancerinvolvement in disease
Early work in deciphering the causes of genetic disease uncovered several examples in
which mutations in regulatory regions, rather than coding sequence, was associated with disease
phenotypes (reviewed in (Kleinjan and Lettice, 2008; Kleinjan and van Heyningen, 2005)). For
example, thalassemia, a blood disorder resulting from imbalanced levels of the oxygen-carrying
factors a- and P-globin, can be caused by mutations in the protein coding regions of these genes,
or by alterations that affect regulatory regions. For example, P-thalassemia can be caused by
translocations that remove the P-globin locus control region, a well-characterized distal enhancer
responsible for driving high level expression of the P-globin gene (Driscoll et al., 1989; Kioussis
et al., 1983). In another example, point mutations or translocations affecting a regulatory region
upstream of the sonic hedgehog (SHH) gene, an important regulator of limb and brain
development, were found to be associated with inherited preaxial polydactyly (Lettice et al.,
2003). Interestingly, mutations or deletions of the coding region of the SHH gene cause a
different disease, holoprosencephaly, a developmental disorder of the brain (Belloni et al., 1996;
Roessler et al., 1996; Roessler et al., 1997). The mutations causing preaxial polydactyly occur in
a limb-specific enhancer, resulting in limb, rather than brain defects (Kleinjan and Lettice, 2008;
Lettice et al., 2003).
Genetic sequence variationin enhancers contributes to complex human traits and disease
27
Many of these early studies investigated relatively simple traits and disorders, involving
single gene, Mendelian inheritance patterns. More recently, genome-wide, high throughput
studies have allowed greater ability to identify the involvement of regulatory regions in disease.
Genome-wide association studies (GWAS) identify common genetic sequence variants (typically
single nucleotide polymorphisms, SNPs) that are associated with complex human traits or
diseases (Stranger et al., 2011). Many hundreds of GWASs have been performed, assessing a
wide spectrum of traits and disorders. In these studies, the vast majority of trait- or diseaseassociated SNPs occur in non-coding regions of the genome (Hindorff et al., 2009; Manolio,
2010). Although many early studies focused on SNPs within coding regions, it has become
apparent that sequence variation in non-coding regions plays an important role in regulating gene
expression. Many disease- or trait- associated SNPs have been mapped to regions termed
expression quantitative trait loci (eQTLs), genomic regions that influence the expression of a
particular gene or genes (reviewed in (Cookson et al., 2009)).
With the ability to obtain global maps of regulatory regions in many cell types, it has
become apparent that the majority of these non-coding SNPs occur in regulatory regions of the
genome, including enhancers (Maurano et al., 2012). Further, disease-associated SNPs tend to
occur in regulatory elements found in cell types that are related to disease pathology. For
example, SNPs associated with diseases such as Alzheimer's disease or bipolar disorder tend to
occur in regulatory regions in brain tissue, while SNPs associated with heart-related traits and
disorders, such as coronary heart disease, tend to occur in heart-specific regulatory regions
(Maurano et al., 2012). The cell-type specific nature of enhancer function explains how genomic
variants, which are present in all cells in the body, can cause cell- or tissue-specific phenotypes.
GWA studies provide only a statistical association between a particular genome region and a trait
28
or disease, and it is often difficult to extend this knowledge to gain an understanding of the
underlying biology resulting in phenotypic differences (Frazer et al., 2009; Fugger et al., 2012).
Overlaying maps of enhancer regions in relevant cell types with disease-associated loci identified
by GWAS may help identify the functional variants (Schaub et al., 2012).
Although unraveling the functional significance of disease- or trait-associated SNPs
remains difficult, several cases have demonstrated that sequence variants can disrupt
transcription factor binding sites within enhancer regions, resulting in altered expression of a
nearby gene (Bauer et al., 2013; Schaub et al., 2012). For example, BCL1 IA is a well-studied
regulator of fetal hemoglobin expression (Esteghamat et al., 2013; Sankaran et al., 2008;
Sankaran et al., 2009; Xu et al., 2013; Xu et al., 2010). Several studies examining fetal
hemoglobin expression levels identified trait-associated SNPs within the second intron of
BCL1 A (Bhatnagar et al., 2011; Lettre et al., 2008; Menzel et al., 2007; Nuinoon et al., 2010;
Solovieff et al., 2010). These SNPs coincided with DNaseI HS regions that were also marked by
typical enhancer-associated histone modifications, and bound by the transcription factors
GATA 1 and TALl in erythroid cells. Closer examination of these enhancer SNP sites revealed
that this variant disrupted a binding motif recognized by GATA 1 and TALl, resulting in reduced
binding of these transcription factors, concomitant with modestly altered expression of BCL 11 A
and fetal hemoglobin (Bauer et al., 2013).
Genetic sequence variants in enhancerelements in cancer
Many of these GWA studies have highlighted genomic variants related to cancer. Again,
many of these cancer susceptibility-related SNPs occur in regulatory regions of the genome,
including promoters, regulatory regions within introns, and distal enhancers that may be many
hundreds of kilobases away from the affected gene. Disease associated sequence variants in
29
regulatory regions, including enhancers, have been identified in a variety of tumor types,
including colorectal (Broderick et al., 2007; Lubbe et al., 2012; Tomlinson et al., 2008), prostate
(Demichelis et al., 2012; Haiman et al., 2007), breast (Easton et al., 2007; Jiang et al., 2011),
nasopharangeal (Yew et al., 2012), renal (Schodel et al., 2012), lung (Liu et al., 2011), and
melanoma (Horn et al., 2013; Huang et al., 2013).
Of particular interest in cancer, many cancer-associated SNPs are located in the 8q24
gene desert, which is located near the MYC oncogene (reviewed in (Sur et al., 2013)). Variations
associated with cancer risk have been mapped to this location in a variety of tumor types,
including prostate, breast, colorectal, bladder, lung, ovary, pancreas, kidney and brain
(Ahmadiyeh et al., 2010; Al Olama et al., 2009; Amundadottir et al., 2006; Couch et al., 2009;
Domagk et al., 2007; Fletcher et al., 2008; Ghoussaini et al., 2008; Gruber et al., 2007;
Gudmundsson et al., 2013; Gudmundsson et al., 2007; Haiman et al., 2007; Kiemeney et al.,
2008; Li et al., 2008; Park et al., 2008; Pomerantz et al., 2009; Poynter et al., 2007; Shete et al.,
2009; Tenesa et al., 2008; Tomlinson et al., 2008; Turnbull et al., 2010; Tuupanen et al., 2009;
Yeager et al., 2009; Zanke et al., 2007; Zhang et al., 2014).
In colorectal cancer, several studies have further investigated the mechanisms by which
sequence variation in the 8q24 regions leads to cancer risk (Jia et al., 2009; Pomerantz et al.,
2009; Tuupanen et al., 2009). In one example, ChIP assays found that the colorectal cancer
predisposition SNP rs6983267 was located in an enhancer region, marked by H3K4mel and
p300 binding in a colorectal cancer cell line (Pomerantz et al., 2009). Reporter assays indicated
that this region had active enhancer function, and that the risk allele led to significantly greater
gene activation. Further examining the sequence context of this SNP found it was located in a
binding site for the transcription factor TCF7L2 (also known as TCF4), part of the Wnt signaling
30
pathway, known to be important in colon cancer pathogenesis (Polakis, 2000; Pomerantz et al.,
2009). The risk allele corresponds to the consensus motif for this TCF7L2, and ChIP of TCF7L2
indicated greater binding to the risk allele. Although this enhancer region is located nearly
400kb away from the MYC gene, physical interaction between the two was confirmed by
chromosome conformation capture (3C) experiments. Interestingly, mice lacking this enhancer
region are resistant to developing intestinal tumors, further underscoring the importance of this
enhancer in the development of colon cancer (Sur et al., 2012). This data provides another
example of the molecular mechanism by which sequence variation may lead to phenotypic
differences, in this case, altering binding of TCF7L2, leading to differential activity of an
enhancer.
Somatic changes in enhancers in tumor cells
In addition to these links between human sequence variation and cancer, there are several
examples of somatic mutation involving enhancer regions in cancer. Many B-cell tumors have
translocations linking immunoglobulin enhancer regions to oncogenes, such as MYC, and BCL2.
These translocation events can be the result of aberrant activation of the immunoglobulin gene
remodeling process, involving V(D)J recombination, somatic hypermutation, and class switch
recombination (Aplan, 2006; Kuppers and Dalla-Favera, 2001). Burkitt lymphoma, for example,
is characterized by translocations between the MYC oncogene and either the immunoglobulin
heavy locus, or less frequently, the immunoglobulin light chain ic or X loci (Boerma et al., 2009;
Hecht and Aster, 2000).
Several recent studies have indicated that epigenetic changes in the enhancer landscape
may play a significant role in tumor biology. One study examined the variation in enhancer
regions between colorectal carcinoma and normal colon epithelial cells by ChIP-seq of
31
H3K4me 1 (Akhtar-Zaidi et al., 2012). This identified thousands of enhancer loci that were
either gained or lost in colon cancer cells, corresponding to the expression levels of nearby
genes. These variant enhancer loci (VELs) were highly enriched in SNPs linked to colon cancer
risk, suggesting that the changes in enhancer regions may be a key factor in colon cancer
pathogenesis. Another study examined patterns of DNA methylation in cancer cells and found
that hypomethylation of enhancer regions was highly correlated with upregulation of cancerrelated genes, while hypermethylation of enhancer sites was correlated with down-regulation of
expression of cell-type-specific genes (Aran and Hellman, 2013; Aran et al., 2013).
Surprisingly, the methylation state of enhancer regions was better correlated with gene
expression changes than the methylation state of promoters, further underscoring the importance
of enhancer regions in driving gene expression changes in cancer.
Dysregulation of enhancer-acting factors in disease
In addition to sequence variation and mutation within enhancer sequences themselves,
many diseases have been linked to mutations or dysregulation of the transcription factors and
transcriptional coactivators that function at enhancer regions (reviewed in (Herz et al., 2014; Lee
and Young, 2013)). Many oncogenes overexpressed in cancer are transcription factors. For
example, the TALI transcription factor is overexpressed in half of T cell acute lymphoblastic
leukemias (T-ALLs), and drives the expression of a key network of genes involved in T-ALL
pathogenesis (Sanda et al., 2012). The MYC oncogene, which is the most frequently amplified
oncogene in cancer, is a basic helix loop helix (bHLH) transcription factor, which binds to a six
nucleotide E-box sequence (Beroukhim et al., 2010; Blackwood et al., 1991; Nesbit et al., 1999).
MYC tends to bind to the core promoter region of genes, but when it is overexpressed, it begins
to bind to lower affinity E-box-like regions in both enhancer and promoter (Lin et al., 2012).
32
This increased promoter binding and enhancer invasion of MYC is accompanied by global
amplification of transcription in the tumor cell, possibly providing more raw material needed to
drive tumor growth and proliferation (Lin et al., 2012; Nie et al., 2012).
Cofactors and chromatin regulators that act at enhancers can also be deregulated in
cancer. Mutations in the Mediator coactivator complex can also occur in cancer; the MED12
subunit is frequently mutated in uterine leiomyomas and leiomyosarcomas, and prostate cancer
(Barbieri et al., 2012; Makinen et al., 2011a; Makinen et al., 2011b). CDK8, part of the same
Mediator module as MED12, is amplified, and acts as an oncogene in colon cancer and
melanoma (Firestein et al., 2008; Kapoor et al., 2010; Morris et al., 2008). MLL3 and MLL4,
histone methyltransferases that catalyze H3K4meI at enhancers (Herz et al., 2014; Hu et al.,
2013), are frequently mutated in a wide range of cancers, including colorectal cancer, breast
cancer, liver cancer, bladder carcinoma, medulloblastoma, non-Hodgkin lymphoma, and gastric
cancer (Ashktorab et al., 2010; Ellis and Perou, 2013; Fujimoto et al., 2012; Gui et al., 2011;
Jones et al., 2012; Kandoth et al., 2013; Morin et al., 2011; Parsons et al., 2011; Watanabe et al.,
2011; Zhang et al., 2012).
Super-enhancers in human disease
Disease-associated sequence variation and somatic changes in super-enhancer regions
may play an important role in human disease. Disease- and trait-associated SNPs are particularly
enriched in the super-enhancers of relevant cell types, when compared to typical enhancers
(Hnisz et al., 2013). In one example, two SNPs associated with Alzheimer's disease risk are
found in a super-enhancer associated with the BIN] gene in brain tissue. BIN] expression has
recently been linked to Alzheimer's disease risk, and in the same study, an insertion in the BIN]
super-enhancer was also linked to Alzheimer's risk (Chapuis et al., 2013). This enrichment of
33
disease-associated variation in super-enhancer regions was supported by work from Parker and
colleagues, who found an enrichment of type 2 diabetes-associated variation in stretch enhancers,
large enhancer regions, of islet cells (Parker et al., 2013).
As discussed further in chapter two, super-enhancers in tumor cells are associated with
many key oncogenes, including MYC, the most commonly overexpressed oncogene in cancer
(Beroukhim et al., 2010; Loven et al., 2013). Comparisons of super-enhancers between tumor
cells and corresponding healthy tissue revealed that a majority of the genes acquired by tumor
cells had functions that were related to cancer hallmark traits, such as evasion of growth
suppressors, or activation of invasion and metastasis (Hnisz et al., 2013). Tumor cells can
acquire super-enhancers through several mechanisms, including translocation, focal
amplification, and overexpression of transcription factors. Examples of these cases include
multiple myeloma, where the highly active IGH enhancer can be translocated near the MYC
oncogene (Dib et al., 2008; Hnisz et al., 2013; Shou et al., 2000); T cell acute lymphoblastic
leukemia (T-ALL), where the master transcription factor TALl is overexpressed, and is bound to
a super-enhancer region near MYC (Bash et al., 1995; Brown et al., 1990; Hnisz et al., 2013); and
small cell lung cancer (SCLC), where focal amplifications of the MYC gene and surrounding
enhancer regions commonly occur (Iwakawa et al., 2013).
Many recent studies have identified super-enhancers in tumor cells, supporting the
observation that key oncogenes are often associated with super-enhancer regions. Chapuy et al.,
found that super-enhancers in diffuse large b-cell lymphoma (DLBCL) were associated with
several known oncogenes, and also identified OCA-B as a previously unknown oncogene driver
of DLBCL (Chapuy et al., 2013). Shi et al. identified a larger super-enhancer in the 8q24 locus
downstream of the MYC gene in acute myeloid leukemia (AML); this region was frequently
34
subject to focal amplifications in patient samples (Shi et al., 2013). Groeschel et al. found that a
translocation in AML places the EVIl gene under the control of a super-enhancer from the
GA TA2 gene, leading to aberrant expression of EVIl (Groeschel et al., 2014). Finally, Walker et
al. found that translocations at the 8q24 region in multiple myeloma often place super-enhancers
near the MYC gene, increasing expression of the MYC oncogene. In addition to the commonly
translocated immunoglobulin enhancers, including IGH,they found that other translocations
involve important myeloma genes, such as XBP1, along with their super-enhancers (Walker et
al., 2014).
Targeting enhancerfunction through small molecule inhibitorsof transcriptional
coactivators
Given the recent appreciation that disruptions in transcriptional regulators and regulatory
regions can play a key role in disease, there has been increasing interest in developing
therapeutics that target transcriptional regulators (reviewed in (Dawson and Kouzarides, 2012;
Stellrecht and Chen, 2011; Villicana et al., 2014)). Many small molecules have been developed
targeting chromatin modifying enzymes. Several of these types of inhibitors are FDA approved
for cancer treatment, including the DNA methyltransferase inhibitors azacytidine and decitabine,
and the histone deacetylase inhibitors vorinostat and romidepsin. Many additional small
molecules are under investigation, including additional DMNT and HDAC inhibitors, as well as
inhibitors of histone acetyltransferases (HATs) such as p300, lysine methyltransferases such as
the H3K79 methyltransferase DOT1 L, or the Polycomb complex member EZH2, and lysine
demethylases such as LSD1 (Reviewed in (Helin and Dhanak, 2013; Villicana et al., 2014).
Small molecule inhibitors targeting the histone reader, BRD4, have also shown promise as
35
therapeutic agents in many cancer types (Dawson et al., 2011; Delmore et al., 2011;
Filippakopoulos et al., 2010; Mertz et al., 2011; Zuber et al., 2011).
In addition to targeting chromatin regulators, other transcriptional inhibitors target the
general transcription factors and other complexes closely associated with RNA Pol II.
Triptolide, and its analog minnelide, inhibits the activity of the TFIIH general transcription factor
complex by binding covalently to the XPB subunit (Iben et al., 2002; Manzo et al., 2012; Titov
et al., 2011; Vispe et al., 2009). Many inhibitors target CDK kinases involved in the
transcriptional cycle, including CDK7, a component of TFIIH, CDK8, which forms a complex
with Mediator, and CDK9, the kinase component of P-TEFb (Villicana et al., 2014).
Flavopiridol, for example, is a pan-CDK inhibitor that has been investigated in the treatment of
chronic lymphocytic leukemia (CLL) (Christian et al., 2009). Many other CDK inhibitors are
under clinical investigation in a variety of cancers (Villicana et al., 2014). Most of these small
molecules are relatively non-specific inhibitors of CDK activity, and many can target non-CDK
kinases. Further research focused on the development of inhibitors selective for individual
CDKs may potentially limit side effects due to off target inhibition of non-transcriptional kinases
(Villicana et al., 2014; Wesierska-Gadek and Krystof, 2009).
A critical feature of cancer therapy is the ability to selectively target tumor cells, and the
oncogenic properties that define them. This raises the conundrum of how global inhibitors of
transcription can achieve a gene-selective and tumor-selective effect. Despite this, many of these
inhibitors, including small molecules targeting BRD4, appear able to selectively inhibit the
expression of key oncogenes in tumor cells, such as MYC (Bandopadhayay et al., 2014; Dawson
and Kouzarides, 2012; Dawson et al., 2011; Delmore et al., 2011; Filippakopoulos et al., 2010;
Mertz et al., 2011; Zuber et al., 2011). This selective effect on the MYC oncogene is especially
36
important, given the difficulty in targeting transcription factors, such as MYC, with conventional
small molecule therapies (Damell, 2002).
There are several properties of oncogenes, and of tumor cells, that may allow
transcriptional inhibitors to act in a gene-specific and tumor-specific manner. First, many
oncogenes and pro-apoptotic factors have short half-lives, resulting in rapid down-regulation of
these transcripts upon global inhibition of transcription. Tumor cells often become addicted to
certain oncogenes, rendering tumor cells particularly sensitive to reduced levels of oncogene
expression. In addition, tumor cells may require higher sustained levels of transcription than
normal cells, as suggested by the transcriptional amplification effected by the MYC oncogene.
Lastly, in the second chapter of this thesis, I describe a mechanism by which disruption of superenhancers by the BRD4 inhibitor JQI leads to the selective downregulation of tumor oncogenes.
In multiple myeloma cells, key oncogenes, including MYC, are associated with large, BRD4occupied super-enhancers. These super-enhancers are particularly sensitive to JQ 1 treatment,
resulting in a disproportionate loss of BRD4 occupancy at super-enhancer regions. Loss of
BRD4 binding at super-enhancers is concomitant with a loss of transcriptional elongation at
associated genes, and a decrease in mRNA levels of these key oncogenes. Again, recent work
has corroborated these findings, lending support to the idea that some transcriptional inhibitors
may be able to selectively target super-enhancers and associated genes. As described above,
Chapuy et al. investigated super-enhancer regions in DLBCL. In this study, they observed that
super-enhancer-associated genes, including OCA-B, were particularly sensitive to BRD4
inhibition by JQI (Chapuy et al., 2013). Groeschel et al. (2014) found that the expression of
EVIl was especially sensitive to JQI treatment in AML cell lines in which EVII is driven by a
translocated, highly BRD4-occupied super-enhancer. They also observed greater loss of BRD4
37
occupancy in response to JQ 1 at super-enhancers when compared to typical enhancers in these
cell lines (Groschel et al., 2014). Wang et al. examined the response of T-lymphoblastic
leukemia (T-LL) to a gamma-secretase inhibitor (GSI), which prevents nuclear localization of
the transcription factor Notch. They found that regions showing a dynamic change in Notch
occupancy upon GSI treatment overlapped significantly with super-enhancers. These dynamic
super-enhancer regions in turn correlated with changes in gene expression of associated genes
upon drug treatment (Wang et al., 2014).
Conclusion
Together, this evidence provides a strong impetus for further study of enhancer regions,
and super-enhancers in particular, as a therapeutic target in cancer and other human disease,
which will be discussed further in chapter three. Enhancer regions play an important role in
controlling gene expression programs through their association with sequence-specific
transcription factors and looping interactions with promoter regions. New sequencing
technologies have allowed for the genome-wide mapping of enhancer regions in a wide range of
cell types, giving new appreciation to the role of these regions in human development and
disease. The identification of super-enhancers may serve to further focus future research by
marking the critical genes involved in cell or disease-state. Because super-enhancers are
associated with many critical tumor oncogenes in the cancers thus far investigated, identifying
the super-enhancers in a given tumor may provide valuable insight into its biology. In addition,
given the greater ability to target transcriptional regulators and other enhancer-associated factors,
and the particular sensitivity super-enhancers display in response to many perturbations, this
strategy may be an effective way to selectively target important super-enhancer-associated
oncogenes.
38
References
Adhya, S. (1989). Multipartite genetic control elements: communication by DNA loop. Annu
Rev Genet 23, 227-250.
Agalioti, T., Lomvardas, S., Parekh, B., Yie, J., Maniatis, T., and Thanos, D. (2000). Ordered
recruitment of chromatin modifying and general transcription factors to the IFN-beta promoter.
Cell 103, 667-678.
Ahmadiyeh, N., Pomerantz, M.M., Grisanzio, C., Herman, P., Jia, L., Almendro, V., He, H.H.,
Brown, M., Liu, X.S., Davis, M., et al. (2010). 8q24 prostate, breast, and colon cancer risk loci
show tissue-specific long-range interaction with MYC. Proc Natl Acad Sci U S A 107, 97429746.
Akhtar-Zaidi, B., Cowper-Sal-lari, R., Corradin, 0., Saiakhova, A., Bartels, C.F.,
Balasubramanian, D., Myeroff, L., Lutterbaugh, J., Jarrar, A., Kalady, M.F., el al. (2012).
Epigenomic enhancer profiling defines a signature of colon cancer. Science 336, 736-739.
Al Olama, A.A., Kote-Jarai, Z., Giles, G.G., Guy, M., Morrison, J., Severi, G., Leongamornlert,
D.A., Tymrakiewicz, M., Jhavar, S., Saunders, E., et al. (2009). Multiple loci on 8q24 associated
with prostate cancer susceptibility. Nat Genet 41, 1058-1060.
Almada, A.E., Wu, X., Kriz, A.J., Burge, C.B., and Sharp, P.A. (2013). Promoter directionality
is controlled by Ul snRNP and polyadenylation signals. Nature 499, 360-363.
Amundadottir, L.T., Sulem, P., Gudmundsson, J., Helgason, A., Baker, A., Agnarsson, B.A.,
Sigurdsson, A., Benediktsdottir, K.R., Cazier, J.B., Sainz, J., et al. (2006). A common variant
associated with prostate cancer in European and African populations. Nat Genet 38, 652-658.
Antoniou, M., deBoer, E., Habets, G., and Grosveld, F. (1988). The human beta-globin gene
contains multiple regulatory regions: identification of one promoter and two downstream
enhancers. EMBO J 7, 377-384.
Aplan, P.D. (2006). Causes of oncogenic chromosomal translocation. Trends Genet 22, 46-55.
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, R2 1.
Arnold, C.D., Gerlach, D., Stelzer, C., Boryn, L.M., Rath, M., and Stark, A. (2013). Genomewide quantitative enhancer activity maps identified by STARR-seq. Science 339, 1074-1077.
Ashktorab, H., Schaffer, A.A., Daremipouran, M., Smoot, D.T., Lee, E., and Brim, H. (2010).
Distinct genetic alterations in colorectal cancer. PLoS One 5, e8879.
39
Augereau, P., and Chambon, P. (1986). The mouse immunoglobulin heavy-chain enhancer:
effect on transcription in vitro and binding of proteins present in HeLa and lymphoid B cell
extracts. EMBO J 5, 1791-1797.
Bandopadhayay, P., Bergthold, G., Nguyen, B., Schubert, S., Gholamin, S., Tang, Y., Bolin, S.,
Schumacher, S.E., Zeid, R., Masoud, S., et al. (2014). BET bromodomain inhibition of MYCamplified medulloblastoma. Clin Cancer Res 20, 912-925.
Banerji, J., Olson, L., and Schaffner, W. (1983). A lymphocyte-specific cellular enhancer is
located downstream of the joining region in immunoglobulin heavy chain genes. Cell 33, 729740.
Banerji, J., Rusconi, S., and Schaffner, W. (1981). Expression of a beta-globin gene is enhanced
by remote SV40 DNA sequences. Cell 27, 299-308.
Bannister, A.J., Zegerman, P., Partridge, J.F., Miska, E.A., Thomas, J.O., Allshire, R.C., and
Kouzarides, T. (2001). Selective recognition of methylated lysine 9 on histone H3 by the HP 1
chromo domain. Nature 410, 120-124.
Barbieri, C.E., Baca, S.C., Lawrence, M.S., Demichelis, F., Blattner, M., Theurillat, J.P., White,
T.A., Stojanov, P., Van Allen, E., Stransky, N., et al. (2012). Exome sequencing identifies
recurrent SPOP, FOXAI and MED12 mutations in prostate cancer. Nat Genet 44, 685-689.
Bash, R.O., Hall, S., Timmons, C.F., Crist, W.M., Amylon, M., Smith, R.G., and Baer, R.
(1995). Does activation of the TALl gene occur in a majority of patients with T-cell acute
lymphoblastic leukemia? A pediatric oncology group study. Blood 86, 666-676.
Bauer, D.E., Kamran, S.C., Lessard, S., Xu, J., Fujiwara, Y., Lin, C., Shao, Z., Canver, M.C.,
Smith, E.C., Pinello, L., et al. (2013). An erythroid enhancer of BCLI IA subject to genetic
variation determines fetal hemoglobin level. Science 342, 253-257.
Behringer, R.R., Hammer, R.E., Brinster, R.L., Palmiter, R.D., and Townes, T.M. (1987). Two 3'
sequences direct adult erythroid-specific expression of human beta-globin genes in transgenic
mice. Proc Natl Acad Sci U S A 84, 7056-7060.
Beisel, C., and Paro, R. (2011). Silencing chromatin: comparing modes and mechanisms. Nat
Rev Genet 12, 123-135.
Belloni, E., Muenke, M., Roessler, E., Traverso, G., Siegel-Bartelt, J., Frumkin, A., Mitchell,
H.F., Donis-Keller, H., Helms, C., Hing, A.V., et al. (1996). Identification of Sonic hedgehog as
a candidate gene responsible for holoprosencephaly. Nat Genet 14, 353-356.
Berger, S.L. (2007). The complex language of chromatin regulation during transcription. Nature
447, 407-412.
Bernstein, B.E., Birney, E., Dunham, I., Green, E.D., Gunter, C., and Snyder, M. (2012). An
integrated encyclopedia of DNA elements in the human genome. Nature 489, 57-74.
40
Beroukhim, R., Mermel, C.H., Porter, D., Wei, G., Raychaudhuri, S., Donovan, J., Barretina, J.,
Boehm, J.S., Dobson, J., Urashima, M., el al. (2010). The landscape of somatic copy-number
alteration across human cancers. Nature 463, 899-905.
Bhatnagar, P., Purvis, S., Barron-Casella, E., DeBaun, M.R., Casella, J.F., Arking, D.E., and
Keefer, J.R. (2011). Genome-wide association study identifies genetic variants influencing F-cell
levels in sickle-cell patients. J Hum Genet 56, 316-323.
Blackwood, E.M., Luscher, B., Kretzner, L., and Eisenman, R.N. (1991). The Myc:Max protein
complex and cell growth regulation. Cold Spring Harb Symp Quant Biol 56, 109-117.
Boerma, E.G., Siebert, R., Kluin, P.M., and Baudis, M. (2009). Translocations involving 8q24 in
Burkitt lymphoma and other malignant lymphomas: a historical review of cytogenetics in the
light of todays knowledge. Leukemia 23, 225-234.
Bonn, S., Zinzen, R.P., Girardot, C., Gustafson, E.H., Perez-Gonzalez, A., Delhomme, N.,
Ghavi-Helm, Y., Wilczynski, B., Riddell, A., and Furlong, E.E. (2012). Tissue-specific analysis
of chromatin state identifies temporal signatures of enhancer activity during embryonic
development. Nat Genet 44, 148-156.
Borggrefe, T., and Yue, X. (2011). Interactions between subunits of the Mediator complex with
gene-specific transcription factors. Semin Cell Dev Biol 22, 759-768.
Broderick, P., Carvajal-Carmona, L., Pittman, A.M., Webb, E., Howarth, K., Rowan, A., Lubbe,
S., Spain, S., Sullivan, K., Fielding, S., et al. (2007). A genome-wide association study shows
that common alleles of SMAD7 influence colorectal cancer risk. Nat Genet 39, 1315-1317.
Brown, L., Cheng, J.T., Chen, Q., Siciliano, M.J., Crist, W., Buchanan, G., and Baer, R. (1990).
Site-specific recombination of the tal- 1 gene is a common occurrence in human T cell leukemia.
EMBO J 9, 3343-3351.
Buenrostro, J.D., Giresi, P.G., Zaba, L.C., Chang, H.Y., and Greenleaf, W.J. (2013).
Transposition of native chromatin for fast and sensitive epigenomic profiling of open chromatin,
DNA-binding proteins and nucleosome position. Nat Methods 10, 1213-1218.
Bulger, M., and Groudine, M. (1999). Looping versus linking: toward a model for long-distance
gene activation. Genes Dev 13, 2465-2477.
Bulger, M., and Groudine, M. (2011). Functional and mechanistic diversity of distal transcription
enhancers. Cell 144, 327-339.
Burdach, J., O'Connell, M.R., Mackay, J.P., and Crossley, M. (2012). Two-timing zinc finger
transcription factors liaising with RNA. Trends Biochem Sci 37, 199-205.
Carey, M. (1998). The enhanceosome and transcriptional synergy. Cell 92, 5-8.
Carey, M., Lin, Y.S., Green, M.R., and Ptashne, M. (1990). A mechanism for synergistic
activation of a mammalian gene by GAL4 derivatives. Nature 345, 361-364.
41
Cedar, H., and Bergman, Y. (2012). Programming of DNA methylation patterns. Annu Rev
Biochem 81, 97-117.
Chapuis, J., Hansmannel, F., Gistelinck, M., Mounier, A., Van Cauwenberghe, C., Kolen, K.V.,
Geller, F., Sottejeau, Y., Harold, D., Dourlen, P., et al. (2013). Increased expression of BINl
mediates Alzheimer genetic risk by modulating tau pathology. Mol Psychiatry 18, 1225-1234.
Chapuy, B., McKeown, M.R., Lin, C.Y., Monti, S., Roemer, M.G., Qi, J., Rahl, P.B., Sun, H.H.,
Yeda, K.T., Doench, J.G., el al. (2013). Discovery and characterization of super-enhancerassociated dependencies in diffuse large B cell lymphoma. Cancer Cell 24, 777-790.
Chen, X., Xu, H., Yuan, P., Fang, F., Huss, M., Vega, V.B., Wong, E., Orlov, Y.L., Zhang, W.,
Jiang, J., el al. (2008). Integration of external signaling pathways with the core transcriptional
network in embryonic stem cells. Cell 133, 1106-1117.
Cheng, B., and Price, D.H. (2007). Properties of RNA polymerase II elongation complexes
before and after the P-TEFb-mediated transition into productive elongation. J Biol Chem 282,
21901-21912.
Chepelev, I., Wei, G., Wangsa, D., Tang, Q., and Zhao, K. (2012). Characterization of genomewide enhancer-promoter interactions reveals co-expression of interacting genes and modes of
higher order chromatin organization. Cell Res 22, 490-503.
Choi, O.R., and Engel, J.D. (1986). A 3' enhancer is required for temporal and tissue-specific
transcriptional activation of the chicken adult beta-globin gene. Nature 323, 731-734.
Christian, B.A., Grever, M.R., Byrd, J.C., and Lin, T.S. (2009). Flavopiridol in chronic
lymphocytic leukemia: a concise review. Clin Lymphoma Myeloma 9 Suppl 3, S 179-185.
Churchman, L.S., and Weissman, J.S. (2011). Nascent transcript sequencing visualizes
transcription at nucleotide resolution. Nature 469, 368-373.
Conaway, R.C., and Conaway, J.W. (2011). Function and regulation of the Mediator complex.
Curr Opin Genet Dev 21, 225-230.
Cookson, W., Liang, L., Abecasis, G., Moffatt, M., and Lathrop, M. (2009). Mapping complex
disease traits with global gene expression. Nat Rev Genet 10, 184-194.
Core, L.J., and Lis, J.T. (2008). Transcription regulation through promoter-proximal pausing of
RNA polymerase II. Science 319, 1791-1792.
Core, L.J., Waterfall, J.J., and Lis, J.T. (2008). Nascent RNA sequencing reveals widespread
pausing and divergent initiation at human promoters. Science 322, 1845-1848.
Couch, F.J., Wang, X., McWilliams, R.R., Bamlet, W.R., de Andrade, M., and Petersen, G.M.
(2009). Association of breast cancer susceptibility variants with risk of pancreatic cancer. Cancer
Epidemiol Biomarkers Prev 18, 3044-3048.
42
Crawford, G.E., Holt, I.E., Whittle, J., Webb, B.D., Tai, D., Davis, S., Margulies, E.H., Chen, Y.,
Bernat, J.A., Ginsburg, D., et al. (2006). Genome-wide mapping of DNase hypersensitive sites
using massively parallel signature sequencing (MPSS). Genome Res 16, 123-131.
Creyghton, M.P., Cheng, A.W., Welstead, G.G., Kooistra, T., Carey, B.W., Steine, E.J., Hanna,
J., Lodato, M.A., Frampton, G.M., Sharp, P.A., et al. (2010). Histone H3K27ac separates active
from poised enhancers and predicts developmental state. Proc Natl Acad Sci U S A 107, 2193121936.
Darnell, J.E., Jr. (2002). Transcription factors as targets for cancer therapy. Nat Rev Cancer 2,
740-749.
Davidson, I., Fromental, C., Augereau, P., Wildeman, A., Zenke, M., and Chambon, P. (1986).
Cell-type specific protein binding to the enhancer of simian virus 40 in nuclear extracts. Nature
323, 544-548.
Dawson, M.A., and Kouzarides, T. (2012). Cancer epigenetics: from mechanism to therapy. Cell
150, 12-27.
Dawson, M.A., Prinjha, R.K., Dittmann, A., Giotopoulos, G., Bantscheff, M., Chan, W.I.,
Robson, S.C., Chung, C.W., Hopf, C., Savitski, M.M., et al. (2011). Inhibition of BET
recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. Nature 478, 529533.
De Santa, F., Barozzi, I., Mietton, F., Ghisletti, S., Polletti, S., Tusi, B.K., Muller, H., Ragoussis,
J., Wei, C.L., and Natoli, G. (2010). A large fraction of extragenic RNA pol II transcription sites
overlap enhancers. PLoS Biol 8, e1000384.
de Villiers, J., Olson, L., Tyndall, C., and Schaffner, W. (1982). Transcriptional 'enhancers' from
SV40 and polyoma virus show a cell type preference. Nucleic Acids Res 10, 7965-7976.
Dekker, J., Rippe, K., Dekker, M., and Kleckner, N. (2002). Capturing chromosome
conformation. Science 295, 1306-1311.
Delmore, J.E., Issa, G.C., Lemieux, M.E., Rahl, P.B., Shi, J., Jacobs, H.M., Kastritis, E.,
Gilpatrick, T., Paranal, R.M., Qi, J., et al. (2011). BET bromodomain inhibition as a therapeutic
strategy to target c-Myc. Cell 146, 904-917.
Demichelis, F., Setlur, S.R., Banerjee, S., Chakravarty, D., Chen, J.Y., Chen, C.X., Huang, J.,
Beltran, H., Oldridge, D.A., Kitabayashi, N., el al. (2012). Identification of functionally active,
low frequency copy number variants at 15q21.3 and 12q21.31 associated with prostate cancer
risk. Proc Natl Acad Sci U S A 109, 6686-6691.
Deng, W., Lee, J., Wang, H., Miller, J., Reik, A., Gregory, P.D., Dean, A., and Blobel, G.A.
(2012). Controlling long-range genomic interactions at a native locus by targeted tethering of a
looping factor. Cell 149, 1233-1244.
43
Dey, A., Ellenberg, J., Farina, A., Coleman, A.E., Maruyama, T., Sciortino, S., LippincottSchwartz, J., and Ozato, K. (2000). A bromodomain protein, MCAP, associates with mitotic
chromosomes and affects G(2)-to-M transition. Mol Cell Biol 20, 6537-6549.
Dib, A., Gabrea, A., Glebov, O.K., Bergsagel, P.L., and Kuehl, W.M. (2008). Characterization of
MYC translocations in multiple myeloma cell lines. J Natl Cancer Inst Monogr, 25-31.
Domagk, D., Schaefer, K.L., Eisenacher, M., Braun, Y., Wai, D.H., Schleicher, C., DialloDanebrock, R., Bojar, H., Roeder, G., Gabbert, H.E., et al. (2007). Expression analysis of
pancreatic cancer cell lines reveals association of enhanced gene transcription and genomic
amplifications at the 8q22.1 and 8q24.22 loci. Oncol Rep 17, 399-407.
Driscoll, M.C., Dobkin, C.S., and Alter, B.P. (1989). Gamma delta beta-thalassemia due to a de
novo mutation deleting the 5' beta-globin gene activation-region hypersensitive sites. Proc Natl
Acad Sci U S A 86, 7470-7474.
Easton, D.F., Pooley, K.A., Dunning, A.M., Pharoah, P.D., Thompson, D., Ballinger, D.G.,
Struewing, J.P., Morrison, J., Field, H., Luben, R., et al. (2007). Genome-wide association study
identifies novel breast cancer susceptibility loci. Nature 447, 1087-1093.
Eberhardy, S.R., and Farnham, P.J. (200 1). c-Myc mediates activation of the cad promoter via a
post-RNA polymerase II recruitment mechanism. J Biol Chem 276, 48562-48571.
Eberhardy, S.R., and Farnham, P.J. (2002). Myc recruits P-TEFb to mediate the final step in the
transcriptional activation of the cad promoter. J Biol Chem 277, 40156-40162.
Ellis, M.J., and Perou, C.M. (2013). The genomic landscape of breast cancer as a therapeutic
roadmap. Cancer Discov 3, 27-34.
Ernst, J., Kheradpour, P., Mikkelsen, T.S., Shoresh, N., Ward, L.D., Epstein, C.B., Zhang, X.,
Wang, L., Issner, R., Coyne, M., el al. (2011). Mapping and analysis of chromatin state
dynamics in nine human cell types. Nature 473, 43-49.
Esteghamat, F., Gillemans, N., Bilic, I., van den Akker, E., Cantu, I., van Gent, T., Klingmuller,
U., van Lom, K., von Lindern, M., Grosveld, F., et al. (2013). Erythropoiesis and globin
switching in compound Klfl::Bcll la mutant mice. Blood 121, 2553-2562.
Feng, Y., Yang, Y., Ortega, M.M., Copeland, J.N., Zhang, M., Jacob, J.B., Fields, T.A., Vivian,
J.L., and Fields, P.E. (2010). Early mammalian erythropoiesis requires the DotIL
methyltransferase. Blood 116, 4483-4491.
Ferreira, R., Ohneda, K., Yamamoto, M., and Philipsen, S. (2005). GATAI function, a paradigm
for transcription factors in hematopoiesis. Mol Cell Biol 25, 1215-1227.
Filippakopoulos, P., Qi, J., Picaud, S., Shen, Y., Smith, W.B., Fedorov, 0., Morse, E.M., Keates,
T., Hickman, T.T., Felletar, I., et al. (2010). Selective inhibition of BET bromodomains. Nature
468, 1067-1073.
44
Firestein, R., Bass, A.J., Kim, S.Y., Dunn, I.F., Silver, S.J., Guney, I., Freed, E., Ligon, A.H.,
Vena, N., Ogino, S., et al. (2008). CDK8 is a colorectal cancer oncogene that regulates betacatenin activity. Nature 455, 547-551.
Fletcher, 0., Johnson, N., Gibson, L., Coupland, B., Fraser, A., Leonard, A., dos Santos Silva, I.,
Ashworth, A., Houlston, R., and Peto, J. (2008). Association of genetic variants at 8q24 with
breast cancer risk. Cancer Epidemiol Biomarkers Prev 17, 702-705.
Flynn, R.A., Almada, A.E., Zamudio, J.R., and Sharp, P.A. (2011). Antisense RNA polymerase
II divergent transcripts are P-TEFb dependent and substrates for the RNA exosome. Proc Natl
Acad Sci U S A 108, 10460-10465.
Forneris, F., Binda, C., Vanoni, M.A., Battaglioli, E., and Mattevi, A. (2005). Human histone
demethylase LSD1 reads the histone code. J Biol Chem 280, 41360-41365.
Forrester, W.C., Epner, E., Driscoll, M.C., Enver, T., Brice, M., Papayannopoulou, T., and
Groudine, M. (1990). A deletion of the human beta-globin locus activation region causes a major
alteration in chromatin structure and replication across the entire beta-globin locus. Genes Dev 4,
1637-1649.
Frazer, K.A., Murray, S.S., Schork, N.J., and Topol, E.J. (2009). Human genetic variation and its
contribution to complex traits. Nat Rev Genet 10, 241-251.
Fugger, L., McVean, G., and Bell, J.I. (2012). Genomewide association studies and common
disease--realizing clinical utility. N Engl J Med 367, 2370-2371.
Fujimoto, A., Totoki, Y., Abe, T., Boroevich, K.A., Hosoda, F., Nguyen, H.H., Aoki, M.,
Hosono, N., Kubo, M., Miya, F., et al. (2012). Whole-genome sequencing of liver cancers
identifies etiological influences on mutation patterns and recurrent mutations in chromatin
regulators. Nat Genet 44, 760-764.
Fullwood, M.J., Liu, M.H., Pan, Y.F., Liu, J., Xu, H., Mohamed, Y.B., Orlov, Y.L., Velkov, S.,
Ho, A., Mei, P.H., et al. (2009). An oestrogen-receptor-alpha-bound human chromatin
interactome. Nature 462, 58-64.
Gargano, B., Amente, S., Majello, B., and Lania, L. (2007). P-TEFb is a crucial co-factor for
Myc transactivation. Cell Cycle 6, 2031-2037.
Ghoussaini, M., Song, H., Koessler, T., Al Olama, A.A., Kote-Jarai, Z., Driver, K.E., Pooley,
K.A., Ramus, S.J., Kjaer, S.K., Hogdall, E., et al. (2008). Multiple loci with different cancer
specificities within the 8q24 gene desert. J Natl Cancer Inst 100, 962-966.
Giese, K., Kingsley, C., Kirshner, J.R., and Grosschedl, R. (1995). Assembly and function of a
TCR alpha enhancer complex is dependent on LEF- 1-induced DNA bending and multiple
protein-protein interactions. Genes Dev 9, 995-1008.
Giniger, E., and Ptashne, M. (1988). Cooperative DNA binding of the yeast transcriptional
activator GAL4. Proc Natl Acad Sci U S A 85, 382-386.
