Discovery of potent 1H-imidazo[4,5-b]pyridine-based c-Met kinase inhibitors via mechanism-directed structural optimization

Article abstract of DOI:10.1016/j.bmcl.2014.11.070

Starting from our previously identified novel c-Met kinase inhibitors bearing 1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-one scaffold, a global structural exploration was conducted to furnish an optimal binding motif for further development, directed by the

Full text of DOI:10.1016/j.bmcl.2014.11.070

Bioorganic & Medicinal Chemistry Letters 25 (2015) 708–716
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of potent 1H-imidazo[4,5-b]pyridine-based c-Met kinase
inhibitors via mechanism-directed structural optimization
a,b,
, Hongyan Liu b, , Zhong-Liang Xu , Yi Jin , Xia Peng , Ying-Ming Yao , Meiyu Geng b,
⇑
,
b
c
b
a,
⇑
Xiao-De An
b,
⇑
Ya-Qiu Long
a
Key Laboratory of Organic Synthesis of Jiangsu Province, College of Chemistry, Chemical Engineering and Materials Science, Soochow University, Suzhou 215123, China
State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203, China
School of Chemical Science and Technology, Yunnan University, Kunming 650091, China
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Starting from our previously identified novel c-Met kinase inhibitors bearing 1H-imidazo[4,5-
h][1,6]naphthyridin-2(3H)-one scaffold, a global structural exploration was conducted to furnish an
optimal binding motif for further development, directed by the enzyme inhibitory mechanism. First
round SAR study picked two imidazonaphthyridinone frameworks with 1,8- and 3,5-disubstitution
pattern as class I and class II c-Met kinase inhibitors, respectively. Further structural optimization on type
II inhibitors by truncation of the imidazonaphthyridinone core and incorporation of an N-phenyl cyclo-
propane-1,1-dicarboxamide pharmacophore led to the discovery of novel imidazopyridine-based c-Met
kinase inhibitors, displaying nanomolar enzyme inhibitory activity and improved Met kinase selectivity.
More significantly, the new chemotype c-Met kinase inhibitors effectively inhibited Met phosphorylation
and its downstream signaling as well as the proliferation of Met-dependent EBC-1 human lung cancer
cells at submicromolar concentrations.
Received 4 November 2014
Revised 24 November 2014
Accepted 26 November 2014
Available online 4 December 2014
Keywords:
c-Met tyrosine kinase
Small molecule inhibitor
Imidazonaphthyridinone
Imidazopyridine
Binding mode
Pyridone-3-carboxamide
Cyclopropane-1,1-dicarboxamide
Ó 2014 Elsevier Ltd. All rights reserved.
The MET tyrosine kinase is a cell surface receptor for the hepa-
tocyte growth factor (HGF), a pleiotropic cytokine controlling
pro-migratory, anti-apoptotic and mitogenic signals.1 Normal
activation of c-Met kinase is essential for wound healing and
embryonic development.2 Aberrant c-Met due to specific genetic
lesions, including transcriptional upregulation, gene amplification,
activating mutations, increased autocrine or paracrine ligand med-
U-shaped conformation to the ATP-binding site at the entrance of
the kinase pocket and wrap around Met1211, while type II inhibi-
tors bind to c-Met with an extended conformation that stretches
from the ATP-binding site (hinge region) to the deep hydrophobic
Ile 1145 pocket near the C-helix region. Some representative
chemical structures from each class are illustrated in Figure 1. In
general, type I inhibitors block c-Met kinase activity with high
selectivity against other kinases, whereas a majority of type II mol-
ecules are multikinase inhibitors, but expected to be more effective
against the mutations of c-Met active site that disrupt the type I
–4
5
iated stimulation, occurred in many types of cancers. Importantly,
deregulated c-Met activation has been associated with poor
6
,7
clinical outcomes. Furthermore, compelling evidence has linked
c-Met overactivation to mediating intrinsic or acquired resistance
1
6,17
binding mode.
8
,9
to targeted therapies. All of these emphasize c-Met as an attrac-
tive target for cancer therapy. Indeed, quite a few c-Met inhibitors
are now under clinical investment, which further proved the feasi-
bility of c-Met inhibition method in cancer therapy.
Recently, we identified a new class of type II c-Met kinase
inhibitors with 1,3,5-trisubstituted-1H-imidazo[4,5-h][1,6]naph-
thyridin-2(3H)-one scaffold (Fig. 2, exemplified by LXM-22), by
employing drug repurposing and pharmacophore incorporation
1
8
In the past decade, an impressive number of small molecule c-
Met inhibitors have been reported,10–14 which have basically been
categorized into two classes based on their structure and binding
modes in the c-Met kinase domain.2 Type I inhibitors bind in a
strategies.
The new scaffold inhibitors displayed promising
pharmacological property and cellular efficacy, but the inhibitory
activity fell in micromolar range. According to the binding mode
,15
18
predicted by molecular modeling,
the 1H-imidazo[4,5-
h][1,6]naphthyridin-2(3H)-one-based c-Met kinase inhibitor
adopts an extended conformation in the ATP-binding pocket of
the enzyme with the activation loop in an inactive, DFG-out
conformation. However, the heterotricyclic core resides in the
middle of the channel between the hinge region and the DFG motif,
⇑
Corresponding authors. Tel./fax: +86 21 50806876.
