Fluorescence polarization for the evaluation of small-molecule inhibitors of PCAF BRD/Tat-AcK50 association

Article abstract of DOI:10.1002/cmdc.201300499

A fluorescence polarization competitive assay was developed to efficiently screen and evaluate inhibitors of PCAF bromodomain/Tat-AcK50 protein-peptide interaction. A series of pyridine 1-oxide derivatives were synthesized and evaluated. Some of the novel

Full text of DOI:10.1002/cmdc.201300499

CHEMMEDCHEM
COMMUNICATIONS
DOI: 10.1002/cmdc.201300499
Fluorescence Polarization for the Evaluation of Small-
Molecule Inhibitors of PCAF BRD/Tat-AcK50 Association
Ping Hu,[a] Xinghui Wang,[a] Baiqun Zhang,[a] Shuai Zhang,[a] Qiang Wang,*[b] and
Zhiyong Wang*[a]
A fluorescence polarization competitive assay was developed
to efficiently screen and evaluate inhibitors of PCAF bromodo-
main/Tat-AcK50 protein–peptide interaction. A series of pyri-
dine 1-oxide derivatives were synthesized and evaluated. Some
of the novel compounds, 2-(3-aminopropylamino) pyridine 1-
oxide derivatives, could be effective inhibitors of PCAF bromo-
domain/Tat-AcK50 association. Specifically, 2-(3-aminopropyla-
mino)-5-(hydroxymethyl)pyridine 1-oxide hydrochloride (15)
and the 5-((3-aminopropylamino)methyl) derivative (20) were
found to be effective ligands for the PCAF BRD pocket. First
preliminary cellular studies indicate that these small-molecule
inhibitors have lower cytotoxicities and are potential leads for
the anti-HIV/AIDS therapeutic strategy by targeting host-cell
protein PCAF BRD to block HIV replication.
In preceding research, 2D NMR and enzyme-linked immuno-
sorbent assay (ELISA) were applied for the evaluation of ligand
binding affinity towards PCAF BRD in vitro.[4,5] However, as
these methods are rather inefficient and time consuming, it is
necessary to develop new assays for inhibitor screening and
evaluation. Fluorescence polarization (FP) is a technique espe-
cially applied to study molecular interactions. It gives a direct
and nearly instantaneous measurement of the ratio of bound
to free tracer.[13] In recent years, as a homogeneous assay, the
fluorescence polarization assay (FPA) was widely applied in sys-
tems that involve binding of ligands to receptors, in which flu-
orescently labelled ligands are used as a tracers. Because of its
simplicity, sensitivity and speed, FPA is readily suitable for
probing biological associations such as ligand–protein,[6] enzy-
matic activity,[7] peptide–protein binding,[8] DNA hybridiza-
tion,[7,9] DNA–protein binding,[10] and even for Hg2+ ion deter-
mination.[11] In addition, outstanding advances of FPA, showing
the measurement of DNA wrapping, the close contact involved
in protein–DNA binding and the application in immunoassays,
have been reported.[12] Recent developments in the field indi-
cate that FP is a useful tool in high-throughput screening
(HTS) and in small-molecule drug discovery.[8c,13] Inspired by
former research and using the FP principle, we finally devel-
oped a novel FPA involving the interaction between PCAF BRD
and fluorescein isothiocyanate (FITC)-labelled Tat peptide.
More importantly, a series of pyridine 1-oxide derivatives were
synthesized and evaluated.
Human p300/CBP associated factor bromodomain (PCAF BRD),
a relatively conserved host-cell protein, can selectively bind to
Tat-AcK50 (lysine 50 acetylated trans-activator). Former re-
search showed that the transcription of the integrated HIV pro-
virus required the interaction of HIV Tat and human co-activa-
tor PCAF BRD.[1,2] PCAF BRD adopts a highly conserved struc-
tural fold of a left-handed four-helix bundle (aZ, aA, aB, and
aC). The ZA and BC loops are at one end of the bundle, form-
ing a hydrophobic pocket, which is suitable for the binding of
Tat-AcK50.[3] Previously, we developed a series of N1-aryl-pro-
pane-1,3-diamine compounds, which could bind to the hydro-
phobic pocket and inhibit the interaction of PCAF BRD and
Tat-AcK50.[4] Based on this former research, we are attempting
to discover novel small molecules that bind to PCAF BRD. Tar-
geting an essential host-cell protein for HIV-1 replication rather
than enzymes or proteins encoded in the viral gene, we as-
sumed that this target could minimize drug resistance result-
ing from rapid mutations of the HIV gene.
A schematic illustration of the FP competition assay principle
is presented in Figure 1. In FP competitive binding assays, the
[a] P. Hu, X. Wang, B. Zhang, S. Zhang, Prof. Z. Wang
Hefei National Laboratory for Physical Sciences at Microscale
CAS Key Laboratory of Soft Matter Chemistry and Department of Chemistry
University of Science and Technology of China
Hefei, Anhui, 230026 (P.R. China)
E-mail: zwang3@ustc.edu.cn
[b] Q. Wang
Research assistant Qiang Wang
High Magnetic Field Laboratory Chinese Academy of Sciences Hefei
Hefei, Anhui, 230031 (P.R. China)
E-mail: wq2013@hmfl.ac.cn
Figure 1. Schematic illustration of the FP competition assay. When the fluo-
rescence tracer binds to the target protein, high FP signal is observed. When
small-molecule inhibitors completely disrupt the peptide–protein interac-
tions, the FP signal is shifted to low values.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201300499.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 928 – 931 928

