Discovery of dihydroquinoxalinone acetamides containing bicyclic amines as potent Bradykinin B1 receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2008.07.055

Replacement of the core β-amino acid in our previously reported piperidine acetic acid and β-phenylalanine-based Bradykinin B1 antagonists by dihydroquinoxalinone acetic acid increases the in vitro potency and metabolic stability. The most potent compounds from this series have IC₅₀s < 0.2 nM in a human B1 receptor functional assay. A molecular modeling study of the binding modes of key compounds, based on a B1 homology model, explains the structure-activity relationship (SAR) for these analogs.

Full text of DOI:10.1016/j.bmcl.2008.07.055

Bioorganic & Medicinal Chemistry Letters 18 (2008) 4477–4481
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of dihydroquinoxalinone acetamides containing bicyclic amines
as potent Bradykinin B1 receptor antagonists
a
a
b
d
Jian Jeffrey Chen a, , Wenyuan Qian , Kaustav Biswas , Vellarkad N. Viswanadhan , Benny C. Askew ,
*
Stephen Hitchcock a, Randall W. Hungate a, Leyla Arik c, Eileen Johnson c
a Chemistry Research and Development, Amgen Inc., One Amgen Center Drive, MS-B29-4-1-B, Thousand Oaks, CA 91320, USA
b Molecular Structure and Design, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320, USA
c Neuroscience, Amgen Inc., One Amgen Center Drive, Thousand Oaks, CA 91320, USA
d Department of Medicinal Chemistry, Serono Research Institute, Rockland, MA, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Replacement of the core b-amino acid in our previously reported piperidine acetic acid and b-phenylal-
anine-based Bradykinin B1 antagonists by dihydroquinoxalinone acetic acid increases the in vitro
potency and metabolic stability. The most potent compounds from this series have IC50s < 0.2 nM in a
human B1 receptor functional assay. A molecular modeling study of the binding modes of key com-
pounds, based on a B1 homology model, explains the structure–activity relationship (SAR) for these
analogs.
Received 26 June 2008
Revised 9 July 2008
Accepted 14 July 2008
Available online 17 July 2008
Keywords:
Dihydroquinoxalinone acetamides
Bradykinin B1
Ó 2008 Elsevier Ltd. All rights reserved.
Receptor antagonist
Pain
Inflammation
Sulfonamides
Kinins are released at sites of tissue injury inducing pain and
inflammation via activation of two G protein-coupled receptors
termed Bradykinin B2 and B1. The B2 subtype is expressed in most
cell types. On the other hand, the B1 receptor is absent or ex-
pressed at very low levels under normal conditions and is induced
during tissue injury and inflammation. It has been implicated in
the humoral, cellular, and neuronal mechanisms underlying
chronic inflammatory pain.1,2 Peripherally restricted B1 peptide
antagonists are efficacious in reversing neurogenic pain induced
by capsaicin as well as inflammatory pain induced by UV irradia-
tion, carrageenan, CFA or LPS. B1 receptor knockout mice exhibit
hypoalgesia to chemical and thermal noxious stimuli. They show
attenuated inflammatory responses and reduced neutrophil accu-
mulation following tissue injury. Therefore, B1 receptor antago-
nists are potentially useful in the treatment of chronic pain and
inflammation.3–5
first generation leads served as excellent proof of concept mole-
cules that showed in vitro and in vivo efficacy in B1 antagonism as-
says. Subsequent studies focused on further improving the in vitro
functional potency, the metabolic stability, and reducing the
human plasma protein binding.
