Identification of 3 H -naphtho[1,2,3- de ]quinoline-2,7-diones as inhibitors of apoptosis signal-regulating kinase 1 (ASK1)

Article abstract of DOI:10.1021/jm200117h

Apoptosis signal-regulating kinase 1 (ASK1) has recently emerged as an attractive therapeutic target for the treatment of cardiac and neurodegenerative disorders. The selective inhibitors of ASK1 may become important compounds for the development of clini

Full text of DOI:10.1021/jm200117h

ARTICLE
pubs.acs.org/jmc
Identification of 3H-Naphtho[1,2,3-de]quinoline-2,7-diones as
Inhibitors of Apoptosis Signal-Regulating Kinase 1 (ASK1)
Galyna P. Volynets,† Maksym O. Chekanov,†,‡ Anatoliy R. Synyugin,†,‡ Andriy G. Golub,†
Oleksandr P. Kukharenko,† Volodymyr G. Bdzhola,† and Sergiy M. Yarmoluk*,†,‡
†Institute of Molecular Biology and Genetics, NAS of Ukraine, 150 Zabolotnogo Str., 03143, Kyiv, Ukraine
‡Otava, Ltd., 150 Zabolotnogo Str., 03143, Kyiv, Ukraine
ABSTRACT: Apoptosis signal-regulating kinase 1 (ASK1) has
recently emerged as an attractive therapeutic target for the
treatment of cardiac and neurodegenerative disorders. The selec-
tive inhibitors of ASK1 may become important compounds for the
development of clinical agents. We have identified the ASK1
inhibitor among 3H-naphtho[1,2,3-de]quinoline-2,7-diones
using receptor-based virtual screening. In vitro kinase assay
revealed that ethyl 2,7-dioxo-2,7-dihydro-3H-naphtho[1,2,3-
de]quinoline-1-carboxylate (NQDI-1) inhibited ASK1 with a
Ki of 500 nM. The competitive character of inhibition is
demonstrated in LineweaverꢀBurk plots. In our preliminary
selectivity study this compound exhibited strong specific in-
hibitory activity toward ASK1.
’ INTRODUCTION
induced septic shock.6 ASK1 can also be regarded as an important
therapeutic target for the treatment of malignant fibrous histiocytoma.7
Thus, the potent and selective inhibitors of ASK1 would be
important compounds for the development of clinical agents.
However, there have been few reports regarding ASK1
inhibitors.8ꢀ10 The main aim of our research is to identify the
small molecule inhibitors of ASK1.
Apoptosis signal-regulating kinase 1 (ASK1) is the ubiqui-
tously expressed mitogen-activated protein kinase kinase kinase 5
that has a broad range of biological functions depending on the
cell type and cellular context. ASK1 is associated with several
diseases and may be a target for pharmaceutical intervention. For
example, ASK1 was identified as an essential component in the
neuronal death signaling pathway induced by expanded poly-
glutamine (polyQ) repeats.1 Reactive oxygen species (ROS)
mediated ASK1 activation by amyloid β (Aβ) protein is an
important step in Alzheimer’s disease pathogenesis.2 Recently, it
has been revealed that ASK1 is also involved in motor neuron cell
death during familial amyotrophic lateral sclerosis.3 Therefore,
inhibitors of ASK1 can be considered as therapeutic remedies for
patients with chronic neurodegenerative diseases.
New experimental results clearly showed that ASK1 deficiency
attenuated cardiac inflammation and fibrosis, as well as vascular
endothelial dysfunction and remodeling induced by the obesity/
diabetes. ASK1 deficiency also abolished hepatic steatosis in
diabetic mice.4 Moreover, the inhibition of ASK1 by the recom-
binant adeno-associated virus expressing a dominant-negative
ASK1ΔN-KR mutant was capable of suppressing heart failure
progression by preventing cardiomyocyte apoptosis.5 Hence, the
inhibition of ASK1 activity may provide a valid approach for the
treatment of cardiovascular diseases. In addition, ASK1 inhibitors
could have broad therapeutic potential for many acute syn-
dromes such as cardiac infarction or stroke as inhibitors of
apoptosis.
’ RESULTS AND DISCUSSION
The in silico approach is an important part of protein kinase
inhibitor rational design. The docking process involves the
prediction of ligand conformation within a targeted binding site
and energy calculations (electrostatic and van der Waals) during
complex formation.11 The most important application of dock-
ing software is a virtual screening that allows selection of the most
interesting and promising molecules from the existing com-
pound collections for further evaluation and investigation.
In order to discover novel ASK1 inhibitors, we have performed
a screening program, using both in silico and in vitro approaches.
The system based on DOCK 4.0 package was used for
receptorꢀligand flexible docking.12ꢀ15 The virtual screening
experiments were carried out targeting the ATP binding site of
ASK1 by browsing the commercially available compound library
of 156 000 organic compounds.16 The most promising 172
compounds having a docking score less than ꢀ50 kcal/mol have
been selected and taken for the kinase assay study. In vitro experi-
ments revealed that ethyl 2,7-dioxo-2,7-dihydro-3H-naphtho
In vivo study supports the role of ASK1 in the immune response,
since ASK1-deficient mice were resistant to lipopolysaccharide (LPS)
Received: November 10, 2010
Published: March 30, 2011
r
2011 American Chemical Society
2680
dx.doi.org/10.1021/jm200117h J. Med. Chem. 2011, 54, 2680–2686
|

Journal of Medicinal Chemistry
ARTICLE
Figure 1. Kinetic analysis of 1ꢀASK1 complexation consistent with a reversible and competitive mechanism of inhibition. ASK1 activity was
determined as described in the Experimental Section either in the absence or in the presence of the indicated concentrations. The data represent the
mean of triplicate experiments with SEM never exceeding 15%. (A) LineweaverꢀBurk linearized plots of the substrate concentration curves for ASK1 at
different concentrations of the compound 1. Concentration of inhibitor varied from 0 to 4 μM. (B) Calculation of Ki from linear regression analysis of
LineweaverꢀBurk double-reciprocal plots.
