A series of potent crebbp bromodomain ligands reveals an induced-fit pocket stabilized by a cation-π interaction

Article abstract of DOI:10.1002/anie.201402750

The benzoxazinone and dihydroquinoxalinone fragments were employed as novel acetyl lysine mimics in the development of CREBBP bromodomain ligands. While the benzoxazinone series showed low affinity for the CREBBP bromodomain, expansion of the dihydroquino

Full text of DOI:10.1002/anie.201402750

Angewandte
Chemie
DOI: 10.1002/anie.201402750
Ligand Discovery
A Series of Potent CREBBP Bromodomain Ligands Reveals an
Induced-Fit Pocket Stabilized by a Cation–p Interaction**
Timothy P. C. Rooney, Panagis Filippakopoulos, Oleg Fedorov, Sarah Picaud,
Wilian A. Cortopassi, Duncan A. Hay, Sarah Martin, Anthony Tumber, Catherine M. Rogers,
Martin Philpott, Minghua Wang, Amber L. Thompson, Tom D. Heightman, David C. Pryde,
Andrew Cook, Robert S. Paton, Susanne Mꢀller, Stefan Knapp, Paul E. Brennan, and
Stuart J. Conway*
Abstract: The benzoxazinone and dihydroquinoxalinone
fragments were employed as novel acetyl lysine mimics in the
development of CREBBP bromodomain ligands. While the
benzoxazinone series showed low affinity for the CREBBP
bromodomain, expansion of the dihydroquinoxalinone series
resulted in the first potent inhibitors of a bromodomain outside
the BET family. Structural and computational studies reveal
that an internal hydrogen bond stabilizes the protein-bound
conformation of the dihydroquinoxalinone series. The side
chain of this series binds in an induced-fit pocket forming
a cation–p interaction with R1173 of CREBBP. The most
potent compound inhibits binding of CREBBP to chromatin
in U2OS cells.
binding protein (CREBBP, Figure 1A) bromodomain have
been previously identified, they have relatively low binding
affinities or selectivities.[7–12] Here we report the development
S
mall molecules that modulate a disease phenotype in
preclinical assays are important tools in the validation of
putative therapeutic targets.[1] Such compounds have proved
invaluable in elucidating the biology of the acetyl-lysine
(KAc)-binding bromodomain and extra C-terminal domain
(BET) family of bromodomain-containing proteins (BCPs),
which have emerged as important therapeutic targets for
cancer and inflammation.[2–7] To date there are few potent
bromodomain ligands with nanomolar binding affinity for
BCPs that are not part of the BET family. Although ligands
for the cAMP response element binding protein (CREB)
Figure 1. A) The domain architecture of CREBBP.[13] B) The internal
hydrogen bonds of the benzoxazinone and dihydroquinoxalinone series
À
determine the vector of the C N bond. C) The dihydroquinoxalinone
ligands, (R)-1 and (R)-2, bind to the CREBBP BRD in an induced-fit
pocket, forming a cation–p interaction with R1173.
[*] T. P. C. Rooney, W. A. Cortopassi, Dr. A. L. Thompson,
Prof. Dr. R. S. Paton, Prof. Dr. S. J. Conway
[$] Current address: EnVivo Pharmaceuticals
500 Arsenal Street, Watertown, MA 02472 (USA)
Department of Chemistry, Chemistry Research Laboratory
University of Oxford, Mansfield Road, Oxford, OX1 3TA (UK)
E-mail: stuart.conway@chem.ox.ac.uk
[**] We thank Pfizer Neusentis and the EPSRC for a studentship to
T.P.C.R. The authors receive funding from the SGC, a registered
charity (number 1097737) that receives funds from AbbVie,
Boehringer Ingelheim, the Canada Foundation for Innovation, the
Canadian Institutes for Health Research, Genome Canada, Glax-
oSmithKline, Janssen, Lilly Canada, the Novartis Research Foun-
dation, the Ontario Ministry of Economic Development and
Innovation, Pfizer, Takeda, and the Wellcome Trust [092809/Z/10/
Z]. P.F. and S.P. are supported by a Wellcome Trust Career
Development Fellowship (095751/Z/11/Z). W.A.C. is supported by
a Science Without Borders (CAPES) scholarship. We thank Abid
Masood for technical assistance.
