Assessment of Bromodomain Target Engagement by a Series of BI2536 Analogues with Miniaturized BET-BRET

Article abstract of DOI:10.1002/cmdc.201600502

Evaluating the engagement of a small molecule ligand with a protein target in cells provides useful information for chemical probe optimization and pharmaceutical development. While several techniques exist that can be performed in a low-throughput manner, systematic evaluation of large compound libraries remains a challenge. In-cell engagement measurements are especially useful when evaluating compound classes suspected to target multiple cellular factors. In this study we used a bioluminescent resonant energy transfer assay to assess bromodomain engagement by a compound series containing bromodomain- and kinase-biasing polypharmacophores based on the known dual BRD4 bromodomain/PLK1 kinase inhibitor BI2536. With this assay, we discovered several novel agents with bromodomain-selective specificity profiles and cellular activity. Thus, this platform aids in distinguishing molecules whose cellular activity is difficult to assess due to polypharmacologic effects.

Full text of DOI:10.1002/cmdc.201600502

DOI: 10.1002/cmdc.201600502
Communications
Assessment of Bromodomain Target Engagement by
a Series of BI2536 Analogues with Miniaturized BET-BRET
Luke W. Koblan+,[c] Dennis L. Buckley+,[a, b] Christopher J. Ott+,[a, b, c] Mark E. Fitzgerald+,[c, h]
Stuart W. J. Ember,[d, g] Jin-Yi Zhu,[d] Shuai Liu,[e] Justin M. Roberts,[a] David Remillard,[a]
Sarah Vittori,[a, c] Wei Zhang,[e] Ernst Schonbrunn,[d] and James E. Bradner*[a, b, c, f]
Evaluating the engagement of a small molecule ligand with
a protein target in cells provides useful information for chemi-
cal probe optimization and pharmaceutical development.
While several techniques exist that can be performed in a low-
throughput manner, systematic evaluation of large compound
libraries remains a challenge. In-cell engagement measure-
ments are especially useful when evaluating compound classes
suspected to target multiple cellular factors. In this study we
used a bioluminescent resonant energy transfer assay to assess
bromodomain engagement by a compound series containing
bromodomain- and kinase-biasing polypharmacophores based
on the known dual BRD4 bromodomain/PLK1 kinase inhibitor
BI2536. With this assay, we discovered several novel agents
with bromodomain-selective specificity profiles and cellular ac-
tivity. Thus, this platform aids in distinguishing molecules
whose cellular activity is difficult to assess due to polypharma-
cologic effects.
dues. Owing to their structurally distinct hydrophobic acetylly-
sine recognition pocket, bromodomains have proven tractable
targets for synthetic small-molecule ligand development.[1] The
first reported bromodomain inhibitors were shown to potently
and selectively displace the bromodomain-and-extra terminal
domain (BET) family bromodomains from acetyllysine-bearing
histone peptides in vitro and from regions of acetylated chro-
matin in cells.[2–4] The BET bromodomain protein family has
four members—BRD2, BRD3, BRD4, and BRDT—each involved
in a number of biological processes including epigenetic regu-
lation and control of gene expression. Since the original publi-
cation of BET bromodomain ligands, a variety of inhibitors
with diverse scaffolds have been developed with several being
studied in clinical trials for indications such as cancer and car-
diovascular disease.[5–9] Pre-clinical development of these inhib-
itors has been enabled by a number of cell-free in vitro assays
for compound hit discovery and lead optimization; however,
the typical assays for assessing cellular compound activity, in-
cluding fluorescence recovery after photo-bleaching (FRAP)
and chromatin immunoprecipitation (ChIP), are much lower
throughput and can suffer from poor dynamic range. In fact,
general and robust high-throughput methods to determine
ligand association with target proteins in live cells are not thor-
oughly used. Fluorescence or bioluminescence resonance
energy transfer (FRET/BRET) systems with paired energy
donors and acceptors can be used to detect protein–protein
and protein–ligand interactions, though it can be difficult to
achieve the necessary sensitivity and dynamic range required
for screening assays. Here we sought to evaluate the use of
a cell-permeable fluorophore-tagged BET bromodomain ligand
for use in a miniaturized bioluminescence resonance energy
transfer (BET-BRET) assay. This assay is based on technology re-
cently developed by the Promega Corporation whereby the