μM. Compound 3 with ethylamine in this position has IC50 = 10
μM, and compounds 5 and 6 substituted with 2-hydroxyethyla-
mine and benzoyl have IC50 equal to 25 and 35 μM, respectively.
The nature of the R3 substituent also significantly affects the
inhibitory activity toward ASK1. For instance, in derivative 4
(IC50 = 15 μM), a substitution of the hydrogen atom with a
methyl group (derivative 7) results in over 2-fold decrease in
activity (IC50 = 35 μM)).
Table 1. Residual Activity of Protein Kinases (%) in the
Presence of 1 at 25 μMa
ASK1 Aurora A ROCK HGFR FGFR1 Tie2 JNK3 CK2
12.5
79
100
62
44
84
100
78
a The values of residual activity represent the mean of at least three
independent experiments with SEM never exceeding 15%. Final con-
centration of ATP in the experiment was 100 μM.
The R4 substituent of the scaffold is of special interest. All
active compounds have hydrogen in this position. The substitu-
tion of hydrogen atom with any other chemical groups results in a
complete loss of inhibition efficiency (derivatives 8ꢀ15).
The compounds in this class are reversible ATP-competitive
inhibitors of ASK1 and, as expected, occupy the ATP binding
region between the N-terminal and C-terminal lobes of the ASK1
enzyme. Our molecular docking investigations have demon-
strated that active derivatives of 3H-naphtho[1,2,3-de]quino-
line-2,7-dione display a very good steric and chemical
complementarity with the ATP-binding pocket. Apolar interac-
tions with hydrophobic residues Leu686, Val694, Ala707,
Met754, Val757, Gly759, and Gly760 are the most important
noncovalent forces that contribute to the ligand recognition
process. A peculiarity of the active inhibitor binding mode in
comparison to those displayed by other investigated compounds
is the formation of a hydrogen bond with Val757 in the hinge
region. All active compounds (1ꢀ7) have R4 = H, whereas all the
nonactive compounds have various substituents in this position
adjacent to the carbonyl group. The substitution of the hydrogen
atom with any other chemical group results in a complete loss of
inhibition efficiency (derivatives 8ꢀ15). According to the visual
inspection of the complexes obtained with docking, R4 = H
allows formation of hydrogen bonds with Val757 in the hinge
region for compounds 1ꢀ7, while all the inactive compounds do
not form this hydrogen bond. The loss of the activity may be
explained as due to the introduction of steric hindrance into this
area of the binding pocket that, in consequence, results in
destabilization of the ligand in the ASK1 active site.
[1,2,3-de]quinoline-1-carboxylate (NQDI-1) inhibited ASK1
with an IC50 of 3 μM.
The LineweaverꢀBurk plots, which are the linearized trans-
formations of the MichaelisꢀMenten curves, demonstrate the
competitive character of inhibition for compound 1 (NQDI-1),
as the inhibitor decreases KM without affecting Vmax (Figure 1A).
Ki of 500 nM has been calculated from linear regression analysis
of LineweaverꢀBurk double reciprocal plots (Figure 1B).
The selectivity of 1 was evaluated in vitro on four serine/
threonine protein kinases (protein kinase CK2 (CK2), c-Jun
N-terminal kinase 3 (JNK3), Rho-associated protein kinase 1
(Rock1), and Aurora A) and three tyrosine protein kinases
(fibroblast growth factor receptor 1 (FGFR1), hepatocyte
growth factor receptor (HGFR), and endothelial TEK tyrosine
kinase (Tie2)). The results of our study have shown that 1 seems
to be a selective inhibitor of ASK1 (Table 1). The activity of
FGFR1 protein kinase is inhibited by 1 (residual activity of 44%);
nevertheless, we suppose that this derivative of 3H-naphtho-
[1,2,3-de]quinoline-2,7-dione is a promising lead for extended
selectivity evaluation and optimization.
In order to examine the structureꢀactivity relationships of the
derivatives of 3H-naphtho[1,2,3-de]quinoline-2,7-dione, the
structural analogues were selected from the Otava, Ltd. com-
pound library and tested in vitro.
Several important structural features of 3H-naphtho[1,2,3-
de]quinoline-2,7-dione derivatives can be identified from a
qualitative analysis of their activity on ASK1 (Table 2). The ASK1
inhibitory activity of the studied derivatives depends on the nature of
R1 substituents in this heterocyclic system. The most active com-
pound 1 with R1 = acetic acid ethyl ester has IC50 = 3 μM. The
presence of an acetyl group in this position has a negative effect
on the inhibitory activity of derivative 4 (IC50 = 15 μM).
Compound 2, wherein R1 is a methylamine, has IC50 = 4.7
Since the protein molecule remains rigid during the docking,
we performed a 10 ns molecular dynamics (MD) simulation to
more accurately predict the binding mode of 1 with the ATP-
binding site of protein kinase ASK1. The binding mode calcu-
lated with MD is similar to the one obtained with docking, and
2681
dx.doi.org/10.1021/jm200117h |J. Med. Chem. 2011, 54, 2680–2686

Journal of Medicinal Chemistry
ARTICLE
Table 2. Structures and in Vitro ASK1 Inhibitory Activity for
3H-Naphtho[1,2,3-de]quinoline-2,7-dione Derivatives
Figure 2. The rmsd of compound 1 in complex with ASK1 during MD
simulation (smoothed by SavitzkyꢀGolay method, 13 steps).
where it interacts with the hydrophobic pocket formed by the
residues of the hinge region and adjacent residues of C-terminal
and N-terminal lobes and forms hydrogen bonds with Gln756
and Val757. These interactions are present almost all the time
during MD simulation. The hydrogen bond existence map
calculated by GROMACS is represented in Figure 3.
