Small Molecule Control of Chromatin Remodeling
Transcript of Small Molecule Control of Chromatin Remodeling
Chemistry & Biology
Review
Small Molecule Control of Chromatin Remodeling
Aidan Finley1 and Robert A. Copeland1,*1Epizyme, Inc., 400 Technology Square, 4th Floor, Cambridge, MA 02139, USA*Correspondence: [email protected]://dx.doi.org/10.1016/j.chembiol.2014.07.024
Control of cellular transcriptional programs is based on reversible changes in chromatin conformation thataffect access of the transcriptional machinery to specific gene promoters. Chromatin conformation is inturn controlled by the concerted effects of reversible, covalent modification of the DNA and histone com-ponents of chromatin, along with topographical changes in DNA-histone interactions; all of these chro-matin-modifying reactions are catalyzed by specific enzymes and are communicated to the transcriptionalmachinery by proteins that recognize and bind to unique, covalent modifications at specific chromatin sites(so-called reader proteins). Over the past decade, considerable progress has been made in the discovery ofpotent and selective small molecule modulators of specific chromatin-modifying proteins. Here we reviewthe progress that has been made toward small molecule control of these mechanisms and the potential clin-ical applications of such small molecule modulators of chromatin remodeling.
All somatic cells within multicellular organisms contain the full
complement of DNA that constitutes the genome of that
organism. Nevertheless, beyond early embryonic development,
multicellular organisms are distinguished from single cell and
colony-forming organisms by the presence of a broad spectrum
of fit-for-purpose cells that have differentiated themselves from
their common progenitors. Cellular differentiation depends on
precise control of gene transcription programs, such that certain
genes are transcriptionally active whereas others are repressed,
depending on the cell type, cell cycle stage, and timing relative to
cellular division.
The regulatory system used to control transcriptional pro-
grams is commonly referred to as epigenetics (Allis et al.,
2007; Holliday, 1990), but is more correctly referred to as chro-
matin remodeling (Ptashne, 2013). At the molecular level, chro-
matin remodeling involves controlling the local structure of
chromatin at specific gene promoters by the concerted action
of a spectrum of chromatin-modifying proteins (CMPs) (Arrow-
smith et al., 2012; Copeland et al., 2009; Wilkins et al., 2014).
Chromatin is the amalgam of DNA and histone proteins that
make up chromosomes. Chromosomal DNA exists as stretches
of unassociated, double-stranded DNA, interspersed with nu-
cleosomes—loci of �140–150 base-pair stretches of DNA
coiled twice around a spool-like octamer core of histones, con-
sisting of two copies each of four core histone proteins (H2A,
H2B, H3, and H4). The coiling of DNA around the histone
core provides a mechanism for compaction of the ca. 2 m of
DNA that needs to be fit into the small volume of a cell nucleus.
Fully compacted chromatin is referred to a heterochromatin.
This conformational state, however, severely limits steric ac-
cess of gene promoter regions to transcription factors, poly-
merases and the rest of the transcriptional machinery. A more
relaxed (i.e., less compacted) state of chromatin, referred to
as euchromatin, also exists; in this latter state, the gene pro-
moter region is readily accessible to the transcriptional machin-
ery (Figure 1). Methylation of CpG islands of chromosomal DNA
directly affects gene transcription, whereas the conformational
switch between euchromatin and heterochromatin states in
1196 Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Lt
proximity to particular gene locations is effected by modifica-
tions to the histone proteins; these modifications involve a
collection of highly specific enzymes and associated recogni-
tion proteins, as detailed next.
CMPsThree major categories of chromatin modifications work in con-
cert to effect transcriptional regulation. These are: (1) methyl-
ation of chromosomal DNA, (2) posttranslational modifications
of specific amino acids on histones, and (3) ATP hydrolysis-
dependent alteration of DNA-histone interactions.
Methylation of chromosomal DNA is catalyzed by the DNA
methyltransferases (DNMTs) and methylation is reversed by
the ten-eleven translocation (TET) proteins (Figure 2). DNA
methylation results in transcriptional silencing of genes, in-
cluding a number of tumor suppressors (Baylin, 2005). The
DNMTs have been targeted for small molecule inhibitors as
cancer therapeutics; indeed, two DNMT-targeted drugs have
been approved as treatments for myelodysplastic syndrome,
decitibine (Dacogen) and azacitidine (Vidaza). DNA methylation
and the drugs targeting this modification have been extensively
reviewed in the literature (see, for example, Baylin, 2005 and
Dhe-Paganon et al., 2011). Among the covalent modifications
of histone proteins, the protein kinases that phosphorylate
histones and their inhibitors have also been extensively re-
viewed in the recent literature (see, for example, Rossetto
et al., 2012 and Suganuma and Workman, 2012). Hence, we
shall focus our attention for the remainder of the current
review on the various CMPs involved in histone acetylation/
deacetylation and methylation/demethylation reactions and on
the CMPs involved in ATP hydrolysis-dependent chromatin
remodeling.
Posttranslational Modification of Histone Proteins
Site-specific covalent modification of histones can affect the
conformation of chromatin by altering interactions between the
histones and the nucleosomal DNA (Arrowsmith et al., 2012;
Copeland et al., 2009). These modifications include covalent
attachment of small proteins, such as ubiquitin and SUMO, to
d All rights reserved
Figure 1. Cartoon of the ConformationalTransition of Chromatin from the MoreOpen, Uncondensed Euchromatin State tothe More Condensed HeterochromatinStatePurple and red spheres represent individual his-tone proteins comprising the nucleosome corearound which the DNA (turquoise) is wrapped.Small circles are meant to represent covalentmodifications to the DNA and histones.
Chemistry & Biology
Review
specific lysines, as well as covalent attachment of chemical
groups to specific amino acids; these latter reactions include
lysine/arginine methylation, lysine acetylation, and serine/
threonine/tyrosine phosphorylation. All of these modifications
affect chromatin structure, hence gene transcription. The en-
zymes involved in placement of acetyl and methyl groups on
histones (referred to as writers) and the enzymes involved in
the specific removal of these chemical groups (referred to as
erasers) have been the most studied and are enzymes for which
small molecule modulators have been reported. As stated
above, we shall focus the remainder of our discussion on these
enzymes and the reactions catalyzed by them, together with
the proteins that recognize and bind to specific histonemodifica-
tions (referred to as readers) as initiating events in transcriptional
control.
Histone Acetylation. Histone lysine residues are acetylated by
the histone acetyltransferases (HATs). HATs acetylate lysine
residues using acetyl-CoA as a common acetyl group donor.
This results in charge neutralization of the lysine side chain,
altering the electrostatic interactions that can be made with
the negatively charged DNA (Figure 3A). Histone acetylation
generally results in exposure of gene promoter sites to tran-
scriptional machinery and activation of gene transcription
(Rice and Allis, 2001). Acetylation also creates a locus for
bromodomains to recognize ‘‘transcription ready’’ chromatin,
and thus recruitment of transcriptional machinery (Dey et al.,
2003).
There are 18 human HATs, divided into distinct families based
on sequence homology and structure. Type A (nuclear) HATs are
divided into three subfamilies, GNAT, p300/CB, and MYST
(Hodawadekar and Marmorstein, 2007; Andreoli et al., 2013).
Type B HATs are found in the cytoplasm and acetylate histones
prior to incorporation into chromatin (Andreoli et al., 2013).MYST
and p300/CPB enzymes also acetylate nonhistone proteins
(Friedmann and Marmorstein, 2013). HATs are large proteins
with multiple domains, and the complexity of structure is impor-
tant for substrate targeting and catalytic function (Lee and
Workman, 2007). Structures of several HAT catalytic domains
have been solved (Andreoli et al., 2013) and feature a conserved
core binding to acetyl-CoA despite low sequence conservation
and differing surrounding structures (Liu et al., 2008; Hodawade-
kar and Marmorstein, 2007; Marmorstein, 2001). This structural
understanding is bolstered by detailed kinetic and biochemical
studies of the HAT reaction mechanism. GCN5 family members
require a ternary complex (enzyme * acetyl-CoA * H3 histone)
before catalysis can occur, and this is conserved for the human
and yeast enzymes. The catalytic mechanism is, however, likely
specific to each HAT family/subfamily, as kinetic mechanisms
are different among the P300/CBP, MYST, and GNAT subfam-
ilies and tool inhibitory peptides specific for PCAF (another
Chemistry & Biology 2
type A HAT that associates with p300/CBP) were unable to
inhibit P300/CBP (Roth et al., 2001; Hodawadekar and Marmor-
stein, 2007).
Deacetylation of lysine is catalyzed by the histone deacety-
lases (HDACs). There are 18 human HDACs and these are sub-
divided into four subclasses (Gregoretti et al., 2004). Class I
contains the enzymes HDAC 1, 2, 3, and 8. Class IIa consists
of HDACs 4, 5, 7, and 9. Class IIb contains HDACs 6 and 10.
Class III is made up of the Sirtuins 1–7, whereas class IV contains
a single enzyme, HDAC 11. With the exception of the Sirtuins, all
of these enzymes use an active site zinc atom to active a coordi-
nated water molecule that, together with active site histidine and
aspartate side chains facilitate nucleophilic catalysis on the
acetyl-lysine substrate (Lombardi et al., 2011; Figure 3B). The
Sirtuins, on the other hand, use a distinct mechanism of catalysis
involving NAD+ attack of the bound acetyl-lysine substrate
(Walsh, 2006; Milne and Denu, 2008). For both the metal-utilizing
HDACs and the Sirtuins, substrate specificity is moderate and is
conferred by recognition elements outside the catalytic active
site of these enzymes.