use of nanoluciferase (Nluc) as a bioluminescent BRET donor
can substantially increase assay signals in live-cell protein–pro-
tein interaction studies, including BRD4 binding to histones.[10]
We used this strategy for a simplified bromodomain ligand-dis-
placement assay system by appending a fluorescent acceptor
nonchloro TOM (NCT) dye onto an accessible site of JQ1, a de-
finitive BET bromodomain probe ligand (Figure 1A,B).[2] The
NCT dye has minimal spectral overlap with Nluc, and has re-
cently been described as an effective BRET tracer dye when
conjugated to histone deacetylase and bromodomain-binding
small molecule ligands.[11,12] Thus, we considered that adapta-
tion of this assay with a JQ1-conjugate in miniaturized screen-
ing-plate format may be a useful addition to bromodomain in-
Bromodomains are highly conserved protein modules that rec-
ognize post-translationally modified e-acetylated lysine resi-
[a] Dr. D. L. Buckley,+ Dr. C. J. Ott,+ J. M. Roberts, D. Remillard, S. Vittori,
Prof. Dr. J. E. Bradner
Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA
02215 (USA)
E-mail: james_bradner@dfci.harvard.edu
[b] Dr. D. L. Buckley,+ Dr. C. J. Ott,+ Prof. Dr. J. E. Bradner
Department of Medicine, Harvard Medical School, Boston, MA 02115 (USA)
[c] L. W. Koblan,+ Dr. C. J. Ott,+ Dr. M. E. Fitzgerald,+ S. Vittori,
Prof. Dr. J. E. Bradner
Center for the Science of Therapeutics, Broad Institute, Cambridge, MA
02142 (USA)
[d] Dr. S. W. J. Ember, Dr. J.-Y. Zhu, Prof. Dr. E. Schonbrunn
Drug Discovery Department, Moffitt Cancer Center, Tampa, FL 33612 (USA)
[e] S. Liu, Prof. Dr. W. Zhang
Department of Chemistry, University of Massachusetts-Boston, Boston, MA
02125 (USA)
[f] Prof. Dr. J. E. Bradner
Current address: Novartis Institutes for Biomedical Research, Cambridge,
MA 02139 (USA)
[g] Dr. S. W. J. Ember
Current address: Reaction Biology Corporation, Malvern, PA 19355 (USA)
[h] Dr. M. E. Fitzgerald+
Current address: C4 Therapeutics, Cambridge, MA 02142 (USA)
[+] These authors contributed equally to this work.
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under http://dx.doi.org/10.1002/
cmdc.201600502.
ChemMedChem 2016, 11, 1 – 8
1
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
Figure 1. BET-BRET assay design with JQ1-NCT. A) Schematic of the assay.
Nluc-BET fusion proteins are transiently transfected into cells (i); cells are
treated with JQ1-NCT tracer (ii); Nluc substrate is added (iii); and BRET signal
is recorded (iv). B) JQ1-NCT dye structure. C) Spectral characteristics of JQ1-
NCT. Nluc emission as determined in Machleidt et al.[10]
hibitor development to identify JQ1-competitive compounds.
The BET-BRET assay consists of transient expression of full-
length BET proteins amino-terminally tagged with Nluc. The
JQ1-NCT probe is added to live cells, followed by addition of
the Nluc substrate furimazine to generate a BRET signal. Similar
to other NCT-conjugated dye systems described by Wood and
colleagues,[10–12] the JQ1-NCT probe displayed fluorescent prop-
erties that notably achieve 165 nm of spectral emission separa-
tion from the Nluc donor (Figure 1C). To characterize cellular
penetration of the dye and optimal BET-BRET conditions, vary-
ing ratios of transfected Nluc-BRD4 and JQ1-NCT were evaluat-
ed (Figure 2A). Using a 96-well assay format, we found that
250 nm of JQ1-NCT produced the highest signal to background
ratio at all concentrations of transfected Nluc-BRD4 tested. For
subsequent experiments, a ratio of 50 ng Nluc-BET with
250 nm JQ1-NCT was selected as it generated the most consis-
tent data. Using these conditions, JQ1-NCT generated signifi-
cant BRET signal with all four members of the BET protein
family (Figure 2B). Some variation in maximal signal intensity is
Figure 2. Evaluation of BET-BRET assay parameters with JQ1-NCT. A) Determi-
nation of optimal BET-BRET acceptor/donor ratios. A four-point titration of
donor Nluc-BRD4 (full-length) plasmid with four concentrations of acceptor
JQ1-NCT dye. Signal was evaluated as described in the Experimental Section.
The ratio used for subsequent experiments is denoted with a red asterisk.
B) Evaluation of mBRET signal across all BET family members (***p<0.001,
****p<0.0001). C) BET ligands of varying potencies evaluated in the BRD4
BET-BRET assay. Cells were treated at a concentration of 1 mm for 2 h. D) Eval-
uation of JQ1 dose–response in the BRD4 BET-BRET assay. E) 384-well Z’-
factor assessment comparing two concentrations of JQ1 inhibition. Z’-factor
is reported for DMSO versus 10 nm and 1000 nm JQ1. Each condition was
tested with n=40 samples.
