Synthesis and SAR of 1,9-dihydro-9-hydroxypyrazolo[3,4-b]quinolin-4-ones as novel, selective c-Jun N-terminal kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2006.02.046

A novel class of 1,9-dihydro-9-hydroxypyrazolo[3,4-b]quinolin-4-ones as c-Jun-N-terminal kinase (JNK) inhibitors is described. These compounds were synthesized via the condensation of 2-nitrobenzaldehydes and hydroxypyrazoles. The structure-activity relationships (SAR) and kinase selectivity profile of the inhibitors are also discussed. Compound 16 was identified as a potent JNK inhibitor with good cellular potency.

Full text of DOI:10.1016/j.bmcl.2006.02.046

Bioorganic & Medicinal Chemistry Letters 16 (2006) 2590–2594
Synthesis and SAR of
1,9-dihydro-9-hydroxypyrazolo[3,4-b]quinolin-4-ones
as novel, selective c-Jun N-terminal kinase inhibitors
Mei Liu,* Zhili Xin, Jill E. Clampit, Sanyi Wang, Rebecca J. Gum, Deanna L. Haasch,
James M. Trevillyan, Cele Abad-Zapatero, Elizabeth H. Fry, Hing L. Sham and Gang Liu
Metabolic Disease Research, Global Pharmaceutical Research and Development, Abbott Laboratories,
100 Abbott Park Road, Abbott Park, IL 60064-6098, USA
Received 23 December 2005; revised 15 February 2006; accepted 16 February 2006
Available online 9 March 2006
Abstract—A novel class of 1,9-dihydro-9-hydroxypyrazolo[3,4-b]quinolin-4-ones as c-Jun-N-terminal kinase (JNK) inhibitors is
described. These compounds were synthesized via the condensation of 2-nitrobenzaldehydes and hydroxypyrazoles. The struc-
ture–activity relationships (SAR) and kinase selectivity profile of the inhibitors are also discussed. Compound 16 was identified
as a potent JNK inhibitor with good cellular potency.
Ó 2006 Elsevier Ltd. All rights reserved.
Type 2 diabetes has become the most prevalent meta-
bolic disease worldwide.1 Patients with type 2 diabetes
are insulin resistant, a condition in which the body fails
to respond to insulin properly. c-Jun N-terminal kinase-
1 (JNK1), a member of the mitogen-activated protein
(MAP) kinase family, has recently emerged as an attrac-
tive target for diabetes therapy, since JNK1 is believed
to play a key role in linking obesity and insulin resis-
tance.2 JNK1 is activated by inflammatory cytokines,
and negatively regulates insulin signaling via serine
(307) phosphorylation of the insulin receptor substrate
(IRS-1) which leads to the degradation of IRS-1.2,3
JNK1À/À mice show marked reduction in both plasma
glucose and insulin concentrations relative to their
wild-type littermates, and thus are protected from diet-
induced obesity.2a In addition, JNK1 activity is elevated
in adipocytes of type 2 diabetic patients.4 Inhibitors of
JNK1 can potentially increase insulin sensitivity, and
therefore could be useful as therapeutics for the treat-
ment of type 2 diabetes, as well as serving as valuable
tools for the evaluation of other clinical benefits that
these inhibitors may provide.5
Research related to JNK inhibitors has gained increas-
ing attention in the past few years, and as a result a
number of JNK inhibitors have recently appeared in
the literature.6 We wish to report here our efforts toward
the identification of 1,9-dihydro-9-hydroxypyrazolo[3,4-
b]quinolin-4-ones as novel ATP competitive, pan-JNK
inhibitors. A high-throughput screening (HTS) of the
Abbott compound collection identified two 9-hydroxy-
pyrazoloquinolinones 1, 2 (Fig. 1) as ATP-competitive
JNK1 inhibitors with IC50 values around 1 lM. JNK1
inhibitors need to be cell permeable to reach the intra-
cellular targets, therefore a phospho c-Jun cell-based
ELISA assay was also utilized to test the compounds
cellular potency.7 In this assay, compound 2 was mar-
ginally active with an IC50 of 16.4 lM. Additionally,
both compounds showed enzymatic IC50s of greater
than 50 lM against p38 and ERK2, two closely related
MAP kinases. The novelty of the core structure8 as well
O
O
CF3
N
Cl
Cl
N
N
N
N
N
OH
OH
1
2
Keywords: JNK; Diabetes; Obesity; Pyrazoloquinolinones.
