Synthesis and AMPA receptor antagonistic activity of a novel class of quinoxalinecarboxylic acid with a substituted phenyl group at the C-7 position

Article abstract of DOI:10.1016/S0960-894X(03)00740-6

The synthesis and biological properties of a novel class of 7-heterocycle-substituted quinoxalinecarboxylic acids, which bear a substituted phenyl group through a urethane linkage at the C-7 position, are described. One of the synthesized compounds, 15l,

Full text of DOI:10.1016/S0960-894X(03)00740-6

Bioorganic & Medicinal Chemistry Letters 13 (2003) 3521–3525
Synthesis and AMPA Receptor Antagonistic Activity of a Novel
Class of Quinoxalinecarboxylic Acid with a Substituted Phenyl
Group at the C-7Position
Yasuo Takano,* Futoshi Shiga, Jun Asano, Naoki Ando, Hideharu Uchiki and
Tsuyosi Anraku
Discovery Research Laboratories, Kyorin Pharmaceutical Co., Ltd., 2399-1, Nogi, Nogi-machi, Simotsuga-gun,
Tochigi 329-0114, Japan
Received 13 March 2003; accepted 10 July 2003
Abstract—The synthesis and biological properties of a novel class of 7-heterocycle-substituted quinoxalinecarboxylic acids, which
bear a substituted phenyl group through a urethane linkage at the C-7 position, are described. One of the synthesized compounds,
15l, which has a 4-carboxyphenyl carbamoyloxymethyl imidazole group at the C-7 position and is water-soluble, was found to
possess high potency in vitro and showed excellent neuroprotective efficacy in vivo.
# 2003 Elsevier Ltd. All rights reserved.
Glutamic acid, an excitatory amino acid (EAA), is a
major excitatory neurotransmitter in the mammalian
central nervous system. However, overstimulation of
the postsynaptic glutamate receptors by release of EAA
at greater-than-normal physiological levels results in
neuronal death. Consequently, it is thought that this
process may induce neurodegenerative disorders such as
stroke, epilepsy,1 Huntington’s disease, and Alzheimer’s
disease.2
YM-8726 (Fig. 1), and have been shown to exhibit good
AMPA-R antagonistic activity. However, among these
compounds the first-generation agents have been shown
to cause kidney toxicity as a result of their physico-
chemical properties (particularly, low water solubility),7
and have been limited to use in clinical trials. On the
other hand, in second-generation agents such as YM-
872these undesirable properties have been ameliorated
by introducing a hydrophilic functional group, for
example acetic acid, into the quinoxalinedione skeleton
by medicinal chemistry, and this compound has been
shown to have neuroprotective effects in animal models
of focal cerebral ischemia.6b Therefore, it seemed to us
that little scope remained for development of new sec-
ond-generation AMPA-R antagonists based on chemical
Recent studies of EAA receptors have shown that a-
amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid
receptor (AMPA-R) antagonists have no side effects
such as schizophrenia3 and protect effectively against
neuronal death even in post-ischemia animal models.4
Additionally, many quinoxalinedione derivatives with
competitive AMPA-R antagonistic activity have been
synthesized and tested against the EAA receptor. How-
ever, there have been no conclusive clinical trials show-
ing that compounds with a quinoxalinedione structure
are therapeutic in settings such as acute cerebral ische-
mia. These quinoxalinedione compounds can be divided
into first-generation compounds, such as NBQX4 and
YM-90K,5 and second-generation compounds, such as
*Corresponding author. Tel.: +81-280-56-2201; fax: +81-280-57-
1293; e-mail: yasuo.takano@mb.kyorin-pharm.co.jp
Figure 1.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00740-6

3522
Y. Takano et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3521–3525
modification of the quinoxalinedione structure. Thus,
our research efforts have focused on designing and syn-
thesizing novel third-generation compounds with no
quinoxalinedione skeleton that have potent AMPA-R
affinities and potent neuroprotective effects in vivo and
that in addition are water-soluble, allowing their use as
injections.
moiety with a 78% yield, followed by sequential
nucleophilic substitution of the fluoro group and con-
version to 12a–j by hydrolysis with a 5–81% yield from
9. Next, 4-urethane substituted imidazole derivatives
12k–l, 15a–l, 15n–q were prepared from key inter-
mediate 11i; 4-hydroxymethyl imidazole 11i was treated
with various isocyanates followed by hydrolyses to pro-
vide corresponding 4-urethane derivatives with a 11–
81% yield from 11i as shown in Scheme 1. Furthermore,
the 7-pyridone compound 12l was prepared from 3-
hydroxymethyl-4-pyridone10 instead of 4-hydroxy-
methylimidazole 11i. Then, 15m was also prepared from
4-hydroxyethylimidazole11 instead of 4-hydroxy-
methylimidazole 11i.