&
ChemMedChem 2016, 11, 1 – 8
www.chemmedchem.org
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
observed between individual BET proteins, which may reflect
differences in plasmid-based expression levels between the
assays, or unique cellular contexts for each BET protein. In ex-
periments directly comparing JQ1-NCT BRET signal with that
generated from BRD4–histone interactions, we observe higher
signal and increased dynamic range with competitive inhibi-
tion by JQ1 using the JQ1-NCT tracer (Supporting Information
Figure S1). We then evaluated the capability of BET-BRET to
report on competitive inhibition of the dye-bromodomain in-
teraction using a panel of structurally diverse BET ligands at
a dose of 1 mm (RVX208, UMB32, IBET151, JQ1). Following two
hours of treatment, each inhibitor displays previously reported
trends of in vitro potency against the BET bromodomains; JQ1
displayed the most potent target engagement for most BETs
tested, followed by the dimethyl isoxazole-based scaffolds
IBET151 and UMB32 (Figure 2C).[2,5,7,9] Notably, RVX208 did not
show significant target engagement in most of the assays
tested, which corresponds with in vitro observations of this
compound as a weak BET ligand and reveals our BET-BRET
system most reliably reports on the cellular activity of com-
pounds with sub-micromolar in vitro affinity. Robust dose–re-
sponse characteristics could be obtained with inhibitor treat-
ment and EC50 values that correspond to concentrations
known to effect cancer cell phenotypes with longer treatment
times (Figure 2D). Maximal inhibition of the assay results in
a slightly greater than 50% overall signal inhibition relative to
vehicle control-treated samples. This corresponds to observed
dose–response characteristics of BRD4-BRET systems using his-
tone co-expression.[10] In an effort to use BET-BRET for high-
throughput assessment of larger compound libraries, we evalu-
ated its performance in a 384-well plate format (Figure 2E).
Using JQ1 as a competitive inhibitor of the assay, robust Z’-
factor scores were achieved at moderate inhibitor concentra-
tions.
Figure 3.
A BET-BRET screen of a BI2536 scaffold library. A) Replicate concordance of
two BRD4 BET-BRET assays. A 5 mm treatment of each compound was used
in 384-well format (n=95 compounds). BI2536 is shown along with two
compounds used for further evaluation. Inset shows BI2536 structure with
highlighted regions of derivatization. B) Dose–response activity of BI2536
scaffold library. Highlighted are BI2536 (red line), 2 (blue line), and 7 (black
line). Data are the mean of duplicate experiments. C) mBRET signal (5 mm)
versus AlphaScreen IC50 values for each compound tested. Highlighted are
BI2536, 2, and 7. Shown are 92 of the 95 compounds used in the BET-BRET
screen for which AlphaScreen data were available. Data are the mean of du-
plicate experiments.
While previous studies of BET inhibitors have used cellular
viability assays as a high-throughput surrogate for cellular ac-
tivity, this is less feasible for a recently reported class of com-
pounds that can inhibit both BET proteins and selected kinase
signaling proteins.[13–15] This includes BI2536 that possesses
dual activity against BET proteins and PLK1. Each of these ac-
tivities likely contributes to decreased cellular viability. There-
fore, to assess the cellular inhibition of BRD4 in a high-
throughput fashion, we tested BI2536 and a collection of struc-
turally analogous derivatives predicted to be JQ1-competitive
in the BET-BRET assay. These compounds included derivatives
of BI2536, with varying substituents on the dihydropteridinone
core and off of the central aryl ring. As shown in Figure 3A,
replicate experiments are highly correlated with an R2 value of
0.794. Compounds tested have a range of activities in the
assay, including active BRD4-engaging compounds (e.g.,
BI2536, compound 2), inactive compounds (e.g., compound 7),
and compounds that display apparent artefactual BRET activity.
Active compounds can be seen to have dose-specific effects in
the assay including BI2536 and 2 (Figure 3B). We compared
data obtained from the JQ1-NCT BRET screen with biochemical
AlphaScreen data in an attempt to identify interesting trends.
We observed that the most potent compounds in the BET-
BRET experiments also tended to display superior in vitro inhib-
itory activity (Figure 3C). However, some compounds that dis-
play high potency in the in vitro assay show low or minimal in-
hibition in the BET-BRET assay, perhaps an indication of lower
cell permeability or poor stability in culture media. Thus, the
paired analysis affords the ability to stratify and triage cell-
active, selective BET ligands in a clear and rapid manner.
To confirm the observations made from the BET-BRET screen-
ing, we selected several compounds for further characteriza-
tion by assessing PLK1 inhibitory activity and dose–response
effects on viability of the MV4;11 human leukemia cell line
(Table 1). Included in this set are four compounds recently de-
scribed by Chen et al. in similar in vitro BRD4 and PLK1 activity
and cell viability assays (1/39q, 3/39p, 5/39b, 7/39a).[15] Our
ChemMedChem 2016, 11, 1 – 8
www.chemmedchem.org
3
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
Table 1. Structures and activities of BRD4/PLK1 inhibitors.