Homepage: http://conway.chem.ox.ac.uk/
Dr. P. Filippakopoulos, Dr. O. Fedorov, Dr. S. Picaud, D. A. Hay,
Dr. S. Martin, Dr. A. Tumber, Dr. C. M. Rogers, Dr. M. Philpott,
Dr. M. Wang, Dr. T. D. Heightman,[+] Dr. S. Mꢀller,
Prof. Dr. S. Knapp, Dr. P. E. Brennan
Nuffield Department of Clinical Medicine, Structural Genomics
Consortium, University of Oxford, Old Road Campus Research
Building
Roosevelt Drive, Oxford, OX3 7DQ (UK)
Dr. D. C. Pryde, Dr. A. Cook[$]
Neusentis, The Portway Building
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201402750.
Granta Park, Great Abington, Cambridge, CB21 6GS (UK)
[+] Current address: TDH: Astex Pharmaceuticals
436 Cambridge Science Park, Cambridge, CB4 0QA (UK)
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Table 1: AlphaScreen data for compounds 1, 2, and 5–8.
of a dihydroquinoxalinone-based series of compounds, (R)-
1 and (R)-2 (Figure 1B), that show nanomolar affinities for
the CREBBP bromodomain.
Using an AlphaScreen assay we identified N-methyl-2-
pyrrolidone (NMP) as a weak (IC50 = 1.9 mm) but ligand-
efficient (0.54) CREBBP bromodomain ligand.[14,15] Further
screening of NMP analogues revealed that 3,4-dihydro-3-
methyl-2(1H)-quinazolinone also binds to the CREBBP
bromodomain, mimics the interactions made by KAc, and
forms an additional hydrogen bond between the NH of the
dihydroquinazolinone ring and the carbonyl oxygen atom of
N1168 in CREBBP (Figure S1A in the Supporting Informa-
tion). This fragment was employed in the development of
BET bromodomain ligands by Chung et al.[16] and Fish et al.[5]
However, our previous work[15,17] revealed that this dihydro-
quinazolinone is susceptible to oxidation at the benzylic
position, and consequently we sought to develop a KAc
mimic that addressed this problem. We initially investigated
the benzoxazinone moiety (3, 4, Figure 1B), in which an
oxygen atom replaces the oxidation-prone methylene group.
Synthesis and biological evaluation of benzoxazinone-based
fragments indicated that 8-position substitution was tolerated
(Table S1). Additionally, the high-resolution X-ray crystal
structure of dihydroquinazolinone bound to the CREBBP
bromodomain indicated that modification at this position
would direct substituents into the ZA channel (Figure S1B).
However, at that time we were unaware that this channel is
permanently occupied by water molecules. Ranking of the
results from an in silico screen of approximately 6000
commercially available amines, conjugated to an 8-position
carboxylic acid on the benzoxazinone moiety, identified 50
target compounds. Parallel synthesis and biological evalua-
tion of these compounds (AlphaScreen) showed the most
active compounds 3 and 4 as having IC50 values of 77 mm and
51 mm, respectively, against CREBBP (Table S2).
The low affinity of these compounds for the CREBBP
bromodomain led us to redesign the KAc mimic giving
consideration to two key points. Firstly, the X-ray crystal
structure of dihydroquinazolinone bound to the CREBBP
bromodomain indicated that the benzoxazinone oxygen atom
would be placed ca. 3.5 ꢀ from the backbone carbonyl group
of P1110, potentially causing an unfavorable interaction
(Figure S1A). To overcome this problem, we employed the
dihydroquinoxalinone group as the KAc mimic [(R)-1, (R)-2;
Figure 1B]. In this group, the oxygen atom is replaced with
a secondary amine, and hence any unfavorable interaction
with P1110 will be ameliorated. Secondly, a structure of
unbound (R)-3 determined using high-resolution single-
crystal X-ray diffraction data confirmed that an internal
hydrogen bond [NH···O 2.696(2) ꢀ] is formed between the
amide and the benzoxazinone (Figure 1B and Figure S2).
Therefore, a consequence of the change to the KAc mimic
was the altered nature of the internal hydrogen bond that
exists in both of these molecules. Both enantiomers of
compound 1, with the tetrahydroquinoline side chain from 4
attached to the dihydroquinoxalinone KAc mimic, were
synthesized (Scheme S1 and S4). Pleasingly, compound (R)-
1 showed an increase in CREBBP potency, with IC50 = 758 nm
(Table 1).
[a] Peptide and protein concentration=25 nm. [b] Isothermal titration
calorimetry.