45
Giresi, P.G., Kim, J., McDaniell, R.M., Iyer, V.R., and Lieb, J.D. (2007). FAIRE
(Formaldehyde-Assisted Isolation of Regulatory Elements) isolates active regulatory elements
from human chromatin. Genome Res 17, 877-885.
Griggs, D.W., and Johnston, M. (1991). Regulated expression of the GAL4 activator gene in
yeast provides a sensitive genetic switch for glucose repression. Proc Natl Acad Sci U S A 88,
8597-8601.
Groschel, S., Sanders, M.A., Hoogenboezem, R., de Wit, E., Bouwman, B.A., Erpelinck, C., van
der Velden, V.H., Havermans, M., Avellino, R., van Lom, K., et al. (2014). A Single Oncogenic
Enhancer Rearrangement Causes Concomitant EVIl and GATA2 Deregulation in Leukemia.
Cell.
Grosveld, F., Antoniou, M., van Assendelft, G.B., de Boer, E., Hurst, J., Kollias, G., MacFarlane,
F., and Wrighton, N. (1987). The regulation of expression of human beta-globin genes. Prog Clin
Biol Res 251, 133-144.
Gruber, S.B., Moreno, V., Rozek, L.S., Rennerts, H.S., Lejbkowicz, F., Bonner, J.D., Greenson,
J.K., Giordano, T.J., Fearson, E.R., and Rennert, G. (2007). Genetic variation in 8q24 associated
with risk of colorectal cancer. Cancer Biol Ther 6, 1143-1147.
Gudmundsson, J., Sulem, P., Gudbjartsson, D.F., Masson, G., Petursdottir, V., Hardarson, S.,
Gudjonsson, S.A., Johannsdottir, H., Helgadottir, H.T., Stacey, S.N., et al. (2013). A common
variant at 8q24.21 is associated with renal cell cancer. Nat Commun 4, 2776.
Gudmundsson, J., Sulem, P., Manolescu, A., Amundadottir, L.T., Gudbjartsson, D., Helgason,
A., Rafnar, T., Bergthorsson, J.T., Agnarsson, B.A., Baker, A., et al. (2007). Genome-wide
association study identifies a second prostate cancer susceptibility variant at 8q24. Nat Genet 39,
631-637.
Guenther, M.G., Levine, S.S., Boyer, L.A., Jaenisch, R., and Young, R.A. (2007). A chromatin
landmark and transcription initiation at most promoters in human cells. Cell 130, 77-88.
Gui, Y., Guo, G., Huang, Y., Hu, X., Tang, A., Gao, S., Wu, R., Chen, C., Li, X., Zhou, L., et al.
(2011). Frequent mutations of chromatin remodeling genes in transitional cell carcinoma of the
bladder. Nat Genet 43, 875-878.
Hah, N., Murakami, S., Nagari, A., Danko, C.G., and Kraus, W.L. (2013). Enhancer transcripts
mark active estrogen receptor binding sites. Genome Res 23, 1210-1223.
Haiman, C.A., Patterson, N., Freedman, M.L., Myers, S.R., Pike, M.C., Waliszewska, A.,
Neubauer, J., Tandon, A., Schirmer, C., McDonald, G.J., et al. (2007). Multiple regions within
8q24 independently affect risk for prostate cancer. Nat Genet 39, 638-644.
Hansen, U., and Sharp, P.A. (1983). Sequences controlling in vitro transcription of SV40
promoters. EMBO J 2, 2293-2303.
46
Hargreaves, D.C., and Crabtree, G.R. (2011). ATP-dependent chromatin remodeling: genetics,
genomics and mechanisms. Cell Res 21, 396-420.
Hecht, J.L., and Aster, J.C. (2000). Molecular biology of Burkitt's lymphoma. J Clin Oncol 18,
3707-3721.
Heintzman, N.D., Stuart, R.K., Hon, G., Fu, Y., Ching, C.W., Hawkins, R.D., Barrera, L.O., Van
Calcar, S., Qu, C., Ching, K.A., et al. (2007). Distinct and predictive chromatin signatures of
transcriptional promoters and enhancers in the human genome. Nat Genet 39, 311-318.
Helin, K., and Dhanak, D. (2013). Chromatin proteins and modifications as drug targets. Nature
502, 480-488.
Herz, H.M., Hu, D., and Shilatifard, A. (2014). Enhancer Malfunction in Cancer. Mol Cell 53,
859-866.
Hesse, J.E., Nickol, J.M., Lieber, M.R., and Felsenfeld, G. (1986). Regulated gene expression in
transfected primary chicken erythrocytes. Proc Natl Acad Sci U S A 83, 4312-4316.
Hindorff, L.A., Sethupathy, P., Junkins, H.A., Ramos, E.M., Mehta, J.P., Collins, F.S., and
Manolio, T.A. (2009). Potential etiologic and functional implications of genome-wide
association loci for human diseases and traits. Proc Natl Acad Sci U S A 106, 9362-9367.
Hnisz, D., Abraham, B.J., Lee, T.I., Lau, A., Saint-Andre, V., Sigova, A.A., Hoke, H.A., and
Young, R.A. (2013). Super-enhancers in the control of cell identity and disease. Cell 155, 934947.
Ho, L., and Crabtree, G.R. (2010). Chromatin remodelling during development. Nature 463, 474484.
Horn, S., Figl, A., Rachakonda, P.S., Fischer, C., Sucker, A., Gast, A., Kadel, S., Moll, I.,
Nagore, E., Hemminki, K., et al. (2013). TERT promoter mutations in familial and sporadic
melanoma. Science 339, 959-961.
Hu, D., Gao, X., Morgan, M.A., Herz, H.M., Smith, E.R., and Shilatifard, A. (2013). The
MLL3/MLL4 branches of the COMPASS family function as major histone H3K4
monomethylases at enhancers. Mol Cell Biol 33, 4745-4754.
Huang, B., Yang, X.D., Zhou, M.M., Ozato, K., and Chen, L.F. (2009). Brd4 coactivates
transcriptional activation of NF-kappaB via specific binding to acetylated RelA. Mol Cell Biol
29, 1375-1387.
Huang, F.W., Hodis, E., Xu, M.J., Kryukov, G.V., Chin, L., and Garraway, L.A. (2013). Highly
recurrent TERT promoter mutations in human melanoma. Science 339, 957-959.
Iben, S., Tschochner, H., Bier, M., Hoogstraten, D., Hozak, P., Egly, J.M., and Grummt, I.
(2002). TFIIH plays an essential role in RNA polymerase I transcription. Cell 109, 297-306.
47
Iwakawa, R., Takenaka, M., Kohno, T., Shimada, Y., Totoki, Y., Shibata, T., Tsuta, K.,
Nishikawa, R., Noguchi, M., Sato-Otsubo, A., ei al. (2013). Genome-wide identification of
genes with amplification and/or fusion in small cell lung cancer. Genes Chromosomes Cancer
52, 802-816.
Jang, M.K., Mochizuki, K., Zhou, M., Jeong, H.S., Brady, J.N., and Ozato, K. (2005). The
bromodomain protein Brd4 is a positive regulatory component of P-TEFb and stimulates RNA
polymerase I-dependent transcription. Mol Cell 19, 523-534.
Jeong, K.W., Kim, K., Situ, A.J., Ulmer, T.S., An, W., and Stallcup, M.R. (2011). Recognition
of enhancer element-specific histone methylation by TIP60 in transcriptional activation. Nat
Struct Mol Biol 18, 1358-1365.
Jia, L., Landan, G., Pomerantz, M., Jaschek, R., Herman, P., Reich, D., Yan, C., Khalid, 0.,
Kantoff, P., Oh, W., et al. (2009). Functional enhancers at the gene-poor 8q24 cancer-linked
locus. PLoS Genet 5, e1000597.
Jiang, Y., Shen, H., Liu, X., Dai, J., Jin, G., Qin, Z., Chen, J., Wang, S., Wang, X., and Hu, Z.
(2011). Genetic variants at Ip 11.2 and breast cancer risk: a two-stage study in Chinese women.
PLoS One 6, e21563.
Jiang, Y.W., Veschambre, P., Erdjument-Bromage, H., Tempst, P., Conaway, J.W., Conaway,
R.C., and Kornberg, R.D. (1998). Mammalian mediator of transcriptional regulation and its
possible role as an end-point of signal transduction pathways. Proc Natl Acad Sci U S A 95,
8538-8543.
John, S., Sabo, P.J., Canfield, T.K., Lee, K., Vong, S., Weaver, M., Wang, H., Vierstra, J.,
Reynolds, A.P., Thurman, R.E., et al. (2013). Genome-scale mapping of DNase I
hypersensitivity. Curr Protoc Mol Biol Chapter27, Unit 21 27.
Jones, P.A. (2012). Functions of DNA methylation: islands, start sites, gene bodies and beyond.
Nat Rev Genet 13, 484-492.
Jones, W.D., Dafou, D., McEntagart, M., Woollard, W.J., Elmslie, F.V., Holder-Espinasse, M.,
Irving, M., Saggar, A.K., Smithson, S., Trembath, R.C., et al. (2012). De novo mutations in
MLL cause Wiedemann-Steiner syndrome. Am J Hum Genet 91, 358-364.
Kagey, M.H., Newman, J.J., Bilodeau, S., Zhan, Y., Orlando, D.A., van Berkum, N.L., Ebmeier,
C.C., Goossens, J., Rahl, P.B., Levine, S.S., et al. (2010). Mediator and cohesin connect gene
expression and chromatin architecture. Nature 467, 430-435.
Kanazawa, S., Soucek, L., Evan, G., Okamoto, T., and Peterlin, B.M. (2003). c-Myc recruits PTEFb for transcription, cellular proliferation and apoptosis. Oncogene 22, 5707-5711.
Kandoth, C., McLellan, M.D., Vandin, F., Ye, K., Niu, B., Lu, C., Xie, M., Zhang, Q.,
McMichael, J.F., Wyczalkowski, M.A., et al. (2013). Mutational landscape and significance
across 12 major cancer types. Nature 502, 333-339.
48
Kapoor, A., Goldberg, M.S., Cumberland, L.K., Ratnakumar, K., Segura, M.F., Emanuel, P.O.,
Menendez, S., Vardabasso, C., Leroy, G., Vidal, C.I., et al. (2010). The histone variant
macroH2A suppresses melanoma progression through regulation of CDK8. Nature 468, 11051109.
Kharchenko, P.V., Alekseyenko, A.A., Schwartz, Y.B., Minoda, A., Riddle, N.C., Ernst, J.,
Sabo, P.J., Larschan, E., Gorchakov, A.A., Gu, T., et al. (2011). Comprehensive analysis of the
chromatin landscape in Drosophila melanogaster. Nature 471, 480-485.
Kiemeney, L.A., Thorlacius, S., Sulem, P., Geller, F., Aben, K.K., Stacey, S.N., Gudmundsson,
J., Jakobsdottir, M., Bergthorsson, J.T., Sigurdsson, A., et al. (2008). Sequence variant on 8q24
confers susceptibility to urinary bladder cancer. Nat Genet 40, 1307-1312.
Kim, T.K., Hemberg, M., Gray, J.M., Costa, A.M., Bear, D.M., Wu, J., Harmin, D.A.,
Laptewicz, M., Barbara-Haley, K., Kuersten, S., et al. (2010). Widespread transcription at
neuronal activity-regulated enhancers. Nature 465, 182-187.
Kim, T.K., and Maniatis, T. (1997). The mechanism of transcriptional synergy of an in vitro
assembled interferon-beta enhanceosome. Mol Cell 1, 119-129.
Kioussis, D., Vanin, E., deLange, T., Flavell, R.A., and Grosveld, F.G. (1983). Beta-globin gene
inactivation by DNA translocation in gamma beta-thalassaemia. Nature 306, 662-666.
Kleinjan, D.A., and Lettice, L.A. (2008). Long-range gene control and genetic disease. Adv
Genet 61, 339-388.
Kleinjan, D.A., and van Heyningen, V. (2005). Long-range control of gene expression: emerging
mechanisms and disruption in disease. Am J Hum Genet 76, 8-32.
Koch, F., and Andrau, J.C. (2011). Initiating RNA polymerase II and TIPs as hallmarks of
enhancer activity and tissue-specificity. Transcription 2, 263-268.
Kollias, G., Hurst, J., deBoer, E., and Grosveld, F. (1987). The human beta-globin gene contains
a downstream developmental specific enhancer. Nucleic Acids Res 15, 5739-5747.
Kornberg, R.D. (2005). Mediator and the mechanism of transcriptional activation. Trends
Biochem Sci 30, 235-239.
Kornberg, R.D., and Lorch, Y. (1999). Chromatin-modifying and -remodeling complexes. Curr
Opin Genet Dev 9, 148-151.
Kouzarides, T. (2007). Chromatin modifications and their function. Cell 128, 693-705.
Krueger, B.J., Varzavand, K., Cooper, J.J., and Price, D.H. (2010). The mechanism of release of
P-TEFb and HEXIM 1 from the 7SK snRNP by viral and cellular activators includes a
conformational change in 7SK. PLoS One 5, e12335.
49
Kuppers, R., and Dalla-Favera, R. (2001). Mechanisms of chromosomal translocations in B cell
lymphomas. Oncogene 20, 5580-5594.
Kurdistani, S.K., Robyr, D., Tavazoie, S., and Grunstein, M. (2002). Genome-wide binding map
of the histone deacetylase Rpd3 in yeast. Nat Genet 31, 248-254.
Lachner, M., O'Carroll, D., Rea, S., Mechtler, K., and Jenuwein, T. (2001). Methylation of
histone H3 lysine 9 creates a binding site for HPl proteins. Nature 410, 116-120.
Lai, F., Orom, U.A., Cesaroni, M., Beringer, M., Taatjes, D.J., Blobel, G.A., and Shiekhattar, R.
(2013). Activating RNAs associate with Mediator to enhance chromatin architecture and
transcription. Nature 494, 497-501.
Lai, F., and Shiekhattar, R. (2014). Enhancer RNAs: the new molecules of transcription. Cuff
Opin Genet Dev 25C, 38-42.
Lam, M.T., Cho, H., Lesch, H.P., Gosselin, D., Heinz, S., Tanaka-Oishi, Y., Benner, C.,
Kaikkonen, M.U., Kim, A.S., Kosaka, M., et al. (2013). Rev-Erbs repress macrophage gene
expression by inhibiting enhancer-directed transcription. Nature 498, 511-515.
Lam, M.T., Li, W., Rosenfeld, M.G., and Glass, C.K. (2014). Enhancer RNAs and regulated
transcriptional programs. Trends Biochem Sci 39, 170-182.
Lee, M.G., Wynder, C., Bochar, D.A., Hakimi, M.A., Cooch, N., and Shiekhattar, R. (2006).
Functional interplay between histone demethylase and deacetylase enzymes. Mol Cell Biol 26,
6395-6402.
Lee, T.I., and Young, R.A. (2000). Transcription of eukaryotic protein-coding genes. Annu Rev
Genet 34, 77-137.
Lee, T.I., and Young, R.A. (2013). Transcriptional regulation and its misregulation in disease.
Cell 152, 1237-1251.
Lee, W., Haslinger, A., Karin, M., and Tjian, R. (1987). Activation of transcription by two
factors that bind promoter and enhancer sequences of the human metallothionein gene and SV40.
Nature 325, 368-372.
Lejeune, E., and Allshire, R.C. (2011). Common ground: small RNA programming and
chromatin modifications. Curr Opin Cell Biol 23, 258-265.
LeRoy, G., Rickards, B., and Flint, S.J. (2008). The double bromodomain proteins Brd2 and
Brd3 couple histone acetylation to transcription. Mol Cell 30, 51-60.
Lettice, L.A., Heaney, S.J., Purdie, L.A., Li, L., de Beer, P., Oostra, B.A., Goode, D., Elgar, G.,
Hill, R.E., and de Graaff, E. (2003). A long-range Shh enhancer regulates expression in the
developing limb and fin and is associated with preaxial polydactyly. Hum Mol Genet 12, 17251735.
50
Lettre, G., Sankaran, V.G., Bezerra, M.A., Araujo, A.S., Uda, M., Sanna, S., Cao, A.,
Schlessinger, D., Costa, F.F., Hirschhorn, J.N., ef al. (2008). DNA polymorphisms at the
BCL1 A, HBS IL-MYB, and beta-globin loci associate with fetal hemoglobin levels and pain
crises in sickle cell disease. Proc Natl Acad Sci U S A 105, 11869-11874.
Levine, M. (2010). Transcriptional enhancers in animal development and evolution. Curr Biol
20, R754-763.
Li, B., Carey, M., and Workman, J.L. (2007). The role of chromatin during transcription. Cell
128, 707-719.
Li, G., Ruan, X., Auerbach, R.K., Sandhu, K.S., Zheng, M., Wang, P., Poh, H.M., Goh, Y., Lim,
J., Zhang, J., el al. (2012). Extensive promoter-centered chromatin interactions provide a
topological basis for transcription regulation. Cell 148, 84-98.
Li, L., Plummer, S.J., Thompson, C.L., Merkulova, A., Acheson, L.S., Tucker, T.C., and Casey,
G. (2008). A common 8q24 variant and the risk of colon cancer: a population-based case-control
study. Cancer Epidemiol Biomarkers Prev 17, 339-342.
Li, Q., Peterson, K.R., Fang, X., and Stamatoyannopoulos, G. (2002). Locus control regions.
Blood 100, 3077-3086.
Lin, C., Garruss, A.S., Luo, Z., Guo, F., and Shilatifard, A. (2013). The RNA Pol II elongation
factor E113 marks enhancers in ES cells and primes future gene activation. Cell 152, 144-156.
Lin, C.Y., Loven, J., Rahl, P.B., Paranal, R.M., Burge, C.B., Bradner, J.E., Lee, T.I., and Young,
R.A. (2012). Transcriptional amplification in tumor cells with elevated c-Myc. Cell 151, 56-67.
Liu, G., Gramling, S., Munoz, D., Cheng, D., Azad, A.K., Mirshams, M., Chen, Z., Xu, W.,
Roberts, H., Shepherd, F.A., el al. (2011). Two novel BRM insertion promoter sequence variants
are associated with loss of BRM expression and lung cancer risk. Oncogene 30, 3295-3304.
Loven, J., Hoke, H.A., Lin, C.Y., Lau, A., Orlando, D.A., Vakoc, C.R., Bradner, J.E., Lee, T.I.,
and Young, R.A. (2013). Selective inhibition of tumor oncogenes by disruption of superenhancers. Cell 153, 320-334.
Lubbe, S.J., Pittman, A.M., Olver, B., Lloyd, A., Vijayakrishnan, J., Naranjo, S., Dobbins, S.,
Broderick, P., Gomez-Skarmeta, J.L., and Houlston, R.S. (2012). The 14q22.2 colorectal cancer
variant rs4444235 shows cis-acting regulation of BMP4. Oncogene 31, 3777-3784.
Luo, S.D., Shi, G.W., and Baker, B.S. (2011). Direct targets of the D. melanogaster DSXF
protein and the evolution of sexual development. Development 138, 2761-2771.
Luo, Z., Lin, C., and Shilatifard, A. (2012). The super elongation complex (SEC) family in
transcriptional control. Nat Rev Mol Cell Biol 13, 543-547.
51
Madisen, L., and Groudine, M. (1994). Identification of a locus control region in the
immunoglobulin heavy-chain locus that deregulates c-myc expression in plasmacytoma and
Burkitt's lymphoma cells. Genes Dev 8, 2212-2226.
Makinen, N., Heinonen, H.R., Moore, S., Tomlinson, I.P., van der Spuy, Z.M., and Aaltonen,
L.A. (201la). MED 12 exon 2 mutations are common in uterine leiomyomas from South African
patients. Oncotarget 2, 966-969.
Makinen, N., Mehine, M., Tolvanen, J., Kaasinen, E., Li, Y., Lehtonen, H.J., Gentile, M., Yan,
J., Enge, M., Taipale, M., et al. (2011 b). MED12, the mediator complex subunit 12 gene, is
mutated at high frequency in uterine leiomyomas. Science 334, 252-255.
Malik, S., and Roeder, R.G. (2010). The metazoan Mediator co-activator complex as an
integrative hub for transcriptional regulation. Nat Rev Genet 11, 761-772.
Manolio, T.A. (2010). Genomewide association studies and assessment of the risk of disease. N
Engl J Med 363, 166-176.
Manzo, S.G., Zhou, Z.L., Wang, Y.Q., Marinello, J., He, J.X., Li, Y.C., Ding, J., Capranico, G.,
and Miao, Z.H. (2012). Natural product triptolide mediates cancer cell death by triggering
CDK7-dependent degradation of RNA polymerase II. Cancer Res 72, 5363-5373.
Matthews, K.S. (1992). DNA looping. Microbiol Rev 56, 123-136.
Maurano, M.T., Humbert, R., Rynes, E., Thurman, R.E., Haugen, E., Wang, H., Reynolds, A.P.,
Sandstrom, R., Qu, H., Brody, J., et al. (2012). Systematic localization of common diseaseassociated variation in regulatory DNA. Science 337, 1190-1195.
Melo, C.A., Drost, J., Wijchers, P.J., van de Werken, H., de Wit, E., Oude Vrielink, J.A., Elkon,
R., Melo, S.A., Leveille, N., Kalluri, R., et al. (2013). eRNAs are required for p53-dependent
enhancer activity and gene transcription. Mol Cell 49, 524-535.
Menzel, S., Jiang, J., Silver, N., Gallagher, J., Cunningham, J., Surdulescu, G., Lathrop, M.,
Farrall, M., Spector, T.D., and Thein, S.L. (2007). The HBS1L-MYB intergenic region on
chromosome 6q23.3 influences erythrocyte, platelet, and monocyte counts in humans. Blood
110, 3624-3626.
Mercola, M., Goverman, J., Mirell, C., and Calame, K. (1985). Immunoglobulin heavy-chain
enhancer requires one or more tissue-specific factors. Science 227, 266-270.
Merika, M., and Thanos, D. (2001). Enhanceosomes. Curr Opin Genet Dev 11, 205-208.
Mertz, J.A., Conery, A.R., Bryant, B.M., Sandy, P., Balasubramanian, S., Mele, D.A., Bergeron,
L., and Sims, R.J., 3rd (2011). Targeting MYC dependence in cancer by inhibiting BET
bromodomains. Proc Natl Acad Sci U S A 108, 16669-16674.
Mikkelson, G.M., Gonzalez, A., and Peterson, G.D. (2007). Economic inequality predicts
biodiversity loss. PLoS One 2, e444.
52
Moazed, D. (2009). Small RNAs in transcriptional gene silencing and genome defence. Nature
457, 413-420.
Morin, R.D., Mendez-Lago, M., Mungall, A.J., Goya, R., Mungall, K.L., Corbett, R.D., Johnson,
N.A., Severson, T.M., Chiu, R., Field, M., et al. (2011). Frequent mutation of histone-modifying
genes in non-Hodgkin lymphoma. Nature 476, 298-303.
Morris, E.J., Ji, J.Y., Yang, F., Di Stefano, L., Herr, A., Moon, N.S., Kwon, E.J., Haigis, K.M.,
Naar, A.M., and Dyson, N.J. (2008). E2Fl represses beta-catenin transcription and is
antagonized by both pRB and CDK8. Nature 455, 552-556.
Musselman, C.A., Avvakumov, N., Watanabe, R., Abraham, C.G., Lalonde, M.E., Hong, Z.,
Allen, C., Roy, S., Nunez, J.K., Nickoloff, J., et al. (2012). Molecular basis for H3K36me3
recognition by the Tudor domain of PHFl. Nat Struct Mol Biol 19, 1266-1272.
Nakayama, T., and Takami, Y. (2001). Participation of histones and histone-modifying enzymes
in cell functions through alterations in chromatin structure. J Biochem 129, 491-499.
Nesbit, C.E., Tersak, J.M., and Prochownik, E.V. (1999). MYC oncogenes and human neoplastic
disease. Oncogene 18, 3004-3016.
Ng, H.H., and Surani, M.A. (2011). The transcriptional and signalling networks of pluripotency.
Nat Cell Biol 13, 490-496.
Ng, S.Y., Bogu, G.K., Soh, B.S., and Stanton, L.W. (2013). The long noncoding RNA RMST
interacts with SOX2 to regulate neurogenesis. Mol Cell 51, 349-359.
Nie, Z., Hu, G., Wei, G., Cui, K., Yamane, A., Resch, W., Wang, R., Green, D.R., Tessarollo, L.,
Casellas, R., et al. (2012). c-Myc is a universal amplifier of expressed genes in lymphocytes and
embryonic stem cells. Cell 151, 68-79.
Ntini, E., Jarvelin, A.I., Bornholdt, J., Chen, Y., Boyd, M., Jorgensen, M., Andersson, R., Hoof,
I., Schein, A., Andersen, P.R., et al. (2013). Polyadenylation site-induced decay of upstream
transcripts enforces promoter directionality. Nat Struct Mol Biol 20, 923-928.
Nuinoon, M., Makarasara, W., Mushiroda, T., Setianingsih, I., Wahidiyat, P.A., Sripichai, 0.,
Kumasaka, N., Takahashi, A., Svasti, S., Munkongdee, T., et al. (2010). A genome-wide
association identified the common genetic variants influence disease severity in beta0thalassemia/hemoglobin E. Hum Genet 127, 303-314.
Ong, C.T., and Corces, V.G. (2011). Enhancer function: new insights into the regulation of
tissue-specific gene expression. Nat Rev Genet 12, 283-293.
Orkin, S.H., and Hochedlinger, K. (2011). Chromatin connections to pluripotency and cellular
reprogramming. Cell 145, 835-850.
53
Orom, U.A., Derrien, T., Beringer, M., Gumireddy, K., Gardini, A., Bussotti, G., Lai, F.,
Zytnicki, M., Notredame, C., Huang, Q., et al. (2010). Long noncoding RNAs with enhancer-like
function in human cells. Cell 143, 46-58.
Orom, U.A., and Shiekhattar, R. (2013). Long noncoding RNAs usher in a new era in the
biology of enhancers. Cell 154, 1190-1193.
Panne, D., Maniatis, T., and Harrison, S.C. (2007). An atomic model of the interferon-beta
enhanceosome. Cell 129, 1111-1123.
Park, S.L., Chang, S.C., Cai, L., Cordon-Cardo, C., Ding, B.G., Greenland, S., Hussain, S.K.,
Jiang, Q., Liu, S., Lu, M.L., et al. (2008). Associations between variants of the 8q24
chromosome and nine smoking-related cancer sites. Cancer Epidemiol Biomarkers Prev 17,
3193-3202.
Parker, S.C., Stitzel, M.L., Taylor, D.L., Orozco, J.M., Erdos, M.R., Akiyama, J.A., van Bueren,
K.L., Chines, P.S., Narisu, N., Black, B.L., et al. (2013). Chromatin stretch enhancer states drive
cell-specific gene regulation and harbor human disease risk variants. Proc Natl Acad Sci U S A
110, 17921-17926.
Parsons, D.W., Li, M., Zhang, X., Jones, S., Leary, R.J., Lin, J.C., Boca, S.M., Carter, H.,
Samayoa, J., Bettegowda, C., et al. (2011). The genetic landscape of the childhood cancer
medulloblastoma. Science 331, 435-439.
Polakis, P. (2000). Wnt signaling and cancer. Genes Dev 14, 1837-1851.
Pomerantz, M.M., Ahmadiyeh, N., Jia, L., Herman, P., Verzi, M.P., Doddapaneni, H., Beckwith,
C.A., Chan, J.A., Hills, A., Davis, M., et al. (2009). The 8q24 cancer risk variant rs6983267
shows long-range interaction with MYC in colorectal cancer. Nat Genet 41, 882-884.
Poynter, J.N., Figueiredo, J.C., Conti, D.V., Kennedy, K., Gallinger, S., Siegmund, K.D., Casey,
G., Thibodeau, S.N., Jenkins, M.A., Hopper, J.L., et al. (2007). Variants on 9p24 and 8q24 are
associated with risk of colorectal cancer: results from the Colon Cancer Family Registry. Cancer
Res 67, 11128-11132.
Preker, P., Almvig, K., Christensen, M.S., Valen, E., Mapendano, C.K., Sandelin, A., and
Jensen, T.H. (2011). PROMoter uPstream Transcripts share characteristics with mRNAs and are
produced upstream of all three major types of mammalian promoters. Nucleic Acids Res 39,
7179-7193.
Preker, P., Nielsen, J., Kammler, S., Lykke-Andersen, S., Christensen, M.S., Mapendano, C.K.,
Schierup, M.H., and Jensen, T.H. (2008). RNA exosome depletion reveals transcription upstream
of active human promoters. Science 322, 1851-1854.
Proudfoot, N.J. (2011). Ending the message: poly(A) signals then and now. Genes Dev 25, 17701782.
54
Ptashne, M. (1986). Gene regulation by proteins acting nearby and at a distance. Nature 322,
697-701.
Ptashne, M., and Gann, A. (1998). Imposing specificity by localization: mechanism and
evolvability. Curr Biol 8, R897.
Rada-Iglesias, A., Bajpai, R., Swigut, T., Brugmann, S.A., Flynn, R.A., and Wysocka, J. (2011).
A unique chromatin signature uncovers early developmental enhancers in humans. Nature 470,
279-283.
Rahl, P.B., Lin, C.Y., Seila, A.C., Flynn, R.A., McCuine, S., Burge, C.B., Sharp, P.A., and
Young, R.A. (2010). c-Myc regulates transcriptional pause release. Cell 141, 432-445.
Rahman, S., Sowa, M.E., Ottinger, M., Smith, J.A., Shi, Y., Harper, J.W., and Howley, P.M.
(2011). The Brd4 extraterminal domain confers transcription activation independent of pTEFb by
recruiting multiple proteins, including NSD3. Mol Cell Biol 31, 2641-2652.
Reyes-Turcu, F.E., and Grewal, S.I. (2012). Different means, same end-heterochromatin
formation by RNAi and RNAi-independent RNA processing factors in fission yeast. Curr Opin
Genet Dev 22, 156-163.
Rhee, H.S., and Pugh, B.F. (2011). Comprehensive genome-wide protein-DNA interactions
detected at single-nucleotide resolution. Cell 147, 1408-1419.
Roessler, E., Belloni, E., Gaudenz, K., Jay, P., Berta, P., Scherer, S.W., Tsui, L.C., and Muenke,
M. (1996). Mutations in the human Sonic Hedgehog gene cause holoprosencephaly. Nat Genet
14, 357-360.
Roessler, E., Belloni, E., Gaudenz, K., Vargas, F., Scherer, S.W., Tsui, L.C., and Muenke, M.
(1997). Mutations in the C-terminal domain of Sonic Hedgehog cause holoprosencephaly. Hum
Mol Genet 6, 1847-1853.
Saiz, L., and Vilar, J.M. (2006). DNA looping: the consequences and its control. Curr Opin
Struct Biol 16, 344-350.
Sanda, T., Lawton, L.N., Barrasa, M.I., Fan, Z.P., Kohlhammer, H., Gutierrez, A., Ma, W.,
Tatarek, J., Ahn, Y., Kelliher, M.A., et al. (2012). Core transcriptional regulatory circuit
controlled by the TAL1 complex in human T cell acute lymphoblastic leukemia. Cancer Cell 22,
209-221.
Sankaran, V.G., Menne, T.F., Xu, J., Akie, T.E., Lettre, G., Van Handel, B., Mikkola, H.K.,
Hirschhorn, J.N., Cantor, A.B., and Orkin, S.H. (2008). Human fetal hemoglobin expression is
regulated by the developmental stage-specific repressor BCL1 IA. Science 322, 1839-1842.
Sankaran, V.G., Xu, J., Ragoczy, T., Ippolito, G.C., Walkley, C.R., Maika, S.D., Fujiwara, Y.,
Ito, M., Groudine, M., Bender, M.A., et al. (2009). Developmental and species-divergent globin
switching are driven by BCL1 IA. Nature 460, 1093-1097.
55
Sassone-Corsi, P., Dougherty, J.P., Wasylyk, B., and Chambon, P. (1984). Stimulation of in vitro
transcription from heterologous promoters by the simian virus 40 enhancer. Proc Natl Acad Sci
U S A 81, 308-312.
Saunders, A., Core, L.J., and Lis, J.T. (2006). Breaking barriers to transcription elongation. Nat
Rev Mol Cell Biol 7, 557-567.
Schaub, M.A., Boyle, A.P., Kundaje, A., Batzoglou, S., and Snyder, M. (2012). Linking disease
associations with regulatory information in the human genome. Genome Res 22, 1748-1759.
Schirm, S., Weber, F., Schaffner, W., and Fleckenstein, B. (1985). A transcription enhancer in
the Herpesvirus saimiri genome. EMBO J 4, 2669-2674.
Schleif, R. (1992). DNA looping. Annu Rev Biochem 61, 199-223.
Schodel, J., Bardella, C., Sciesielski, L.K., Brown, J.M., Pugh, C.W., Buckle, V., Tomlinson,
I.P., Ratcliffe, P.J., and Mole, D.R. (2012). Common genetic variants at the 11q13.3 renal cancer
susceptibility locus influence binding of HIF to an enhancer of cyclin DI expression. Nat Genet
44, 420-425, S421-422.
Scholer, H.R., and Gruss, P. (1985). Cell type-specific transcriptional enhancement in vitro
requires the presence of trans-acting factors. EMBO J 4, 3005-3013.
Seila, A.C., Calabrese, J.M., Levine, S.S., Yeo, G.W., Rahl, P.B., Flynn, R.A., Young, R.A., and
Sharp, P.A. (2008). Divergent transcription from active promoters. Science 322, 1849-1851.
Sen, R., and Baltimore, D. (1986). Multiple nuclear factors interact with the immunoglobulin
enhancer sequences. Cell 46, 705-716.
Sergeant, A., Bohmann, D., Zentgraf, H., Weiher, H., and Keller, W. (1984). A transcription
enhancer acts in vitro over distances of hundreds of base-pairs on both circular and linear
templates but not on chromatin-reconstituted DNA. J Mol Biol 180, 577-600.
Shandilya, J., and Roberts, S.G. (2012). The transcription cycle in eukaryotes: from productive
initiation to RNA polymerase II recycling. Biochim Biophys Acta 1819, 391400.
Shen, Y., Yue, F., McCleary, D.F., Ye, Z., Edsall, L., Kuan, S., Wagner, U., Dixon, J., Lee, L.,
Lobanenkov, V.V., et al. (2012). A map of the cis-regulatory sequences in the mouse genome.
Nature 488, 116-120.
Shete, S., Hosking, F.J., Robertson, L.B., Dobbins, S.E., Sanson, M., Malmer, B., Simon, M.,
Marie, Y., Boisselier, B., Delattre, J.Y., et al. (2009). Genome-wide association study identifies
five susceptibility loci for glioma. Nat Genet 41, 899-904.
Shi, J., Whyte, W.A., Zepeda-Mendoza, C.J., Milazzo, J.P., Shen, C., Roe, J.S., Minder, J.L.,
Mercan, F., Wang, E., Eckersley-Maslin, M.A., et al. (2013). Role of SWI/SNF in acute
leukemia maintenance and enhancer-mediated Myc regulation. Genes Dev 27, 2648-2662.
56
Shlyueva, D., Stampfel, G., and Stark, A. (2014). Transcriptional enhancers: from properties to
genome-wide predictions. Nat Rev Genet 15, 272-286.
Shou, Y., Martelli, M.L., Gabrea, A., Qi, Y., Brents, L.A., Roschke, A., Dewald, G., Kirsch, I.R.,
Bergsagel, P.L., and Kuehl, W.M. (2000). Diverse karyotypic abnormalities of the c-myc locus
associated with c-myc dysregulation and tumor progression in multiple myeloma. Proc Natl
Acad Sci U S A 97, 228-233.
Sigova, A.A., Mullen, A.C., Molinie, B., Gupta, S., Orlando, D.A., Guenther, M.G., Almada,
A.E., Lin, C., Sharp, P.A., Giallourakis, C.C., et al. (2013). Divergent transcription of long
noncoding RNA/mRNA gene pairs in embryonic stem cells. Proc Natl Acad Sci U S A 110,
2876-2881.
Simon, J.M., Giresi, P.G., Davis, I.J., and Lieb, J.D. (2012). Using formaldehyde-assisted
isolation of regulatory elements (FAIRE) to isolate active regulatory DNA. Nat Protoc 7, 256267.
Smith, E., Lin, C., and Shilatifard, A. (2011). The super elongation complex (SEC) and MLL in
development and disease. Genes Dev 25, 661-672.
Solovieff, N., Milton, J.N., Hartley, S.W., Sherva, R., Sebastiani, P., Dworkis, D.A., Klings,
E.S., Farrer, L.A., Garrett, M.E., Ashley-Koch, A., et al. (2010). Fetal hemoglobin in sickle cell
anemia: genome-wide association studies suggest a regulatory region in the 5' olfactory receptor
gene cluster. Blood 115, 1815-1822.
Song, L., and Crawford, G.E. (2010). DNase-seq: a high-resolution technique for mapping active
gene regulatory elements across the genome from mammalian cells. Cold Spring Harb Protoc
2010, pdb prot5384.
Spandidos, D.A., and Wilkie, N.M. (1983). Host-specificities of papillomavirus, Moloney
murine sarcoma virus and simian virus 40 enhancer sequences. EMBO J 2, 1193-1199.
Spitz, F., and Furlong, E.E. (2012). Transcription factors: from enhancer binding to
developmental control. Nat Rev Genet 13, 613-626.
Stellrecht, C.M., and Chen, L.S. (2011). Transcription inhibition as a therapeutic target for
cancer. Cancers (Basel) 3, 4170-4190.
Stranger, B.E., Stahl, E.A., and Raj, T. (2011). Progress and promise of genome-wide association
studies for human complex trait genetics. Genetics 187, 367-383.
Su, W., Porter, S., Kustu, S., and Echols, H. (1990). DNA-looping and enhancer activity:
association between DNA-bound NtrC activator and RNA polymerase at the bacterial glnA
promoter. Proc Natl Acad Sci U S A 87, 5504-5508.
Sur, I., Tuupanen, S., Whitington, T., Aaltonen, L.A., and Taipale, J. (2013). Lessons from
functional analysis of genome-wide association studies. Cancer Res 73, 4180-4184.
57
Sur, I.K., Hallikas, 0., Vaharautio, A., Yan, J., Turunen, M., Enge, M., Taipale, M., Karhu, A.,
Aaltonen, L.A., and Taipale, J. (2012). Mice lacking a Myc enhancer that includes human SNP
rs6983267 are resistant to intestinal tumors. Science 338, 1360-1363.
Taatjes, D.J. (2010). The human Mediator complex: a versatile, genome-wide regulator of
transcription. Trends Biochem Sci 35, 315-322.
Tan-Wong, S.M., Zaugg, J.B., Camblong, J., Xu, Z., Zhang, D.W., Mischo, H.E., Ansari, A.Z.,
Luscombe, N.M., Steinmetz, L.M., and Proudfoot, N.J. (2012). Gene loops enhance
transcriptional directionality. Science 338, 671-675.
Tenesa, A., Farrington, S.M., Prendergast, J.G., Porteous, M.E., Walker, M., Haq, N., Barnetson,
R.A., Theodoratou, E., Cetnarskyj, R., Cartwright, N., et al. (2008). Genome-wide association
scan identifies a colorectal cancer susceptibility locus on 1 q23 and replicates risk loci at 8q24
and 18q21. Nat Genet 40, 63 1-637.
Thanos, D., and Maniatis, T. (1995). Virus induction of human IFN beta gene expression
requires the assembly of an enhanceosome. Cell 83, 1091-1100.
Titov, D.V., Gilman, B., He, Q.L., Bhat, S., Low, W.K., Dang, Y., Smeaton, M., Demain, A.L.,
Miller, P.S., Kugel, J.F., et al. (2011). XPB, a subunit of TFIIH, is a target of the natural product
triptolide. Nat Chem Biol 7, 182-188.
Tolhuis, B., Palstra, R.J., Splinter, E., Grosveld, F., and de Laat, W. (2002). Looping and
interaction between hypersensitive sites in the active beta-globin locus. Mol Cell 10, 1453-1465.
Tomlinson, I.P., Webb, E., Carvajal-Carmona, L., Broderick, P., Howarth, K., Pittman, A.M.,
Spain, S., Lubbe, S., Walther, A., Sullivan, K., et al. (2008). A genome-wide association study
identifies colorectal cancer susceptibility loci on chromosomes IOp14 and 8q23.3. Nat Genet 40,
623-630.
Trudel, M., and Costantini, F. (1987). A 3' enhancer contributes to the stage-specific expression
of the human beta-globin gene. Genes Dev 1, 954-961.
Turnbull, C., Ahmed, S., Morrison, J., Pernet, D., Renwick, A., Maranian, M., Seal, S.,
Ghoussaini, M., Hines, S., Healey, C.S., el al. (2010). Genome-wide association study identifies
five new breast cancer susceptibility loci. Nat Genet 42, 504-507.
Tuupanen, S., Turunen, M., Lehtonen, R., Hallikas, 0., Vanharanta, S., Kivioja, T., Bjorklund,
M., Wei, G., Yan, J., Niittymaki, I., et al. (2009). The common colorectal cancer predisposition
SNP rs6983267 at chromosome 8q24 confers potential to enhanced Wnt signaling. Nat Genet 41,
885-890.
Vakoc, C.R., Letting, D.L., Gheldof, N., Sawado, T., Bender, M.A., Groudine, M., Weiss, M.J.,
Dekker, J., and Blobel, G.A. (2005). Proximity among distant regulatory elements at the betaglobin locus requires GATA-1 and FOG-1. Mol Cell 17, 453-462.
58
Villicana, C., Cruz, G., and Zurita, M. (2014). The basal transcription machinery as a target for
cancer therapy. Cancer Cell Int 14, 18.
Visel, A., Blow, M.J., Li, Z., Zhang, T., Akiyama, J.A., Holt, A., Plajzer-Frick, I., Shoukry, M.,
Wright, C., Chen, F., et al. (2009). ChIP-seq accurately predicts tissue-specific activity of
enhancers. Nature 457, 854-858.
Vispe, S., DeVries, L., Creancier, L., Besse, J., Breand, S., Hobson, D.J., Svejstrup, J.Q.,
Annereau, J.P., Cussac, D., Dumontet, C., et al. (2009). Triptolide is an inhibitor of RNA
polymerase I and II-dependent transcription leading predominantly to down-regulation of shortlived mRNA. Mol Cancer Ther 8, 2780-2790.
Wall, L., deBoer, E., and Grosveld, F. (1988). The human beta-globin gene 3' enhancer contains
multiple binding sites for an erythroid-specific protein. Genes Dev 2, 1089-1100.
Wamstad, J.A., Alexander, J.M., Truty, R.M., Shrikumar, A., Li, F., Eilertson, K.E., Ding, H.,
Wylie, J.N., Pico, A.R., Capra, J.A., et al. (2012). Dynamic and coordinated epigenetic
regulation of developmental transitions in the cardiac lineage. Cell 151, 206-220.
Wang, A., Kurdistani, S.K., and Grunstein, M. (2002). Requirement of Hos2 histone deacetylase
for gene activity in yeast. Science 298, 1412-1414.
Wang, H., Zang, C., Taing, L., Arnett, K.L., Wong, Y.J., Pear, W.S., Blacklow, S.C., Liu, X.S.,
and Aster, J.C. (2014). NOTCH 1-RBPJ complexes drive target gene expression through dynamic
interactions with superenhancers. Proc Natl Acad Sci U S A 111, 705-710.
Wang, Z., Zang, C., Cui, K., Schones, D.E., Barski, A., Peng, W., and Zhao, K. (2009). Genomewide mapping of HATs and HDACs reveals distinct functions in active and inactive genes. Cell
138, 1019-1031.