It has been reported that the 2,3-dione moiety on the
quinoxalinedione skeleton is required to form a cou-
lombic interaction with the AMPA-R as an essential
part of the proton acceptor.8a On the other hand, it
seemed to us that introduction of an acetic acid group
onto the quinoxalinedione skeleton, for example YM-
872, results in improved affinity for the AMPA-R as
compared with YM-90K. We considered, therefore, that
the acetic acid group on the quinoxaline skeleton affects
the ability of the compound to interact with the AMPA-
R as well as improving water solubility. Thus, we
designed and synthesized new compounds, 3-oxo-2-quin-
oxalinecarboxylic acids, with the hydrophilic group on
the quinoxaline nucleus instead of a 2,3-dione moiety
with an acetic acid group. As a result, we have dis-
covered a novel compound with a 4-carboxyphenyl car-
bamoyloxymethyl imidazole substitution of the C-7 side
chain of quinoxalinecarboxylic acid (Fig. 1). In this
communication, we wish to describe the synthesis of
novel quinoxalinecarboxylic acid derivatives and their
biological effects.
The affinities of the synthesized compounds for the
AMPA and NMDA receptors were assessed by measur-
3
ing their ability to displace, respectively, H-AMPA (5
nM) and 3H-CGS-19755 (10 nM) bound to crude synap-
tic membranes prepared from rat cerebral cortex.12a,13
We selected a nitro group at the C-6 position because
this structure has been shown to confer high AMPA-R
affinity on quinoxalinedione derivatives,8b and per-
formed chemical modification at the C-7 position of the
6-nitro-3-oxo-2-quinoxalinecarboxylic acid compound
(Table 1). From these results, except for the imidazole
group and 4-pyridone group, none of the amino deri-
vatives was significantly different in its AMPA-R affi-
nity at the C-7 position. The AMPA-R affinity of the
heterocycle derivatives (12e, f) was generally better than
that of the alkyl amine derivatives (12a–d). Next, we
compared AMPA-R affinity between various sub-
stituents on the imidazole moiety. Unfortunately,
AMPA-R affinity was reduced by the introduction of
substituents into the imidazole moiety (12g–i). How-
ever, the carboxymethyl compound (12j) maintained an
AMPA-R affinity equal to that of the non-substituted
imidazole compound (12f). From these results, we con-
sidered that the carbonyl group on the imidazole moiety
may facilitate interaction with AMPA-R, and thus we
synthesized a series of carbonyl compounds based on
the hydroxymethyl group of 12i.
The syntheses of novel quinoxalinecarboxylic acid deri-
vatives are outlined in Scheme 1. The 7-fluoroquinoxa-
line (7) was prepared from 4-fluoroaniline using an
improvement of the standard procedure.9 The nitroani-
line 4 was treated with ethyl malonyl chloride (Et3N,
DMF), followed by intermolecular cyclization (t-BuOK,
EtOH) and deoxygenation (PBr3, DMF) to provide 7-
fluoroquinoxalinecarboxylate 6, and then, converted to
the nitro compound by selective nitrozation (f. HNO3,
AcOH) at the C-6 position and hydrolysis of an ester
group (KOH, EtOH–H2O) with an overall yield of 29%
from 4. The conversion of the 7-fluoro ester 8 to the
imino ether 9 was performed by ethylation of the amide
t
Scheme 1. (a) Ethyl malonyl–Cl, Et3N, DMF; (b) BuO-K, EtOH; (c) PBr3, DMF; (d) f. HNO3, AcOH; (e) KOH, EtOH–H2O; (f) EtBr, Ag2O,
toluene; (g) R02-NH or imidazole 10, Et3N, THF or CH2Cl2 or benzene; (h) KOH, EtOH–H2O followed by aq HCl; (i) 48% HBr or c.HCl, AcOH;
(j) R4-NCO, Et3N, CH2Cl2 or benzene; (k) c.HCl, AcOH followed by 1N-LiOH.

