Chromodomain antagonists that target the polycomb-group methyllysine reader protein chromobox homolog 7 (CBX7)

Article abstract of DOI:10.1021/jm401487x

We report here a peptide-driven approach to create first inhibitors of the chromobox homolog 7 (CBX7), a methyllysine reader protein. CBX7 uses its chromodomain to bind histone 3, lysine 27 trimethylated (H3K27me3), and this recognition event is implicated in silencing multiple tumor suppressors. Small trimethyllysine containing peptides were used as the basic scaffold from which potent ligands for disruption of CBX7-H3K27me3 complex were developed. Potency of ligands was determined by fluorescence polarization and/or isothermal titration calorimetry. Binding of one ligand was characterized in detail using 2D NMR and X-ray crystallography, revealing a structural motif unique among human CBX proteins. Inhibitors with a ～200 nM potency for CBX7 binding and 10-fold/400-fold selectivity over related CBX8/CBX1 proteins were identified these are the first reported inhibitors of any chromodomain.

Full text of DOI:10.1021/jm401487x

Article
pubs.acs.org/jmc
Chromodomain Antagonists That Target the Polycomb-Group
Methyllysine Reader Protein Chromobox Homolog 7 (CBX7)
Chakravarthi Simhadri,†,⊥ Kevin D. Daze,†,⊥ Sarah F. Douglas,† Taylor T. H. Quon,† Amarjot Dev,†
Michael C. Gignac,† Fangni Peng,‡ Markus Heller,§ Martin J. Boulanger,‡ Jeremy E. Wulff,†
and Fraser Hof*,†
†Department of Chemistry, University of Victoria, P.O. Box 3065, Victoria, British Columbia, V8W 3V6, Canada
‡Department of Biochemistry and Microbiology, University of Victoria, P.O. Box 3065, Victoria, British Columbia, V8W 3V6, Canada
§Centre for Drug Research and Development, 2405 Wesbrook Mall, Vancouver, British Columbia, V6T 1Z3, Canada
S
* Supporting Information
ABSTRACT: We report here a peptide-driven approach to create first inhibitors of the chromobox homolog 7 (CBX7), a
methyllysine reader protein. CBX7 uses its chromodomain to bind histone 3, lysine 27 trimethylated (H3K27me3), and this
recognition event is implicated in silencing multiple tumor suppressors. Small trimethyllysine containing peptides were used as
the basic scaffold from which potent ligands for disruption of CBX7-H3K27me3 complex were developed. Potency of ligands was
determined by fluorescence polarization and/or isothermal titration calorimetry. Binding of one ligand was characterized in detail
using 2D NMR and X-ray crystallography, revealing a structural motif unique among human CBX proteins. Inhibitors with a
∼200 nM potency for CBX7 binding and 10-fold/400-fold selectivity over related CBX8/CBX1 proteins were identified. These
are the first reported inhibitors of any chromodomain.
these proteins are predicted to be druggable.22 The first
inhibitors of any methyllysine reader were reported to target a
INTRODUCTION
■
A complex language of histone modifications is the basis for the
few similar members of a single family (the malignant brain
epigenetic control of gene expression. In the metaphorical
tumour (MBT) proteins) and have moved quickly through
“histone code” there are “writer” and “eraser” proteins that add
subsequent rounds of optimization.23−27 As with bromodomain
and remove modifications including methylation, acetylation,
phosphorylation, ubiquitylation, SUMOylation, and others.1
inhibitors, the first tool compound against an MBT domain has
been demonstrated to have a potent influence on the biology of
There are also “reader” proteins that recognize and bind to
cancer cells.28 Inhibitors of a PHD (plant homeodomain)
certain histone modifications, including acetyllysine, methyl-
methyllysine reader have also been recently disclosed.29
lysines, and methylarginines.2−4 Inhibitors of writer proteins
like histone methyltransferases (HMTs) have entered the
clinic,5−12 and inhibitors of histone deacetylase (HDAC) eraser
proteins are approved and used clinically against multiple
malignancies.13−15
We are primarily interested in developing tool compounds to
probe the activities of protein chromodomains (“chromatin
organization modifier’” domains), which are a family of
methyllysine readers generally associated with chromatin
remodeling.30−32 Thirty-one human chromodomains exist,20
and each is predicted to act as a modular unit for the
recognition of a mono-, di-, or trimethyllysine modifications in
different structural contexts. One major family of human
chromodomains are the paralogs of the Drosophila protein
Very few reader proteins have yet been successfully targeted
by small molecule agents. Acetyllysine marks are read by
bromodomain-family proteins, and bromodomain inhibitors
(which were the first inhibitors of any reader protein) have
shown powerful and diverse biological activities.16−19 Methyl-
lysine marks are read by a diverse set of hundreds of human
proteins, with the major groups being the MBT, PHD, Tudor,
PWWP, WDR, and chromodomain families.4,20,21 Many of
Received: September 25, 2013
Published: March 13, 2014
© 2014 American Chemical Society
2874
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
Figure 1. (a) The product of a methyltransferase (e.g., EZH2) acting on a histone lysine (H3K27) is a methylated residue (H3K27me3) that is
recognized and bound by a reader protein (CBX7). (b) Competitive fluorescence polarization (FP) data can report on the disruption of the complex
between recombinant CBX7 chromodomain and H3K27me3. Data shown arise from addition of compound 1 and compound 16 (as two examples
of active compounds), compound 2, unmethylated control, or BSA (additional control) to a solution containing CBX7 chromodomain (8.68 μM)
and FITC-H3K27me3 probe peptide (500 nM).
Polycomb, called chromobox homolog (CBX) proteins CBX2,
CBX4, CBX6, CBX7, and CBX8.31 Polycomb group genes are
transcriptional repressors that play essential roles in develop-
ment, cancer progression, and stem cell maintenance.33
Polycomb group proteins form multiprotein polycomb
repressive complexes (PRC1/2), members of which are
frequently dysregulated in metastatic cancers in which tumor
suppressor genes are silenced and stemlike phenotypes are
acquired.33−35 Gene suppression is initiated by the writer
protein enhancer of zeste homolog 2 (EZH2) of PRC2, which
trimethylates lysine 27 on histone H3 to create the mark
H3K27me3. The resulting gene-specific trimethyl marks then
provide the binding sites for the reader components of PRC1
(CBX proteins) which cause gene silencing and ultimately lead
to stable DNA methylation. The five human paralogs CBX2, -4,
-6, -7, and -8 participate in a modular way within different
PRC1 complexes.36,37 Each recognizes and binds to H3K27me3
mark laid down by EZH2 using its chromodomain but has
distinct genetic targets, overall domain structure, and functional
outcomes.36,37 CBX7 is the best-studied of the Polycomb
paralogs; it is a master controller that extends cellular lifespan,
delays senescence, drives proliferation, and bestows pluripo-
tency to adult and embryonic stem cells.38−40
Polycomb group proteins have been highlighted as being
important in different malignancies.41 Multiple lines of evidence
suggest promise for CBX7 as a therapeutic target in certain
forms of cancer. CBX7 expression is strongly proliferative in
various stem cell and stemlike cancer cell lines.36,38−40,42−44
The depletion or knockdown of CBX7 in both androgen-
dependent (AD) and androgen-independent (AI) prostate
cancer cell lines induces a senescent phenotype with reduced
cell proliferation by repression of the p16INK4a/p14ARF tumor
suppressor locus.39,40 CBX7 expression in the prostate increases
as AD prostate cancer progresses to AI status; in a panel of 27
diverse cancer cell lines, higher CBX7 expression was observed
in two lines of aggressive, AI cells (PC3, DU145) than in the
AD line LNCaP (very high CBX7 was also seen in metastatic
lymphoma (K299) and glioblastoma (U87MG) lines).39 CBX7
expression significantly increases in clinical samples of
hormone-resistant prostate cancer relative to hormone-depend-
ent prostate cancer45 and was recently linked to poor prognosis
in ovarian clear cell adenocarcinoma.44 CBX7 also plays a
central role in some blood malignancies. It initiates and
accelerates lymphomagenesis and cooperates with Myc to
generate aggressive lymphomas.43 Overexpression of CBX7 in
hematopoietic stem cells drives proliferation and genesis of T-
cell leukemia (while its close relative CBX8 does not).46 CBX7
is consistently shown to be strongly proliferative in androgen-
independent prostate cancer cells, embryonic and adult stem
cells, and hematopoiesis/lymphomagenisis.36,38−40,42,43,46
These cell types are linked by their stemlike gene regulation
and phenotypes.