E-mail addresses: yaoym@suda.edu.cn (Y.-M. Yao), mygeng@simm.ac.cn
(
M. Geng), yqlong@simm.ac.cn (Y.-Q. Long).
These authors contributed to this work equally.
http://dx.doi.org/10.1016/j.bmcl.2014.11.070
960-894X/Ó 2014 Elsevier Ltd. All rights reserved.
0

X.-D. An et al. / Bioorg. Med. Chem. Lett. 25 (2015) 708–716
709
NH
Type l
N N
O
F
Cl
F
N
H
N
N
N
N
O
N
HN
NH
O
N
N
N
N
NH
2
N
Cl
N
N
N
AMG-337
PF2341066 (Crizotinib)
INCB028060
Type ll
H
N
H
O
O
N
H
H
N
H
N
F
N
N
O
O
O
N
F
O
O
O
F
O
O
O
O
F
F
O
O
Cl
N
N
H
2
N
N
O
XL-184 (Cabozantinib)
XL-880 (Foretinib)
BMS-777607
Figure 1. Representative structures of type I and type II c-Met kinase inhibitors launched or under clinical trials.
Hydrophobic interaction
within the hinge region
Type I
O
R2
5
N
N
7
8
N
R2
N
NH
O
O
N
R
3
1
N
O
N
Mechanism-based
HN
1
R1
N
structural modification
2a-g
1
a-g
N
N
O
N
Cl
O
N
Type II
F
NH
O
Cl
F
H
H
N
N
F
O
H
2
N
Asp1222
LXM-22
O
O
Hydrophobic pocket
N
F
O
H
N
R1
O
Solvent area
IC50 = 2.6 uM
R1
N
N
HN
N
N
H
3
a-f
4a-e
O
Figure 2. The mechanism-based structural optimization on the 1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-one scaffold to develop novel type I and II c-Met kinase inhibitors.
due to the bulky size, thus losing the key hydrogen bonding inter-
actions with Met1160. Herein, we tried to improve the potency of
imidazonaphthyridinone c-Met inhibitors based on the inhibitory
mechanism through two approaches: one is evolved into type I
inhibitor by using the 1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-
one as a U-shape inducing moiety with two proper substituents
deep into the solvent and hydrophobic regions, respectively
group furnished the cyclic urea 6. However, the approach to install
the 3-substituent was dependent on the structure of the coupling
partner. Mitsunobu reaction and nucleophilic substitution reaction
were applied for an alcohol and a halide to convert the 1H-imida-
zol-2(3H)-one compound 6 into 1,3-disubstituted intermediates
7(b, d, e) and 7(a, c), respectively. Further introduction of 5-substi-
tuent was achieved via Williamson etherification reaction between
5-bromo intermediates 7 and the corresponding alcohols to yield
8b–e or Buchwald–Hartwig cross coupling with amines to give
(
Fig. 2, 1a–g, 2a–g); the other is optimized into a bona fide type
1
II inhibitor by incorporating the typical pharmacophore of N -phe-
3
19
1
nyl-N -(4-fluorophenyl)malonamide or its bioisostere
and
8a. The removal of the N -4-methoxybenzyl group by TFA and
further truncation of the core structure to fit the ATP binding site
and DFG motif pocket (Fig. 2, 3a–f, 4a–e).
With respect to the U-shaped type I inhibitor design, the opti-
mal substitution pattern was investigated by synthesizing 3,5-
and 1,8-disubstituted analogs with hydrophobic and hydrophilic
groups incorporated rotationally at the two positions (1a–g, and
3
TfOH mixture and further demethylation by BBr afforded the
designed 3,5-disubstituted 1H-imidazo[4,5-h][1,6]naphthyridin-
2(3H)-one derivatives 1a–e. Interestingly, by tuning the ratio of
TfOH and TFA ranging from 10% to 50%, selective debenzylation
1
3
on the N -position without impacting N -benzyl group was
achieved. Notably, when a privileged structure of 1-(2,6-dichloro-
3-fluorophenyl)ethoxy group was installed at 5-position, a differ-
ent synthetic route was employed. The 5-bromo intermediate 7b
was converted into a phenol 10, followed by Mitsunobu reaction
with 1-(2,6-dichloro-3-fluorophenyl)ethanol to generate 1f–g.