Y. Takano et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3521–3525
Table 1. 7-Halogenated and 7-heterocycle derivatives
3523
benzene moiety joined through a urethane linkage
(Table 2). There was a significant increase in AMPA-R
affinity after introduction of various substituents into
the terminal benzene ring, except for the 2-carboxylic
derivatives (15j), which were inferior to the other com-
pounds. On the other hand, the introduction of various
substituents into the 2- or 3-position of the terminal
benzene ring resulted in significantly better selectivity.
However, compounds substituted at the 4-position
showed low selectivity, except when a carboxylic acid
group was introduced to the benzene ring. Furthermore,
we investigated the influence of the methylene length of
the C-7 side chain with respect to AMPA-R affinity and
selectivity. When the methylene group was inserted
between the imidazole group and the urethane group
there was significantly lower AMPA-R affinity (15l vs
15m). On the other hand, when the methylene group
was extended at the terminal substituted benzene moiety
the AMPA-R affinity or the selectivity were reduced
(15p vs 15e and 15o vs 15d, respectively), although a
non-substituted phenyl compound exhibited the same
AMPA-R affinity and selectivity (12k vs 15n). Conse-
quently, we concluded that methylene length between
the imidazole ring and the urethane linkage, and exten-
sion between the urethane linkage and the terminal
substituted benzene, influence AMPA-R affinity and
selectivity.
Compd
R
AMPA-R affinity
Ki (nM)12a
7
12a
F
(CH3)2N–
2000
910
12b
12c
12d
12e
12f
12g
12h
12i
3800
2000
2500
330
560
1300
>10,000
2600
530
The neuroprotective effects of the selected compounds,
and those of NBQX, YM-90K and YM-872, were
examined using the permanent focal ischemia model in
rats described by Tamura et al.14 The results are shown
in Table 3.15 The compounds showed better neuropro-
tective effects in vivo, as well as AMPA-R antagonistic
activity in vitro, than NBQX, YM-90K and YM-872.
The AMPA-R antagonistic activity in vitro confirmed
the inhibitory effect on AMPA-induced DC potential in
rat cortical slices.12b In a comparison of the biological
activities of the non-substituted imidazole 12f and the
imidazole urethane-linked at the C-7 position 15l, com-
pound 15l had markedly more potent activity in vivo
than compound 12f.
12j
12k
12l
86
170
We found that compound 12k, with the urethane link-
age, exhibited more potent AMPA-R affinity than did
the alcohol compound (12i) and the carboxylic com-
pound (12j) (Table 1). On the other hand, we attempted
the introduction of a similar substituent group, a phenyl
group through a urethane linkage, to a 4-pyridone
compound (12e). Despite having the same N-phenyl
urethane group as a substituent group at the C-7 posi-
tion of the quinoxaline nucleus, the 4-pyridone com-
pound (12l) showed weaker AMPA-R affinity than the
imidazole compound (12k). From these results, we con-
cluded that an imidazole group is optimal at the 7-
position of the quinoxaline ring, and that the imidazole
group is needed to confer functional activity through
the urethane linkage for enhancement of AMPA-R affi-
nity. Moreover, we concluded that the urethane moiety
has the same proton acceptor effect as the acid moiety
of compound 12j.
It was noteworthy that the AMPA-R affinity and neu-
roprotective effects of 12f were inferior to those of 15k
and 15l, despite the presence of a quinoxalinecarboxylic
acid nucleus. Compound 12f lacked the urethane link-
age on the imidazole group. Based on these results, this
novel substituent at the C-7 position, namely a sub-
stituted benzene ring with a urethane linkage to imid-
azole, not only confers potent AMPA-R affinity but
also contributes to therapeutic efficacy in animal mod-
els. In particular, the neuroprotective effect of 15l in the
rat model was superior to that of any previously repor-
ted compounds, with a relatively low iv infusion rate of
2.5 mg/kg/h for 4 h, and its aqueous solubility was
higher than that of the first-generation compounds.16
These characteristics qualify this compound, in an
injectable formulation, for use in the treatment of acute
cerebral ischemia.