Bromodomains (BRDs) are protein domains of approximately
110 amino acids in length with a predominantly a-helical struc-
ture (Haynes et al., 1992) configured as a left-handed up-and-
down helical bundle. Bromodomain-containing proteins are
found across several phyla from fruit flies to man, and in humans
there are 61 different BRD structures (Filippakopoulos et al.,
2012). Although the bromodomain family, including the bromo-
domain and extra-terminal domain (BET) proteins is large and
relatively diverse, with low overall sequence homology, all bro-
modomain proteins are defined by a well-ordered, deep, hydro-
phobic pocket into which the acetyl-lysine side chain binds
(Filippakopoulos and Knapp, 2014). Surface residues in the var-
iable loop regions contribute to substrate recognition, allowing
for considerable diversity in binding partners. In contrast to the
surface residues, the binding site for the acetylated lysine is
well conserved, with critical hydrophobic and aromatic residues
(Owen et al., 2000). These residues interact with the aliphatic
side chain of the lysine while a conserved asparagine residue
forms a key hydrogen bond with the acetyl carbonyl. Water mol-
ecules in the binding pocket also bridge interactions between the
acetylated lysine and the protein (Owen et al., 2000; Mujtaba
et al., 2007). These common features notwithstanding, BRDs
display considerable diversity of amino acid composition within
the binding pocket and this has provided a basis for the develop-
ment of domain-selective BRD inhibitors, as discussed below.
Histone Methylation. Lysine and arginine residues within his-
tones are methylated by the protein methyltransferases (PMTs;
Copeland et al., 2009). This enzyme class bifurcates into families
(Richon et al., 2011): the protein lysine methyltransferases
(PKMTs) and the protein arginine methyltransferases (PRMTs).
1, September 18, 2014 ª2014 Elsevier Ltd All rights reserved 1197
Figure 2. Reversible Methylation of Cytosine within CpG Islandsof DNAMethylation is catalyzed by the DNMT class of enzymes and reversed, in amultistep process, by the TET class of enzymes. See text for further details.SAM, S-adenosyl methionine; SAH, S-adenosyl homocysteine; a-KG, alpha-ketoglutarate.
Chemistry & Biology
Review
All PMTs use SAM as a universal methyl donor and perform
protein methylation via a SN2 reaction mechanism, involving
formation of a ternary enzyme-SAM-protein complex prior to
direct methyl transfer from SAM to the nitrogen atom of the
lysine or arginine side chain (Figures 4A and 4B). Despite
the common chemical reaction catalyzed by these enzymes,
the active site architecture varies considerably among them,
thus providing a structural basis for substrate specificity. As
discussed below, the structural diversity of PMT active sites
also provides a basis for potent and selective modulation of
these enzymes by small molecule ligands (Basavapathruni
et al., 2012).
Substrate permissiveness varies among the PMT enzymes,
as does the degree of enzyme redundancy for methylation of
a particular histone site (Copeland, 2013; Copeland et al.,
2009; Richon et al., 2011). For example, the enzyme DOT1L
is the only human enzyme known to methylate histone H3
at lysine 79 (H3K79) and this position is the only site
methylated by DOT1L. Likewise, the multi-protein complex
polycomb repressive complex 2 (PRC2), containing the cata-
lytic subunits EZH1 or EZH2, represent the only H3K27 methyl-
transferases in humans; PRC2 is also known to methylate only
H3K27. On the other hand, some histone sites are methylated
by several PMTs (e.g., H3K36). While the PKMTs tend to
demonstrate a high degree of substrate specificity, the PRMTs
tend to display greater permissiveness of substrate utilization,
sometimes methylating multiple arginine residues on multiple
proteins.
The transcriptional consequences of histone methylation
depend on the specific site being methylated and also on the
number of methyl groups placed on the amino acid. Lysine can
accept 1, 2 or 3 methyl groups and arginine can be monomethy-
lated on one nitrogen, symmetrically dimethylated, or asymmet-
rically dimethylated (Figure 4B); each of these states of
methylation can confer different biological consequences.
Methylation of histone sites can be transcriptionally activating
or repressive, depending on the site of the methylation. For
1198 Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Lt
example, trimethylation at H3K27 suppresses gene transcrip-
tion, whereas trimethylation of H3K79 is transcriptionally acti-
vating (Copeland, 2013; Copeland et al., 2009).
Methyl groups are oxidatively released from lysine by the
lysine demethylases (KDMs; Arrowsmith et al., 2012; Forneris
et al., 2005; Tsukada et al., 2006). The KDM class clusters into
two families: the lysine-specific demethylase (LSD) family and
the Jumonji domain-containing family (JmjC; Arrowsmith et al.,
2012). Two flavin-dependent enzymes comprise the LSD family,
LSD1 and LSD2 (Figure 4C). The JmjCs constitute a larger family
of 27 enzymes that all use iron and 2-oxoglutarate to oxygenate
the methyl group (Figure 4D). KDMs act primarily on histones,
but can also demethylate nonhistone proteins. At least one
JmJC family KDM, JMJD6, has been reported to directly de-
methylate methyl-arginine residues (Chang et al., 2007) and the
protein arginine deiminases (PADs) can also lead to demethyla-
tion of arginine via conversion of arginine to citrulline (Cuthbert
et al., 2004; Wang et al., 2004).
Methyl-lysine is an abundant histone modification, and a
diversity of methyl-lysine reader domains exist (Wigle and Cope-
land, 2013). Over 200 methyl-lysine reader domains have been
described to date (James and Frye, 2013). The methyl-lysine
readers (KMe readers) are divided into two families: PHD zinc
finger domains, and the ‘‘Royal Family,’’ comprising Tudor,
Agenet, MBT, CHROMO domain, WD40 repeats (WDR5 and
EED) and PWWP domains (Wigle and Copeland, 2013; James
and Frye, 2013). Recently, another protein domain, the bromo
adjacent homology (BAH) domain of the ORC1 protein has
been shown to also be a KMe reader specific to H4K20 dimethy-
lation (Kuo et al., 2012). The KMe reader domains are typically
comprised of less than 100 residues and fold to form pockets
that contain an electron-rich cage of two to four aromatic
residues (the aromatic cage) that interact with the lysine
through p-cation, hydrogen bonding, and van der Waals inter-
actions. The methylation state of the lysine is recognized by
the presence of from zero to two acidic amino acids, at the
base of the pocket, that interact with the nitrogen of the lysine
side-chain.
ATP Hydrolysis-Dependent Changes to DNA-Histone
Interactions
ATP hydrolysis-dependent chromatin remodeling is exemplified
by the SWI/SNF complexes (Shain and Pollack, 2013; Wang
et al., 2014; Wilson and Roberts, 2011). These complexes are
composed of 9–12 protein subunits in humans; the exact
composition of subunits defines the specific complex and its
gene targets. Subunits of the SWI/SNF complex consist of: (1)
one of two mutually exclusive ATP-hydrolyzing subunits known
as BRM (also known as SMARCA2) or BRG1 (also known as
SMARCA4), (2) a set of highly conserved core subunits, such
as SNF5 (also known as SMARCB5), INI1, and BAF47, and (3)
a variable set of subunits involved in targeting, assembly and
cell lineage-specific functions.
SWI/SNF complexes remodel chromatin by catalyzing sliding
and ejection or insertion of histone octamer cores (Wilson and
Roberts, 2011). Sliding refers to the following sequence of steps.
First, the SWI/SNF complex binds at a specific position on nucle-
osomal DNA. Disruption of key contacts between the DNA and
histone proteins then ensues. This is followed by translocation
of the DNA, using the energy derived from ATP hydrolysis. A
d All rights reserved
Figure 3. The Chemistry of Histone Acetylation(A) Reaction mechanism for the histone acetyltransferases (HATs).(B) Reaction mechanism for the metal-dependent histone deacetylases (HDACs).(C) Reaction mechanism for the Sirtuin histone deacetylases.
Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Ltd All rights reserved 1199
Chemistry & Biology
Review
A
B
C
D
Figure 4. The Chemistry of Histone Methylation(A) The common methyl group donor SAM is converted to SAH by all proteinmethyltransferases (PMTs).(B) Stepwise methylation of the ε-nitrogen of lysine side chains by the proteinlysine methyltransferases (PKMTs) and of the side chain guanidino group ofarginine by the protein arginine methyltransferases (PRMTs).(C) Reaction mechanism for the flavin-dependent lysine demethylases (LSDs).(D) Reaction mechanism for the iron-dependent lysine demethylases (JmjC).
Chemistry & Biology
Review
DNA loop is thus formed and this can be propagated around
the nucleosome to create sites of increased access to DNA
binding factors. Thus, transcription factors and the like can
1200 Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Lt
bind to specific gene promoter locations because of SWI/SNF
catalysis. The SWI/SNF complex is also known to catalyze the
ejection and insertion of histone octamers into nucleosomes,
although the mechanistic details of these reactions are not well
understood.