Recently, Su and co-workers reported that dihydroquinoxali-
none-based compounds such as 3 and its analogs are highly potent
B1 antagonists.9 Molecular modeling studies of
3 and its
analogs,9–12 using a B1 homology model, revealed putative binding
modes of these compounds. These models were supported by data
from site-directed mutagenesis of B1 receptor.11
Through high throughput screening we identified a related
compound,
2-(1-(mesitylsulfonyl)-3-oxo-1,2,3,4-tetrahydro-
noxalin-2-yl)-N-(2-methoxyphenyl)acetamide 4, as a weak B1
antagonist in the receptor binding assay (Ki = 500 nM). Struc-
turally these compounds are benzo-fused b-amino acid deriva-
tives. It would be interesting to investigate whether the
replacement of the right-hand amine in 4 with the chiral bicy-
clic amines from 1 to 2 might result in enhanced potency. Our
efforts led to a number of highly potent (IC50s < 0.2 nM) meta-
bolically more stable compounds. Furthermore, they are less
protein bound than previous compounds. Herein, we report
the detailed synthesis, SAR, and molecular modeling studies
of this new class of B1 antagonists.
We recently disclosed a series of sulfonylated b-amino acid
derivatives containing chiral chroman and tetralin diamine moie-
ties, such as 16,7 and 2,8 respectively, as novel B1 antagonists. These
compounds were generated from conformational restriction of
acyclic amine-based initial hits to improve in vitro potency. These
* Corresponding author.
E-mail address: jianc@amgen.com (J.J. Chen).
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.07.055

4478
J. J. Chen et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4477–4481
F3C
SO2
N
SO2
HN
H
H
N
N
O
O
O
2
1
Ki =23 nM
IC50 = 13 nM
ref. 8
N
Ki = 0.3 nM
IC50 = 1.2 nM
ref. 6 and 7
N
H
Cl
Cl
SO2
N
H
SO2
OMe
H
N
4
N
N
Ki = 500 nM
IC50 > 40000 nM
H
O
N
O
N
H
O
N
H
O
N
3
Ki = 0.034 nM
IC50 = 0.18 nM
ref. 9
Ar
SO2
H
N
N
O
X
N
H
O
5
NR1R2
Scheme 1 outlines the general synthetic routes for the dihydro-
quinoxalinone acetamides 5. The dihydroquinoxalinone acetate
was prepared according to modified literature.13 2-Fluoronitroben-
Encouraged by these findings, we investigated the effect of
introducing various amino groups. The primary amine 5a had an
IC50 = 1.2 nM and is about 52-fold more potent than the corre-
sponding alcohol 12 in the functional assay. Interestingly, the
introduction of a single NH2 group not only improves the potency
dramatically but also has the beneficial effect of increasing the
zene 6 was condensed with commercially available (D)-aspartic
acid dimethyl ester 7 to give the SNAr adduct. Hydrogenation of
the nitro group followed by intramolecular cyclization gave methyl
dihydroquinoxalinone acetate 8. Sulfonylation of 8 with different
aryl sulfonyl chlorides in neat pyridine yielded the sulfonamides
9. Under normal basic ester hydrolysis conditions, the sulfona-
mides in 9 are prone to elimination and racemization. Acid-cata-
lyzed hydrolysis avoided these side reactions and gave the
desired acids 10, although the reaction was often slow and did
not go to completion. The crude acids were used directly in the
next coupling step without any trouble.
metabolic stability (e.g., compare HLM value of 264 vs 127 lL/
min/mg for alcohol 12 and amine 5a). Mono-methylation of NH2
improves the binding affinity slightly but has minimal effect on
functional potency (compare 5a and 5b, Table 1). Further methyl-
ation of the amino group to give 5c did not yield any significant
improvement in potency. The metabolic stability was, however,
compromised.
Replacement of the NHMe of 5b with a more lipophilic amine
moiety such as those in 5d–g improved the binding potency only
slightly. For example, the piperidine analog 5g is only 8-fold more
potent than 5b (compare 5b, 5d–g, Table 1). Interestingly, in the
b-phenylalanine series such as 1, the potency increases with lipo-
philicity on the right-hand side amino group. For example, replace-
ment of NHMe by piperidine improves the binding affinity by 460-
fold.7 In order to gain better understanding of the interactions be-
tween these antagonists and the receptor, we carried out the
molecular modeling of these compounds using a B1 homology
model.
The amino bicyclic chroman and tetralin alcohol 116–8 were
coupled with the acid 10 to give the amide 12. The oxidation of
the benzylic alcohol 12 with MnO2 yielded the aldehyde 13. Reduc-
tive amination of 13 with a variety of primary and secondary
amines gave the desired product 5. The piperidine derivatives
5g–h and 5j–k were prepared from the coupling of the acid 10
and bicyclic piperidine 15.