Epigenetic targets like the CBX proteins often have strongly
tissue- and context-selective functions, and the literature on
CBX7 outside of the prostate, embryonic, hematapoietic, or
stem-cell contexts contains some results that give pause.33,47−51
But many such functional studies on CBX7 involve the
knockout/knockdown of the whole, multidomain protein that
do not predict the outcome of chemical chromodomain
antagonism. Critically, multiple reports show that the prolifer-
ative/prosurvival functions of CBX7 depend narrowly on its
H3K27me3 binding ability; when a single Kme3-binding residue in
its chromodomain is mutated, CBX7-driven proliferation dis-
appears.39,40,52 Accordingly, our goal is to create tool
compounds that antagonize the chromodomain of CBX7 in
order to test the antiproliferative potential of this novel
epigenetic target.
2875
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
RESULTS AND DISCUSSION
Table 1. H3K27me3-CBX7 Disruption by FALKme3S
Analogs Bearing Leucine (−1) Replacements
■
Our early efforts at virtual screening failed to yield potent lead
compounds against the CBX7 chromodomain. A high-
throughput screen against the somewhat related chromodo-
main of CBX1 (HP1β) was completed at the National
Chemical Genomics Center but failed to produce tractable
hits and was not published.53 We were inspired to start on a
search for peptidic antagonists of the CBX7 chromodomain by
a report of the binding by CBX7 of a 25 amino acid long
peptide sequence derived from the protein SETDB1, which was
reported to have slightly higher in vitro affinity for CBX7 than
the H3K27me3 sequence itself.31 We truncated the methylated
SETDB1 sequence to arrive at a five amino acid long peptide
(1) that demonstrated reasonable potency (IC50 = 12 μM) in a
competitive fluorescence polarization (FP) assay for CBX7-
H3K27me3 disruption (Figure 1). In a variety of in vitro and in
vivo studies, CBX7 binding has been shown to depend strongly
on the methylation of the lysine residue that binds in CBX7’s
aromatic cage motif made up of Phe5, Trp26, and Trp29
(numbering derived from the CBX7 chromodomain construct
used in this study). This allowed us to demonstrate the
specificity of the initial assay result by a control titration with
the unmethylated analog (2), which showed no ability to
displace H3K27me3 from CBX7 under the assay conditions
(Figure 1). Thus, we used the small peptide (1) as a scaffold in
order to establish structure−function relationships for CBX7
antagonism.
We built a simple, energy-minimized molecular model of the
complex of CBX7 with Ac-FALKme3S-NH2 (1) starting with
the published, NMR-derived structure 2L1B that shows CBX7
bound to its native histone tail partner, H3K27me3 (whose
sequence at the trimethyllysine site is ...AARKme3S...).31 As in
the native complex, the trimethyllysine residue makes multiple
cation−π contacts to the aromatic cage motif of CBX7 (Phe5,
Trp26, Trp29), and the adjacent (+1) Ser residue donates
hydrogen bonds from side chain OH and backbone NH to the
carboxylate side chain of Glu37. The molecular recognition
requirements for the peptide’s (−1) Leu residue were not
immediately clear to us because, in addition to a hydrophobic
contact with a well-ordered Val10 residue, it is close to the N-
terminal CBX7 residues (Gly1/Glu2/Gln3) that are poorly
ordered in the NMR structure. Thus, we first made a family of
analogs of the peptide Ac-FALKme3S-NH2 that varied at the
(−1) Leu position and tested them for their ability to disrupt
the H3K27me3-CBX7 complex as indicated by fluorescence
polarization assay (Table 1). A variety of hydrophobic and/or
aromatic residues at this position provided strong (<10 μM)
inhibitors. A preference for hydrophobic, β-branched residues
can be clearly seen in the cases of compounds 3, 4, and 16. n-
Alkyl inhibitors 6 and 7 are weak inhibitors, as is the overly
bulky tert-butylglycine inhibitor 5. Aromatics are generally well
tolerated.
a
IC50 values were determined as the inhibitor concentration required
for the half-dissociation of the complex between CBX7 and FITC-
labeled H3K27me3. Data are mean values of two or three independent
experimental trails. Errors reported are 95% confidence intervals.
Figure 2. Cocrystal structure of CBX7 chromodomain bound to ligand
11 as determined by X-ray crystallography. CBX7 is shown as a gray
solvent-accessible surface, and the ligand is shown as yellow sticks
overlaid with the electron density contours (Coot). The hydrophobic
clasp between Val4 and Leu43 is common to all chromodomains. The
Mg2+-mediated bridge between Glu2 and His41 appears to be unique
to CBX7 among human CBX proteins.
We solved the X-ray structure of the complex of CBX7 with
compound 11 to 1.54 Å resolution (Figure 2). The structure
confirms the ligand binding pose predicted by modeling. The
significant recognition elements are the trimethyllysine side
chain participating in multiple cation−π interactions, extensive
backbone-to-backbone β-sheet-like hydrogen bonding network
between protein and ligand, the Ser(+1) residue of the ligand
donating multiple H-bonds from side chain OH and backbone
NH to Glu37 of the protein, and the Ala(−2) side chain of the
ligand being nestled tightly into the small hydrophobic pocket
of CBX7 that is largely responsible for its substrate-binding
selectivity. The exposure of the (−1) substituent (in this ligand,
Tyr) to solvent provides a good explanation for the tolerance of
this position to different substitutions (Table 1). Unexpectedly,
the X-ray structure also revealed a significantly more closed
2876
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
binding pocket than is depicted in the existing NMR structure
of the CBX7-H3K27me3 complex.31 In addition to the
“hydrophobic clasp,” which is a major side chain to side
chain bridge that is formed by Val4/Leu43 in all Polycomb
paralogs,31 the new structure revealed an unexpected metal-
mediated bridge between Glu2 and His41 as an additional
clasping motif that envelops the bound ligand. Neither the
purification buffers nor the crystallization solutions contained
any divalent metal, which was modeled as Mg2+. Our choice of
Mg2+ for this model is supported indirectly by the facts that the
tetrahedral coordination geometry54 is that which would be
observed for Mg2+, and that Mg2+ concentrations are high in
chromatin. The biological relevance of the bound metal is
unclear, but it is intriguing that CBX7 is unique among the
eight human CBX proteins from both Polycomb and HP1
families in having a metal-binding His residue at position 41; all
others have a Asn at this position and adopt a more open
conformation (see Supporting Information).22 In solution, we
found only small differences in ligand binding to CBX7 in the
absence/presence of 2 mM MgCl2 (data not shown),
suggesting that this motif, while structurally unique, is not
necessarily a significant determinant in ligand binding.