For the pedagogic purposes of this review, we have described
the various mechanisms of chromatin modification as discrete
reactions, each with unique biochemical and biological conse-
quences. It is important, however, to recognize that there can
be considerable interplay and crosstalk between these varied
mechanisms. For example, it is clear that DNA methylation
affects the degree and location of histone post-translational
modifications and that the two work in concert to control tran-
scription (Collings et al., 2013; Rivera et al., 2014). Also, methyl-
ation of histone H3 at lysine 79 (H3K79) by DOT1L appears to
depend critically on ubiquitylation of the nucleosome on histone
H2B (Wang et al., 2013). Histone methylation at sites H3K36 and
H3K27 appear to be antagonistic to one another, so that methyl-
ation at one site diminishes the ability of the alternative site to be
methylated (Zheng et al., 2012) The activity of the SWI/SNF com-
plex may be modulated by acetylation/deacetylation by HATs/
HDACs and chromatin modification by the SWI/SNF complex
is antagonistically related to methylation of the H3K27 site by
PRC2 (Knutson et al., 2013; Wilson et al., 2010). Finally, there
are several examples of proteins that contain domains leading
to multiple forms of chromatin interaction and remodeling. For
example, the SWI/SNF complex not only contains an ATP-
hydrolysis subunit, but also contains bromodomains that are
critical for recognition of acetyl-lysines and at least one
CHROMO domain; these histone mark-recognition elements of
SWI/SNF likely play an important role in homing of the complex
to specific gene locations that have been appropriately marked
by acetylation and (perhaps) methylation (Tang et al., 2010;
Wilson et al., 2014). Thus, the various forms of chromatin
modification work not as isolated biochemical reactions, but
rather in concert with one another to provide a combinatorial,
chemical ‘‘code’’ that regulates chromatin structure, hence
gene transcription.
Chromatin Modification and Human DiseaseCMPs are genetically altered in a number of human diseases,
most notably in various human cancers. Hence, the develop-
ment of small molecule modulators of specific CMPs as thera-
peutics has become a focal point for drug discovery research.
The genetic alterations of CMPs in human disease and their
pathogenic significance have been reviewed a number of times
in the recent literature (Arrowsmith et al., 2012; Baylin and Jones,
2011; Campbell and Tummino, 2014; Copeland, 2013; Copeland
et al., 2009, 2010, 2013; Helin and Dhanak, 2013; Wigle and
Copeland, 2013). Table 1 summarizes some examples of patho-
genic genetic alterations of CMPs. The reader is referred to the
review articles cited above for additional information on this
topic.
Small Molecule Modulators of Chromatin-ModifyingProteinsThe importance of chromatin modification in human diseases
has spurred significant interest in the discovery of small molecule
modulators of CMPs. A pivotal start for the field was provided by
d All rights reserved
Table 1. Genetic Alterations of Chromatin-Modifying Proteins in Human Diseases
Target Genetic Alteration or Disease Association Indication Key Referencea
PMT
PRC2 (EZH2) heterozygous activating mutations occurring at Y641,
A677 and A687 that result in hypertrimethylation of
H3K27
lymphoma (1–3)
deletion of miR-101 leads to EZH2 overexpression prostate cancer
deletion of SNF5 leads to EZH2 dependency malignant rhabdoid tumors
heterozygous mutations at Y153, H694Y, and
P132S
Weaver syndrome
PRC2 (EZH2, SUZ12,
EED, RBBP4)
SUZ12 fusion protein (JJAZ1/JAZF1) endometrial stromal sarcoma (4–6)
SUZ12 Inactivating mutations T cell precursor acute lymphoblastic
leukemia (ALL)
EED Inactivating mutations acute myeloid leukemia (AML)
DOT1L 11q23 chromosomal translocations fusing MLL1
(without its catalytic SET domain) to DOT1L binding
partners such as AF4, AF9, AF-10, and ENL leading
to aberrant H3K79 methylation
acute leukemia (both ALL and AML) (7)
CALM-AF10 and SET-NUP214 fusions are known to
mistarget DOT1L
NSD1 t(5;11)(q35;p15.5) translocation create NSD1-NUP98
fusions
AML (9, 10)
exon 22 mutations (p.Asp2119Valfs*31) Sotos syndrome
NSD2 (WHSC1/MMSET) t(4:14)(p16;q32) chromosomal translocations that
places WHSC1 gene under the control of the IGH
promoter and results in the overexpression of
WHSC1
multiple myeloma (8, 97)
Glu1099-to-Lys (E1099K) variant is hyperactivating ALL
NSD3 (WHSC1L1) t(8;11)(p11.2;p15) chromosomal translocations fuses
WHSC1L1 to NUP98
AML (11,12)
8p11–12 focal amplifications breast cancer, squamous cell lung
cancer
SETDB1 1q21 amplifications melanoma (13)
SMYD2 1q32 amplifications esophageal squamous cell
carcinoma
(14)
EHMT2 (G9A) H3K9 gene silencing required for apoptotic hair
cell death
sensorineural hearing loss (15, 16)
regulates HoxA9 transcription AML
SETD2 (HYPB, KMT3A) mutations Sotos-like overgrowth syndrome (98, 99)
clear cell renal cell carcinoma
(ccRCC)
SETD7 polymorphisms Type 1 Diabetes (T1D) and diabetic
complications
(17)
SMYD3 upregulation prostate cancer (18–20,100)
tandem repeat polymorphisms in promoter region
(VNTR)
esophageal squamous cell
carcinoma
tandem repeat polymorphisms in E2F-1 binding
element (CCGCC)
colorectal cancer and hepatocellular
carcinoma (HCC)
Type 1 PRMT (1,2,3, 6,8) PRMT1 methylates SRSF1, upregulated in ALL pediatric ALL (18, 21, 22)
PRMT1 upregulated in asthma asthma
PRMT1 methylation of FUS is linked to
heritable ALS
ALS
PRMT8 upregulated prostate cancer
(Continued on next page)
Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Ltd All rights reserved 1201
Chemistry & Biology
Review
Table 1. Continued
Target Genetic Alteration or Disease Association Indication Key Referencea
CARM1 (PRMT4) methylates SWI/SNF core subunit BAF155 colon, breast, and prostate cancer (23–25)
necessary for AR function glycogen storage diseases
necessary for glycogen gene expression program in
skeletal muscle
acute coronary syndrome
involved in chemokine activation (IP-10, MCP-1, IL-8)
Type II PRMTs (5,7) PRMT5 overexpression glioblastoma (15, 26)
ovarian cancer
Putative PMT
NSUN2 5p amplification breast cancer (27, 28, 102)
homozygous truncating mutation (Q227X) mental retardation
METTL1 SNPs multiple sclerosis 101
KDM
KDM4C (JMJD2C, GASC1) 9p23–24 amplifications esophageal squamous cell
carcinoma, squamous cell lung
cancer
(12, 29, 30, 39–41)
t(9;14)(9p24.1q32) translocations creating fusions
to IGH overexpression
medulloblastoma, basal breast
cancer
SNPs
KDM1B (LSD2, AOF1) 6p22 amplifications urothelial carcinomas (31)
KDM6 (UTX) inactivating mutations of UTX lead to pro-oncogenic
hypertrimethylation of H3K27 and dependence
on EZH2
multiple blood and solid cancers (32)
KDM1A (LSD1) overexpression AML (103, 33–35)
hypomorphic allele (point mutations in tower domain) heart development defects
mRNA overexpression ovarian cancer
SNPs (rs671357, rs587168) HCC, salt-sensitive hypertension
KDM4A (JMJD2A) site-specific copy gain/re-replication ovarian cancer (36–38)
overexpression breast cancer
lung cancer
KDM6B (JMJD3) overexpression driven by EBV oncogene Hodgkin’s lymphoma (42)
KDM5A (JARID1A) amplifications breast cancer (104, 48, 49)
t(11;21;12)(p15;p13;p13) translocations create fusions
of JARID1A PHD domain and NUP98
AML
KDM2B (FBXL10, JHDM1B,
PCCX2)
overexpression pancreatic cancer 105
PAD
PADI2 SNPs rheumatoid arthritis (43–45)
overexpression breast cancer
PADI4 SNPs cutaneous basal cell carcinoma (46, 47)
rheumatoid arthritis
Methyl Reader
PHF23 t(11;17)(p15;13) translocations create fusions of
PHF23’s PHD domain to NUP98
AML (48, 50)
PHF1 t(1;6)(p34;p21) translocation create MEAF6-PHF1
fusions, which may misdirect HAT activity toward
PHF1 targets
endometrial stromal sarcoma (51)
SFMBT1 copy number loss ventriculomegaly (52)
HAT
(A) GCN5 (KAT2A) increase in occupancy at regulated promoters spinocerebellar ataxia type 7
(SCA7)
(53, 54)
loss of function retinopathy
(Continued on next page)
1202 Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Ltd All rights reserved
Chemistry & Biology
Review
Table 1. Continued
Target Genetic Alteration or Disease Association Indication Key Referencea
(A) PCAF (KAT2B) Asn386Ser polymorphism HCC (55, 56)
promoter methylation esophageal squamous cell
carcinoma
(A) P300 (KAT3B) C-terminal truncations without catalytic HAT activity diffuse large B-cell lymphoma
(DLBCL)
(57–59)
inactivating mutations cervical cancer
fusion with MLL (t(11;22)(q23;q13) serous endometrial tumors
frameshift mutation AML, Rubinstein-Taybi syndrome,
Cornelia de Lange syndrome
(A) CBP (CREBBP) translocations (t(11;16)(q23;p13.3) with MLL AML (60–63)
translocations t(8;16) with MOZ Rubinstein-Taybi syndrome
point mutations including truncating protein product
duplication (16p13.3)
mental retardation, DLBCL
C-terminal truncations without catalytic HAT activity relapsed ALL