The in vitro binding (Ki) and functional (IC50) assay data with
the new antagonists are shown in Table 1. Earlier work in the b-
phenylalanine series such as 1 indicated that the basic amino
group is required for potency.7 Interestingly, the intermediate alco-
hol 12 and aldehyde 13 were moderately active in both binding
and functional assays, with 13 (IC50 = 23 nM) about 3-fold more
potent than 12 (IC50 = 62 nM). These results suggest that dihydro-
quinoxalinone acetamides could provide superior potency over the
b-phenylalanine core.
The recently proposed homology models of the B1 receptor by
Ha and co-workers,10,11 based on site-directed mutagenesis exper-
iments and SAR data, were employed to assess the binding modes
of compounds discussed. Conformational analyses of several com-
pounds reported were performed, each of which followed a proce-
dure described recently.14 As an example, the molecular modeling

J. J. Chen et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4477–4481
4479
H
Ar
N
OMe
F
H2N
OMe
SO2
N
b
a
OMe
+
O
O
N
H
O
NO2
MeO
O
O
N
H
O
6
7
8
9
H2N
c
Ar
Ar
X
SO2
N
SO2
N
H
N
OH
11
O
O
X
CH2OH
d
N
H
O
N
H
O
10
12
CH2OH
e
Ar
SO2
N
Ar
H
SO2
N
N
H
N
f
O
O
X
N
H
O
X
N
H
O
5
13
NR1R2
CHO
H2N
H2N
BocHN
g
h
X
X
X
CH2OH
N
N
11
15
14
Ar
SO2
Ar
H
SO2
N
N
N
OH
i
O
X
N
H
O
O
N
H
O
5
10
N
Scheme 1. Reagents and conditions: (a) 1—iPr2NEt, DMSO, 60–70 °C, 60–70%; 2—Pd/C, HCO2NH4, EtOAc/EtOH, 88%; (b) ArSO2Cl, pyridine, 32–76%; (c) 10% HCl, MeOH,
dioxane (1/1/1), reflux; (d) EDCI, HOBt, 11, DMF, 52%; (e) MnO2, CH2Cl2, 100%; (f) 1—R1R2NH, HOAc, CHCl3, then NaBH4, MeOH, 40–70% (for 5b, 5d–f, 5i); or 2—Me2NH,
NaBH(OAc)3, ClCH2CH2Cl, 64% (for 5c); or 3—NH4OAc, Ti(OiPr)4, ClCH2CH2Cl/DMF, then NaBH4, MeOH, 59% (for 5a); (g) 1—Boc2O, NEt3, THF; 2—MnO2, CH2Cl2; 3—piperidine,
NaBH(OAc)3, ClCH2CH2Cl; (h) 1.4 N HCl in dioxane and then NaOH, 40–81% yield from 11 to 15; (i) EDCI, HOBt, 15, DMF, 23–40%. Note: In this scheme, X = O or CH2.
study of 1 is described here, which was carried out using molecular
dynamics/mechanics. A conformational search for the small mole-
cule was carried out using a genetic algorithm. Simulations were
conducted using the program, FLAME,14 which employs the MMFF
forcefield.15 Only low energy conformations lacking intramolecular
H-bonding with the basic nitrogen (consistent with the earlier pro-
posed binding modes of 3 and its analogs) were considered as
putative binding modes and the 10 lowest energy conformations
were considered for docking, visualization and optimization in
the kinin binding pocket of the B1 homology model. These low en-
ergy conformations are also consistent with a conformational anal-
ysis of representative model structures reported recently.6 These
low energy conformations included the two different puckers of
the chroman or tetralin on the right-hand side and different con-
formations of the aliphatic amine side chain on the right-hand side.