As shown in Scheme 2, the 3,5-disubstituted type II inhibitors
3a–f were synthesized similarly, just appending a specified
structure of N-(3-fluoro-4-(oxyphenyl)-1-(4-fluorophenyl)-2-oxo-
1,2-dihydropyridine-3-carboxamide at 5-position via Williamson
2
a–g). As shown in Scheme 1, starting from the 1,6-naphthyridine
precursor, which was prepared according to the known proce-
2
0
dure, the introduction of the 4-methoxybenzylamino group to
-position was readily realized by the aromatic nucleophilic
substitution of 4-methoxybenzylamine. Then the 1H-imidazol-
(3H)-one core was constructed in a similar manner as we previ-
8
2
1
8
ously reported, whereby the Curtius rearrangement produced
an isocyanate and the following attack of the neighboring 8-amino

7
10
X.-D. An et al. / Bioorg. Med. Chem. Lett. 25 (2015) 708–716
Br
Br
R2
R2
Br
Br
a, b
c
N
d or e
N
f or g
N
h or h, i
N
N
N
OH
N
N
R1
R1
R1
NH
O
N
O
N
N
O
N
N
O
N
O
N
N
N
N
HN
O
N
HN
OTs
O
4
5
6
1a-e
7
a-e
8a-e
O
O
O
O
O
R2
OH
2
Br
R
N
N
h
k
e
N
j
N
N
N
O
N
N
O
N
N
N
N
N
N
O
N
N
O
N
N
N
N
N
HN
N
N
O
O
O
O
O
O
7b
9
10
a
8f-g
1f-g
O
O
O
O
1a-g
b
c
d
e
f
g
8
a-g
∗
∗
7
a-e
a
b
c
d
e
R1
N
∗
∗
∗
∗
∗
R1
N
O
∗
N
N
N
O
O
OH
NO
2
O
F
F
F
O
N
O
N
R2
H
O
O
N
N
Boc
Cl
Cl Cl
Cl
O
∗
O
∗
O
NH
∗
O
∗
O
∗
∗
Scheme 1. Reagent and conditions: (a) 4-methoxybenzylamine, TEA, THF, 60 °C, 10 h, 83% yield; (b) 2 N LiOH (aq), THF, 50 °C, 12 h, 74% yield; (c) DPPA, toluene, TEA, 100 °C,
0 h, 38% yield; (d) 4-methoxybenzyl chloride or tert-butyl 4-(2-bromoethyl)piperazine-1-carboxylate, K CO , DMF, rt–70 °C, 10–20 h (for 7a,7c); (e) ROH, DEAD, PPh , THF,
rt, 3 h (for 7b, 7d, 7e); (f) ROH, K CO , DMF, 110 °C, 1 h; (g) 1-(aminomethyl)-3-nitrobenzene, Pd(OAc) , dimethylbisdiphenylphosphinoxanthene, Cs
h, 36% yield; (h) TfOH : TFA = 10–50%, rt; (i) BBr , DCM, 0 °C–rt, 4 h; (j) benzyl alcohol, NaH, DMF, rt, 40 min; (k) H
1
2
3
3
2
3
2
2
CO
3
, 1,4-dioxane, 80 °C,
2
3
2
, Pd–C, rt, 30 min, 80% yield.
a
b
F
c, d
F
O
N
O
O
O
N
O
N
F
NH
O
O
O
N
O
O
F
F
OH
O
11
12
HO
13
O
N
Br
N
NH
O
O
N
F
e
f
F
NH
O
R1
O
F
N
N
N
O
O
R1
3a-f
a
b
c
d
e
f
N
N
O
N
N
∗
∗
R1
∗
∗
N
N
O
R1
H
N
N
O
∗
N
HN
O
7a-e
1
4a-e
3
a-f
N
H
O
N
OH
O
Scheme 2. Reagent and conditions: (a) 4-fluoroaniline, EDCI, DMAP, DMF, rt, 10 h, 65% yield. (b) 2 N NaOH (aq), rt, 3 h. (c) SOCl
fluorophenol, K CO , rt, 10 h, 30% overall yield for two steps. (e) 13, K CO , DMF, 110 °C, 1 h; (f) TfOH: TFA = 10–50%, rt.
2
, CH
2
2
Cl , cat. DMF, reflux, 3 h. (d) 4-Amino-2-
2
3
2
3
etherification reaction. This fragment was recognized as a typical
pharmacophore for type II c-Met inhibitors and was prepared
according to the literature method.
For the synthesis of 1,8-disubstituted-1H-imidazo[4,5-h][1,6]
naphthyridin-2(3H)-one derivatives, as described in Scheme 3,
see 6-methylpyridine-2,3-dicarboxylic acid was used as the
starting material subjected to the same procedures as pyridine-
reductive amination reaction, furnishing the 8-substituted target
compounds 2a–g.
21
Further truncated analogs (4a–d) were synthesized from 4/6-
chloro-3-nitropyridine-2-amine, as depicted in Scheme 4. Nucleo-
philic substitution by aminophenol followed by amidation enabled
the incorporation of N-(3-fluoro-4-hydroxyphenyl)-N-(4-fluoro-
phenyl)cyclopropane-1,1-dicarboxamide (21a–b). Reduction of the
nitro group by sodium borohydride in the presence of nickel chloride
afforded the pyridine-2,3-diamine key intermediate (22a–b),
which was converted to the 1H-imidazo[4,5-b]pyridin-2(3H)-one
2
1
,3-dicarboxylic acid, affording the key intermediate of 8-methyl-
H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-one 18.20 Oxidation
with SeO
2
produced aldehyde 19 to undergo a reduction or

X.-D. An et al. / Bioorg. Med. Chem. Lett. 25 (2015) 708–716
711
O
O
ref. 19
b
N
a
N
N
OH
OH
O
O
N
O
N
N
N
N
OH
O
N
OTs
O
HN
R1
O
15
16
1
7
N
N
f or g
c, d
e
N
N
NH
O
NH
O
NH
O
N
O
N
R
R2
N
R
2(a-g)
1
8
R
1
1
1
9
1
2
a
2b
2c
2d
2e
2f
2g
∗
∗
∗
∗
R1
R2
2
∗
∗
∗
HN
N
HN
HN
O
H
N
OH
H
N
∗
∗
N
O
N
OH
Scheme 3. Reagent and conditions: (a) TsCl, DCM, 60 °C, 2 h, 74% yield; (b) amines, TEA, THF, 100 °C, 10 h; (c) 2 N LiOH (aq), THF, 50 °C, 12 h; (d) DPPA, toluene, TEA, 80 °C,
6
2
h; (e) SeO
2
, 1,4-dioxane, 75–100 °C, 3 h; (f) NaBH
4
, ethanol, ice-salt, 10 min (for the synthesis of 2a), 70% yield; (g) RNH
2
, NaBH
3 3 3 4
CN, CH OH, CH COOH, anhydrous MgSO , rt,
h; (h) 2-(trimethylsilyl)-phenyltrifluoromethanesulfonate, CsF, CH
3
CN, 40 °C, 48 h, 43% yield.