Table 2. H3K27me3-CBX7 Disruption by FALKme3S
Analogs Bearing Alanine (−2) Replacements
a
compd
R
IC50 (μM)
1
Ala
Abu
Nle
Ser
Val
11 0.4
18
19
20
21
39
3
120 30
53
3
>500
a
IC50 values were determined as the inhibitor concentration required
for the half-dissociation of the complex between CBX7 and FITC-
labeled H3K27me3. Data are mean values of two or three independent
experimental trails. Errors reported are 95% confidence intervals.
We carried out 2D NMR studies to confirm that the ligand-
binding pose shown by X-ray also operates in solution.
Titration of ligand 11 into an 15N-labeled sample of CBX7
position were made by preparing the C-terminal carboxy
peptide Ac-FALKme3-CO2− on 2-chlorotrityl resin, purification
by HPLC, and subsequent coupling to a variety of different
amines as shown in Scheme 1a. At this position, the installation
of hydrophobic functionality significantly weakened binding
(26, 27, 35), as did the inclusion of a variety of conformation-
ally constrained bis-hydrogen bond donors 26, 27, 28, and 29.
Happily, we found that simple H-bond-donating substituents,
including an aminobenzimidazole (30), a propanediol (33),
and a pendent ammonium group (34), served well in this
region presumably because they can donate hydrogen bonds to
Glu37 in a manner analogous to the ligand Ser residue that
normally occupies this position.
1
was first followed by H−15N HSQC. Mapping the binding-
induced chemical shift changes for assigned backbone nitrogen
resonances onto the structure of CBX7 provides a clear picture
of the binding site that is consistent with X-ray data: Medium/
large chemical shift perturbations occur for residues directly in
contact with the Kme3 side chain (Gln3/Trp26/Trp29/Thr35)
and with the contact partners for the inhibitor’s Phe and Tyr
residues (His41/Leu43/Asp44; see Supporting Information).
Further, the side chain methyl groups of Val4 and Leu43 are
both predicted by models to be engaged with the faces of the
1
ligand’s Tyr residue. Accordingly, H−13C HSQC experiments
The molecular recognition requirements of the extended β-
groove region of CBX7, occupied by the Ac-F portion of Ac-
FALKme3S-NH2, were probed by two different families of
inhibitors. The first group of Ac-FALKme3S-NH2 analogs were
made by replacing Phe with a natural or unnatural fragment
(Table 4). Simple replacement of Phe with other aromatic
amino acids gave peptides with only small changes in potency
(Tyr, Trp, Phg). Installation of non amino acid groups at the
N-terminus was achieved by preparing the peptide Fmoc-
ALKme3S- on resin, with subsequent deprotection of the N-
terminus and reaction with a variety of electrophiles and
coupling partners (Scheme 1b). We found that the retention of
a “backbone” CO in this position, which is strongly ordered
in the parent X-ray structure by accepting a hydrogen bond
from the backbone NH of Val7, is critically important.
Replacement of this CO with a sulfonamide linkage (48,
49) was detrimental, while the use of on-resin reductive
aminations to produce benzylamine-type linkages (52, 53)
completely abrogated ligand binding. The potency of amide-
and carbamate-type compounds that preserve this ligand CO
was generally good, with variations depending on the exact
nature of aromatic substituent and linker directed into the β-
groove.
showed that significant broadening and upfield chemical shift
for these Val/Leu methyl groups occur upon addition of the
peptide, adding specific support to this aspect of the protein−
ligand complex shown in the crystal structure. With a good
understanding of the binding mode in hand, we sought next to
probe the structure−activity relationships in different subpock-
ets of the binding site using a series of peptidic inhibitors.
The side chain of the ligand alanine (−2) position is buried
in a small hydrophobic pocket of CBX7 that is an important
determinant of protein−protein interaction specificity in the
different natural complexes of chromodomains that have been
studied. CBX7 (and all Polycomb homologues) have an
exclusive preference for natural binding partners with an Ala
residue at this position.31 We varied the parent FALKme3S
structure by installation of a few small side chains (Table 2).
We found that even an ethyl side chain (Abu) was too big to be
accommodated in this pocket, and so Ala was retained at the
(−2) position for all subsequent CBX7 ligands.
Both the backbone amide NH and the OH side chain of the
serine (+1) residue donate hydrogen bonds to the side chain of
Glu37, again forming one of the primary determinants for
selection of binding partners by CBX7.31 The switch to D-Ser
(23, Table 3) was tolerated without loss in affinity, suggesting
some freedom of movement for the ligand in this pocket. The
A second group of analogs that probe this binding site were
prepared by leaving the Phe residue in place and installing
additional fragments at the N-terminus of the ligand peptide
(Scheme 1c, Table 5). N-terminal benzamides and phenyl-
acetamides of FALKme3S are generally potent inhibitors (54,
−
switch to C-terminal acid forms (Ac-FALKme3S-CO2 ) was
detrimental to binding for peptides built from both D- and L-Ser
stereoisomers. Other non-amino acid substitutions at the Ser
2877
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
compounds against CBX7 (63 and 64) bind with Kd ≈ 200
nM; the most selective compound (64) is ∼10-fold selective
over CBX8. The affinities and selectivities of the two more
selective compounds most likely arise from the amide-linked
aromatic ester motifs of each at their respective N-termini,
which are designed to engage a region in the β-groove of
CBX7/8 where significant structural divergence occurs between
the two otherwise highly similar chromodomains. All other
human CBX protein chromodomains are less similar to CBX7
than CBX8 (see Supporting Information), with the notable
exception of CBX4. The CBX4 chromodomain is 90% similar
to that of CBX7. The only two dissimilar residues in the
chromodomains of these two proteins are far from the binding
site, making selectivity for CBX7 over CBX4 particularly
challenging. Compounds 1 and 64 showed 1.9- and 1.5-fold
selectivity, respectively, for CBX7 over CBX4, and we anticipate
that CBX4 will remain the most significant challenge for efforts
to optimize selectivity of any CBX7 antagonists going forward.
Selectivity over HP1 family chromodomains seems to be less of
a problem, with initial peptide 1 having 21-fold selectivity for
CBX7 over CBX1 (HP1β) and optimized compound 64
showing >2500-fold selectivity, in spite of these two
chromodomains having 71% similarity scores.
Table 3. H3K27me3-CBX7 Disruption by FALKme3S
Analogs Bearing Serine (+1) Replacements
The FP and ITC activity data collectively reveal a few
examples of small, additive impacts of individual changes: for
example, compound 54’s N-terminal p-bromobenzamide
modification improves binding by 7-fold relative to peptide 1,
with further small improvements observed after inclusion of
two additional substitutions at other positions (62). Elsewhere,
structural features that produced differences when applied
individually did not produce differences in final compounds
(e.g., the two C-terminal functionalizations that create 2.5-fold
differences in compounds 23 and 33 result in no difference
between compounds 63 and 64). The general weakness or
absence of additivities is likely due to the inherent flexibility
remaining in the backbones of these peptidic compounds.