Compound
R1
Me
Me
Me
cPent
iPr
Me
4-(THP)
Me
R2
X
BRD4 IC50 [nm][a]
BRET signal [%][a,b]
PLK1 IC50 [nm][a]
MV4;11 IC50 [nm][c]
BI2536
Me
Me
Me
Me
Me
Et
NH
O
142
415
130
112
96
339
175
9171
67.5
66.2
51.8
60.4
55.0
5
>10000
725
4.7–6.0
265–377
184–218
65.0–78.0
31.2–46.5
99.0–127
69.8–79.7
272–337
1 (DB-1-205)
2 (DB-1-264)
3 (DB-1-285)
4 (DB-2-011)
5 (DB-2-059)
6 (DB-2-073)
7 (DB-2-101)
NMe
NH
NH
NH
NH
NH
19
8
20
58
89.5
51.9
Me
H
>100
129
[a] Values are the mean of duplicate measurements. [b] Treatment at 5 mm. [c] Cell viability reported as 95% confidence intervals (n=4).
data here largely corresponds with previously reported activity
trends for these compounds. We also included in our analysis
several compounds with unique activity profiles reported for
the first time here (including 2 and 7 highlighted in Figure 3).
In particular we observe that 2 maintains potent BRD4 activity
with decreased inhibition of PLK1 when compared with
BI2536. This increased selectivity for BRD4 is due to the meth-
ylation of a key NꢀH with an N-methyl group, which forms
a direct hydrogen bond with the hinge region of PLK1,[16] but
establishes a weaker water-mediated hydrogen bond in crystal
structures of BI2536 bound to the first bromodomain of BRD4
(BRD4-1).[13,14] The affinity of 2 for BRD4 is slightly higher than
that of BI2536, and the crystal structure of 2 bound to BRD4-
1 shows that the aryl ring and tail of the inhibitor is oriented
differently from BI2536 (Supporting Information Figure S2),
suggesting a degree of tolerance in this region (Figure 4A). An
orthogonal cellular thermal shift assay (CETSA) probing for
BRD4 was used to validate the cellular BET inhibitory activity of
2 (Figure 4B). Compound 2 stabilized BRD4 in cells, as did JQ1
and BI2536. Notably, 7 did not stabilize BRD4, correlating with
the BET-BRET observations.
Our use of JQ1-NCT as an intracellular BRET tracer suggests
it as a useful tool for the large-scale characterization of cell-
active BET inhibitors. Furthermore, this approach can be used
to dissect the specific activity of molecules suspected or pre-
dicted to have multiple cellular targets. Notably, similar probes
have also been used to resolve intracellular residence time of
ligand target engagement via dynamic monitoring of BRET
signal in live cells.[12] Thus, we believe this approach is general-
izable and can be applied to many protein classes amenable
to Nluc tagging and anticipate this strategy will be expanded
upon and improved as a tool for target-oriented chemical
screening and target engagement measurements.
Experimental Section
Plasmid construction: BET family cDNAs were obtained from the
Harvard Plasmid repository and were PCR-amplified with primers
containing SgfI (forward) and PmeI (reverse) restriction sites. PCR
products were digested using the Flexi enzyme blend (Promega).
PFN31K vector (Promega) was also digested according to the Flexi
digestion protocol. Ligation of the PCR products and digested
PFN31K vector using T4 ligase (Promega) generated Kanamycin se-
lectable vectors containing BET gene products with N-terminal
Nanoluciferase tags. Vectors were transformed into DH5a sub-clon-
ing efficiency E. coli and colonies were screened for correct DNA
sequences.
While both compounds are active in the BET-BRET assay, 2
had increased potency against BRD4 relative to BI2536. Howev-
er it was far less potent in decreasing the viability of MV4;11
cells. These data suggest that the increased activity of BI2536
in MV4;11 cells is due to its effects on PLK1 inhibition rather
than BRD4 inhibition. To confirm this hypothesis, we analyzed
the effects of 2, BI2536, JQ1 and the selective PLK1 inhibitor
GSK461364 on the cell cycle of MV4;11 leukemia cells (Fig-
ure 4C,D). While JQ1 did not have a strong effect on cell cycle,
GSK461364 induced a G2M arrest. Testing each at their cellular
IC50, BI2536 induced a similar G2M arrest, whereas 2 did not.
These data help to confirm that the effects of BI2356 on
MV4;11 viability are primarily due to PLK1 inhibition (as with
GSK461364) while 2 acts as a selective BET inhibitor similar to
JQ1. Thus, the BET-BRET system helps identify the pertinent in-
tracellular targets of small molecules that lead to their broad
cellular effects.