*
Figure 1. 1,9-Dihydro-9-hydroxypyrazolo[3,4-b]quinolin-4-ones iden-
tified from HTS.
Corresponding author. Tel.: +1 847 935 7106; fax: +1 847 938
1674; e-mail: mei.liu@abbott.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.02.046

M. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2590–2594
2591
as its good selectivity profile motivated us to initiate hit
to lead chemistry to explore the activity of this class of
compounds as JNK1 inhibitors.
at the 6-position of the aromatic ring (Scheme 1). For
example, Suzuki coupling of 10 with arylboronic acids
yielded 12. In most cases only a simple filtration was
needed for purification of the final product.
9-Hydroxypyrazoloquinolinones were prepared in a
straightforward manner through the condensation of
2-nitrobenzaldehydes (6) and hydroxypyrazoles (5) in
the presence of concentrated hydrochloric acid
(Scheme 1). This method allowed for easy modifica-
tion of the N-1 and C-3 positions of the pyrazole ring,
as large numbers of hydroxypyrazoles were accessible
via either commercial sources or the condensation of
b-keto esters (3) with substituted hydrazines (4)
(Scheme 1). A number of substituted nitrobenzalde-
hydes (8) reacted smoothly with hydroxypyrazoles (5)
providing the desired products in good yield. How-
ever, 5-substituted nitrobenzaldehydes reacted poorly,
which resulted in very little or no product. Similarly,
6-bromosubstituted analogs (10) were prepared using
hydrogen bromide in acetic acid as the solvent. The
bromide served as a handle for further fuctionalization
The reaction mechanism for the formation of 1 is pro-
posed in Scheme 2. In the presence of an acid, hydroxy-
pyrazole 13, an enol, reacted with the aldehyde to give
II. As previously reported by Loudon in a mechanisti-
cally similar system,10 the next step was the addition
of chloride to the aromatic ring with subsequent rearo-
matization in acidic media providing the N-hydroxy-
isoxazoline derivative IV.11 This step was consistent
with the observation that 5-substituted nitrobenzalde-
hydes did not react or react poorly with hydroxypyraz-
oles, since the addition site was blocked by the
substituents at that position. Ring opening of IV pro-
duced nitroso intermediate V, which was subjected to
intramolecular addition of the hydroxypyrazole to give
the tricyclic system VI. Finally, intramolecular cycliza-
tion of VI, proton abstraction at the benzylic position
of VII, and subsequent dehydration of VIII would form
the desired pyrazoloquinolone product 1 (path A).
R1
O
Cl
N
Table 1. Enzymatic and cellular activity of analogs with aromatic ring
modifications
N
N
OH
O
O
R2
R1
N
b
c
R1
OEt
CHO
NO2
O
O
a
7
CF3
N
3
+
R
Cl
+
R1
N
O
N
HO
N
H
R2
Br
N
N
N
R2 NH2
N
N
R
OH
12
4
5
6
OH
9
N
N
R2
OH
R1 = CF3, R2 = CH3
10
b
CHO
R1 = R2 = CH3
Compound
R
JNK1
IC50 (lM)
Pc-Jun
IC50 (lM)
R
R-B(OH)2
11
d
NO2
1
—
—
1.22
0.98
0.92
>10
5.14
>10
>10
>10
>10
>10
>30
16.4
>30
NTa
>30
NTa
NTa
NTa
NTa
NTa
8
2
O
O
N
CF3
N
Cl
10
9a
9b
9c
—
R
N
5-Cl
7-Cl
N
N
N
R
OH
OH
8-OMe
7-N(Me)2
H
9
12
9d
12a
12b
12c
Scheme 1. Reagents and conditions: (a) cat. cont. HCl, EtOH, reflux,
overnight; (b) HCl/HOAc, 85 °C, 5 h, 50–60%; (c) HBr (33% in
HOAc)/H2O, 85 °C, 5 h, 52%; (d) FibreCat 1032 (5 mol%),9 K2CO3,
EtOH/H2O, microwave 120 °C, 20 min, 50% plus reduced product
(12a, R = H).
Ph
1H-Pyrazol-3-yl
a Compounds with enzymatic IC50s greater than 10 lM were not tested
in the cellular assay.