CONCLUSION
■
The only currently approved cancer therapies that target
histone modification pathways directly are histone deacetylase
inhibitors (vorinostat and romidepsin). They are relatively
crude, broad spectrum drugs, but their efficacy and promise
have many now creating second-generation agents (a) that are
more target- and disease-specific and/or (b) that inhibit
different families of molecular targets in epigenetic pathways.
The numerous histone reader proteins of the human proteome
offer a diverse set of functional molecular targets that appear to
have great therapeutic potential. The inhibition of methyllysine
reader proteins is still in its infancy, with hundreds of potential
targets and therapeutic hypotheses but only a very small
number of targets yet inhibited by chemical agents. The
peptidic compounds reported here represent the first inhibitors
of any chromodomain. While earlier inhibitors for other classes
of methyllysine reader proteins have come out of high-
throughput screens, we show here that generating relatively
potent submicromolar peptidic leads against reader proteins is
possible even when the reader protein−histone interaction on
which they are initially based is relatively weak. The potential of
CBX proteins as therapeutic targets will almost certainly
depend on cancer and tissue type and is generally an idea that
has elicited significant disagreement in the cancer biology
literature. These peptidic lead compounds are intended to
provide rapid access to studies of the biological effects of CBX
a
IC50 values were determined as the inhibitor concentration required
for the half-dissociation of the complex between CBX7 and FITC-
labeled H3K27me3. Data are mean values of two or three independent
experimental trails. Errors reported are 95% confidence intervals.
55, 56, 59). This family of compounds revealed a broader
tolerance for replacement of the backbone CO group, such
as in sulfonamide 57, which retains <10 μM potency.
Interestingly, the very closely related sulfonamide 58 showed
no activity, revealing a sharp dependence on the ring
substituents in these compounds.
Finally, we prepared a selected set of compounds bearing
multiple well-tolerated modifications to the FALKme3S scaffold
(Figure 3). The competitive FP assay is convenient and
accurate for moderate affinity compounds, but some of these
compounds proved to be potent enough to be below the
theoretical lower limit for accurate IC50 measurements by
competitive fluorescence polarization.55,56 Accordingly, we
subjected the final four compounds, and also a subset of earlier
compounds, to analysis by isothermal titration calorimetry
(ITC) (Table 6). The ITC data indicate some degree of
selectivity for CBX7 over CBX8, in spite of the 88% sequence
similarity between these two chromodomains. The most potent
2878
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
Scheme 1. Synthesis of (a) Compounds Bearing C-Terminal Modifications (See Table 3), (b) Analogs Bearing Phe
Replacements (See Table 4), and (c) Analogs Bearing N-Terminal Additions (See Table 5)
1
576.38, found 576.27. H NMR (D2O, 500 MHz, δ ppm) 7.46−7.35
(m, 5H), 4.37−4.32 (m, 2H), 4.09 (d, 1H, J = 10 Hz), 4.00 (quintet,
1H, J = 5 Hz), 3.73−3.62 (m, 6H), 3.36−3.25 (m, 2H), 3.10 (br s,
10H), 2.16 (sextet, 1H, J = 10 Hz), 1.93−1.70 (m, 5H), 1.69−1.51(m,
6H), 1.48−1.33 (m, 6H), 1.3−1.18 (m, 2H).
protein antagonism, which in turn can help to test and revise
therapeutic hypotheses related to chromodomain-containing
proteins. We will report on these results in due course.
EXPERIMENTAL SECTION
Compound 62. Compound 62 was assembled manually on 2-
chlorotrityl resin. The first amino acid was coupled as above, and the
following three amino acids were attached using standard Fmoc
protocols. After cleavage of Fmoc group on the N-terminus, the resin
was treated with 4-bromobenzoic acid (2 equiv), HBTU (5 equiv),
and DIPEA (10 equiv) for 2 h and then washed with DMF (3 × 5 mL)
and DCM (5 × 5 mL). The peptide was cleaved from the resin by
shaking with 1% TFA in DCM for 5 min, filtering, washing the resin
with DCM (3 × 5 mL), and concentration of the filtrate in vacuo. This
crude peptide was taken up in DMF, and HCTU (5 equiv), DIPEA
(10 equiv), and N-Boc-ethylenediamine were added and stirred at 70
°C overnight. The reaction mixture was diluted with water and EtOAc,
the aqueous layer was extracted with EtOAc (3 × 5 mL), and the
aqueous layer was concentrated in vacuo to yield the crude compound,
which was purified by HPLC as described in Supporting Information.
Compound purity was determined by LC−MS, 98.72% at tR = 21.52
min. LR-ESI−MS (m/z): C37H55N7O5, calculated 756.34/758.34,
found 756.34/758.34. 1H NMR (D2O, 300 MHz, δ ppm), 7.6 (d, 2H, J
= 9 Hz), 7.5 (d, 2H, J = 9 Hz), 4.26−4.16 (m, 1H), 3.96 (d, 2H, J = 12
Hz), 3.52−3.33 (m, 2H), 3.18 (q, 3H, J = 9 Hz), 3.07−3.01 (m, 3H),
2.98 (br s,10H), 1.80−1.61 (m, 4H), 1.60−1.08 (m, 12H).
Compound 63. The peptide FA(cyclopropyl-Gly)Kme3(D-S) was
synthesized on Rink amide resin using CEM Liberty microwave
synthesizer standard Fmoc protocols. The N-terminal deprotected
peptide, on resin, was treated with monomethyl terephthalate (5
equiv) that was preactivated by shaking with HBTU (5 equiv) and
DIPEA (10 equiv) in DMF for 2 min. This solution was added to the
resin and allowed to stir at ambient temperature overnight. The
solution was filtered and resin washed with DMF (3 × 5 mL) and
DCM (3 × 5 mL) and air-dried. The peptide was cleaved from resin
upon treatment with 10 mL 95:2.5:2.5 TFA/H2O/triisopropylsilane
for 2 h. This solution was then concentrated in vacuo and upon
■
Synthesis. Synthesis of all compounds was carried out using
standard Fmoc-based solid-phase peptide synthesis protocols,
modified where necessary as described in the Supporting Information.
Identity and purity of all peptides were determined by HPLC−MS and
were determined to be of >95% purity. Retention times reported are as
determined under the following conditions: Phenomenex Luna C-18
column, 5 μm, 4.6 mm × 250 mm, flow rate of 1 mL/min, gradient
running from 90:10 water (0.1% TFA) and MeCN (0.1% TFA) to
10:90 water (0.1% TFA) and MeCN (0.1% TFA) over 30 min. Final
compounds 61−64 were further characterized by ESI-MS and 1H
NMR as reported below.