The hydrogen-bond existence map shows long-term hydrogen
bonds formed, on one hand, by carbonyl of ligand heterocyclic
system and, on the other hand, by the amino group of Val757
(bond B) and the side chain amino group of Gln756 (bond C).
The hydrogen bond between the ligand and Gln756 is disrupted
at 5500 ps. The loss of this hydrogen bond was particularly
compensated by the hydrogen bond between the carbonyl in the
second position of the compound and hydroxyl group of Ser821
(bond A). At 7600 ps the hydrogen bond between the inhibitor
and Gln756 is formed again.
The described hydrogen bond network correlates well with
the energy of the complex (Figure 4). The average energy value
of ꢀ110 kJ/mol reflects the presence of hydrogen bonds B and C.
Significant slump of the energy curve is directly associated with
the loss of hydrogen bond C. These data suggest an important
role of this intermolecular hydrogen bond in the existence of the
1ꢀASK1 complex.
Thus, using MD simulation we confirmed the formation of
hydrogen bond between 1 and Val757. Also, the hydrogen bond
with Gln756 in the hinge region of ASK1 ATP-binding pocket
has been identified.
The residues Leu686, Val694, Ala707, Met754, Val757,
Gly759, and Gly760 of the ASK1 ATP-binding site are involved
in hydrophobic contacts with 1. The major contributions to the
hydrophobic interactions are provided by Leu686, Met754, and
Leu810. These residues participate in the clamping of the
inhibitor heterocyclic system and therefore in stabilization of
the ligand in the active site. The MD snapshot characterizing the
binding mode of 1 is represented in Figure 5.
Thus, in silico analysis of ASK1 complexes with inhibitors
indicates that steric and hydrophobic effects contribute the
most to the inhibitory activity of the compounds. The com-
pounds providing hydrogen bonding capability with the hinge
region of the ATP-binding site were shown to be potent
inhibitors of ASK1. The results of biological tests allowed us
to conclude that substituents R1, R3, and R4 play a key role in ASK1
inhibitory activity for 3H-naphtho[1,2,3-de]quinoline-2,7-dione de-
rivatives. These features can be used for the development of more
potent inhibitors.
there were no major changes observed. The rmsd plot for 1
demonstrates that no considerable disposition of the inhibitor
occurs during MD simulation (Figure 2). The complex is stable
during the calculation period with no detectable tendencies to
dissociate.
The results of MD simulation allowed us to make the following
deductions. Compound 1 occupies the ASK1 ATP-binding site,
2682
dx.doi.org/10.1021/jm200117h |J. Med. Chem. 2011, 54, 2680–2686

Journal of Medicinal Chemistry
ARTICLE
Figure 3. Hydrogen-bond existence map of 1ꢀASK1 complex during a 10 ns MD simulation.
practice. The goal of this study was to identify small-molecule
inhibitors of protein kinase ASK1. Virtual screening allowed us to
identify a small-molecule inhibitor of human protein kinase
ASK1 among 3H-naphtho[1,2,3-de]quinoline-2,7-diones. In vi-
tro experiments revealed that 1 inhibits ASK1 with a Ki of 500
nM. Inhibition of ASK1 by 1 is competitive with respect to the
phosphodonor substrate ATP. According to the in silico model-
ing results, the mechanism of ASK1 inhibition involves the
hydrogen bond formation between the carbonyl of compound
1 and the ASK1 hinge region. Our preliminary selectivity studies
have demonstrated that 1 is a selective inhibitor of ASK1. These
results strongly suggest that the core structure of this class of
compounds can be used for the development of more potent and
selective inhibitors of ASK1.
Figure 4. Sum of Coulomb and Lennard-Jones interaction energies for
1ꢀASK1 complex during MD simulation (smoothed by Savitz-
kyꢀGolay method, 13 steps).
’ EXPERIMENTAL SECTION
Synthetic Chemistry. Starting materials and solvents were pur-
chased from commercial suppliers and were used without further
1
purification. H NMR spectra were recorded on a Varian VXR 400
instrument at 400 MHz. Chemical shifts are described as parts per
million (δ) downfield from an internal standard of tetramethylsilane,
and spin multiplicities are given as s (singlet), d (doublet), dd (double
doublet), t (triplet), q (quartet), or m (multiplet). HPLCꢀMS analysis
was performed using the Agilent 1100 LC/MSD SL separations module
and Mass Quad G1956B mass detector with electrospray ionization
(þve or ꢀve ion mode as indicated), and HPLC was performed using a
Zorbax SB-C18, Rapid Resolution HT cartridge, 4.6 mm ꢁ 30 mm, 1.8
μm i.d. column (Agilent P/N 823975-902) at a temperature of 40 °C
with gradient elution of 0ꢀ100% CH3CN (with 1 mL/L HCOOH)/
H2O (with 1 mL/L HCOOH) at a flow rate of 3 mL/min and a run time
of 2.8 min. Compounds were detected at 215 nm using a diode array
G1315B detector. All tested compounds gave g95% purity as deter-
mined by this method. All purified synthetic intermediates gave g95%
purity as determined by this method.
Ethyl 2,7-Dioxo-2,7-dihydro-3H-naphtho[1,2,3-de]quino-
line-1-carboxylate (1). Compound 1 was prepared according to the
method previously described.17
Figure 5. Binding mode of 1 in the active site of the ASK1 catalytic
subunit. Hydrogen bonds are shown by the dotted lines.