Watanabe, Y., Castoro, R.J., Kim, H.S., North, B., Oikawa, R., Hiraishi, T., Ahmed, S.S.,
Chung, W., Cho, M.Y., Toyota, M., et al. (2011). Frequent alteration of MLL3 frameshift
mutations in microsatellite deficient colorectal cancer. PLoS One 6, e23320.
Wesierska-Gadek, J., and Krystof, V. (2009). Selective cyclin-dependent kinase inhibitors
discriminating between cell cycle and transcriptional kinases: future reality or utopia? Ann N Y
Acad Sci 1171, 228-241.
Whitehouse, I., Rando, O.J., Delrow, J., and Tsukiyama, T. (2007). Chromatin remodelling at
promoters suppresses antisense transcription. Nature 450, 1031-1035.
Whyte, W.A., Bilodeau, S., Orlando, D.A., Hoke, H.A., Frampton, G.M., Foster, C.T., Cowley,
S.M., and Young, R.A. (2012). Enhancer decommissioning by LSD1 during embryonic stem cell
differentiation. Nature 482, 221-225.
Whyte, W.A., Orlando, D.A., Hnisz, D., Abraham, B.J., Lin, C.Y., Kagey, M.H., Rahl, P.B., Lee,
T.I., and Young, R.A. (2013). Master transcription factors and mediator establish superenhancers at key cell identity genes. Cell 153, 307-319.
59
Wildeman, A.G., Sassone-Corsi, P., Grundstrom, T., Zenke, M., and Chambon, P. (1984).
Stimulation of in vitro transcription from the SV40 early promoter by the enhancer involves a
specific trans-acting factor. EMBO J 3, 3129-3133.
Wu, F., and Yao, J. (2013). Spatial compartmentalization at the nuclear periphery characterized
by genome-wide mapping. BMC Genomics 14, 591.
Wu, S.Y., and Chiang, C.M. (2007). The double bromodomain-containing chromatin adaptor
Brd4 and transcriptional regulation. J Biol Chem 282, 13141-13145.
Wu, X., and Sharp, P.A. (2013). Divergent transcription: a driving force for new gene
origination? Cell 155, 990-996.
Xie, W., Schultz, M.D., Lister, R., Hou, Z., Rajagopal, N., Ray, P., Whitaker, J.W., Tian, S.,
Hawkins, R.D., Leung, D., et al. (2013). Epigenomic analysis of multilineage differentiation of
human embryonic stem cells. Cell 153, 1134-1148.
Xu, J., Bauer, D.E., Kerenyi, M.A., Vo, T.D., Hou, S., Hsu, Y.J., Yao, H., Trowbridge, J.J.,
Mandel, G., and Orkin, S.H. (2013). Corepressor-dependent silencing of fetal hemoglobin
expression by BCL1 IA. Proc Natl Acad Sci U S A 110, 6518-6523.
Xu, J., Sankaran, V.G., Ni, M., Menne, T.F., Puram, R.V., Kim, W., and Orkin, S.H. (2010).
Transcriptional silencing of {gamma}-globin by BCL lIA involves long-range interactions and
cooperation with SOX6. Genes Dev 24, 783-798.
Yamaguchi, Y., Shibata, H., and Handa, H. (2013). Transcription elongation factors DSIF and
NELF: promoter-proximal pausing and beyond. Biochim Biophys Acta 1829, 98-104.
Yang, Z., Yik, J.H., Chen, R., He, N., Jang, M.K., Ozato, K., and Zhou, Q. (2005). Recruitment
of P-TEFb for stimulation of transcriptional elongation by the bromodomain protein Brd4. Mol
Cell 19, 535-545.
Yeager, M., Chatterjee, N., Ciampa, J., Jacobs, K.B., Gonzalez-Bosquet, J., Hayes, R.B., Kraft,
P., Wacholder, S., Orr, N., Berndt, S., et al. (2009). Identification of a new prostate cancer
susceptibility locus on chromosome 8q24. Nat Genet 41, 1055-1057.
Yew, P.Y., Mushiroda, T., Kiyotani, K., Govindasamy, G.K., Yap, L.F., Teo, S.H., Lim, P.V.,
Govindaraju, S., Ratnavelu, K., Sam, C.K., et al. (2012). Identification of a functional variant in
SPLUNC 1 associated with nasopharyngeal carcinoma susceptibility among Malaysian Chinese.
Mol Carcinog 51 Suppl 1, E74-82.
Young, R.A. (2011). Control of the embryonic stem cell state. Cell 144, 940-954.
Zanke, B.W., Greenwood, C.M., Rangrej, J., Kustra, R., Tenesa, A., Farrington, S.M.,
Prendergast, J., Olschwang, S., Chiang, T., Crowdy, E., et al. (2007). Genome-wide association
scan identifies a colorectal cancer susceptibility locus on chromosome 8q24. Nat Genet 39, 989994.
60
Zentner, G.E., Tesar, P.J., and Scacheri, P.C. (2011). Epigenetic signatures distinguish multiple
classes of enhancers with distinct cellular functions. Genome Res 21, 1273-1283.
Zhang, Y., Chen, A., Yan, X.M., and Huang, G. (2012). Disordered epigenetic regulation in
MLL-related leukemia. Int J Hematol 96, 428-437.
Zhang, Y., Wong, C.H., Birnbaum, R.Y., Li, G., Favaro, R., Ngan, C.Y., Lim, J., Tai, E., Poh,
H.M., Wong, E., et al. (2013). Chromatin connectivity maps reveal dynamic promoter-enhancer
long-range associations. Nature 504, 306-3 10.
Zhang, Y., Yi, P., Chen, W., Ming, J., Zhu, B., Li, Z., Shen, N., Shi, W., Ke, J., Zhao, Q., et al.
(2014). Association between polymorphisms within the susceptibility region 8q24 and breast
cancer in a Chinese population. Tumour Biol 35, 2649-2654.
Zuber, J., Shi, J., Wang, E., Rappaport, A.R., Herrmann, H., Sison, E.A., Magoon, D., Qi, J.,
Blatt, K., Wunderlich, M., et al. (2011). RNAi screen identifies Brd4 as a therapeutic target in
acute myeloid leukaemia. Nature 478, 524-528.
61
Chapter 2
Selective Inhibition of Tumor Oncogenes by Disruption of Super-enhancers
Jakob Loven*, Heather A. Hoke' 2 *, Charles Y. Lin' 3 5 *, Ashley Lau' , David A. Orlando',
Christopher R. Vakoc 4, James E. Bradner 5,6, Tong Ihn Lee', and Richard A. Young' 2
Whitehead Institutefor Biomedical Research, 9 Cambridge Center, Cambridge,MA 02142,
2Department ofBiology, MIT, Cambridge,MA
021399,3 Computationaland Systems Biology
Program,MassachusettsInstitute of Technology, Cambridge,MA 02139, 4 Cold Spring Harbor
Laboratory, 1 Bungiown Road, Cold Spring Harbor,New York 11724, USA, SDepartment of
Medical Oncology, Dana-FarberCancerInstitute, 44 Binney Street, Boston, MA 02115, USA,
6Departmentof Medicine, HarvardMedical School, 25 Shattuck Street, Boston, MA 02115, USA
* These authors contributed equally
This chapter originally appeared in Cell 2013 vol. 153 (2) pp. 320-334. Corresponding
Supplementary Material is appended. Supplementary tables and data are available online at
http://dx.doi.org/10.1016/j.cell.2013.03.036.
62
Personal Contribution to the Project
This work was a close collaboration between Jakob Loven, Charles Y. Lin and myself. I
performed the majority of experiments, with assistance from Jakob Loven and Ashley Lau.
Charles Y. Lin performed the analysis. I wrote the manuscript in collaboration with Jakob
Loven, Charles Y. Lin, and Richard A. Young.
63
Summary
Chromatin regulators have become attractive targets for cancer therapy, but it is
unclear why inhibition of these ubiquitous regulators should have gene-specific effects in
tumor cells. Here we investigate how inhibition of the widely-expressed transcriptional
coactivator BRD4 leads to selective inhibition of the MYC oncogene in multiple myeloma
(MM). BRD4 and Mediator were found to co-occupy thousands of enhancers associated
with active genes. They also co-occupied a small set of exceptionally large super-enhancers
associated with genes that feature prominently in MM biology, including the MYC
oncogene. Treatment of MM tumor cells with the BET-bromodomain inhibitor JQ1 led to
preferential loss of BRD4 at super-enhancers and consequent transcription elongation
defects that preferentially impacted genes with super-enhancers, including MYC. Superenhancers were found at key oncogenic drivers in many other tumor cells. These
observations have implications for the discovery of novel cancer therapeutics directed at
components of super-enhancers in diverse tumor types.
64
Control
BRD4 inhibited
Growth arrest
Tumor cells
Typical enhancer driven gene
Coactivators:
BRD4, Mediator
Enhancer elements
Super-enhancer driven gene
M Amp&
WW
"W
=F WE
~.Iuuu
Graphical Abstract
65
Introduction
Chromatin regulators are attractive as therapeutic targets for cancer because they are
deregulated in numerous cancers (Baylin and Jones, 2011; Elsasser et al., 2011; Esteller, 2008;
Feinberg and Tycko, 2004; You and Jones, 2012) and amenable to small molecule inhibition
(Cole, 2008; Dawson and Kouzarides, 2012; Geutjes et al., 2012). Inhibition of some chromatin
regulators has already proven to be efficacious for treatment of certain cancers (Issa and
Kantarjian, 2009; Marks and Xu, 2009). Most chromatin regulators, however, are expressed in a
broad range of healthy cells and contribute generally to gene expression, so inhibition of these
important genome-associated proteins might be expected to adversely affect global gene
expression in healthy cells and thus produce highly toxic effects. Nonetheless, inhibitors of
some chromatin regulators, such as BRD4, have been shown to selectively inhibit transcription
of key oncogenic drivers such as c-MYC (hereafter referred to as MYC) in multiple tumor types
(Dawson et al., 2011; Delmore et al., 2011; Mertz et al., 2011; Zuber et al., 2011). It is important
to understand how inhibition of a widely expressed, general regulator such as BRD4 can exert a
selective effect on the expression of a small number of genes in specific cells.
BRD4 is a member of the bromodomain and extraterminal (BET) subfamily of human
bromodomain proteins, which include BRDT, BRD2, BRD3, and BRD4. These proteins
associate with acetylated chromatin and facilitate transcriptional activation (LeRoy et al., 2008;
Rahman et al., 2011). BRD4 was first identified as an interaction partner of the murine Mediator
coactivator complex (Jiang et al., 1998), and was subsequently shown to associate with Mediator
in a variety of human cells (Dawson et al., 2011; Wu and Chiang, 2007). BRD4 is involved in
the control of transcriptional elongation by RNA polymerase II (RNA Pol II) through its
recruitment of the positive transcription elongation factor P-TEFb (Jang et al., 2005; Yang et al.,
66
2005). Almost all human cells express the BRD4 gene, based on analysis of human tissue
expression data across 90 distinct tissue types (Human body index - transcriptional profiling, see
Supplementary Methods), and BRD4 is found associated with a large population of active genes
in CD4+ T cells (Zhang et al., 2012). It is not yet clear whether the BRD4 protein is generally
involved in the transcription of active genes in tumor cells or if it is selectively associated with a
subset of these genes.
Two recently developed bromodomain inhibitors, JQ 1 and iBET, selectively bind to the
amino-terminal twin bromodomains of BRD4 (Filippakopoulos et al., 2010; Nicodeme et al.,
2010). These BET inhibitors cause selective repression of the potent MYC oncogene in a range
of tumors, including multiple myeloma (MM), Burkitt's lymphoma (BL), acute myeloid
leukemia (AML), and acute lymphoblastic leukemia (ALL) (Dawson et al., 2011; Delmore et al.,
2011; Mertz et al., 2011; Ott et al., 2012; Zuber et al., 2011). The inhibition of MYC apparently
occurs as a consequence of BRD4 depletion at the enhancers that drive MYC expression
(Delmore et al., 2011). Although BRD4 is widely expressed in mouse tissues, mice are
reasonably tolerant of the levels of BET bromodomain inhibition that inhibit certain tumors in
mouse models (Dawson et al., 2011; Delmore et al., 2011; Filippakopoulos et al., 2010; Mertz et
al., 2011; Zuber et al., 2011).
The multiple myeloma cell line (MM 1.S) used to study the effects of JQ 1 has an IgHMYC rearrangement and MYC gene expression is driven by factors associated with the IgH
enhancer (Dib et al., 2008; Shou et al., 2000). Enhancers function through co-operative and
synergistic interactions between multiple transcription factors and coactivators (Carey et al.,
1990; Giese et al., 1995; Kim and Maniatis, 1997; Thanos and Maniatis, 1995). Cooperative
binding and synergistic activation confer increased sensitivity, so that small changes in activator
67
concentration can lead to dramatic changes in activator binding and transcription of associated
genes (Carey, 1998). Furthermore, enhancers with large numbers of transcription factor binding
sites can be more sensitive to small changes in factor concentration than those with smaller
numbers of binding sites (Giniger and Ptashne, 1988; Griggs and Johnston, 1991). This concept
led us to postulate that some feature of the IgH enhancer might account for the selective effect of
BRD4 inhibition.
We show here that BRD4 and Mediator are associated with most active enhancers and
promoters in MM 1.S tumor cells, but exceptionally high levels of these cofactors occur at a
small set of large enhancer regions, which we call super-enhancers. Super-enhancers are
associated with MYC and other key genes that feature prominently in the biology of MM,
including many lineage-specific survival genes. Treatment of MM tumor cells with the BRD4
inhibitor JQ 1 caused a preferential loss of BRD4, Mediator and P-TEFb at super-enhancers and
preferential loss of transcription at super-enhancer-associated genes, including the MYC
oncogene. Tumor cell addiction to high-level expression of these oncogenes may then contribute
to their vulnerability to super-enhancer disruption (Chin et al., 1999; Felsher and Bishop, 1999;
Jain et al., 2002; Weinstein, 2002). We find super-enhancers in additional tumor types, where
they are similarly associated with key oncogenes. Thus, key oncogene drivers of tumor cells are
regulated by super-enhancers, which can confer disproportionate sensitivity to loss of the BRD4
coactivator and thus cause selective inhibition of transcription.
68
Results
BRD4 and Mediator co-occupy promoters of active genes in multiple myeloma
Transcription factors bind to enhancers and recruit the Mediator coactivator, which in
turn becomes associated with RNA polymerase II at the transcription start site, thus forming
DNA loops between enhancers and core promoters (Kagey et al., 2010). BRD4 is known to
associate with Mediator in some mammalian cells (Dawson et al., 2011; Jiang et al., 1998; Wu et
al., 2003). To identify active promoter and enhancer elements and determine how BRD4 and
Mediator occupy the genome in MM 1.S multiple myeloma cells, we used chromatin
immunoprecipitation coupled to high-throughput sequencing (ChIP-Seq) with antibodies against
the Mediator subunit MED 1, BRD4, the enhancer-associated histone modification H3K27Ac,
and the transcription start site (TSS)-associated histone modification H3K4Me3 (Figure 1).
ChIP-Seq signals for both Mediator and the histone modification H3K27Ac have previously
been shown to occur at both enhancers and TSSs (Creyghton et al., 2010; Heintzman et al., 2009;
Rada-Iglesias et al., 2011), and enhancers can be distinguished from TSSs by the absence of TSS
annotation and relatively low levels of H3K4Me3. We found that BRD4 co-occupied enhancers
and TSSs with MEDI throughout the genome (Figure IA and IB) and that the levels of BRD4
and MED 1 were strongly correlated (Figure Sl).
To confirm that BRD4 and Mediator are generally associated with active genes in
MM 1.S cells, we compared the ChIP-Seq data for these regulators with that for RNA Pol II and
the histone modification H3K4Me3. The levels of BRD4 and Mediator correlated with the levels
of RNA Pol II genome-wide (Figure I C). Signals for BRD4 and Mediator were found together
with those for the histone modification H3K4Me3 and RNA Pol II at -10,000 annotated TSSs
and these were considered active TSSs (Table S 1). Signals for BRD4 and the enhancer69
associated histone modification H3K27Ac were found in -8,000 Mediator-occupied regions
either lacking TSSs or extending beyond the immediate vicinity of the TSS, and these were
considered enhancer regions (Table S2, Data S1, Supplementary Methods).
70
A
Enhancer:
6.0
10kb
ChIP-Seq
E.
6.0]
L
1CAhII
I
6.01
MED1
BRD4
-A
A
H3K27Ac
i
6.01
H3K4Me3
chrl9:10,945,355
chrl9:10,904,705
SMARCA4
B
Genome-wide
4.0
avera ge
MED1
4.0-
BRD4
4.0
H3K27Ac
H3K4Me3
4.0
-2.5kb
C
Enhancer +2.5kb
Center
-2.5kb
TSS
+2.5kb
- 10.0
3.0-
- 5.0
1.5 MED1
0
3
BRD4
A 1f%
Genes ranked by
RNA Pol11 levels (rpm/bp)
Figure 1
71
Figure 1: Mediator and BRD4 co-occupy promoters of active genes in multiple myeloma
A) Gene tracks of MED1, BRD4, H3K27Ac, and H3K4Me3 ChIP-Seq occupancy at the
enhancer (left) and promoter (right) of SMARCA4 in MMl.S multiple myeloma cells. The x-axis
shows genomic position and enhancer containing regions are depicted with a white box. The yaxis shows signal of ChIP-Seq occupancy in units of reads per million mapped reads per base
pair (rpm/bp).
B) Meta-gene representation of global MED 1, BRD4, H3K27Ac, and H3K4Me3 occupancy at
enhancers and promoters. The x-axis shows the +/- 2.5kb region flanking either the center of
enhancer regions (left) or the TSS of active genes (right). The y-axis shows the average
background subtracted ChIP-Seq signal in units of rpm/bp.
C) Median MEDI and BRD4 levels in the +/- Ikb region around the TSSs of actively transcribed
genes ranked by increasing RNA Polymerase II (RNA Pol II) occupancy in MML.S cells. Levels
are in units of rpm/bp with the left y-axis showing levels of MED 1 and the right y-axis showing
levels of BRD4. Promoters were binned (50/bin) and a smoothing function was applied to
median levels. See also Figure 51.
72
Super-enhancers are associated with key multiple myeloma genes
Further analysis of the -8,000 enhancer regions revealed that the MED1 signal at 308
enhancers was significantly greater than at all other enhancers and promoters (Figure 2A; Figure
S2A; Table S2). These 308 "super-enhancers" differed from typical enhancers in both size and
Mediator levels (Figure 2B). Remarkably, -40% of all enhancer-bound Mediator and BRD4
occupied these 308 super-enhancers. While the typical enhancer had a median size of 1.3kb, the
super-enhancers had a median size of 19.4kb. These super-enhancers were thus 15-fold larger
than typical enhancers and were occupied, based on ChIP-Seq signal, by 18-fold more Mediator
and 16-fold more BRD4. Similarly high levels of H3K27Ac were observed in these large regions
(Figure 2B). Examples of gene tracks showing super-enhancers at either end of the spectrum of
Mediator occupancy (Figure 2A) are shown in Figure 2C. The largest super-enhancer was found
associated with the IGLL5 gene, which encodes an immunoglobulin lambda peptide expressed at
high levels in these cells.
We next sought to identify the complete set of MM l.S genes that are most likely
associated with super-enhancers. Enhancers tend to loop to and associate with adjacent genes in
order to activate their transcription (Gondor and Ohlsson, 2009; Lelli et al., 2012; Ong and
Corces, 2011; Spitz and Furlong, 2012). Most of these interactions occur within a distance of
-50kb of the enhancer (Chepelev et al., 2012). Using a simple proximity rule, we assigned all
transcriptionally active genes (TSSs) to super-enhancers within a 50kb window, a method shown
to identify a large proportion of true enhancer/promoter interactions in embryonic stem cells
(Dixon et al., 2012). This identified 681 genes associated with super-enhancers (Supplemental
Table S3), and 307 of these had a super-enhancer overlapping a portion of the gene, as shown for
CCND2 in Figure 2C.
73
Super-enhancer-associated genes were generally expressed at higher levels than genes
with typical enhancers and tended to be specifically expressed in MML.S cells (Figure 2D). To
test whether components of super-enhancers confer stronger activity compared to typical
enhancers, we cloned representative super-enhancer or typical enhancer fragments of similar size
into luciferase reporter constructs and transfected these into MML.S cells. Cloned sequence
fragments from super-enhancers generated 2-3 fold higher luciferase activity compared to typical
enhancers of similar size (Figure 2E; Supplemental Methods). These results are consistent with
the notion that super-enhancers help activate high levels of transcription of key genes that
regulate and enforce the MM 1.S cancer cell state.
The super-enhancer-associated genes included most genes that have previously been
shown to have important roles in multiple myeloma biology, including MYC, IRF4, PRDM1/
BLIMP-] and XBPJ (Figure 3A). MYC is a key oncogenic driver in MM (Chng et al., 2011; Dib
et al., 2008; Holien et al., 2012; Shou et al., 2000) and the MMl.S MYC locus contains a
chromosomal rearrangement that places MYC under the control of the IgH enhancer, which
qualifies as a super-enhancer in MM l.S cells. The IRF4 gene encodes a key plasma cell
transcription factor that is frequently deregulated in MM (Shaffer et al., 2008). PRDMJ/BLIMP1 encodes a transcription factor considered a master regulator of plasma cell development, and is
required for the formation of plasma cell tumors in a mouse model (Shapiro-Shelef et al., 2003;
Turner et al., 1994). XBP1 encodes a basic-region leucine zipper (bZIP) transcription factor of
the CREB-ATF family that governs plasma cell differentiation (Reimold et al., 2001). XBPJ is
frequently overexpressed in human MM and can drive the development of MM in a mouse
model (Carrasco et al., 2007; Claudio et al., 2002).
74
Super-enhancers were associated with many additional genes that have important roles in
cancer pathogenesis more generally (Figure 3B). Cyclin D2 (CCND2) is deregulated in many
human cancers, including MM (Bergsagel et al., 2005; Musgrove et al., 2011). The PIMI kinase
has been implicated in the biology of many different cancers (Shah et al., 2008). MCLJ and
BCL-xL, members of the BCL-2 family of apoptosis regulators, are frequently deregulated in
cancer, promoting cell survival and chemoresistance (Beroukhim et al., 2010). We conclude that
super-enhancers are frequently associated with genes that feature prominently in the biology of
multiple myeloma and other human cancers.
75
A
D
500,000 400,000
in
IGLL5
-
IRF4
XBP1
MYC
CCNO2
200,000 100,000
-,-.
04
t
-
"
6
-2
2
\
Genes
associated with:
of
0
2,000 4,000 6,000 8,000
Enhancers ranked by increasing MED1 signal
Genome-wide average
Typical enhancers
Super-enhancer
4.0MEDI
4-0
BRD4
E
-
Genes
associated with:
jk 4b o
0-
B
...
2
00
BCL-xL
300,000 -
0
w
--
10
0
oi
14
.
Luciferase
10- Promoter
ts
8-
Cloned enhancer
region (2kb)
6-
~4-
E.
0,0
0.0-________
Typical
0.
Typical
enhancers
Total Density at
signal constituents
18 unit 3.4 unit
16 unit 3.7 unit
10 unit 1.9 unit
IGLL5 Super-enhancer
TOP1 Enhancer
40.0 MED1
40.0
_5kb
enhancer
,signal
40.
40.0 BRD4
chr20:39,062,569
m
chr22:21,597,576
Cloned regions M
SMARCA4 Enhancer
10.0 MED1
CL
E
CCND2 Super-enhancer
Typical 10.01
enhancer
signal
10.0 BRD4
chrl9:10,902,034
qd 4 CO 0 ' "
308
19,400 bp
Total Density at
signal constituents
MED1:
1 unit
1 unit
BRD4: 1 unit 1 unit
H3K27Ac: 1 unit
1 unit
C
00
VL
ci)
U) W
Number:
7,985
Median size: 1,330 bp
10.01
chrl2:4,241,581
CCND2
Figure 2
76
Super-enhancers
Figure 2: Super-enhancers identified in multiple myeloma
A) Total MED1 ChIP-Seq signal in units of reads per million in enhancer regions for all
enhancers in MM 1.S. Enhancers are ranked by increasing MEDI ChIP-Seq signal.
B) Meta-gene representation of global MEDI (red line) and BRD4 (blue line) occupancy at
typical enhancers and super-enhancers. The x-axis shows the start and end of the enhancer (left)
or super-enhancer (right) regions flanked by +/- 5kb of adjacent sequence. Enhancer and superenhancer regions on the x-axis are relatively scaled. The y-axis shows the average signal in units
of rpm/bp.
C) Gene tracks of MED 1 (top) and BRD4 (bottom) ChIP-Seq occupancy at the typical enhancer
upstream of TOP], the super-enhancer downstream of IGLL5, the typical enhancer upstream of
SMARCA4,
and the super-enhancer overlapping the CCND2 gene TSS. The x-axis shows
genomic position and super-enhancer containing regions are depicted with a grey box. The yaxis shows signal of ChIP-Seq occupancy in units of rpm/bp.
D) Left: Boxplots of expression values for genes with proximal typical enhancers (white), or
with proximal super-enhancers (pink). The y-axis shows expression value in Log2 arbitrary
units. Right: Boxplots of cell type specificity values for genes with proximal typical enhancers
(white), or with proximal super-enhancers (purple). The y-axis shows the Z-score of the JS
divergence statistic for genes, with higher values corresponding to a more cell type specific
pattern of expression. Changes between expression levels are significant (two tailed Welch's ttest, p-value < 2e-16) as are changes between cell type specificity levels (two tailed Welch's ttest, p-value = 1e- 14)
77
E) Bar graph depicting luciferase activity of reporter constructs containing cloned fragments of
typical enhancers and super-enhancers in MML.S cells. 2kb fragments of three super-enhancers,
IGLL5, DUSP5, and SUB 1, and three typical enhancers, PDHX, SERPINB8, and TOP 1, ranked
1, 129, 227, 2352, 4203, and 4794, respectively, in terms of MED1 occupancy, were cloned into
reporter plasmids downstream of the luciferase gene, driven by a minimal MYC promoter.
Luciferase activity is represented as fold over empty vector. Error bars represent standard
deviation of triplicate experiments. See also Figure S2 and Data S1.
78
A
B
Super-enhancers:
Super-enhancers:
CL
ALk
10kb
IgH insertion
der3t(3;8)
15.0
MED1
15.0
15.0
BRD4
chrl4:105,092,200
chrl2:40,77,621
chr8:128,833,087
MYC
BRD4
/
CCND2
i
10kb
15.0.
chrl2:4,313,350
5kb
15,01
MED1
MED1
15.0}
BRD4
15.0-
40kb -4
Al1-
"L-
1
chr6:37,118,592
chr6:361,617
chr6:247,594
15.0
1Okb
MED1
15.0 -10kb
15e
MED1
BRD4
15.0-
BRD4
chr2O:29,785,891
chr6:106,738,042
chr6:106,634,280
chr6:37,260,532
PIMi
IRF4
15.0-
BRD4
chr20:29,706,134
BCL-xL
PRDM1
I
15.0-
15.0'
1.
10kbMD1
C
E
EL
15.0'
BRD4
15.0
chri:148,820,188
chr22:27,512,763
chr22:27,552,905
XBP
I
MCL1
Figure 3
79
10kb
MED1
BRD4
chrl:148,798,243
Figure 3: Super-enhancers are associated with key multiple myeloma genes
A,B) Gene tracks of MEDI and BRD4 ChIP-Seq occupancy at super-enhancers near genes with
important roles in multiple myeloma biology A) or genes with important roles in cancer B).
Super-enhancers are depicted in gray boxes over the gene tracks. The x-axis shows genomic
position and super-enhancer containing regions are depicted with a grey box. The y-axis shows
signal of ChIP-Seq occupancy in units of rpm/bp.
80
Inhibition of BRD4 leads to displacement of BRD4 genome-wide
BRD4 interacts with chromatin-associated proteins such as transcription factors, the
Mediator complex and acetylated histones (Dawson et al., 2011; Dey et al., 2003; Jang et al.,
2005; Jiang et al., 1998; Wu and Chiang, 2007; Wu et al., 2013). Previous studies have shown
that treatment of MM 1 .S cells with JQ 1 leads to reduced levels of BRD4 at the IgH enhancer that
drives MYC expression (Delmore et al., 2011), but it is not clear whether such treatment causes a
general reduction in the levels of BRD4 associated with the genome. We found that treatment of
MML.S cells with 500nM JQ 1 for 6 hours reduced the levels of BRD4 genome-wide by
approximately 70% (Figure 4A and 4B). This reduction in BRD4 occupancy was evident both
by inspection of individual gene tracks (Figure 4C) and through global analysis of the average
effects at enhancers and TSSs (Figure 4D). JQI treatment led to -60% reduction in BRD4 signal
at enhancers and -90% reduction at promoters (Figure 4D). The reduction in BRD4 was more
profound at super-enhancers such as those associated with IgH-MYC and CCND2 (Figure 4E),
where the loss of BRD4 was nearly complete. We conclude that BET bromodomain inhibition
of BRD4 leads to reduced levels of BRD4 at enhancers and promoters throughout the genome in
MM1.S cells.
81
10Mb
A
Chr2l
1
25.0 BRD4 ChIP-Seq
DMSO
EIiI
25.0
B
500nM
JQ1
,,
ii iiiiAiLi
All BRD4 enriched regions
4.0- BRD4
ChIP-Seq
3.0E 2.01.0-
0.0J01 concentration (nM)
C
EnhancerI-I
8.0
BRD4 ChIP-Seq
DMSO
.
8.0
500nM JQ1
chrl9:10,945,355
chrl9:10,904,705
SM FARCA4
D
4.0
BRD4
-2.5kb Enhancer +2.5kb
Center
E
IgH-MYC super-enhancer
15.0
BRD4
MSO
10kb
I rL, .IL
0.L
15.0
O0nMJQ1
-2.5kb
TSS
+2.5kb
CCND2 super-enhancer
10kb
10.0
0]LAL
10.0
CCND2
//
Figure 4
82
Figure 4: Inhibition of BRD4 leads to loss of BRD4 genome-wide
A) Tracks showing BRD4 ChIP-Seq occupancy on the 35mb right arm of chromosome 21 after
DMSO (top) or 500nM JQl (bottom) treatment. The chromosome 21 ideogram is displayed
above the gene tracks with the relevant region highlighted in blue. The x-axis of the gene tracks
shows genomic position and the y-axis shows BRD4 ChIP-Seq signal in units of rpm/bp.
B) Boxplot showing the distributions of BRD4 ChIP-Seq signal at BRD4 enriched regions after
DMSO (left) or 500nM JQl (right) treatment. BRD4 enriched regions were defined in MML.S
cells treated with DMSO. The y-axis shows BRD4 ChIP-Seq signal in units of rpm/bp. The loss
of BRD4 occupancy at BRD4 enriched regions after JQ l is highly significant (p-value < le-16).
C) Gene tracks of BRD4 ChIP-Seq occupancy at the enhancer (left) and promoter (right) of
SMARCA4 in MM L.S cells after DMSO (top) or 500nM JQI (bottom) treatment for 6 hours. The
x-axis shows genomic position and enhancer containing regions are depicted with a white box.
The y-axis shows signal of ChIP-Seq occupancy in units of rpm/bp.
D) Meta-gene representation of global BRD4 occupancy at enhancers and promoters after
DMSO (solid line) or 500nM JQ l (dotted line) treatment. The x-axis shows the +/- 2.5kb region
flanking either the center of enhancer regions (left) or the TSS of active genes. The y-axis shows
the average background subtracted ChIP-Seq signal in units of rpm/bp.
E) Gene tracks of BRD4 binding at super-enhancers after DMSO (top) or 500nM JQ 1 (bottom)
treatment. The x-axis shows genomic position and super-enhancer containing regions are
depicted with a grey box. The y-axis shows signal of ChIP-Seq occupancy in units of rpm/bp.
83
Transcription of super-enhancer-associated genes is highly sensitive to BRD4 inhibition
Enhancers are formed through co-operative and synergistic binding of multiple
transcription factors and coactivators (Carey, 1998; Carey et al., 1990; Giese et al., 1995; Kim
and Maniatis, 1997; Thanos and Maniatis, 1995). As a consequence of this binding behavior,
enhancers bound by many cooperatively-interacting factors lose activity more rapidly than
enhancers bound by fewer factors when the levels of enhancer-bound factors are reduced
(Giniger and Ptashne, 1988; Griggs and Johnston, 1991). The presence of super-enhancers at
MYC and other key genes associated with multiple myeloma led us to consider the hypothesis
that super-enhancers are more sensitive to reduced levels of BRD4 than typical enhancers and
that genes associated with super-enhancers might then experience a greater reduction of
transcription than genes with average enhancers when BRD4 is inhibited (Figure 5A).
To test this hypothesis, we first examined the effects of various concentrations of JQ 1 on
BRD4 occupancy genome-wide (Figure 5B). JQ 1 had little effect on MM 1 .S cell viability when
treated for 6 hours at these various concentrations, while at later time points JQ 1 had a
significant anti-proliferative effect (Figure 5C). As expected, MYC protein levels were
significantly depleted by exposure of MMl .S cells to 50nM or greater doses of JQl for 6 hrs
(Figure 5D) (Delmore et al., 2011). In contrast, JQ1 did not affect total BRD4 protein levels
within the cells, and did not significantly reduce ChIP efficiency (Figure 5E). When BRD4
occupancy was examined genome-wide in cells exposed to the increasing concentrations of JQ 1,
it was evident that super-enhancers showed a greater loss of BRD4 occupancy than typical
enhancer regions (Figure 5F). For example, the IgH super-enhancer showed significantly greater
reduction in BRD4 occupancy in cells treated with 5nM or 50nM JQl than typical enhancer
regions such as that upstream of SMARCA4 (Figure 5G). Ultimately, virtually all BRD4
84
occupancy was lost at the IgH super-enhancer (97% reduction versus DMSO control) after
treatment with 500nM JQ 1, while loss of BRD4 occupancy at the typical enhancer for
SMARCA4 was less pronounced (71% reduction versus DMSO control) (Figure 5G).
We next investigated whether genes associated with super-enhancers might experience a
greater reduction of transcription than genes with average enhancers when BRD4 is inhibited.
As expected, treatment of MML.S cells with 500nM JQ 1 led to progressive reduction in global
mRNA levels over time (Figure 6A; Figure S3A). Similarly, treatment with increasing
concentrations of JQ 1 caused progressive reductions in global mRNA levels (Figure 6A; Figure
S3B). There was a selective depletion of mRNAs from super-enhancer-associated genes that
occurred in both a temporal (Figure 6B) and concentration-dependent manner (Figure 6C).
Notably, MYC and IRF4 mRNA levels were more rapidly depleted than other mRNAs that are
expressed at similar levels (Figure 6D). The levels of transcripts from super-enhancer-associated
genes were somewhat more affected than those from genes that have multiple typical enhancers
bound by BRD4 (Figure S3C and S3D). Thus, BET bromodomain inhibition preferentially
impacts transcription of super-enhancer-driven genes.
To further test the model that super-enhancers are responsible for the special sensitivity
to BRD4 inhibition, we transfected MMl .S cells with luciferase reporter constructs containing
super-enhancer and typical enhancer fragments and examined the effects of various JQ1
concentrations on luciferase activity. Upon treatment with JQl, MM 1.S cells transfected with a
super-enhancer reporter experienced a greater reduction in luciferase activity than those
transfected with a typical enhancer reporter (Figure 6E). Interestingly, the dose-response curve
observed for luciferase activity of the super-enhancer construct is consistent with that expected
for enhancers that are bound cooperatively by multiple factors (Figure 5A) (Giniger and Ptashne,
85
1988; Griggs and Johnston, 1991). These results are also consistent with the model that superenhancers are responsible for the special sensitivity of gene transcription to BRD4 inhibition.
86
A
F
Typical Enhancer
Activator Transcriptional
activity
C
0 .:
100.
MM Enhancers
80.
C
E
60.
DNA binding
site
Activator
concentration
TSS
IRD
Super-enhancer
500 5,000
50
5
J01 concentration (nM)
DMSO
G
+
+++
Activator
concentration
5nM JQ1
ChIP-Seq
5OnM
MM.1S
500nM JQ1
Gene
expression
3
5,OOOnM JQ1
1501
3
0
6 HR
100
24 HR
50.
0
5000
500
50
5
JQ1 concentration (nM)
E
D
Western Blot
Western Blot
MYC
BRD4
Actin
Actin
ChIP-Westem
BRD4
JQ1 (nM)
46
17
50OOnM
...---- --J01
0
58
500nM J31
C
0
100
JQ1
Western blot
50nM JQ1
z
10kb
% BRD4 3.0
BRD4 ChIP-Seq
IDMSO
remaining
100
A L
25nM
lJ 0 -29
DMSO
C
SMARCA4 enhancer
IgH-MYC super-enhancer
15.0
C.
Analyses:
6 hour treatment
C
Typical
Super
20.
0
=Z
B
Enhancer type:
E 40.
N 0M
0
JQ1 (nM)
Figure 5
87
1
29
32
Figure 5: BRD4 occupancy at super-enhancers is highly sensitive to bromodomain
inhibition
A) Schematic example of how cooperative interactions of enhancer-associated factors at superenhancers leads to both higher transcriptional output and increased sensitivity to factor
concentration.
B) Measuring the effects of various concentrations of JQ 1 on genome-wide on BRD4 occupancy.
Schematic depicting the experimental procedure.
C) Short-term JQ1 treatment (6 hours) has little effect on MML.S cell viability. JQl sensitivity
of MML.S cells by measurement of ATP levels (CellTiterGlo) after 6, 24, 48 and 72 hours of
treatment with JQ1 (5, 50, 500 or 5000nM) or vehicle (DMSO, 0.05%).
D) Western blot of relative MYC levels after 6 hours of JQ 1 or DMSO treatment.
E) Western blot of relative BRD4 levels after 6 hours of JQ1 or DMSO treatment.
ChIP-
Western blot of the relative levels of immunoprecipitated BRD4 after 6 hours of JQ1 or DMSO
treatment.
F) Line graph showing the percentage of BRD4 occupancy remaining after 6 hour treatment at
various JQ1 concentrations for typical enhancers (grey line) or super-enhancers (red line). The
y-axis shows the fraction of BRD4 occupancy remaining versus DMSO. The x-axis shows
different JQ1 concentrations (DMSO (none), 5nM, 50nM, and 500nM). Error bars represent
95% confidence intervals of the mean (95% CI).
G) Gene tracks of BRD4 ChIP-Seq occupancy after various concentrations of JQ1 treatment at
the IgH-MYC associated super-enhancer (left) and the SMARCA4-associated typical enhancer
(right). The x-axis shows genomic position and grey boxes depict super-enhancer regions. The
88
y-axis shows signal of ChIP-Seq occupancy in units of rpm/bp. The percent BRD4 remaining
after each concentration of JQ 1 treatment is annotated to the right of the gene tracks.
89
BRD4 inhibition and transcription elongation
At active genes, enhancers and core promoters are brought into close proximity, so
factors associated with enhancers can act on the transcription apparatus in the vicinity of
transcription start sites and thereby influence initiation or elongation. BRD4 is known to interact
with Mediator and P-TEFb and to be involved in the control of transcriptional elongation by
RNA polymerase II (Conaway and Conaway, 2011; Dawson et al., 2011; Jang et al., 2005;
Krueger et al., 2010; Rahman et al., 2011; Yang et al., 2005). This suggests that the preferential
loss of BRD4 from super-enhancers might affect the levels of Mediator and P-TEFb at these sites
and furthermore, that the reduced levels of mRNAs from super-enhancer-associated genes might
be due to an effect on transcription elongation.
To test these predictions, we carried out ChIP-Seq for the Mediator component MED 1,
and the catalytic subunit of the P-TEFb complex CDK9, in MM 1.S cells treated with DMSO or
500nM JQ 1 for 6 hours. In control cells, MED 1 and CDK9 were found at enhancers and
promoters of active genes throughout the MM genome, as expected (Figure IA and IB; Figure
S3E). In cells treated with JQ1, reduced levels of MED 1 and CDK9 were observed primarily at
enhancers, with the greatest loss at super-enhancers (Figure 6F). As many super-enhancers span
contiguous regions that encompass or overlap the TSS, we analyzed MED 1 and CDK9 loss in
either TSS proximal or TSS distal regions of super-enhancers and again observed loss of MED 1
and CDK9 predominantly at TSS distal regions (Figure S3F). We conclude that inhibition of
BRD4 genomic binding leads to a marked reduction in the levels of Mediator and P-TEFb at
genomic regions distal to TSSs, with the greatest reduction occurring at super-enhancers.
To determine whether reduced levels of BRD4 lead to changes in transcription
elongation, we quantified changes in transcription elongation by performing ChIP-Seq of RNA
90
Pol II before and after treatment of MML.S cells with 500nM JQ1. We then calculated the fold
loss of RNA Pol II occupancy in the gene body regions for all transcriptionally active genes and
found that more than half of these genes show a decrease in elongating RNA Pol II density after
JQ 1 treatment (Figure 6G). Importantly, genes associated with super-enhancers showed a
greater decrease of RNA Pol II in their elongating gene body regions compared to genes
associated with typical enhancers (Figure 6H; Figure S3G). Inspection of individual gene tracks
revealed pronounced elongation defects at super-enhancer-associated genes such as MYC and
IRF4, with the greatest effects observed with MYC (Figure 61 and 6J). Thus, the selective effects
of JQ 1 on the transcription of MYC and other super-enhancer-associated genes can be explained,
at least in part, by the sensitivity of super-enhancers to reduced levels of BRD4, which leads to a
pronounced effect on pause release and transcription elongation.
91
A
F
00
T
'0.0
-2.0
-2.0
500nM JQ1
treatment for (hrs)
0
*C
-1.0
-12.0
201 MEDI ChIP-Seq
0
0+
a) M
20 .jCDK9 ChIP-Seq
0
000
JQI concentration
(nM)
-20.
O
4
-40
B
0.0.
Change in expression vs. time
Ar
4
G
-0.2.
-0.4 -
r
Genes
associated
with:
-0.6-
Typical
-0.8.
Super
-1.0Control
0.5
1
500nM JQ1
C
2.0.
1.0.
-_= 0.0.
0.0.
-
00 z.0
04
3
6
treatment for (hrs)
H
Change in expression vs. dose
-0.6 -
C
Super
-
5
50
500 5,000
JOI concentration (nM)
DMso
Genes associated with:
Super-enhancers
0-0
Typical
-0.3 J- enhancers
16.0
Q
Pol 11ChIP-Seq
DMSO
j RNA
-1.0.
I
D
1.0.
-0.1
=+
C ~g-0.2f300.
Iypical
-0.8 -
High expressed genes
-
16.0
5O0nM JQ1
chr8:128,812,310
C 4)
IGLL
IRF4
-2.0
10.0
-4.0
,x
Control
0.5
chr8-128,827,31 1
J
MYC
E
2,000 4,000 6,000 8,000 10,000
Transcriptionally active genes
0.0.
o
Genes
associated
with:
U.U
F4
CY
tM -0.4 -
C
BCLJ
CCND2
MYC
0
-0.2-
20
(D (0
\
3
1
6
10.0
RNA Pol Il ChlP-Seq
JDMSO
2Li
500nM JQI
20
chr6:324,515
15
4
IGLL5
10
Super-enhancer
5
PDHX
Typical
-,
0
enhance
10 25 100 250
JQ1 Concentration (nM)
DMSO
92
IRF4
chr6:359,517
Figure 6
Figure 6: JQ1 causes disproportionate loss of transcription at super-enhancer genes
A) Boxplots showing the Log 2 change in gene expression for all actively transcribed genes in
JQ1 treated vs. control cells for a time course of cells treated with 500nM JQ1 (left) or for a
concentration course of cells treated for 6 hours with varying amounts of JQ 1 (right). The Y-axis
shows the Log 2 change in gene expression vs. untreated control cells (left graph) or control cells
treated with DMSO for 6 hours (right graph).