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
of NP1. In the primary FPA screening, 3 showed promising in-
hibitory potency against the PCAF BRD/Tat-AcK50 interaction.
The IC50 value was assessed by the competitive assay against
FITC-labelled Tat-AcK50 binding to PCAF BRD described above.
Surprisingly, an IC50 value of 4.62Æ0.23 mm was obtained. In-
trigued by this promising lead, a series of analogues of 3 were
synthesized, and their binding affinities with PCAF BRD were
evaluated using the developed FPA. The assay results indicated
several candidate compounds to be effective inhibitors of the
PCAF BRD/Tat-AcK50 protein–peptide interaction, as shown in
Table 1. According to the FPA results, a primary structure–activ-
ity relationship (SAR) study was performed. First of all, the pyri-
dine-oxide moiety of the compounds was very important for
binding to PCAF BRD. The binding affinity was largely de-
creased without this moiety in the molecules (3 versus 22, 6
versus 23). Secondly, the 1,3-diamine moiety on the 2-position
also played a crucial role for activity retention. Lengthening or
shortening the carbon chain weakened the binding affinity (3
versus 2 and 4). Changing the 1,3-diamine moiety to other
chain moieties decreased the inhibitory potency, and most of
Table 1. 2-Aminopyridine 1-oxide derivatives as inhibitors for PCAF BRD/Tat-AcK50 in-
teraction.
Figure 2. A) Binding test: FITC-Tat 3 nm, GST-PCAF BRD concen-
tration: 0 nm, 25 nm, 50 nm, 100 nm, 150 nm, 200 nm, 48C, 6 h in
&
*
either 125 mm Tris, pH 6.85 ( ); 125 mm Tris, pH 7.2 ( ); 100 mm
~
!
PBS, pH 6.85 ( ); 100 mm PBS, pH 7.2 ( ). Excitation: 485 nm,
emission: 535 nm. B) Competitive FP assay for IC50 calculation of
NP1. Assay conditions: 3 nm FITC-Tat, 100 nm GST-PCAF BRD,
125 mm Tris, pH 6.85, 48C, 6 h. Excitation: 485 nm, emission:
535 nm.
Compd[a] R1
R2
IC50 [mm][b]
38.24Æ1.82
EC50 [mm][c]
1
5-Me
À(CH2)2CH3
94.51
52.11
91.52
112.55
87.28
13.94
2
3
4
5
5-Me
5-Me
5-Me
5-Br
À(CH2)2NH2
21.28Æ1.11
4.62Æ0.23
10.69Æ0.26
20.70Æ0.67
11.73Æ0.27
86.66Æ2.31
9.31Æ0.26
À(CH2)3NH2
presence of unlabelled small-molecule inhibitors
result in the displacement of FITC-labelled Tat pep-
tide, thereby increasing their tumbling motion, which
in turn can be detected as a decrease in FP value.
Figure 2A shows the effects for the typical variation
of the concentration of glutathione S-transferase
(GST)-tagged PCAF BRD, and satisfactory FP signals
(in mP) could be achieved. To verify the conditions
for the FP competition assay, we performed the assay
À(CH2)4NH2
À(CH2)3NH2
6
7
8
9
10
11
12
5-CF3
5-Ph
4-Me
5-Me
5-Me
5-Me
5-Me
À(CH2)3NH2
À(CH2)3NH2
61.85
À(CH2)3NH2
180.68
>200
100.12
78.83
ÀBn
>200
À(CH2)2NH(CH2)2OH
À(CH2)2NH(CH2)2NH
À(CH2)2NHCH2CH(OH)CH3
5.51Æ0.19
10.25Æ0.31
31.43Æ1.21
75.28
13
14
5-Me
5-Me
>200
>200
>200
*
in 125 mm Tris, pH 6.85 ( , Figure 2A) to assess the
inhibitory activity of the known small-molecule
ligand NP1. The obtained IC50 value (1.37Æ0.09 mm,
see Figure 2B) was consistent with the previous IC50
value assessed by ELISA in vitro (1.60Æ0.10 mm). The
FP competitive assay results of other compounds
also coincide with our in vitro anti-HIV-1 activity
(Table 1.). These results indicate that our FP competi-
tion assay could be an effective evaluation method
and could be applied for inhibitor screening and
evaluation for PCAF BRD.
>200
15
5-CH2OH
À(CH2)3NH2
1.18Æ0.09
8.90Æ0.31
>200
11.92
66.1
>200
16
5-CH2NH2
À(CH2)3NH2
17
5-CH2OH
ÀCH2CH(OH)CH2NH2
À(CH2)3NH2
18
5-CH2NHCH2Ph
162.61Æ3.68 >200
19
20
21
22[d]
23[d]
5-CH2NH(CH2)2NH2 À(CH2)3NH2
5-CH2NH(CH2)3NH2 À(CH2)3NH2
5-CH2NH(CH2)4NH2 À(CH2)3NH2
2.26Æ0.19
0.93Æ0.12
9.76Æ0.37
71.85
11.52
40.64
5-Me
5-CF3
À(CH2)3NH2
À(CH2)3NH2
>200
>200
181.28
145.20Æ2.89
[a] Compounds bearing primary amine groups were transformed to their hydrochlo-
ride salts for assays. [b] Determined using FP competition assay described above. Data
are the meanÆSD of three experiments performed in triplicate. [c] Determined using
human T lymphocyte cell line C8166 and HIV-1IIIB. Data are the mean of three tripli-
cate experiments, and EC50 values were determined by nonlinear regression analysis
using GraphPad Prism 5.0. [d] Without N-oxide on the pyridine ring.
Due to their structural similarity to NP1 and low
toxicity, 2-(3-aminopropylamino)pyridine 1-oxide de-
rivatives, a novel class of compounds, were specifical-
ly designed and synthesized as PCAF BRD ligands. Ini-
tially, compound 3 was synthesized as an analogue
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 928 – 931 929