Compound 61. Compound 61 was synthesized by using manual
solid phase peptide synthesis on 2-chlorotrityl resin. Coupling of the
first amino acid (Fmoc-Lys(Me)3-OH, 0.6 equiv) was done in 95:5
CH2Cl2/DMF, DIEA (2.4 equiv) as a base with stirring at ambient
temperature for 2.5 h. The next two amino acid attachments were
done by using standard Fmoc protocols. The N-terminal Fmoc group
was deprotected upon treatment with 20% piperidine in DMF (3 × 3
mL × 5 min) and washed with DMF (3 × 5 mL). This resin was
treated with benzoic acid (2 equiv), HBTU (5 equiv), and DIPEA (10
equiv) for 2 h, and the resin was then washed with DMF (3 × 5 mL)
and DCM (3 × 5 mL). The peptide was cleaved from the resin upon
stirring with 1% TFA in DCM for 5 min, filtering, washing with DCM
(5 × 5 mL), and concentration in vacuo. This crude solid was then
dissolved in DMF (5 mL) with HCTU (3 equiv) and DIPEA (10
equiv) and was treated with 2-amino-1,3-propanediol (3 equiv) and
stirred overnight at 70 °C. Water was added (10 mL) and extracted
with EtOAc (3 × 25 mL), and the aqueous layer was evaporated to
yield crude product which was then purified by HPLC as described in
Supporting Information. Compound purity as determined by LC−MS:
97.63% at tR = 16.21 min. LR-ESI−MS (m/z): C30H50N5O6, calculated
2879
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
Table 4. H3K27me3-CBX7 Disruption by FALKme3S
Analogs Bearing Phe (−3) Replacements
Table 5. H3K27me3-CBX7 Disruption by FALKme3S
Analogs Bearing Phe (−3) Replacements
a
IC50 values were determined as the inhibitor concentration required
for the half-dissociation of the complex between CBX7 and FITC-
labeled H3K27me3. Data are mean values of two or three independent
experimental trails. Errors reported are 95% confidence intervals.
treatment with cold diethyl ether yielded crude white precipitate which
was collected by centrifugation. This crude product was then purified
by HPLC (see Supporting Information). Compound purity
determined by LC−MS, purity >99% at tR = 20.8 min. LR-ESI-MS
+
1
(m/z): C40H58N7O9 calculated 780.4, found 780.0. H NMR (D2O,
500 MHz, δ ppm) 8.09 (br d, 2H), 7.76 (br d, 2H) 7.36 (m, 5H), 4.38
(m, 3H), 4.08 (br d, 1H), 3.96 (s, 2H), 3.89 (m, 3H), 3.31 (m, 3H),
3.09 (s, 9H), 2.19 (m, 1H), 1.65 (m, 13H), 1.35 (d, 7.4 Hz, 3H), 1.25
(m, 2H).
Compound 64. Peptide FA(cyclopropyl-Gly)Kme3 was first
synthesized on 2-chlorotrityl resin. Coupling of the first amino acid
(Fmoc-Lys(Me)3-OH, 0.6 equiv) was done in 95:5 CH2Cl2/DMF,
DIEA (2.4 equiv) as a base and stirred at ambient temperature for 2.5
h. Further amino acids attachments, deprotection, and acetylation on
the N-terminus were performed as above. The N-terminal Fmoc group
was deprotected upon treatment with 20% piperidine in DMF (3 × 3
mL × 5 min). While the peptide was still on resin, monomethyl
terephthalate (5 equiv) was preactivated by shaking with HBTU (5
equiv) and DIPEA (10 equiv) in DMF for 2 min. This solution was
added to the resin and allowed to stir at ambient temperature
overnight. The solution was filtered and resin washed with DMF (3 ×
5 mL) and DCM (3 × 5 mL) and air-dried. The peptide was cleaved
from resin upon treatment with 10 mL of 95:2.5:2.5 TFA/H2O/
triisopropylsilane for 2 h. This solution was then concentrated in vacuo
and upon treatment with cold diethyl ether yielded crude white
precipitate that was collected by centrifugation. This crude solid was
then dissolved in DMF (5 mL) with HCTU (3 equiv) and DIPEA (10
a
IC50 values were determined as the inhibitor concentration required
for the half-dissociation of the complex between CBX7 and FITC-
labeled H3K27me3. Data are mean values of two or three independent
experimental trails. Errors reported are 95% confidence intervals.
2880
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
Figure 3. Final set of peptidic CBX7 inhibitors bearing multiple functional group replacements.
background of the blank buffer, giving millipolarization (mP) values
for each reading that were determined and normalized to percentage
of complex formed. Values were graphed using XLfit (IDBS) and fitted
using a sigmoidal curve function from which IC50 values were
extrapolated and standard error was derived. Experiments were
performed in triplicate, and the IC50 values reported are the averages
of all values.
Isothermal Titration Calorimetry. Isothermal titration calorim-
etry was carried out on a VP-ITC (Microcal, Inc.) at 298 K in Tris-
buffered water (20 mM Tris, 250 mM NaCl, 1 mM DTT, pH 8).
Inhibitor concentration in the syringe was 0.75−1.0 mM, which was
titrated into a sample in matched buffer of protein at a concentration
of 50−70 μM. After subtraction of background heats of dilution, curve
fitting was carried out using “one-set-of-sites” binding model of the
manufacturer’s supplied Origin software. See Supporting Information
for raw and fitted data.
Table 6. Kd Values (μM) Arising from ITC Titration of
Selected Compounds into Chromodomains of CBX7, CBX8,
a
CBX4, and CBX1
compd
CBX7
CBX8
CBX4
CBX1
1
2.00 0.20
1.77 0.12
5.4 0.9
14.2 2.0
12.3 1.4
15.0 2.8
1.5 0.3
3.7 0.5
43 2.8
nd
11
23
54
61
62
63
64
nd
nd
nd
0.28 0.05
4.1 0.2
nd
nd
15.0 3.0
0.48 0.04
1.78 0.25
1.89 0.16
nd
nd
0.22 0.02
0.20 0.03
0.20 0.04
nd
nd
nd
nd
0.29 0.02
>500
a
Averaged values determined by duplicate ITC titrations at 298 K in
buffered H2O (20 mM Tris, 250 mM NaCl, 1 mM DTT, pH 8). See
Supporting Information for experimental details, thermodynamic
parameters, and exemplary raw data.
ASSOCIATED CONTENT
* Supporting Information
■
S
Supplementary experimental procedures for peptide synthesis;
1H NMR spectra for compounds 61−64; experimental
procedures and supplementary figures for X-ray crystallography
and 2D-NMR studies; chromodomain sequences, alignments,
and structural comparisons; raw and fitted ITC data. This
material is available free of charge via the Internet at http://
pubs.acs.org.
equiv) and was treated with 2-amino-1,3-propanediol (3 equiv) and
stirred overnight at room temperature. Water was added (10 mL) and
extracted with ethyl acetate (3 × 25 mL), and the aqueous layer was
evaporated to yield crude product, which was purified by HPLC (see
Supporting Information). Compound purity determined by LC−MS
+
>99% at tR = 21.0 min. LR-ESI-MS (m/z): C40H59N6O9 calculated
767.4, found 767.9. 1H NMR (D2O, 500 MHz, δ ppm), 8.1 (d, 7.8 Hz,
2H), 7.77 (d, 7.8 Hz, 2H), 7.35 (m, 5H), 4.34 (m, 2H), 4.05 (d, 7.9
Hz, 1H), 3.96 (s + underlying m, 4H), 3.31 (m, 3H), 3.18 (m, 1H),
3.09 (s + s, 9H), 2.18 (q, 8 Hz, 9 Hz, 1H), 1.70 (m, 12H), 1.35 (d, 6.4
Hz, 3H), 1.25 (m, 2H).
Accession Codes
The coordinates of the complex between CBX7 and ligand 11
have been deposited in the Protein Data Bank (PDB code
4MN3).
Protein Expression and Fluorescence Polarization Assay for
CBX7-H3K27me3 Disruption. Chromodomains were expressed and
purified as previously reported,31 using Addgene plasmids 25241
(CBX7), 25245 (CBX1), and 25237 (CBX4) deposited by C.