sequence or deletion mutations in HAT domain adenoid cystic carcinoma
(A) TIP60 (KAT5) monoallelic loss
Point mutation (Y327F)
lymphoma, H&N, and mammary
carcinomas
(64, 65)
(A) MOZ (MYST3) fusion gene (inv(8)(p11q13) with TIF2 AML (66, 67)
homozygous (C-terminal deletion) and heterozygous
(deletion of exons 3-7) mutations of Moz
DiGeorge syndrome
(A) MORF (MYST4, KAT6B) MORF/CBP fusion gene—t(10;16)(q22;p13) AML (68–70)
heterozygous mutations (nonsense, truncating,
frameshift)
Ohdo syndrome SBBYS variant
genitopatellar syndrome
HDAC
HDAC 1,2,3,8 (class I) HDAC8 loss of function Cornelia de Lange syndrome (71, 72)
X-linked intellectual disability
HDAC 4,5,7,9 (Class IIa) HDAC4 point mutations brachydactyly-mental retardation (73, 74)
HDAC9 amplification CNS lymphoma
HDAC 6, 10 (Class IIb) HDAC6 point mutation adult-onset Alexander disease (75)
HDAC6 SNP (c.281A > T) X-linked dominant chondrodysplasia
Sirtuins 1-7 (Class III) SIRT1 polymorphisms cardiovascular disease (76–78)
Type 1 diabetes (T1D)
Parkinson’s disease
BRD
BRD4 balanced translocation (t15:19) NUT midline carcinoma (79, 80, 106)
localization to N-terminal domain of AR castration-resistant prostate cancer
transcriptional pause release heart failure
SWI/SNF
SWI/SNF (ARID1A, ARID1B,
SMARCA4, SMARCC1,
SMARCC2, SMARCD1,
SMARCD2, SMARCD3,
ACTL6A, ACTL6B)
ADNP mutation autism spectrum disorder (81–85)
AIRD1B (truncating mutations) intellectual disability (Coffin Siris
syndrome)
SMARCB1 mutations (biallelic) rhabdoid tumors
PBRM1, ARID1A (truncating), SMARCA4 mutations ccRCC
AIRD1A, AIRD1B, PRBM1, SMARCA2, SMARCA4
(alterations, deletions, mutations, and rearrangements)
Burkitt’s lymphoma, pancreatic
cancer
DNMT
DNMT1 heterozygous mutations affecting DNMT1 folding autosomal dominant hereditary
sensory neuropathy type IE
(86, 87)
autosomal dominant cerebellar
ataxia, deafness, and narcolepsy
(Continued on next page)
Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Ltd All rights reserved 1203
Chemistry & Biology
Review
Table 1. Continued
Target Genetic Alteration or Disease Association Indication Key Referencea
DNMT3A somatic mutations (Arg883) AML (88–90)
somatic mutations (reduced enzymatic activity or
aberrant affinity to histone H3)
MDS
heterozygous mutations (Arg882) chronic myelomonocytic
DNMT3L polymorphisms (R271Q) leading to hypomethylation impaired spermatogenesis (91, 92)
TET
TET2 loss of function mutation overexpression systemic mastocytosis, MDS,
AML, CLL
(93–95)
TET1 loss of function mutation AML (96)
ALS, amyotrophic lateral sclerosis; TET, ten-eleven translocation; MDS, myelodysplastic syndromes; CLL, chronic lymphocytic leukemia.aPlease see the Supplemental Information for key references associated with this table.
Chemistry & Biology
Review
the crystal structure of a HDAC with the early inhibitors trichos-
tatin A and suberoylanilide hydroxamic acid (SAHA) bound in
the active site (Finnin et al., 1999). Since then, the field has grown
to see potent, selective inhibitors reported for several of the CMP
classes discussed above. Moreover, inhibitors of HDACs and of
DNMTs are now approved drugs for specific cancer indications
and additional CMP inhibitors have recently entered human clin-
ical trials for the treatment of specific cancers.
Below we describe the current state of efforts toward identifi-
cation of potent, selective small molecule modulators of CMPs,
exemplified by the most advanced compounds for several of the
CMP classes. A number of contemporary reviews have been
published on protein kinase inhibitors [see, for example, (Kollar-
eddy et al., 2012)] and DNMT hypomethylating agents (vide
supra). Since these topics have been well covered within recent
reviews, we shall not reiterate this information here. Instead, we
focus our attention on small molecule modulators of histone
acetylation and methylation.
HAT Inhibitors
Relatively few potent HAT inhibitors have been reported. Early
examples of inhibitors were substrate mimetics and natural
product derivatives of modest potency, selectivity and cell
permeation. High throughput library screening has yielded iso-
thiazolones and benzythiazine sulfonamide scaffolds as inter-
esting starting points for inhibitor discovery. Recently, C646
(Figure 5A) a potent (Ki = 400 nM), selective pyrazolone-based in-
hibitor of the HAT p300/CBP was identified. Intracellular inhibi-
tion of p300/CBP by C646 led to cell growth inhibition, providing
a clear rationale for further exploration of p300/CBP and other
HATs as potential therapeutic targets (Bowers et al., 2010;
Dekker et al., 2014). C646 provides a useful tool compound
with some demonstration of intracellular activity. This com-
pound, however, is not a pharmacologically tractable therapeu-
tic agent. Hence, additional inhibitor discovery efforts are clearly
needed for the HAT target class.
HDAC Modulators
Nonselective HDAC inhibitors have been reported that all act
through chelation of the active site zinc atom of these enzymes.
Two such compounds, vorinostat (SAHA, Zolinza) and romidep-
sin (Istodax) have been approved for the treatment of refractory
cutaneous T cell lymphoma and peripheral T cell lymphoma
(Prince et al., 2009). Vorinostat contains a hydroxamic acid as
1204 Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Lt
the zinc-chelating moiety (Finnin et al., 1999), wherease romi-
depsin is administered as a disulfide-containing, cyclic peptide
prodrug (Furumai et al., 2002). Glutathione-mediated reduction
of the disulfide bond of Istodax liberates a free thiol moiety that
acts to chelate the active site zinc of HDACs (Figure 5B). Both
drugs inhibit a broad spectrum of HDAC enzymes; vorinostat is
reported to inhibit essentially all of the zinc-utilizing HDACs
with similar potency, whereas romidepsin is somewhat more
selective, inhibiting all class I HDACs.
The utility of these drugs for cancers beyond T cell lym-
phomas has been limited in part due to a number of dose-
limiting adverse effects, such as thrombocytopenia and anemia
(Mann et al., 2007). This has led to the suggestion that more
enzyme-selective inhibitors may provide efficacy while mini-
mizing adverse side effects, a hypothesis that has yet to be
tested clinically. Additionally, selective HDAC inhibitors may
also be effective therapeutics for indications beyond cancer,
such as autoimmune diseases and schizophrenia (Bridle et al.,
2013; Kurita et al., 2012).
The most advanced, isozyme-selective compound to be re-
ported to date is the HDAC6 selective inhibitor ACY-1215
(Figure 5B). This compound inhibits HDAC6 with a half-maximal
inhibitory concentration (IC50) of 5 nM; it displays >10-fold selec-
tivity with respect to HDAC1, 2, and 3 and shows no inhibitory
activity against other HDACs (Santo et al., 2012). ACY-1215 is
currently being tested in a phase I/II open-label, multicenter
study as monotherapy and in combination with bortezomib and
dexamethasone for the treatment of relapsed or relapsed/refrac-
tory multiple myeloma (http://clinicaltrials.gov, NCT01323751).
Other examples of selective HDAC inhibitors include BRD8430,
a HDAC1/2 selective inhibitor (IC50 = 69 and 560 nM for HDAC1
and 2, respectively) that induces differentiation in neuroblastoma
(Frumm et al., 2013) and PCI-34051 (Balasubramanian et al.,
2008), a selectiveHDAC8 inhibitor (Ki = 10 nM) being investigated
for potential use in T cell lymphomas (Figure 5B).
Bromodomain Inhibitors
TheBRD class includes a family of protein domains referred to as
the bromodomain and BET proteins. The BET family contains the
proteins BRD2, BRD3, BRD4, and BRDT (bromodomain testis-
specific protein). The BRDs are defined by a well-ordered,
deep, hydrophobic pocket into which the acetyl-lysine side chain
binds. This pocket provides a highly favorable locus for small
d All rights reserved
A
B
C
Figure 5. Small Molecule Modulators ofHistone Acetylation(A) Histone acetyltransferase (HAT) inhibitors.(B) Metal-dependent histone deacetylase (HDAC)inhibitors.(C) BET domain protein antagonists.
Chemistry & Biology
Review
molecule modulator binding; indeed all of the BRD modulators
that have been reported bind within the pocket in a manner
competitive with acetyl-lysine.
A number of BRD binding molecules have been reported and
several of these have been advanced into human clinical trials.
The first BRD antagonist to enter clinical study was RVX-208
(Figure 5C), a preferential BRD2 inhibitor that was tested for
use in atherosclerosis. RVX-208 was well tolerated in patients
at the doses tested (Picaud et al., 2013). Optimization of the thie-
notriazolodiazepene core, first identified by Mitsubishi Tanabe
scientists, led to JQ1 (Figure 5C). This compound selectively
binds to BET family BRDs, such as BRD4, with minimal binding
to non-BET family BRD proteins (Filippakopoulos et al., 2010).