96-well BET-BRET: HEK293T cells were seeded onto six-well plates
at a density of 4ꢁ105 cells per mL, 2 mL per well. After 6 h incuba-
tion (378C, 5% CO2) cells were transfected with 50 ng of PFN31K
vectors containing BET protein cDNAs using Xfect reagent per
manufacturer’s protocol (Clontech). After overnight incubation,
transfected cells were split using phenol-free trypsin and phenol-
free DMEM (Gibco) supplemented with 10% FBS into experimental
aliquots with JQ1-NCT dye and control aliquots with equal volumes
of DMSO. JQ1-NCT concentration was 250 nm unless otherwise in-
dicated. Synthesis of JQ1-NCT is described in the Supporting Infor-
mation. Cells were plated onto white-bottomed 96-well plates (Per-
kinElmer) at a density of 10000 cells per well with five DMSO con-
trol wells and five experimental JQ1-NCT wells per treatment.
Drugs were incubated with cells for 2 h at 378C, 5% CO2 at the in-
&
ChemMedChem 2016, 11, 1 – 8
www.chemmedchem.org
4
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
DMSO control cells. Standard deviation of signal/noise was calcu-
lated as the mean multiplied by the square root of the sum of
each sample’s (with and without JQ1-NCT) standard deviation di-
vided by its mean squared. mBRET (%DMSO) was calculated as raw
mBRET signal normalized to the DMSO signal for the corresponding
sample [Equations (1), (2), and (3)]:
Fluorescence
Luminescence
ð1Þ
*
mBRET ¼
1000
mBRET
ðSampleÞ
ðDMSOÞ
mBRETð% DMSOÞ ¼
ð2Þ
mBRET
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ꢀ
ꢁ
ꢀ
ꢁ
2
2 þ
StDev:
StDev
Mean
ðJQ1ꢀNCTÞ
ðJQ1ꢀNCTÞ
ðDMSOÞ
ðDMSOÞ
*
StDev: Signal ¼ Mean Signal
ðNoise
Þ
ðNoise
Þ
Mean
ð3Þ
384-well BET-BRET: HEK293T cells were transfected and split as de-
scribed above. Cells were plated into white-bottomed 384-well
plates (Nunc) at 10000 cells per well in 50 mL final volume. Drugs
were pinned into 384-well plates at the indicated final concentra-
tions using a 100 nL pinhead on a JANUS workstation from Perki-
nElmer; incubation was performed as described above. NanoGlo
Luciferase Assay Substrate was administered at a 1:750 final dilu-
tion. Plates were read as described above.
96-well histone-based BRET: HEK293T cells were transiently trans-
fected with 25 ng of BRD4 PFN31K vector with N-terminally Nano-
luciferase tag and 2 mg of Histone H3.3-HaloTag fusion vector
(Promega). Transfected cells were split, plated, treated with JQ1,
and BRET was generated using HaloTag NanoBRET ligand per man-
ufacturer’s instructions.
AlphaScreen assays: BRD4 AlphaScreen assays were conducted as
described.[7] Briefly, BRD4 bromodomain (site 1) was bound to Al-
phaScreen emission beads via a (His)6 tag. These beads were treat-
ed with compound for 1 h at room temperature. Subsequent to
treatment, biotinylated-JQ1 was added to the mixture concurrently
with streptavidin coated AlphaScreen donor beads. Emissions were
recorded using an Envision plate reader (PerkinElmer).
Cellular viability: Cell lines were incubated with compounds for
72 h at 378C and ATPlite (PerkinElmer) reagent was used to assess
cell viability per the manufacturer’s protocol. Luminescence was
read on an Envision plate reader.
Figure 4. Structural and cellular evaluation of BET-BRET-selected BRD4 inhibi-
tors. A) Crystal structure of 2 in complex with BRD4 bromodomain 1.
Dashed lines represent hydrogen bond interactions with BRD4 Asn140.
B) CETSA analysis of compounds. MV4;11 cells were treated with JQ1,
BI2536, or 2 for 3 h at 1 mm. Ladder bands represent 250 kDa (upper),
150 kDa (middle), and 100 kDa (lower) markers. C) Cell-cycle profile evalua-
tion of MV4;11 cells treated with indicated compounds (PI=propidium
iodide). D) Quantification of MV4;11 cell-cycle profiles.