3-Methyl-1-(methylamino)-3H-naphtho[1,2,3-de]quino-
line-2,7-dione (2), 3-Methyl-1-(ethylamino)-3H-naphtho[1,2,3-
de]quinoline-2,7-dione (3), and 1-[(2-Hydroxyethyl)amino]-3-
methyl-3H-naphtho[1,2,3-de]quinoline-2,7-dione (5). Com-
pounds 2, 3, and 5 were prepared from 3-methylanthrapyridone-1-
sulfonic acid sodium salt (16) that was synthesized accordingly to the
original report.18
’ CONCLUSIONS
Recent investigations of the ASK1 role in the functioning of
the cellular signaling networks determined the connection bet-
ween increased activity of the kinase and development of several
pathologies. This provides the groundwork for the discovery of
small-molecule inhibitors of ASK1 which may be used in medical
2683
dx.doi.org/10.1021/jm200117h |J. Med. Chem. 2011, 54, 2680–2686

Journal of Medicinal Chemistry
ARTICLE
3-Methyl-1-(methylamino)-3H-naphtho[1,2,3-de]quinoli-
ne-2,7-dione (2). A solution of 16 (4.00 g, 0.011 mol) and sodium
acetate (2.71 g, 0.033 mol) in methylamine (40% in H2O, 400 mL) was
heated to reflux for 30 h. The mixture was filtered. The resulting solid
was taken up in H2O (400 mL). The suspension was heated to reflux for
5 min, filtered, and dried to give 2 (2.97 g, 0.010 mol, 93%) as light
brown needle crystals. LCꢀMS m/z 291 [M þ Hþ], tR = 0.97 min. 1H
NMR (DMSO-d6) δ 2.76 (d, J = 5.5 Hz, 3H), 3.84 (s, 3H), 7.23ꢀ7.31
(m, 1H), 7.50 (t, J = 7.3 Hz, 1H), 7.55 (t, J = 7.8 Hz, 1H), 7.69ꢀ7.83 (m,
3H), 8.12 (d, J = 7.7 Hz, 1H), 8.25 (d, J = 7.7 Hz, 1H).
3-Methyl-1-(ethylamino)-3H-naphtho[1,2,3-de]quinoline-2,
7-dione (3). A solution of 16 (4.00 g, 0.011 mol) and sodium acetate
(2.71 g, 0.033 mol) in ethylamine (40% in H2O, 400 mL) was heated to
reflux for 30 h. The mixture was filtered. The resulting solids were
dissolved in H2O (400 mL). The suspension was heated to reflux for 5
min, filtered, and dried to give 3 (2.74 g, 0.093 mol, 82%) as light brown
needle crystals. LCꢀMS m/z 305 [M þ Hþ], tR = 0.96 min. 1H NMR
(DMSO-d6) δ 1.05 (t, J = 5.5 Hz, 3H), 3.85 (s, 3H), 7.20ꢀ7.31 (m, 1H),
7.51 (t, J = 7.3 Hz, 1H), 7.55 (t, J = 7.8 Hz, 1H), 7.67ꢀ7.82 (m, 3H), 8.12
(d, J = 7.7 Hz, 1H), 8.25 (d, J = 7.7 Hz, 1H).
J = 8.1 Hz, 1H), 8.09 (d, J = 7.2 Hz, 1H), 8.39 (d, J = 7.7 Hz, 1H), 12.60
(s, 1H, NH).
1-Acetyl-4-methyl-3H-naphtho[1,2,3-de]quinoline-2,7-
dione (7). Yield 83% (white powder). LCꢀMS m/z 304 [M þ Hþ], tR =
0.95 min. 1H NMR (DMSO-d6) δ 2.60 (s, 3H), 2.67 (s, 3H), 7.60 (d, J =
7.7 Hz, 1H), 7.69ꢀ7.80 (m, 2H), 7.86 (d, J = 7.5 Hz, 1H), 8.01 (d, J = 7.6
Hz, 1H), 8.36 (d, J = 7.5 Hz, 1H), 11.75 (s, 1H, NH).
1-Acetyl-6-chloro-3H-naphtho[1,2,3-de]quinoline-2,7-di-
one (8). Yield 89% (white powder). LCꢀMS m/z 324 [M þ Hþ], tR =
0.98 min. 1H NMR (DMSO-d6) δ 2.67 (s, 3H), 7.63 (d, J = 8.8 Hz, 1H),
7.74ꢀ7.88 (m, 4H), 8.27 (d, J = 7.2 Hz, 1H), 12.64 (s, 1H, NH).
6-Bromo-1-(cyclohexylamino)-3-methyl-3H-naphtho[1,2,
3-de]quinoline-2,7-dione (10). Compound 10 was prepared from
1-(cyclohexylamino)-3-methyl-3H-naphtho[1,2,3-de]quinoline-2,7-
dione (17) that was synthesized by the method previously described.18
Bromine (240 mg, 1.5 mmol) in acetic acid (5 mL) was added dropwise
over 10 min to a solution of 17 (537 mg, 1.5 mmol) in acetic acid (5 mL).
After 1 h, the volatiles were removed under reduced pressure and the
residual solid was recrystallized from ethanol to give 10 (466 mg, 1.06
1
mmol, 71%). LCꢀMS m/z 438 [M þ Hþ], tR = 1.29 min. H NMR
1-[(2-Hydroxyethyl)amino]-3-methyl-3H-naphtho[1,2,3-
de]quinoline-2,7-dione (5). A solution of 16 (4.00 g, 0.011 mol),
sodium acetate (2.71 g, 0.033 mol), and ethanolamine (3.36 g, 0.055
mol) in H2O (400 mL) was heated to reflux for 20 h. The mixture was
filtered and solids were dried to give 5 (2.78 g, 0.009 mol, 79%) as yellow
needle crystals. LCꢀMS m/z 321 [M þ Hþ], tR = 0.98 min. 1H NMR
(DMSO-d6) δ 3.00ꢀ3.08 (m, 2H), 3.78 (t, J = 5.0 Hz, 2H), 3.86 (s, 3H),
3.75 (s, 1H, OH), 6.92 (t, J = 5.5 Hz, 1H), 7.51 (t, J = 7.5 Hz, 1H), 7.58
(t, J = 8.1 Hz, 1H), 7.75ꢀ7.82 (m, 2H), 8.00 (d, J = 8.1 Hz, 1H), 8.13 (d,
J = 7.7 Hz, 1H), 8.26 (d, J = 8.1 Hz, 1H).