NUT midline carcinoma is a rare, genetically defined form of
squamous carcinoma for which there is currently no effective
treatment. Patients with NUT midline carcinoma are identified
by the occurrence of a chromosomal translocation affecting
the nuclear protein in testis (NUT) gene at the 15q14 location
(French, 2010). In approximately two-thirds of patients, the
translocation results in a fusion protein between NUT and
BRD4 (Stelow, 2011). Preclinical studies with JQ1 have shown
that the compound induces tumor regression and a significant
survival advantage in a mouse model of NUT midline carcinoma
involving the NUT-BRD4 fusion. Optimization efforts have led to
the JQ1 analog TEN-010 (also known as JQ2; structure not avail-
able), which is currently in phase I clinical testing (http://
clinicaltrials.gov, NCT01987362). Contemporaneous work by
the group at GlaxoSmithKline led to the identification of GSK
I-BET762 (GSK525762; Figure 5C), a benzodiazepine that also
selectively inhibits members of the BET-family of BRDs. GSK
I-BET762 displays Ki values of 50–60 nM for BRD2, BRD3, and
BRD4 with minimal effect on non-BET family BRDs (Mirguet
et al., 2013). This compound is currently in phase I clinical
testing for NUT midline carcinoma and other cancers (http://
clinicaltrials.gov, NCT01587703). A similar compound, OTX015
(Figure 5C) is also in phase I clinical testing in acute leukemia
and other hematological malignancies (http://clinicaltrials.gov,
Chemistry & Biology 21, September 18, 2014
NCT01713582). The sponsor of this trial
(Oncoethix) recently reported preliminary
results from their dose escalation studies.
Thrombocytopenia was observed as a
dose-limiting toxicity in patients with other
hematological malignancies at a dose of
80 mg every day, but not when dosed at
40 mg twice daily, suggesting that the
toxicity may be schedule dependent.
Two complete responses and one partial
response were observed in patients with
leukemia or lymphoma and some evi-
dence of response was observed in
another four patients out of a total of
42 patients treated (see http://oncoethix.com). A fourth BET-
family selective BRD inhibitor, CPI-0610, has entered phase I
clinical testing in patients with progressive lymphoma (http://
clinicaltrials.gov, NCT01949883); little additional information on
this compound is currently available. All of the BET-family selec-
tive BRD inhibitors have also shown antiproliferative activity in a
spectrum of solid tumor cell lines. These compounds also share
the ability to silence MYCN expression; amplification of MYCN
has been suggested to be a marker of BET-family selective
BRD inhibitor sensitivity. Indeed, MYCN amplification is being
used as a patient stratification biomarker within the GSK
I-BET762 clinical trial.
PMT Inhibitors
Selective inhibitors of specific PMTs have been reported that
bind within the SAM/SAH pocket, within the protein substrate
pocket or within allosteric sites on these proteins (Copeland
et al., 2013; Wigle and Copeland, 2013).
Potent, selective, SAM-competitive inhibitors have been re-
ported for several of the PKMT enzymes. Most notably, SAM-
competitive inhibitors of the enzymes DOT1L and EZH2 have
advanced to human clinical trials in cancer indications (Cope-
land, 2013). DOT1L catalyzes the methylation of the H3K79 res-
idue, leading to transcriptional activation of affected genes. In
MLL-rearranged leukemia (MLL-r), DOT1L is aberrantly recruited
to ectopic gene locations as a result of binding of the enzyme to
the various MLL-fusion proteins that result from the 11q23 chro-
mosomal translocation that is a universal hallmark of the disease.
EPZ-4777 (Figure 6A) was the first potent, selective DOT1L inhib-
itor demonstrated to selectively kill leukemia cells bearing the
11q23 chromosomal translocation in cell culture and in a murine
model of MLL-r (Daigle et al., 2011). Further optimization of
the amino-nucleoside core of EPZ-4777 led to EPZ-5676
(Figure 6A), the first PMT inhibitor to enter human clinical trials.
EPZ-5676 is a ca. 80 pM inhibitor of DOT1L that is >37,000-
fold selective against other human PMTs. In rat subcutaneous
xenografts of human MLL-r cells, EPZ-5676 treatment led to
essentially complete and durable tumor regression (Daigle
ª2014 Elsevier Ltd All rights reserved 1205
A
B
C
Figure 6. Small Molecule Modulators of Histone Methylation(A) Protein methyltransferase (PMT) inhibitors.(B) Lysine demethylase (KDM) inhibitors.(C) KMe reader domain protein antagonists.
1206 Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Ltd All rights reserved
Chemistry & Biology
Review
Chemistry & Biology
Review
et al., 2013). EPZ-5676 is currently in a phase I study in
advanced hematologic malignancies, including acute leukemia
with rearrangement of the MLL gene (http://clinicaltrials.gov,
NCT01684150). Earlier in 2014, the study sponsor (Epizyme) re-
ported objective responses in MLL-r patients within the context
of this ongoing trial.
Several groups have reported potent, selective, SAM-compet-
itive inhibitors of EZH2 (Figure 6A). Among these, EPZ-6438 (also
known as E7438) is the first EZH2 inhibitor to enter human clinical
trials as a single agent in subjects with advances solid tumors or
with B cell lymphomas (http://clinicaltrials.gov, NCT01897571
(Knutson et al., 2013, 2014). More recently, a second EZH2 inhib-
itor, GSK2816126, has likewise entered clinical testing to inves-
tigate the safety, pharmacokinetics, pharmacodynamics, and
clinical activity of the drug in subjects with relapsed/refractory
diffuse large B cell and transformed follicular lymphoma (http://
clinicaltrials.gov, NCT02082977). Both of these compounds are
nanomolar inhibitors of EZH2 with high selectivity against other
human PMTs. In preclinical studies, both EPZ-6438 and
GSK2816126 demonstrated robust tumor growth inhibition in
xenograft models of lymphomas bearing mutations in EZH2
(Knutson et al., 2014; McCabe et al., 2012; Van Aller et al.,
2014). EPZ-6438 is an orally bioavailable compound that is being
administered orally twice per day, while GSK2816126 is being
administered by intravenous infusion over 2 hours, twice weekly
for 3 weeks of a 28-day dosing cycle. In addition to the non-
Hodgkin lymphoma indication, EZH2 inhibitors have been shown
in preclinical models to effect tumor growth inhibition in INI-1-
deficient tumors, such as malignant rhabdoid tumors and have
also been implicated in additional solid tumor malignancies
(Knutson et al., 2013)
Protein substrate-competitive inhibitors of the PKMTs
EHMT1/2 (e.g., UNC0642 and A-366; Liu et al., 2013; Sweis
et al., 2014), SMYD2 (e.g., AZ505; Ferguson et al., 2011) and
SETD7 (e.g., PFI-2; Structural Genomics Consortium, Protein
Data Bank code 4JLG), and of the PRMT CARM1 (e.g., Methyl-
gene compound 7a; Allan et al., 2009) and BMS compound 7f
(Huynh et al., 2009) have been reported (Figure 6A). None of
these compounds, however, have yet been reported to demon-
strated efficacy in animal models of disease.
A final mechanism of selective PMT inhibitor binding is exem-
plified by the recently reported series of urea-containing allo-
steric inhibitor of PRMT3 (Figure 6A). These compounds bind
at a PRMT3 dimer interface and cause structural rearrangements
of the protein that are allosterically communicated to the enzyme
active site. Themost potent inhibitor in this series displays a Ki of
230 nM and is highly selective for PRMT3, owing to the unique
nature of the compound binding pocket (Siarheyeva et al.,
2012). Whether inhibitors of this type can be pharmacologically
optimized and whether this approach can be more generally
applied to the PMT enzyme class remains to be determined.
KDM Inhibitors
The active sites of the flavin-dependent LSDs bear considerable
similarity to the FAD-dependent monoamino oxidases. Hence,
covalent modifiers of the flavin cofactor of monoamine oxidases,
such as tranylcypromine, also inactivate the LSDs by the same
mechanism. This has been capitalized on to develop LSD-
selective inactivators. Two such compounds, ORY-1001 (Maes
et al., 2013) and GSK2879552 (Kruger et al., 2013; Figure 6B),
Chemistry & Biology 2
have entered human clinical trials for relapsed or refractory
acute leukemia (ORY-1001) and for relapsed/refractory small
cell lung carcinoma (GSK2879552; FAD-http://clinicaltrials.gov,
NCT02034123).
The JmjCs are likewise amenable to small molecule inhibition.
In the case of these enzymes, all known inhibitors derive the
majority of their binding energy from iron chelation (e.g., GSK-
J1; Kruidenier et al., 2012; Figure 6B). Enzyme selectivity has
been modest for JmJC inhibitors and no reported compounds
have progressed beyond biochemical and cell culture assays.
Nevertheless, the pathobiological relevance of the JmjCs
makes them important targets for further inhibitor discovery
and optimization.
Methyl Reader Modulators
The KMe reader domains (vide supra) have only recently become
the focus of small molecule modulator discovery efforts. Only a
few examples of KMe reader domain ligands have been reported
(Wigle and Copeland, 2013), the most potent of which is
UNC1215 (James et al., 2013), a substrate-competitive inhibitor
of L3MBTL3 (Kd = 120 nM; Figure 6C).
