Synthesis and biological evaluation of novel 9-heteroaryl substituted 7-chloro-4,5-dihydro-4-oxo-1,2,4-triazolo[1,5-a]quinoxaline-2-carboxylates (TQX) as (R,S)-2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl)propionic acid (AMPA) receptor antagonists

Article abstract of DOI:10.1248/cpb.56.1085

In this paper, we report a study on some new 4,5-dihydro-4-oxo-1,2,4- triazolo[1,5-a]quinoxaline-2-carboxylate derivatives (TQXs), bearing a nitrogen-containing heterocycle at position-9, and designed as (R,S)-2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl)propionic acid (AMPA) receptor antagonists. These compounds ensue from the structural modification of previously reported 8-heteroaryl-TQXs which were endowed with high affinity and selectivity for the AMPA receptor. All the newly synthesized compounds were biologically evaluated for their binding at the AMPA receptor. Gly/N-methyl-D-aspartic acid (NMDA) and kainic acid (KA) high-affinity binding assays were performed to assess the selectivity of the reported derivatives toward the AMPA receptor. This study produced some new TQXs which are less potent than the reference compounds, and endowed with a mixed AMPA and Gly/NMDA receptor binding affinity. To rationalize the experimental findings, a molecular modeling study was performed by docking some TQX derivatives to the AMPA receptor model.