B,C) Line graph showing the Log 2 change in gene expression vs. control cells after JQ1
treatment in a time B) or dose C) dependent manner for genes associated with typical enhancers
(grey line) or genes associated with super-enhancers (red line). The Y-axis shows the Log 2
change in gene expression of JQl treated vs. untreated control cells. X-axis shows time of
500nM JQ 1 treatment B) or JQ 1 treatment concentration at 6 hours C).
D) Graph showing the Log 2 change in gene expression after JQ1 treatment over time for genes
ranked in the top 10% of expression in MM 1. S cells. Each line represents a single gene with the
MYC and IRF4 genes drawn in red. The Y-axis shows the Log 2 change in gene expression of JQ 1
treated vs. untreated control cells. The X-axis shows time of 500nM JQl treatment.
E) Line graph showing luciferase activity after JQ 1 treatment at various concentrations for
luciferase reporter constructs containing either a fragment from the IGLL5 super-enhancer (red
line) or the PDHX typical enhancer (grey line). Y-axis represents relative luciferase activity in
arbitrary units. X-axis shows JQ 1 concentrations. Error bars are standard error of the mean.
F) Bar graphs showing the percentage loss of either MED 1 (top, red) or CDK9 (bottom, green) at
promoters, typical enhancers, and super-enhancers. Error bars represent 95% CI.
93
G) Graph of loss of RNA Pol II density in the elongating gene body region for all
transcriptionally active genes in MML.S cells after 6 hour 500nM JQ1 treatment. Genes are
ordered by decrease in elongating RNA Pol II in units of Log 2 fold loss. Genes with a greater
than 0.5 Log 2 fold change in elongating RNA Pol II are shaded in green (loss) or red (gain). The
amount of RNA Pol II loss is indicated for select genes.
H) Bar graph showing the Log 2 fold change in RNA Pol II density in elongating gene body
regions after 6 hour 500nM JQ1 treatment for genes with typical enhancers (left, grey) or genes
with super-enhancers (red, right). Error bars represent 95% CI.
1,J) Gene tracks of RNA Pol II ChIP-Seq occupancy after DMSO (black) or 500nM JQ1
treatment (red) at the super-enhancer proximal MYC gene I) and IRF4 gene J). The y-axis
shows signal of ChIP-Seq occupancy in units of rpm/bp. See also Figure S3.
94
Super-enhancers are associated with disease-critical genes in other cancers
To map enhancers and determine whether super-enhancers occur in additional tumor types, we
investigated the genome-wide occupancy of Mediator (MED 1), BRD4, and the enhancerassociated histone modification H3K27Ac using ChIP-Seq in glioblastoma multiforme (GBM)
and small cell lung cancer (SCLC) (Figure 7). Mediator (MED 1) occupancy was used to identify
enhancer elements because enhancer-bound transcription factors bind directly to Mediator
(Borggrefe and Yue, 2011; Conaway and Conaway, 2011; Kornberg, 2005; Malik and Roeder,
2010; Taatjes, 2010) and because it has proven to produce high-quality evidence for enhancers in
mammalian cells (Kagey et al., 2010). Global occupancy of BRD4 and H3K27Ac were used as
corroborative evidence to identify enhancer elements (Figure S4; Table S4). Analysis of the
regions occupied by Mediator revealed that, as in MM 1.S cells, large genomic domains were
occupied by this coactivator in both GBM and SCLC (Figure 7A, 7B, 7D, and 7E). The median
super-enhancer was 30kb in GBM cells and I lkb in SCLC cells (Figure 7B and 7E). As in
MM 1. S cells, these GBM and SCLC super-enhancers were an order of magnitude larger and
showed a commensurate increase in MED 1, BRD4 and H3K27Ac levels when compared to
normal enhancers (Figure 7B and 7E).
The super-enhancers in GBM and SCLC were found to be associated with many well-known
tumor-associated genes (Figure 7C and 7F; Table S5). In GBM, super-enhancers were
associated with genes encoding three transcription factors (R UNXJ, FOSL2, BHLHE40) critical
for mesenchymal transformation of brain tumors (Carro et al., 2010); the super-enhancers
associated with BHLHE40 are shown in Figure 7C. BCL3, which associates with NFKB and is
deregulated in many blood and solid tumor types, is associated with an upstream, and an
intergenic super-enhancer in GBM (Figure 7C) (Maldonado and Melendez-Zajgla, 2011). In
95
SCLC, a super-enhancer is associated with the INSMJ gene, which encodes a transcription factor
involved in neuronal development that is highly expressed in neuroendocrine tissue and tumors
such as SCLC (Figure 7F) (Pedersen et al., 2003). A super-enhancer is also associated with the
ID2 gene, which is highly expressed in SCLCs and encodes a protein that interacts with the wellknown retinoblastoma tumor suppressor (Figure 7F) (Pedersen et al., 2003; Perk et al., 2005).
These results indicate that super-enhancers are likely to associate with critical tumor oncogenes
in diverse tumor types.
96
Small Cell Lung Cancer (SCLC)
Glioblastoma Multiforme (GBM)
200,000C
C,C
150.000-
40,000-
I,',
-'-I
.0.
30,000.
C
*0
S
20,000
0
U)
-
100,000.
Cu
'C
10,000.
50,000.
0
2
0
0
0
5,000
10,000 15,000
Enhancers ranked
by increasing MEDI signal
E
B
2,000 4,000 6,100
Enhancers ranked
by increasing MEDI signal
Genome-wide average
Super-enhancer
Typical enhancers
Genome-wide average
Super-enhancer
Typical enhancers
1.5- MEDI 1.5BRD4
.
0.0.
D4
2.0
D
.
10.0
0.0
0.0 t
Total Density at
signal constituents
1 unit
1 unit
MEDI:
1 unit
1 unit
BRD4:
I unit
H3K27Ac: I unit
5,912
Number:
Median size: 1,060bp
528
30,100 bp
F
Super-enhancers:
8
.0
I
4.0
chr3:4,983,336
E1kb
-
~AL~L
8.01
BRD4
chr2O:20,424.566
chr20:20,294,546
chr3:5,063,576
INSM1~
10kb
'[I
chrl9:49,9
MED1
8.01
BRD4
4.0
10kb
8.0.
MED1
chrl9:49,929,859
BCL3
10kb
MED1
MEDI
BRD4
Total Density at
signal constituents
8.9 unit 2.1 unit
9 unit 2.4 unit
9.2 unit 2.1 unit
Super-enhancers:
8.0
BHLHE40r
8.0
274
11,400 bp
Total Density at
signal constituents
1 unit
1 unit
MED1:
1 unit
1 unit
BRD4:
I unit
H3K27Ac: 1 unit
Total Density at
signal constituents
17 unit 3.0 unit
16 unit 2.7 unit
11 unit 2.0 unit
W
CO
(nw
15,024
Number:
Median size: 1,480 bp
BRD4
I
0,066
--
chr2:8,750,153
chr2:8,730,147
ID2
Figure 7
97
Figure 7: Super-enhancers are associated with key genes in other cancers
A,D) Total MEDI ChIP-Seq signal in units of reads per million in enhancer regions for all
enhancers in A) the glioblastoma multiforme (GBM) cell line U-87 MG or D) the small cell lung
cancer (SCLC) cell line H2171. Enhancers are ranked by increasing MED 1 ChIP-Seq signal.
B,E) Meta-gene representation of global MED 1 and BRD4 occupancy at B) typical GMB
enhancers and super-enhancers or E) typical SCLC enhancers and super-enhancers. The x-axis
shows the start and end of the enhancer (left) or super-enhancer (right) regions flanked by +/5kb of adjacent sequence. Enhancer and super-enhancer regions on the x-axis are relatively
scaled. The y-axis shows the average signal in units of rpm/bp.
C,F) Gene tracks of MEDI and BRD4 ChIP-Seq occupancy at C) super-enhancers near
BHLHE40 and BCL3, genes with important roles in GBM or at F) super-enhancers near INSM1
and ID2, genes with important roles in SCLC. Super-enhancers are depicted in gray boxes over
the gene tracks. See also Figure S4.
98
Discussion
Chromatin regulators have become attractive targets for cancer therapy, but many of
these regulators are expressed in a broad range of healthy cells and contribute generally to gene
expression. Thus, it is unclear how inhibition of a global chromatin regulator such as BRD4
might produce selective effects, such as at the MYC oncogene (Delmore et al., 2011). We have
found that key regulators of tumor cell state in MML.S cells are associated with large enhancer
domains, characterized by disproportionately high levels of BRD4 and Mediator. These superenhancers are more sensitive to perturbation than typical enhancers and the expression of the
genes associated with super-enhancers is preferentially affected. Thus, the preferential loss of
BRD4 at super-enhancers associated with the MYC oncogene and other key tumor-associated
genes can explain the gene-selective effects of JQl treatment in these cells.
BRD4 is an excellent example of a chromatin regulator that is expressed in a broad range
of healthy cells and contributes generally to gene expression. Most cell types for which RNASeq data is available express the BRD4 gene. ChIP-seq data revealed that BRD4 generally
occupies the enhancer and promoter elements of active genes with the Mediator coactivator in
MMl .S cells (Figure 1). These results eliminate the model that BRD4 is exclusively associated
with a small set of genes that are thereby rendered inactive by the BRD4 inhibitor JQ 1 and
instead suggest that the gene-specific effects of the small molecule have other causes.
We have found that -3% of the enhancers in MMl .S cells are exceptionally large and are
occupied by remarkably high amounts of BRD4 and Mediator. These super-enhancers are
generally an order of magnitude larger and contain an order of magnitude more BRD4, Mediator
and histone marks associated with enhancers (H3K27Ac) than typical enhancers. Our results
suggest that super-enhancers are collections of closely spaced enhancers that can collectively
99
facilitate high levels of transcription from adjacent genes. Importantly, the super-enhancers are
associated with the MYC oncogene and additional genes such as IGLL5, IRF4, PRDMJ/BLIMP1, and XBP1 that feature prominently in MM biology.
Co-operative and synergistic binding of multiple transcription factors and coactivators
occurs at enhancers. Enhancers bound by many cooperatively-interacting factors can lose activity
more rapidly than enhancers bound by fewer factors when the levels of enhancer-bound factors
are reduced (Giniger and Ptashne, 1988; Griggs and Johnston, 1991). The presence of superenhancers at MYC and other key genes associated with multiple myeloma led us to test the
hypothesis that super-enhancers are more sensitive to reduced levels of BRD4 than average
enhancers. We found that treatment of these tumor cells with the BET-bromodomain inhibitor
JQ 1 leads to preferential loss of BRD4 at super-enhancers. In addition, this decrease in BRD4
occupancy is accompanied by a corresponding loss of MED 1 and CDK9 at super-enhancers.
Consequent transcription elongation defects and mRNA decreases preferentially impact superenhancer-associated genes, with an especially profound effect at the MYC oncogene.
Super-enhancers are not restricted to MM cells. We have identified super-enhancers in
two additional tumor types, small cell lung cancer and glioblastoma multiforme. Superenhancers identified in these cell types have characteristics similar to those found in MMl .S;
they span large genomic regions and contain exceptional amounts of Mediator and BRD4. These
super-enhancers are also associated with important tumor genes in both cell types. In GBM
cells, BHLHE40 and BCL3 are known to be important in tumor biology, and are each associated
with super-enhancers in this cell type. In H2171 SCLC cells, super-enhancers are associated with
INSMJ and ID2, which are frequently overexpressed in SCLC.
100
Our results demonstrate that super-enhancers occupied by BRD4 regulate critical
oncogenic drivers in multiple myeloma and show that BRD4 inhibition leads to preferential
disruption of these super-enhancers. This insight into the mechanism by which BRD4 inhibition
causes selective loss of oncogene expression in this highly malignant blood cancer may have
implications for future drug development in oncology. Tumor cells frequently become addicted
to oncogenes, thus becoming unusually reliant on high level expression of these genes (Cheung
et al., 2011; Chin et al., 1999; Felsher and Bishop, 1999; Garraway and Sellers, 2006; Garraway
et al., 2005; Jain et al., 2002; Weinstein, 2002). Thus, preferential disruption of super-enhancer
function may be a general approach to selectively inhibiting the oncogenic drivers of many
tumor cells.
Experimental Procedures
Cell Culture
MML.S multiple myeloma cells (CRL-2974 ATCC) and U-87 MG glioblastoma cells (HTB-14
ATCC) were purchased from ATCC. H2171 small cell lung carcinoma cells (CRL-5929 ATCC)
were kindly provided by John Minna, UT Southwestern. MM 1.S and H2171 cells were
propagated in RPMI- 1640 supplemented with 10% fetal bovine serum and 1% GlutaMAX
(Invitrogen, 35050-061). U-87 MG cells were cultured in Eagle's Minimum Essential Medium
(EMEM) modified to contain Earles Balanced Salt Solution, nonessential amino acids, 2 mM Lglutamine, 1 mM sodium pyruvate, and 1500 mg/l sodium bicarbonate. Cells were grown at
37'C and 5% CO 2 .
101
For JQ 1 treatment experiments, cells were resuspended in fresh media containing JQ 1 (5nM,
50nM, 500nM, 5000nM) or vehicle (DMSO, 0.05%) and treated for a duration of 6 hours, unless
otherwise indicated.
ChIP-Seq
ChIP was carried out as described in Lin et al. (2012). Additional details are provided in
supplemental methods. Antibodies used are as follows: total RNA Pol II (Rpb 1 N-terminus):
Santa Cruz sc-899 lot# KO I11; MEDI: Bethyl Labs A300-793A lot#A300-793A-2; BRD4:
Bethyl Labs A301-985A lot# A301-985A-1; CDK9: Santa Cruz Biotechnology sc-484, lot
D1612. ChIP-Seq datasets of H3K4Me3 and H3K27Ac in MML.S and MEDI and H3K27Ac in
U-87 MG and H2717 were previously published (Lin et al., 2012).
Luciferase Reporter Assays
A minimal Myc promoter was amplified from human genomic DNA and cloned into the Sac and
HindlIl sites of the pGL3 basic vector (Promega). Enhancer fragments were likewise amplified
from human genomic DNA and cloned into the BamHI and SalI sites of the pGL3-pMyc vector.
All cloning primers are listed in Table S6. Constructs were transfected into MM .S cells using
Lipofectamine 2000 (Invitrogen). The pRL-SV40 plasmid (Promega) was cotransfected as a
normalization control. Cells were incubated for 24 hours, and luciferase activity was measured
using the Dual-Luciferase Reporter Assay System (Promega). For the JQI concentration course,
cells were resuspended in fresh media containing various concentrations of JQ 1 24 hours after
transfection, and incubated for an additional 6 hours before harvesting. Luminescence
measurements were made using the Dual-Luciferase Reporter Assay System (Promega) on a
Wallac EnVision (Perkin Elmer) plate reader.
102
Cell Viability Assays
Cell viability was measured using the CellTiterGlo Assay kit (Promega, G7571). MML.S cells
were resupsended in fresh media containing JQI (5nM, 50nM, 500nM, I0OnM) or vehicle
(DMSO, 0.05%), then plated in 96-well plates at 10,000 cells/well in a volume of I0uL.
Viability was measured after 6, 24, 48 and 72 hour incubations by addition of CellTiter Glo
reagent and luminescence measurement on a Tecan Safire 2 plate reader.
Western Blotting
Western blots were carried out using standard protocols. Antibodies used are as follows: c-Myc
(Epitomics, cat. #: 1472-1), BRD4 (Epitomics, cat.#: 5716-1) or
s-Actin (Sigma, clone AC-15,
A544 1).
Data Analysis
All ChIP-Seq datasets were aligned using Bowtie (version 0.12.2) (Langmead et al., 2009) to
build version NCBI36/HG 18 of the human genome. Aligned and raw data can be found online
associated with the GEO Accession ID GSE42355 (www.ncbi.nlm.nih.gov/geo/). Individual
dataset GEO Accession IDs and background datasets used can be found in Table S7.
ChIP-Seq read densities in genomic regions was calculated as in (Lin et al. 2012). We used the
MACS version 1.4.1 (Model based analysis of ChIP-Seq) (Zhang et al., 2008) peak finding
algorithm to identify regions of ChIP-Seq enrichment over background. A p-value threshold of
enrichment of 1 e-9 was used for all datasets.
103
Active enhancers were defined as regions of ChIP-Seq enrichment for the Mediator complex
component MEDI outside of promoters (e.g. a region not contained within +/- 2.5kb region
flanking the promoter). In order to accurately capture dense clusters of enhancers, we allowed
MEDI regions within 12.5kb of one another to be stitched together. To identify super-enhancers,
we first ranked all enhancers by increasing total background subtracted ChIP-Seq occupancy of
MED 1 (x-axis), and plotted the total background subtracted ChIP-Seq occupancy of MED 1 in
units of total rpm (y-axis). This representation revealed a clear inflection point in the distribution
of MED 1 at enhancers. We geometrically defined the inflection point and used it to establish the
cut off for super-enhancers (See supplemental methods).
Acknowledgments
We thank Zi Peng Fan for bioinformatics support; Michael R. McKeown for help with cell
viability assays; Jun Qi for providing JQ 1; Tom Volkert, Jennifer Love, Sumeet Gupta, and
Jeong-Ah Kwon at the Whitehead Genome Technologies Core for Solexa sequencing; and
members of the Young, Bradner and Vakoc labs for helpful discussion. This work was supported
by a Swedish Research Council Postdoctoral Fellowship VR-B0086301 (J.L.), the DamonRunyon Cancer Research Foundation (J.E.B), a Burroughs-Welcome CAMS award and NCI
Cancer Center Support Grant Development Fund CA45508 (C.R.V.), and National Institutes of
Health grants HG002668 (R.A.Y) and CA146445 (R.A.Y, T.L).
104
References
Baylin, S.B., and Jones, P.A. (2011). A decade of exploring the cancer epigenome - biological
and translational implications. Nat Rev Cancer 11, 726-734.
Bergsagel, P.L., Kuehl, W.M., Zhan, F., Sawyer, J., Barlogie, B., and Shaughnessy, J., Jr. (2005).
Cyclin D dysregulation: an early and unifying pathogenic event in multiple myeloma. Blood 106,
296-303.
Beroukhim, R., Mermel, C.H., Porter, D., Wei, G., Raychaudhuri, S., Donovan, J., Barretina, J.,
Boehm, J.S., Dobson, J., Urashima, M., et al. (2010). The landscape of somatic copy-number
alteration across human cancers. Nature 463, 899-905.
Borggrefe, T., and Yue, X. (2011). Interactions between subunits of the Mediator complex with
gene-specific transcription factors. Semin Cell Dev Biol 22, 759-768.
Carey, M. (1998). The enhanceosome and transcriptional synergy. Cell 92, 5-8.
Carey, M., Leatherwood, J., and Ptashne, M. (1990). A potent GAL4 derivative activates
transcription at a distance in vitro. Science 247, 710-712.
Carrasco, D.R., Sukhdeo, K., Protopopova, M., Sinha, R., Enos, M., Carrasco, D.E., Zheng, M.,
Mani, M., Henderson, J., Pinkus, G.S., el al. (2007). The differentiation and stress response
factor XBP-I drives multiple myeloma pathogenesis. Cancer Cell 11, 349-360.
Carro, M.S., Lim, W.K., Alvarez, M.J., Bollo, R.J., Zhao, X., Snyder, E.Y., Sulman, E.P., Anne,
S.L., Doetsch, F., Colman, H., et al. (2010). The transcriptional network for mesenchymal
transformation of brain tumours. Nature 463, 318-325.
Chepelev, I., Wei, G., Wangsa, D., Tang, Q., and Zhao, K. (2012). Characterization of genomewide enhancer-promoter interactions reveals co-expression of interacting genes and modes of
higher order chromatin organization. Cell Res 22, 490-503.
Cheung, H.W., Cowley, G.S., Weir, B.A., Boehm, J.S., Rusin, S., Scott, J.A., East, A., Ali, L.D.,
Lizotte, P.H., Wong, T.C., et al. (2011). Systematic investigation of genetic vulnerabilities across
cancer cell lines reveals lineage-specific dependencies in ovarian cancer. Proc Natl Acad Sci U S
A 108, 12372-12377.
Chin, L., Tam, A., Pomerantz, J., Wong, M., Holash, J., Bardeesy, N., Shen, Q., O'Hagan, R.,
Pantginis, J., Zhou, H., et al. (1999). Essential role for oncogenic Ras in tumour maintenance.
Nature 400, 468-472.
Chng, W.J., Huang, G.F., Chung, T.H., Ng, S.B., Gonzalez-Paz, N., Troska-Price, T., Mulligan,
G., Chesi, M., Bergsagel, P.L., and Fonseca, R. (2011). Clinical and biological implications of
MYC activation: a common difference between MGUS and newly diagnosed multiple myeloma.
Leukemia 25, 1026-1035.
105
Claudio, J.O., Masih-Khan, E., Tang, H., Goncalves, J., Voralia, M., Li, Z.H., Nadeem, V.,
Cukerman, E., Francisco-Pabalan, 0., Liew, C.C., et al. (2002). A molecular compendium of
genes expressed in multiple myeloma. Blood 100, 2175-2186.
Cole, P.A. (2008). Chemical probes for histone-modifying enzymes. Nat Chem Biol 4, 590-597.
Conaway, R.C., and Conaway, J.W. (2011). Function and regulation of the Mediator complex.
Curr Opin Genet Dev 21, 225-230.
Creyghton, M.P., Cheng, A.W., Welstead, G.G., Kooistra, T., Carey, B.W., Steine, E.J., Hanna,
J., Lodato, M.A., Frampton, G.M., Sharp, P.A., et al. (2010). Histone H3K27ac separates active
from poised enhancers and predicts developmental state. Proc Natl Acad Sci U S A 107, 2193121936.
Dawson, M.A., and Kouzarides, T. (2012). Cancer epigenetics: from mechanism to therapy. Cell
150, 12-27.
Dawson, M.A., Prinjha, R.K., Dittmann, A., Giotopoulos, G., Bantscheff, M., Chan, W.I.,
Robson, S.C., Chung, C.W., Hopf, C., Savitski, M.M., et al. (2011). Inhibition of BET
recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. Nature 478, 529533.
Delmore, J.E., Issa, G.C., Lemieux, M.E., Rahl, P.B., Shi, J., Jacobs, H.M., Kastritis, E.,
Gilpatrick, T., Paranal, R.M., Qi, J., et al. (2011). BET bromodomain inhibition as a therapeutic
strategy to target c-Myc. Cell 146, 904-917.
Dey, A., Chitsaz, F., Abbasi, A., Misteli, T., and Ozato, K. (2003). The double bromodomain
protein Brd4 binds to acetylated chromatin during interphase and mitosis. Proc Natl Acad Sci U
S A 100, 8758-8763.
Dib, A., Gabrea, A., Glebov, O.K., Bergsagel, P.L., and Kuehl, W.M. (2008). Characterization of
MYC translocations in multiple myeloma cell lines. J Natl Cancer Inst Monogr, 25-31.
Dixon, J.R., Selvaraj, S., Yue, F., Kim, A., Li, Y., Shen, Y., Hu, M., Liu, J.S., and Ren, B.
(2012). Topological domains in mammalian genomes identified by analysis of chromatin
interactions. Nature 485, 376-380.
Elsasser, S.J., Allis, C.D., and Lewis, P.W. (2011). Cancer. New epigenetic drivers of cancers.
Science 331, 1145-1146.
Esteller, M. (2008). Epigenetics in cancer. N Engl J Med 358, 1148-1159.
Feinberg, A.P., and Tycko, B. (2004). The history of cancer epigenetics. Nat Rev Cancer 4, 143153.
Felsher, D.W., and Bishop, J.M. (1999). Reversible tumorigenesis by MYC in hematopoietic
lineages. Mol Cell 4, 199-207.
106
Filippakopoulos, P., Qi, J., Picaud, S., Shen, Y., Smith, W.B., Fedorov, 0., Morse, E.M., Keates,
T., Hickman, T.T., Felletar, I., et al. (2010). Selective inhibition of BET bromodomains. Nature
468, 1067-1073.
Garraway, L.A., and Sellers, W.R. (2006). Lineage dependency and lineage-survival oncogenes
in human cancer. Nat Rev Cancer 6, 593-602.
Garraway, L.A., Widlund, H.R., Rubin, M.A., Getz, G., Berger, A.J., Ramaswamy, S.,
Beroukhim, R., Milner, D.A., Granter, S.R., Du, J., et al. (2005). Integrative genomic analyses
identify MITF as a lineage survival oncogene amplified in malignant melanoma. Nature 436,
117-122.
Geutjes, E.J., Bajpe, P.K., and Bernards, R. (2012). Targeting the epigenome for treatment of
cancer. Oncogene 31, 3827-3844.
Giese, K., Kingsley, C., Kirshner, J.R., and Grosschedl, R. (1995). Assembly and function of a
TCR alpha enhancer complex is dependent on LEF- 1-induced DNA bending and multiple
protein-protein interactions. Genes Dev 9, 995-1008.
Giniger, E., and Ptashne, M. (1988). Cooperative DNA binding of the yeast transcriptional
activator GAL4. Proc Natl Acad Sci U S A 85, 382-386.
Gondor, A., and Ohlsson, R. (2009). Chromosome crosstalk in three dimensions. Nature 461,
212-217.
Griggs, D.W., and Johnston, M. (1991). Regulated expression of the GAL4 activator gene in
yeast provides a sensitive genetic switch for glucose repression. Proc Natl Acad Sci U S A 88,
8597-8601.
Heintzman, N.D., Hon, G.C., Hawkins, R.D., Kheradpour, P., Stark, A., Harp, L.F., Ye, Z., Lee,
L.K., Stuart, R.K., Ching, C.W., et al. (2009). Histone modifications at human enhancers reflect
global cell-type-specific gene expression. Nature 459, 108-112.
Holien, T., Vatsveen, T.K., Hella, H., Waage, A., and Sundan, A. (2012). Addiction to c-MYC in
multiple myeloma. Blood 120, 2450-2453.
Issa, J.P., and Kantarjian, H.M. (2009). Targeting DNA methylation. Clin Cancer Res 15, 39383946.
Jain, M., Arvanitis, C., Chu, K., Dewey, W., Leonhardt, E., Trinh, M., Sundberg, C.D., Bishop,
J.M., and Felsher, D.W. (2002). Sustained loss of a neoplastic phenotype by brief inactivation of
MYC. Science 297, 102-104.
Jang, M.K., Mochizuki, K., Zhou, M., Jeong, H.S., Brady, J.N., and Ozato, K. (2005). The
bromodomain protein Brd4 is a positive regulatory component of P-TEFb and stimulates RNA
polymerase II-dependent transcription. Mol Cell 19, 523-534.
107
Jiang, Y.W., Veschambre, P., Erdjument-Bromage, H., Tempst, P., Conaway, J.W., Conaway,
R.C., and Kornberg, R.D. (1998). Mammalian mediator of transcriptional regulation and its
possible role as an end-point of signal transduction pathways. Proc Natl Acad Sci U S A 95,
8538-8543.
Kagey, M.H., Newman, J.J., Bilodeau, S., Zhan, Y., Orlando, D.A., van Berkum, N.L., Ebmeier,
C.C., Goossens, J., Rahl, P.B., Levine, S.S., et al. (2010). Mediator and cohesin connect gene
expression and chromatin architecture. Nature 467, 430-435.
Kim, T.K., and Maniatis, T. (1997). The mechanism of transcriptional synergy of an in vitro
assembled interferon-beta enhanceosome. Mol Cell 1, 119-129.
Kornberg, R.D. (2005). Mediator and the mechanism of transcriptional activation. Trends
Biochem Sci 30, 235-239.
Krueger, B.J., Varzavand, K., Cooper, J.J., and Price, D.H. (2010). The mechanism of release of
P-TEFb and HEXIM 1 from the 7SK snRNP by viral and cellular activators includes a
conformational change in 7SK. PLoS One 5, e12335.
Langmead, B., Trapnell, C., Pop, M., and Salzberg, S.L. (2009). Ultrafast and memory-efficient
alignment of short DNA sequences to the human genome. Genome Biol 10, R25.
Lelli, K.M., Slattery, M., and Mann, R.S. (2012). Disentangling the many layers of eukaryotic
transcriptional regulation. Annu Rev Genet 46, 43-68.
LeRoy, G., Rickards, B., and Flint, S.J. (2008). The double bromodomain proteins Brd2 and
Brd3 couple histone acetylation to transcription. Mol Cell 30, 51-60.
Lin, C.Y., Loven, J., Rahl, P.B., Paranal, R.M., Burge, C.B., Bradner, J.E., Lee, T.I., and Young,
R.A. (2012). Transcriptional Amplification in Tumor Cells with Elevated c-Myc. Cell 151, 5667.
Maldonado, V., and Melendez-Zajgla, J. (2011). Role of Bcl-3 in solid tumors. Mol Cancer 10,
152.
Malik, S., and Roeder, R.G. (2010). The metazoan Mediator co-activator complex as an
integrative hub for transcriptional regulation. Nat Rev Genet 11, 761-772.
Marks, P.A., and Xu, W.S. (2009). Histone deacetylase inhibitors: Potential in cancer therapy. J
Cell Biochem 107, 600-608.
Mertz, J.A., Conery, A.R., Bryant, B.M., Sandy, P., Balasubramanian, S., Mele, D.A., Bergeron,
L., and Sims, R.J., 3rd (2011). Targeting MYC dependence in cancer by inhibiting BET
bromodomains. Proc Natl Acad Sci U S A 108, 16669-16674.
Musgrove, E.A., Caldon, C.E., Barraclough, J., Stone, A., and Sutherland, R.L. (2011). Cyclin D
as a therapeutic target in cancer. Nat Rev Cancer 11, 558-572.
108
Nicodeme, E., Jeffrey, K.L., Schaefer, U., Beinke, S., Dewell, S., Chung, C.W., Chandwani, R.,
Marazzi, I., Wilson, P., Coste, H., et al. (2010). Suppression of inflammation by a synthetic
histone mimic. Nature 468, 1119-1123.
Ong, C.T., and Corces, V.G. (2011). Enhancer function: new insights into the regulation of
tissue-specific gene expression. Nat Rev Genet 12, 283-293.
Ott, C.J., Kopp, N., Bird, L., Paranal, R.M., Qi, J., Bowman, T., Rodig, S.J., Kung, A.L.,
Bradner, J.E., and Weinstock, D.M. (2012). BET bromodomain inhibition targets both c-Myc
and IL7R in high-risk acute lymphoblastic leukemia. Blood 120, 2843-2852.
Pedersen, N., Mortensen, S., Sorensen, S.B., Pedersen, M.W., Rieneck, K., Bovin, L.F., and
Poulsen, H.S. (2003). Transcriptional gene expression profiling of small cell lung cancer cells.
Cancer Res 63, 1943-1953.
Perk, J., lavarone, A., and Benezra, R. (2005). Id family of helix-loop-helix proteins in cancer.
Nat Rev Cancer 5, 603-614.
Rada-Iglesias, A., Bajpai, R., Swigut, T., Brugmann, S.A., Flynn, R.A., and Wysocka, J. (2011).
A unique chromatin signature uncovers early developmental enhancers in humans. Nature 470,
279-283.
Rahman, S., Sowa, M.E., Ottinger, M., Smith, J.A., Shi, Y., Harper, J.W., and Howley, P.M.
(2011). The Brd4 extraterminal domain confers transcription activation independent of pTEFb by
recruiting multiple proteins, including NSD3. Mol Cell Biol 31, 2641-2652.
Reimold, A.M., Iwakoshi, N.N., Manis, J., Vallabhajosyula, P., Szomolanyi-Tsuda, E.,
Gravallese, E.M., Friend, D., Grusby, M.J., Alt, F., and Glimcher, L.H. (2001). Plasma cell
differentiation requires the transcription factor XBP-1. Nature 412, 300-307.
Shaffer, A.L., Emre, N.C., Lamy, L., Ngo, V.N., Wright, G., Xiao, W., Powell, J., Dave, S., Yu,
X., Zhao, H., el al. (2008). IRF4 addiction in multiple myeloma. Nature 454, 226-23 1.
Shah, N., Pang, B., Yeoh, K.G., Thorn, S., Chen, C.S., Lilly, M.B., and Salto-Tellez, M. (2008).
Potential roles for the PIM 1 kinase in human cancer - a molecular and therapeutic appraisal. Eur
J Cancer 44, 2144-2151.
Shapiro-Shelef, M., Lin, K.I., McHeyzer-Williams, L.J., Liao, J., McHeyzer-Williams, M.G.,
and Calame, K. (2003). Blimp- 1 is required for the formation of immunoglobulin secreting
plasma cells and pre-plasma memory B cells. Immunity 19, 607-620.
Shou, Y., Martelli, M.L., Gabrea, A., Qi, Y., Brents, L.A., Roschke, A., Dewald, G., Kirsch, I.R.,
Bergsagel, P.L., and Kuehl, W.M. (2000). Diverse karyotypic abnormalities of the c-myc locus
associated with c-myc dysregulation and tumor progression in multiple myeloma. Proc Natl
Acad Sci U S A 97, 228-233.
Spitz, F., and Furlong, E.E. (2012). Transcription factors: from enhancer binding to
developmental control. Nat Rev Genet 13, 613-626.
109
Taatjes, D.J. (2010). The human Mediator complex: a versatile, genome-wide regulator of
transcription. Trends Biochem Sci 35, 315-322.
Thanos, D., and Maniatis, T. (1995). Virus induction of human IFN beta gene expression
requires the assembly of an enhanceosome. Cell 83, 1091-1100.
Turner, C.A., Jr., Mack, D.H., and Davis, M.M. (1994). Blimp-1, a novel zinc finger-containing
protein that can drive the maturation of B lymphocytes into immunoglobulin-secreting cells. Cell
77, 297-306.
Weinstein, I.B. (2002). Cancer. Addiction to oncogenes--the Achilles heal of cancer. Science
297, 63-64.
Wu, S.Y., and Chiang, C.M. (2007). The double bromodomain-containing chromatin adaptor
Brd4 and transcriptional regulation. J Biol Chem 282, 13141-13145.
Wu, S.Y., Lee, A.Y., Lai, H.T., Zhang, H., and Chiang, C.M. (2013). Phospho Switch Triggers
Brd4 Chromatin Binding and Activator Recruitment for Gene-Specific Targeting. Mol Cell.
Wu, S.Y., Zhou, T., and Chiang, C.M. (2003). Human mediator enhances activator-facilitated
recruitment of RNA polymerase II and promoter recognition by TATA-binding protein (TBP)
independently of TBP-associated factors. Mol Cell Biol 23, 6229-6242.
Yang, Z., Yik, J.H., Chen, R., He, N., Jang, M.K., Ozato, K., and Zhou, Q. (2005). Recruitment
of P-TEFb for stimulation of transcriptional elongation by the bromodomain protein Brd4. Mol
Cell 19, 535-545.
You, J.S., and Jones, P.A. (2012). Cancer genetics and epigenetics: two sides of the same coin?
Cancer Cell 22, 9-20.
Zhang, W., Prakash, C., Sum, C., Gong, Y., Li, Y., Kwok, J.J., Thiessen, N., Pettersson, S.,
Jones, S.J., Knapp, S., el al. (2012). Bromodomain-containing protein 4 (BRD4) regulates RNA
polymerase II serine 2 phosphorylation in human CD4+ T cells. J Biol Chem 287, 43137-43155.
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.
Zuber, J., Shi, J., Wang, E., Rappaport, A.R., Herrmann, H., Sison, E.A., Magoon, D., Qi, J.,
Blatt, K., Wunderlich, M., et al. (2011). RNAi screen identifies Brd4 as a therapeutic target in
acute myeloid leukaemia. Nature 478, 524-528.
110
Chapter 3
Conclusions and future directions
Transcriptional enhancer regions play a critical role in human evolution, development,
and disease. In the previous chapters, I discussed the functional properties of enhancer regions,
how they influence transcription, and how these regions can be misregulated in disease. I
described the identification of super-enhancers, and how these hubs of transcriptional regulation
may be targets for therapeutic intervention in cancer. In this chapter, I describe further how
super-enhancer regions may be utilized to identify key genes and to map the core regulatory
circuitry in both healthy and diseased tissue, and how these enhancer regions may play an
important role in the mechanism of action of drugs targeting transcription. Given the increasing
promise of many inhibitors against chromatin regulators and transcriptional cofactors,
understanding how super-enhancer regions influence drug response may inform and improve our
ability to use these drugs in the treatment of human disease.
Super-enhancers mark key genes
Markers of key genes in healthy tissue
In both healthy and diseased tissues, super-enhancers are associated with many key genes
that are known to control and define cell state. These include both transcription factors involved
in directing cell-state-specific gene expression programs as well as genes involved directly in
cell function. For example, in heart tissue, super-enhancer-associated genes include three
transcription factors known to be key regulators of cardiac development (GA TA4, NKX2.5, and
TBX5), as well as several genes involved in cardiac muscle function, such as MYH6, a cardiac111
specific component of the motor protein, myosin; myoglobin (MB), the oxygen carrying protein
in muscle cells; and KCNQJ, a voltage-gated potassium channel involved in repolarization of
cardiac muscle (Hnisz et al., 2013; McCulley and Black, 2012; Nerbonne and Kass, 2005). In
another example, the nuclear receptor transcription factor PPARG and C/EBP transcription factor
family members, including CEBPB and CEBPD,are important regulators of adipocyte
differentiation, and are associated with super-enhancers in adipose tissue, as is lipoprotein lipase
(LPL), an enzyme involved in the import of fatty acids for storage in adipose tissue (Hnisz et al.,
2013; Rosen and MacDougald, 2006; Wang and Eckel, 2009). Thus, identification of superenhancers in cell or tissue types where the functional and regulatory characteristics are less well
understood may help us to uncover the basic biology of these tissues.
Super-enhancers mark key genes in cancer
Super-enhancers are also associated with key genes in cancer cells. Many known
oncogenes, anti-apoptotic genes, cell cycle regulators, as well as genes involved in functions
associated with the cell or origin are driven by super-enhancer regions. For example, in multiple
myeloma, a cancer originating from antibody-producing plasma cells, the immunoglobulin light
chain gene IGLL5 is associated with a super-enhancer, as is the transcription factor IRF4, which
is involved in plasma cell development and is also important in multiple myeloma biology
(Loven et al., 2013). In addition, many cancer-related genes appear to acquire super-enhancers.
For example, anti-apoptotic factors, such as the Bcl2 family member BCL2LJ (Bcl-xL), are
associated with a super-enhancer in T cell acute lymphoblastic leukemia (T-ALL), but not
normal T cells (Hnisz et al., 2013). Genes for cell cycle regulators, such as cyclins, may also
acquire super-enhancers: in colon carcinoma, CCNDJ (cyclin Dl) gains a super-enhancer, while
in T-ALL, CCND3 (cyclin D3) becomes super-enhancer-associated (Hnisz et al., 2013). These
112
classes of genes are also associated with super-enhancers in multiple myeloma, with superenhancers at both BCL2 and BCL2LJ as well as CCND2 (cyclin D2) (Loven et al., 2013).
Acquired super-enhancers in tumor cells often fall into categories of genes involved in cancer
hallmark traits, such as immune avoidance, induction of angiogenesis, evasion of growth
suppressors, or activation of invasion or metastasis (Hnisz et al., 2013). For example, the
highest-ranked super-enhancer acquired in the colon carcinoma cell line HCT- 116 is PCDH7,
protocadherin-7, an integral membrane protein involved in cell adhesion that has been strongly
linked to metastasis in breast cancer (Bos et al., 2009; Hnisz et al., 2013; Li et al., 2013; Yoshida
et al., 1998). Therefore, profiling of super-enhancers in cancer cells may help to identify key
genes involved in tumor function and pathogenesis.
The identification of super-enhancers in tumor cells may improve our understanding of
tumor cell function, and also suggest existing or novel therapeutic targets. In glioblastoma
multiforme (GBM), for example, the super-enhancer-associated vascular endothelial growth
factor (VEGFA) is targeted by the clinically approved monoclonal antibody Bevacizumab (brand
name Avastin) (Friedman et al., 2009; Hnisz et al., 2013). In acute myeloid leukemia (AML),
the highest-ranked super-enhancer is associated with CD47, which encodes a transmembrane
protein that acts as a "don't eat me" signal to macrophages, allowing tumor cells to avoid
immune destruction (analysis unpublished, data available from (Shi et al., 2013); (Grimsley and
Ravichandran, 2003)). Monoclonal antibodies targeting CD47 have shown promise in models of
human AML and may soon be investigated clinically (Majeti et al., 2009). Thus, identifying the
roughly 100-1000 super-enhancer-associated genes in a given tumor type may help to reduce the
search space when investigating potential therapeutic targets.
Identifying super-enhancersin patient samples
113
Typical ChIP-seq experiments, such as those thus far used to identify super-enhancers,
require the use of as many as 10 million cells, far more than possible to obtain from typical
patient tumor samples. Continuing improvements in ChIP-seq technology, allowing for the use
of small sample sizes and high throughput analysis, may improve our ability to identify superenhancers in primary tumor cells (Adli and Bernstein, 2011; Blecher-Gonen et al., 2013;
Gilfillan et al., 2012). In addition, new technologies that can be used for identifying enhancer
regions, such as ATAC-Seq, may allow for the profiling of super-enhancers in as few as 500
cells. Modification of these techniques to allow for the identification of enhancer and superenhancer regions in formalin-fixed, paraffin-embedded (FFPE) samples would also greatly
increase the ability to assess patient samples (Fanelli et al., 2011; Fanelli et al., 2010). Lastly, it
may be possible to identify super-enhancers through alternate means, such as the measurement of
RNA produced by these regions. Given that single-cell methods of RNA sequencing are
currently being investigated, this type of technology may dramatically improve our ability to use
super-enhancers to understand basic tumor biology and cellular heterogeneity in the context of
clinically relevant samples (Picelli et al., 2013; Ramskold et al., 2012).
Mapping regulatory circuitry with super-enhancers
As described previously, and in chapter 1, super-enhancers are often associated with key
transcription factors that are known to regulate cell-state-specific gene expression programs,
such as the well-characterized embryonic stem cell master transcription factors Oct4, Sox2, and
Nanog. Being able to decipher the transcriptional regulatory circuitry in less well-understood
tissue types would not only contribute to our understanding of the basic biology of these cells,
but would also be useful in applications such as the reprogramming of cells for regenerative
medicine. The human genome encodes over one thousand transcription factors, hundreds of
114
which may be expressed in a given cell type, making it difficult to predict which of these may
form a small set of master regulators (Vaquerizas et al., 2009). Using super-enhancer regions
may help to focus on the key transcription factors for a given cell type, facilitating the prediction
of core circuitry in any cell type.
In mouse embryonic stem cells, the Oct4, Sox2 and Nanog master transcription factors
form an interconnected autoregulatory loop that contributes to the regulation of the pluripotent
state (Boyer et al., 2005). In this loop, each transcription factor is directly involved in activation
of its own expression and that of the others, forming a core regulatory circuit. This type of
regulatory motif can also occur in cancer. For example, in T-ALL, the oncogenic transcription
factor TALI forms a core circuit with the transcription factors GATA3 and RUNX1, driving the
malignant state in these cells (Sanda et al., 2012). Many known master transcription factors are
driven by super-enhancers (Hnisz et al., 2013). As observed in mouse embryonic stem cells,
Oct4, Sox2 and Nanog are associated with super-enhancer regions, which are in turn occupied by
each of these transcription factors (Whyte et al., 2013). Thus, identification of super-enhancer
regions may aid in building maps of the core regulatory circuitry in other cell types by
identifying the set of transcription factors whose genes are associated with super-enhancers, and
who occupy their own super-enhancers and those of the other transcription factors in the circuit.