Future DirectionsInitiation of clinical trials for isozyme-selective HDAC inhibitors,
BRD antagonists, LSD1 inhibitors, and PMT inhibitors ushers in
an exciting period of testing the safety, pharmacokinetics, and
ultimately the effectiveness of these novel therapeutic modal-
ities. These will hopefully augment the existing CMP-targeted
drugs approved for used in clinical oncology that today consists
exclusively of DNMT hypomethylating agents and broad-spec-
trum HDAC inhibitors. It will be interesting to see how these
CMP drugs, which display such promising effectiveness in pre-
clinical animal models, affect disease as single agents in the
context of human clinical settings.
Beyond the use of these various CMP-targeted drugs as single
agents, there is a growing body of data to suggest that these
agents may be even more effective in treating human cancers
when they are appropriately combined with other CMP modal-
ities, existing standard of care drugs and even drugs that act
to modulate allied intracellular signaling cascades (Johnston
et al., 2013; Klaus et al., 2013). The power of such combinations
of CMP modulators with other therapeutic agents is only begin-
ning to be realized, and deserves considerably more preclinical
and clinical exploration.
Beyond the direct treatment of diseases, such as genetically
defined cancers, small molecule CMPmodulators show promise
as constituents of mixed small molecule/macromolecule cock-
tails that are capable of inducing pluripotent stem cells (iPSCs)
by reprogramming of differentiated, somatic cells. Reprogram-
ming to iPSCs was first shown with a cocktail of four transcrip-
tion factors: OCT4, SOX2, KLF4, and c-Myc (the Yamanaka
factors). Several groups have demonstrated that small molecule
modulators of the kinases ALK5 and MEK and of specific CMPs
can replace individual Yamanaka factors to effect iPSC reprog-
ramming. For example, the PMT inhibitor BIX-01294 can substi-
tute for SOX2 and valproic acid, a pan-HDAC inhibitor, can
substitute for c-Myc (Moschidou et al., 2012). More recently
(Onder et al., 2012) the DOT1L inhibitor EPZ-4777 was shown
to substitute for two Yamanaka Factors, KIF4 and c-Myc. This
same group suggested that selective inhibitors of another PMT
1, September 18, 2014 ª2014 Elsevier Ltd All rights reserved 1207
Chemistry & Biology
Review
enzyme, SUV39H1, would also be effective in promoting iPSC
formation, based on results from shRNA studies. The ability to
induce cellular reprogramming to form iPSCs using small,
organic compounds that can be readily synthesized, in large
scale and at reasonable costs, suggests a practical approach
to the development of regenerative medicine applications.
The continued discovery and development of potent, selec-
tive, small molecule modulators of CMPs may provide new
avenues to novel therapeutic agents for a spectrum of human
diseases. In addition to these clinical applications, the availability
of such CMP modulators will continue to create new opportu-
nities to interrogate the biological consequences of chromatin
modification as well as revealing unanticipated interfaces be-
tween chromatin modification and other cellular signaling
pathways.
SUPPLEMENTAL INFORMATION
Supplemental Information includes references for Table 1 and can be foundwith this article online at http://dx.doi.org/10.1016/j.chembiol.2014.07.024.
ACKNOWLEDGMENTS
The authors thank Drs. Edward Olhava, Tim Wigle, P. Ann Boriack-Sjodin,Robert Gould, and James E. Bradner for helpful discussions and commentson this work. Both authors are employees and stockholders of Epizyme, Inc.
REFERENCES
Allan,M., Manku, S., Therrien, E., Nguyen, N., Styhler, S., Robert, M.F., Goulet,A.C., Petschner, A.J., Rahil, G., Robert Macleod, A., et al. (2009). N-Benzyl-1-heteroaryl-3-(trifluoromethyl)-1H-pyrazole-5-carboxamides as inhibitors ofco-activator associated arginine methyltransferase 1 (CARM1). Bioorg. Med.Chem. Lett. 19, 1218–1223.
Allis, C.D., Jenuwein, T., Reinberg, D., and Caparros, M. (2007). Epigenetics.(New York: Cold Spring Harbor Press).
Andreoli, F., Barbosa, A.J., Parenti, M.D., and Del Rio, A. (2013). Modulation ofepigenetic targets for anticancer therapy: clinicopathological relevance, struc-tural data and drug discovery perspectives. Curr. Pharm. Des. 19, 578–613.
Arrowsmith, C.H., Bountra, C., Fish, P.V., Lee, K., and Schapira, M. (2012).Epigenetic protein families: a new frontier for drug discovery. Nat. Rev. DrugDiscov. 11, 384–400.
Balasubramanian, S., Ramos, J., Luo, W., Sirisawad, M., Verner, E., andBuggy, J.J. (2008). A novel histone deacetylase 8 (HDAC8)-specific inhibitorPCI-34051 induces apoptosis in T-cell lymphomas. Leukemia 22, 1026–1034.
Basavapathruni, A., Jin, L., Daigle, S.R., Majer, C.R., Therkelsen, C.A., Wigle,T.J., Kuntz, K.W., Chesworth, R., Pollock, R.M., Scott, M.P., et al. (2012).Conformational adaptation drives potent, selective and durable inhibition ofthe human protein methyltransferase DOT1L. Chem. Biol. Drug Des. 80,971–980.
Baylin, S.B. (2005). DNA methylation and gene silencing in cancer. Nat. Clin.Pract. Oncol. 2 (Suppl 1 ), S4–S11.
Baylin, S.B., and Jones, P.A. (2011). A decade of exploring the cancer epige-nome - biological and translational implications. Nat. Rev. Cancer 11,726–734.
Bowers, E.M., Yan, G., Mukherjee, C., Orry, A., Wang, L., Holbert, M.A.,Crump, N.T., Hazzalin, C.A., Liszczak, G., Yuan, H., et al. (2010). Virtual ligandscreening of the p300/CBP histone acetyltransferase: identification of a selec-tive small molecule inhibitor. Chem. Biol. 17, 471–482.
Bridle, B.W., Chen, L., Lemay, C.G., Diallo, J.S., Pol, J., Nguyen, A., Capretta,A., He, R., Bramson, J.L., Bell, J.C., et al. (2013). HDAC inhibition suppressesprimary immune responses, enhances secondary immune responses, andabrogates autoimmunity during tumor immunotherapy. Mol. Ther. 21,887–894.
1208 Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Lt
Campbell, R.M., and Tummino, P.J. (2014). Cancer epigenetics drug discoveryand development: the challenge of hitting the mark. J. Clin. Invest. 124, 64–69.
Chang, B., Chen, Y., Zhao, Y., and Bruick, R.K. (2007). JMJD6 is a histone argi-nine demethylase. Science 318, 444–447.
Collings, C.K., Waddell, P.J., and Anderson, J.N. (2013). Effects of DNAmethylation on nucleosome stability. Nucleic Acids Res. 41, 2918–2931.
Copeland, R.A. (2013). Molecular pathways: protein methyltransferases incancer. Clin. Cancer Res. 19, 6344–6350.
Copeland, R.A., Solomon, M.E., and Richon, V.M. (2009). Protein methyltrans-ferases as a target class for drug discovery. Nat. Rev. DrugDiscov. 8, 724–732.
Copeland, R.A., Olhava, E.J., and Scott, M.P. (2010). Targeting epigenetic en-zymes for drug discovery. Curr. Opin. Chem. Biol. 14, 505–510.
Copeland, R.A., Moyer, M.P., and Richon, V.M. (2013). Targeting genetic alter-ations in protein methyltransferases for personalized cancer therapeutics.Oncogene 32, 939–946.
Cuthbert, G.L., Daujat, S., Snowden, A.W., Erdjument-Bromage, H., Hagiwara,T., Yamada, M., Schneider, R., Gregory, P.D., Tempst, P., Bannister, A.J., andKouzarides, T. (2004). Histone deimination antagonizes arginine methylation.Cell 118, 545–553.
Daigle, S.R., Olhava, E.J., Therkelsen, C.A., Majer, C.R., Sneeringer, C.J.,Song, J., Johnston, L.D., Scott, M.P., Smith, J.J., Xiao, Y., et al. (2011). Selec-tive killing of mixed lineage leukemia cells by a potent small-molecule DOT1Linhibitor. Cancer Cell 20, 53–65.
Daigle, S.R., Olhava, E.J., Therkelsen, C.A., Basavapathruni, A., Jin, L., Bor-iack-Sjodin, P.A., Allain, C.J., Klaus, C.R., Raimondi, A., Scott, M.P., et al.(2013). Potent inhibition of DOT1L as treatment of MLL-fusion leukemia. Blood122, 1017–1025.
Dekker, F.J., van den Bosch, T., and Martin, N.I. (2014). Small molecule inhib-itors of histone acetyltransferases and deacetylases are potential drugs for in-flammatory diseases. Drug Discov. Today 19, 654–660.
Dey, A., Chitsaz, F., Abbasi, A., Misteli, T., and Ozato, K. (2003). The doublebromodomain protein Brd4 binds to acetylated chromatin during interphaseand mitosis. Proc. Natl. Acad. Sci. USA 100, 8758–8763.
Dhe-Paganon, S., Syeda, F., and Park, L. (2011). DNA methyl transferase 1:regulatory mechanisms and implications in health and disease. Int J BiochemMol Biol 2, 58–66.