Cell-cycle analysis: After treatment, cells were washed with PBS
and then fixed overnight at ꢀ208C in 70% ethanol. Cells were
washed once more with PBS and then incubated at 378C for
20 min in propidium iodide staining solution: 20 mgmLꢀ1 propidi-
um iodide (Life Technologies), 0.1% (v/v) Triton X-100 in PBS, sup-
plemented with 200 mgmLꢀ1 RNase A (Roche). Flow cytometry
analyses were performed on a LSRFortessa X-20 flow cytometer
(BD Biosciences) and all data analyzed with FlowJo software (v10,
Tree Star).
dicated concentrations. Subsequently, cells were incubated with
NanoGlo Luciferase Assay Substrate (Promega, 1:1000 final dilu-
tion) for 15 min at room temperature before samples were read on
an Envision plate reader (PerkinElmer). Donor luminescence was
read at 460 nm and acceptor fluorescence was read at 615 nm;
milliBRET (mBRET) was calculated as the ratio of fluorescent to lu-
minescent signal multiplied by 1000. The mean and standard devi-
ation of each set of five measurements taken per sample are re-
ported. Fold signal-to-noise ratios were calculated for each concen-
tration of BET plasmid transfected. Signal-to-noise ratios were cal-
culated as the ratio of the means of JQ1-NCT treated cells to
BRD4 CETSA: MV4;11 cells were incubated with compounds of in-
terest at 1 mm for 3 h at 378C. Following incubation, samples were
centrifuged at 500 g for 3 min to pellet cellular material. Pellets
were washed with 200 mL PBS and re-centrifuged; 190 mL superna-
tant was removed, and the remaining sample was heated at
47.58C for 3 min and incubated at room temperature for another
3 min. The samples were then resuspended in lysis buffer (50 mm
Tris·HCl pH 7.5, 5% glycerol, 100 mm NaCl, 1.5 mm MgCl2, 0.2%
NP-40) supplemented with protease inhibitor cocktail and lysed by
three cycles of freeze–thaw using liquid nitrogen. Lysed samples
ChemMedChem 2016, 11, 1 – 8
www.chemmedchem.org
5
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Communications
were then spun at 20000 g for 20 min at 48C to pellet; 30 mL su-
pernatant were mixed with gel loading buffer and the gel was run
using BOLT running conditions (Thermo Fisher). Blots were then
transferred with the iBLOT transfer station, blocked for 1 h at room
temperature using Odyssey blocking buffer, and developed using
an anti-BRD4 antibody (Bethyl). A LICOR gel visualizing system was
used for imaging the blots.
grant P30-CA076292). E.S. acknowledges support from the NIH/
NICHD contract HHSN275201300017C. W.Z. and J.E.B. acknowl-
edge US NIH National Cancer Institute (NIH/NCI) grant
(1U54CA156732-05). J.E.B. acknowledges support from the Wil-
liam Lawrence & Blanche Hughes Foundation, the Dana-Farber
Cancer Institute Accelerator Fund, and NIH/NICHD grant U01-
HD076508.
PLK1 activity assay: The enzymatic activities against PLK1 were
tested in Z-Lyte assays with ATP concentrations of KM for each
kinase per manufacturer’s protocol (Thermo Fisher).
Keywords: bromodomain inhibition
·
drug design
·
epigenetics · fluorescence · in-cell target engagement assays ·
luminescence
Characterization of JQ1-NCT probe: Fluorescent and luminescent
characterization of JQ1-TOM probe was done using a TECAN
SAFIRE II plate reader using a 5 mm solution resuspended in PBS.
Fluorescence excitation scanning was conducted between wave-
lengths 500 nm to 650 nm with a 1 nm wavelength step size and
emission readings beginning at 690 nm; gain was set to 140. Ab-
sorbance parameters were obtained by scanning with an excitation
wavelength beginning at 550 nm and a scanning window for emis-
sion from 600 nm to 800 nm. A 1 nm step size was used with gain
set to 140.
[1] P. Filippakopoulos, S. Knapp, Nat. Rev. Drug Discovery 2014, 13, 337–
356.
[2] P. Filippakopoulos, J. Qi, S. Picaud, Y. Shen, W. B. Smith, O. Fedorov,
E. M. Morse, T. Keates, T. T. Hickman, I. Felletar, M. Philpott, S. Munro,
M. R. McKeown, Y. Wang, A. L. Christie, N. West, M. J. Cameron, B.
Schwartz, T. D. Heightman, N. La Thangue, C. A. French, O. Wiest, A. L.
Kung, S. Knapp, J. E. Bradner, Nature 2010, 468, 1067–1073.
[3] J. Zuber, J. Shi, E. Wang, A. R. Rappaport, H. Herrmann, E. A. Sison, D.