In lieu of obtaining ChIP-seq data for each transcription factor, it may also be possible to use the
presence of a particular transcription factor's binding motif in a super-enhancer region to
indicate binding. This type of prediction would allow us to use super-enhancers to map the core
regulatory circuitry in both healthy and diseased tissue, helping to identify the master regulators
of gene expression in any cell type.
Super-enhancers and transcriptional therapeutics
115
Disruptionof super-enhancersby bromodomain inhibitorsin multiple myeloma
In addition to using super-enhancers to identify potential therapeutic targets, these
enhancers may themselves be targets for therapeutic intervention through drugs acting on the
transcriptional regulators that function at super-enhancers. As I described in chapter 2, in
multiple myeloma, the BET bromodomain inhibitor, JQ 1, appears to function through the
selective disruption of highly BRD4-occupied super-enhancer regions, possibly due to the
cooperative and synergistic interactions between the cofactors and transcriptional apparatus
likely occurring at these large, clustered enhancer regions (Loven et al., 2013). Preferential loss
of BRD4 at these regions is accompanied by a loss of transcriptional elongation at superenhancer-associated genes and a concomitant loss in mRNA expression levels. The MYC
oncogene is associated with the large IGH super-enhancers in multiple myeloma, and
experiences a profound loss in transcription elongation, as evidenced by loss of RNA Pol II
binding in the gene body, as well as a dramatic reduction in mRNA levels. Overexpression of
MYC is capable of partially rescuing tumor cells from the effects of JQ 1 treatment, suggesting
that the loss of this potent oncogene plays an important role in JQ 1 response, although it is likely
that other factors are involved (Delmore et al., 2011). Because super-enhancers are associated
with many genes involved in tumor pathogenesis, it is likely that loss of several of these
important genes may contribute to the effect of JQ 1. These results are significant, in part
because of the difficulty of directly targeting an oncogenic transcription factor such as c-MYC
with conventional small molecule inhibitors. Despite the fact that JQ 1 targets transcription
elongation, a general transcriptional process, this inhibitor appears to have a relatively selective
effect on gene expression, with a profound effect on expression of the MYC oncogene in
multiple myeloma (Delmore et al., 2011; Loven et al., 2013).
116
Bromodomain inhibitorsact on super-enhancersin many tumor types
The mechanism described in chapter 2 in multiple myeloma may be a general principle
for understanding the mechanism of action of bromodomain inhibitors in other cell types. For
example, similar results were found in diffuse large B cell lymphoma (DLBCL), where MYC and
several other tumor-associated genes, including BCL6 and POU2AF1 (OCA-B), are associated
with super-enhancers, and are also down-regulated by JQl treatment (Chapuy et al., 2013).
Similarly, in AML, the EVIl gene is driven by the GA TA2 super-enhancer through a
translocation event, and both the enhancer region and the EVIl gene are profoundly affected by
JQ1, experiencing a loss in BRD4 occupancy as well as transcription (Groeschel et al., 2014). In
other AML cell lines, where EVII was expressed, but not occupied by a super-enhancer
containing such high levels of BRD4, the gene was not sensitive to JQ 1 treatment (Groeschel et
al., 2014). Therefore, bromodomain inhibitors may act through the disruption of superenhancers in many other cancer types in addition to multiple myeloma, as described in chapter 2.
The role ofsuper-enhancersin other transcriptionaltherapeutics
As described in previous chapters, there is currently great interest in developing small
molecule inhibitors targeted against factors involved in the transcription cycle. In addition to
bromodomain inhibitors, such as JQ1, many inhibitors target other chromatin regulators,
including histone deacetylases (HDACs), histone methyl transferases such as EZH2 or DOT1 L,
and histone demethylases such as LSD 1. In addition, many drugs are being developed to target
kinases involved in the transcriptional cycle, including CDK7, CDK8, and CDK9 (Stellrecht and
Chen, 2011; Villicana et al., 2014). Understanding how these drugs affect transcription will be
117
important in making further improvements in designing these inhibitors and employing them in
cancer therapy.
A major challenge in the use of these therapeutics is understanding how transcriptional
inhibition can achieve a favorable therapeutic index, allowing for the specific targeting of tumor
cells. Several mechanisms have been suggested by which these transcriptional inhibitors may act
on tumor cells. First, tumor cells could be generally dependent on high levels of transcription, as
evidenced by the transcriptional amplification effect induced by over expression of the c-MYC
oncogene. Second, many oncogenes have short-lived mRNA and protein products, so these may
be among the first, and most profoundly affected genes in response to inhibition of transcription.
For example, the c-MYC oncogene produces exceptionally short-lived products, with both
mRNA and protein half-lives of around 30 minutes (Dani et al., 1984; Ramsay et al., 1984).
However, our evidence suggests that certain transcriptional inhibitors, such as JQ 1, may not act
by profoundly downregulating global transcription, but rather function by targeting superenhancer regions in the genome, leading to a more selective down-regulation of specific genes.
As described previously, JQ 1 treatment induces defects in transcription elongation that are more
pronounced at super-enhancer-associated genes, suggesting that the disruption of super-enhancer
regions has a role in JQ 1's action, and mRNA and protein half-life alone may not account for the
selective effect of this inhibitor (Loven et al., 2013). At a cellular level, tumor cells may become
addicted or dependent on certain oncogenes or genes involved in other tumor-promoting
pathways, as described above (Luo et al., 2009). Because super-enhancers are associated with
many of these key categories of tumor genes, their disruption by JQ 1 may lead to adverse effects
on tumor cells.
118
This mechanism may extend to other transcriptional inhibitors. For example, inhibition
of the Notch transcription factor by a gamma secretase inhibitor (GSI) produced similar results in
T-lymphoblastic leukemia (T-LL), where super-enhancer regions and their associated genes were
the most affected by GSI treatment (Wang et al., 2014). Thus, super-enhancers may represent
regions of the genome with heightened requirements for transcriptional activators, and
heightened sensitivity to inhibition of these factors. ChIP-Seq profiling of transcriptional drug
targets, and identification of super-enhancers based on these factors, may help to key in on
genomic regions and genes that may be particularly sensitive to inhibitor treatment. In addition,
the recent development of Chem-Seq technology allows the profiling of drug-DNA interactions,
which may aid in finding those regions of the genome that are highly targeted by transcriptional
inhibitors (Anders et al., 2014). Examining the effect of drug treatment on transcription itself,
either by profiling RNA Pol II occupancy, or nascent transcription by GRO-Seq, may also
improve our understanding of how these drugs act (Core et al., 2008). Measuring transcription
directly rather than measuring the resultant steady-state mRNA levels may reveal any selective
effects of these inhibitors that are separate from the half-life of the gene products. This further
investigation of the many promising transcriptional inhibitors should indicate the extent to which
super-enhancers play a common role in the mechanism of action of each of the varying classes of
these drugs.
Conclusion
In summary, super-enhancer regions are associated with key genes that have important
roles in cell function and in the control of cell-type-specific gene expression programs. In
addition, super-enhancer regions can be particularly sensitive to perturbation by inhibitors
targeting transcription. Super-enhancers may therefore act as biomarkers, allowing us to identify
119
genes that play critical roles in cellular function, both in healthy tissues and in tumor cells. In
addition, by identifying transcription factors that are both driven by super-enhancers, and act at
super-enhancers, we may be able to map the core regulatory circuitry in any human cell type.
Thus, profiling super-enhancers could contribute to our knowledge of the basic function and
regulation of a particular cell or tissue type, and in tumor cells, may suggest important targets for
therapeutic intervention. Super-enhancers themselves may play an important role in the
mechanism of action of transcriptional inhibitors, a major new category of potential cancer
therapeutics. Better understanding of the basic properties of super-enhancers, their location in
healthy and diseased tissues, and how these regions are affected by transcriptional inhibitors may
lead to an improved ability to characterize and treat human disease.
120
References
Adli, M., and Bernstein, B.E. (2011). Whole-genome chromatin profiling from limited numbers
of cells using nano-ChIP-seq. Nat Protoc 6, 1656-1668.
Anders, L., Guenther, M.G., Qi, J., Fan, Z.P., Marineau, J.J., Rahl, P.B., Loven, J., Sigova, A.A.,
Smith, W.B., Lee, T.I., et al. (2014). Genome-wide localization of small molecules. Nat
Biotechnol 32, 92-96.
Blecher-Gonen, R., Barnett-Itzhaki, Z., Jaitin, D., Amann-Zalcenstein, D., Lara-Astiaso, D., and
Amit, I. (2013). High-throughput chromatin immunoprecipitation for genome-wide mapping of
in vivo protein-DNA interactions and epigenomic states. Nat Protoc 8, 539-554.
Bos, P.D., Zhang, X.H., Nadal, C., Shu, W., Gomis, R.R., Nguyen, D.X., Minn, A.J., van de
Vijver, M.J., Gerald, W.L., Foekens, J.A., et al. (2009). Genes that mediate breast cancer
metastasis to the brain. Nature 459, 1005-1009.
Boyer, L.A., Lee, T.I., Cole, M.F., Johnstone, S.E., Levine, S.S., Zucker, J.P., Guenther, M.G.,
Kumar, R.M., Murray, H.L., Jenner, R.G., et al. (2005). Core transcriptional regulatory circuitry
in human embryonic stem cells. Cell 122, 947-956.
Chapuy, B., McKeown, M.R., Lin, C.Y., Monti, S., Roemer, M.G., Qi, J., Rahl, P.B., Sun, H.H.,
Yeda, K.T., Doench, J.G., el al. (2013). Discovery and characterization of super-enhancerassociated dependencies in diffuse large B cell lymphoma. Cancer Cell 24, 777-790.
Core, L.J., Waterfall, J.J., and Lis, J.T. (2008). Nascent RNA sequencing reveals widespread
pausing and divergent initiation at human promoters. Science 322, 1845-1848.
Dani, C., Blanchard, J.M., Piechaczyk, M., El Sabouty, S., Marty, L., and Jeanteur, P. (1984).
Extreme instability of myc mRNA in normal and transformed human cells. Proc Natl Acad Sci U
S A 81, 7046-7050.
Delmore, J.E., Issa, G.C., Lemieux, M.E., Rahl, P.B., Shi, J., Jacobs, H.M., Kastritis, E.,
Gilpatrick, T., Paranal, R.M., Qi, J., et al. (2011). BET bromodomain inhibition as a therapeutic
strategy to target c-Myc. Cell 146, 904-917.
Fanelli, M., Amatori, S., Barozzi, I., and Minucci, S. (2011). Chromatin immunoprecipitation
and high-throughput sequencing from paraffin-embedded pathology tissue. Nat Protoc 6, 19051919.
Fanelli, M., Amatori, S., Barozzi, I., Soncini, M., Dal Zuffo, R., Bucci, G., Capra, M., Quarto,
M., Dellino, G.I., Mercurio, C., et al. (2010). Pathology tissue-chromatin immunoprecipitation,
coupled with high-throughput sequencing, allows the epigenetic profiling of patient samples.
Proc Natl Acad Sci U S A 107, 21535-21540.
121
Friedman, H.S., Prados, M.D., Wen, P.Y., Mikkelsen, T., Schiff, D., Abrey, L.E., Yung, W.K.,
Paleologos, N., Nicholas, M.K., Jensen, R., et al. (2009). Bevacizumab alone and in combination
with irinotecan in recurrent glioblastoma. J Clin Oncol 27, 4733-4740.
Gilfillan, G.D., Hughes, T., Sheng, Y., Hjorthaug, H.S., Straub, T., Gervin, K., Harris, J.R.,
Undlien, D.E., and Lyle, R. (2012). Limitations and possibilities of low cell number ChIP-seq.
BMC Genomics 13, 645.
Grimsley, C., and Ravichandran, K.S. (2003). Cues for apoptotic cell engulfmnent: eat-me, don't
eat-me and come-get-me signals. Trends Cell Biol 13, 648-656.
Groschel, S., Sanders, M.A., Hoogenboezem, R., de Wit, E., Bouwman, B.A., Erpelinck, C., van
der Velden, V.H., Havermans, M., Avellino, R., van Lom, K., et al. (2014). A Single Oncogenic
Enhancer Rearrangement Causes Concomitant EVIL and GATA2 Deregulation in Leukemia.
Cell.
Hnisz, D., Abraham, B.J., Lee, T.I., Lau, A., Saint-Andre, V., Sigova, A.A., Hoke, H.A., and
Young, R.A. (2013). Super-enhancers in the control of cell identity and disease. Cell 155, 934947.
Li, A.M., Tian, A.X., Zhang, R.X., Ge, J., Sun, X., and Cao, X.C. (2013). Protocadherin-7
induces bone metastasis of breast cancer. Biochem Biophys Res Commun 436, 486-490.
Loven, J., Hoke, H.A., Lin, C.Y., Lau, A., Orlando, D.A., Vakoc, C.R., Bradner, J.E., Lee, T.I.,
and Young, R.A. (2013). Selective inhibition of tumor oncogenes by disruption of superenhancers. Cell 153, 320-334.
Luo, J., Solimini, N.L., and Elledge, S.J. (2009). Principles of cancer therapy: oncogene and nononcogene addiction. Cell 136, 823-837.
Majeti, R., Chao, M.P., Alizadeh, A.A., Pang, W.W., Jaiswal, S., Gibbs, K.D., Jr., van Rooijen,
N., and Weissman, I.L. (2009). CD47 is an adverse prognostic factor and therapeutic antibody
target on human acute myeloid leukemia stem cells. Cell 138, 286-299.
McCulley, D.J., and Black, B.L. (2012). Transcription factor pathways and congenital heart
disease. Curr Top Dev Biol 100, 253-277.
Nerbonne, J.M., and Kass, R.S. (2005). Molecular physiology of cardiac repolarization. Physiol
Rev 85, 1205-1253.
Picelli, S., Bjorklund, A.K., Faridani, O.R., Sagasser, S., Winberg, G., and Sandberg, R. (2013).
Smart-seq2 for sensitive full-length transcriptome profiling in single cells. Nat Methods 10,
1096-1098.
Ramsay, G., Evan, G.I., and Bishop, J.M. (1984). The protein encoded by the human protooncogene c-myc. Proc Natl Acad Sci U S A 81, 7742-7746.
122
Ramskold, D., Luo, S., Wang, Y.C., Li, R., Deng, Q., Faridani, O.R., Daniels, G.A.,
Khrebtukova, I., Loring, J.F., Laurent, L.C., et al. (2012). Full-length mRNA-Seq from singlecell levels of RNA and individual circulating tumor cells. Nat Biotechnol 30, 777-782.
Rosen, E.D., and MacDougald, O.A. (2006). Adipocyte differentiation from the inside out. Nat
Rev Mol Cell Biol 7, 885-896.
Sanda, T., Lawton, L.N., Barrasa, M.I., Fan, Z.P., Kohlhammer, H., Gutierrez, A., Ma, W.,
Tatarek, J., Ahn, Y., Kelliher, M.A., et al. (2012). Core transcriptional regulatory circuit
controlled by the TALl complex in human T cell acute lymphoblastic leukemia. Cancer Cell 22,
209-221.
Shi, J., Whyte, W.A., Zepeda-Mendoza, C.J., Milazzo, J.P., Shen, C., Roe, J.S., Minder, J.L.,
Mercan, F., Wang, E., Eckersley-Maslin, M.A., el al. (2013). Role of SWI/SNF in acute
leukemia maintenance and enhancer-mediated Myc regulation. Genes Dev 27, 2648-2662.
Stellrecht, C.M., and Chen, L.S. (2011). Transcription inhibition as a therapeutic target for
cancer. Cancers (Basel) 3, 4170-4190.
Vaquerizas, J.M., Kummerfeld, S.K., Teichmann, S.A., and Luscombe, N.M. (2009). A census
of human transcription factors: function, expression and evolution. Nat Rev Genet 10, 252-263.
Villicana, C., Cruz, G., and Zurita, M. (2014). The basal transcription machinery as a target for
cancer therapy. Cancer Cell Int 14, 18.
Wang, H., and Eckel, R.H. (2009). Lipoprotein lipase: from gene to obesity. Am J Physiol
Endocrinol Metab 297, E271-288.
Wang, H., Zang, C., Taing, L., Arnett, K.L., Wong, Y.J., Pear, W.S., Blacklow, S.C., Liu, X.S.,
and Aster, J.C. (2014). NOTCH 1-RBPJ complexes drive target gene expression through dynamic
interactions with superenhancers. Proc Natl Acad Sci U S A 111, 705-710.
Whyte, W.A., Orlando, D.A., Hnisz, D., Abraham, B.J., Lin, C.Y., Kagey, M.H., Rahl, P.B., Lee,
T.I., and Young, R.A. (2013). Master transcription factors and mediator establish superenhancers at key cell identity genes. Cell 153, 307-319.
Yoshida, K., Yoshitomo-Nakagawa, K., Seki, N., Sasaki, M., and Sugano, S. (1998). Cloning,
expression analysis, and chromosomal localization of BH-protocadherin (PCDH7), a novel
member of the cadherin superfamily. Genomics 49, 458-461.
123
Appendix A
Supplementary material for chapter 2
SUPPLEMENTAL INFORMATION
Supplemental Data
Table SI: Actively transcribed genes in studied cell lines
Table S2: Active enhancers in multiple myeloma
Table S3: Super-enhancer-associated genes in multiple myeloma
Table S4: Active enhancers in glioblastoma multiforme and small cell lung
cancer
Table S5: Super-enhancer-associated genes in glioblastoma multiforme and
small cell lung cancer
Table S6: List of primers used to clone enhancer regions
Table S7: ChIP-Seq datasets and backgrounds
Data S1: Enhancer and super-enhancer .bed file
Figure S1: Mediator and BRD4 co-occupy promoters of active genes in
multiple myeloma
Figure S2: Super-enhancers identified in multiple myeloma
Figure S3: Transcription of super-enhancer-associated
genes is highly
sensitive to BRD4 inhibition
Figure S4: Super-enhancers are associated with key genes in other cancers
Supplemental Experimental Procedures
124
Cell Culture
Chromatin Immunoprecipitation (ChIP)
Illumina Sequencing and Library Generation
Luciferase Reporter Assays
Cell Viability Assays
Western Blotting
RNA Extraction and Synthetic RNA Spike-In
Microarray Sample Preparation and Analysis
Data Analysis
Supplemental References
125
Tables and Figures
Table S1, related to Figures 1, 2, 4, 6 and 7: Actively transcribed genes in multiple
myeloma
Table of actively transcribed genes in the multiple myeloma cell line MM 1 .S.
Table S2, related to Figures 1, 2, 3, 4, 5, and 6: Active enhancers in multiple myeloma
Table of active enhancers in the multiple myeloma cell line MM1.S. Genomic coordinates are
provided for all identified enhancers. Enhancers that have been classified as super-enhancers are
indicated.
Total background subtracted mediator (MED1) signal at each enhancer is also
provided.
Table S3, related to Figures 2, 3, and 6: Super-enhancer-associated genes in multiple
myeloma
Table of super-enhancer-associated actively transcribed genes in the multiple myeloma cell line
MML.S (see supplemental methods). For each gene, the coordinates of the proximal superenhancers are provided. Genes that are annotated as altered in cancer in either (Patel et al., 2013)
or (Futreal et al., 2004) are indicated.
Table S4, related to Figure 7: Active enhancers in glioblastoma multiforme and small cell
lung cancer
Table of active enhancers in the glioblastoma multiforme (GBM) cell line U-87 and the small
cell lung cancer (SCLC) cell line H2171. Genomic coordinates are provided for all identified
enhancers.
Enhancers that have been classified as super-enhancers are indicated.
background subtracted mediator (MED 1) signal at each enhancer is also provided.
126
Total
Table S5, related to Figure 7: Super-enhancer-associated genes in glioblastoma multiforme
and small cell lung cancer
Table of super-enhancer-associated actively transcribed genes in the glioblastoma multiforme
(GBM) cell line U-87 and the small cell lung cancer (SCLC) cell line H2171 (see supplemental
methods). For each gene, the coordinates of the proximal super-enhancers are provided. Genes
that are annotated as altered in cancer in either (Patel et al., 2013) or (Futreal et al., 2004) are
indicated.
Table S6, related to Figure 2 and Figure 6: Primers used to clone enhancer regions
This table lists primer sequences for all cloned regions used in reporter assays.
Table S7, related to Figures 1 through Figure 7: ChIP-Seq datasets and backgrounds
This table lists all ChIP-Seq datasets used in this study, their GEO accession number, and the
appropriate background used to define enriched regions.
Data Si: Enhancer and super-enhancer .bed file
This file is a UCSC genome browser .bed format file (genome.ucsc.edu) containing the genomic
coordinates for all enhancers and super-enhancers in profiled cell lines. Coordinates are in
genome build hgl 8. For each cell line, individual tracks are provided for all enhancers (colored
in black) or super-enhancers only (colored in red).
127
Figure S1, related to Figure 1: Mediator and BRD4 co-occupy promoters of active genes in
multiple myeloma
A) Scatter plots showing the relationship between MEDI (x-axis) and BRD4 (y-axis) ChIP-Seq
occupancy at individual enhancers in MML.S cells. Units are in total ChIP-Seq reads per million
in enhancer regions. The Pearson correlation score between MED 1 and BRD4 is displayed (p
=
0.91).
B) Scatter plots showing the relationship between MED1 (x-axis) and BRD4 (y-axis) ChIP-Seq
occupancy at individual promoters of actively transcribed genes in MML.S cells. Units are in
total ChIP-Seq reads per million in enhancer regions. The Pearson correlation score between
MED1 and BRD4 is displayed (p = 0.81).
C) Pie chart showing the overlap of BRD4 regions with either MEDI or H3K27Ac regions.
BRD4 regions that overlap either a MED1 or H3K27Ac enriched region are shaded in red.
BRD4 regions without any overlap are shaded in grey.
Figure S2, related to Figure 2: Super-enhancers identified in multiple myeloma
A) Left: Total MEDI ChIP-Seq signal in units of reads per million in enhancer regions for all
enhancers in MM l.S. Enhancers are ranked by increasing MED I ChIP-Seq signal. Right: Total
MED1 ChIP-Seq signal in units of reads per million in promoter regions for all promoters of
actively transcribed genes in MML.S that do not overlap with a super-enhancer. Promoters are
ranked by increasing MEDI ChIP-Seq signal. In both plots, Y-axis shows MEDI signal from 0
to 500,000.
128
Figure S3, related to Figure 6: Transcription of super-enhancer-associated genes is highly
sensitive to BRD4 inhibition
A,B) Heatmap of changes in gene expression for all actively transcribed genes after treatment
with 500nM JQ1 at various time points A) or after treatment with various doses of JQl for 6
hours B). Each line in the heatmap shows a single gene and is shaded relative to the change in
gene expression compared to control untreated cells. Genes are ranked by average fold loss
across the time course. The key on the right shows the relationship between color and change in
gene expression with green indicative of loss of gene expression.
C,D) Line graph showing the Log 2 change in gene expression vs. control cells after JQ1
treatment in a time C) or dose D) dependent manner for genes associated with multiple typical
enhancers (grey line) or genes associated with super-enhancers (red line). The Y-axis shows the
Log 2 change in gene expression of JQ1 treated vs. untreated control cells. X-axis shows time of
500nM JQ 1 treatment C) or JQ 1 treatment concentration at 6 hours D).
E) Meta-gene representation of global CDK9 occupancy at enhancers and promoters. The x-axis
shows the +/- 2.5kb region flanking either the center of enhancer regions (left) or the TSS of
active genes. The y-axis shows the average background subtracted ChIP-Seq signal in units of
rpm/bp.
F) Bar graphs showing the percentage loss of either MED I (left, red) or CDK9 (right, green) at
super-enhancer regions that are TSS proximal (left) or TSS distal (right). Error bars represent
95% confidence intervals of the mean.
G) Histograms showing the distribution of changes in RNA Pol II density in the elongating gene
body region of genes for typical enhancer-associated genes (top, grey) or super-enhancer129
associated genes (bottom, red). The median of the distribution is shown with a line. X-axis
shows the Log 2 change in RNA Pol II density in the elongating gene body region. Y-axis shows
probability density of the distribution.
The difference is distributions is significant by a
Kolmogorov-Smirnov (KS) test (p-value < 2e-16).
Figure S4, related to Figure 7: Super-enhancers are associated with key genes in other
cancers
A) Meta-gene representation of global MED1, BRD4, H3K27Ac, and H3K4Me3 occupancy at
enhancers and promoters in the glioblastoma multiforme cell line U-87. The x-axis shows the
+/- 2.5kb region flanking either the center of enhancer regions (left) or the transcription start site
(TSS) of active genes. The y-axis shows the average background subtracted ChIP-Seq signal in
units of rpm/bp.
B) Meta-gene representation of global MED1, BRD4, H3K27Ac, and H3K4Me3 occupancy at
enhancers and promoters in the small cell lung cancer cell line H2171. The x-axis shows the +/2.5kb region flanking either the center of enhancer regions (left) or the transcription start site
(TSS) of active genes. The y-axis shows the average background subtracted ChIP-Seq signal in
units of rpm/bp.
130
A
5,000
-
E
p = 0.91
500 50
c
5-
>
o
-
0.50.5
5
50
500
5,000
MED1 signal at enhancers (rpm)
B
5,000
-
p =0.81
E 5000
50CL
cc
5
Co,
0.5
0.5
5
50
500
5,000
MED1 signal at promoters (rpm)
C
18,726 BRD4 regions
BRD4 only (12%)
Overlap with MED1 or H3K27Ac
(88%)
Figure S1
131
500,000
400,000
COA300,000
-C
C
0
--
400,000 E
2
0- 300,000-
Cn
C
,
0300,000
2 00,000
-
(D
c
200,000.
CD
c 2000000)
Uj
-
CD
S100,0000-
0
2,000 4,000
6,000
w
"
j
100,000.
MI r
0.
0
8,000
2,000
4,000
6,000
Promoters ranked by increasing MED1 signal
Enhancers ranked by increasing MEDI signal
Figure S2
132
A
E
0
4O
2
2.0
CDK9
Genome-wide average
E
C Xl
-2.5kb Enhancer +2.5kb
Center
C
CI
-
Lu
cC
2,
>
-a)
i
F
20
Mi
1
500nM JQ1
B
.5
0
-4+
-20-
-20-
~ 40
-40]
treatment for (hrs)
Super-enhancer regions:
TSS
TSS
proximal
distal
0
(1)
0
2
CI
X
C
-
CI
01
CU
>
G
1.5
Median: -0.07
r
1.0
a
.;-c
OC)
-
LM(C10C)
I
0.5
00
5
C
0- 0.0
50
500 5,000
JQ1 concentration (nM)
0.0
-
-2
Change in expression vs. time
-0.2.
C
GiL
1.5
-
1.0
-0.8-
with:
Multiple
e"h"'cer
-1.0-
Super
enhancers
0
-0.4-
associated
-0.6-
Control
D
0,0.
CC
-0
" "
""
0
1
-1
Log 2 fold change in
elongating region RNA Pol
2
11
Genes with super-enhancers
y
Genes
01g
00
+2.5kb
Super-enhancer regions:
TSS
TSS
distal
proximal
Genes with typical enhancers
C
TSS
CDK9 ChIP-Seq
20
0-
0
C
MED1 ChIP-Seq
-
U(C1~~
U.0
2
-2.5kb
a)
(hi
0.5
0-
0.5
1
3
6
500nM JQ1 treatment for (hrs)
Genes
associated
-0.6-
Multiple
with.
-0.8-
Super
enhancers
-1.0-
-0 23
-1
0
1
2
Log 2 fold change in
elongating region RNA Pot I
Change in expression vs. dose
-0.4-
2
2
-0.2.
M Oh
01Lu
0.0
I Median:
DMSO 5
50
500 5,000
JQ1 concentration (nM)
Figure S3
133
A
2.0
B
Genome-wide average
MED1
.
2.
Genome-wide average
MEDi
0.o
2.0
BRD4
2.0
BRD4
2.0
H3K27Ac
2.0
H3K27Ac
15.0
H3K4Me3
10.0
H3K4Me3
-2.5kb Enhancer +2.5kb
Center
-2.5kb
TSS
-2.5kb Enhancer +2.5kb
Center
+2.5kb
Figure S4
134
-2.5kb
TSS
+2.5kb
Materials and Methods
Experimental Procedures
Cell Culture
MML.S multiple myeloma cells (CRL-2974 ATCC) and U-87 MG glioblastoma cells (HTB-14
ATCC) were purchased from ATCC. H2171 small cell lung carcinoma cells (CRL-5929 ATCC)
were kindly provided by John Minna, UT Southwestern. MM l.S and H2171 cells were
propagated in RPMI- 1640 supplemented with 10% fetal bovine serum and 1% GlutaMAX
(Invitrogen, 35050-061). U-87 MG cells were cultured in Eagle's Minimum Essential Medium
(EMEM) modified to contain Earles Balanced Salt Solution, nonessential amino acids, 2 mM Lglutamine, 1 mM sodium pyruvate, and 1500 mg/l sodium bicarbonate. Cells were grown at
37*C and 5% CO 2 .
JQ 1 was kindly provided by James Bradner, Dana Farber Cancer Institute. For JQ 1 dose
dependence experiments, cells were resuspended in fresh media containing JQ 1 (5nM, 50nM,
500nM, 5000nM) or vehicle (DMSO, 0.05%) for a duration of 6 hours, unless otherwise
indicated. For JQ 1 time course experiments, cells were resuspended in fresh media containing
500nM JQ 1 and sampled at various time points.
Chromatin Immunoprecipitation (ChIP)
135
Cells were crosslinked for 10 minutes at room temperature by the addition of one-tenth of the
volume of 11% formaldehyde solution (11% formaldehyde, 50mM Hepes pH 7.3, 100mM NaCl,
ImM EDTA pH 8.0, 0.5mM EGTA pH 8.0) to the growth media followed by 5 minutes
quenching with 100mM glycine. Cells were washed twice with PBS, then the supernatant was
aspirated and the cell pellet was flash frozen in liquid nitrogen. Frozen crosslinked cells were
stored at -80'C. 100ul of Dynal magnetic beads (Sigma) were blocked with 0.5% BSA (w/v) in
PBS. Magnetic beads were bound with lOug of the indicated antibody.
For BRD4 occupied genomic regions, we performed ChIP-Seq experiments using a Bethyl
Laboratories (A301-985A, lot A301-985A-1) antibody. The affinity purified antibody was
raised in rabbit against an epitope corresponding to amino acids 1312-1362 of human BRD4.
Antibody specificity was previously determined (Dawson et al., 2011). For MED 1
(CRSP1/TRAP220) occupied genomic regions, we performed ChIP-Seq experiments using a
Bethyl Laboratories (A300-793A, lot A300-783A) antibody. The affinity purified antibody was
raised in rabbit against an epitope corresponding to amino acids 1523-1581 mapping at the Cterminus of human MED 1. For CDK9 occupied genomic regions, we performed ChIP-Seq
experiments using a Santa Cruz Biotechnology (sc-484, lot D1612) antibody. For RNA
polymerase II occupied genomic regions, we performed ChIP-Seq experiments using a Santa
Cruz Biotechnology (sc-899, lot KOl 11) antibody. The affinity purified antibody was raised in
rabbit against an epitope mapping to the N-terminus of murine RBP 1, the largest subunit of RNA
Pol II.
136
For MM1.S, crosslinked cells were lysed with lysis buffer 1 (50mM Hepes pH 7.3, 140mM
NaCl, ImM EDTA, 10% glycerol, 0.5% NP-40, and 0.25% Triton X-100) and resuspended and
sonicated in sonication buffer (20mM Tris-HCl pH 8.0, 150mM NaCl, 2mM EDTA pH 8.0,
0.1% SDS, and 1% TritonX-100). Cells were sonicated for 10 cycles at 30 seconds each on ice
(18-21 watts) with 60 seconds on ice between cycles. Sonicated lysates were cleared and
incubated overnight at 4'C with magnetic beads bound with antibody to enrich for DNA
fragments bound by the indicated factor. Beads were washed two times with sonication buffer,
one time with sonication buffer with 500mM NaCl, one time with LiCl wash buffer (10mM Tris
pH 8.0, 1mM EDTA, 250mM LiCl, 1% NP-40,) and one time with TE with 50mMNaCl. DNA
was eluted in elution buffer (50mM Tris-HCL pH 8.0, 10mM EDTA, 1% SDS). Cross-links were
reversed overnight. RNA and protein were digested using RNAse A and Proteinase K,
respectively, and DNA was purified with phenol chloroform extraction and ethanol precipitation.
For BRD4 ChIPs in U-87 and H2171, crosslinked cells were lysed with lysis buffer 1 (50 mM
HEPES [pH 7.3], 140 mM NaCl, 1 mM EDTA, 10% glycerol, 0.5% NP-40, and 0.25% Triton X100) and washed with lysis buffer 2 (10 mM Tris-HCl [pH 8.0], 200 mM NaCl, 1 mM EDTA
[pH 8.0] and 0.5 mM EGTA [pH 8.0]). Cells were resuspended and sonicated in sonication
buffer (50 mM Tris-HCl [pH 7.5], 140 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X100, 0.1% SDS) for nine cycles at 30 s each on ice (18-21 W) with 60 s on ice between cycles.
Sonicated lysates were cleared and incubated overnight at 4*C with magnetic beads bound with
antibody to enrich for DNA fragments bound by the indicated factor. Beads were washed three
times with sonication buffer, one time with sonication buffer with 500 mM NaCl, one time with
LiCl wash buffer (20 mM Tris [pH 8.0], 1 mM EDTA, 250 mM LiCl, 0.5% NP-40, 0.5% Na-
137
deoxycholate) and one time with TE. DNA was eluted in elution buffer. Crosslinks were
reversed overnight. RNA and protein were digested using RNase A and Proteinase K,
respectively, and DNA was purified with phenol chloroform extraction and ethanol precipitation.
ChIP-Seq datasets of H3K4Me3 and H3K27Ac in MM1.S, and MEDI and H3K27Ac in U-87
MG and H2171 were previously published (Lin et al., 2012).
Illumina Sequencing and Library Generation
Purified ChIP DNA was used to prepare Illumina multiplexed sequencing libraries. Libraries for
Illumina sequencing were prepared following the Illumina TruSeqrm DNA Sample Preparation
v2 kit protocol with the following exceptions. After end-repair and A-tailing,
Immunoprecipitated DNA (-10-50ng) or Whole Cell Extract DNA (50ng) was ligated to a 1:50
dilution of Illumina Adapter Oligo Mix assigning one of 24 unique indexes in the kit to each
sample. Following ligation, libraries were amplified by 18 cycles of PCR using the HiFi NGS
Library Amplification kit from KAPA Biosystems. Amplified libraries were then size-selected
using a 2% gel cassette in the Pippin PrepTM system from Sage Science set to capture fragments
between 200 and 400 bp. Libraries were quantified by qPCR using the KAPA Biosystems
Illumina Library Quantification kit according to kit protocols. Libraries with distinct TruSeq
indexes were multiplexed by mixing at equimolar ratios and running together in a lane on the
Illumina HiSeq 2000 for 40 bases in single read mode.
Luciferase Reporter Assays
138
A minimal Myc promoter was amplified from human genomic DNA and cloned into the SacI
and Hindill sites of the pGL3 basic vector (Promega). Enhancer fragments were likewise
amplified from human genomic DNA and cloned into the BamHIl and SalI sites of the pGL3pMyc vector. All cloning primers are listed in Table S6. Constructs were transfected into MM l.S
cells using Lipofectamine 2000 (Invitrogen). The pRL-SV40 plasmid (Promega) was
cotransfected as a normalization control. Cells were incubated for 24 hours, and luciferase
activity was measured using the Dual-Luciferase Reporter Assay System (Promega). For the JQ 1
concentration course, cells were resuspended in fresh media containing various concentrations of
JQ 1 24 hours after transfection, and incubated for an additional 6 hours before harvesting.
Luminescence measurements were made using the Dual-Luciferase Reporter Assay System
(Promega) on a Wallac EnVision (Perkin Elmer) plate reader.
Cell Viability Assays
Cell viability was measured using the CellTiterGlo Assay kit (Promega, G7571). MM1.S cells
were resuspended in fresh media containing JQl (5nM, 50nM, 500nM, 1 OOOnM) or vehicle
(DMSO, 0.05%), then plated in 96-well plates at 10,000 cells/well in a volume of 1 OOuL.
Viability was measured after 6, 24, 48 and 72 hour incubations by addition of CellTiter Glo
reagent and luminescence measurement on a Tecan Safire 2 plate reader.
Western Blotting
Total cell lysates for immunoblotting were prepared by pelleting lx106 cells from each cell line
at 4'C (1,200 rpm) for 5 min using a Sorvall Legend centrifuge (Thermo Fisher Scientific). After
collecting the cells the pellets were washed IX with ice-cold IX PBS and the pellet after the
139
final wash was resuspended in RIPA lysis buffer (Sigma, R0278) containing 150 mM NaCl,
1.0% IGEPAL@ CA-630, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris [pH 8.0], 2X Halt
Protease inhibitors (Pierce, Thermo Fisher Scientific), and 2X Phosphatase inhibitor cocktail 2
(Sigma, P5726) and 3 (Sigma, P0044). Total protein lysates were collected and snap-frozen in
liquid nitrogen before being stored at -80'C. Protein concentrations were determined by using
the Bradford protein assay (Bio-Rad, #500-0006). Total cell lysates were loaded per well onto
NuPAGE NOVEX 4%-12% Bis-Tris Mini Gel (Invitrogen, Carlsbad, CA) and separated by
electrophoreses at 200 V for 1 hr. The gels were then transferred onto a PVDF membrane
(Immobilon-P; Millipore, Billerica, MA) by semi-dry transfer system (TransBlot SD; Bio-Rad)
and blocked by incubation with 5% dry milk in TBST (TBS with 0.2% Tween-20). Membranes
were probed using antibodies raised against c-Myc (Epitomics, cat. #: 1472-1), BRD4
(Epitomics, cat.#: 5716-1) or $-Actin (Sigma, clone AC-15, A5441). Chemiluminescent
detection was performed with appropriate secondary antibodies and developed using highresolution BioMax MR film (Kodak, cat. #: 870 13012).
RNA Extraction and Synthetic RNA Spike-In
Total RNA and sample preparation was performed as previously descibed (Loven et al., 2012).
Briefly, MM l.S cells were were incubated in JQI-containing media (5nM, 50nM, 500nM,
5000nM) or vehicle (DMSO, 0.05%) for a duration of 6 hours. Cell numbers were determined by
manually counting cells using C-Chip disposable hemocytometers (Digital Bio, DHC-NO 1) prior
to lysis and RNA extraction. Biological duplicates (equivalent to ten million cells per replicate)
were subsequently collected and homogenized in 1 mL of TRIzol Reagent (Life Technologies,
15596-026), purified using the mirVANA miRNA isolation kit (Ambion, AM 1560) following
140
the manufacturer's instructions and re-suspended in IOOpL nuclease-free water (Ambion,
AM9938). Total RNA was spiked-in with ERCC RNA Spike-In Mix (Ambion, 4456740), treated
with DNA-freeTM DNase I (Ambion, AM 1906) and analyzed on Agilent 2100 Bioanalyzer for
integrity. RNA with the RNA Integrity Number (RIN) above 9.8 was hybridized to GeneChip@
PrimeView Human Gene Expression Arrays (Affymetrix).
Microarray Sample Preparation and Analysis
For microarray analysis, 1 00ng of total RNA containing ERCC RNA Spike-In Mix (see above)
was used to prepare biotinylated aRNA (cRNA) according to the manufacturer's protocol (3'
IVT Express Kit, Affymetrix 901228). GeneChip arrays (Primeview, Affymetrix 901837) were
hybridized and scanned according to standard Affymetrix protocols. All samples were processed
in technical duplicate. Images were extracted with Affymetrix GeneChip Command Console
(AGCC), and analyzed using GeneChip Expression Console. A Primeview CDF that included
probe information for the ERCC controls, provided by Affymetrix, was used to generate .CEL
files. We processed the CEL files using standard tools available within the affy package in R.
The CEL files were processed with the expresso command to convert the raw probe intensities to
probeset expression values. The parameters of the expresso command were set to generate
Affymetrix MAS5-normalized probeset values. We used a loess regression to re-normalize these
MAS5 normalized probeset values, using only the spike-in probesets to fit the loess. The affy
package provides a function, loess.normalize,which will perform loess regression on a matrix of
values (defined using the parameter mat) and allows for the user to specify which subset of data
to use when fitting the loess (defined using the parameter subset, see the affy package
documentation for further details). For this application the parameters mat and subset were set as
141
the MAS5-normalized values and the row-indices of the ERCC control probesets, respectively.
The default settings for all other parameters were used. The result of this was a matrix of
expression values normalized to the control ERCC probes. The probeset values from the
duplicates were averaged together and the Log 2 fold change comparing the control to the JQI
treated samples are shown. Spike-in normalized gene expression tables can be found online
associated with the GEO Accession ID GSE44929 (www.ncbi.nlm.nih.gov/geo/).
142
Data Analysis
Accessing data generated in this manuscript
All ChIP-Seq and expression data generated in this manuscript can be found online associated
with GEO Accession ID GSE44931 (www.ncbi.nlm.nih.gov/geo/).
Gene sets and annotations
All analysis was performed using RefSeq (NCBI36/HG18) (Pruitt et al., 2007) human gene
annotations.
BRD4 expression in human tissues
Processed gene expression values for 90 distinct normal tissues were retrieved from the Human
Body Index (GSE7307). BRD4 expression values were calculated by averaging expression
values for three distinct probes that aligned to the sense strand of the BRD4 transcript. Since in
any given tissue, approximately 50% of all genes are expressed (Ramskold et al., 2009), we
classified BRD4 as expressed in a given tissue if the average expression values of its probes was
greater than the median expressed probe in the respective tissue. Using this metric, BRD4 was
determined to be expressed in 95% of all human tissues.
ChiP-Seq dataprocessing
143
All ChIP-Seq datasets were aligned using Bowtie (version 0.12.2) (Langmead et al., 2009) to
build version NCBI36/HG18 of the human genome. Alignments were performed using the
following criteria: -n2, -e70, -ml, -kI, --best. These criteria preserved only reads that mapped
uniquely to the genome with 1 or fewer mismatches. Aligned and raw data can be found online
associated with the GEO Accession ID GSE42355 (www.ncbi.nlm.nih.pov/geo/).
Calculatingreaddensity
We developed a simple method to calculate the normalized read density of a ChIP-Seq dataset in
any region. ChIP-Seq reads aligning to the region were extended by 200bp and the density of
reads per basepair (bp) was calculated. In order to eliminate PCR bias, multiple reads of the
exact same sequence aligning to a single position were collapsed into a single read. Only
positions with at least 2 overlapping extended reads contributed to the overall region density.
The density of reads in each region was normalized to the total number of million mapped reads
producing read density in units of reads per million mapped reads per bp (rpm/bp)
Identifying ChIP-Seq enrichedregions
We used the MACS version 1.4.1 (Model based analysis of ChIP-Seq) (Zhang et al., 2008) peak
finding algorithm to identify regions of ChIP-Seq enrichment over background. A p-value
threshold of enrichment of le-9 was used for all datasets. The GEO accession number and
144
background used for each dataset can be found in the accompanying supplementary file (Table
S6).