Full text of DOI:10.1248/cpb.56.1085

August 2008
Chem. Pharm. Bull. 56(8) 1085—1091 (2008)
1085
Synthesis and Biological Evaluation of Novel 9-Heteroaryl Substituted
7-Chloro-4,5-dihydro-4-oxo-1,2,4-triazolo[1,5-a]quinoxaline-
2-carboxylates (TQX) as (R,S)-2-amino-3-(3-hydroxy-5-methylisoxazol-
4-yl)propionic Acid (AMPA) Receptor Antagonists
Daniela CATARZI, ,a Vittoria COLOTTA,a Flavia VARANO,a Guido FILACCHIONI,a Paola GRATTERI,a
*
b
Jacopo SGRIGNANI,a Alessandro GALLI,b and Chiara COSTAGLI
a Dipartimento di Scienze Farmaceutiche, Laboratorio di Progettazione, Sintesi e Studio di Eterocicli Biologicamente
Attivi, Universita’ degli Studi di Firenze; Polo Scientifico, Via U. Schiff, 6, 50019 Sesto Fiorentino (Firenze), Italy: and
b Dipartimento di Farmacologia Preclinica e Clinica, Universita’ degli Studi di Firenze; Viale G. Pieraccini, 6, 50134
Firenze, Italy. Received December 19, 2007; accepted April 30, 2008; published online May 8, 2008
In this paper, we report a study on some new 4,5-dihydro-4-oxo-1,2,4-triazolo[1,5-a]quinoxaline-2-carboxyl-
ate derivatives (TQXs), bearing a nitrogen-containing heterocycle at position-9, and designed as (R,S)-2-amino-
3-(3-hydroxy-5-methylisoxazol-4-yl)propionic acid (AMPA) receptor antagonists. These compounds ensue from
the structural modification of previously reported 8-heteroaryl-TQXs which were endowed with high affinity and
selectivity for the AMPA receptor. All the newly synthesized compounds were biologically evaluated for their
binding at the AMPA receptor. Gly/N-methyl-D-aspartic acid (NMDA) and kainic acid (KA) high-affinity binding
assays were performed to assess the selectivity of the reported derivatives toward the AMPA receptor. This study
produced some new TQXs which are less potent than the reference compounds, and endowed with a mixed
AMPA and Gly/NMDA receptor binding affinity. To rationalize the experimental findings, a molecular modeling
study was performed by docking some TQX derivatives to the AMPA receptor model.
Key words ionotropic glutamate receptor; competitive antagonist; triazoloquinoxaline; (R,S)-2-amino-3-(3-hydroxy-5-
methylisoxazol-4-yl)propionic acid
Glutamate (Glu) is the major excitatory neurotransmitter AMPA receptor antagonists,22—35) we have published works
in the central nervous system (CNS), where it is involved in which report the synthesis and pharmacological studies on
the physiological regulation of processes such as learning, 4,5-dihydro-4-oxo-1,2,4-triazolo[1,5-a]quinoxaline-2-car-
memory and synaptic plasticity.1—3)
boxylates (TQXs) bearing different substituents on the fused
Glu activates specific receptors which belong to the classes benzo moiety (Series A, Fig. 1).26,27,29,31) These studies pro-
of metabotropic (mGluRs, coupled to G-protein) and vide evidence on the structural requirements which are im-
ionotropic receptors (iGluRs, ligand-gated ion channel), the portant for obtaining potent and selective AMPA receptor an-
latter consisting of three major subclasses: N-methyl-D-aspar- tagonists: i) a NH proton donor that binds to a proton accep-
tic acid (NMDA), kainic acid (KA) and (R,S)-2-amino-3-(3- tor of the receptor; ii) the 3-nitrogen atom and the oxygen
hydroxy-5-methylisoxazol-4-yl)propionic acid (AMPA) re- atom of the 4-carbonyl group that are d-negatively charged
ceptors which are classified according to their preferential heteroatoms able to form a coulombic interaction with a pos-
synthetic agonists.2,4)
itive site of the receptor; iii) a carboxylate function at posi-
It is well known that many neurological disorders, such as tion-2 able to engage a strong hydrogen-bond interaction
cerebral ischemia, epilepsy, amiotrophic lateral sclerosis and with a cationic proton donor site of the receptor; iv) an elec-
Parkinson’s diseases,1) are caused by excessive release of Glu tron-withdrawing substituent (EWG) at position-7; and v) a
from presynaptic terminals which overstimulates postsynap- N3-nitrogen-containing heterocycle at position-8 of the TQX
tic GluRs, thus leading to neurotoxicity.5—12)
framework, which is an essential feature for selective AMPA
An approach to antagonize the overstimulation of postsy- receptor antagonists. These structural requirements are in ac-
naptic iGluRs by excessive endogenous Glu is represented by cordance with those emphasized in the pharmacophore mod-
the use of AMPA receptor antagonists that, together with els of the AMPA receptor reported in the literature for the
other Glu receptor antagonists, have been proposed as poten- binding of quinoxalinedione and heterocyclic-fused quinox-
tial useful neuroprotectives for the prevention and treatment alinone antagonists.10,16—18,36)
of the above mentioned neurological disorders.10—20) The
Among the previously reported TQX series, the 7-chloro-
success of AMPA receptor antagonists as potential therapeu- 4,5-dihydro-4-oxo-8-(1,2,4-triazol-4-yl)-1,2,4-triazolo[1,5-
tic agents is in part due to their greater clinical potential with a]quinoxaline-2-carboxylic acid TQX-173 and its corre-
respect to other pharmacologically well-characterized iGluR sponding ethyl ester (TQX-197) (Fig. 1), both bearing a
antagonists: for example, AMPA receptor antagonists do not (1,2,4-triazol-4-yl) moiety at the crucial 8-position, emerged
produce the adverse psychotomimetic and cardiovascular ef- as two of the most active and selective TQX compounds to-
fects observed for competitive NMDA receptor antago- ward the AMPA receptor.27,29) Thus, it can be hypothesized
nists.21) These data have consequently added great impetus that the 8-heteroaryl group on the TQX framework interacts
for the development of AMPA receptor antagonists as re- with a hydrophobic receptor region which also contains a
search tools.
hydrogen bond donor site.10,16—18,36)
In the course of our efforts to find novel competitive
Pursuing our studies on the SAR of the TQX series, the 8-
∗ To whom correspondence should be addressed. e-mail: daniela.catarzi@unifi.it
© 2008 Pharmaceutical Society of Japan