1-Benzoyl-3-methyl-3H-naphtho[1,2,3-de]quinoline-2,7-
dione (6). A solution of 1-(methylamino)anthraquinone (9.49 g, 0.040
mol), sodium carbonate (424 mg, 4.0 mmol), and ethyl benzoylacetate
(7.69 g, 0.100 mol) in xylene (36 mL) was stirred at 140ꢀ150 °C for 8 h.
To complete the reaction, the ethanol and water formed during the
reaction were removed by azeotropic distillation with xylene. The
residue was cooled, and methanol (36 mL) was added. The mixture
was stirred at 30 °C for 30 min, cooled, filtered, washed with methanol
(36 mL), and dried to give 6 (13.31 g, 0.036 mol, 91%) as white needle
crystals. LCꢀMS m/z 366 [M þ Hþ], tR = 1.02 min. 1H NMR (DMSO-
d6) δ 3.78 (s, 3H), 7.49ꢀ7.60 (m, 3H), 7.62ꢀ7.72 (m, 2H), 7.78 (d, J =
8.1 Hz, 1H), 7.90ꢀ8.05 (m, 4H), 8.22 (d, J = 7.4 Hz, 1H), 8.36 (d, J = 7.8
Hz, 1H).
1-Acetyl-6-hydroxy-3H-naphtho[1,2,3-de]quinoline-2,7-
dione (9). A solution of 1-amino-4-hydroxyanthraquinone (9.57 g,
0.040 mol), sodium carbonate (424 mg, 4.0 mmol), and ethyl acetoa-
cetate (13.01 g, 0.100 mol) in xylene (36 mL) was stirred at 140ꢀ150 °C
for 8 h. To complete the reaction, the ethanol and water formed during
the reaction were removed by azeotropic distillation with xylene. The
residue was evaporated, and methanol (12 mL) was added. The mixture
was stirred at 30 °C for 30 min, cooled, filtered, washed with methanol
(12 mL), and dried to give 9 (9.38 g, 0.031 mol, 77%) as white crystals.
LCꢀMS m/z 306 [M þ Hþ], tR = 0.87 min. 1H NMR (DMSO-d6) δ
2.68 (s, 3H), 7.33 (d, J = 8.8 Hz, 1H), 7.66ꢀ8.02 (m, 4H), 8.51 (d, J = 7.4
Hz, 1H), 12.69 (s, 1H, OH), 13.60 (s, 1H, NH).
1-Acetyl-3H-naphtho[1,2,3-de]quinoline-2,7-dione (4),
1-Acetyl-4-methyl-3H-naphtho[1,2,3-de]quinoline-2,7-dio-
ne (7), and 1-Acetyl-6-chloro-3H-naphtho[1,2,3-de]quino-
line-2,7-dione (8). Compounds 4, 7, and 8 were prepared from
appropriate 1-aminoanthraquinones according to the method described
for above-mentioned compound 9.
(DMSO-d6) δ 0.98ꢀ1.86 (m, 10H), 3.22ꢀ3.36 (m, 1H), 6.31 (d, J = 9.9
Hz, 1H, NH), 7.53 (t, J = 7.4 Hz, 1H), 7.65 (d, J = 9.0 Hz, 1H), 7.71 (t, J
= 7.4 Hz, 1H), 7.76 (d, J = 8.8 Hz, 1H), 8.10ꢀ8.23 (dd, 2H).
General Procedures for the Synthesis of Compounds
11ꢀ14. Compounds 11ꢀ14 were prepared from amino-3-methyl-
3H-naphtho[1,2,3-de]quinoline-2,7-dione (18) that was synthesized
according to the method previously described.14 A solution of 18
(276 mg, 1.0 mmol) in pyridine (2 mL) was treated with the appropriate
sulfonyl chloride (1.0 mmol), and the mixture was stirred at room
temperature for 24 h. The reaction mixture was diluted with water
(10 mL) and filtered. The solid was washed with H2O (5 mL), isopropyl
alcohol (IPA) (2 mL) and air-dried.
N-(3-Methyl-2,7-dioxo-2,7-dihydro-3H-naphtho[1,2,3-de-
]quinolin-6-yl)benzenesulfonamide (11). Yield 89% (white
powder). LCꢀMS m/z 417 [M þ Hþ], tR = 1.04 min. 1H NMR
(DMSO-d6) δ 3.74 (s, 3H), 7.50ꢀ7.64 (m, 3H), 7.77 (t, J = 7.7 Hz, 1H),
7.85ꢀ7.97 (m, 4H), 7.97ꢀ8.09 (m, 2H), 8.40 (d, J = 7.8 Hz, 1H), 8.52
(d, J = 8.2 Hz, 1H), 12.83 (s, 1H, NH).
4-Fluoro-N-(3-methyl-2,7-dioxo-2,7-dihydro-3H-napht-
ho[1,2,3-de]quinolin-6-yl)benzenesulfonamide (12). Yield
93% (white powder). LCꢀMS m/z 435 [M þ Hþ], tR = 1.06 min.
1H NMR (DMSO-d6) δ 3.74 (s, 3H), 7.30 (t, J = 8.4 Hz, 2H), 7.77 (t, J =
7.7 Hz, 1H), 7.83ꢀ7.91 (m, 2H), 7.96ꢀ8.07 (m, 4H), 8.37 (d, J = 7.9
Hz, 1H), 8.50 (d, J = 8.2 Hz, 1H), 12.80 (s, 1H, NH).