Magoon, J. Qi, K. Blatt, M. Wunderlich, M. J. Taylor, C. Johns, A. Chicas,
J. C. Mulloy, S. C. Kogan, P. Brown, P. Valent, J. E. Bradner, S. W. Lowe,
C. R. Vakoc, Nature 2011, 478, 524–528.
[4] E. Nicodeme, K. L. Jeffrey, U. Schaefer, S. Beinke, S. Dewell, C.-W. Chung,
R. Chandwani, I. Marazzi, P. Wilson, H. Coste, J. White, J. Kirilovsky, C. M.
Rice, J. M. Lora, R. K. Prinjha, K. Lee, A. Tarakhovsky, Nature 2010, 468,
1119–1123.
[5] M. A. Dawson, R. K. Prinjha, A. Dittmann, G. Giotopoulos, M. Bantscheff,
W.-I. Chan, S. C. Robson, C.-W. Chung, C. Hopf, M. M. Savitski et al.,
Nature 2011, 478, 529–533.
[6] S. Picaud, D. Da Costa, A. Thanasopoulou, P. Filippakopoulos, P. V. Fish,
M. Philpott, O. Fedorov, P. Brennan, M. E. Bunnage, D. R. Owen, J. E.
Bradner, P. Taniere, B. O’Sullivan, S. Mꢂller, J. Schwaller, T. Stankovic, S.
Knapp, Cancer Res. 2013, 73, 3336–3346.
[7] M. R. McKeown, D. L. Shaw, H. Fu, S. Liu, X. Xu, J. J. Marineau, Y. Huang,
X. Zhang, D. L. Buckley, A. Kadam, Z. Zhang, S. C. Blacklow, J. Qi, W.
Zhang, J. E. Bradner, J. Med. Chem. 2014, 57, 9019–9027.
[8] G. Zhang, S. G. Smith, M.-M. Zhou, Chem. Rev. 2015, 115, 11625–11668.
[9] S. Picaud, C. Wells, I. Felletar, D. Brotherton, S. Martin, P. Savitsky, B.
Diez-Dacal, M. Philpott, C. Bountra, H. Lingard, O. Fedorov, S. Mꢂller,
P. E. Brennan, S. Knapp, P. Filippakopoulos, Proc. Natl. Acad. Sci. USA
2013, 110, 19754–19759.
[10] T. Machleidt, C. C. Woodroofe, M. K. Schwinn, J. Mꢃndez, M. B. Robers, K.
Zimmerman, P. Otto, D. L. Daniels, T. A. Kirkland, K. V. Wood, ACS Chem.
Biol. 2015, 10, 1797–1804.
[11] T. A. Kirkland, M. G. McDougall, S. Dwight (Promega Corporation, Madi-
son, WI) Int. PCT Pub. No. WO2013078244, 2013.
X-ray crystallography: Reagents and compounds for crystallo-
graphic experiments were purchased from Hampton Research
unless otherwise indicated. BRD4-1 was purified as described previ-
ously.[14] Crystals of BRD4-1 were grown by vapor-diffusion in hang-
ing drops using 0.2m (NH4)2SO4, 0.1m Tris·HCl (pH 8.5) and 25%
(w/v) PEG 3350 as precipitant supplemented with 1 mm DB-1-264-
2 and 10% (v/v) DMSO.[14] Crystals were harvested in cryoprotec-
tant (three parts precipitant including 0.5 mm DB-1-264-2 and one
part 100% (v/v) ethylene glycol) and flash frozen in a stream of ni-
trogen gas. X-ray diffraction data were recorded at ꢀ1808C using
CuKa X-rays produced by a Rigaku Micro-Max 007-HF X-ray genera-
tor, focused by mirror optics and equipped with a Rigaku CCD
Saturn 944 system. Data were reduced and scaled with XDS.[17]
PHENIX was employed for phasing and refinement, and model
building was performed using Coot.[18,19] The structure was solved
by molecular replacement using Phaser with the monomer of PDB
ID: 4O7A as the search model.[20] An initial model of the inhibitor
was generated using MarvinSketch (ChemAxon) with ligand re-
straints from eLBOW of the PHENIX suite. All structures were vali-
dated by MolProbity and phenix.model_vs_data.[21,22] Figure were
prepared using PyMOL (Schrçdinger). Atomic coordinates and
structure factors have been deposited in the Protein Data Bank
(PDB) under accession code 5KJ0.
[12] M. B. Robers, M. L. Dart, C. C. Woodroofe, C. A. Zimprich, T. A. Kirkland, T.
Machleidt, K. R. Kupcho, S. Levin, J. R. Hartnett, K. Zimmerman, A. L.
Niles, R. F. Ohana, D. L. Daniels, M. Slater, M. G. Wood, M. Cong, Y.-Q.
Cheng, K. V. Wood, Nat. Commun. 2015, 6, 10091.
[13] P. Ciceri, S. Mꢂller, A. O’Mahony, O. Fedorov, P. Filippakopoulos, J. P.