Arrowsmith, Structural Genomics Consortium, Toronto, Canada,
and an analogous plasmid for CBX8 generously donated directly to us
by C. Arrowsmith. The conditions for competitive FP assay were
adapted from those used in an earlier report of a direct protein-into-
peptide titration.31 The peptide FITC-H3K27me3, BSA, and inhibitors
were resuspended in distilled H2O. The assay was performed in black,
96-well plates (NUNC) in a buffer containing 20 mM Tris-HCl, pH
8.0, 250 mM NaCl, 1 mM DTT, 1 mM benzamidine, 1 mM PMSF,
0.01% Tween. Constant concentrations of CBX7 and FITC-K27me3
were 8.68 μM and 500 nM, respectively. Inhibitor concentrations
varied from 0 to 2 mM, and all wells were made up to a final volume of
100 μL. Plates were incubated for 15 min in darkness prior to reading
AUTHOR INFORMATION
Corresponding Author
*Phone: 1-250-721-7193. E-mail: fhof@uvic.ca.
■
Author Contributions
⊥C.S. and K.D.D. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by the West Coast Ride to Live
Grants for Prostate Cancer Research and a Grant from Prostate
Cancer Canada. K.D.D. acknowledges fellowship support from
the West Coast Ride to LiveVancouver Island and from the
Prostate Cancer Foundation of British Columbia. M.J.B.,
with a SpectraMax M5 plate reader (Molecular Devices) with λexc
=
450 nm, λobs = 530 nm, and an instrument cutoff of 515 nm. The
parallel and perpendicular intensities of emission were adjusted for
2881
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
(10) 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.; Richon, V. M.; Pollock, R. M.; Kuntz, K. W.;
Keilhack, H. Durable tumor regression in genetically altered malignant
rhabdoid tumors by inhibition of methyltransferase EZH2. Proc. Natl.
Acad. Sci. U.S.A. 2013, 110, 7922−7927.
J.E.W., and F.H. are Scholars of the Michael Smith Foundation
for Health Research and Canada Research Chairs. We thank O.
Granot and C. Barr for expert assistance with MS and NMR
instrumentation and experiments. We thank C. Arrowsmith,
Structural Genomics Consortium Toronto, for contributing the
chromodomain plasmids.
(11) Wigle, T. J.; Copeland, R. A. Drugging the human methylome:
an emerging modality for reversible control of aberrant gene
transcription. Curr. Opin. Chem. Biol. 2013, 17, 369−378.
(12) Yao, Y.; Chen, P.; Diao, J.; Cheng, G.; Deng, L.; Anglin, J. L.;
Prasad, B. V.; Song, Y. Selective inhibitors of histone methyltransferase
DOT1L: design, synthesis, and crystallographic studies. J. Am. Chem.
Soc. 2011, 133, 16746−16749.
ABBREVIATIONS USED
■
Abu, 2-aminobutyric acid; BSA, bovine serum albumin; CBX,
chromobox homolog; DCM, dichloromethane; DIPEA, ethyl-
diisopropylamine; EZH2, enhancer of zeste homolog 2; FP,
fluorescence polarization; H3K27me3, histone 3, lysine 27
trimethylated; HBTU, 1H-benzotriazol-1-yl[bis-
(dimethylamino)methylene]oxonium hexafluorophospate;
HCTU, [(6-chloro-1H-benzotriazol-1-yl)oxy]-
(dimethylamino)-N,N-dimethyliminium hexafluorophospate;
ITC, isothermal titration calorimetry; Nle, norleucine; 3-PAL,
3-pyridylalanine; Phg, phenylglycine; PRC1/2, polycomb
repressive complex 1/2; SETDB1, histone-lysine N-methyl-
transferase SETDB1 (KMT1E)
(13) Marks, P. A. Discovery and development of SAHA as an
anticancer agent. Oncogene 2007, 26, 1351−1356.
(14) Stimson, L.; Wood, V.; Khan, O.; Fotheringham, S.; La
Thangue, N. B. HDAC inhibitor-based therapies and haematological
malignancy. Ann. Oncol. 2009, 20, 1293−1302.
(15) New, M.; Olzscha, H.; La Thangue, N. B. HDAC inhibitor-
based therapies: Can we interpret the code? Mol. Oncol. 2012, 6, 637−
656.
(16) Matzuk, M. M.; McKeown, M. R.; Filippakopoulos, P.; Li, Q.;
Ma, L.; Agno, J. E.; Lemieux, M. E.; Picaud, S.; Yu, R. N.; Qi, J.;
Knapp, S.; Bradner, J. E. Small-molecule inhibition of BRDT for male
contraception. Cell 2012, 150, 673−684.
(17) Banerjee, C.; Archin, N.; Michaels, D.; Belkina, A. C.; Denis, G.
V.; Bradner, J.; Sebastiani, P.; Margolis, D. M.; Montano, M. BET
bromodomain inhibition as a novel strategy for reactivation of HIV-1.
J. Leukocyte Biol. 2012, 92, 1147−1154.
REFERENCES
■
(1) Strahl, B. D.; Allis, C. D. The language of covalent histone
modifications. Nature 2000, 403, 41−45.
(2) de la Cruz, X.; Lois, S.; Sanchez-Molina, S.; Martinez-Balbas, M.
A. Do protein motifs read the histone code? BioEssays 2005, 27, 164−
175.
(18) Filippakopoulos, P.; Qi, J.; Picaud, S.; Shen, Y.; Smith, W. B.;
Fedorov, O.; Morse, E. M.; Keates, T.; Hickman, T. T.; Felletar, I.;
Philpott, M.; Munro, S.; McKeown, M. R.; Wang, Y.; Christie, A. L.;
West, N.; Cameron, M. J.; Schwartz, B.; Heightman, T. D.; La
Thangue, N.; French, C. A.; Wiest, O.; Kung, A. L.; Knapp, S.;
Bradner, J. E. Selective inhibition of BET bromodomains. Nature
2010, 468, 1067−1073.
(19) Loven, J.; Hoke, H. A.; Lin, C. Y.; Lau, A.; Orlando, D. A.;
Vakoc, C. R.; Bradner, J. E.; Lee, T. I.; Young, R. A. Selective inhibition
of tumor oncogenes by disruption of super-enhancers. Cell 2013, 153,
320−334.
(20) Liu, L.; Zhen, X. T.; Denton, E.; Marsden, B. D.; Schapira, M.
ChromoHub: a data hub for navigators of chromatin-mediated
signalling. Bioinformatics 2012, 28, 2205−2206.
(21) Yap, K. L.; Zhou, M. M. Structure and mechanisms of lysine
methylation recognition by the chromodomain in gene transcription.
Biochemistry 2011, 50, 1966−1980.
(22) Santiago, C.; Nguyen, K.; Schapira, M. Druggability of methyl-
lysine binding sites. J. Comput.-Aided. Mol. Des. 2011, 25, 1171−1178.
(23) Kireev, D.; Wigle, T. J.; Norris-Drouin, J.; Herold, J. M.; Janzen,
W. P.; Frye, S. V. Identification of non-peptide malignant brain tumor
(MBT) repeat antagonists by virtual screening of commercially
available compounds. J. Med. Chem. 2010, 53, 7625−7631.
(24) Gao, C.; Herold, J. M.; Kireev, D.; Wigle, T.; Norris, J. L.; Frye,
S. Biophysical probes reveal a “compromise” nature of the methyl-
lysine binding pocket in L3MBTL1. J. Am. Chem. Soc. 2011, 133,
5357−5362.