4-Methyl-N-(3-methyl-2,7-dioxo-2,7-dihydro-3H-naph-
ztho[1,2,3-de]quinolin-6-yl)-3-nitrobenzenesulfonamide
(13). Yield 91% (yellow powder). LCꢀMS m/z 476 [M þ Hþ], tR =
1.08 min. 1H NMR (DMSO-d6) δ 3.74 (s, 3H), 7.66 (d, J = 8.1 Hz, 1H),
7.66 (t, J = 7.5 Hz, 1H), 7.81ꢀ7.91 (m, 2H), 7.96ꢀ8.07 (m, 2H), 8.11
(d, J = 7.8 Hz, 1H), 8.37 (d, J = 7.8 Hz, 1H), 8.46 (s, 1H), 8.49 (d, J = 7.8
Hz, 1H), 12.88 (s, 1H, NH).
2-Methyl-N-(3-methyl-2,7-dioxo-2,7-dihydro-3H-naphth-
o[1,2,3-de]quinolin-6-yl)benzenesulfonamide (14). Yield 88%
(white powder). LCꢀMS m/z 431 [M þ Hþ], tR = 1.09 min. 1H NMR
(DMSO-d6) δ 2.64 (s, 3H), 3.70 (s, 3H), 7.15ꢀ7.65 (m, 3H),
7.65ꢀ8.06 (m, 5H), 8.11 (d, J = 7.8 Hz, 1H), 8.42 (d, J = 7.8 Hz,
1H), 8.50 (d, J = 7.8 Hz, 1H), 13.05 (s, 1H, NH).
6-(1,3-Benzothiazol-2-ylthio)-3-methyl-3H-naphtho[1,2,
3-de]quinoline-2,7-dione (15). Compound 15 was prepared from
6-bromo-3-methyl-3H-naphtho[1,2,3-de]quinoline-2,7-dione 19 that
was synthesized according to the method previously described.19
A
1-Acetyl-3H-naphtho[1,2,3-de]quinoline-2,7-dione (4).
Yield 85% (white powder). LCꢀMS m/z 290 [M þ Hþ], tR = 0.91
min. 1H NMR (DMSO-d6) δ 2.68 (s, 3H), 7.66ꢀ7.84 (m, 4H), 7.87 (d,
solution of 19 (408 mg, 1.2 mmol), 2-mercaptobenzothiazole (200 mg,
1.2 mmol), and triethylamine (121 mg, 1.2 mmol) in DMF (5 mL) was
heated at 100 °C for 48 h. The mixture was diluted with H2O (15 mL)
2684
dx.doi.org/10.1021/jm200117h |J. Med. Chem. 2011, 54, 2680–2686

Journal of Medicinal Chemistry
ARTICLE
and filtered. The solid was washed with IPA (10 mL) and dried to
The stock solutions of inhibitors were prepared in DMSO, and the
concentration of inhibitor was 1 mM. The concentration of DMSO in
the reaction did not exceed 3%. Then 1 μL of inhibitor solution in
DMSO was added to each tube and mixed by pipetting. ATP solution
was prepared separately. For each sample 0.05 mCi γ-[P32]ATP was
taken (specific activity of 100 μCi/μM).
provide 15 (425 mg, 1.0 mmol, 83%). LCꢀMS m/z 427 [M þ Hþ], tR =
1
1.16 min. H NMR (DMSO-d6) δ 3.71 (s, 3H), 7.49ꢀ7.68 (m, 3H),
7.73ꢀ7.84 (m, 3H), 7.67 (t, J = 7.4 Hz, 1H), 8.08 (d, J = 7.9 Hz, 2H),
8.34 (d, J = 7.7 Hz, 1H), 8.48 (d, J = 8.2 Hz, 1H).
Receptor Preparation for Flexible Docking. The crystal
structure of human protein kinase ASK1 was obtained from the
Brookhaven Protein Data Bank (PDB code 2CLQ).20 The catalytic
subunit has been extracted from the PDB file, and the staurosporine has
been removed from the ASK1ꢀstaurosporine complex. In the crystal
structure some amino acids side chain atoms are missing. The homology
model of ASK1 has been built using the SWISS-MODEL workspace.21
A receptor molecule has been minimized in water with the GRO-
MACS molecular dynamics simulation package (GROMACS force field,
steepest descent algorithm, 1000 steps, em_tolerance = 100, em_step =
0.001). The partial atom charges of the receptor molecule were taken
from the Amber force field. Active site spheres were calculated with
DOCK sphgen software. The spheres with positions outside the ATP-
binding site were deleted manually. Connolly MS and Grid programs
from the DOCK package were used to generate receptor Connolly
surface and energy grids. Surface and grid calculations were performed
with parameter settings detailed in ref 22. The grid spacing was set to 0.3.
Ligand Database Processing. Calculations of ligand geometry
were performed using YFF force field described in ref 23. Partial atomic
charges of the ligands were calculated with the Kirchhoff method.24
Flexible Docking. DOCK program has been used for recep-
torꢀligand flexible docking. DOCK input parameters have been set as
described previously.22 In our virtual screening experiments the “multi-
ple anchors” parameter was set as follows: the minimum of heavy atoms
in the anchor was set to 6; the maximum number of orientations was set
to 1000; the “all atoms” model has been chosen. 172 compounds with
scores from ꢀ50 to ꢀ90 kcal/mol have been taken for the kinase assay
analysis.
The total concentration of labeled and unlabeled ATP was 100 μM.
The reaction was started with addition of ATP solution (150 μM ATP,
30 mM MgCl2, 15 mM MOPS, pH 7.2). The time of reaction was 20 min
at 30 °C. The reaction was stopped by adding 20 μL of 0.5 M
orthophosphoric acid. Then the reaction mixture was loaded on the
20 mm filter disks of the cellulose phosphate paper (“Whatman”, Great
Britain). Filters were washed three times with 0.075 M orthophosphoric
acid at room temperature and dried. For detection of products, dried
filters were counted by PerkinElmer model Tri-Carb 2800-TR liquid
scintillation analyzer. Then 1 μL of DMSO was added to the reaction
volume instead of the inhibitor stock solution for a positive control.