An X-ray crystal structure of (R)-1 bound to the
CREBBP bromodomain revealed that the dihydroquinoxali-
none was acting as the KAc mimic (Figure 2A). The amide of
the dihydroquinoxalinone formed two hydrogen bonds with
N1168, and the methyl group resided at the base of the KAc-
binding pocket (Figure 2A).
The internal hydrogen bond between the dihydroquinox-
alinone and the amide is evident from the conformation of the
ligand (Figure 2A). The amide carbonyl oxygen also interacts
with the bottom ZA channel water molecule. These two
interactions orient the tetrahydroquinoline moiety away from
the water-occupied ZA channel into an induced-fit pocket
below R1173 (Figure 2B). This pocket was not evident in
previous CREBBP bromodomain structures (Figure 2C) and
hence this interaction was impossible to predict in our initial
docking studies. The 3.5–4.6 ꢀ distance between the tetrahy-
droquinoline ring and R1173 (Figure 2B) suggests that ligand
binding is stabilized by a cation–p interaction.[19] To inves-
tigate this hypothesis, and potentially optimize the interac-
tion, we synthesized a range of tetrahydroquinoline deriva-
tives that possessed electron-donating groups on the aromatic
ring (Scheme S4 and S11). AlphaScreen data revealed that
(R)-2, which has a methoxy group at the 7-position of the
tetrahydroquinoline ring, showed enhanced CREBBP bro-
modomain affinity, with IC50 = 323 nm and KD = 390 nm (iso-
thermal titration calorimetry, ITC).
(R)-2 has modest selectivity over BRD4(1) [KD = 1.4 mm
(ITC)] and good selectivity over seven other phylogentically
diverse BCPs (Figure S4). Compound (R)-5, which has a 3,4-
dihydro-2H-benzo[b][1,4]oxazine side chain, also displays
enhanced CREBBP affinity, whereas the two fluorine-con-
taining derivatives (R)-7 and (R)-8 as well as (S)-1 were less
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
Figure 2.
A) Overlaid X-ray crystal structures of compound (R)-1 (PDB code 4NYW, carbon: yellow) and KAc (PDB code 3P1C, carbon: purple)[18] both
bound to the CREBBP bromodomain. The dihydroquinoxalinone moiety forms two hydrogen bonds with N1168 and a water-mediated hydrogen
bond with Y1125. B) The carbonyl group of (R)-1 forms a hydrogen bond with one of the conserved ZA channel water molecules. This contact,
coupled with the internal hydrogen bond, directs the tetrahydroquinoline side chain into an induced-fit pocket under R1173 that is stabilized by
a cation–p interaction. C) R1173 alters its conformation in the KAc-bound structure (PDB code 3P1C, carbon: gray)[18] to accommodate the
tetrahydroquinoline side chain of (R)-1. D) The side chain of (R)-2 (PDB code 4NYX, carbon: orange) moves relative to that of (R)-1 (PDB code
4NYW, carbon: yellow) maximizing the interaction between the positively charged R1173 and the electron-donating methoxy group.
potent (Table 1). Single-crystal X-ray diffraction studies
confirmed that the dihydroquinoxalinone moieties of (R)-2
and (R)-1 bind in identical positions. However, the side chain
of (R)-2 has moved relative to the position of (R)-1, allowing
interaction of the methoxy group oxygen atom with the
positively charged R1173 (Figure 2D). Compound (R)-6, with
the methoxy group in the 6-position of the tetrahydroquino-
line, shows lower CREBBP affinity than (R)-1, suggesting
that the methoxy group in this position cannot interact with
R1173 without experiencing a steric clash with the protein.
To confirm that the crystallographically observed cation–
p interaction is an important component of ligand binding in
the solution phase we undertook molecular dynamics (MD)
simulations. In an aqueous environment, the interaction
between R1173 and the aromatic ring of (R)-2 was observed
to form spontaneously during the simulation (0.1 ms, Fig-
ure S6). Using a cut-off distance of 6 ꢀ, as proposed by
Dougherty,[19] the cation–p interaction was maintained for
approximately 40% of the trajectory time. Using counter-
poise-corrected dispersion corrected DFT calculations we
estimated the strength of the cation–p interaction to be in the
region of 3.2–4.7 kcalmolÀ1 (Figure S8). These calculated
values agree well with the experimentally measured average
strengths of cation–p interactions involving either Lys
(À3.3 Æ 1.5 kcalmolÀ1) or Arg (À2.9 Æ 1.4 kcalmolÀ1) (see
Supporting Information).[19] Hence the cation–p interaction
appears to be of relevance to the protein–ligand binding in
solvent. This observation is consistent with literature reports
in which optimization of cation–p interactions between
ligands and proteins has resulted in enhanced binding
affinities.[20–23] The MD simulations also identified that the
hydrogen bonds formed between (R)-2 and N1168 and
dispersive interactions between V1174 and the ring face of
the KAc mimic inside the binding pocket contribute to the
binding energy (Figure S6). The combination of AlphaScreen
data, structural information and MD studies demonstrate that
Figure 3.