In order to find an active compound with high AMPA-
R affinity and better selectivity for the AMPA-R than
the NMDA-R (N-methyl-d-aspartate receptor), we
investigated the effects of substituents in the terminal
In conclusion, the molecular design of novel AMPA-R
antagonists based on the quinoxalinecarboxylic acid

3524
Y. Takano et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3521–3525
Table 2. 7-Substituted imidazole derivatives
Compd
R4
n
AMPA-R affinity12a
(Ki, nM)
Selectivity13
À
Á
NMDA-R
AMPA-R
3
YM-872—
40
62
2
12k
15aa
15ba
15c
15d
15ea
15f
15g
15h
15ia
15j
Ph
2-Cl–Ph
3-Cl–Ph
4-Cl–Ph
2-Br–Ph
3-Br–Ph
4-Br–Ph
2-Me–Ph
3-Me–Ph
4-Me–Ph
2-CO2H–Ph
3-CO2H–Ph
4-CO2H–Ph
1
1
1
1
1
1
1
1
1
1
1
1
1
86
20
18
30
30
16
3238
16
>100
210
370
43
280
440
370
290
70
>9.9
>330
>400
22
30
>900
27
22
15k
15l
15m
4-CO70H–Ph
2
22
—
15nb
15o
15p
15q
PhCH2
1
1
1
1
67
29
100
47
>110
160
69
2-Br–PhCH2
3-Br–PhCH2
4-Br–PhCH2
62
aHCl salt.
bSodium salt.
Table 3. Pharmacological data of 7-heterocycle-substituted quinoxalinecarboxylic acid derivatives
Compd
AMPR-R affinity12a
(Ki, nM)
Selectivity13
AMPA-R
antagonism12b
(DC potential)
Protective effects in
focal ischemia model14,15
(dose; mg/kg/h for 4 h, iv)
À
Á
NMDA-R
AMPA-R
NBQX (1)
YM-90K (2)
YM-872( 34)0
65
100
(6+22)
>1300
430
N.T.
(+)
2.3 (30)
2.8 (15)
0.5 (30)
12f
15k
15l
560
27
22
480
>330
>400
(+)
(+)
(+)
0.2(20)
2.3 (5.0)
3.0 (2.5)
N.T., not tested.
Acknowledgements
We are grateful to Mr. K. Fukuchi for running the
radioligand-binding assay, and we would like to thank
Dr. K. Obi and Dr. T. Ishizaki in Kyorin Pharmaceu-
tical Co. Ltd. for helpful discussion.
Figure 2. Four-point scale.
References and Notes
nucleus has created novel quinoxaline compounds with
good biological activities. By studying structure–activity
relationships and evaluating the properties of various
compounds, we identified compound 15l (GRA-293)17
as a novel AMPA-R antagonist that possesses high
potency and good selectivity in vitro as well as more
potent neuroprotective effects in an animal model in
vivo than known quinoxalinedione compounds.
1. Satkowski, M.; Attwell, D. Trends Neurosci. 1994, 17, 359.
2. Gorter, J. A.; Petrozzino, J. J.; Aronica, E.; Rosenbaum,
D. M.; Opits, T.; Bennet, M. V. L.; Connor, J. A.; Zukin, R. S.
J. Neurosci. 1997, 17, 6179.
3. Advokat, C.; Pellegrin, A. I. Neurosci. Biobehav. Rev. 1992,
16, 13. Wouter, K.; James, H. W.; Gail, D. W. J. Pharmacol.
Exp. Ther. 1958, 245, 969.
4. Sheardown, M. J.; Nielsen, E. O.; Hansen, A. J.; Jacobsen, P.;

Y. Takano et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3521–3525
3525
Honore, T. Science 1990, 247, 571. Judge, M. E.; Sheardown,
M. J.; Jacobsen, P.; Honore, T. Neurosci. Lett. 1991, 133, 291.
5. Ohmori, J.; Sakamoto, S.; Kubota, H.; Shimizu-Sasamata,
M.; Okada, M.; Kawasaki, S.; Hidaka, K.; Togami, J.; Fur-
uya, T.; Murase, K. J. Med. Chem. 1994, 37, 467.
6. (a) Shishikura, J.; Tsukamoto, S.; Inami, H.; Fujii, M.;
Okada, M.; Sasamata, M.; Sakamoto, S. WO 9610023; Chem.