Cl-
OH
O
OH
N
N
N
N
H
N
N
N
N
Cl
Cl
H
O
N
O
O
H
O
O
N
OH
O
OH
O
O
N
O
N
O
N
O
N
O
N
O
H
6
13
I
II
III
IV
H
OH
OH
O
Path
A
O
N
N
N
H
N
N
OH
Cl
Cl
Cl
H
N
N
Cl
Cl
N
N
O
O
N
N
N
O
N
OH
N
OH
H
OH
N
OH
VII
OH
OH
V
VI
VIII
1
Path B
Scheme 2. Proposed mechanism for the formation of 9-hydroxypyrazolo[3,4-b]quinolin-4-one 1.

2592
M. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2590–2594
Alternatively, proton abstraction at the benzylic posi-
tion could occur first, followed by rearrangement to give
VIII directly, which then lead to the formation of the fi-
nal product 1 (path B).
chlorine was replaced with hydrogen, was not active
against JNK1, whereas there was only a slight decrease
in activity when chloride was replaced with bromide
(10). However, replacement of chlorine by an aryl group
at the same position gave only inactive compounds
(12b,c). Most analogs (9a–9d) with substitution at the
5, 7, and 8 positions of the ring were inactive as well.
All the compounds were tested in our JNK1 enzymatic
inhibition assay as well as a cell-based assay measuring
inhibition of TNFa stimulated phosphorylation of
c-Jun in HepG2 cells.12 The effect of different substitu-
tion on the phenyl ring was studied first, and these
results are summarized in Table 1. Analog 12a, in which
The role of the N-hydroxyl group was investigated
next. As summarized in Table 2, O-alkylated analogs
generally maintained or slightly decreased (2- to 4-fold)
binding potency when R was small alkyl groups (14a–
14c, Table 2), whereas bulky R group resulted in com-
Table 2. Enzymatic and cellular activity of 9-(alk)oxy analogs
O
O
Cl
Cl
NHBoc
COOH
N
N
O
N
N
N
OR
14
N
H
15
1.ClCO2Et/Et3N
O
Cl
Cl
2. NaN3
3. tBuOH
N
N
N
N
N
N
OH
7i
OH
Compound
R
JNK1 IC50 (lM) Pc-Jun IC50 (lM)
not obtained
1
H
1.22
4.59
2.78
5.43
>10
>30
19.4
10.6
5.6
NTa
NTa
NTa
NTa
O
14a
14b
14c
14d
14e
14f
15
Me
Et
Cl
NH
N
Pr
i-Pr
O
N
N
O
Cyclopentyl >10
actual product:16
JNK1 IC50 = 0.29 μM
Pc-Jun IC50 = 0.85 μM
Bn
—
>10
>10
a Compounds with enzymatic IC50s greater than 10 lM were not tested
in the cellular assay.
Figure 2. Cell permeable JNK inhibitor 16.
Table 3. Enzymatic and cellular activity of analogs with C-1 and N-3 modifications
R1
O
Cl
N
N
N
OH
7
R2
Compound
R1
R2
JNK1 IC50 (lM)
Pc-Jun IC50 (lM)
1
Me
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
1.22
0.96
1.31
0.52
2.64
>10
>10
>10
2.41
>10
2.74
2.83
0.75
0.78
0.50
0.43
4.41
7.64
8.77
7.67
4.19
1.75
>30
32
7a
7b
7c
7d
7e
7f
n-Pr
n-Bu
26.6
>30
>30
NTa
NTa
NTa
14.7
NTa
>30
>30
>30
>30
>30
>30
>30
>30
>30
>30
>30
>30
i-Pr
t-Bu
COOMe
4-NH2–Ph
CH2CH2OMe
(CH2)2CO2H
(CH2)3CO2H
CH2CH2NH2
(CH2)3CONHMe
(CH2)2NHCOMe
(CH2)3NHCONHEt
(CH2)2NHCO2Et
Me
7g
7h
7i
7j
7k
7l
7m
7n
7o
7p
7q
7r
7s
7t
CH2CO2H
CH2CO2Et
CH2CONHMe
CH2CH2OH
Et
Me
Me
Me
Me
7u
Me
Ph
a Compounds with enzymatic IC50s greater than 10 lM were not tested in the cellular assay.

M. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2590–2594
2593
M108
O
NH
S
E109
small specificity
pocket
NH
O
O
Cl
L110
N
O
M111
O
N
O
V40
NH
O
O
N
O
N
D112
O
N
NH
O
A113
N114
Figure 3. X-ray structure of compound 1.
plete loss of activity (14d–14f). Interestingly, compounds
with small alkyl groups exhibited increased cellular
activity, especially compound 14c (R = Pr), which
showed an almost 4-fold improvement in cellular po-
tency presumably due to increased cell permeability in
the absence of the acidic hydroxyl proton. Removal of
the hydroxyl group yielded inactive compound 15,
indicating that the hydroxyl group may serve as a
H-bond donor within the active site of the enzyme.
ture, the N-9 oxygen binds to the hinge region of JNK1
via a hydrogen bond with Met 111. The chlorine of the
aromatic ring extends toward the specificity pocket,14
thus additional interaction with the protein may result
from the substituents at C-5 and C-6, and possibly pro-
duce JNK1-selective inhibitors over JNK2 and JNK3.
On the other hand, the pyrazole ring protrudes from
the binding pocket, with N-1 and C-3 substituents
having minimal contact with the enzyme, therefore the
modification of these two positions showed minimal im-
pact on potency.
A number of compounds with different substitution at
the N-1 and C-3 positions of the pyrazole ring were
also prepared and tested (Table 3). Elongation of the
C-3 methyl group to ethyl (7a), propyl (7b), and butyl
(7c) groups appeared to have some impact on potency,
in particular 7c gave a 2.5 fold boost in JNK1 activity.
Branching at the a-carbon was not favorable, leading
to decreased JNK1 activity or inactive analogs
(7d–7g). To further optimize the SAR at the C-3 posi-
tions, a number of ethers, acids, amines, amides, ureas
as well as carbamates were prepared (see representative
examples 7h–7o in Table 3), which yielded several po-
tent compounds with submicromolar potency. How-
ever, most of these analogs showed no or minimal
cellular activity possibly due to low ClogP values
(<1.00) which may lead to low cell permeability. On
the other hand, the effect of N-1 substitution appeared
to be less profound. A variety of substituents (7p–7u)
including carboxylates, amides, alcohols, and other
hydrophobic groups were tolerated at this position,
but the fact that all analogs exhibited similar IC50s
seemed to suggest that it might not be a critical point
of contact with the enzyme.
The kinase selectivity of several pyrazoloquinolinone
JNK inhibitors was evaluated as well, and they generally
showed no significant activity at 10 lM against a panel
of 81 kinases.15 These compounds were also metaboli-
cally stable, for example, 100% of parent compound 1
was recovered after 30 min incubation with human liver
microsome. The oral bioavailability for compounds 1
and 2 was 15% and 56%, respectively.16
In summary, we have discovered a novel series of 1,9-
dihydro-9-hydroxypyrazolo[3,4-b]quinolin-4-ones
as
JNK inhibitors with submicromolar potency, good
pharmacokinetic profiles, and cell permeability. The
one-step procedure for the synthesis of 9-hydroxypyra-
zoloquinolinones greatly facilitated the SAR study for
this series of compounds. X-ray crystal structure data
suggest that further extension at the C-5 or C-6 position
of the aromatic ring, or core modification of this ring to
introduce H-bond acceptors may lead to more potent
and selective JNK inhibitors.
We also serendipitously discovered a novel tetracyclic
analog 16, which was obtained under Curtius rearrange-
ment conditions in an attempt to convert carboxylic acid
7i to amines (Fig. 2). To our delight, compound 16
showed binding IC50 of 0.29 lM with submicromolar
cellular potency, again suggesting that the molecule is
more cell permeable in the absence of the acidic
hydroxyl proton.
Acknowledgment
Dr. Charles Hutchins’ assistance in preparing the com-
puter image in Figure 3 is greatly appreciated.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.
2006.02.046.
An X-ray structure of 1 bound into the ATP site of
JNK1 has been obtained (Fig. 3)13. In the crystal struc-

2594
M. Liu et al. / Bioorg. Med. Chem. Lett. 16 (2006) 2590–2594
8. No similar core structures have been reported in the
References and notes
literature via substructure search in SciFinder.
9. FibreCat anchored homogeneous Pd catalyst purchased
from Johnson Matthey Catalysts.