H
H
N
H
N
H
N
F
H
N
H
H
2
N
F
F
N
N
F
Cl
O
O
NO
2
O
F
O
O
F
O
O
O
O
NO
2
NH
2
F
O
a
NO
d
H
N
b
2
c
N
NH
2
22a-b
4a-b
2
1a-b
O
N
NH2
a: 6-substituted
b: 4-substituted
N
NH2
a: 5-substituted
b: 7-substituted
2
0a-b
N
a: 6-substituted
b: 4-substituted
NH2
4/6-chloro-3-
N
N
H
a: 6-substituted
nitropyridin-2-amine
b: 4-substituted
H
H
N
H
N
H
N
F
N
F
f
H
N
H
O
O
N
F
O
O
O
F
O
O
H
N
e
F
O
4e
NH
O
2
OH
4c
O
2b
O
F
O
NH
2
N
NH
2
e,f
N
F
N
H
Cl
H
N
H
2
N
N
NH2
O
F
O
N
4d
N
N
Scheme 4. Reagents and conditions: (a) 4-amino-2-fluorophenol, t-BuOK, DMF, 80 °C, N
HOBt, EDC, DIEA, DMF, rt, 74–80% yield; (c) NiCl O, THF/MeOH (v:v = 1:1), NaBH , 0 °C, 85–92% yield; (d) bis(trichloromethyl)carbonate, cat. DMF, THF, 0 °C, 85–90%
ꢀ6H
yield; (e) 2-chloroacetyl chloride (for the preparation of 4c, 87% yield) or 5-bromopentanoyl chloride (for the preparation of 4d, yield 29%), TEA, THF, 0 °C; (f) AcOH, 110 °C,
3% yield.
2
, 80–85% yield; (b) 1-((4-fluorophenyl)carbamoyl)cyclopropanecarboxylic acid,
2
2
4
4
derivatives (4a–b) by reaction with bis(trichloromethyl)carbonate.
Treatment of the pyridine-2,3-diamine intermediate with haloalka-
noyl chloride delivered the 6,7,8,9-tetrahydroimidazo[1,2-a:5,4-
hand, the 1,8-disubstitution pattern with a heteroaromatic or
hydrophilic moiety at C-8 position and the hydrophobic group at
N-1 position turned out more effective for the Met kinase
inhibition (Table 1, 2a–g). When benzyl group was installed at
N-1 position and 2-methoxypyridin-4-amino or quinolin-6-amino
group at C-8 position, the resulting 1,8-disubstituted imidazo-
naphthyridinone derivatives displayed low micromolar inhibition
0
b ]dipyridine analog (4d) and the aminopyridine derivatives (4c, 4e).
All the synthesized 1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-
one derivatives were evaluated with the c-Met kinase inhibition.
An initial SAR study was focused on an optimal binding motif
exploration. For the designed type I inhibitors using imidazonaph-
thyridinone as the U-shape inducing module, both 3,5-disubstitu-
tion and 1,8-disubstitution pattern proved unfavorable for
binding to c-Met kinase domain, with the latter being slightly more
advantageous. As shown in Table 1, most of the 3,5-disubstituted
against c-Met enzyme (2b, IC50 = 3.9 lM; 2d, IC50 = 12.7 lM). The
extension of the phenyl ring by one more methylene from N-1
position (2e), or the replacement of the aromatic or heteroaromatic
ring at position N-1 or C-8 by aliphatic ring (2f) or heterocycle (2g)
abolished the Met inhibitory activity, indicating the important key
of the aromatic ring in the Met kinase binding.
1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-one derivatives (1a–g)
exhibited weak to moderate inhibition against Met enzyme, even
though the privileged structures were incorporated into N-3 and
C-5 positions. We selected isopropyl morpholine or phenol as the
hydrophilic substituent and the benzene or heteroaromatic ring
Gratifyingly, the attempt to optimize the imidazonaphthyridi-
none-based type II Met inhibitors proved successful when utilizing
the privileged structures presented by potent c-Met inhibitors
entering clinical trials (Table 2, 3a–f). For example, the fragment
0
as the hydrophobic group since they were frequently found in
of N,N -diphenyl cyclopropane-1,1-dicarboxamide was commonly
potent c-Met inhibitors.2
2,23
The best compound in this series
present in type II c-Met kinase inhibitors such as XL-184,
1
1
was exemplified by 1c with hydrophobic group at N-3 and
hydrophilic group at C-5 position (1c, IC50 = 26 M). On the other
XL-880. Furthermore, its conformationally constrained isostere,
that is, N-phenyl-2-pyridone-3-carboxamide gave favorable
l

7
12
X.-D. An et al. / Bioorg. Med. Chem. Lett. 25 (2015) 708–716
Table 1
c-Met enzymatic activity of 3,5- and 1,8-disubstituted 1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-one derivatives as type I inhibitorsa
R2
N
N
R1
N
NH
O
N
N
O
R2
N
R1
HN
1a-g
2a-g
Compd
R1
R2
c-Met inhibition
%
at 10
lM
IC50b
—
(lM)
NO2
OH
OH
OH
1
1
a
∗
∗
∗
33
∗
NH
F
b
30.6
—
O
∗
O
N
N
N
1
c
H
65.8
26.0 ± 1.8
O
O
∗
O
O
∗
∗
N
H
1
1
d
e
37.4
19.7
—
—
O
O
∗
N
O
N
N
O
∗
O
∗
∗
F
O
O
Cl
Cl
1
1
f
39.7
38.8
—
—
O
∗
F
N
Cl
Cl
g
O
∗
∗
∗
2
2
a
–OH
21.2
96.2
—
O
b
3.9 ± 0.2
HN
N
∗
∗
∗
∗
2
2
c
19.4
44.9
—
HN
OH
∗
HN
d
12.7 ± 4.8
N
∗
∗
HN
2
2
2
e
f
6.4
—
—
—
N
∗
∗
HN
O
1.4
∗
∗
HN
g
15.7
N
a
See Supporting information for the structural characterization of all tested compounds and a description of assay conditions.