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
docking structure (DOCK soft-
ware)[14] of PCAF BRD with 3 and
found that 3 interacts with the
PCAF BRD side chain residues of
W746, E750 and E756, similar to
the interaction with NP1.
In conclusion, we developed
an inhibitor screening and evalu-
ation method for PCAF BRD
using a FP competitive assay.
Furthermore, a series of pyridine
N-oxide 1,3-diamine small mole-
cules were synthesized and
screened by this FP competitive
assay. Some effective lead com-
pounds were selected to inhibit
the PCAF BRD/Tat-AcK50 associa-
tion in vitro, and the corre-
sponding SAR was investigated.
Among these compounds, 20
and 15 were found to be effec-
tive ligands for the PCAF BRD
pocket. Importantly, the primary
Figure 3. Structural basis of ligand recognition of PCAF BRD. A) A docking model of the BRD/compound 3 com-
plex, showing the binding site of 3 (magenta). B) GRASP view of the binding pocket of 3. C) A docking model of
the BRD/3 (magenta) and BRD/NP1 (yellow) complex, showing the binding sites of 3 and NP1. D) GRASP view of 3
and NP1 binding pocket.
the changes greatly impaired the binding affinity (3 versus 1,
17, 9–14). Moreover, it was noted that large substitution
groups at the 2-position of the pyridyl ring abolished activity,
with IC50 >200 mm (9, 13 and 14), which implied that big
volume of the substituent at the 2-position of pyridine N-oxide
would block the pocket recognition of PCAF BRD to the small
molecule. Thirdly, the substituent at the 5-position had a great
influence on the biding affinity. For instance, when 5-Me (3)
was changed to 4-Me (8), the corresponding IC50 value de-
creased to 9.31Æ0.26 mm. The bioactivity was almost disrupted
with an IC50 value of more than 80 mm when 5-Me (3) was re-
placed with 5-Ph (7). In other cases, the variation of the 5-posi-
tion substitution could also have a positive influence on the
bioactivity, as shown with 15, 19 and 20. In particular, when
a CH2OH group was located at the 5-position of pyridine N-
oxide (15), the corresponding IC50 value was increased nearly
fourfold compared to that of 5-Me substitution (3). Similarly,
with 5-CH2NH(CH2)2NH2 (19) and 5-CH2NH(CH2)3NH2 (20), which
differ in the number of methylene groups only, the binding af-
finity was increased remarkably. The bioactivity of 20 was the
best compound with an increased IC50 value of nearly fivefold
compared to that of 3. Finally, when a chiral center in the cen-
tral carbon of the 1,3-diamine moiety of 15 was introduced by
adding a hydroxy group (17), the binding affinity was abolish-
ed with an IC50 >200 mm after this minor modification. This in-
dicates that the 1,3-diamine moiety at the 5-position of pyri-
dine N-oxide played an important role in the binding affinity
of these small molecules.
SAR will facilitate our efforts to design new inhibitors to recog-