’ AUTHOR INFORMATION
Corresponding Author
*Phone/Fax: þ38 044 522 2458. E-mail: yarmoluksm@gmail.
com.
’ ACKNOWLEDGMENT
This work was supported by the National Academy of Sciences
of Ukraine (Grant 0107U000337).
’ ABBREVIATIONS USED
ASK1, apoptosis signal-regulating kinase 1; polyQ, polygluta-
mine; ROS, reactive oxygen species; Aβ, amyloid β; LPS, lipopo-
lysaccharide; CK2, protein kinase CK2;JNK3, c-Jun N-terminal
kinase 3; ROCK1, Rho-associated protein kinase 1; FGFR1, fi-
broblast growth factor receptor 1; HGFR, hepatocyte growth
factor receptor; Tie2, TEK tyrosine kinase, endothelial; MD,
molecular dynamics; IPA, isopropyl alcohol
The duration of the virtual screening experiment was around 9 days.
We used 21 processors, 3 GHz.
The structures were visualized using the program ViewerLite, version
4.2.25
Molecular Dynamics. The starting position of the ligand in the
active site of the receptor was generated with the DOCK software
package. GROMACS package and GROMOS96 force field were used
for experiments of MD.26 Energy minimization of molecular complexes
was carried out in explicit water environment with steepest descent
energy minimization algorithm for 1000 relaxation steps. The minimized
structure proceeded to position restrain dynamics, and the relaxation
time was 20 ps. On the next stage an MD of 10 ns was calculated. The
integration of the equations of motion was performed using the leapfrog
algorithm.27 In the MD simulation, the canonical (NVT) ensemble was
used. This ensemble involves keeping the number of particles (N),
system volume (V), and temperature (T) constant/conserved. There-
fore, no pressure coupling was used. V-rescale thermostat at 300 K was
taken to control the simulation temperature. For electrostatics treat-
ment, particle mesh Ewald algorithm was applied.28
’ REFERENCES
(1) Nishitoh, H.; Matsuzawa, A.; Tobiume, K.; Saegusa, K.; Takeda,
K.; Inoue, K.; Hori, S.; Kakizuka, A.; Ichijo, H. ASK1 is essential for
endoplasmic reticulum stress-induced neuronal cell death triggered by
expanded polyglutamine repeats. Genes Dev. 2002, 16, 1345–1355.
(2) Hashimoto, Y.; Niikura, T.; Chiba, T.; Tsukamoto, E.; Kadowaki,
H.; Nishitoh, H.; Yamagishi, Y.; Ishizaka, M.; Yamada, M.; Nawa, M.;
Terashita, K.; Aiso, S.; Ichijo, H.; Nishimoto, I. The cytoplasmic domain
of Alzheimer’s amyloid-beta protein precursor causes sustained apop-
tosis signal-regulating kinase 1/c-Jun NH2-terminal kinase-mediated
neurotoxic signal via dimerization. J. Pharmacol. Exp. Ther. 2003,
306, 889–902.
(3) Nishitoh, H.; Kadowaki, H.; Nagai, A.; Maruyama, T.; Yokota,
T.; Fukutomi, H.; Noguchi, T.; Matsuzawa, A.; Takeda, K.; Ichijo, H.
ALS-linked mutant SOD1 induces ER stress- and ASK1-dependent
motor neuron death by targeting Derlin-1. Genes Dev. 2008, 22,
1451–1464.
(4) Yamamoto, E.; Dong, Y. F.; Kataoka, K.; Yamashita, T.; Toku-
tomi, Y.; Matsuba, S.; Ichijo, H.; Ogawa, H.; Kim-Mitsuyama, S.
Olmesartan prevents cardiovascular injury and hepatic steatosis in
obesity and diabetes, accompanied by apoptosis signal regulating
kinase-1 inhibition. Hypertension 2008, 52, 573–580.
Biochemical Testing. Enzyme activity of human protein kinases
ASK1, Aurora A, ROCK1, HGFR, FGFR1, Tie2, JNK3, and CK2 was
determined using in vitro kinase assay (γ-32P-ATP method). Each
reaction mixture contained 6 μL of buffer solution (25 mM MOPS,
pH 7.2, 2.5 mM EGTA, 2.5 mM EDTA, 0.5 mM DTT, 0.25 mg/mL
BSA, 20 mM β-glycerophosphate), 3 μL of substrate solution (MBP,
Kemptide, Long S6 kinase substrate peptide, KKKSPGEYVNIEFG,
IGF-IRtide (12-527), TK substrate 2, JNK3tide, or RRRDDDSDDD for
each kinase, respectively) (5.0 μg/μL), 0.3 μL of enzyme (protein kinase
catalytic subunit, 0.1 μg/μL ≈ 32 mU/μL) (“Upstate/Millipore”), and
10.25 μL of H2O. The reaction mixture (total volume of 19 μL) was
quickly added to 1.5 mL tubes at room temperature.
(5) Hikoso, S.; Ikeda, Y.; Yamaguchi, O.; Takeda, T.; Higuchi, Y.;
Hirotani, S.; Kashiwase, K.; Yamada, M.; Asahi, M.; Matsumura, Y.;
Nishida, K.; Matsuzaki, M.; Hori, M.; Otsu, K. Progression of heart
2685
dx.doi.org/10.1021/jm200117h |J. Med. Chem. 2011, 54, 2680–2686

Journal of Medicinal Chemistry
ARTICLE
failure was suppressed by inhibition of apoptosis signal-regulating kinase
1 via transcoronary gene transfer. J. Am. Coll. Cardiol. 2007, 50, 453–462.