Ferguson, A.D., Larsen, N.A., Howard, T., Pollard, H., Green, I., Grande, C.,Cheung, T., Garcia-Arenas, R., Cowen, S., Wu, J., et al. (2011). Structural basisof substrate methylation and inhibition of SMYD2. Structure 19, 1262–1273.
Filippakopoulos, P., and Knapp, S. (2014). Targeting bromodomains: epige-netic readers of lysine acetylation. Nat. Rev. Drug Discov. 13, 337–356.
Filippakopoulos, P., Qi, J., Picaud, S., Shen, Y., Smith, W.B., Fedorov, O.,Morse, E.M., Keates, T., Hickman, T.T., Felletar, I., et al. (2010). Selective inhi-bition of BET bromodomains. Nature 468, 1067–1073.
Filippakopoulos, P., Picaud, S., Mangos, M., Keates, T., Lambert, J.P., Bar-syte-Lovejoy, D., Felletar, I., Volkmer, R., Muller, S., Pawson, T., et al.(2012). Histone recognition and large-scale structural analysis of the humanbromodomain family. Cell 149, 214–231.
Finnin, M.S., Donigian, J.R., Cohen, A., Richon, V.M., Rifkind, R.A., Marks,P.A., Breslow, R., and Pavletich, N.P. (1999). Structures of a histone deacety-lase homologue bound to the TSA and SAHA inhibitors. Nature 401, 188–193.
Forneris, F., Binda, C., Vanoni, M.A., Mattevi, A., and Battaglioli, E. (2005). His-tone demethylation catalysed by LSD1 is a flavin-dependent oxidative pro-cess. FEBS Lett. 579, 2203–2207.
French, C.A. (2010). Demystified molecular pathology of NUT midline carci-nomas. J. Clin. Pathol. 63, 492–496.
Friedmann, D.R., and Marmorstein, R. (2013). Structure and mechanism ofnon-histone protein acetyltransferase enzymes. FEBS J. 280, 5570–5581.
Frumm, S.M., Fan, Z.P., Ross, K.N., Duvall, J.R., Gupta, S., VerPlank, L., Suh,B.C., Holson, E., Wagner, F.F., Smith, W.B., et al. (2013). Selective HDAC1/HDAC2 inhibitors induce neuroblastoma differentiation. Chem. Biol. 20,713–725.
d All rights reserved
Chemistry & Biology
Review
Furumai, R., Matsuyama, A., Kobashi, N., Lee, K.H., Nishiyama, M., Nakajima,H., Tanaka, A., Komatsu, Y., Nishino, N., Yoshida, M., and Horinouchi, S.(2002). FK228 (depsipeptide) as a natural prodrug that inhibits class I histonedeacetylases. Cancer Res. 62, 4916–4921.
Gregoretti, I.V., Lee, Y.M., and Goodson, H.V. (2004). Molecular evolution ofthe histone deacetylase family: functional implications of phylogenetic anal-ysis. J. Mol. Biol. 338, 17–31.
Haynes, S.R., Dollard, C., Winston, F., Beck, S., Trowsdale, J., and Dawid, I.B.(1992). The bromodomain: a conserved sequence found in human, Drosophilaand yeast proteins. Nucleic Acids Res. 20, 2603.
Helin, K., and Dhanak, D. (2013). Chromatin proteins andmodifications as drugtargets. Nature 502, 480–488.
Hodawadekar, S.C., and Marmorstein, R. (2007). Chemistry of acetyl transferby histone modifying enzymes: structure, mechanism and implications foreffector design. Oncogene 26, 5528–5540.
Holliday, R. (1990). Mechanisms for the control of gene activity during devel-opment. Biol. Rev. Camb. Philos. Soc. 65, 431–471.
Huynh, T., Chen, Z., Pang, S., Geng, J., Bandiera, T., Bindi, S., Vianello, P.,Roletto, F., Thieffine, S., Galvani, A., et al. (2009). Optimization of pyrazoleinhibitors of coactivator associated arginine methyltransferase 1 (CARM1).Bioorg. Med. Chem. Lett. 19, 2924–2927.
James, L.I., and Frye, S.V. (2013). Targeting chromatin readers. Clin. Pharma-col. Ther. 93, 312–314.
James, L.I., Barsyte-Lovejoy, D., Zhong, N., Krichevsky, L., Korboukh, V.K.,Herold, J.M., MacNevin, C.J., Norris, J.L., Sagum, C.A., Tempel, W., et al.(2013). Discovery of a chemical probe for the L3MBTL3 methyllysine readerdomain. Nat. Chem. Biol. 9, 184–191.
Johnston, L.D., Knutston, S.K., Warholic, N., Klaus, C.R., Wigle, T.J., Iwano-wicz, D., Littlefield, B.A., Porter Scott, M., Smith, J.J., Daigle, S.R., et al.(2013). EZH2 inhibitor EPZ-6438 synergizes with anti-lymphoma therapies inpreclinical models. Blood 122, 4416.
Klaus, C.R., Daigle, S.R., Iwanowics, D., Johnston, D., Klaus, C.R., PorterScott, M., Warholic, N., Knutston, S.K., Smith, J.J., Wigle, T.J., et al. (2013).DOT1L inhibitor EPZ-5676 displays synergistic antiproliferative activity incombination with standard of care drugs or DNA hypomethylating agents inMLL-rearranged leukemia cells. Blood 122, 3930.
Knutson, S.K., Warholic, N.M., Wigle, T.J., Klaus, C.R., Allain, C.J., Raimondi,A., Porter Scott, M., Chesworth, R., Moyer, M.P., Copeland, R.A., et al. (2013).Durable tumor regression in genetically altered malignant rhabdoid tumors byinhibition of methyltransferase EZH2. Proc. Natl. Acad. Sci. USA 110, 7922–7927.
Knutson, S.K., Kawano, S., Minoshima, Y., Warholic, N.M., Huang, K.C., Xiao,Y., Kadowaki, T., Uesugi, M., Kuznetsov, G., Kumar, N., et al. (2014). Selectiveinhibition of EZH2 by EPZ-6438 leads to potent antitumor activity in EZH2-mutant non-Hodgkin lymphoma. Mol. Cancer Ther. 13, 842–854.
Kollareddy, M., Zheleva, D., Dzubak, P., Brahmkshatriya, P.S., Lepsik, M., andHajduch, M. (2012). Aurora kinase inhibitors: progress towards the clinic.Invest. New Drugs 30, 2411–2432.
Kruger, R.G., Mohammad, H., Smitheman, K., Cusan, M., Liu, Y., Pappalardi,M., Federowicz, K., Van Aller, G., Kasparec, J., Tian, X., et al. (2013). Inhibitionof LSD1 as a therapeutic strategy for the treatment of acute myeloid leukemia.Blood 122, 3964.
Kruidenier, L., Chung, C.W., Cheng, Z., Liddle, J., Che, K., Joberty, G., Bant-scheff, M., Bountra, C., Bridges, A., Diallo, H., et al. (2012). A selective jumonjiH3K27 demethylase inhibitor modulates the proinflammatory macrophageresponse. Nature 488, 404–408.
Kuo, A.J., Song, J., Cheung, P., Ishibe-Murakami, S., Yamazoe, S., Chen, J.K.,Patel, D.J., and Gozani, O. (2012). The BAH domain of ORC1 links H4K20me2to DNA replication licensing andMeier-Gorlin syndrome. Nature 484, 115–119.
Kurita, M., Holloway, T., Garcıa-Bea, A., Kozlenkov, A., Friedman, A.K., Mor-eno, J.L., Heshmati, M., Golden, S.A., Kennedy, P.J., Takahashi, N., et al.(2012). HDAC2 regulates atypical antipsychotic responses through the modu-lation of mGlu2 promoter activity. Nat. Neurosci. 15, 1245–1254.
Lee, K.K., and Workman, J.L. (2007). Histone acetyltransferase complexes:one size doesn’t fit all. Nat. Rev. Mol. Cell Biol. 8, 284–295.
Chemistry & Biology 2
Liu, X., Wang, L., Zhao, K., Thompson, P.R., Hwang, Y., Marmorstein, R., andCole, P.A. (2008). The structural basis of protein acetylation by the p300/CBPtranscriptional coactivator. Nature 451, 846–850.
Liu, F., Barsyte-Lovejoy, D., Li, F., Xiong, Y., Korboukh, V., Huang, X.P., Allali-Hassani, A., Janzen, W.P., Roth, B.L., Frye, S.V., et al. (2013). Discovery of anin vivo chemical probe of the lysine methyltransferases G9a and GLP. J. Med.Chem. 56, 8931–8942.
Lombardi, P.M., Cole, K.E., Dowling, D.P., and Christianson, D.W. (2011).Structure, mechanism, and inhibition of histone deacetylases and related met-alloenzymes. Curr. Opin. Struct. Biol. 21, 735–743.
Maes, T., Tirapu, I., Mascaro, C., Ortega, A., Estiarte, A., Nuria, V., Castro-Palomino, J., Buesa Arjol, C., Kurz, G., and Genomics, O. (2013). Preclinicalcharacterization of a potent and selective inhibitor of the histone demethylaseKDM1A for MLL leukemia. J. Clin. Oncol. 31, e13543.
Mann, B.S., Johnson, J.R., Cohen, M.H., Justice, R., and Pazdur, R. (2007).FDA approval summary: vorinostat for treatment of advanced primary cuta-neous T-cell lymphoma. Oncologist 12, 1247–1252.