nize the PCAF BRD pocket. First preliminary cellular studies in-
dicate that these small-molecule inhibitors have lower cytotox-
icities and are potential leads for the anti-HIV/AIDS therapeutic
strategy by targeting a cellular protein PCAF BRD to block HIV
replication.
Acknowledgements
We gratefully acknowledge the University of Science and Technol-
ogy of China and the Natural Science Foundation of China
(91213303, 21272222, J1030412, 21172205).
Keywords: fluorescence polarization
· HIV Tat · PCAF ·
protein–peptide interactions · small-molecule inhibitors
[1] S. Mujtaba, Y. He, L. Zeng, A. Farooq, J. E. Carlson, M. Ott, E. Verdin,
M. M. Zhou, Mol. Cell 2002, 9, 575.
[2] C. Hetzer, W. Dormeyer, M. Schnolzer, M. Ott, Microbes Infect. 2005, 7,
1364.
[3] C. Dhalluin, J. E. Carlson, L. Zeng, C. He, A. K. Aggarwal, M. M. Zhou,
Nature 1999, 399, 491.
[4] L. Zeng, J. Li, M. Muller, S. Yan, S. Mujtaba, C. Pan, Z. Wang, M. M. Zhou,
J. Am. Chem. Soc. 2005, 127, 2376.
[5] C. F. Pan, M. Mezei, S. Mujtaba, M. Muller, L. Zeng, J. M. Li, Z. Y. Wang,
M. M. Zhou, J. Med. Chem. 2007, 50, 2285.
[6] N. Rezvani, R. J. Mayer, W. C. Chan, Chem. Commun. 2010, 46, 2043.
[7] M. Zhang, Y. M. Guan, B. C. Ye, Chem. Commun. 2011, 47, 3478.
[8] a) P. Wu, M. Brasseur, U. Schindler, Anal. Biochem. 1997, 249, 29; <
lit b>J. W. Choi, D. K. Kang, H. Park, A. J. deMello, S. Chang, Anal.
Chem. 2012, 84, 3849; c) J. Saupe, Y. Roske, C. Schillinger, N. Kamdem, S.
Radetzki, A. Diehl, H. Oschkinat, G. Krause, U. Heinemann, J. Rademann,
ChemMedChem 2011, 6, 1411.
[9] A. Murakami, M. Nakaura, Y. Nakatsuju, S. Nagahara, Q. Tran-Cong, K.
Makino, Nucleic Acids Res. 1991, 19, 4097.
[10] M. S. Ozers, J. J. Ervin, K. Hill, J. R. Wood, A. M. Nardulli, C. A. Royer, J. J.
Gorski, Biol. Chem. 1997, 272, 30405.
In order to find how pyridine N-oxide 1,3-diamine derivatives
bound to PCAF BRD, we attempted to analyze the interaction
by molecular docking (Figure 3). It was found that the docked
molecule took a better scoring pose with a higher level of
sampling than the previous binding pocket of NP1. Based on
the structure of complex PCAF BRD/NP1,[4] we obtained the
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 928 – 931 930

CHEMMEDCHEM
COMMUNICATIONS
www.chemmedchem.org
[11] B. C. Ye, B. C. Yin, Angew. Chem. 2008, 120, 8514; Angew. Chem. Int. Ed.
2008, 47, 8386.
[14] T. J. Ewing, S. Makino, A. G. Skillman, I. D. Kuntz, J. Comput.-Aided. Mol.
Des. 2001, 15, 411.
[12] a) H. Wang, M. Lu, M. S. Tang, B. Van Houten, J. B. Ross, M. Weinfeld,
X. C. Le, Proc. Natl. Acad. Sci. USA 2009, 106, 12849; b) D. Zhang, M. Lu,
H. Wang, J. Am. Chem. Soc. 2011, 133, 9188.
Received: December 2, 2012
[13] W. A. Lea, A. Simeonov, Expert Opin. Drug Discovery 2011, 6, 17.
Published online on January 28, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2014, 9, 928 – 931 931