1086
Vol. 56, No. 8
Fig. 1. Previously (Series A) and Currently (Series B) Reported TQX Derivatives
Chart 2
Chart 1
Chart 3
heteroaryl substituent of previously reported TQX deriva-
tives was moved to position-9 of the fused benzo ring yield- intermediate, represented by the 9-amino-substituted deriva-
ing compounds of Series B (Fig. 1). In fact, on the basis of tive 1a, was synthesized as reported in Chart 1.
the above mentioned AMPA receptor pharmacophore models
Diazotization with NaNO2 in conc. H2SO4 of the commer-
reported in the literature,10,17,18) the hydrophobic task which cially available 4-chloro-2,6-dinitroaniline, followed by reac-
binds the R8 substituent should be large enough to well ac- tion with sodium tetrafluoborate, yielded the not isolated di-
commodate also bulky groups at position-9 of the TQX azonium tetrafluoborate. The diazonium salt was directly re-
framework.
Moreover, the effect of the presence of the electron-donat- chloroacetate 6, which was transformed with ammonia into
acted with ethyl 2-chloro-3-oxobutanoate to afford the N2-
ing NH2 group (Series B, Fig. 1) was evaluated.
its corresponding N2-oxamidrazonate 7. By reacting 7 with
All of the newly synthesized compounds were biologically ethyloxalyl chloride, the N3-ethoxalyl derivative 8 was ob-
evaluated for their binding at the AMPA receptor. Gly/ tained, which was transformed into the diethyl 1-(4-chloro-
NMDA and KA high-affinity binding assays were performed 2,6-dinitrophenyl)-1,2,4-triazole-3,5-dicarboxylate 9 by de-
to assess the selectivity of the reported compounds toward hydration with concentrated H2SO4. Reduction of both the
the AMPA receptor. A selected compound was also tested for nitro groups of 9 and contemporary cyclization of the inter-
its functional antagonist activity at the AMPA receptor.
mediate afforded the ethyl 9-amino-4,5-dihydro-4-oxo-1,2,4-
Chemistry The syntheses of the 9-substituted-4,5-dihy- triazolo[1,5-a]quinoxaline-2-carboxylate 1a. By reacting
dro-4-oxo-1,2,4-triazolo[1,5-a]quinoxaline-2-carboxylates 1a with diformylhydrazine, the 9-(1,2,4-triazol-4-yl) ester
1a—4a, 1b—4b; 5a are illustrated in Charts 1—3. The key 2a was prepared. The esters 9-(pyrrol-1-yl)- and 9-(3-