Abstr. 1996, 125, 114689. (b) Kawasaki-Yatsugi, S.; Yatsugi,
S.; Takahashi, M.; Toya, T.; Ichiki, C.; Shimizu-Sasamata,
M.; Yamaguchi, T.; Minematsu, K. Brain. Res. 1998, 793,
39.
7. Xue, D.; Huang, Z.-G.; Barnes, K.; Lesiuk, H. J.; Smith,
K. E.; Buchan, A. M. J. Cereb. Blood Flow Metab. 1994, 14,
251.
8. (a) Bigge, C. F.; Boxer, P. A.; Ortwine, D. F. Curr. Pharm.
Des. 1996, 2, 397. (b) Bigge, C. F.; Nikam, S. S. Exp. Opin.
Ther. Pat. 1997, 7, 1099.
14. Tamura, A.; Graham, D. I.; McCulloch, J.; Teasdale,
G. M. J. Cereb. Blood Flow Metab. 1981, 1, 53.
15. Evaluation of the rat focal ischemia model: After 24 h of
MCA occlusion as described by Tamura et al.,14 brains were
removed and sliced into five coronal (2mm thick) sections
with the use of a rat brain matrix (a manual slicer). Slices were
placed in 2% (w/v) triphenyltetrazolium chloride (TTC) solu-
tion, and then in 10% (v/v) phosphate-buffered formalin. Tis-
sue damage (areas not stained with TTC) was scored on a
four-point scale (see Fig. 2). Each test compound was admi-
nistered by continuous iv infusion for 4 h, starting immedi-
ately after occlusion of the MCA. Control rats received saline
only, and its four-point scale value was less than 1.0.
16. Solubility test: The solubility of test compounds was
determined in 0.1 M phosphate buffer (pH 7.4) at room tem-
perature. After centrifugation, the concentrations in the
supernatant were assayed by HPLC. The results at pH 7.4
were as follows: compounds 1 and 2, <1 mg/mL; compound
3, >5 mg/mL; compound 15l, 4.86 mg/mL: Takeru, H.; Fong-
Mei, L. S.; Toshikiro, K.; Ryttingj, J. H. J. Pharm. Sci. 1979,
68, 1267.
17. The newly-synthesized compound GRA-293 has the fol-
lowing 1H NMR and MS spectral characteristics: 7-[4-[N-[4-
carboxyphenyl]carbamoyloxy]methyl]imidazolyl]-3,4-dihydro-
6-nitro-3-oxo-quinoxaline-2-carboxylic acid (15l): pale-brown
solid; mp 268–270 ꢀC; 1H NMR (DMSO-d6) d 5.10 (2H, s),
7.58 (2H, d, J=8.8 Hz), 7.58 (1H, s), 7.86 (2H, d, J=8.8 Hz),
7.99 (1H, s), 8.03 (1H, s), 8.20 (1H, s), 10.2 (1H, s), 12.6 (1H,
bs); LRMS (FABꢁ) m/z 493 (MHꢁ); HRMS (FABꢁ) m/z
calcd for C21H13O9N6 (MHꢁ) 493.3690, Found 493.0744.
Anal. calcd for C21H13O9N6–2H2O: C, 47.55; H, 3.42; N,
15.65. Found: C, 47.46; H, 3.33; N, 15.65.
9. Jeanne, H. S.; Graham, J.; Joseph, L. L.; Charles, M. T.;
Michael, N. P. WO 9211,425;
101927t.
Chem. Abstr. 1992, 118,
10. This compound, 3-hydroxymethyl-4-pyridone, was syn-
thesized from reduction by sodium borohydride of 3-formyl-4-
hydroxypyridine, which was prepared by lithiation of 4-
chloropyridine followed by treatment of N,N-dimethylform-
amide, and hydroxylation with NaOH solution.
11. Bloemhoff, W.; Kerling, K. E. T. Recl. Trav. Chim. Pays-
Bas 1970, 89, 1181.
12. (a) Johansen, T. H.; Drejer, D.; Watjen, F.; Nielsen, E. O.
Eur. J. Pharmacol. 1993, 246, 195. (b) Neil, L. H.; Michael,
A. S. Br. J. Pharmacol. 1985, 84, 381.
13. Murphy, D. E.; Hutchison, A. J.; Hurt, S. D.; Williams,
M.; Sills, M. A. Br. J. Pharmacol. 1988, 95, 932.