IC50 values are reported as the mean of at least two independent determinations with eight concentrations each.
b

X.-D. An et al. / Bioorg. Med. Chem. Lett. 25 (2015) 708–716
713
Table 2
1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-one caused a big drop
of the enzyme potency (Table 3, 4a). But switching the dicarbox-
amide moiety from C-5 to C-7 position on the 1H-imidazo[4,5-
b]pyridin-2(3H)-one core remarkably enhanced the Met kinase
inhibition in both the biochemical and cell-based assays (4b,
IC50 = 209 nM, EBC-1 IC50 = 240 nM). Synthetic attempt to intro-
duce additional substituent on the imidazopyridine ring resulted
in the formation of aminopyridine derivatives, which displayed
improved biochemical potency but inferior cellular activity (4c,
c-Met enzymatic activity and antiproliferative effect of 3,5-disubstituted-1H-imi-
dazo[4,5-h][1,6]naphthyridin-2(3H)-one derivatives as type II inhibitorsa
O
N
F
NH
O
F
O
IC50 = 19.5 nM, 4e, IC50 = 20.2 nM; EBC-1, 10 lM > IC50 > 2 lM).
N
The inconsistence between the enzyme and cellular potency might
be attributed to the polar property of the molecule and thus poor
cell membrane permeability. The fused piperidino benzimidazole
analog gave similar biochemical inhibition (4d, IC50 = 260 nM)
but was 50-fold less potent in cells compared to the parent
N
R1
N
O
HN
Compd
R1
c-Met inhibition
ECB-1 IC50c
(lM)
1H-imidazo[4,5-b]pyridin-2(3H)-one inhibitor 4b.
%
at 10
lM
IC50b
(lM)
The cellular active inhibitors (4b–c,e) were further assessed for
3
a
H
20.3
—
—
their ability to inhibit c-Met signaling in the Met-driven EBC-1
cells. Phosphorylation levels of the Met protein in EBC-1 cell
lysates were determined by Western blotting after a 2 h incubation
period. As illustrated in Figure 3, phosphorylation of the Met
receptor was obviously inhibited by the selected compounds in a
concentration-dependent manner, consistent with the observed
antiproliferative activity in EBC-1 cells being a result of Met kinase
inhibition. Furthermore, Erk1/2 and AKT, the key downstream mol-
ecules of c-Met that play important roles in Met-driven cellular
N
NH
3
b
50.9
51.5
61.3
20.8
42.7
2.0 ± 0.1
3.6 ± 2.7
>50
>50
∗
OH
3c
3d
3e
3f
>50
∗
∗
N
—
O
N
O
—
—
∗
∗
4
,26
proliferation and survival,
were also significantly inhibited as
N
2.9 ± 0.1
19.8 ± 1.2
a result of our inhibitor treatment. These data support the finding
that compounds 4b–c,e inhibit c-Met signaling and, in turn, sup-
press c-Met dependent cell proliferation.
a
See Supporting information for the structural characterization of all tested
compounds and a description of assay conditions.
IC50 values are reported as the mean of at least two independent determinations
with eight concentrations each.
b
Since the parent 1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-
one scaffold was disclosed to possess multiple kinase inhibitory
c
Met-dependent human non-small lung cancer cell line, Met amplification.
18
activity, we were interested in investigating the kinase selectivity
of the newly developed 1H-imidazo[4,5-b]pyridin-2(3H)-one inhib-
itor. The lead compound 4b was selected for kinase profiling against
a panel of 15 oncogenic kinases (Table 4). Interestingly, at the con-
c-Met inhibitory activity both in enzyme and cell levels.21 There-
fore, we directly appended the 2-pyridone-3-carboxamide moiety
to the imidazonaphthyridinone core at C-5 position (Table 2),
meanwhile slimming the original 1,3,5-trisubstitution core of lead
LXM-22 into 3,5-disubstitution one. When a hydrophilic heterocy-
cle such as piperazine, piperidine or morpholine was attached on
the N-3 position through an alkyl chain, the imidazonaphthyridi-
none-based type II inhibitors displayed low micromolar inhibitory
centration of 1 lM, 4b was just effective in inhibiting the c-MET
subfamily member RON and phylogenetically related kinase Axl,
and EGFR, while it was almost inactive in all other 12 receptor
and nonreceptor kinases in the panel. This result demonstrated
the 1H-imidazo[4,5-b]pyridin-2(3H)-one scaffold achieved much
higher Met kinase selectivity than the original 1H-imidazo
1
8
[4,5-h][1,6]naphthyridin-2(3H)-one chemotype, thus warranting
a further preclinical development.