(25) Herold, J. M.; Wigle, T. J.; Norris, J. L.; Lam, R.; Korboukh, V.
K.; Gao, C.; Ingerman, L. A.; Kireev, D. B.; Senisterra, G.; Vedadi, M.;
Tripathy, A.; Brown, P. J.; Arrowsmith, C. H.; Jin, J.; Janzen, W. P.;
Frye, S. V. Small-molecule ligands of methyl-lysine binding proteins. J.
Med. Chem. 2011, 54, 2504−2511.
(3) Migliori, V.; Muller, J.; Phalke, S.; Low, D.; Bezzi, M.; Mok, W.
C.; Sahu, S. K.; Gunaratne, J.; Capasso, P.; Bassi, C.; Cecatiello, V.; De
Marco, A.; Blackstock, W.; Kuznetsov, V.; Amati, B.; Mapelli, M.;
Guccione, E. Symmetric dimethylation of H3R2 is a newly identified
histone mark that supports euchromatin maintenance. Nat. Struct. Mol.
Biol. 2012, 19, 136−144.
(4) Taverna, S. D.; Li, H.; Ruthenburg, A. J.; Allis, C. D.; Patel, D. J.
How chromatin-binding modules interpret histone modifications:
lessons from professional pocket pickers. Nat. Struct. Mol. Biol. 2007,
14, 1025−1040.
(5) Daigle, S. R.; Olhava, E. J.; Therkelsen, C. A.; Basavapathruni, A.;
Jin, L.; Boriack-Sjodin, P. A.; Allain, C. J.; Klaus, C. R.; Raimondi, A.;
Scott, M. P.; Waters, N. J.; Chesworth, R.; Moyer, M. P.; Copeland, R.
A.; Richon, V. M.; Pollock, R. M. Potent inhibition of DOT1L as
treatment of MLL-fusion leukemia. Blood 2013, 122, 1017−1025.
(6) Anglin, J. L.; Song, Y. A medicinal chemistry perspective for
targeting histone H3 lysine-79 methyltransferase DOT1L. J. Med.
Chem. 2013, 56, 8972−8983.
(7) Takawa, M.; Masuda, K.; Kunizaki, M.; Daigo, Y.; Takagi, K.;
Iwai, Y.; Cho, H.-S.; Toyokawa, G.; Yamane, Y.; Maejima, K.; Field, H.
I.; Kobayashi, T.; Akasu, T.; Sugiyama, M.; Tsuchiya, E.; Atomi, Y.;
Ponder, B. A. J.; Nakamura, Y.; Hamamoto, R. Validation of the
histone methyltransferase EZH2 as a therapeutic target for various
types of human cancer and as a prognostic marker. Cancer Sci. 2011,
102, 1298−1305.
(8) 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.; LaFrance, L. V.; Mellinger, M.; Duquenne, C.; Tian,
X.; Kruger, R. G.; McHugh, C. F.; Brandt, M.; Miller, W. H.; Dhanak,
D.; Verma, S. K.; Tummino, P. J.; Creasy, C. L. EZH2 inhibition as a
therapeutic strategy for lymphoma with EZH2-activating mutations.
Nature 2012, 492, 108−112.
(26) James, L. I.; Korboukh, V. K.; Krichevsky, L.; Baughman, B. M.;
Herold, J. M.; Norris, J. L.; Jin, J.; Kireev, D. B.; Janzen, W. P.;
Arrowsmith, C. H.; Frye, S. V. Small-molecule ligands of methyl-lysine
binding proteins: optimization of selectivity for L3MBTL3. J. Med.
Chem. 2013, 56, 7358−7371.
(27) Camerino, M. A.; Zhong, N.; Dong, A.; Dickson, B.; James, L.;
Baughman, B.; Norris, J.; Kireev, D.; Janzen, W. P.; Arrowsmith, C.;
Frye, S. V. The structure−activity relationships of L3MBTL3
(9) Verma, S. K.; Tian, X. R.; LaFrance, L. V.; Duquenne, C.; Suarez,
D. P.; Newlander, K. A.; Romeril, S. P.; Burgess, J. L.; Grant, S. W.;
Brackley, J. A.; Graves, A. P.; Scherzer, D. A.; Shu, A.; Thompson, C.;
Ott, H. M.; Van Aller, G. S.; Machutta, C. A.; Diaz, E.; Jiang, Y.;
Johnson, N. W.; Knight, S. D.; Kruger, R. G.; McCabe, M. T.; Dhanak,
D.; Tummino, P. J.; Creasy, C. L.; Miller, W. H. Identification of
potent, selective, cell-active inhibitors of the histone lysine
methyltransferase EZH2. ACS Med. Chem. Lett. 2012, 3, 1091−1096.
2882
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883

Journal of Medicinal Chemistry
Article
inhibitors: a second series of potent compounds which bind the
L3MBTL3 dimer. MedChemComm 2013, 4, 1501−1507.
(28) 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.; Marcon, E.; Guo, H. B.; Gao, C.; Huang, X. P.;
Duan, S. L.; Emili, A.; Greenblatt, J. F.; Kireev, D. B.; Jin, J.; Janzen, W.
P.; Brown, P. J.; Bedford, M. T.; Arrowsmith, C. H.; Frye, S. V.
Discovery of a chemical probe for the L3MBTL3 methyllysine reader
domain. Nat. Chem. Biol. 2013, 9, 184−191.
(29) Wagner, E. K.; Nath, N.; Flemming, R.; Feltenberger, J. B.;
Denu, J. M. Identification and characterization of small molecule
inhibitors of a plant homeodomain finger. Biochemistry 2012, 51,
8293−8306.
(30) Eissenberg, J. C. Structural biology of the chromodomain: form
and function. Gene 2012, 496, 69−78.
(31) Kaustov, L.; Ouyang, H.; Amaya, M.; Lemak, A.; Nady, N.;
Duan, S.; Wasney, G. A.; Li, Z.; Vedadi, M.; Schapira, M.; Min, J.;
Arrowsmith, C. H. Recognition and specificity determinants of the
human cbx chromodomains. J. Biol. Chem. 2011, 286, 521−529.
(32) Blus, B. J.; Wiggins, K.; Khorasanizadeh, S. Epigenetic virtues of
chromodomains. Crit. Rev. Biochem. Mol. Biol. 2011, 46, 507−526.
(33) Gieni, R. S.; Hendzel, M. J. Polycomb group protein gene
silencing, non-coding RNA, stem cells, and cancer. Biochem. Cell Biol.
2009, 87, 711−746.
(34) Bracken, A. P.; Helin, K. Polycomb group proteins: navigators of
lineage pathways led astray in cancer. Nat. Rev. Cancer 2009, 9, 773−
784.
(35) Sauvageau, M.; Sauvageau, G. Polycomb group proteins: multi-
faceted regulators of somatic stem cells and cancer. Cell Stem Cell
2010, 7, 299−313.
pathway regulation. Int. J. Cancer [Online early access]. DOI:
10.1002/ijc.28692. Published Online: Jan 10, 2014.
(45) Wyatt, A.; Collins, C. Vancouver Prostate Centre, personal
communication.