Defining actively transcribedgenes
In MML.S and U-87 cells, a gene was defined as actively transcribed if enriched regions for
H3K4me3, RNA Polymerase II (RNA Pol II), BRD4, and Mediator were located within +/- 5kb
of the TSS. In H2171 cells, a gene was defined as actively transcribed if enriched regions for
H3K4me3, RNA Polymerase II (RNA Pol II), and BRD4, were located within +/- 5kb of the
TSS. Mediator was omitted in H2171 transcribed gene definitions due to the lower numbers of
enriched regions for the factor. H3K4me3 is a histone modification associated with transcription
initiation (Guenther et al., 2007) and BRD4 and Mediator have been shown to also occupy
promoters of active genes (Kagey et al., 2010; Zhang et al., 2012). Using these criteria, we
identified transcriptionally active genes in MM1.S, U-87, and H2171 cells (Table Sl).
Defining active enhancers
Active enhancers form loops with promoters that are facilitated by the Mediator complex (Kagey
et al., 2010), thus we used regions of ChIP-Seq enrichment for the Mediator complex component
MEDI outside of promoters (e.g. a region not contained within +/- 2.5kb region flanking the
promoter) to define active enhancers. In order to accurately capture dense clusters of enhancers,
we allowed MEDI regions within 12.5kb of one another to be stitched together (distance criteria
established in the accompanying mouse embryonic stem cell manuscript). We noticed that in
145
some cases stitched regions in gene dense areas spanned multiple transcriptionally active genes
and were likely not representative of enhancer clusters. Thus, stitched regions spanning more
than two promoters of active genes were excluded from the analysis. This analysis defined
enhancer regions in the MML.S genome (Table S2). The same method was used to identify
enhancers in the glioblastoma multiforme cell line U-87 and in the small cell lung cancer cell
line H2171 (Table S4). Data SI contains hgl8 genome coordinate tracks for all enhancers in all
profiled cell types in the form of a UCSC genome browser .bed file that can be uploaded and
viewed at genome.ucsc.edu (Data Si).
Creatingmeta representationsof ChiP-Seq occupancy at active enhancers and promoters
Genome-wide average "meta" representations of ChIP-Seq occupancy at active enhancers and
promoters were created by mapping ChIP-Seq read density to the 5kb regions flanking the center
of active enhancer regions or transcription start sites (TSS) of active genes. Each active
enhancer or TSS region was split into one hundred 50bp bins. All active enhancer or TSS
regions were then aligned and the average background subtracted ChIP-Seq factor density in
each bin was calculated to create a meta genome-wide average in units of rpm/bp. See Figure
IB; Figure 4D; Figure S3E; Figure S4A and S4B.
CorrelatingBRD4 and MED] occupancy with RNA Pol II at promoters
All genes were ranked by increasing density of RNA Pol II ChIP-Seq reads in the promoter
region (+/- Ikb of transcription start site (TSS) and binned in increments of 100 genes. The
146
median ChIP-Seq density for promoter regions within each bin was calculated in rpm/bp for both
RNA Pol II and BRD4 and plotted (BRD4) or shaded by binding density (RNA Pol II). The same
method was applied for MED I in place of BRD4. See Figure IC.
Defining promoter boundaries in MML.S cells
Promoter regions are typically defined as the +/- several kb region centered on the TSS. In order
to produce a more accurate and gene specific definition of promoters, we used the promoter
centric H3K4Me3 histone modification mark to delineate the span of promoters for actively
transcribed genes in MMl .S. At these genes, the boundaries of the directly overlapping
H3K4Me3 ChIP-Seq enriched region were used to define the upstream/downstream ends of the
promoter.
Calculating total ChIP-Seq occupancy signal at enhancers andpromoters
The total ChIP-Seq occupancy signal at enhancers was calculated by first determining the
average ChIP-Seq read density in the entire enhancer region (rpm/bp). This value was multiplied
by the length of the region to produce total ChIP-Seq occupancy in units of total rpm. See Figure
2A; Figure S2A; Figure 7A and 7B. For all cell lines used in this study the total MEDI ChIPSeq occupancy signal at enhancers can be found in Table S2 (MM 1.S) or Table S4 (GBM and
SCLC).
147
The total ChIP-Seq occupancy signal at promoters was calculated similarly. In order to
accurately compare signal at promoters to that at enhancers, promoters that overlapped superenhancers were excluded from analysis (Figure S2A).
Correlations of total BRD4 and MED 1 occupancy signal at enhancers and promoters were
calculated using a Pearson correlation statistic (Figure S 1 A and S 1 B).
Identifying super-enhancers
We observed disproportionately high occupancy of factors such as MED 1 and BRD4 at a subset
of enhancers that we termed "super-enhancers". To formally identify super-enhancers, we first
ranked all enhancers by increasing total background subtracted ChIP-Seq occupancy of MED 1
(x-axis), and plotted the total background subtracted ChIP-Seq occupancy of MED 1 in units of
total rpm (y-axis). This representation revealed a clear inflection point in the distribution of
MED 1 at enhancers (Figure 2A; Figure 7A and 7D). We geometrically defined the inflection
point and used it to establish the cut off for super-enhancers. Enhancers that were classified as
super enhancers are noted in Table S2 (MM 1.S) and Table S4 (GBM and SCLC).
Creatingmeta representationsof ChIP-Seq occupancy at typical enhancers and super-enhancers
Genome-wide average "meta" representations of MEDI and BRD4 ChIP-Seq occupancy at
typical enhancers and super-enhancers were created by mapping MED 1 and BRD4 ChIP-Seq
read density to the enhancer regions and their corresponding +/-5kb flanking regions. Each
148
enhancer region was split into 100 equally sized bins. The flanking +/- 5kb regions were split
into 50bp bins. This split all enhancer regions, regardless of their size, into 300 bins. All typical
enhancer or super-enhancer regions were then aligned and the average MED 1 and BRD4 ChIPSeq density in each bin was calculated to create a meta genome-wide average in units of rpm/bp.
In order to visualize the length disparity between typical and super-enhancer regions, the
enhancer region (between its actual start and end) was scaled relative to its average length. See
Figure 2B (MM 1.S), Figure 7B (GBM) and 7E (SCLC).
Characterizingsuper-enhancers
We calculated the total ChIP-Seq occupancy for MED 1, BRD4, and H3K27Ac at superenhancers and typical enhancers and determined the fold difference between the average signal
at super-enhancers versus typical enhancers. ChIP-Seq density for MED 1, BRD4, and
H3K27Ac were determined by calculating the average density of each factor in constituent
regions of every enhancer. See Figure 2B (MM 1.S), Figure 7B (GBM) and 7E (SCLC).
Assigning genes to super-enhancers
Transcriptionally active genes were assigned to enhancers using the following method:
Enhancers tend to loop to and associate with adjacent genes in order to activate their
transcription (Ong and Corces, 2011). Most of these interactions occur within a distance of
-50kb of the enhancer (Chepelev et al., 2012). Using a simple proximity rule, we assigned all
transcriptionally active genes (TSSs) to super-enhancers within a 50kb window, a method shown
149
to identify a large proportion of true enhancer/promoter interactions in embryonic stem cells
(Dixon et al., 2012). This identified 681 genes associated with super-enhancers in MM 1.S cells
(Table S3). The same method was applied to identify super-enhancer-associated genes in GBM
and SCLC (Table S5).
Calculatingthe cell type specificity of enhancer-associatedgenes
Spike-in normalized gene expression levels for all genes in MMl .S cells were determined as
previously described (RNA Extraction and Synthetic RNA Spike-In and Microarray Sample
Preparation and Analysis sections). The expression levels of genes associated with typical
enhancers or super-enhancers were compared using a Welch's t-test and found be statistically
significant (p-value < 2e-16) (Figure 2D).
To evaluate the cell-type specificity of SE and typical enhancer -associated genes, we compared
MMl.S gene expression from the dataset (GSE17385) (Carrasco et al., 2007) to the largest
matched microarray expression dataset of normal human tissues (GSE7307) on Affymetrix U133
plus 2.0 arrays from Gene Expression Omnibus (GEO). We used the three positive controls from
GSE17385 and the 504 samples of normal human tissues from GSE7307 for the analysis. All
.CEL files were MAS5 normalized together and probe level intensities were reported using
standard Affymetrix CDF. We used the entropy-based measure Jensen-Shannon (JS) divergence
to quantify cell-type specificity (Cabili et al., 2011). For each triplicate of MML.S expression
dataset from GSE7307, the JS divergence quantified the similarity between a probe's expression
pattern across the MML.S dataset and the 504 samples of normal human tissues and a predefined pattern that represents an extreme case in which the probe is expressed only in MM1.S.
150
The JS divergence statistic was converted to a Z-score to give a relative measure of the cell type
specificity of genes in MML.S compared to the average gene. A similar comparison of cell type
specificity Z-scores also showed a statistically significant difference between genes associated
with typical or super-enhancers (p-value = 1 e- 14) (Figure 2D).
Quantifying effects ofJQ] treatment on BRD4 occupancy genome wide
In order to quantify the effects of JQ 1 treatment on the genomic occupancy of BRD4, we first
calculated average BRD4 occupancy (in units of rpm/bp) at BRD4 enriched regions in DMSO
treated cells. We next calculated the average BRD4 occupancy in those same regions in cells
treated with 500nM JQ 1 for 6 hours. The distributions of BRD4 at enriched regions are plotted
(Figure 4B). The difference in the distributions of BRD4 was compared using a Welch's t-test
and found to be statistically significant (p-value < 1 e- 16) (Figure 4B)
Quantifying effects of JQJ treatment on BRD4 occupancy at active enhancers
In order to quantify the effects of JQ 1 treatment on the genomic occupancy of BRD4 at active
enhancers, we first calculated the BRD4 occupancy at typical enhancers or super-enhancers in
DMSO and various concentrations of JQ 1. The loss of BRD4 at typical enhancers or superenhancers was quantified as the fraction of BRD4 occupancy remaining compared to DMSO.
The average loss of BRD4 at typical enhancers and super-enhancers is plotted for DMSO and
5nM, 50nM, 500nM JQ 1treatment in Figure 5F. 95% confidence intervals were determined by
performing 10,000 random samplings (with replacement) of either super-enhancers or typical
enhancers. Sample sizes were identical between typical enhancers and super-enhancers. The
151
differences in the mean loss of BRD4 between typical enhancers and super-enhancers were also
deemed significant (p-value < 1 e-8) at 5nM, 50nM, and 500nM JQ 1 treatment (Figure 5F) using
a Welch's /-test.
Quantifying effects ofJQJ treatment on gene expression in MML.S
Spike-in normalized gene expression levels for all genes were determined as previously
described (RNA Extraction and Synthetic RNA Spike-In and Microarray Sample Preparation and
Analysis sections). In order to quantify changes in gene expression as a function of JQ 1
treatment in a time or dose dependent manner, expression levels were first calculated for either
super-enhancer or typical enhancer-associated genes (Figure 6B and 6C). For time dependent
changes, the Log 2 fold change in expression was calculated relative to untreated control cells.
For dose dependent changes, the Log 2 fold change in expression was calculated relative to
DMSO treated cells. The average change in gene expression at super-enhancer or typical
enhancer-associated genes was calculated at each timepoint or dose. In order to determine the
significance of changes in the mean between super-enhancer and typical enhancer-associated
genes, 95% confidence intervals were determined by performing 10,000 random samplings (with
replacement) of the mean change for each set of genes. Sample sizes were identical between
typical enhancer and super-enhancer-associated genes. We also determined the significance of
the difference between distributions of gene expression values for super-enhancer and typical
enhancer -associated genes at each time point or dose. Changes were significant over time at 3
and 6 hours after 500nM JQI treatment (p-value < le-8, Figure 6B). Changes were significant
after 6 hour JQI treatments at doses of 50nM, 500nM, and 5,OOOnM (p-value < le-8, Figure 6C).
152
Identical analysis was performed comparing changes in gene expression between superenhancer-associated genes with multiple typical enhancers (Figure S3C and S3D). Changes
were significant over time at 6 hours after 500nM JQI treatment (p-value = 0.03, Figure S3C).
Changes were significant after 6 hour JQl treatments at a dose of 500nM (p-value = 0.052,
Figure S3D).
Quantifying effects ofJQJ treatment on CDK9 and MED] occupancy at promoters and
enhancers
We quantified changes in CDK9 and MED 1 occupancy by calculating changes in CDK9 and
MED 1 ChIP-Seq occupancy at promoters, typical enhancers, and super-enhancers. Since superenhancers often contain TSS overlapping regions, we used H3K4Me3 promoter definitions
(Figure S2B) to exclude TSS overlapping super-enhancers. The change in CDK9 and MED1
occupancy was quantified as the fraction of factor occupancy remaining after 500nM JQl
treatment for 6 hours compared to DMSO. The average change in CDK9 and MEDI at
promoters, typical enhancers, and super-enhancers is plotted in Figure 6F. To show the
significance of changes in the mean, 95% confidence intervals were determined by performing
10,000 random samplings (with replacement) of the mean change at either super-enhancers or
typical enhancers.
In order to determine the changes in CDK9 and MEDI occupancy at TSS overlapping and TSS
distal portions of super-enhancers, we split super-enhancers into TSS proximal and TSS distal
regions (Figure S3F). The average change in CDK9 and MED 1 at super-enhancer TSS proximal
153
and TSS distal regions is plotted in Figure S3F. To show the significance of changes in the mean,
95% confidence intervals were determined by performing 10,000 random samplings (with
replacement) of the mean change at either super-enhancers or typical enhancers.
Quantifyjing effects ofJQJ treatment on transcriptionelongation
We quantified the effects of JQ 1 treatment on transcription elongation of genes by measuring the
change in RNA Pol II density in the elongating gene body region of genes. The elongating
region of genes was defined as the region beginning 300bp downstream of the TSS and
extending 3000bp past the annotated 3' end of the transcript. RNA Pol II density (rpm/bp) was
calculated in this region for all transcribed genes in cells treated with DMSO or 500nM JQ 1.
The Log2 fold change in RNA Pol II density +/- 500nM JQ 1 treatment was calculated for all
actively transcribed genes (Figure 6G).
To quantify the effects of JQ 1 treatment on RNA Pol II elongation at typical or super-enhancerassociated genes, the Log 2 fold change in RNA Pol II density +/- 500nM JQ 1 treatment was
calculated for typical enhancer or super-enhancer-associated genes. The distributions of changes
in RNA Pol II occupancy were statistically different as determined by a two sample
Kolmogorov-Smirnov (KS) test (Figure S3G, p-value < 2e-16). The average change in RNA Pol
II between the two groups of genes was also examined (Figure 6H) and the significance of
change was shown using plotted 95% confidence intervals determined by performing 10,000
random samplings (with replacement) of the mean change for either super-enhancer or typical
enhancer-associated genes.
154
References
Cabili, M.N., Trapnell, C., Goff, L., Koziol, M., Tazon-Vega, B., Regev, A., and Rinn, J.L.
(2011). Integrative annotation of human large intergenic noncoding RNAs reveals global
properties and specific subclasses. Genes Dev 25, 1915-1927.
Carrasco, D.R., Sukhdeo, K., Protopopova, M., Sinha, R., Enos, M., Carrasco, D.E., Zheng, M.,
Mani, M., Henderson, J., Pinkus, G.S., et al. (2007). The differentiation and stress response
factor XBP-1 drives multiple myeloma pathogenesis. Cancer Cell 11, 349-360.
Chepelev, I., Wei, G., Wangsa, D., Tang, Q., and Zhao, K. (2012). Characterization of genomewide enhancer-promoter interactions reveals co-expression of interacting genes and modes of
higher order chromatin organization. Cell Res 22, 490-503.
Dawson, M.A., Prinjha, R.K., Dittmann, A., Giotopoulos, G., Bantscheff, M., Chan, W.I.,
Robson, S.C., Chung, C.W., Hopf, C., Savitski, M.M., et al. (2011). Inhibition of BET
recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. Nature 478, 529533.
Dixon, J.R., Selvaraj, S., Yue, F., Kim, A., Li, Y., Shen, Y., Hu, M., Liu, J.S., and Ren, B.
(2012). Topological domains in mammalian genomes identified by analysis of chromatin
interactions. Nature 485, 376-380.
Futreal, P.A., Coin, L., Marshall, M., Down, T., Hubbard, T., Wooster, R., Rahman, N., and
Stratton, M.R. (2004). A census of human cancer genes. Nat Rev Cancer 4, 177-183.
Guenther, M.G., Levine, S.S., Boyer, L.A., Jaenisch, R., and Young, R.A. (2007). A chromatin
landmark and transcription initiation at most promoters in human cells. Cell 130, 77-88.
Kagey, M.H., Newman, J.J., Bilodeau, S., Zhan, Y., Orlando, D.A., van Berkum, N.L., Ebmeier,
C.C., Goossens, J., Rahl, P.B., Levine, S.S., et al. (2010). Mediator and cohesin connect gene
expression and chromatin architecture. Nature 467, 430-435.
Langmead, B., Trapnell, C., Pop, M., and Salzberg, S.L. (2009). Ultrafast and memory-efficient
alignment of short DNA sequences to the human genome. Genome Biol 10, R25.
Lin, C.Y., Loven, J., Rahl, P.B., Paranal, R.M., Burge, C.B., Bradner, J.E., Lee, T.I., and Young,
R.A. (2012). Transcriptional amplification in tumor cells with elevated c-Myc. Cell 151, 56-67.
Loven, J., Orlando, D.A., Sigova, A.A., Lin, C.Y., Rahl, P.B., Burge, C.B., Levens, D.L., Lee,
T.I., and Young, R.A. (2012). Revisiting global gene expression analysis. Cell 151, 476-482.
Ong, C.T., and Corces, V.G. (2011). Enhancer function: new insights into the regulation of
tissue-specific gene expression. Nat Rev Genet 12, 283-293.
Patel, M.N., Halling-Brown, M.D., Tym, J.E., Workman, P., and Al-Lazikani, B. (2013).
Objective assessment of cancer genes for drug discovery. Nat Rev Drug Discov 12, 35-50.
155
Pruitt, K.D., Tatusova, T., and Maglott, D.R. (2007). NCBI reference sequences (RefSeq): a
curated non-redundant sequence database of genomes, transcripts and proteins. Nucleic Acids
Res 35, D61-65.
Ramskold, D., Wang, E.T., Burge, C.B., and Sandberg, R. (2009). An abundance of ubiquitously
expressed genes revealed by tissue transcriptome sequence data. PLoS Comput Biol 5,
e1000598.
Zhang, W.S., Prakash, C., Sum, C., Gong, Y., Li, Y., Kwok, J.J., Thiessen, N., Pettersson, S.,
Jones, S.J., Knapp, S., et al. (2012). Bromodomain-Containing-Protein 4 (BRD4) Regulates
RNA Polymerase II Serine 2 Phosphorylation in Human CD4+ T Cells. J Biol Chem.
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.
156
Appendix B
Enhancer Decommissioning By LSD1 During
Embryonic Stem Cell Differentiation
Warren A. Whyte, 2 *, Steve Bilodeaul*, David A. Orlando', Heather A. Hoke, 2 , Garrett M.
Frampton, 2 , Charles T. Foster' 4 , Shaun M. Cowley 4 , and Richard A. Young' 2
Whitehead Institutefor Biomedical Research, 9 Cambridge Center, Cambridge, Massachusetts
02142, USA. 2Departmentof Biology, MassachusettsInstitute of Technology, Cambridge,
Massachusetts 02139, USA.
3Department of MolecularBiology, Adolf-Butenandt Institut,
Ludwig-Maximilians-UniversitdiMfinchen, Munich, Germany 80336. 4 Departmentof
Biochemistry, University of Leicester, Leicester LEI 9HN, United Kingdom.
*These authors contributed equally to this work
This chapter originally appeared in Nature 2012 vol. 482 pp. 221-225. Corresponding
Supplementary Material is included. Supplementary tables and data are available online at
http://dx.doi.org/ 10.1038/naturel0805.
157
Personal Contribution to the Project
I performed ChIP-seq of HDAC 1 and HDAC2 in mouse ESCs, and assisted with knockdown
experiments. I also assisted in preparation of the manuscript.
158
Summary
Transcription factors and chromatin modifiers play important roles in
programming and reprogramming of cellular states during development (Graf and Enver,
2009; Young, 2011). Transcription factors bind to enhancer elements and recruit
coactivators and chromatin modifying enzymes to facilitate transcription initiation (Fuda
et al., 2009; Li et al., 2007). During differentiation, a subset of these enhancers must be
silenced, but the mechanisms underlying enhancer silencing are poorly understood. Here
we show that the H3K4/K9 histone demethylase LSD1 (Shi et al., 2004) plays an essential
role in decommissioning enhancers during differentiation of embryonic stem cells (ESCs).
LSD1 occupies enhancers of active genes critical for control of ESC state. However, LSD1
is not essential for maintenance of ESC identity. Instead, ESCs lacking LSD1 activity fail
to fully differentiate and ESC-specific enhancers fail to undergo the histone demethylation
events associated with differentiation. At active enhancers, LSD1 is a component of the
NuRD complex, which contains additional subunits that are necessary for ESC
differentiation. We propose that the LSD1-NuRD complex decommissions enhancers of the
pluripotency program upon differentiation, which is essential for complete shutdown of the
ESC gene expression program and the transition to new cell states.
159
The histone H3K4/K9 demethylase LSD1 (Lysine-specific-demethylase- 1, KDM 1 A) is
among the chromatin regulators that have been implicated in control of early embryogenesis
(Wang et al., 2007; Wang et al., 2009a; Foster at al., 2010). Loss of LSD1 leads to embryonic
lethality and ESCs lacking LSD 1 function fail to differentiate into embryoid bodies (Wang et al.,
2007; Wang et al., 2009a; Foster at al., 2010). These results suggest that LSD1 contributes to
changes in chromatin that are critical to differentiation of ESCs, but LSD I's role in this process
is not yet understood. To investigate the function of LSD1 in ESCs, we first identified the sites it
occupies in the genome using chromatin immunoprecipitation coupled with massively parallel
DNA sequencing (ChIP-Seq; Fig. 1, and Supplementary Fig. 1). The results revealed that LSD 1
occupies the enhancers and core promoters of a substantial population of actively transcribed and
bivalent genes (Fig. la, b, and Supplementary Table 1). Inspection of individual gene tracks
showed that LSD 1 co-occupies well-characterized enhancer regions with the ESC master
transcription factors Oct4, Sox2 and Nanog and the Mediator coactivator (Fig. Ib, and
Supplementary Fig. 1). Loci bound by Oct4, Sox2 and Nanog are generally associated with
Mediator and p300 coactivators and have enhancer activity (Kagey et al., 2010; Chen et al.,
2008). A global view of Oct4, Sox2, Nanog and Mediator -occupied enhancer regions confirmed
that 97% of the 3,838 high confidence enhancers were co-occupied by LSD1 (p < 10-9) (Fig. Ic,
and Supplementary Table 2). This is consistent with evidence that LSD 1 can interact with Oct4
(Pardo et al., 2010; van den Berg et al., 2010). LSD1 signals were also observed at core promoter
regions with RNA Polymerase II (Pol II) and TBP (Fig. Id). The density of LSD1 signals at
enhancers was higher than at core promoters (p < 10-16; Supplementary Fig. 1), indicating that
LSD 1 is associated predominantly with the enhancers of actively transcribed genes in ESCs.
It was striking to find that LSDI is associated with active genes in ESCs because
160
previous studies have shown that LSD 1 is not essential for maintenance of ESC state, but is
required for normal differentiation (Wang et al., 2007; Wang et al., 2009a; Foster et al., 2010).
We used an ESC differentiation assay to further investigate the involvement of LSD 1 in cell state
transitions (Fig. 2a, b). Prolonged depletion of Oct4 in ZHBTc4 ESCs using doxycycline causes
loss of pluripotency and differentiation into trophectoderm (Niwa et al., 2000). As expected, loss
of Oct4 expression led to a rapid loss of ESC morphology and a marked reduction in SSEA- 1
and alkaline phosphatase, two markers of ESCs (Fig. 2c, and Supplementary Fig. 2). When these
ESCs were treated with the LSDI inhibitor tranylcypromine (TCP) during Oct4 depletion, they
failed to undergo the morphological changes associated with differentiation of ESCs (Fig. 2c).
Instead, the TCP-treated cells formed small colonies resembling those of untreated ESCs and
maintained expression of SSEA-I and alkaline phosphatase (Fig. 2c, and Supplementary Fig. 3).
Very similar results were obtained in LSD 1 knockout ESCs (Supplementary Fig. 4, 5), and in
cells treated with another LSD 1 inhibitor, pargyline (Prg), or an shRNA against LSD 1
(Supplementary Fig. 2, 3). LSD1 inhibition also caused an increase in cell death during
differentiation, as has been observed with cells lacking LSD 1 in other assays (Wang et al.,
2009a; Foster et al., 2010). These results suggest that LSDI may be required for ESCs to
completely silence the ESC gene expression program.
Further analysis of ESCs forced to differentiate in the absence of LSD 1 activity
confirmed that these cells failed to fully transition from the ESC gene expression program; while
key genes of the trophectoderm gene expression program were activated, including Cdx2 and
EsxJ (Rossant and Cross, 2001), there was incomplete repression of many ESC genes, including
Sox2 and Fbox15 (Fig. 2d). A global analysis confirmed that a set of genes neighboring LSD I occupied enhancers in ESCs are repressed upon differentiation, and that repression of this set of
161
genes is partially relieved in the presence of TCP (Fig. 2e, and Supplementary Table 3). Similar
results were obtained with LSD1 knockout cells (Supplementary Fig. 4, 5), and cells treated with
either pargyline or an shRNA against LSD 1 (Supplementary Fig. 3). These results indicate that
the trophectoderm differentiation program can be induced in cells lacking LSD1 function, but the
ESC program is not fully silenced in these cells.
To gain further insight into the role of LSD 1 in ESC differentiation, we investigated
whether LSDI is associated with previously described complexes, including NuRD (Nucleosome
remodeling and histone deacetylase), CoREST (Cofactor of REST) and the AR/ER (Androgen
receptor/ Estrogen receptor) complexes (Foster et al., 2010; Metzger et al., 2005; Shi et al., 2005;
Wang et al., 2009b). We first studied whether the LSDI found at Oct4-occupied genes is a
component of NuRD because Oct4 and Nanog have been reported to interact with several
components of NuRD (Pardo et al., 2010; van den Berg et al., 2010; Liang et al., 2008). ChIPSeq experiments confirmed that NuRD subunits Mi-2p, HDAC 1 and HDAC2 co-occupy sites
with LSD 1 at enhancers (p < 10-9; Fig. 3 and Supplementary Table 1). Immunoprecipitation of
LSD 1 confirmed its association with Mi-2p, HDAC 1 and HDAC2 (Fig. 3b, c). We then
investigated whether LSD 1 is associated with CoREST; ChIP-Seq data revealed that a minor
fraction of LSD 1 co-occupies sites with CoREST and Rest (2% and 6%,
respectively)(Supplementary Fig. 6 and Supplementary Table 1). As expected, LSD 1-REST
sites were frequently found associated with neuronal genes (Supplementary Fig. 7 and
Supplementary Table 4). Immunoprecipitation experiments confirmed that LSD I is associated
with CoREST (Fig. 3b, c). AR and ER are not expressed in ESCs based on the lack of histone
H3K79me2 and H3K36me3 (modifications associated with transcriptional elongation) at the
genes encoding these proteins (Supplementary Table 1). Further examination of the ChIP-Seq
162
data revealed that enhancers were significantly more likely to be occupied by the LSD 1 and
NuRD proteins as compared to REST and CoREST (p < 10-9) (Fig. 3d and Supplementary Fig.
8). Multiple components of NuRD are dispensable for ESC state but required for normal
differentiation (Wang et al., 2007; Dovey et al., 2010; Kaji et al., 2006; Scimone et al., 2010).
ESCs with reduced levels of the core NuRD ATPase Mi-23 failed to differentiate properly and
partially maintained expression of SSEA-1, alkaline phosphatase and ESC genes (Supplementary
Fig. 9), which are the same phenotypes we observed with reduced levels of LSD 1. These results
indicate that LSD 1 at enhancers is associated with a NuRD complex that is essential for normal
cell state transitions.
Nucleosomes with histone H3K4mel are commonly found at enhancers of active genes,
and are a substrate for LSDl (Heintzman et al., 2007; Shi et al., 2004). If LSD1-dependent
H3K4mel demethylase activity is involved in enhancer silencing during ESC differentiation,
LSD 1 inhibition should cause retention of H3K4me 1 levels at active ESC enhancers when
differentiation is induced. During trophectoderm differentiation with control ESCs, we found
p300 and H3K27ac levels reduced at a set of active ESC enhancers, suggesting these enhancers
were being silenced (Supplementary Fig. 10). The levels of H3K4me 1 at enhancers were also
reduced, as seen for example at Lefty] (Fig. 4a, and Supplementary Table 5), while the levels of
H3K4mel increased at newly active trophectoderm genes such as Gata2 (Fig. 4b). In contrast,
H3K4me 1 signals were higher at LSD 1- occupied enhancers in differentiating ESCs treated with
TCP than in control cells, including Lefty] and Sox2 (Fig. 4a, c). The majority of enhancers
(1,722 of 2,755) that were occupied by LSDI and that experienced reduced levels of H3K4mel
during differentiation retained H3K4me 1 in TCP-treated ESCs compared to untreated control
differentiating ESCs (Fig. 4d, e). These results are consistent with the model that LSD 1
163
demethylates H3K4me 1 at the enhancers of ESC-specific genes during differentiation, and that
this activity is essential to fully repress the genes associated with these enhancers.
Our results indicate that an LSD 1 -NuRD complex is required for silencing of ESC
enhancers during differentiation, which is essential for complete shutdown of the ESC gene
expression program and the transition to new cell states. These results, together with those of
previous studies on NuRD function (Liang et al., 2008; Scimone et al., 2010; Forneris et al.,
2005; Lee et al., 2006), suggest the following model for LSDI-NuRD in enhancer
decommissioning. LSD1-NuRD complexes occupy Oct4-regulated active enhancers in ESCs,
but do not substantially demethylate histone H3K4 because LSD1 's H3K4 demethylase activity
is inhibited in the presence of acetylated histones (Forneris et al., 2005; Lee et al., 2006).
Enhancers occupied by Oct4, Sox2 and Nanog are co-occupied by the HAT p300 and
nucleosomes with acetylated histones (Supplementary Fig. 10 and Chen et al., 2008). Thus, as
long as the enhancer-bound transcription factors recruit HATs to enhancers, the net effect of
having both HATs and NuRD-associated HDACs present is to have sufficient levels of
acetylated histones to suppress LSD 1 demethylase activity. During ESC differentiation, the
levels of Oct4 and p300 are reduced, thus reducing the level of acetylated histones, which in turn
permits demethylation of H3K4 by LSD 1. Consistent with this model, we find that the shutdown
of Oct4 leads to reduced levels of p300 and histone H3K27ac at enhancers that are occupied by
Oct4 and LSD 1 (Supplementary Fig. 10, 11), and this is coincident with reduced levels of
methylated H3K4 (Fig. 4, and Supplementary Figs. 12, 13). This model would explain why key
components of LSD 1 -NuRD complexes are not essential for maintenance of ESC state, but are
essential for normal differentiation, when the active enhancers must be silenced. Additional
HATs expressed in ESCs may also contribute to the dynamic balance of nucleosome acetylation.
164
Future biochemical analysis of HAT, HDAC and demethylase complexes at enhancers will be
valuable for testing this model and for further understanding how enhancers are regulated during
differentiation.
We conclude that LSD I -NuRD complexes present at active promoters in ESCs are
essential for normal differentiation, when the active enhancers must be silenced. Given evidence
that LSD1 is required for differentiation of multiple cell types (Wang et al., 2007; Musri et al.,
2010; Choi et al., 2010), LSDl is likely to be generally involved in enhancer silencing during
differentiation. The ESC gene expression program can be maintained in the absence of many
other chromatin regulators (Young, 2011), and it is possible that some of these also play key
roles in the transition from one transcriptional program to another during differentiation.
165
Methods Summary
ESC Cell Culture Conditions
ESCs were grown on irradiated murine embryonic fibroblasts (MEFs) and passaged as
previously described (Kagey et al., 2010). In drug treatment experiments, ESCs were split off
MEFs and treated with tranylcypromine (TCP, ImM) or pargyline (Prg, 3mM) to inhibit LSD1
activity. Lentiviral constructs were purchased from Open Biosystems and produced according to
the Trans-lentiviral shRNA Packaging System (TLP4614).
Differentiation assay, Immunofluorescence, and Alkaline PhosphataseStaining
ZHBTc4 ESCs were split off MEFs in ESC media containing 2pg/ml doxycycline to reduce Oct4
expression levels. For Immunofluorescence, ESCs were crosslinked, blocked and permeabilized
before incubation with Oct4 (Santa Cruz, sc-9081x; 1:200 dilution) or SSEA 1 (mc-480, DHSB,
1:20 dilution) antibodies. Alexa-conjugated secondary antibodies were used for detection.
Staining of ESCs for alkaline phosphatase was achieved using the Alkaline Phosphatase
Detection Kit (Millipore, SCRO04). Cells were harvested at indicated time points for ChIP-Seq,
qPCR or expression array analyses.
ChIP-Seq
Chromatin immunoprecipitations (ChIPs) were performed and analyzed as previously described
(Kagey et al., 2010). The following antibodies were used: LSDl (Abcam, ab1772 1), Mi-2b
166
(Abcam, ab72418), HDAC1 (Abcam, ab7028), HDAC2 (Abcam, ab7029), REST (Millipore, 07579), CoREST (Abcam, ab3263 1), H3K4mel (Abcam, ab8895), p300 (Santa-Cruz, sc-584) and
H3K27Ac (Abcam, ab4729).
For ChIP-Seq analyses, reads were aligned with Bowtie and analyzed as described in
Supplemental Information.
Full Methods and any associated references are available in Supplementary Information of the
paper at www.nature.com/nature.
Accession Numbers
ChIP-Seq and GeneChip expression data have been deposited in Gene Expression Omnibus with
accession number GSE27844.
Acknowledgements
We thank Jakob Loven, Michael H. Kagey, Jill Dowen, Alan C. Mullen, Alla Sigova, Peter B.
Rahl, Tony Lee, and members of Yang Shi's lab for experimental assistance, reagents and
helpful discussions; and Jeong-Ah Kwon, Vidya Dhanapal, Jennifer Love, Sumeet Gupta,
Thomas Volkert, Wendy Salmon and Nicki Watson for assistance with ChIP-Seq, expression
arrays, and immunofluorescence imaging acquisition. This work was supported by a Canadian
Institutes of Health Research Fellowship (SB), a Career Development Award from the Medical
167
Research Council (SMC), and by NIH grants HG002668 and NS055923 (RY).
Author Contributions
The ChIP-Seq, immunofluorescence and expression experiments were designed, conducted and
interpreted by W.A.W., S.B., H.A.H., C.T.F., S.M.C and R.A.Y. W.A.W., D.A.O. and G.M.F.
performed data analysis. The manuscript was written by S.B., W.A.W. D.A.O., H.A.H., G.M.F.
and R.A.Y.
168
References
Chen, X., Yuan, P., Fang, F., Huss, M., Vega V.B., Wong, E., Orlov Y.L., Zhang, W., Jiang, J.,
Loh, Y.H., et al. (2008). Integration of external signaling pathways with the core transcriptional
network in embryonic stem cells. Cell 133, 1106-1117.
Choi, J., Jang, H., Kim, H., Kim, S.T., Cho, E.J., and Youn, H.D. (2010). Histone demethylase
LSD 1 is required to induce skeletal muscle differentiation by regulating myogenic factors.
Biochem Biophys Res Commun 401, 327-332.
Dovey, O.M., Foster, C.T., and Cowley, S.M. (2010). Histone deacetylase 1 (HDACI), but not
HDAC2, controls embryonic stem cell differentiation. Proc Natl Acad Sci U S A 107, 82428247.
Forneris, F., Binda, C., Vanoni, M.A., Battaglioli, E., and Mattevi, A. (2005). Human histone
demethylase LSD1 reads the histone code. J Biol Chem 280, 41360-41365.
Foster, C.T., Dovey, O.M., Lezina, L. Luo, H.L. Gant, T.W., Barlev, N., Bradley, A., and
Cowley, S.M. (2010). Lysine-specific demethylase 1 regulates the embryonic transcriptome and
CoREST stability. Mol Cell Biol 30, 4851-4863.
Fuda, N.J., Ardehali, M.B., and Lis, J.T. (2009). Defining mechanisms that regulate RNA
polymerase II transcription in vivo. Nature 461, 186-192.
Graf, T. and Enver, T. (2009). Forcing cells to change lineages. Nature 462, 587-594.
Heintzman, N.D., Stuart, R.K., Hon, G., Fu, Y., Ching, C.W., Hawkins, R.D., Barrera, L.O., Calcar, S.V.,
Qu, C., Ching, K.A., el al. (2007). Distinct and predictive chromatin signatures of transcriptional
promoters and enhancers in the human genome. Nat Genet 39, 311-318.
Kagey, M.H., Newman, J.J., Bilodeau, S., Zhan, Y., Orlando, D.A., van Berkum, N.L., Ebmeier, C.C.,
Goossens, J., Rahl, P.B., Levine, S.S., el al. (2010). Mediator and cohesin connect gene expression and
chromatin architecture. Nature 467, 430-435.
Kaji, K., Cabellero, I.M., MacLeod, R., Nichols, J., Wilson, V.A., and Hendrich, B. (2006). The NuRD
component Mbd3 is required for pluripotency of embryonic stem cells. Nat Cell Biol 8, 285-292.
Lee, M.G., Wynder, C., Bochar, D.A., Hakimi, M.A., Cooch, N., and Shiekhattar, R. (2006). Functional
interplay between histone demethylase and deacetylase enzymes. Mol Cell Biol 26, 6395-6402.
Li, B., Carey, M., and Workman, J.L. (2007). The role of chromatin during transcription. Cell 128, 707719.
Liang, J., Wan, M., Zhang, Y., Gu, P., Xin, H., Jung, S.Y., Qin, J., Wong, J., Cooney, A.J., Liu, D., et al.
(2008). Nanog and Oct4 associate with unique transcriptional repression complexes in embryonic stem
cells. Nat Cell Biol 10, 731-739.
169
Metzger, E., Wissmann, M., Yin, N., Muller, J.M., Schneider, R., Peters, A.H., Gunther, T. Buettner, R.,
and Schule, R. (2005). LSD 1 demethylates repressive histone marks to promote androgen-receptordependent transcription. Nature 437, 436-439.
Musri, M.M., Carmona, M.C., Hanzu, F.A., Kaliman, P., Gomis, R., and Parrizas, M. (2010). Histone
demethylase LSD I regulates adipogenesis. J Biol Chem 285, 30034-30041.
Nakatake, Y., Fukui, N., Iwamatsu, Y., Masui, S., Takahashiki, K., Yagi, R., Yagi, K., Miyazaki, J.,
Matoba, R., Ko, M.S., el al. (2006). Klf4 cooperates with Oct3/4 and Sox2 to activate the Leftyl core
promoter in embryonic stem cells. Mol Cell Biol 26, 7772-7782.
Niwa, H., Miyazaki, J., and Smith, A.G. (2000). Quantitative expression of Oct-3/4 defines
differentiation, dedifferentiation or self-renewal of ES cells. Nat Genet 24, 372-376.
Okumura-Nakanishi, S., Saito, M., Niwa, H., and Ishikawa, F. (2005). Oct-3/4 and Sox2 regulate Oct-3/4
gene in embryonic stem cells. J Biol Chem 280, 5307-5317.
Pardo, M., Lang, B., Prosser, H., Bradley, A., Babu, M.M., and Choudhary, J. (2010). An expanded Oct4
interaction network: implications for stem cell biology, development, and disease. Cell Stem Cell 6, 382-
395.
Ray, S., Dutta D., Rumi, M.A., Kent, L.N., Soares, M.J., and Paul, S. (2009). Context-dependent function
of regulatory elements and a switch in chromatin occupancy between GATA3 and GATA2 regulate Gata2
transcription during trophoblast differentiation. J Biol Chem 284, 4978-4988.
Rossant, J. and Cross, J.C. (200 1). Placental development: lessons from mouse mutants. Nat Rev Genet 2,
538-548.
Scimone, M.L., Meisel, J., and Reddien, P.W. (2010). The Mi-2-like Smed-CHD4 gene is required for
stem cell differentiation in the planarian Schmidtea mediterranea. Development 137, 1231-1241.
Shi, Y., Lan, F., Matson, C., Mulligan, P., Whetstine, J.R., Cole, P.A., Casero, R.A., and Shi, Y. (2004).
Histone demethylation mediated by the nuclear amine oxidase homolog LSD 1. Cell 119, 941-953.
Shi, Y.J., Matson, C., Lan, F., Iwase, S., and Shi, Y. (2005). Regulation of LSDI histone demethylase
activity by its associated factors. Mol Cell 19, 857-864.
van den Berg, D.L., Snoek, T., Mullin, N.P., Yates, A., Bezstarosti, K., Demmers, J., Chambers, I., and
Poot, R.A. (2010). An Oct4-centered protein interaction network in embryonic stem cells. Cell Stem Cell
6,369-381.
Wang, J., Scully, K., Zhu, X., Cai, L., Zhang, J., Prefontaine, G.G., Krones, A., Ohgi, K.A., Zhu, P.,
Garcia-Bassets, I., ef al. (2007). Opposing LSDI complexes function in developmental gene activation
and repression programmes. Nature 446, 882-887.
Wang, J., Hevi, S., Kurash, J.K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., ei al.
(2009a). The lysine demethylase LSD 1 (KDM 1) is required for maintenance of global DNA methylation.
Nat Genet 41, 125-129.
170
Wang, Y., Zhang, H., Chen, Y., Sun, Y., Yang, F., Yu, W., Liang, J., Sun, L., Yang, X., Shi, L., el al.
(2009b). LSDl is a subunit of the NuRD complex and targets the metastasis programs in breast cancer.
Cell 138, 660-672.
Yeom, Y.I., Fuhrmann, G., Ovitt, C.E., Brehm, A., Ohbo, K., Gross, M., Hubner, K., and Scholer, H.R.
(1996). Germline regulatory element of Oct-4 specific for the totipotent cycle of embryonal cells.
Development 122, 881-894.
Young, R.A. (2011). Control of the embryonic stem cell state. Cell 144, 940-954.
171
FIGURE LEGENDS
Figure 1: LSD1 is associated with enhancer and core promoter regions of active genes in
ESCs
a, LSD 1 occupies a substantial population of actively transcribed genes in murine ESCs. Pie
charts depict active (green), bivalent (yellow) and silent (red) genes, and the proportion (black
lines) occupied by either LSD 1, Pol II, or the Polycomb protein Suz 12 (Supplementary Table 1,
and Supplementary Information). The numbers depict the number of genes bound over the total
number of genes in each of the active, bivalent, and silent classes. LSD 1 ChIP-Seq data is from
combined biological replicates using an antibody specific for LSD 1 as determined by
knockdown experiments (Supplementary Fig. 1). The p-value for each category was determined
by a hypergeometric test.
b, LSD 1 occupies enhancers and core promoter regions of actively transcribed genes. ChIP-Seq
binding profiles (reads/million) for ESC transcription factors (Oct4, Sox2, Nanog), coactivator
(Med 1), chromatin regulator (LSD 1), the transcriptional apparatus (Pol II, TBP) and histone
modifications (H3K4me 1, H3K4me3, H3K79me2, H3K36me3) at the Oct4 (Pou5f]) and Lefty]
loci in ESCs, with the y-axis floor set to 1. Gene models, and previously described enhancer
regions (Okumura-Nakanishi et al., 2005; Yeom et al., 1996; Nakatake et al., 2006) are depicted
below the binding profiles.
c, LSD 1 occupies enhancer sites. Density map of ChIP-Seq data at Oct4, Sox2, Nanog, and
Medl co-occupied enhancer regions. Data is shown for ESC regulators (Oct4), coactivators
(Med 1 and p300) and a chromatin regulator (LSD1) in ESCs. Enhancers were defined as Oct4,
172
Sox2, Nanog and Mediator co-occupied regions. Over 96% of the 3,838 high confidence
enhancers were co-occupied by LSDl (p < 10-9). Color scale indicates ChIP-seq signal in reads
per million.
d, LSD 1 occupies core promoter sites. Density map of ChIP-Seq data at transcriptional start sites
(TSS) of genes neighboring the 3,838 previously defined enhancers (Fig. ic). Data is shown for
components of the transcription apparatus (Pol II and TBP) and the chromatin regulator LSDl in
ESCs. Core promoters were defined as the closest TSS from each enhancer. Color scale indicates
ChIP-Seq signal in reads per million.