activity against c-Met enzyme (Table 2, 3b, IC50 = 2.0
IC50 = 3.6 M; 3f, IC50 = 2.9 M), providing a new motif for further
structural optimization of type II c-Met kinase inhibitors.
lM; 3c,
l
l
Based on the X-ray cocrystal structures of c-Met inhibitors
2
7
21
bound to the Met kinase domain (PDB: 3CCN, 3F82 ), we estab-
lished the binding modes of the new scaffold inhibitors (2b, 3f and
4b) by autodocking. As demonstrated by Figure 4, the 1,8-disubsti-
tuted-imidazonaphthyridinone based c-Met inhibitor 2b binds to
unphosphorylated c-Met kinase domain in a bend ‘U-shape’ con-
formation with the inhibitor wrapped around Met1211 as charac-
teristic of type I inhibitor. The [1,6]naphthyridine core of 2b is
sandwiched into a narrow cleft formed by the P-loop of the N-lobe
The global structural exploration on the imidazonaphthyridi-
none-based c-Met kinase inhibitors identified an optimal motif
bearing a pyridone-3-carboxamide pharmacophore at C-5 position
and a hydrophilic heterocycle at N-3 position as type II inhibitors.
However, this series of inhibitors gave disappointing results in the
Met-driven EBC-1 cell proliferation assay (Table 2). The EBC-1
human lung cancer cell line expresses high levels of constitu-
2
4,25
tively-activated Met kinase due to Met gene amplification.
and the activation loop of the C-lobe, forming stable p–p stacking
According to the binding mode of the best compound 3f in this
series, predicted by the molecular modeling, the imidazolone moi-
ety was involved in the H-bonding with the hinge region, but the
moiety of 2-pyridone-3-carboxamide was not properly positioned
in the DFG motif binding region (described in detail later in Figure 4
and the following binding mode interpretation). Therefore, further
structural optimization was steered toward the re-shaping of the
core and the side-chain structures.
Truncating the tricyclic core and replacing the rigid 2-pyridone-
carboxamide fragment with a relatively flexible isostere of cyclo-
propane-1,1-dicarboxamide were performed for an improved
activity (Table 3). Simply deduction of the pyridine ring from the
interaction with residue Tyr1230 and a H-bond between the car-
bonyl oxygen of [1,6]naphthyridine and the phenolic hydroxyl of
Tyr1230. The 2-methoxy-pyridin-4-amine side chain faces to the
solvent accessible region, and its amino unit makes an additional
hydrogen bond with the carboxyl oxygen of residue Asp1164.
As type II c-Met inhibitors, compounds 3f and 4b also occupy
the ATP-binding pocket but further extend into a second pocket
that is formed by the residues of the activation loop in an inactive,
DGF-out conformation. For compound 3f, the m-fluoroaniline
fragment resides in the middle of the tunnel between the hinge
region and DFG motif. Aromatic ring of [1,6]naphthyridine forms
stable
p–p stacking interaction with residue Tyr1159. The

7
14
X.-D. An et al. / Bioorg. Med. Chem. Lett. 25 (2015) 708–716
Table 3
c-Met enzymatic and cellular activities of 1H-imidazo[4,5-b]pyridin-2(3H)-one derivatives as type II inhibitorsa
Compd
Structure
c-Met inhibition
EBC-1 IC50c
(lM)
b
%
at 1.0
lM
IC50 (nM)
H
N
H
N
F
O
O
O
F
F
4
4
a
N
33%@ 10
lM
—
—
NH
O
HN
H
N
H
N
F
N
O
O
b
70
209 ± 45
0.24
O
NH
O
HN
H
N
H
N
F
N
O
4
4
c
O
O
O
O
Cl
84
71
19.5 ± 2.0
260 ± 35
<10
>10
F
F
O
N
H
NH
2
H
N
H
N
F
N
d
O
N
N
H
N
H
N
F
N
O
O
F
O
NH
2
4
e
87
20.2 ± 3.7
<10
NH
O
HO
a
b
c
See Supporting information for the structural characterization of all tested compounds and a description of assay conditions.
IC50 values are reported as the mean of at least two independent determinations with eight concentrations each.
Met-dependent human non-small lung cancer cell line, Met amplification.
Table 4
Effect of 4b on a panel of PTKs at 1.0 l
M concentrationa
RTK
Inhibition rate (%)
Met
70 (IC50: 209 nM)
Ron
Axl
FGFR1
FGFR2
ALK
43.2
41.6
7.4
37.2
2.5
Flt-1
KDR
13.3
0.0
IGF1R
PDGFR-
PDGFR-b
EGFR
ErbB4
c-Src
33.6
22.6
0.0
49.5
24.3
0.0
a
ABL
0.0
EPH-A2
18.8
a
The kinase profiling assay was conducted by using ELISA kinase assay. Each
reference compound and its inhibition rate for every kinase at the concentration of
M (unless otherwise specified) was indicated below: XL-184 (0.1 M) for RON
(100%) and Axl (100%); AZD4547 (10 nM) for FGFR1 (95.1%) and FGFR2 (82.51%);
AEW541 for IGF1R (56.6%); Su11248 for PDGFR- (68.4%), PDGFR-b (80.2%), Flt-1
66.6%), KDR (81.8%); BIBW2992 for EGFR (84.7%) and ErbB4 (84.2%); Dasatinib for
c-Src (63.2%), ABL (56.7%) and EPH-A2 (76.2%).