Energy minimization of the B1 homology model with each of
the starting structures was performed using the AMBER force-
field16 for proteins and the GAFF forcefield17 for small molecules,
to optimize the complex and estimate MMGB binding energies.18
A mild positional restraint was placed on the ligand during the
minimization/dynamics run. Protein residues within 2.0 Å of any
atom of the docked ligand (including hydrogens) were treated as
flexible, along with other residues in the close proximity of the li-
gand. These include Ile97, Trp98, Ile113, Asn114, Ile117, Lys118,
Tyr266, His267, Glu273, Gln295, Ile290, Asn298, and Phe302.
Site-directed mutagenesis studies of several kinins and small mol-
ecule B1 antagonists9,10,12 showed that most of these residues are
involved in the interactions with ligands.
Figure 1a and b shows the putative binding modes for 1 and 5i.
These models suggest that the phenyl group of the b-phenylalanine

4480
J. J. Chen et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4477–4481
Table 1
In vitro data for the B1 antagonists
Ar
SO2
H
N
N
O
X
N
H
O
5
R
Compound #
Ar
X
R
Binding Ki SEMa (nM)
Functional IC50 SEMa (nM)
62 15
HLMb
(lL/min/mg)
12
13
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
3,4-Cl2Ph
3,4-Cl2Ph
3,4-Cl2Ph
3,4-Cl2Ph
3,4-Cl2Ph
3,4-Cl2Ph
3,4-Cl2Ph
3,4-Cl2Ph
3,4-Cl2Ph
3,4-Cl2Ph
Ph
O
O
O
O
O
O
O
O
O
CH2
O
O
O
CH2OH
CHO
CH2NH2
CH2NHMe
CH2NMe2
CH2NHiPr
CH2NHiBu
CH2NHtBu
CH2piperidine
CH2piperidine
CH2NHtBu
CH2piperidine
CH2piperidine
86 43
57 18
2.6 0.6
1.2 0.2
264
407
127
107
208
156
264
174
389
417
26
23
4
1.2 0.2
0.8 0.2
1.0 0.3
0.86 0.2
1.4 0.2
1.7 0.3
0.48 0.09
0.55 0.08
0.37 0.12
0.12 0.1
0.14 0.2
1.7
1
0.7 0.3
0.44 0.16
1.1 0.6
0.15 0.02
0.10 0.01
1.8 0.8
Ph
4-MePh
0.19 0.05
0.24 0.01
76
128
a
The average of two or more independent experiments with duplicates at each concentration. For assays see Refs. 6–8.
The experiments are performed with two replicates for each compound. The percent difference between replicates must be less than or equal to 20% in order for the result
b
to be considered valid. The average value of the two replicates is then reported as the percent turnover.
Figure 1. Predicted binding modes of 5i (a) and 1 (b) in the putative binding site of the B1 receptor. For the protein, the carbons are colored gray. For the ligand 5i, the carbons
are colored green and for 1, they are colored orange. The protein ribbon is shown in cyan color. Protein residues interacting with the ligand are identified.
series or the fused aryl ring of dihydroquinoxalinones interact with
Phe302 and Ile117, whereas the aryl group of aryl sulfonamide
interacts with Trp98, Ile97, and Ile113. The –SO2 oxygens are seen
to interact with Asn298 and Gln278 side chain amides. In the case
of 5i, an additional interaction of the endocyclic amide group of
dihydroquinoxalinone ring with Asn114 is also observed in the
model, leading to enhanced in vitro potency for these compounds
and making them less sensitive to the changes in the other portion
of the molecule (e.g., lipophilicity of the aliphatic amino group).
Not surprisingly, these interactions are similar to those proposed
in the models by Ha and co-workers10,11 for 3 and its analogs.
Consistent with the observation that the basic amino group in
the amide portion increases the in vitro potency, modeling shows
that the amino group in 1 and 5i interacts with the Glu273 acid

J. J. Chen et al. / Bioorg. Med. Chem. Lett. 18 (2008) 4477–4481
4481
functionality. Furthermore, the alkyl part of the amino group can
form hydrophobic contacts with Ile290 and the aliphatic part of
Glu273 side chain.
We have identified a number of highly potent and selective B1
antagonists that will serve as the basis for further studies that will
be reported in due course.