(6) Matsuzawa, A.; Saegusa, K.; Noguchi, T.; Sadamitsu, C.;
Nishitoh, H.; Nagai, S.; Koyasu, S.; Matsumoto, K.; Takeda, K.; Ichijo,
H. ROS-dependent activation of the TRAF6-ASK1-p38 pathway is
selectively required for TLR4-mediated innate immunity. Nat. Immunol.
2005, 6, 587–592.
(7) Chibon, F.; Mariani, J.; Derre, A.; Mairal, J. M.; Condre, L.;
Guillou, X.; Sastre, F.; Pedeteuor, F.; Aurias, A. ASK1 (MAP3K5) as a
potential therapeutic target in malignant fibrous histiocytomas with
12q14-q15 and 6q23 amplifications. Genes, Chromosomes Cancer 2004,
40, 32–37.
solvation: the GROMOS force-field parameter sets 53A5 and 53A6. J.
Comput. Chem. 2004, 25 (13), 1656–1676.
(27) Hockney, R. W.; Goel, S. P.; Eastwood, J. Quiet high resolution
computer models of a plasma. J. Comput. Phys. 1974, 14, 148–158.
(28) Darden, T.; Perera, L.; Li, L.; Pedersen, L. New tricks for
modelers from the crystallography toolkit: the particle mesh Ewald
algorithm and its use in nucleic acid simulations. Structure 1999,
7, 55–60.
(8) Uchikawa, O; Sakai, N.; Terao, Y.; Suzuki, H. WO 2008016131
A1, 2008; EP 2058309 A1, 2009; US 20100029619, 2010.
(9) Volynets, G. P.; Bdzhola, V. G.; Kukharenko, O. P.; Yarmoluk,
S. M. Identification of ASK1 small-molecule inhibitors. Ukr. Biokhim. Zh.
2010, 82, 41–50.
(10) Okamoto, M.; Saito, N.; Kojima, H.; Okabe, T.; Takeda, K.;
Ichijo, H.; Furuya, T.; Nagano, T. Identification of novel ASK1 inhibitors
using virtual screening. Bioorg. Med. Chem. 2011, 19, 486–489.
(11) Kitchen, D. B.; Decornez, H.; Furr, J. R.; Bajorath, J. Docking
and scoring in virtual screening for drug discovery: methods and
applications. Nat. Rev Drug Discovery 2004, 3, 935–949.
(12) Stoichet, B. K.; Stroud, R. M.; Santi, D. V.; Kuntz, I. D.; Perry,
K. M. Structure-based discovery of inhibitors of thymidylate synthase.
Science 1993, 259, 1445–1450.
(13) Bodian, D. L.; Yamasaki, R. B.; Buswell, R. L.; Stearns, J. F.;
White, J. M.; Kuntz, I. D. Inhibition of the fusion-inducing conforma-
tional change of influenza hemagglutinin by benzoquinones and hydro-
quinones. Biochemistry 1993, 32, 2967–2978.
(14) Ring, C. S.; Sun, E.; McKerrow, J. H.; Lee, G. K.; Rosenthal,
P. J.; Kuntz, I. D.; Cohen, F. E. Structure-based inhibitor design by using
protein models for the development of antiparasitic agents. Proc. Natl.
Acad. Sci. U.S.A. 1993, 90, 3583–3587.
(15) Ewing, T. J.; Makino, S.; Skillman, A. G.; Kuntz, I. D. DOCK
4.0: search strategies for automated molecular docking of flexible
molecule databases. J. Comput.-Aided Mol. Des. 2001, 15, 411–428.
(16) We used a database of commercially available compounds
(Otava Ltd., http://www.otavachemicals.com/).
(17) Bu, X.; Deady, L. W.; Finlay, G. J.; Baguley, B. C.; Denny, W. A.
Synthesis and cytotoxic activity of 7-oxo-7H-dibenz[f,ij]isoquinoline
and 7-oxo-7H-benzo[e]perimidine derivatives. J. Med. Chem. 2001,
44, 2004–2014.
(18) Kazankov, M. V.; Putsa, G. I. Synthesis of anthradipyridone
derivatives. Chem. Heterocycl. Compd. (N. Y., NY, U. S.) 1973,
9, 763–767.
(19) Allen, C. F. H.; Wilson, C. V. Nitroanthrapyridones. J. Org.
Chem. 1945, 10, 594–602.
(20) Bunkoczi, G.; Salah, H.; Filippakopoulos, P.; Fedorov, O.;
M€uller, S.; Sobott, F.; Parker, S. A.; Zhang, H.; Min, W.; Turk, B. E.;
Knapp, S. Structural and functional characterization of the human
protein kinase ASK1. Structure 2007, 15, 1215–1226.
(21) Homology model of ASK1 was built using the SWISS-MODEL
workspace (http://swissmodel.expasy.org/).
(22) Bursulaya, B. D.; Totrov, M.; Abagyan, R.; Brooks, C. L., III.
Comparative study of several algorithms for flexible ligand docking. J.
Comput.-Aided Mol. Des. 2003, 17, 755–763.
(23) Yakovenko, O. Ya.; Oliferenko, A. A.; Golub, A. G.; Bdzhola, V. G.;
Yarmoluk, S. M. The new method of distribution integrals evaluations for
high throughput virtual screening. Ukr. Bioorg. Acta 2007, 1, 52–62.
(24) Yakovenko, O. Ya.; Oliferenko, A. A.; Bdzhola, V. G.; Palyulin,
V. A.; Zefirov, N. S. Kirchhoff atomic charges fitted to multipole
moments: implementation for a virtual screening system. J. Comput.
Chem. 2008, 29, 1332–1343.
(25) The structures were visualized using the program ViewerLite,
version 4.2. (Accelrys Inc., http://www.accelrys.com).
(26) Oostenbrink, C.; Villa, A.; Mark, A. E.; Van Gunsteren, W. F. A
biomolecular force field based on the free enthalpy of hydration and
2686
dx.doi.org/10.1021/jm200117h |J. Med. Chem. 2011, 54, 2680–2686