Figure 2: LSD1 inhibition results in incomplete silencing of ESC genes during
differentiation
a, Schematic representation of trophectoderm differentiation assay using doxycycline-inducible
Oct4 shutdown murine ESC line ZHBTc4. Treatment with doxycycline for 48 hours leads to
depletion of Oct4 and early trophectoderm specification. Cells were treated with DMSO
(control) or the LSD1 inhibitor tranylcypromine (TCP) for 6 hours before 2pgg/ml doxycycline
was added for an additional 24 or 48 hours.
b, Treatment of ZHBTc4 ESCs with doxycycline leads to loss of Oct4 proteins. Oct4 and LSD 1
protein levels in nuclear extracts (NE) determined by Western blot (WB) before and after
treatment of ZHBTc4 ESCs with 2pg/ml doxycycline. Tubulin served as loading control.
c, Doxcycline-treated cells treated with TCP maintained SSEA- I cell surface marker expression.
Cells were stained for Hoechst (Hoe), Oct4 and SSEA- 1. Scale bar = 1 OOiM
173
d, Expression of selected ESC and trophectodermal genes 48 hours after Oct4 depletion in
DMSO- versus TCP-treated cells (black versus grey bars, respectively). Treatment of TCP
partially relieved repression of ESC genes, but did not affect upregulation of trophectodermal
genes. Error bars reflect standard deviation from biological replicates.
e, Genes neighboring LSD 1-occupied enhancers are less downregulated during ESC
differentiation following TCP treatment. Mean fold change in expression of the 630
downregulated (at least 1.25 fold; p < 0.01) genes nearest LSD 1- occupied enhancers (Fig. I c)
during differentiation of TCP-treated and untreated control cells. Alleviation of repression is
significantly higher (p < 0.005) for LSD 1 enhancer-bound repressed genes compared to all
repressed genes.
Figure 3: LSD1 is associated with a NuRD complex at active enhancers in ESCs
a, NuRD components occupy enhancers and core promoter regions of actively transcribed genes.
ChIP-Seq binding profiles (reads/million) for transcription factors (Oct4, Sox2, Nanog),
coactivator (Medi), and chromatin regulators (LSD1, Mi-2, HDAC1, HDAC2), at the Oct4
(Pou5fl) and Lefty] loci in ESCs, with the y-axis floor set to 1. Gene models, and previously
described enhancer regions (Okumura-Nakanishi et al., 2005; Yeom et al., 1996; Nakatake et al.,
2006) are depicted below the binding profiles.
b, LSD 1 is associated with NuRD components Mi-2p, HDAC 1, HDAC2, as well as CoREST.
LSD 1 and HDAC 1 are detected by Western blot (WB) after immunoprecipitation of crosslinked
chromatin using LSD1, HDAC1, HDAC2, Mi-20, or CoREST antibodies. IgG is shown as a
control.
174
c, LSD 1 and HDAC 1 are detected by Western blot (WB) after immunoprecipitation of
uncrosslinked nuclear extracts (NE) using LSD1, HDAC 1, HDAC2, Mi-20, or CoREST
antibodies. IgG is shown as a control.
d, The occupancy of enhancers by NuRD proteins (Mi-2, HDAC 1 and HDAC2) is significantly
greater than the occupancy by CoREST or REST (p < 10-9). The height of the bars represents the
percentage of the 3,838 enhancers co-occupied by LSD1, NuRD proteins (Mi-2, and either
HDAC 1 or HDAC2), CoREST and REST.
Figure 4: LSD1 is required for H3K4mel removal at ESC enhancers
a, H3K4me 1 levels are reduced at LSD 1-occupied enhancers upon ESC differentiation, and this
effect is partially blocked upon TCP treatment.
b, TCP treatment does not affect the increase in H3K4me 1 levels at trophectodermal genes
during differentiation. ChIP-Seq binding profiles (reads/million) for Oct4 and LSD1 at the Lefty]
and Gata2 loci in ESCs. Below these profiles, histone H3K4mel levels are shown for ZHBTc4
control ESCs, cells treated with doxycycline for 48 hours to repress Oct4 and induce
differentiation (ESCs +Dox), and ESCs treated with doxycycline and TCP (ESCs +Dox,TCP).
For appropriate normalization, ChIP-Seq data for histone H3K4meI is shown as rank normalized
reads/million with the y-axis floor set to 1 (Supplementary Information). Gene models, and
previously described enhancer regions (Nakatake et al., 2006; Ray et al., 2009) are depicted
below the binding profiles.
c, Sum of the normalized H3K4meI density +/- 250 nucleotides surrounding LSDI-occupied
175
enhancer regions before and during trophectoderm differentiation in the presence or absence of
TCP. The associated genes were identified based on their proximity to the LSD 1-occupied
enhancers.
d, Sum of the normalized H3K4meI density +/- 250 nucleotides surrounding 1,722 LSD1occupied enhancers before and during differentiation in the presence or absence of TCP. Of the
2,755 LSD 1-occupied enhancers having reduced levels of H3K4mel upon differentiation, 63%
(1,722) display higher H3K4me l levels after TCP treatment (p < 10-16).
e, Heatmap displaying the sum of the normalized H3K4me 1 density +/- 250 nucleotides
surrounding the 1,722 LSD 1-occupied enhancers that retained H3K4me 1 in TCP-treated ESCs
compared to untreated control differentiating ESCs. Color scale indicates ChIP-Seq signal in
normalized reads per million.
176
a
Suz12
Pol 11
Active genes
6,899/7,339
P < 10-300
LSD1
Active genes
6,844/7,339
P< 10-300
Active genes
147/7,339
P= 1
Silent genes
140/6,976
P= 1
Silent genes
349/6,976
P=1
Silent genes
1,583/6,976
P= 1
b
15kb
701
Transcription factors
A
LI
Li
Li
I
Coactivator [15
Chromatin regulator [10.
15kb
Oct4
Sox2
Ii
ILI
70
70
Bivalent genes
2,413/3,094
P< 10-300
Bivalent genes
681/3,094
P= 1
Bivalent genes
2,003/3,094
P= 0.173
L
Nanog
A
Med1
LSD1
Pol 11
Transcription 0
apparatus r10 1
I
TBP
=
H3K4me1
.,H
Histonemodifications
H3K79nme2
II
Enhancers
Core promoters
I
I
I
=4
C
Oct4 Med
P300 LSD1
d
T
LSD1
0-
Tf
0
C
E
0
w
H3K4me3
-g
+5 -5
Distance from Oct4, Sox2, Nanog
0-
I
+5 -5
Distance from TSS (kb)
and Mediator-occupied site (kb)
Figure 1
177
eftyl
a
b
Oj*e
ZHBTc4 ESCs
Oct4ON
-6 h
0 -I
TCP
Time (h)
Early trophectoderm
24 h
-
+
24
'0 24 480
87
WB Oct4
48 h
WB LSD1
WB Tubulin
Doxycycline:
c
24 h
I
IPhase I Hoe II Oct4 IISSEA-1I FPhase II Hoe II Oct4 IISSEA-1l
U
U)
wU +
0-
U)0
U,'
d
ai)
Zic3
C c
Sox2
ESC genes
Fbox15 Dppa5 Octl
Trophectoderm genes
Nanog
0
e
4.5
6
15
2 30
4
0
111.51
L)
0
3
O
C0)
cC
-24
-9
Repressed
genes
0
-4.5
-3
1
-2.1
ESCs + Dox
ESCs + Dox,TCP
-0.5
-1.5
0-
0)--
-2.5
Figure 2
178
0
Tead2
CIdn4
21 11
Plac1
DkkIl1
1
5
Cdx2
Esx1
a
15kb
I 4
Transcription factors
I A.
il
70
IA
i
70
I
Coactivator [15
-
Nanog
A.
Uk
A
.
Medi
LSD 1
Mi-23
--
LkL
I i~
5
[
[
II
HDAC2
I
G"Lfty1
d
C
Crosslinked
HDAG1
IAL
I A.
b
WCE
Sox2
L
Chromatin regulators
Core promoter
Oct4
A
I
[ I,.,.
5
A
IA
IA
i~
10
Enhancers
15kb
70
NE
m--
Uncrosslinked
i
i
100
0.
x
10% -/0MI
[
I
10%
WB LSD1
4"
ER
C
WB LSD1
WB HDAC 1
M
50
C
w
WB HDAC1
(< q c
0
Figure 3
179
a
ab
15 kb
Transcription factor
Chromatin regulator
[70
ESC
7
E SC
SDox C
7
Enhancers
Core promoter
--------
[10
H3K4me1
AL
-
F
.,.
80
120
E6= 60
90
60
IEE40
22
i
I
0o +
Dox-++
TCP -- +
e
SalI4
160
120
3I1
Trim28
80
160
130
100
40
7
-++
-++
--
--
+
+
Esrrb
+
+
H3K4me1
H3K4mel
H3K4me1
120
100
80
60
40
20
60
0 -++
--
+
0
Gata2
d 1,722 LSD1Akt2
LIFR
Dppal 5 occupied enhancers
100
160
120
140
120
90
(P ?- 75
105
80
60
70
40
30
35
0
0
-++
-++
-++
--
--
---
+
+
C\j
-
H3K4me1
Phc3
H' 3K4me1
+
_LSD1
H3K4me1
Leftyl
0.
Dox TCP -
20 kb
Oct4
1
I
E
Sox2
4
10
7
LSD1
C
Leftyl
1
Oct4
Figure 4
180
-++
---
C 2 50
teo '
-+
-++
--
+
25
Dox -++
TCP -- +
SUPPLEMENTARY INFORMATION
Enhancer Decommissioning by LSD1 During Embryonic Stem Cell Differentiation
Warren A. Whyte, Steve Bilodeau, David A. Orlando, Heather A. Hoke, Garrett M. Frampton,
Charles T. Foster, Shaun M. Cowley, and Richard A. Young
CONTENTS
SuDmlementary Tables
Suiolementary Data Files
Supplementary Discussion
Supplementary Experimental Procedures
Cell Culture Conditions and Differentiation Assays
181
Embryonic Stem Cells (ESCs)
Generationof LSDJ Knockout ESC lines (SupplementaryFig. 4, 5)
Generationand Expression ofLSDJ Expression Constructsin ESCs
(Supplementary Fig. 5)
Differentiation of ESCs (Fig.2, 4, and Supplementary Fig.2, 3, 9, and 10)
Inhibition ofp300 using HAT inhibitor C-646 (SupplementaryFig. 12)
Lentiviral Productionand Infection (SupplementaryFig.2, 3, 4, 5, and 9)
Immunofluorescence (Fig.2b and Supplementary Fig. 3, 4, 5, and 9)
Alkaline PhosphataseStaining (SupplementaryFig. 2, 4, 5, and 9)
RNA Extraction, cDNA, and TaqMan Expression Analysis (Supplementary Fig. 1,
3, 5, and 9)
Chromatin Immunoprecipitation (ChIP)
Antibody Specificity (SupplementaryFig. 1, 6, and 9)
ChIPProtocol
Gene Specific ChIPs (SupplementaryFig. 9, 10, 11, 12)
ChIP-Seq Sample Preparation and Analysis
Sample Preparation
Polony Generationand Sequencing
182
ChIP-Seq DataAnalysis
Assigning ChIP-SeqEnrichedRegions to Genes (Supplementary
Table 1)
Enrichment of LSD1, Pol II, and Suz12 at Genes (Fig. 1)
Definition of Enhancerand Core Promoter (Fig.1, 4, Supplementary Fig. 8, 13,
and Supplementary Table 1)
DeterminingLSD] density at Enhancerand Core Promoter (Supplementary Fig.
1)
ChIP-Seq Density Heatmaps (Fig.1, 4, Supplementary Fig. 8, and 13)
Calculation of the StatisticalSignificance of the Overlap Between Sets of
Genomic Regions (Fig.1, 3, 4, Supplementary Fig. 8, and 12)
Rank Normalization of H3K4meJ ChIP-Seq Data (Fig.4, SupplementaryFig. 13,
and Supplementary Table 5)
Gene Ontology (GO) Analysis (SupplementaryFig. 7)
Public availability of ChIP-Seq Datasets
Microarray Analysis (Fig. 2)
Cell Cultureand RNA isolation
183
Microarrayhybridization and Analysis
Publicavailabilityof MicroarrayGene Expression Datasets
Protein Extraction and Western Blot Analysis (Fig. 2, Supplementary Fig. 1, 4, 5, 6,
and 9)
ChIP-Western and Co-Immunoprecipitation (Fig. 3)
Supplementary References
184
Supplementary Tables
Supplementary Table 1 - Summary of occupied genes and regions
Supplementary Table 2 - Summary of LSD I occupancy at active and poised enhancers
Supplementary Table 3 - Gene expression changes 48 hours after Oct4 repression in ZHBTc4
DMSO (control)- and TCP-treated cells
Supplementary Table 4 - Genes co-occupied by LSD 1 and REST
Supplementary Table 5 - Enhancer decommissioning 48 hours after Oct4 depletion in ZHBTc4
DMSO (control)- and TCP-treated cells
Supplementary Data File 1
Supplementary Data File 1 contains ChIP-Seq data in compressed WIG format (WIG.GZ) for
upload into the UCSC genome browser (Kent et al., 2002). This file contains data for
mESH3K4mel, DMSOH3K4mel, DMSO_48HRH3K4mel, TCP_48HRH3K4mel,
mESH3K4me3, mESH3K79me2, mESH3K36me3, mES_Oct4, mESSox2, mESNanog,
mESPol II, mESTBP, mESLSD1, mESp300, mESREST, mESCoREST, mES_Suzl2,
mESH3K27ac, mESHDAC1, mESHDAC2, mESMi-2b, mESDMSOH3K27ac,
mESDMSOH3K4meI, mES_C646_H3K27ac, mESC646_H3K4meI, WCEDMSO,
185
WCE_48HRDMSO, WCE_48HRTCP, and WCEmES.
The first track for each data set contains the ChIP-Seq density across the genome in 25bp bins.
The minimum ChIP-Seq density shown in these files is 1.0 reads per million total reads. For
DMSOH3K4mel, DMSO_48HRH3K4mel, and TCP_48HRH3K4mel datasets, the
minimum ChIP-Seq density is 1 normalized read per million total reads. Subsequent tracks
identify genomic regions identified as enriched at a p-value threshold of 10-9.
Supplementary Discussion
To address LSD 1 function in ESCs and during ESC differentiation, cells were treated with the
monoamine oxidase inhibitors (MAOIs) tranylcypromine (TCP) or pargyline (Prg).
Mechanistically, LSD 1 is unique relative to other demethylases, as it demethylates lysine
residues via a flavin-adenine dinucleotide-dependent reaction (Forneris et al., 2005; Shi et al.,
2004). This reaction is inhibited by MAOIs, which are used in the treatment of certain
psychiatric and neurological disorders (Lee et al., 2006; Schmidt and McCafferty, 2007; Yang et
al., 2007; Mimasu et al., 2010; Liang et al., 2009). Furthermore, TCP inhibits LSD1 at levels
comparable to MAO inhibition of clinical mitochondrial MAOI targets (Yang et al., 2007;
Mimasu et al., 2010). Finally, the effects of inhibition by TCP and Prg were highly similar in our
assays. Therefore, we found TCP and Prg suitable to study LSD 1 activity during differentiation
of ES cells.
186
Previous studies demonstrated that LSDl is required for mouse development beyond e6.5 (Foster
et al., 2010). The differentiation defect is recapitulated in vitro in embryoid body (EB) formation
assays, where ESCs are plated in suspension in media lacking leukemia inhibitory factor (LIF),
causing ESC differentiation. In these assays, deletion of LSD 1 leads to considerable cell death
and formation of small EBs, with the remaining cells expressing markers of all three germ layers
(Foster et al., 2010; Wang et al., 2009a). The differentiation defect is also recapitulated in vitro
in the Oct4 depletion assay used in the present study, where ESCs preferentially differentiate into
the trophectodermal lineage. In this assay, inhibition of LSD 1 with either TCP or Prg leads to
considerable cell death, and surviving cells express trophectodermal differentiation markers.
Thus, very similar effects are observed in these two assays.
In the Oct4 depletion assay used in the present study, full repression of key ESC regulators such
as Nanog is not achieved in the presence of LSD 1 inhibitors 48 hours after Oct4 depletion is
initiated. Similar results were obtained for Nanog in the assay used by Foster et al. (2010) with
LSD1 mutant ESCs 48 hours after LIF removal (Foster et al., 2010). In this latter assay, Nanog
levels were ultimately fully reduced in ESCs lacking LSD 1. Our interpretation of all these data is
that inhibition of LSD 1 delays differentiation because there is a delay in reducing levels of key
ESC regulators whose enhancers are occupied by LSD 1.
Previous studies also report that deletion of LSD I protein destabilizes Co-REST, which leads to
a less active LSD I-CoREST-HDAC complex (Foster et al., 2010). Accordingly, developmental
regulators targeted by the LSD 1 -CoREST-HDAC complex are de-repressed, and this may
187
explain the defect observed in ESC differentiation. In experiments using small molecule
inhibitors of LSD 1 activity, we did not observe significant changes in genes associated with
LSD 1-REST co-occupied sites. One explanation for this discrepancy may be that loss of LSD 1
proteins destabilizes the CoREST-HDAC complex, while small molecules inhibiting LSD 1
enzymatic activity may not alter the stability of the complex, thereby giving a different
transcriptional phenotype. Thus, the different methods (gene deletion versus enzymatic
inhibition) of LSD 1 inhibition may explain the discrepancy in gene expression.
Differences were observed for LSD1 function in mouse and human ESCs. Although mouse ESCs
fail to differentiate into embryoid bodies in absence of LSD1 (Foster et al., 2010; Wang et al.,
2009a; Wang et al., 2007), human ESCs differentiate when LSD 1 levels are reduced (Adamo et
al., 2011). In both mouse and human ESCs, LSD1 is found at Oct4 and Nanog -occupied regions.
Similar to our observations in the Oct4 depletion assay, a subpopulation of differentiating human
ESCs with reduced LSD 1 levels retained the expression of pluripotency markers reduced
(Adamo et al., 2011).
Supplementary Experimental Procedures
Cell Culture Conditions and Differentiation Assays
Embryonic Stem Cells
188
V6.5, ZHBTc4, and E14 murine ESCs were grown on irradiated murine embryonic fibroblasts
(MEFs), unless otherwise stated. Cells were grown under standard ESC conditions as described
previously (Boyer et al., 2005). Briefly, cells were grown on 0.2% gelatinized (Sigma, G1890)
tissue culture plates in ESC media; DMEM-KO (Invitrogen, 10829-018) supplemented with 15%
fetal bovine serum (Hyclone, characterized SH3007103), 1000 U/mL LIF (ESGRO, ESG 1106),
100 PM nonessential amino acids (Invitrogen, 11140-050), 2 mM L-glutamine (Invitrogen,
25030-081), 100 U/mL penicillin, 100 pg/mL streptomycin (Invitrogen, 15140-122), and 8
nL/mL of 2-mercaptoethanol (Sigma, M7522).
Generation ofLSDJ knockout ESC lines (Supplementary Fig. 4, 5)
An E14 ESC line expressing a Cre-estrogen receptor fusion protein from the ROSA26 locus was
used to generate cells in which LSD1 can be inactivated conditionally. Briefly, LSD 1 Lox/A3
ESCs were generated as previously described (Adamo et al., 2011), in which one LSD1 allele has
exon 3 flanked by LoxP sites (floxed) and the second has exon 3 deleted. Induction of Cre
activity by addition of 4-hydroxytamoxifen (4-OHT, Sigma, H7904) to the growth medium
resulted in complete recombination of the remaining floxed allele (to generate LSD 1 A3/A3)
within 6 hours. Cells were grown in the absence of irradiated MEFs under standard ESC
conditions as described previously (Boyer et al., 2005).
Generationand Expression of LSD] Expression Constructs in ESCs (Supplementary Fig. 5)
189
Two forms of human LSD 1 were generated by PCR with primers containing tails of family D
vector homology for cloning into pCAGGs expression vectors. An enhanced green fluorescent
protein (EGFP) tag was added to the N-terminus of either the full-length LSD 1 protein, or the
catalytically inactive form of LSD 1 with a lysine to alanine mutation introduced at residue 661
by site-directed mutagenesis (Binda et al., 1999).
For expression of LSDI constructs in ESCs, LSDl Lox/A3 and A3/A3 E14 ESCs were plated in
6-well culture plates. The following day, cells were transfected with 2gg pCAGGs construct and
2pg of the pMONO-hygro plasmid containing the hygromycin resistance gene (Invivogen,
pmonoh-mcs) using 1 Opl Lipofectamine 2000 (Invitrogen, 11668-019). After 24 hours, the
media was removed and replaced with ESC media containing 400pg/ml hygromycin. 24 hours
after hygromycin addition, hygromycin concentration was reduced to 200pg/ml for the duration
of the experiment.
Differentiationof ESCs (Fig.2, 4, and Supplementary Fig. 2, 3, 9, and 10)
For immunofluorescence, monitoring H3K4meI levels and ESC expression analysis during
differentiation following either DMSO, TCP or Prg treatment, ZHBTc4 ESCs were split off
MEFs, placed in a tissue culture dish for 45 minutes to selectively remove the MEFs, and plated
on either 6-well or 15cm cell culture plates. The following day, cells were treated with either
DMSO, ImM TCP or 3mM Prg for 6 hours in ESC media. After 6 hours, the media was
removed and replaced with ESC media containing 2pg/ml doxycyline and either DMSO, TCP or
190
Prg. 24 hours after doxycycline addition, the media was replaced with ESC media containing
doxycycline and either DMSO, TCP or Prg for another 24 hours (48 hours total).
Inhibition ofp300 using HAT inhibitorC-646 (SupplementaryFig. 12)
During differentiation, it is presently unclear which acetylated histone residue(s) need to be
deacetylated to trigger activation of LSD 1 demethylation activity. While p300 and H3K27Ac are
very good candidates at enhancers, many other HATs (CBP, GCN5, Tip60, Elp3, Myst3, Myst4)
are active in ESCs (Lin et al., 2007; Ura et al., 2011; Miyabayashi et al., 2007). Therefore,
regulation of acetylation levels at enhancers is most likely to be the result of a combination of
HATs and HDACs complexes.
To test if acetylation of H3K27 prevents demethylation of H3K4 by LSD 1, we treated ESCs with
the p300-specific acteyltransferase inhibitor C-646 (Bowers et al., 2010). V6.5 ESCs were split
off MEFs, placed in a tissue culture dish for 45 minutes to selectively remove the MEFs, and
plated in 15cm cell culture plates. The following days, cells were treated with either DMSO
(control) or C-646 (Tocris, 4200) for 24 hours. Cells were then crosslinked and collected to
generate ChIP-seq datasets for H3K27ac and H3K4me 1.
LentiviralProductionand Infection (SupplementaryFig. 2, 3, 4, 5, and 9)
Lentivirus was produced according to Open Biosystems Trans-lentiviral shRNA Packaging
System (TLP4614). The shRNA constructs targeting LSD1, Oct4, CoREST and Mi-2b are listed
191
below. The shRNAs targeting either GFP (RHS4459) or Luciferase (SHCO07) were used as
controls.
GFP
RHS4459
Luciferase
SHCO07
LSD1
TRCN0000071373
Oct4
TRCN0000009611
Mi-2b #1
TRCN0000086143
Mi-2b #2
TRCN0000086145
CoREST
TRCN0000071371
For GFP, LSD 1 and Mi-2b, ZHBTc4 ESCs were split off MEFs, placed in a tissue culture dish
for 45 minutes to selectively remove the MEFs, and plated on either 6-well or 12-well cell
culture plates. The following day, cells were infected in ESC media containing 8 pg/ml
polybrene (Sigma, H9268-10G). After 24 hours the media was removed and replaced with ESC
media containing 3.5 gg/mL puromycin (Sigma, P8833). ESC media with puromycin was
changed daily. Three days post infection, the media was removed and replaced with ESC media
containing 2pg/ml doxycycline to shutdown Oct4 and induce differentiation. Cells were
harvested 48 hours later.
For Luciferase and Oct4, LSD 1 Lox/A3 or A3/A3 E14 ESCs were plated on 6-well culture plates.
The following day, cells were infected in ESC media containing 8 pg/ml polybrene. Cells
192
transfected with LSD 1 expression constructs were infected in ESC media lacking antibiotics and
polybrene. After 16 hours the media was removed and replaced with ESC media containing 3.5
pg/mL puromycin (Sigma, P8833). The media of ESCs transfected with the LSD 1 expression
constructs also contained hygromycin, as described earlier. ESC media with either puromycin, or
puromycin and hygromycin, was changed daily. Cells were harvested 72 hours later.
Immunofluorescence (Fig.2b and SupplementaryFig. 3, 4, 5, and 9)
Following crosslinking, the cells were washed once with PBS, twice with blocking buffer (PBS
with 0.25% BSA, Sigma, A3059-10G) and then permeabilized for 15 minutes with 0.2% Triton
X-100 (Sigma, T8797-100ml). After two washes with blocking buffer cells were stained
overnight at 4 degrees C for either Oct4 (Santa Cruz Biotechnology, sc-908 lx; 1:200 dilution) or
SSEA 1 (mc-480, DHSB, 1:20 dilution) and washed twice with blocking buffer. Cells were
incubated for 1 hour at room temperature with either goat anti-rabbit-conjugated Alexa Fluor
488, goat anti-rabbit-conjugated Alexa Fluor 647 or goat anti-mouse conjugated 568 (Invitrogen;
1:1000 dilution) and Hoechst 33342 (Invitrogen; 1:2000 dilution). Finally, cells were washed
twice with blocking buffer and twice with PBS before imaging. Images were acquired on either
Nikon Inverted TE300 or Zeiss Axiovert 200m Inverted microscopes with a Hamamatsu Orca
camera. Openlab (http://www.improvision.com/products/openlab/) was used for image
acquisition.
Alkaline PhosphataseStaining (SupplementaryFig. 2, 4, 5, and 9)
Staining of ZHBTc4 cells for alkaline phosphatase was achieved using the Alkaline Phosphatase
Detection Kit (Millipore, SCRO04). Briefly, cells were crosslinked at various timepoints before
193
addition of Fast Red Violet solution and Napthol AS-BI phosphate solution. Cells were
visualized on a Nikon Inverted TE300 with a Hamamatsu Orca camera. Openlab
(http://www.improvision.com/products/openlab/) was used for image acquisition.
RNA Extraction, cDNA, and TaqMan Expression Analysis (SupplementaryFig. 1, 3, 5, and 9)
RNA utilized for real-time qPCR was extracted with TRIzol according to the manufacturer
protocol (Invitrogen, 15596-026). Purified RNA was reverse transcribed using Superscript III
(Invitrogen) with oligo dT primed first-strand synthesis following the manufacturer protocol.
Real-time qPCR were carried out on the 7000 ABI Detection System using the following
Taqman probes according to the manufacturer protocol (Applied Biosystems).
Gapdh Mm99999915_gl
LSDI1
MmO 1181030_g1
Mi-2b
MmI190896_ml
CoREST
Mm03053471_sl
Sox2
Mm00488369_sI
Nanog
Mm02019550_sl
Leftyl
Mm00438615_ml
194
Lefty2
Mm00774547_ml
Esrrb
Mm0044241I_ml
Trim28Mm00495594_ml
Klf4
MmO0516104_g1
Oct4
Mm03053917_gI
Apobec 1
Mm00482895_ml
Dppa3
MmO I184198_g1
TclI
Mm00493477_ml
Expression levels were normalized to Gapdh levels. All comparisons were made relative to
either DMSO (control)-treated, Luciferase, or GFP-infected cells.
Chromatin Immunoprecipitation (ChIP)
Antibody Specificity (Supplementary Fig. 1, 6, and 9)
For LSD 1 (AOF2/KDM 1 A)-occupied genomic regions, we performed ChIP-Seq experiments
using Abcam ab17721 rabbit polyclonal antibody. The antibody was raised with a synthetic
peptide conjugated to KLH derived from within residues 800 to the C-terminus of human LSD1.
Antibody specificity was determined by shRNA-mediated knockdown (Open Biosystems) of
195
LSD 1 proteins, followed by Western blot analysis. Knockdowns were carried out using the
following shRNAs according to the manufacturer protocol (Open Biosystems). Cells were
infected with indicated short hairpins for 5 days, followed by protein extraction and Western
blotting.
GFP
RHS4459
LSDI1
TRCN0000071374
For Mi-2b (CHD4)-occupied genomic regions, we performed ChIP-Seq experiments using
Abcam ab72418 rabbit polyclonal antibody. The antibody was raised with a synthetic peptide
conjugated to KLH derived from within residues 25 to 75 of human CHD4. Antibody specificity
was determined by shRNA-mediated knockdown (Open Biosystems) of Mi-2b proteins,
followed by Western blot analysis. Knockdowns were carried out using the following shRNAs
according to the manufacturer protocol (Open Biosystems). Cells were infected with indicated
short hairpins for 5 days, followed by protein extraction and Western blotting.
GFP
RHS4459
shMi-2b #1
TRCN0000086143
shMi-2b #2
TRCN0000086147
196
For HDAC 1-occupied genomic regions, we performed ChIP-Seq experiments using Abcam
ab7028 rabbit polyclonal antibody. The antibody was raised with a synthetic peptide conjugated
to KLH derived from within residues 466 to 482 of human HDAC 1. Antibody specificity was
previously determined (Wang et al., 2009b).
For HDAC2-occupied genomic regions, we performed ChIP-Seq experiments using Abcam
ab7029 rabbit polyclonal antibody. The antibody was raised with a synthetic peptide conjugated
to KLH derived from within residues 471 to 488 of human HDAC2. Antibody specificity was
previously determined (Wang et al., 2009b).
For REST (NRSF)-occupied genomic regions, we performed ChIP-Seq experiments using
Millipore 07-579 rabbit polyclonal antibody. The antibody was raised with a GST fusion protein
corresponding to residues 801-1097 of human REST. Antibody specificity was previously
determined (Singh et al., 2008).
For CoREST-occupied genomic regions, we performed ChIP-Seq experiments using Abcam
ab32631 rabbit polyclonal antibody. The antibody was raised with a synthetic peptide
conjugated to KLH derived from within residues 450 to the C-terminus of human CoREST.
Antibody specificity was determined by shRNA-mediated knockdown (Open Biosystems) of
CoREST proteins, followed by Western blot analysis. Knockdowns were carried out using the
following shRNAs according to the manufacturer protocol (Open Biosystems). Cells were
infected with indicated short hairpins for 5 days, followed by protein extraction and Western
blotting.
197
GFP
RHS4459
CoREST
TRCN0000071371
For H3K4me 1-occupied genomic regions, we performed ChIP-Seq experiments using Abcam
ab8895 rabbit polyclonal antibody. The antibody was raised with a synthetic peptide
conjugated to KLH derived from within residues 1 to 100 of human H3K4me 1. Antibody
specificity was previously determined (Meissner et al, 2008).
For p300-occupied genomic regions, we performed ChIP-PCR experiments using Santa Cruz sc584 rabbit polyclonal antibody. The antibody was raised with a synthetic peptide derived
from the N-terminus of human p300. Antibody specificity was previously determined (Chen et
al., 2008).
For H3K27Ac-occupied genomic regions, we performed ChIP-PCR experiments using Abcam
ab4729 rabbit polyclonal antibody. The antibody was raised with a synthetic peptide
conjugated to KLH derived from within residues 1 - 100 of human histone H3, acetylated at K27.
Antibody specificity was previously determined (Creyghton et al., 2010).
ChIP Protocol
Protocols describing chromatin immunoprecipitation materials and methods have been
previously described (Boyer et al., 2006). v6.5 or ZHBTc4 ESCs were grown to a final count of
198
5-10 x 107 cells for each ChIP experiment. Cells were chemically crosslinked by the addition of
one-tenth volume of fresh 11% formaldehyde solution for 15 minutes at room temperature. Cells
were rinsed twice with IX PBS and harvested using a silicon scraper and flash frozen in liquid
nitrogen. Cells were stored at -80'C prior to use. Cells were resuspended, lysed in lysis buffers
and sonicated to solubilize and shear crosslinked DNA. Sonication conditions vary depending on
cells, culture conditions, crosslinking and equipment.
For LSD1, Mi-2b, HDAC1, HDAC2, REST, CoREST, p300, H3K27Ac, and H3K4mel the
sonication buffer was 20mM Tris-HCl pH8, 150mM NaCl, 2mM EDTA, 0.1% SDS, 1% Triton
X-100. We used a Misonix Sonicator 3000 and sonicated at approximately 24 watts for 10 x 30
second pulses (60 second pause between pulses). Samples were kept on ice at all times. The
resulting whole cell extract was incubated overnight at 4 degrees C with 100ul of Dynal Protein
G magnetic beads that had been pre-incubated with approximately 10 ug of the appropriate
antibody. Beads were washed IX with the sonication buffer, IX with 20mM Tris-HCl pH8,
500mM NaCl, 2mM EDTA, 0.1% SDS, I%Triton X-100, 1X with 10mM Tris-HCl pH8,
250mM LiCl, 2mM EDTA, 1% NP40 and IX with TE containing 50 mM NaCl.
Bound complexes were eluted from the beads (50 mM Tris-Hcl, pH 8.0, 10 mM EDTA and 1%
SDS) by heating at 65 degrees C for 1 hour with occasional vortexing and crosslinking was
reversed by overnight incubation at 65'C. Whole cell extract DNA reserved from the sonication
step was also treated for crosslink reversal.
199
Gene Specific ChIPs (Supplementary Fig. 9, 10, and 1)
For gene specific ChIPs carried out in ZHBTc4 ESCs, approximately 5x 107 ES cells were grown
on two 15cm plates. Plates were treated or not treated with 2ng/ml doxycyline for 12 hours, and
crosslinked. SYBR Green real-time qPCR was carried out on the 7000 ABI Detection System
according to the manufacturer protocol (Applied Biosystems). Data was normalized to the whole
cell extract and control regions. Primer pairs are listed below.
Leftyl
5'-GTAGCCAGCAGACAGGACAA-3'
5'-ATCCCCAATCCACATTCACT-3'
Lefty2
5'-AGGCCTAGCTTTTGCATCAC-3'
5'-TCTCCCAGAGTCGATCTTCC-3'
Sal14
5'- GAAATAAACATCTGGGAGAAGGA-3'
5'- GGAAACCCCAGATTGAGAGA-3'
200
Sox2
5'- TGGCGAGTGGTTAAACAGAG-3'
5'- TAGCGAGAACTAGCCAAGCA-3'
Trim28
5'- GGTCTGCAATTGAAGGAAGG-3'
5'- TTAAACAGCAGGGGGTAAGG-3'
Esrrb
5'- CGAGCTTCAGCTGGCTATTT-3'
5'- GAGCTCCAGATCCCCTACAC-3'
Nanog
5'-GGAATTTCCTTCCCAGGTTT-3'
5'- GGTTGGAACTAGCTGTGTGG-3'
201
Ctrl (Olfr460)
5'-AACTGTTATTGTGCCCGTGA -3'
5'-CATTGCTCCAAGCAAAGAAA -3'
ChIP-Seq Sample Preparation and Analysis
All protocols for Illumina/Solexa sequence preparation, sequencing and quality control are
provided by Illumina (http://www.illumina.com/pages.ilmn?ID=203). A brief summary of the
technique and minor protocol modifications are described below.
Sample Preparation
Purified chromatin immunoprecipitated (ChIP) DNA was prepared for sequencing according to a
modified version of the Illumina/Solexa Genomic DNA protocol. Approximately 200 ng of ChIP
DNA was prepared for ligation of Solexa linkers by repairing the ends and adding a single
adenine nucleotide overhang to allow for directional ligation. A 1:200 dilution of the Adaptor
Oligo Mix (Illumina) was used in the ligation step. A subsequent PCR step with 18 amplification
cycles added additional linker sequence to the fragments to prepare them for annealing to the
Genome Analyzer flow-cell. Amplified material was purified by Qiaquick MinElute (Qiagen,
Valencia, CA) and a narrow range of fragment sizes was selected by separation on a 2% agarose
gel and excision of a band between 150-300bp, representing ChIP fragments between 50 and 200
202
nt in length and -l00bp of primer sequence. The DNA was purified from the agarose and diluted
to 10nM for loading on the flow cell.
For multiplexed samples, libraries were prepared using the Illumina TruSeq adapters (to enable
multiplexing) and prepared using Beckman-Coulter's SPRIworks system. For library
preparations, ChIP samples were used in their entirety and whole-cell extracts control samples
were prepared using 100ng. Adapters were diluted to 1:200. Size selection was 200-400bp
before PCR, and the samples were amplified using KAPA Hi-Fi polymerase and 18 cycles of
PCR according to manufacturer's cycling recommendations. Amplified material was purified
using Agencourt Ampure XP beads using a 0.93 ratio of beads to sample.
Polony Generation and Sequencing
The DNA library (2-5pM) was applied to the flow-cell (8 samples per flow-cell) using the
Cluster Station device from Illumina. The concentration of library applied to the flow-cell was
calibrated such that polonies generated in the bridge amplification step originate from single
strands of DNA. Multiple rounds of amplification reagents were flowed across the cell in the
bridge amplification step to generate polonies of approximately 1,000 strands in 1mm diameter
spots. Double-stranded polonies were visually checked for density and morphology by staining
with a 1:5000 dilution of SYBR Green I (Invitrogen) and visualizing with a microscope under
fluorescent illumination. Validated flow-cells were stored at 4*C until sequencing. Flow-cells
were removed from storage and subjected to linearization and annealing of sequencing primer on
the Cluster Station. Primed flow-cells were loaded into the Illumina Genome Analyzer II or Hi203
seq. After the first base was incorporated in the Sequencing by-Synthesis reaction the process
was paused for a quality control checkpoint. A small section of each lane was imaged and the
average intensity value for all four bases was compared to minimum thresholds. Flow-cells with
low first base intensities were reprimed and if signal was not recovered the flow-cell was
aborted. Flow-cells with signal intensities meeting the minimum thresholds were resumed and
sequenced for 26 cycles. For multiplexed samples Truseq V2.5 kits were used to cluster them on
the cBot and Truseq V2 were used to do a multiplex 40+7 cycle run on the Hi-seq.
ChIP-Seq Data Analysis
Images acquired from the Illumina/Solexa sequencer were processed through the bundled Solexa
image extraction pipeline which identified polony positions, performed base-calling and
generated QC statistics.
ChIP-Seq reads were aligned using the software Bowtie (Langmead et al., 2009) to NCBI build
36 (mm8) of the mouse genome with default settings. Sequences uniquely mapping to the
genome with zero or one mismatch were used in further analysis
When multiple reads mapped to the same genomic position, a maximum of two reads mapping to
the same position were used. ChIP-Seq datasets profiling the genomic occupancy of H3K36me2
(Marson et al., 2008), H3K79me2 (Marson et al., 2008), H3K4me3 (Marson et al., 2008),
H3K27me3 (Mikkelsen et al., 2007), Oct4 (Marson et al., 2008), Sox2 (Marson et al., 2008),
204
Nanog (Marson et al., 2008), RNA Polymerase 2 (Seila et al., 2008), TBP (Kagey et al., 2010),
Medi (Kagey et al., 2010), p300 (Chen et al., 2008), H3K4mel (Meissner et al., 2008),
H3K27ac (Creyghton et al., 2010) and Suzl2 (Marson et al., 2008) in mouse ESCs were
obtained from previous publications. Below is the list of ChIP-Seq datasets used and
corresponding GEO Accession numbers.
Dataset
GEO Accession Numbers
H3K27ac
GSM594578, GSM594579
H3K27me3
GSM307619
H3K36me3
GSM307152, GSM307153
H3K4meI1
GSM281695
H3K4me3
GSM307146
H3K79me2
GSM307150, GSM307151
Medl
GSM560347, GSM560348
WCE mES
GSM307154, GSM307155, GSM560357
WCE mES (matching H3K4me 1 mES)
GSM307625
Nanog
GSM307140, GSM307141
Oct4
GSM307137
p300
GSM288359
205
RNA Pol I11
GSM318444
Sox2
GSM307138, GSM307139
Suz 12
GSM307144, GSM307145
TBP
GSM555160, GSM555162
DMSO H3K4mel
This Paper
48HR DMSO H3K4mel
This Paper
48HR TCP H3K4mel
This Paper
HDAC1
This Paper
HDAC2
This Paper
LSD1
This Paper
Mi-2b
This Paper
REST
GSM656525, This Paper
CoREST
This Paper
WCE - DMSO
This Paper
WCE - 48HR DMSO
This Paper
WCE - 48HR TCP
This Paper
All ChIP-Seq datastsets, including those obtained elsewhere, were analyzed using the methods
described below.
206
Analysis methods were derived from previously published methods (Marson et al., 2008;
Mikkelsen et al., 2007; Guenther et al., 2008; Johnson et al., 2007). Sequence reads from
multiple flow cells for each IP target and/or biological replicates were combined. For all
datasets, excluding H3K4mel, H3K4me3, H3K79me2, H3K36me3, H3K27ac, and H3K27me3
each read was extended 200bp, towards the interior of the sequenced fragment, based on the
strand of the alignment. For H3K4meI, H3K4me3, H3K79me2, H3K36me3, H3K27ac, and
H3K27me3 datasets, each read was extended 600bp towards the interior and 400bp towards the
exterior of the sequenced fragment, based on the strand of the alignment. Across the genome, in
25 bp bins, the number of extended ChIP-Seq reads was tabulated. The 25bp genomic bins that
contained statistically significant ChIP-Seq enrichment were identified by comparison to a
Poissonian background model. Assuming background reads are spread randomly throughout the
genome, the probability of observing a given number of reads in a genomic bin can be modeled
as a Poisson process in which the expectation can be estimated as the number of mapped reads
multiplied by the number of bins (8 for all sequences datasets except H3K4me 1, H3K4me3,
H3K79me2, H3K36me3, H3K27ac, and H3K27me3, which was 40) into which each read maps,
divided by the total number of bins available (we estimated 70% of the genome). Enriched bins
within 200bp of one another were combined into regions.
The Poissonian background model assumes a random distribution of background reads, however
we have observed significant deviations from this expectation. Some of these non-random
events can be detected as sites of apparent enrichment in negative control DNA samples and can
207
create many false positives in ChIP-Seq experiments. To remove these regions, we compared
genomic bins and regions that meet the statistical threshold for enrichment to a set of reads
obtained from Solexa sequencing of DNA from whole cell extract (WCE) in matched cell
samples. We required that enriched bins and enriched regions have five-fold greater ChIP-Seq
density in the specific IP sample, compared with the control sample, normalized to the total
number of reads in each dataset. This served to filter out genomic regions that are biased to
having a greater than expected background density of ChIP-Seq reads. A summary of the bound
regions and genes for each antibody is provided (Supplementary Table 1).