A) A plot of CPCM-wB97XD/TZVP relative energetics (kcalmolÀ1) in implicit water for rotation about the Ccarbonyl–Caryl bond of the model compounds (R)-9
(green) and (R)-10 (orange). Compound (R)-9 adopts a single lowest energy conformation, stabilized by an internal hydrogen bond between the O atom of
the benzoxazinone ring and the amide NH. This conformation is similar to that observed in the small-molecule X-ray crystal structure of (R)-3. B) The two
lowest energy conformations of (R)-10 are stabilized by internal hydrogen bonds between the NH of the dihydroquinoxalinone and the amide O atom.
These conformations are similar to that observed in the CREBBP bromodomain-bound X-ray crystal structure of (R)-2. C) The reversed internal hydrogen
bond changes the amide bond vector in benzoxazinone and dihydroquinoxalinone analogues (blue arrows).
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
the tetrahydroquinoline ring is forming a cation–p interaction
with R1173 through binding in an induced-fit pocket. It seems
that this interaction might be important for ligands to show
high affinity for the CREBBP bromodomain.
To explore the significance of the internal hydrogen bond
in pre-organizing the ligand conformation for bromodomain
binding, we performed DFT (wB97XD/TZVP) calculations
on simplified model compounds (R)-9 and (R)-10 (Figure 3
and Table S6). These calculations indicate that benzoxazi-
none-derived (R)-9 will adopt a single conformer at room
temperature stabilized by the internal hydrogen bond (Fig-
ure 3A). The DFT-calculated
bromodomain, as orienting the side chain in this direction
would result in a clash with the protein and ZA channel
waters and low binding affinity. To rotate the Ccarbonyl–Caryl
bond away from the optimal 1868 observed in compound (R)-
3 and avoid this clash, the internal hydrogen bond must be
broken, resulting in a low affinity CREBBP bromodomain
ligand.
To determine the utility of (R)-2 in a cellular setting
a FRAP assay was conducted. Data from this assay demon-
strated that (R)-2 can displace the CREBBP bromodomain
from chromatin (Figure 4 and S9) in U2OS cells in a dose-
conformation
(Ccarbonyl–Caryl
dihedral angle q = 2108) is sim-
ilar to the conformation
adopted by the high-resolution
crystal structure [q = 186.4
(2)8]. In the dihydroquinoxali-
none, the internal hydrogen
bond forms between the NH of
the dihydroquinoxalinone and
the amide carbonyl oxygen
atom, reversing the sense of
the internal hydrogen bond
and consequently changing the
vector of the amide substituent
(Figure 3C). DFT calculations
indicate that two of the lowest
energy conformations adopted
by (R)-10 (q = 308 and 3308)
possess the internal hydrogen
bond (Figure 3B). When bound
to CREBBP (R)-2 adopts a con-
formation that is intermediate
between these two (q = 28). An
MD simulation (250 ns) in
explicit water indicated that
the values observed for the O-
C-C-C dihedral angle are fully
consistent with the DFT com-
puted energy profile, with the
value maintained close to the
global energy minimum due to
the existence of an intramolec-
ular hydrogen bond (Figure S7).
This reinforces the idea that this
H-bond plays a key role in
determining the solution-phase
conformation of (R)-10. There-
fore, while the internal hydro-
gen bond in (R)-1 and (R)-2
orients the tetrahydroquinoline
towards R1173 and the induced-
fit pocket, the internal hydrogen
bond in (R)-3 directs the side
chain in the opposite direction.
This observation explains the
low affinity of the benzoxazi-
Figure 4.