(46) Klauke, K.; Radulovic, V.; Broekhuis, M.; Weersing, E.; Zwart,
E.; Olthof, S.; Ritsema, M.; Bruggeman, S.; Wu, X.; Helin, K.;
Bystrykh, L.; de Haan, G. Polycomb Cbx family members mediate the
balance between haematopoietic stem cell self-renewal and differ-
entiation. Nat. Cell Biol. 2013, 15, 353−362.
(47) Perry, A. S. Prostate cancer epigenomics. J. Urol. 2013, 189, 10−
11.
(48) Pallante, P.; Terracciano, L.; Carafa, V.; Schneider, S.; Zlobec, I.;
Lugli, A.; Bianco, M.; Ferraro, A.; Sacchetti, S.; Troncone, G.; Fusco,
A.; Tornillo, L. The loss of the CBX7 gene expression represents an
adverse prognostic marker for survival of colon carcinoma patients.
Eur. J. Cancer 2010, 46, 2304−2313.
(49) Pallante, P.; Federico, A.; Berlingieri, M. T.; Bianco, M.; Ferraro,
A.; Forzati, F.; Iaccarino, A.; Russo, M.; Pierantoni, G. M.; Leone, V.;
Sacchetti, S.; Troncone, G.; Santoro, M.; Fusco, A. Loss of the CBX7
gene expression correlates with a highly malignant phenotype in
thyroid cancer. Cancer Res. 2008, 68, 6770−6778.
(50) Federico, A.; Pallante, P.; Bianco, M.; Ferraro, A.; Esposito, F.;
Monti, M.; Cozzolino, M.; Keller, S.; Fedele, M.; Leone, V.; Troncone,
G.; Chiariotti, L.; Pucci, P.; Fusco, A. Chromobox protein homologue
7 protein, with decreased expression in human carcinomas, positively
regulates E-cadherin expression by interacting with the histone
deacetylase 2 protein. Cancer Res. 2009, 69, 7079−7087.
(51) Forzati, F.; Federico, A.; Pallante, P.; Abbate, A.; Esposito, F.;
Malapelle, U.; Sepe, R.; Palma, G.; Troncone, G.; Scarfo, M.; Arra, C.;
Fedele, M.; Fusco, A. CBX7 is a tumor suppressor in mice and
humans. J. Clin. Invest. 2012, 122, 612−623.
(36) Morey, L.; Pascual, G.; Cozzuto, L.; Roma, G.; Wutz, A.;
Benitah, S. A.; Di Croce, L. Nonoverlapping functions of the
Polycomb group Cbx family of proteins in embryonic stem cells.
Cell Stem Cell 2012, 10, 47−62.
(37) O’Loghlen, A.; Munoz-Cabello, A. M.; Gaspar-Maia, A.; Wu, H.
A.; Banito, A.; Kunowska, N.; Racek, T.; Pemberton, H. N.; Beolchi,
P.; Lavial, F.; Masui, O.; Vermeulen, M.; Carroll, T.; Graumann, J.;
Heard, E.; Dillon, N.; Azuara, V.; Snijders, A. P.; Peters, G.; Bernstein,
E.; Gil, J. MicroRNA regulation of Cbx7 mediates a switch of
Polycomb orthologs during ESC differentiation. Cell Stem Cell 2012,
10, 33−46.
́
(52) Klauke, K.; Radulovic, V.; Broekhuis, M.; Weersing, E.; Zwart,
E.; Olthof, S.; Ritsema, M.; Bruggeman, S.; Wu, X.; Helin, K.;
Bystrykh, L.; de Haan, G. Polycomb Cbx family members mediate the
balance between haematopoietic stem cell self-renewal and differ-
entiation. Nat. Cell Biol. 2013, 15, 353−362.
(53) qHTS Validation Assay for Inhibitors of HP1-beta Chromodo-
main Interactions with Methylated Histone Tails. http://pubchem.
ncbi.nlm.nih.gov/assay/assay.cgi?aid=488953.
(54) Completing the coordination of the Mg2+ are a symmetry
related Glu/His pair from another monomer. This suggests indirectly
that multimerization, either with another copy of CBX7 or with
another member of the PRC1 complex, might play a role in stabilizing
the bound metal in vivo.
(38) Gil, J.; Bernard, D.; Martinez, D.; Beach, D. Polycomb CBX7
has a unifying role in cellular lifespan. Nat. Cell Biol. 2004, 6, 67−72.
(39) Bernard, D.; Martinez-Leal, J. F.; Rizzo, S.; Martinez, D.;
Hudson, D.; Visakorpi, T.; Peters, G.; Carnero, A.; Beach, D.; Gil, J.
CBX7 controls the growth of normal and tumor-derived prostate cells
by repressing the Ink4a/Arf locus. Oncogene 2005, 24, 5543−5551.
(55) Huang, X. Fluorescence polarization competition assay: the
range of resolvable inhibitor potency is limited by the affinity of the
fluorescent ligand. J. Biomol. Screening 2003, 8, 34−38.
(56) The lower limit for any IC50 determined by competitive
fluorescence polarization assay is approximately the Kd value for the
dissociation of probe peptide from protein target. This value for CBX7
and the H3K27me3 probe peptide is 7 μM under our experimental
conditions, and so compounds with greater potency than Kd ≈ 7 μM
generate IC50 values that are only marginally lower than this regardless
of how potent they are (see ref 55 for a detailed explanation). Direct
comparison of IC50 values determined by competitive FP and Kd
values determined by ITC confirms the existence of the lower limit for
competitive FP measurements. For example, compound 54 has an IC50
value measured at 4.6 μM by competitive FP but is shown to have a
true Kd of 280 nM when measured by ITC.
(40) Yap, K. L.; Li, S.; Munoz-Cabello, A. M.; Raguz, S.; Zeng, L.;
̃
Mujtaba, S.; Gil, J.; Walsh, M. J.; Zhou, M.-M. Molecular interplay of
the noncoding RNA ANRIL and methylated histone H3 lysine 27 by
Polycomb CBX7 in transcriptional silencing of INK4a. Mol. Cell 2010,
38, 662−674.
(41) NCT01897571 information available on http://www.
clinicaltrials.gov.
(42) van den Boom, V.; Rozenveld-Geugien, M.; Bonardi, F.;
Malanga, D.; van Gosliga, D.; Heijink, A. M.; Viglietto, G.; Morrone,
G.; Fusetti, F.; Vellenga, E.; Schuringa, J. J. Non-redundant and locus-
specific gene repression functions of PRC1 paralog family members in
human hematopoietic stem/progenitor cells. Blood 2013, 121, 2452−
2461.
(43) Scott, C. L.; Gil, J.; Hernando, E.; Teruya-Feldstein, J.; Narita,
M.; Martinez, D.; Visakorpi, T.; Mu, D.; Cordon-Cardo, C.; Peters, G.;
Beach, D.; Lowe, S. W. Role of the chromobox protein CBX7 in
lymphomagenesis. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 5389−5394.
(44) Shinjo, K.; Yamashita, Y.; Yamamoto, E.; Akatsuka, S.; Uno, N.;
Kamiya, A.; Niimi, K.; Sakaguchi, Y.; Nagasaka, T.; Takahashi, T.;
Shibata, K.; Kajiyama, H.; Kikkawa, F.; Toyokuni, S. Expression of
chromobox homolog 7 (CBX7) is associated with poor prognosis in
ovarian clear cell adenocarcinoma via TRAIL-induced apoptotic
2883
dx.doi.org/10.1021/jm401487x | J. Med. Chem. 2014, 57, 2874−2883