Similar to the observation made in the case of 1,6 replacement
of the chroman unit by a tetralin has minimal effect on the in vitro
potency (compare 5g–h, Table 1).
References and notes
1. Calixto, J. B.; Medeiros, R.; Fernandes, E. S.; Ferreira, J.; Cabrini, D. A.; Campos,
M. M. Br. J. Pharmacol. 2004, 143, 803.
In the piperidine acetic acid series such as 2, the lipophilic sul-
fonamides are required for potency. For example, substitution of
3-chloro-4-methyl-phenylsulfonamide with simple unsubtituted
phenyl sulfonamide decreases both binding and functional potency
by 100-fold.8 Initially 3,4-dichlorophenyl sulfonamide was also
used in the current series to maximize potency. Since compounds
in this series are at least 10-fold more potent than the correspond-
ing analogs in the piperidine acetic acid series, it is reasonable to
expect that the in vitro potency in the current series would be less
sensitive to the lipophilicity change of the sulfonyl group. The sim-
ple phenyl sulfonamides 5i–j are indeed highly potent, both in
binding and in functional assays. For example, the tert-butyl amine
5i is basically equally potent with the corresponding 3,4-dichloro-
phenyl sulfonamide analog 5f in the binding assay, but it is 3-fold
more potent than 5f in functional assays. Likewise the piperidine 5j
has binding affinity similar to 5g but its functional potency im-
proves by 2-fold compared to 5g. These results suggested that a
reduction of the lipophilicity of the substituted phenyl sulfon-
amide has beneficial effect on functional potency in the current
series. Furthermore, due to the reduced lipophilicity, the metabolic
stability improved significantly. Compounds 5j represents one of
the most functionally potent and metabolically stable B1 antago-
nists. As expected, switching from phenylsulfonamide to p-tolu-
enesulfonamide has miminal effect on the binding and functional
potency (compare 5j and 5k).
The additional binding interactions provided by the endocyclic
amide in the case of 5i gave us more flexibility in optimizing the
other portion of the molecule for pharmacokinetic properties while
still maintaining good overall potency. For example, we were able
to reduce the lipophilicity of both sulfonamide and terminal amine
portions to improve metabolic stability such as in 5j–k. The pres-
ence of the ring amide bond in the dihydroquinoxalinone moiety
also decreases the protein binding, which could be an important
factor in determining the in vivo efficacy. For example, 5f is
79.8% human serum protein bound versus 94.1% for the
corresponding b-phenylalanine analog.
2. Marceau, F.; Regoli, D. Nat. Rev. 2004, 3, 845.
3. Campos, M. M.; Leal, P. C.; Yunes, R. A.; Calixto, J. B. TrendsPharmacol. Sci. 2006,
27, 646.
4. Chen, J. J.; Johnson, E. J. Expert Opin. Ther. Targets 2007, 11, 21.
5. Kuduk, S. D.; DiMarco, C. N.; Chang, R. K.; Wood, M. R.; Schirripa, K. M.; Kim, J.
J.; Wai, J. M. C.; DiPardo, R. M.; Murphy, K. L.; Ransom, R. W.; Harrell, C. M.;
Reiss, D. R.; Holahan, M. A.; Cook, J.; Hess, J. F.; Sain, N.; Urban, M. O.; Tang, C.;
Prueksaritanont, T.; Pettibone, D. J.; Bock, M. G. J. Med. Chem. 2007, 50, 272.
6. D’Amico, D. C.; Aya, T.; Human, J.; Fotsch, C.; Chen, J. J.; Biswas, K.; Riahi, B.;
Norman, M. H.; Willoughby, C. A.; Hungate, R.; Reider, P. J.; Biddlecome, G.;
Lester-Zeiner, D.; VanStaden, C.; Johnson, E.; Kamassah, A.; Arik, L.; Wang, J.;
Viswanadhan, V. N.; Groneberg, R. D.; Zhan, J.; Suzuki, H.; Toro, A.; Mareska, D.
A.; Clarke, D. E.; Harvey, D. M.; Burgess, L. E.; Laird, E. R.; Askew, B.; Ng, G.