Marmorstein, R. (2001). Structure and function of histone acetyltransferases.Cell. Mol. Life Sci. 58, 693–703.
McCabe, M.T., Ott, H.M., Ganji, G., Korenchuk, S., Thompson, C., Van Aller,G.S., Liu, Y., Graves, A.P., Della Pietra, A., 3rd, Diaz, E., et al. (2012). EZH2 in-hibition as a therapeutic strategy for lymphoma with EZH2-activating muta-tions. Nature 492, 108–112.
Milne, J.C., and Denu, J.M. (2008). The Sirtuin family: therapeutic targets totreat diseases of aging. Curr. Opin. Chem. Biol. 12, 11–17.
Mirguet, O., Gosmini, R., Toum, J., Clement, C.A., Barnathan, M., Brusq, J.M.,Mordaunt, J.E., Grimes, R.M., Crowe,M., Pineau, O., et al. (2013). Discovery ofepigenetic regulator I-BET762: lead optimization to afford a clinical candidateinhibitor of the BET bromodomains. J. Med. Chem. 56, 7501–7515.
Moschidou, D., Mukherjee, S., Blundell, M.P., Drews, K., Jones, G.N., Abdul-razzak, H., Nowakowska, B., Phoolchund, A., Lay, K., Ramasamy, T.S., et al.(2012). Valproic acid confers functional pluripotency to human amniotic fluidstem cells in a transgene-free approach. Mol. Ther. 20, 1953–1967.
Mujtaba, S., Zeng, L., and Zhou, M.M. (2007). Structure and acetyl-lysinerecognition of the bromodomain. Oncogene 26, 5521–5527.
Onder, T.T., Kara, N., Cherry, A., Sinha, A.U., Zhu, N., Bernt, K.M., Cahan, P.,Marcarci, B.O., Unternaehrer, J., Gupta, P.B., et al. (2012). Chromatin-modi-fying enzymes as modulators of reprogramming. Nature 483, 598–602.
Owen, D.J., Ornaghi, P., Yang, J.C., Lowe, N., Evans, P.R., Ballario, P., Neu-haus, D., Filetici, P., and Travers, A.A. (2000). The structural basis for therecognition of acetylated histone H4 by the bromodomain of histone acetyl-transferase gcn5p. EMBO J. 19, 6141–6149.
Ptashne, M. (2013). Epigenetics: core misconcept. Proc. Natl. Acad. Sci. USA110, 7101–7103.
Picaud, S., Wells, C., Felletar, I., Brotherton, D., Martin, S., Savitsky, P., Diez-Dacal, B., Philpott, M., Bountra, C., Lingard, H., et al. (2013). RVX-208, aninhibitor of BET transcriptional regulators with selectivity for the second bro-modomain. Proc. Natl. Acad. Sci. USA 110, 19754–19759.
Prince, H.M., Bishton, M.J., and Harrison, S.J. (2009). Clinical studies of his-tone deacetylase inhibitors. Clin. Cancer Res. 15, 3958–3969.
Rice, J.C., and Allis, C.D. (2001). Histone methylation versus histone acetyla-tion: new insights into epigenetic regulation. Curr. Opin. Cell Biol. 13, 263–273.
Richon, V.M., Johnston, D., Sneeringer, C.J., Jin, L., Majer, C.R., Elliston, K.,Jerva, L.F., Scott, M.P., and Copeland, R.A. (2011). Chemogenetic analysisof human protein methyltransferases. Chem. Biol. Drug Des. 78, 199–210.
Rivera, C., Gurard-Levin, Z.A., Almouzni, G., and Loyola, A. (2014).Histone lysine methylation and chromatin replication. Biochim. Biophys.Acta. Published online March 28, 2014. http://dx.doi.org/10.1016/j.bbagrm.2014.03.009.
Rossetto, D., Avvakumov, N., and Cote, J. (2012). Histone phosphorylation: achromatin modification involved in diverse nuclear events. Epigenetics 7,1098–1108.
1, September 18, 2014 ª2014 Elsevier Ltd All rights reserved 1209
Chemistry & Biology
Review
Roth, S.Y., Denu, J.M., and Allis, C.D. (2001). Histone acetyltransferases.Annu. Rev. Biochem. 70, 81–120.
Santo, L., Hideshima, T., Kung, A.L., Tseng, J.C., Tamang, D., Yang,M., Jarpe,M., van Duzer, J.H., Mazitschek, R., Ogier, W.C., et al. (2012). Preclinical activ-ity, pharmacodynamic, and pharmacokinetic properties of a selective HDAC6inhibitor, ACY-1215, in combination with bortezomib in multiple myeloma.Blood 119, 2579–2589.
Shain, A.H., and Pollack, J.R. (2013). The spectrum of SWI/SNF mutations,ubiquitous in human cancers. PLoS ONE 8, e55119.
Siarheyeva, A., Senisterra, G., Allali-Hassani, A., Dong, A., Dobrovetsky, E.,Wasney, G.A., Chau, I., Marcellus, R., Hajian, T., Liu, F., et al. (2012). Anallosteric inhibitor of protein arginine methyltransferase 3. Structure 20,1425–1435.
Stelow, E.B. (2011). A review of NUT midline carcinoma. Head Neck Pathol. 5,31–35.
Suganuma, T., and Workman, J.L. (2012). MAP kinases and histone modifica-tion. J. Mol. Cell Biol. 4, 348–350.
Sweis, R.F., Pliushchev, M., Brown, P.J., Guo, J., Li, F., Maag, D., Petros, A.M.,Soni, N.B., Tse, C., Vedadi, M., et al. (2014). Discovery and development ofpotent and selective inhibitors of histone methyltransferase g9a. ACS MedChem Lett 5, 205–209.
Tang, L., Nogales, E., and Ciferri, C. (2010). Structure and function of SWI/SNFchromatin remodeling complexes and mechanistic implications for transcrip-tion. Prog. Biophys. Mol. Biol. 102, 122–128.
Tsukada, Y., Fang, J., Erdjument-Bromage, H., Warren, M.E., Borchers, C.H.,Tempst, P., and Zhang, Y. (2006). Histone demethylation by a family of JmjCdomain-containing proteins. Nature 439, 811–816.
Van Aller, G.S., Pappalardi, M.B., Ott, H.M., Diaz, E., Brandt, M., Schwartz,B.J., Miller, W.H., Dhanak, D., McCabe, M.T., Verma, S.K., et al. (2014).Long residence time inhibition of EZH2 in activated polycomb repressive com-plex 2. ACS Chem. Biol. 9, 622–629.
Walsh, C.T. (2006). Posttranslational Modification of Proteins: ExpandingNature’s Inventory. (Englewood, CO: Roberts & Company).
1210 Chemistry & Biology 21, September 18, 2014 ª2014 Elsevier Lt
Wang, Y., Wysocka, J., Sayegh, J., Lee, Y.-H., Perlin, J.R., Leonelli, L., Son-buchner, L.S., McDonald, C.H., Cook, R.G., Dou, Y., et al. (2004). HumanPAD4 regulates histone arginine methylation levels via demethylimination.Science 306, 279–283.
Wang, E., Kawaoka, S., Yu, M., Shi, J., Ni, T., Yang, W., Zhu, J., Roeder,R.G., and Vakoc, C.R. (2013). Histone H2B ubiquitin ligase RNF20 isrequired for MLL-rearranged leukemia. Proc. Natl. Acad. Sci. USA 110,3901–3906.
Wang, X., Haswell, J.R., and Roberts, C.W. (2014). Molecular pathways: SWI/SNF (BAF) complexes are frequently mutated in cancer—mechanisms andpotential therapeutic insights. Clin. Cancer Res. 20, 21–27.
Wigle, T.J., and Copeland, R.A. (2013). Drugging the human methylome: anemerging modality for reversible control of aberrant gene transcription. Curr.Opin. Chem. Biol. 17, 369–378.
Wilkins, B.J., Rall, N.A., Ostwal, Y., Kruitwagen, T., Hiragami-Hamada, K.,Winkler, M., Barral, Y., Fischle, W., and Neumann, H. (2014). A cascade of his-tone modifications induces chromatin condensation in mitosis. Science 343,77–80.
Wilson, B.G., and Roberts, C.W. (2011). SWI/SNF nucleosome remodellersand cancer. Nat. Rev. Cancer 11, 481–492.
Wilson, B.G., Wang, X., Shen, X., McKenna, E.S., Lemieux, M.E., Cho, Y.-J.,Koellhoffer, E.C., Pomeroy, S.L., Orkin, S.H., and Roberts, C.W.M. (2010).Epigenetic antagonism between polycomb and SWI/SNF complexes duringoncogenic transformation. Cancer Cell 18, 316–328.
Wilson, B.G., Helming, K.C., Wang, X., Kim, Y., Vazquez, F., Jagani, Z., Hahn,W.C., and Roberts, C.W. (2014). Residual complexes containing SMARCA2(BRM) underlie the oncogenic drive of SMARCA4 (BRG1) mutation. Mol.Cell. Biol. 34, 1136–1144.
Zheng, Y., Sweet, S.M., Popovic, R., Martinez-Garcia, E., Tipton, J.D.,Thomas, P.M., Licht, J.D., and Kelleher, N.L. (2012). Total kinetic anal-ysis reveals how combinatorial methylation patterns are established onlysines 27 and 36 of histone H3. Proc. Natl. Acad. Sci. USA 109,13549–13554.
d All rights reserved