A) Fluorescence recovery after photobleaching (FRAP) demonstrating compound (R)-2 can displace the
CREBBP bromodomain from chromatin. Nuclei of U2OS cells transfected with plasmids encoding GFP
chimerized to wild-type (wt) or mutant (N1168F) multimerized CREBBP bromodomain, with or without
2.5 mm SAHA (indicated by †) and the inhibitor (R)-2 at the indicated concentrations. The bleached area
is indicated by a red circle. B) Half times of fluorescence recovery (t = ) of U2OS cells expressing green
fluorescent protein (GFP) chimerized to multimerized CREBBP bromodomains with or without 2.5 mm
1
2
1
SAHA (indicated by †). Bars represent the mean t = calculated from individual recovery curves of 15 cells
2
none series for the CREBBP per group. * significantly different from wt, p<0.0001. ** significantly different from wt†, p<0.0001.
4
www.angewandte.org ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
[5] P. V. Fish, P. Filippakopoulos, G. Bish, P. E. Brennan, M. E.
dependent manner. Measurement of the time taken for
fluorescence recovery, in a bleached area of cells expressing
GFP chimerized to multimerized CREBBP bromodomains,
Bunnage, A. S. Cook, O. Federov, B. S. Gerstenberger, H. Jones,
S. Knapp, et al., J. Med. Chem. 2012, 55, 9831 – 9837.
[6] X. Lucas, D. Wohlwend, M. Hꢂgle, K. Schmidtkunz, S. Gerhardt,
R. Schꢂle, M. Jung, O. Einsle, S. Gꢂnther, Angew. Chem. 2013,
125, 14305 – 14309; Angew. Chem. Int. Ed. 2013, 52, 14055 –
14059.
[7] D. S. Hewings, T. P. C. Rooney, L. E. Jennings, D. A. Hay, C. J.
Schofield, P. E. Brennan, S. Knapp, S. J. Conway, J. Med. Chem.
2012, 55, 9393 – 9413.
1
shows that t = is dose-dependently reduced in the presence of
2
(R)-2 compared to wild-type cells, in the presence of
suberanilohydroxamic acid (SAHA/Vorinostat) (2.5 mm).
1
The t = in the presence of 5 and 10 mm (R)-2 is similar to the
2
N1168F mutant, which shows reduced binding to KAc. These
data indicate that (R)-2 interacts with the KAc binding site of
the CREBBP bromodomain to displace it from chromatin.
In summary, we have developed a series of dihydroqui-
noxalinone-derived CREBBP bromodomain ligands. MD and
DFT calculations indicate that an internal hydrogen bond
helps to preorganize the solution phase conformation of (R)-2
in a manner that is favorable for CREBBP bromodomain
binding, resulting in high-affinity ligands. The tetrahydroqui-
noline side chain of these compounds binds in a previously
unknown induced-fit pocket, and this interaction is stabilized
by a cation–p interaction. MD simulations and DFT calcu-
lations indicate that this cation–p interaction forms sponta-
neously in water and contributes to the ligandꢁs affinity for the
CREBBP bromodomain. Compound (R)-2 inhibits binding of
the CREBBP bromodomain to chromatin in U2OS cells.
These compounds demonstrate that it is possible to develop
potent “drug-like” ligands (Table S7) for non-BET bromo-
domains. They will be invaluable tools to validate inhibition
of the CREBBP bromodomain as a therapeutic target in
diseases including leukemias, Rubinstein–Taybi syndrome,
ovarian, breast and lung cancers and systemic lupus eryth-
ematosus.
[8] Sachchidanand, L. Resnick-Silverman, S. Yan, S. Mutjaba, W.-J.
Liu, L. Zeng, J. J. Manfredi, M.-M. Zhou, Chem. Biol. 2006, 13,
81 – 90.
[9] G. Gerona-Navarro, Y. Rodrꢃguez, S. Mujtaba, A. Frasca, J.
Patel, L. Zeng, A. N. Plotnikov, R. Osman, M.-M. Zhou, J. Am.
Chem. Soc. 2011, 133, 2040 – 2043.
[10] J. C. Borah, S. Mujtaba, I. Karakikes, L. Zeng, M. Muller, J.
Patel, N. Moshkina, K. Morohashi, W. Zhang, G. Gerona-
Navarro, et al., Chem. Biol. 2011, 18, 531 – 541.
[11] D. Hay, O. Fedorov, P. Filippakopoulos, S. Martin, M. Philpott, S.
Picaud, D. S. Hewings, S. Uttakar, T. D. Heightman, S. J.
Conway, et al., Med. Chem. Commun. 2013, 4, 140 – 144.
[12] O. Fedorov, H. Lingard, C. Wells, O. P. Monteiro, S. Picaud, T.