J. Med. Chem. 2007, 50, 607.
7. Biswas, K.; Li, A.; Chen, J. J.; D’Amico, D. C.; Fotsch, C.; Han, N.; Human, J.; Liu,
Q.; Norman, M. H.; Riahi, B.; Yuan, C.; Suzuki, H.; Mareska, D. A.; Zhan, J.;
Clarke, D. E.; Toro, A.; Groneberg, R. D.; Burgess, L. E.; Lester-Zeiner, D.;
Biddlecome, G.; Manning, B. H.; Arik, L.; Dong, H.; Huang, M.; Kamassah, A.;
Loeloff, R.; Sun, H.; Hsieh, F. Y.; Kumar, G.; Ng, G. Y.; Hungate, R. W.; Askew, B.
C.; Johnson, E. J. Med. Chem. 2007, 50, 2200.
8. Fotsch, C.; Biddlecome, G.; Biswas, K.; Chen Jian, J.; D’Amico Derin, C.;
Groneberg Robert, D.; Han Nianhe, B.; Hsieh, F.-Y.; Kamassah, A.; Kumar, G.;
Lester-Zeiner, D.; Liu, Q.; Mareska David, A.; Riahi Babak, B.; Wang Yueh-Ju, J.;
Yang, K.; Zhan, J.; Zhu, J.; Johnson, E.; Ng, G.; Askew Benny, C. Bioorg. Med.
Chem. Lett. 2006, 16, 2071.
9. Su, D.-S.; Markowitz, M. K.; DiPardo, R. M.; Murphy, K. L.; Harrell, C. M.;
O’Malley, S. S.; Ransom, R. W.; Chang, R. S. L.; Ha, S.; Hess, F. J.; Pettibone, D.
J.; Mason, G. S.; Boyce, S.; Freidinger, R. M.; Bock, M. G.J. Am. Chem. Soc. 2003,
125, 7516.
10. Ha, S. N.; Hey, P. J.; Ransom, R. W.; Bock, M. G.; Su, D.-S.; Murphy, K. L.; Chang,
R.; Chen, T.-B.; Pettibone, D.; Hess, J. F. Biochemistry 2006, 45, 14355.
11. Ha, S. N.; Hey, P. J.; Ransom, R. W.; Harrell, C. M.; Murphy, K. L.; Chang, R.;
Chen, T.-B.; Su, D.-S.; Markowitz, M. K.; Bock, M. G.; Freidinger, R. M.; Hess, F. J.
Biochem. Biophys. Res. Commun. 2005, 331, 159.
12. Kuduk, S. D.; Di Marco, C. N.; Chang, R. K.; Wood, M. R.; Kim, J. J.; Schirripa, K.
M.; Murphy, K. L.; Ransom, R. W.; Tang, C.; Torrent, M.; Ha, S.; Prueksaritanont,
T.; Pettibone, D. J.; Bock, M. G. Bioorg. Med. Chem. Lett. 2006, 16, 2791.
13. Su, D.-S.; Markowitz, M. K.; Murphy, K. L.; Wan, B.-L.; Zrada, M. M.; Harrell, C.
M.; O’Malley, S. S.; Hess, J. F.; Ransom, R. W.; Chang, R. S.; Wallace, M. A.; Raab,
C. E.; Dean, D. C.; Pettibone, D. J.; Freidinger, R. M.; Bock, M. G. Bioorg. Med.
Chem. Lett. 2004, 14, 6045.
14. Cho, S.-J.; Sun, Y. J. Chem. Info. Model. 2006, 46, 298.
15. Halgren, T. A. J. Comput. Chem. 1996, 17, 490.
16. Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.; Ferguson, D. M.;
Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.; Kollman, P. A. J. Am. Chem. Soc. 1995,
117, 5179.
17. Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. . J. Comput.
Chem. 2004, 25, 1157.
In summary, replacement of the b-amino acid in 1–2 by dihy-
droquinoxalinone acetic acid afforded compounds with much im-
proved potency, metabolic stability, and reduced protein binding.
18. Lee, M.; Sun, Y. J. Chem. Theor. Comput. 2007, 3, 1106.