Figure 3. Effects of the selected compounds on Met phosphorylation and down-
stream signaling in Met-driven lung cancer EBC-1 cells.
1
l
l
a
(
backbone carbonyl and N–H of residue Met1160 form two hydro-
gen bonds with the N–H of imidazole ring and nitrogen atom of
[
1,6]naphthyridine, respectively. The oxygen atom of pyridone
fragment forms an additional hydrogen bond with backbone N–H
of residue Asp1222. The p-fluoroaniline tail falls into the hydro-
phobic pocket formed by residues Ala1221, Leu1140, Val1139
and Leu1195 in DFG motif region. The potent and selective
c-Met inhibitor 4b adopts a similar binding pattern to 3f. Signifi-
cantly, the simplified 1H-imidazo[4,5-b]pyridin-2(3H)-one core is
responsible for the strong interaction with c-Met by forming two
hydrogen bonds with the hinge residue Met1160 as well as the
pi-stacking interaction with Tyr1159. Furthermore, the two amide
groups of the cyclopropane-1,1-dicarboxamide moiety are
involved in three H-bonds with Asp1222 and Lys1110, which
greatly enhances the potency. Hydrophobic interactions of the
p-fluoroaniline moiety with DFG motif pocket also contribute to
the tight binding.

X.-D. An et al. / Bioorg. Med. Chem. Lett. 25 (2015) 708–716
715
Figure 4. Proposed binding modes of selected compounds 2b, 3f and 4b to the Met kinase domain, based on the X-ray co-crystal structure of c-Met in complex with a small
molecule inhibitor.2
1,27
In conclusion, starting from our newly identified imidazonaph-
thyridinone c-Met inhibitors, based on the inhibitory mechanism,
we conducted a global structural exploration with respect to the
substitution pattern and the core structure size to furnish an opti-
mal scaffold with improved potency and selectivity. A series of 3,5/
References and notes
1
2
3
4
.
.
.
.
Bottaro, D. P.; Rubin, J. S.; Faletto, D. L.; Chan, A. M. L.; Kmiecik, T. E.;
Vandewoude, G. F.; Aaronson, S. A. Science 1991, 251, 802.
Gherardi, E.; Birchmeier, W.; Birchmeier, C.; Woude, G. V. Nat. Rev. Cancer 2012,
12, 89.
Naldini, L.; Vigna, E.; Narsimhan, R. P.; Gaudino, G.; Zarnegar, R.;
Michalopoulos, G. K.; Comoglio, P. M. Oncogene 1991, 6, 501.
Birchmeier, C.; Birchmeier, W.; Gherardi, E.; Vande Woude, G. F. Nat. Rev. Mol.
Cell Biol. 2003, 4, 915.
1
,8-disubstituted-1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-one
derivatives were designed, synthesized and evaluated with privi-
leged structures incorporated into N-1/C-8 or N-3/C-5 positions.
3
,5-Disubstituted 1H-imidazo[4,5-h][1,6]naphthyridin-2(3H)-one
5. Migliore, C.; Giordano, S. Eur. J. Cancer 2008, 44, 641.
6
7
8
.
.
.
Trusolino, L.; Bertotti, A.; Comoglio, P. M. Nat. Rev. Mol. Cell Biol. 2010, 11, 834.
Comoglio, P. M.; Giordano, S.; Trusolino, L. Nat. Rev. Drug Disc. 2008, 7, 504.
Engelman, J. A.; Zejnullahu, K.; Mitsudomi, T.; Song, Y. C.; Hyland, C.; Park, J. O.;
Lindeman, N.; Gale, C. M.; Zhao, X. J.; Christensen, J.; Kosaka, T.; Holmes, A. J.;
Rogers, A. M.; Cappuzzo, F.; Mok, T.; Lee, C.; Johnson, B. E.; Cantley, L. C.; Janne,
P. A. Science 2007, 316, 1039.
turned out novel type II c-Met inhibitors with low micromolar
IC50 values. Further truncation of the core structure afforded potent
and selective 1H-imidazo[4,5-b]pyridin-2(3H)-one Met inhibitors
with nanomolar inhibition against the enzyme and the growth of
Met-dependent EBC-1 human lung cancer cells, providing a prom-
ising new lead for further structural optimization and development
of targeted cancer therapeutics.
9.
Straussman, R.; Morikawa, T.; Shee, K.; Barzily-Rokni, M.; Qian, Z. R.; Du, J. Y.;
Davis, A.; Mongare, M. M.; Gould, J.; Frederick, D. T.; Cooper, Z. A.; Chapman, P.
B.; Solit, D. B.; Ribas, A.; Lo, R. S.; Flaherty, K. T.; Ogino, S.; Wargo, J. A.; Golub, T.
R. Nature 2012, 487, 500.
1
0. Zhu, K. K.; Kong, X. Q.; Zhao, D.; Liang, Z. J.; Luo, C. Expert Opin. Ther. Patents
Acknowledgments
2014, 24, 217.