Hunt, E. A. Lasater, G. Pallares, S. Picaud, C. Wells, S. Martin, L. M. Wo-
dicka, N. P. Shah, D. K. Treiber, S. Knapp, Nat. Chem. Biol. 2014, 10, 305–
312.
[14] S. W. J. Ember, J.-Y. Zhu, S. H. Olesen, M. P. Martin, A. Becker, N. Berndt,
G. I. Georg, E. Schçnbrunn, ACS Chem. Biol. 2014, 9, 1160–1171.
[15] L. Chen, J. L. Yap, M. Yoshioka, M. E. Lanning, R. N. Fountain, M. Raje,
J. A. Scheenstra, J. W. Strovel, S. Fletcher, ACS Med. Chem. Lett. 2015, 6,
764–769.
Acknowledgements
We thank Dr. Ralph Mazitschek for assistance with spectral char-
acterizations, and Dr. Dannette Daniels for helpful advice on
BRET assay development. We also thank Dr. Jennifer Perry for crit-
ical reading of the manuscript. L.W.K. is supported by a US Na-
tional Science Foundation (NSF) GRFP fellowship (DGE1144152)
as well as the US NIH National Institute of General Medical Scien-
ces (NIH/NIGMS) (5T32M095450-04). C.J.O. acknowledges support
from a Fellow Award from the Leukemia and Lymphoma Society
(5667-13) and by an NCI Pathway to Independence Award
(K99A190861). D.L.B. is a Merck Fellow of the Damon Runyon
Cancer Research Foundation (DRG-2196-14). S.W.J.E., E.S., and J.-
Y.Z. thank the Moffitt Chemical Biology Core for use of the pro-
tein crystallography facility (US National Cancer Institute (NCI)
[16] M. Kothe, D. Kohls, S. Low, R. Coli, G. R. Rennie, F. Feru, C. Kuhn, Y.-H.
Ding, Chem. Biol. Drug Des. 2007, 70, 540–546.
[17] W. Kabsch, Acta Crystallogr. Sect. D 2010, 66, 125–132.
[18] P. D. Adams, P. V. Afonine, G. Bunkꢄczi, V. B. Chen, I. W. Davis, N. Echols,
J. J. Headd, L.-W. Hung, G. J. Kapral, R. W. Grosse-Kunstleve, A. J. McCoy,
N. W. Moriarty, R. Oeffner, R. J. Read, D. C. Richardson, J. S. Richardson,
T. C. Terwilliger, P. H. Zwart, Acta Crystallogr. Sect. D 2010, 66, 213–221.
&
ChemMedChem 2016, 11, 1 – 8
www.chemmedchem.org
6
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communications
[19] P. Emsley, B. Lohkamp, W. G. Scott, K. Cowtan, Acta Crystallogr. Sect. D
2010, 66, 486–501.
[20] A. J. McCoy, R. W. Grosse-Kunstleve, P. D. Adams, M. D. Winn, L. C. Storo-
ni, R. J. Read, J. Appl. Crystallogr. 2007, 40, 658–674.
[22] P. V. Afonine, R. W. Grosse-Kunstleve, V. B. Chen, J. J. Headd, N. W. Mor-
iarty, J. S. Richardson, D. C. Richardson, A. Urzhumtsev, P. H. Zwart, P. D.
Adams, J. Appl. Crystallogr. 2010, 43, 669–676.
[21] V. B. Chen, W. B. Arendall, J. J. Headd, D. A. Keedy, R. M. Immormino,
G. J. Kapral, L. W. Murray, J. S. Richardson, D. C. Richardson, Acta Crystal-
logr. Sect. D 2010, 66, 12–21.
Received: October 4, 2016
Revised: October 21, 2016
Published online on && &&, 0000
ChemMedChem 2016, 11, 1 – 8
www.chemmedchem.org
7
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

COMMUNICATIONS
L. W. Koblan, D. L. Buckley, C. J. Ott,
M. E. Fitzgerald, S. W. J. Ember, J.-Y. Zhu,
S. Liu, J. M. Roberts, D. Remillard,
S. Vittori, W. Zhang, E. Schonbrunn,
J. E. Bradner*
&& – &&
Assessment of Bromodomain Target
Engagement by a Series of BI2536
Analogues with Miniaturized BET-BRET
Place your BETs on BRET: Target en-
gagement assays for evaluating small
molecules in cells provide valuable in-
formation for the development of drugs
and probes. Here, a high-throughput
bioluminescence resonance energy
transfer assay is used to study bromo-
domain inhibition. Novel bromodomain-
specific inhibitors are identified from
a series of bromodomain- and kinase-
binding polypharmacophores whose
cellular activity can obscure the perti-
nent biological target of the com-
pounds.
&
ChemMedChem 2016, 11, 1 – 8
www.chemmedchem.org
8
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!