Assigning ChIP-Seq EnrichedRegions to Genes (Supplementary Table 1)
The complete set of RefSeq genes was downloaded from the UCSC table browser
(http://genome.ucsc.edu/cgi-bin/hgTables) on June 1, 2010. For all datasets except H3K4mel,
H3K4me3, H3K79me2, H3K36me3, H3K27ac, and H3K27me3, genes with enriched regions
within 10kb of their transcription start site were called bound. For H3K4meI, H3K4me3,
H3K79me2, H3K36me3, H3K27ac, and H3K27me3 datasets, genes with enriched regions within
the gene body were called bound.
Enrichment ofLSDJ, Pol II, and Suzl 2 at genes (Fig.1)
Each gene in the mouse genome was classified into active, bivalent, or silent groups based on the
presence of co-occupancy of H3K4me3 (GSE 11724), H3K79me2 (GSE 11724), and H3K27me3
(GSM307619). See the table below for a description of the number of genes in each category as
208
well as the classification rules. Using a hypergeometric test, the p-value for enrichment of LSD 1,
Pol II, and Suz 12 binding to the genes in each of these classes was determined.
Classification rules
H3K4me3
H3K79me2
H3K27me3
Number
within +/- 2kb
within first
within +/-
of genes
of the TSS
5kb of gene
5kb of the
body
TSS
Active
Required
Required
-
7339
Bivalent
Required
-
Required
3094
Silent
Absent
Absent
-
6976
Definition of Enhancerand Core Promoter (Fig.1, 4, Supplementary Fig. 8, 13, and
Supplementary Table 1)
An enhancer was defined as a MedI high-confidence enriched region that is overlapped at least
lbp by enriched regions of Oct4, Sox2, and Nanog in mES cells. Using this definition we
identified 3,838 enhancers, and the genomic locations of those enhancers are listed in
Supplementary Table 1. These enhancers were assigned to the nearest genes to determine the
neighboring core promoter regions.
209
Determining LSD] density at Enhancer and Core Promoter (Supplementary Fig. 1)
The mean LSDl density at enhancers (15.91) and core promoters (13.46) was determined by
calculating the sum of LSD 1 density within 5kb of each enhancer and core promoter, and then
taking the average across the 3,838 enhancers and core promoters. A t-test was then performed to
obtain a p-value (p < 10-16) indicating that LSD 1 density was statistically larger at enhancers
compared to core promoters.
ChIP-Seq Density Heatmaps (Fig. 1, 4, Supplementary Fig. 8, and 13)
Selected enriched regions were aligned with each other according to the position of either Med 1
(Fig. Ic, 4e, and Supplementary Fig. 5b) or the TSS (Fig. Id). For each experiment, the ChIPSeq density profiles were normalized to the density per million total reads. Heatmaps were
generated using Java Treeview (http://itreeview.sourceforge.net/) with color saturation as
indicated.
Calculation of the Statistical Significance of the Overlap Between Sets of Genomic Regions (Fig.
1, 3, 4, and Supplementary Fig. 8)
In the manuscript, when assessing the overlap between two sets of genomic regions, we report pvalues calculated using a Monte Carlo method. In order to make this type of comparison, a
number of different statistical methods could be applied. One could use a chi-square test, and
compare the total size of the overlap of two sets of genomic regions to the size of this overlap
that would be expected at random based on the total size of the genome and the total size of the
210
genomic regions in each group. This method accurately models the expected overlap between
two datasets if they were truly randomly associated. Unfortunately, since the genome is so large,
this method will call the overlap between two datasets statistically significant even if it is not
much larger than would be expected at random.
Another way to assess the statistical significance of the overlap between two sets of genomic
regions is to use a Monte Carlo method. In Monte Carlo methods, a tremendous number of
simulated datasets are created and the overlaps between the simulated datasets are tabulated. The
statistical significant of the actual overlap is then estimated based on the frequency of an overlap
at least that large occurring in the simulated datasets. We used a Monte Carlo method to assess
the statistical significance of the overlap between two sets of genomic regions in this manuscript.
To create the simulated datasets, we shifted each of the genomic regions in one of the datasets a
random distance between -2,000bp and +2,000bp from its original position. For this manuscript,
Monte Carlo simulations were run one billion times.
We noted that, if two sets of genomic regions tend to occur in the same place, then this overlap is
statistically significant. This will be true whether the regions occur together 10% of the time, or
100% of the time. Thus, the test for statistical significance is not a very good at describing the
degree of overlap between two datasets. Consequently, in each instance that we assessed the
statistical significance of the overlap between two sets of genomic regions, we also calculated
the fold enrichment of the overlap between the two datasets. This was calculated as the total size
of the actual overlap between the two sets of regions and the overlap between the two datasets
211
that would be expected at random.
Rank Normalization of H3K4mel ChIP-Seq Data (Fig.4, Supplementary Fig. 13, and
Supplementary Table 5)
In order to compare the levels of H3K4me 1 between three conditions a rank normalization
method was used. The three cells types and conditions in which H3K4me 1 occupancy was
profiled were ZHBTc4 ESCs, ZHBTc4 ESCs treated with doxycycline for 48 hours, and
ZHBTc4 ESCs treated with doxycycline for 48 hours and TCP for 54 hours. In each of the three
datasets, the genomic bin with the greatest ChIP-Seq signal was identified. The average of these
three values was calculated, and the greatest bin in each dataset was assigned this average
value. This was repeated for all genomic bins from the greatest signal to the least, assigning each
the average ChIP-Seq signal for all bins of that rank across all datasets. Subsequently, the total
number of ChIP-Seq reads in the three H3K4mel datasets was calculated, and this value was
used to scale each dataset to the units, rank normalized reads/million.
Using this method, we computed the rank normalized reads/million density in each of the three
datasets in 10 bp bins across the genome. The H3K4me1 signal at an enhancer was then
calculated as the sum of the observed densities in the 50 bins (+/- 250bp) surrounding each of the
3,838 identified enhancers. This number captures the normalized H3K4mel signal in dataset,
and the value for each enhancer in each of the three datasets is reported in Supplementary Table
5.
212
Gene Ontology (GO) Analysis (SupplementaryFig. 7)
Gene ontology analysis was performed using the web tool David Bioinformatics Database
(http://david.abcc.ncifcrf.gov/home.jsp) (Huang et al., 2009a; Huang et al., 2009b). The
complete set of all RefSeq genes was used as a background.
Public availabilityof ChiP-Seq datasets
ChIP-Seq data have been submitted to the Gene Expression Omnibus Database
(http://www.ncbi.nlm.nih.gov/geo) with accession number GSE27844
(http://www.ncbi.nlm.nih.gov/geo/querv/acc.cgi?token=ifcvlkoimyymyzm&acc=GSE27844).
Microarray Analysis (Fig. 2)
Cell Culture and RNA isolation
For ESC expression analysis during differentiation following either DMSO or TCP treatment,
ZHBTc4 ESCs were split off MEFs, placed in a tissue culture dish for 45 minutes to selectively
remove the MEFs and plated in 6-well plates. The following day cells were treated with either
DMSO or ImM TCP for 6 hours in ESC media. After 6 hours a subset of cells from DMSO
treated wells was removed for TRIzol RNA isolation, the media was removed and replaced with
ESC media containing 2pg/ml doxycyline and either DMSO or ImM TCP. 24 hours after
doxycycline addition, a subset of cells was removed for RNA isolation and the media was
replaced with ESC media containing doxycycline and either DMSO or TCP. After another 48
hours, RNA was isolated using TRIzol (Invitrogen, 15596-026) from all time points (at indicated
213
times), further purified with RNeasy columns (Qiagen, 74104) and DNase treated on column
(Qiagen, 79254) following the manufacturer's protocols.
Micorarrayhybridizationand Analysis
For microarray analysis, Cy3 and Cy5 labeled cRNA samples were prepared using Agilent's
QuickAmp sample labeling kit starting with lug total RNA. Briefly, double-stranded cDNA was
generated using MMLV-RT enzyme and an oligo-dT based primer. In vitro transcription was
performed using T7 RNA polymerase and either Cy3-CTP or Cy5-CTP, directly incorporating
dye into the cRNA.
Agilent mouse 4x44k expression arrays were hybridized according to our laboratory's standard
method, which differs slightly from the standard protocol provided by Agilent. The hybridization
cocktail consisted of 825 ng cy-dye labeled cRNA for each sample, Agilent hybridization
blocking components, and fragmentation buffer. The hybridization cocktails were fragmented at
60 degrees C for 30 minutes, and then Agilent 2X hybridization buffer was added to the cocktail
prior to application to the array. The arrays were hybridized for 16 hours at 60 degrees C in an
Agilent rotor oven set to maximum speed. The arrays were treated with Wash Buffer #1 (6X
SSPE / 0.005% n-laurylsarcosine) on a shaking platform at room temperature for 2 minutes, and
then Wash Buffer #2 (0.06X SSPE) for 2 minutes at room temperature. The arrays were then
dipped briefly in acetonitrile before a final 30 second wash in Agilent Wash 3 Stabilization and
Drying Solution, using a stir plate and stir bar at room temperature.
214
Arrays were scanned using an Agilent DNA microarray scanner. Array images were quantified
and statistical significance of differential expression for each hybridization was calculated using
Agilent's Feature Extraction Image Analysis software with the default two-color gene expression
protocol. For each gene in the RefSeq gene list (see ChIP-Seq analysis section), the log10 ratio
values (ESC+Dox / ESC, or ESC+DOX+TCP / ESC) and p-value for that gene were determined
by averaging log 10 ratios and p-values of all Agilent Features annotated to that gene. Genes with
no annotated features were reported as NA (Supplementary Table 3). Genes with a log 10 ratio of
at least 0.0969 (1.25x) and a p-value less than or equal to 102 were determined to have a
significant change in expression.
Public availabilityof microarraygene expression datasets
Microarray gene expression data have been submitted to the Gene Expression Omnibus Database
(http://www.ncbi.nlm.nih.gov/geo/) with accession number GSE27844
(http://www.ncbi.nlm.nih.gov/geo/guery/acc.cgi?token=ifcvlkoimyymyzm&acc=GSE27844).
Protein Extraction and Western Blot Analysis (Fig. 2, Supplementary Fig. 1, 4, 5, 6, and 9)
ESCs were lysed with CelLytic Reagent (Sigma, C2978-50ml) containing protease inhibitors
(Roche). After SDS-PAGE, Western blots were revealed with antibodies against Oct4 (Santa
Cruz Biotechnology, sc-5279), LSDI (Abcam, ab17721), Mi-2b (Abcam, ab72418), CoREST
(Abcam, ab3263 1), p30 0 (Santa Cruz, sc-584), GAPDH (Abcam, ab9484) or Tubulin (Millipore,
05-661).
215
ChIP-Western and Co-Immunoprecipitation (Fig. 3)
For ChIP-Western, the same conditions as for ChIP-Seq were used. For co-immunoprecipitation,
murine ES cells were harvested in cold PBS and extracted for 30 min at 4 degrees C in
TNEN250 (50 mM Tris pH 7.5, 5 mM EDTA, 250 mM NaCl, 0.1% NP-40) with protease
inhibitors. After centrifugation, supernatant was mixed to 2 volumes of TNENG (50 mM Tris pH
7.5, 5 mM EDTA, 100 mM NaCl, 0.1% NP-40, 10% glycerol). Protein complexes were
immunoprecipitated overnight at 4 degrees C using 5 micrograms of either LSD 1 (Abcam,
ab17721), HDACI (Abcam, ab7028), HDAC2 (Abcam, ab7029), Mi-2b (Abcam, ab72418),
CoREST (Abcam, ab3263 1) or Rabbit IgG (Upstate, 12-370), bound to 50ul of Dynabeads@.
Immunoprecipitates were washed three times with TNEN 125 (50 mM Tris pH 7.5, 5 mM EDTA,
125 mM NaCl, 0.1% NP-40). For both ChIP-Western and co-immunoprecipitation, beads were
boiled for 10 minutes in XT buffer (Biorad) containing 100mM DTT to elute proteins. After
SDS-PAGE, Western blots were revealed with antibodies against either LSD 1 (Abcam,
ab1772 1), or HDAC1 (Abcam, ab7028).
216
Supplementary References
Adamo, A., Sese B., Castano, J., Paramonov, I., Barrero, M.J., and Izpisua Belmonte, J.C.
(2011). LSD 1 regulates the balance between self-renewal and differentiation in human
embryonic stem cells. Nat Cell Biol 13, 652-660.
Binda, C., Coda, A., Angelini, R., Federico, R., Ascenzi, P., and Mattevi, A. (1999). A 30angstrom-long U-shaped catalytic tunnel in the crystal structure of polyamine oxidase. Structure
7, 265-276.
Bowers, E.M., Yan, G., Mukherjee, C., Orry, A., Wang, L., Holbert, M.A., Crump, N.T.,
Hazzalin, C.A., Liszcaak, G., Yuan, H., el al. (2010). Virtual ligand screening of the p300/CBP
histone acetyltransferase: identification of a selective small molecule inhibitor. Chemistry &
biology 17,471-482.
Boyer, L.A., Lee, T.I., Cole, M.F., Johnstone, S.E., Levine, S.S., Zucker, J.P., Guenther, M.G.,
Kumar, R.M., Murray, H.L., Jenner, R.G., et al. (2005). Core transcriptional regulatory circuitry
in human embryonic stem cells. Cell 122, 947-956.
Boyer, L.A., Plath, K., Zeitlinger, J., Brambrink, T., Mederos, L.A., Lee, T.I., Levine, S.S.,
Wernig, M., Tajonar, A., Ray, M.K., et al. (2006). Polycomb complexes repress developmental
regulators in murine embryonic stem cells. Nature 441, 349-353.
Chen, X., Yuan, P., Fang, F., Huss, M., Vega V.B., Wong, E., Orlov Y.L., Zhang, W., Jiang, J.,
Loh, Y.H., et al. (2008). Integration of external signaling pathways with the core transcriptional
network in embryonic stem cells. Cell 133, 1106-1117.
Creyghton, M.P., Cheng, A.W., Welstead, G.G., Kooistra, T., Carey, B.W., Steine, E.J., Hanna,
J., Lodato, M.A., Frampton, G.M., Sharp, P.A., et al. (2010). Histone H3K27ac separates active
from poised enhancers and predicts developmental state. Proc Natl Acad Sci U S A 107, 2193121936.
Forneris, F., Binda, C., Vanoni, M.A., Mattevi, A., and Battaglioli, E. (2005). Histone
demethylation catalysed by LSD1 is a flavin-dependent oxidative process. FEBS Lett 579, 22032207.
Foster, C.T., Dovey, O.M., Lezina, L. Luo, H.L. Gant, T.W., Barlev, N., Bradley, A., and
Cowley, S.M. (2010). Lysine-specific demethylase 1 regulates the embryonic transcriptome and
CoREST stability. Mol Cell Biol 30, 4851-4863.
Guenther, M.G., Lawton, L.N., Rozovskaia, T., Frampton, G.M., Levine, S.S., Volkert, T.L.,
Croce, C.M., Nakamura, T., Canaani, E., and Young, R.A. (2008). Aberrant chromatin at genes
encoding stem cell regulators in human mixed-lineage leukemia. Genes Dev 22, 3403-3408.
Huang da, W., Sherman, B.T., and Lempicki, R.A. (2009a). Bioinformatics enrichment tools:
paths toward the comprehensive functional analysis of large gene lists. Nucleic acids research
37, 1-13.
217
Huang da, W., Sherman, B.T., and Lempicki, R.A. (2009b). Systematic and integrative analysis
of large gene lists using DAVID bioinformatics resources. Nature protocols 4, 44-57.
Johnson, D.S., Mortazavi, A., Myers, R.M., and Wold, B. (2007). Genome-wide mapping of in
vivo protein-DNA interactions. Science 316, 1497-1502.
Kagey, M.H., Newman, J.J., Bilodeau, S., Zhan, Y., Orlando, D.A., van Berkum, N.L., Ebmeier,
C.C., Goossens, J., Rahl, P.B., Levine, S.S., et al. (2010). Mediator and cohesin connect gene
expression and chromatin architecture. Nature 467, 430-435.
Kent, W.J., Sugnet, C.W., Furey, T.S., Roskin, K.M., Pringle, T.H., Zahler, A.M., and Haussler, D.
(2002). The human genome browser at UCSC. Genome Res 12, 996-1006.
Langmead, B., Trapnell, C., Pop, M., and Salzberg, S.L. (2009). Ultrafast and memory-efficient
alignment of short DNA sequences to the human genome. Genome biology 10, R25.
Lee, M.G., Wynder, C., Schmidt, D.M., McCafferty, D.G., and Shiekhattar, R. (2006). Histone H3 lysine
4 demethylation is a target of nonselective antidepressive medications. Chem Biol 13, 563-567.
Liang, Y., Vogel, J.L., Narayanan, A., Peng, H., and Kristie, T.M. (2009). Inhibition of the histone
demethylase LSD 1 blocks alpha-herpesvirus lytic replication and reactivation from latency. Nat Med 15,
1312-1317.
Lin, W., Srajer, G., Evrard, Y.A., Phan, H.M., Furuta, Y., and Dent, S.Y. (2007). Developmental potential
of Gcn5(-/-) embryonic stem cells in vivo and in vitro. Dev Dyn 236, 1547-1557.
Marson, A., Levine, S.S., Cole, M.F., Frampton, G.M., Brambrink, T., Johnstone, S., Guenther, M.G.,
Johnston, W.K., Wernig, M., Newman, J., el al. (2008). Connecting microRNA genes to the core
transcriptional regulatory circuitry of embryonic stem cells. Cell 134, 521-533.
Meissner, A., Mikkelsen, T.S., Gu, H., Wernig, M., Hanna, J., Sivachenko, A., Zhang, X., Bernstein,
B.E., Nusbaum, C., Jaffe, D.B., el al. (2008). Genome-scale DNA methylation maps of pluripotent and
differentiated cells. Nature 454, 766-770.
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. (2007). Genome-wide maps of chromatin state in pluripotent and
lineage-committed cells. Nature 448, 553-560.
Mimasu, S., Umezawa, N., Sato, S., Higuchi, T., Umehara, T., and Yokoyama, S. (2010). Structurally
designed trans-2-phenylcyclopropylamine derivatives potently inhibit histone demethylase LSD 1 /KDM 1.
Biochemistry 49, 6494-6503.
Miyabayashi, T., Teo, J.L., Yamamoto, M., McMillan, M., Nguyen, C., and Kahn, M. (2007). Wnt/betacatenin/CBP signaling maintains long-term murine embryonic stem cell pluripotency. Proc Natl Acad Sci
U S A 104, 5668-5673.
Schmidt, D.M. and McCafferty, D.G. (2007). trans-2-Phenylcyclopropylamine is a mechanism-based
inactivator of the histone demethylase LSD 1. Biochemistry 46, 4408-4416.
Seila, A.C., Calabrese, J.M., Levine, S.S., Yeo, G.W., Rahl, P.B., Flynn, R.A., Young, R.A., and Sharp,
P.A. (2008). Divergent transcription from active promoters. Science 322, 1849-1851.
218
Shi, Y., Lan, F., Matson, C., Mulligan, P., Whetstine, J.R., Cole, P.A., Casero, R.A., and Shi, Y. (2004).
Histone demethylation mediated by the nuclear amine oxidase homolog LSD1. Cell 119, 941-953.
Singh, S.K., Kagalwala, M.N., Parker-Thornburg, J., Adams, H., and Majumder, S. (2008). REST
maintains self-renewal and pluripotency of embryonic stem cells. Nature 453, 223-227.
Ura, H., Murakami, K., Akagi, T., Kinoshita, K., Yamaguchi, S., Masui, S., Niwa, H., Koide, H., and
Yokota, T. (2011). Eed/Sox2 regulatory loop controls ES cell self-renewal through histone methylation
and acetylation. EMBO J 30, 2190-2204.
Wang, J., Scully, K., Zhu, X., Cai, L., Zhang, J., Prefontaine, G.G., Krones, A., Ohgi, K.A., Zhu, P.,
Garcia-Bassets, I., et al. (2007). Opposing LSD 1 complexes function in developmental gene activation
and repression programmes. Nature 446, 882-887.
Wang, J., Hevi, S., Kurash, J.K., Lei, H., Gay, F., Bajko, J., Su, H., Sun, W., Chang, H., Xu, G., el al.
(2009a). The lysine demethylase LSD 1 (KDM 1) is required for maintenance of global DNA methylation.
Nat Genet 41, 125-129.
Wang, Z., Zang, C., Cui, K., Schones, D.E., Barski, A., Peng, W., and Zhao, K. (2009b). Genome-wide
mapping of HATs and HDACs reveals distinct functions in active and inactive genes. Cell 138, 1019-
1031.
Yang, M., Culhane, J.C., Szewczuk, L.M., Jalili, P., Ball, H.L., Machius, M., Cole, P.A., and Yu, H.
(2007). Structural basis for the inhibition of the LSDI histone demethylase by the antidepressant trans-2phenylcyclopropylamine. Biochemistry 46, 8058-8065.
219
Supplementary Figure Legends
Supplementary Fig. 1: Validation and characterization of LSD1 ChIP-Seq dataset
a, Validation of shRNA targeting LSD1. qPCR of LSDl transcripts in v6.5 ESCs infected with
short hairpins (Open Biosystems) against either LSD1 or GFP (control) for 5 days. The data is
normalized to Gapdh. The error bars represent the standard deviation of triplicate PCR reactions.
b, LSD 1 antibody is specific. Western blot of LSD 1 protein levels in GFP or LSD 1 knockdown
cells, showing a decrease in signal of one band corresponding to LSD 1 molecular weight (MW =
92.9 kDa). Tubulin served as loading control. c, LSD 1 ChIP-Seq signals are above experimental
noise. ChIP-Seq binding profiles (reads/million) for LSD 1 at the OcI4 (Pou5fl) and Lefty] loci
in ESCs, with the y-axis floor set to 1. Gene models, and previously described enhancer
regions 1-3 are depicted below the binding profiles. d, ChIP-Seq data, at high resolution, showing
coincidental binding of Oct4, Sox2, and Nanog with LSD 1. High-resolution ChIP-Seq binding
profiles (reads/million) at the Oct4 (Pou~fl) and Lefty] enhancer loci in ESCs, with y-axis floor
at 1.0 reads/million. e, Metagene showing mean LSD1 density is significantly higher at the 3,838
ESC enhancers compared to 3,838 neighboring core promoters (p<10-16).
Supplementary Fig. 2: LSD1 is required for doxcycline-induced differentiation of ZHBTc4
ESCs
a, Schematic representation of trophectoderm differentiation assay using doxycycline-inducible
Oct4 depletion murine ESC line ZHBTc4. Treatment with doxycyline for 48 hours leads to
depletion of Oct4 and early trophectoderm specification. ZHBTC4 cells were treated with
DMSO (control) or the LSD 1 inhibitors Tranylcypromine (TCP) and Pargyline (Prg) for 6 hours
before 2pg/ml doxycycline was added for an additional 48 hours. b, c, ESCs treated with either
220
(b) ImM tranylcypromine (TCP) or (c) 3mM pargyline (Prg) maintained ESC colony
morphology and alkaline phosphatase (AP) staining in doxycycline-treated cells despite
depletion of Oct4 protein levels. Scale bar = 20pM. d, Schematic representation of
trophectoderm differentiation assay using doxycycline-inducible Oct4 depletion murine ESC line
ZHBTc4. ESCs were infected with a GFP or LSD 1 shRNA for 72 hours before 2pg/ml
doxycycline was added for an additional 48 hours. e, ESCs infected with an shRNA targeting
LSD 1 maintained ESC colony morphology and alkaline phosphatase (AP) staining in
doxycycline-treated cells despite depletion of Oct4 protein levels. Scale bar = 20pM.
Supplementary Fig. 3: LSD1 is required for repression of ESC genes
a, Treatment of Prg partially maintained SSEA- 1 cell surface marker expression in doxycyclinetreated cells. Cells were stained for Hoechst (Hoe), Oct4 and SSEA- 1. Scale bar = 1 OOpM. b,
Treatment of Prg partially relieved repression of ESC genes 48 hours after Oct4 depletion in
DMSO- versus Prg-treated cells. Error bars reflect standard deviation of either duplicate or
triplicate PCR reactions. c, ESCs with decreased LSDI levels partially maintained SSEA-I cell
surface marker expression in doxycycline-treated cells. Cells were stained for Hoechst (Hoe),
Oct4 and S SEA- 1. Scale bar = 20gM. d, Reduced LSD 1 levels partially relieved repression of
ESC genes 48 hours after Oct4 depletion in GFversus LSD 1 shRNA. Error bars reflect standard
deviation of either duplicate or triplicate PCR reactions.
Supplementary Fig. 4: Characterization of LSD1 knockout ESCs
221
a, Beginning with an E14 ESC line (WT), cells were engineered to express an inducible Cre-ER
fusion protein from the endogenous ROSA26 locus 12. Sequential gene targeting produced cells
in which one LSD 1 allele has exon 3 flanked by LoxP sites (floxed) and the second has exon 3
deleted (Lox/!3). Induction of Cre activity by addition of 4-hydroxytamoxifen (4-OHT) to the
growth medium resulted in complete recombination of the remaining floxed LSD 1 allele,
generating LSD1 homozygous null A3/A3) cells. Loss of exon 3 disrupts the open reading frame
of LSD 1 such that a premature stop codon is introduced into exon 4, resulting in the progressive
loss of LSD 1 protein by 48 hours, as detected by Western blot (WB). Gapdh served as a loading
control. b, LSD1 heterozygous (Lox/A3) and homozygous null (A3/A3) ESCs maintain Oct4 and
SSEA-I cell surface marker expression compared to E14 wild type (WT) cells. Cells were
stained for Hoechst (Hoe), Oct4 and SSEA-1. Scale bar = 1 OpM. c, LSD 1 heterozygous
(Lox/A3) or homozygous null (A3/A3) ESCs maintain ESC morphology and alkaline
phosphatase (AP) staining compared to E14 wild type (WT) ESCs. Scale bar = IOOM. LSDI
Lox/A3 ESCs are used as a control throughout the remainder of the study. d, Schematic
representation of trophectoderm differentiation assay. LSD 1 Lox/A3 cells were treated for 48
hours with either ethanol (control) or 4-OHT to generate LSDI A3/A3 ESCs before being
infected with a GFP (shRNAGFP, control) or Oct4 (shRNAOct4) shRNA to knockdown Oct4
and induce differentiation. e, LSD 1 A3/A3 ESCs partially maintained SSEA- 1 cell surface
marker expression when Oct4 levels are decreased. Cells were stained for Hoechst (Hoe), Oct4
and SSEA-1. Scale bar =100gM. f, LSDI A3/A3 ESCs partially maintained ESC colony
morphology and alkaline phosphatase (AP) staining when Oct4 levels are decreased.
Supplementary Fig. 5: Catalytic activity of LSD1 is required for repression of ESC genes
222
a, Overexpression of LSDl wildtype and catalytic mutant in LSDl homozygous null (A3/A3)
ESCs. Homozygous null (A3/A3) cells were transfected with LSD I wildtype (EGFP-LSDIwt)
and a catalytic mutant (EGFP-LSDlmut) fused to EGFP. LSDl expression in A3/A3 ESCs are
relative to signal in LSD1 heterozygous(Lox/A3) cells. The data is normalized to Gapdh. The
error bars represent the standard deviation of triplicate PCR reactions. b, Western blot (WB)
showing homozygous null (A3/A3) ESCs transfected with LSD 1 transgenes express high levels
of LSD1 protein. c, Homozygous null (A3/A3) and homozygous null (A3/A3) ESCs expressing
the LSD1 transgenes maintain Oct4 and SSEA-1 cell surface marker expression compared to
LSD 1 heterozygous (Lox/ A3) ESCs. Cells were stained for Hoechst (Hoe), Oct4 and SSEA- 1.
Scale bar = 100"M. d, Homozygous null (A3/A3) ESCs expressing LSD1 transgenes maintain
ESC morphology and alkaline phosphatase (AP) staining compared to LSD 1 heterozygous (Lox
/A3) ESCs. Scale bar = 100pM. e, Schematic representation of trophectoderm differentiation
assay. LSD 1 Lox/A3 cells were treated for 48 hours with either ethanol (control) or 4-OHT to
generate LSD 1 A3/A3 ESCs before transfected with either an LSD 1 wildtype or catalytic mutant
(K661A) transgene fused to EGFP (EGFP-LSDlwt or EGFP-LSDlmut). Cells were transduced
with an shRNA targeting either Luciferase (control) or Oct4 (shRNAOct4) to knockdown Oct4
and induce differentiation. f, Validation of shRNA targeting Oct4. qPCR of Oct4 transcripts in
LSD 1 heterozygous (Lox/A3) ESCs infected with a short hairpin (Open Biosystems) against
either Oct4 or Luciferase (control) for 72 hours. The data is normalized to Gapdh. The error bars
represent the standard deviation of triplicate PCR reactions. g, Western blot of Oct4 protein
levels in Luciferase (control) or Oct4 knockdown cells, showing a decrease in signal of one band
corresponding to Oct4 molecular weight (MW = 38 kDa). Tubulin served as loading control. h,
Expression of ESC genes was downregulated during differentiation when wildtype, but not the
223
catalytic mutant, LSDl was introducted into homozygous null (A3/A3) cells. Expression of ESC
genes was maintained in differentiating LSDl null cells compared to control cells.
Overexpression of wildtype LSD 1, but not the catalytic mutant, rescued this phenotype. Error
bars reflect standard deviation of triplicate PCR reactions.
Supplementary Fig. 6: Validation of CoREST antibody
a, Validation of shRNA targeting CoREST. qPCR of CoREST transcripts in v6.5 ESCs infected
with short hairpins (Open Biosystems) against either CoREST or GFP (control) for 5 days. The
data is normalized to Gapdh. The error bars represent the standard deviation of triplicate PCR
reactions. b, CoREST antibody is specific. Western blot of CoREST protein levels in GFP or
CoREST knockdown cells showing a decrease in signal of one band corresponding to CoREST
molecular weight (MW = 70 kDa). Gapdh served as loading control.
Supplementary Fig. 7: LSD1 occupies a subset of genomic sites with REST
a, LSD 1 occupies a subset of genomic sites with REST. ChIP-Seq binding profiles
(reads/million) for transcription factors (Oct4, Sox2, Nanog, REST), chromatin regulator
(LSD1), the transcriptional apparatus (Pol II, TBP) and histone modifications (H3K4me1,
H3K4me3, H3K79me2, H3K36me3) at the NeuroD4 and VGF loci in ESCs. Gene models are
depicted below the binding profiles. b, GO analysis on the top 500 (based on peak height) REST
bound genes co-occupied by LSD 1 reveals significant enrichment for genes involved in
neurogenesis.
224
Supplementary Fig. 8: LSD1 occupies ESC enhancers with NuRD
LSD 1 occupies ESC enhancer sites with Mi-2b and either HDAC 1 or HDAC2. Density map of
ChIP-Seq data at Oct4, Sox2, Nanog, and MedI co-occupied enhancer regions. Data is shown for
chromatin regulators (LSD1, Mi-2b, and either HDAC I or HDAC2) in ESCs. Over 70% of the
3,838 high confidence enhancers were co-occupied by LSD1 and NuRD (p < 10-9). Color scale
indicates ChIP-seq signal in reads per million.
Supplementary Fig. 9: NuRD is required for doxcycline-induced differentiation of ZHBTc4
ESCs
a, Validation of shRNAs targeting Mi-2b. qPCR of Mi-2b transcripts in v6.5 ESCs infected with
short hairpins (Open Biosystems) against either Mi-2b or GFP (control) for 5 days. The data is
normalized to Gapdh. The error bars represent the standard deviation of triplicate PCR reactions.
b, Mi-2b antibody is specific. Western blot of Mi-2b protein levels in GFP or Mi-2b knockdown
cells, showing a decrease in signal of two bands corresponding to Mi-2b molecular weight (MW
= 280 and 218 kDa). Gapdh served as loading control. c, ESCs infected with shRNAs targeting
Mi-2b partially maintained ES cell colony morphology and alkaline phosphatase (AP) staining in
doxycycline-treated cells despite depletion of Oct4 protein levels. Scale bar = 20pM. d, ESCs
with decreased Mi-2b levels partially maintained SSEA- 1 cell surface marker expression in
doxycyclinetreated cells. ESCs were infected with Mi-2b shRNA as represented in
Supplementary Fig. 2d and were stained for Hoechst (Hoe), Oct4 and SSEA- 1. Scale bar =
20pM. e, Mi-2b is required for downregulation of ESC genes. Expression of ESC genes was
225
maintained 48 hours after Oct4 depletion in GFP versus Mi-2b shRNA treated cells. Error bars
reflect standard deviation of either duplicate or triplicate PCR reactions.
Supplementary Fig. 10: Reduced p300 and H3K27Ac levels at ESC enhancers during
differentiation
a, b, Reduced levels of transcriptional coactivator p300 and H3K27Ac at enhancers during ESC
differentiation. ChIP-PCR analysis of p300 and H3K27Ac enrichment at enhancer loci in
doxycycline-treated (+dox) and untreated (-dox) ZHBTc4 cells. Error bars reflect standard
deviation of either duplicate or triplicate PCR reactions.
Supplementary Fig. 11: LSD1 occupies ESC enhancers during differentiation
a, b, LSD I is approximately 0-5% reduced at 12 hours and 30-50% at 48 hours following Oct4
depletion and differentiation of ESCs. ChIP-PCR analysis of LSD 1 enrichment at enhancer loci
in doxycycline-treated (+dox) and untreated (-dox) ZHBTc4 cells. Error bars reflect standard
deviation from biological replicates.
Supplementary Figure 12: p3 0 0 inhibition results in reduced H3K27ac and H3K4mel
levels at ESC enhancers
a, Reduced levels of H3K27ac and H3K4mel at enhancers of ESCs treated with p300 inhibitor
C-646. ChIP-Seq binding profiles (reads/million) for chromatin regulators (p300, LSD 1) at the
226
Lefty] locus in ESCs. Below these profiles, histone H3K27ac and H3K4me 1 levels are shown for
DMSO treated (control) ESCs, and cells treated with p300 inhibitor C-646 for 24 hours. For
appropriate normalization, ChIP-Seq data for histone H3K27ac and H3K4mel is shown as rank
normalized reads/million with the y-axis floor set to 1 (Supplementary Information). Gene
models, and previously described enhancer regions3 are depicted below the binding profiles. b,
Mean of the normalized H3K27ac and H3K4me l density +/- 250 nucleotides surrounding LSDIoccupied enhancer regions in the presence or absence of C-646. The associated genes were
identified based on their proximity to the LSD1 -occupied enhancers (Supplementary
Information). c, Mean of the normalized H3K27ac and H3K4me 1 density +/- 250 nucleotides
surrounding 1,234 LSD 1-occupied enhancers showing reduced levels of H3K27ac in the
presence of C-646. Of the 1,234 LSD 1-occupied enhancers having reduced levels of H3K27ac
upon C-646 treatment, 63% (773) display reduced H3K4mel levels after C-646 treatment (p <
10-3).
Supplementary Fig. 13: LSD1 is required for H3K4mel removal at ESC enhancers
Heatmap of normalized H3K4meI density +/- 250 nucleotides surrounding the 2,755 LSD1
occupied enhancers having reduced levels of H3K4mel upon differentiation. Of the 2,755 LSDI
occupied enhancers having reduced levels of H3K4me 1 upon differentiation, 63% (1,722)
display higher H3K4me 1 levels after TCP treatment (p < 10-16). Color scale indicates ChIP-Seq
signal in normalized reads per million.
227
b
a
MW (kDa)
LSD1 Fold Change
0 .25 .50 .75 1.0
R8100-
shRNAGFP
WB: LSD1
7555-
shRNALsDl
50-
C
WB: Tubulin
15kb
15kb
Chromatin r10 F
Regulator L
LSD1
WCE
I
Enhancers
I
Core Promoter
sr.OCt4
800bp
70
efty1
800bp
70
Oct4
70O
70
Sox2
70[
70F
Nanog
,or
LSD1
jd&im.
Transcription
Factors
Chromatin
Regulator
EMOM.-
10
I
nIAd
Oct4 enhancer
Leftyl enhancer
e
2.0
ID
1.5
- Enhancers
- Core promoters
1.0
C,
0.5
0
5
Distance from LSD1 peak (kb)
-5
Figure S1
228
a
ZHBTc4 ESCs
Early Trophectoderm
Oct4ON
-6 hrs
0 hrs
48 hrs
Prg
Doxycycine
b
C
0 hrs
8
U)
Uj
X
o
C/) a
LLJ +
U) a.
0 .
mg
W
L
U x
d
ZHBTc4 ES cells
Oct4ON
ZHBTc4 ES cells
Oct40*
Early Trophectoderm
0 hrs
-72 hrs
48 hrs
:shRNA
Doxycycline
e
I
Oh
C,
I.A.
x
0o
tikft
Ar
COf
+~
Figure S2
229
a
K]
LI]j
W .2
ESC Genes
0.
Sox2
-x
LL
Esrrb
-0.8
-2
-1.
-2.4
-4
-61
Trim28
Leftyl
Lefty2
-0.8
-2-.
-2.41 -121
* ESCs +Dox
ESCs +Dox,Prg
C
d
ESC Genes
(1)cc .
-oO 0
U- U
Tcl1
Esrrb
0 Sox2
-5
-6
C0.8
081
-1.6
-2.4
-12
-18
-10
-15
Lefty1 Left 2
-5
-10
-1
ESCs +Dox; shRNAGFP
ESCs +Dox: shRNALSD1
Figure S3
230
b
a
WT
4-OHT:
(hrs)
0
Lox/A3
48
0
WB: LSD1
C/)
WB: Gapdh
7
C
w
d
ESCs'oxIA
3
ESCsM
Oct4N
0 hrs
Oct4ON
-48 hrs
3
Early TrophectoderrT
72 hrs
-4-OHT
shRNA4
:sh
RNA
e
f
Figure S4
231
4
b
a
Lox/A3
15
10
C0
-x
ow
ESCSA*
EGFP-LSD1
ESCSANas - EGFP-LSD1
ESCsAA
0
U-
3
EGFP-LSD1:
-
A3/A3
wt mut.
-
-
*
'
~WB: LSD1
mid
MOM
LSD1
C
d
e
f
-
WB: Gapdh
Oct4 Fold Change
0 .25 .50 .75 1.0
shRNALoeras
4
shRNA~O
ESCsLOxA3
ESCsA'^3
Oct4ON
Oct4ON
-56 hrs
-8 hrs 0 hrs
:4-OHT
Early Trophectoderm
_
:EGFP:LSD1 wt
0
+
MW (kDa)
9
72 hrs
mu
50-
:shRNA~ct 4
25 37 -
h
H
WB: Oct4
WB: Tubulin
ESC Genes
24
16
.
.24
9
S8
0 0
T
-8
75
LL
-16
.12
3
-.1]
-3
-12
-.24
-6
-.24
-,3E
-9
0
0
.36
.24
.12
6
3
15
10
0 ESCsLXIA
0 ESCsa&'s
* ESCsVA,3 - EGFP-LSD1 w
5
0 ESCSAT3/
-241
Leftyl
Sox2
-5
-10
-.36
Esrrb
-19
Dppa3
Figure S5
232
Apobeci
- EGFP-LSD1 mid
a
CoREST Fold Change
0 .25 .50 .75 1.0
shRNAGFP
shRNACoREST
b
MW (kDa)
7 50 25 -
WB:CoREST
37 -
WB: Gapdh
Figure S6
233
15kb
15kb
Transcription
Factors
70F
Oct4
70
Sox2
70
Nanog
F
15
r
'
Regulator L
Chromatin
REST
10
LSD1
1
[30
F
3
15
Transcription
Apparatus
~Pol111
TBP
10
H3K4me1
80
H3K4me3
80
H3K79me2
Histone
Modifications
Core Promoter
[
I
Neurod4
-
*
b
Vgf
-Log (P-val)
0
GO Categories
Synaptic transmission
Regulation of neurotransmission
Neurotransmitter secretion
Neuron development
Generation of neurons
Neurogenesis
3
6
9
Figure S7
234
12
LSD1 Mi-2 HDAC1/2
0
)
ccl
C
cvi
.J
-5 0 5
Distance from Oct4, Sox2, Nanog
Mediator-occupied site (kb)
Figure S8
235
a
C
Mi-2p Fold Change
0 .25 .50 .75 1.0
shRNAGFP
2
shRNAmI- 0 #1
shRNAmI-20 #2
WCCz
0
x
b
MW (kDa)
0 O
WB: Mi-23
75 -
CL)
55-
37-~
WB: Gapdh
d
I
e
1.5
3
U
-9
2
1
6
1u
0
.5
3
ESC Genes
12
8
3
2
3
4
1
1
.5
-3
-4
-1
-1
-1
-3
-1.51
-6
-91
-8
-12
Lefty2
Leftyl
.2
-3
-2
-3
Nanog
KIf4
236
ESCs-shRNAI-2P
ESCs-shRNAMI-2P
2
-1
-2
Sox2
SESCsshGFP + Dox
Essrb
*1+ Dox
*2
+ Dox
Figure S9
237
a
.
8
15
12
35
28
20
16
20
16
50
40
12
12
30
8
20
10
9
21
Cw
6
14
o
3
7
U. L
OI
8
4L
01
-- O
Sox2
Trim28
Esrrb
* ESCs
ESCs +Dox (48hrs)
01 .
Lefty2
Leftyl
b
4
E
LL
15
40
10
3.C
15
12
9
8
6
4
3
0
8
0
2
O
2.4
1.8
1.2
0.6
12
9
6
32
24
16
Esrrb
Nanog
M
6i
3
I
I
Leftyl
Sox2
Figure S10
238
* ESCs
ESCs +Dox (48hrs)
Lefty2
a
30
24
20
16
5
E 4
4
ESCs
3
2
18
12
12
8
3
2
ESCs +Dox (1 2hrs)
1
6
4
1
5
0~
C)
0
U-
b
10
8
6
4
2
5
4
3
2
1
40
32
24
16
8
10
E 8
6
4
020 2
U-
Sox2
Leftyl
Trim28
Sal4
0..
ESCs
ESCs +Dox (48hrs)
00
Sa114
Trim28
Sox2
Lefty1
Figure S11
239
a
20kb
50
Chromatin
Regulators
b
Sox2
p300
10
LSD1
ES[s
b 200
120
150
100
90
50
30
C646:
-
+
E
17
11
- &
1LI
H3K4me1
ESCs r10F
= fty1
aC
U
I
E
170I
16n
+
Ebf2
48
36
400
300
40
20
24
12
20C
10C
-+
-+
- +
175
100
100
175
14C
8C
80
140
105
60
60
105
70
4C
40
20
-
+
-
+
Nanog
80
60
C646:
240
+
240
5
Figure S12
-
Zfp831
-+
-
70
500
60
+
105
60
0646:
1234 Enhancers
200
200
190
1901
180
E 180
60
C -0
0646: - +
60
-
175
14C
100
S 120
170
150
12
+
300
180
I
40
20
Of
-
150
12C
Esrrb
Core Promoter
+
-
Enhancers
I
-
C646
+C-646LAs.
C
0
Smad2
100
80
60
33
22
11
of
41
j
HDAC2
55
44
60
280
226
ESCs
(control)M
+C-646
(coE t ol
150
01
H3K27ac
10
CHD6
250
+
70
35
-
+
-
+
+ +
io
2,755 LSD1 occupied enhancer regions
CD
3