1. Must, A.; Spadano, J.; Coakley, E. H.; Field, A. E.;
Colditz, G.; Dietz, W. H. JAMA 1999, 282, 1523.
2. (a) Hirosumi, J.; Tuncman, G.; Chang, L.; Go¨rgun, C. Z.;
¨
Uysal, K. T.; Maeda, K.; Karin, M.; Hotamisligil, G. S.
Nature 2002, 420, 333; (b) Bennett, B. L.; Satoh, Y.; Lewis,
A. J. Curr. Opin. Pharmacol. 2003, 3, 420; (c) Wellen, K.
E.; Hotamisligil, G. S. J. Clin. Invest. 2005, 115, 1111.
3. Aguirre, V.; Uchida, T.; Yenush, L.; Davis, R.; White, M.
F. J. Biol. Chem. 2000, 275, 9047.
4. Waeber, G.; Delplanque, J.; Bonny, C.; Mooser, V.;
Steinmann, M.; Widmann, C.; Maillard, A.; Miklossy, J.;
Dina, C.; Hani, E. H.; Vionnet, N.; Nicod, P.; Boutin, P.;
Froguel, P. Nat. Genet. 2000, 24, 291.
5. Manning, A. M.; Davis, R. J. Nat. Rev. Drug Disc. 2003,
2, 554.
10. (a) Loudon, J. D.; Wellings, I. J. Chem. Soc. 1960, 3470;
(b) Loudon, J. D.; Tennant, G. J. Chem. Soc. 1962, 3092.
11. Kim, J. N. have proposed a mechanism for the formation
of hydroxyquinoline N-oxide involving similar N-
hydroxyisoxazoline intermediate (a) Kim, J. N.; Lee, K.
Y.; Kim, H. S.; Kim, T. Y. Org. Lett. 2000, 2, 343; (b) Lee,
K. Y.; Kim, J. M.; Kim, J. N. Tetrahedron 2003, 59, 385.
12. Some of the potent analogs were also tested in a JNK2
inhibition assay. In general, they are about few folds less
potent against JNK2 as compared to JNK1. For example:
JNK2 IC50s for compounds 1 and 2 were 3.1 and 3.0 lM,
respectively.
13. Co-crystals of JNK1 complexed with 1 belong to space
˚
group P3221 with cell constants a = b = 150.6, c = 119.0 A
6. (a) Bain, J.; McLauchlan, H.; Elliott, M.; Cohen, P.
Biochem. J. 2003, 371, 199; (b) Gaillard, P.; Jeanclaude-
Etter, I.; Ardissone, V.; Arkinstall, S.; Cambet, Y.;
Camps, M.; Chabert, C.; Church, D.; Cirillo, R.;
Gretener, D.; Halazy, S.; Nichols, A.; Szyndralewiez,
C.; Vitte, P.-A.; Gotteland, J.-P. J. Med. Chem. 2005, 48,
4596; (c) Stocks, M. J.; Barber, S.; Ford, R.; Leroux, F.;
St.-Gallay, S.; Teague, S.; Xue, Y. Bioorg. Med. Chem.
Lett. 2005, 15, 3459; (d) Swahn, B.-M.; Huerta, F.;
Kallin, E.; Malmstro¨m, J.; Weigelt, T.; Viklund, J.;
˚
and diffracted weakly to about 3.5 A, using synchrotron
radiation (Advanced Photon Source, APS, IMCA-CAT,
17-ID). The structure was solved by molecular replace-
ment and refined by conventional methods. Crystallo-
graphic coordinates of JNK1 complexed with 1 have been
deposited at the Protein Data Bank (www.rcsb.org) with
Accession code 2g01.
14. (a) Scapin, G.; Patel, S. B.; Lisnock, J.; Becker, J. W.;
LoGrasso, P. V. Chem. Biol. 2003, 10, 705; (b) Traxler, P.
Exp. Opin. Ther. Patents 1998, 8, 1599.
15. A detailed kinase selectivity table was attached in the
Supplementary material for reference.
¨
Wonack, P.; Xue, Y.; Ohberg, L. Bioorg. Med. Chem.
Lett. 2005, 15, 5095.
7. Gupta, S.; Campbell, D.; Derijard, B.; Davis, R. J. Sci.
1995, 267, 389, Also see Supplementary material for
detailed assay procedures.
16. See Supplementary material for detailed PK data for
compounds 1 and 2.