1
1
1
1. Porter, J. Expert Opin. Ther. Patents 2010, 20, 159.
2. Cui, J. J. Expert Opin. Ther. Patents 2007, 17, 1035.
3. Liu, X. D.; Newton, R. C.; Scherle, P. A. Expert Opin. Invest. Drugs 2011, 20, 1225.
This work was supported by grants from National Science
Foundation of China (81325020, 21302201, 81321092 and
14. Cui, J. J. J. Med. Chem. 2014, 57, 4427.
1
1
5. Dussault, I.; Bellon, S. F. Anti-Cancer Agents Med. Chem. 2009, 9, 221.
6. Berthou, S.; Aebersold, D. M.; Schmidt, L. S.; Stroka, D.; Heigl, C.; Streit, B.;
Stalder, D.; Gruber, G.; Liang, C. X.; Howlett, A. R.; Candinas, D.; Greiner, R. H.;
Lipson, K. E.; Zimmer, Y. Oncogene 2004, 23, 5387.
9
1229205), National Program on Key Basic Research Project of
China (No. 2012CB910704), National S&T Major Projects of China
2012ZX09301001-007) and Science and Technology Commission
(
of Shanghai Municipality, China (13ZR1447700).
17. Norman, M. H.; Liu, L. B.; Lee, M.; Xi, N.; Fellows, I.; D’Angelo, N. D.;
Dominguez, C.; Rex, K.; Bellon, S. F.; Kim, T. S.; Dussault, I. J. Med. Chem.
2
012, 55, 1858.
Supplementary data
18. Wang, Y.; Xu, Z. L.; Ai, J.; Peng, X.; Lin, J. P.; Ji, Y. C.; Geng, M. Y.; Long, Y. Q. Org.
Biomol. Chem. 2013, 11, 1545.
1
9. Williams, D. K.; Chen, X. T.; Tarby, C.; Kaltenbach, R.; Cai, Z. W.; Tokarski, J. S.;
An, Y. M.; Sack, J. S.; Wautlet, B.; Gullo-Brown, J.; Henley, B. J.; Jeyaseelan, R.;
Kellar, K.; Manne, V.; Trainor, G. L.; Lombardo, L. J.; Fargnoli, J.; Borzilleri, R. M.
Bioorg. Med. Chem. Lett. 2010, 20, 2998.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.11.
0
70.

7
16
X.-D. An et al. / Bioorg. Med. Chem. Lett. 25 (2015) 708–716
2
0. Zeng, L. F.; Wang, Y.; Kazemi, R.; Xu, S. L.; Xu, Z. L.; Sanchez, T. W.; Yang, L. M.;
Debnath, B.; Odde, S.; Xie, H.; Zheng, Y. T.; Ding, J.; Neamati, N.; Long, Y. Q. J.
Med. Chem. 2012, 55, 9492.
1. Schroeder, G. M.; An, Y. M.; Cai, Z. W.; Chen, X. T.; Clark, C.; Cornelius, L. A. M.;
Dai, J.; Gullo-Brown, J.; Gupta, A.; Henley, B.; Hunt, J. T.; Jeyaseelan, R.; Kamath,
A.; Kim, K.; Lippy, J.; Lombardo, L. J.; Manne, V.; Oppenheimer, S.; Sack, J. S.;
Schmidt, R. J.; Shen, G.; Stefanski, K.; Tokarski, J. S.; Trainor, G. L.; Wautlet, B. S.;
Wei, D.; Williams, D. K.; Zhang, Y. R.; Zhang, Y. P.; Fargnoli, J.; Borzilleri, R. M. J.
Med. Chem. 2009, 52, 1251.
2. Wang, Y. X.; Ai, J.; Wang, Y.; Chen, Y.; Wang, L.; Liu, G.; Geng, M. Y.; Zhang, A. J.
Med. Chem. 2011, 54, 2127.
3. Zhang, D. Y.; Ai, J.; Liang, Z. J.; Zhu, W.; Peng, X.; Chen, X. J.; Ji, Y. C.; Jiang, H. L.;
Luo, C.; Geng, M. Y.; Liu, H. Bioorg. Med. Chem. Lett. 2013, 23, 2408.
24. Park, M.; Dean, M.; Cooper, C. S.; Schmidt, M.; Obrien, S. J.; Blair, D. G.;
Vandewoude, G. F. Cell 1986, 45, 895.
25. Sattler, M.; Pride, Y. B.; Ma, P.; Gramlich, J. L.; Chu, S. C.; Quinnan, L. A.;
Shirazian, S.; Liang, C. X.; Podar, K.; Christensen, J. G.; Salgia, R. Cancer Res.
2003, 63, 5462.
26. Bertotti, A.; Burbridge, M. F.; Gastaldi, S.; Galimi, F.; Torti, D.; Medico, E.;
Giordano, S.; Corso, S.; Rolland-Valognes, G.; Lockhart, B. P.; Hickman, J. A.;
Comoglio, P. M.; Trusolino, L. Sci. Signal. 2009, 2, ra80.
27. Albrecht, B. K.; Harmange, J. C.; Bauer, D.; Berry, L.; Bode, C.; Boezio, A. A.; Chen, A.;
Choquette, D.; Dussault, I.; Fridrich, C.; Hirai, S.; Hoffman, D.; Larrow, J. F.; Kaplan-
Lefko, P.; Lin, J.; Lohman, J.; Long, A. M.; Moriguchi, J.; O’Connor, A.; Potashman, M.
H.; Reese, M.; Rex, K.; Siegmund, A.; Shah, K.; Shimanovich, R.; Springer, S. K.;
Teffera, Y.; Yang, Y. J.; Zhang, Y. H.; Bellon, S. F. J. Med. Chem. 2008, 51, 2879.
2
2
2