Keates, C. Yapp, M. Philpott, S. J. Martin, I. Felletar, et al., J.
Med. Chem. 2014, 57, 462 – 476.
[13] F. Wang, C. B. Marshall, M. Ikura, Cell. Mol. Life Sci. 2013, 70,
3989 – 4008.
[14] M. Philpott, J. Yang, T. Tumber, O. Fedorov, S. Uttarkar, P.
Filippakopoulos, S. Picaud, T. Keates, I. Felletar, A. Ciulli, et al.,
Mol. BioSyst. 2011, 7, 2899 – 2908.
[15] D. S. Hewings, M. Wang, M. Philpott, O. Fedorov, S. Uttarkar, P.
Filippakopoulos, S. Picaud, C. Vuppusetty, B. Marsden, S.
Knapp, et al., J. Med. Chem. 2011, 54, 6761 – 6770.
[16] C.-W. Chung, A. W. Dean, J. M. Woolven, P. Bamborough, J.
Med. Chem. 2012, 55, 576 – 586.
[17] D. S. Hewings, O. Fedorov, P. Filippakopoulos, S. Martin, S.
Picaud, A. Tumber, C. Wells, M. M. Olcina, K. Freeman, A. Gill,
et al., J. Med. Chem. 2013, 56, 3217 – 3227.
Received: February 24, 2014
Published online: && &&, &&&&
[18] P. Filippakopoulos, S. Picaud, M. Mangos, T. Keates, J.-P.
Lambert, D. Barsyte-Lovejoy, I. Felletar, R. Volkmer, S.
Muller, T. Pawson, et al., Cell 2012, 149, 214 – 231.
[19] J. P. Gallivan, D. A. Dougherty, Proc. Natl. Acad. Sci. USA 1999,
96, 9459 – 9464.
[20] P. Sçrme, P. Arnoux, B. Kahl-Knutsson, H. Leffler, J. M. Rini,
U. J. Nilsson, J. Am. Chem. Soc. 2005, 127, 1737 – 1743.
[21] L. M. Salonen, C. Bucher, D. W. Banner, W. Haap, J.-L. Mary, J.
Benz, O. Kuster, P. Seiler, W. B. Schweizer, F. Diederich, Angew.
Chem. 2009, 121, 825 – 828; Angew. Chem. Int. Ed. 2009, 48, 811 –
814.
Keywords: bromodomain · CREBBP · enzyme inhibitors ·
epigenetics · ligand discovery
.
[1] M. E. Bunnage, Nat. Chem. Biol. 2011, 7, 335 – 339.
[2] P. Filippakopoulos, J. Qi, S. Picaud, Y. Shen, W. B. Smith, O.
Fedorov, E. M. Morse, T. Keates, T. T. Hickman, I. Felletar,
et al., Nature 2010, 468, 1067 – 1073.
[3] E. Nicodeme, K. L. Jeffrey, U. Schaefer, S. Beinke, S. Dewell, C.-
W. Chung, R. Chandwani, I. Marazzi, P. Wilson, H. Coste, et al.,
Nature 2010, 468, 1119 – 1123.
[4] 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.
[22] L. M. Salonen, M. Ellermann, F. Diederich, Angew. Chem. 2011,
123, 4908 – 4944; Angew. Chem. Int. Ed. 2011, 50, 4808 – 4842.
[23] A. S. Mahadevi, G. N. Sastry, Chem. Rev. 2013, 113, 2100 – 2138.
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Ligand Discovery
T. P. C. Rooney, P. Filippakopoulos,
O. Fedorov, S. Picaud, W. A. Cortopassi,
D. A. Hay, S. Martin, A. Tumber,
C. M. Rogers, M. Philpott, M. Wang,
A. L. Thompson, T. D. Heightman,
D. C. Pryde, A. Cook, R. S. Paton,
S. Mꢁller, S. Knapp, P. E. Brennan,
S. J. Conway*
&&&& — &&&&
Ligands made to fit: A series of dihydro-
quinoxalinone fragments were employed
as acetyl lysine mimics in the develop-
ment of CREBBP bromodomain ligands.
The side chain of these ligands binds in
an induced-fit pocket and forms a cation–
p interaction with R1173 of CREBBP. The
most potent compound inhibits binding
of CREBBP to chromatin in U2OS cells.
A Series of Potent CREBBP Bromodomain
Ligands Reveals an Induced-Fit Pocket
Stabilized by a Cation–p Interaction
6
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!