August 2008
1087
formylpyrrol-1-yl)-substituted 3a and 4a were obtained by pound is that bearing, on the fused benzo ring, the 1,2,4-tri-
reacting 1a with 2,5-diethoxytetrahydrofuran and 2,5- azol-4-yl moiety, i.e., compound 2b.29) However, it emerged
dimethoxy-3-tetrahydrofurancarboxaldehyde, respectively, in that moving the 8-heteroaryl substituent to position-9 caused
glacial AcOH.
a modest reduction in potency: in fact, compounds 2a and
Oxidation of the 9-(3-formylpyrrol-1-yl) derivative 4a 2b, bearing the (1,2,4-triazol-4-yl) moiety at position-9, were
with potassium permanganate yielded the 9-(3-carboxy- about 3-fold less active than the corresponding 8-substituted
pyrrol-1-yl) ester 5a in very low yield (Chart 2).
TQX-197 and TQX-173. In general, this can be observed
Finally, the hydrolysis of the esters 1a—4a to afford the also when comparison was performed among the other TQX
corresponding 2-carboxylic acids 1b—4b was performed derivatives bearing different heteroaryl substituents (i.e.
(Chart 3). Due to the small amount of compound 5a, its hy- pyrrol-1-yl, 3-formylpyrrol-1-yl) on the fused benzo moi-
drolysis was not attempted.
ety.29)
Affinities of the 9-heteroaryl-substituted derivatives 2a—
4a, 2b—4b, 5a, for both the Gly/NMDA and KA receptors,
Results and Discussion
The triazoloquinoxalines 1a—4a, 1b—4b, 5a, together are generally lower than those for the AMPA receptor. Only
with NBQX (2,3-dihydroxy-6-nitro-7-sulphamoylbenzo[f ]- compounds 2a, 2b and 3b showed a Gly/NMDA binding
quinoxaline) and DCKA (5,7-dichlorokinurenic acid) as affinity in the micromolar range: the same order of potency
standard compounds, were tested for their ability to displace toward the KA receptor is peculiar for the 9-(pyrrol-1-yl)-
tritiated AMPA and Gly from their specific binding in rat substituted compound 3b showing an IC50 value below
cortical membranes. High-affinity KA binding assays were 10 mM.
also performed. The binding results are shown in Table 1 to-
gether with those of previously reported TQX derivatives in- boxylic group at position-2 does not seem to be an essential
cluded as reference compounds.27,29)
feature for the anchoring at the AMPA receptor, but it posi-
As previously observed,26,27,29,31) the presence of a free car-
The binding results indicated that movement of the het- tively influences the potency of these compounds. In fact, all
eroaryl substituent from position-8 of the TQX framework to the 9-heteroaryl-2-carboxylic acids 2b—4b are four to eight
position-9 yielded some new AMPA receptor antagonists. In fold more active than the corresponding ethyl esters 2a—4a
fact, compounds 2b and 3b are endowed with AMPA recep- at the AMPA receptor. The differences between the binding
tor affinities in the low micromolar range. It has to be noted affinities of the esters with respect to those of the correspond-
that, also in this 9-substituted series, the most active com- ing acids does not seem to be highly dependent on the nature
Table 1. Binding Affinity at AMPA, Gly/NMDA and KA High-Affinity Receptors of 9-Substituted TQX Derivativesa)
Ki (mM)b) or I%c)
[3H]AMPA
IC50 (mM)d) or I%c)
Compd.
R
R9
[3H]Gly
[3H]KA
1a
1b
Et
H
NH2
NH2
27%
3.2ꢀ0.3
0%
0.61ꢀ0.13
3%
34.1ꢀ3
2a
2b
3a
3b
4a
Et
2.2ꢀ0.2
0.4ꢀ0.07
6.8ꢀ1.0
5.9ꢀ0.7
2.3ꢀ0.4
40%
30%
37ꢀ2
38%
H
Et
H
0.80ꢀ0.05
3.4ꢀ0.4
1.6ꢀ0.2
91ꢀ10
8.1ꢀ0.8
33%
Et
4b
5a
H
1.2ꢀ0.08
8.9ꢀ0.7
35ꢀ6
21ꢀ3
Et
36ꢀ4.6
35%
TQX-197e)
TQX-173e)
NBQX
—
—
—
—
—
—
—
—
0.70ꢀ0.13
0.14ꢀ0.02
0.07ꢀ0.06
5%
15%
33.5ꢀ5.3
3%
42%
11.3ꢀ1.3
7.0ꢀ1.1
8%
DCKA
0.09ꢀ0.02
a) The tested compounds were dissolved in DMSO and then diluted with the appropriate buffer. b) Inhibition constant (Ki) values were meansꢀS.E.M. of three or four sep-
arate determinations in triplicate. c) Percentage of inhibition (I%) of specific binding at 100 mM concentration. d) Concentration necessary for 50% inhibition (IC50). The IC50
values were meansꢀS.E.M. of three or four separate determinations in triplicate. e) Refs. 27, 29.

1088
Vol. 56, No. 8
of the heteroaryl substituent as it was in the previously re- and TQX-173, and the newly synthesized compounds 2a and
ported 8-substituted TQX series.29,31)
2b (Fig. 2). The interactions between the ligands and the
Replacement of the heteroaryl substituent with an amino residues in D1 and D2 involve the highly conserved residue
group caused a decrease of the AMPA binding affinity. In triad among iGluRs: R485, T480 and P478. In particular, the
fact, compound 1b was eight fold less potent than the corre- 4-carbonyl group interacts with R485 and T480, and the NH
sponding 9-(1,2,4-triazol-4-yl) substituted derivative 2b. In at position-5 engages a hydrogen bond interaction with the
contrast, 1b was the most active compound at the Gly/ D1 residue P478. Moreover, the triazoloquinoxaline scaffold
NMDA receptor among the reported series.
is stabilized by a p–p stacking interaction with the aromatic
Compound 2b, which was the most potent compound of Y450 residue (distanceꢁabout 4.2 Å), located at the top of
this series at the AMPA receptor, together with the well the binding pocket. The OH group of the S654 in D2 is hy-
known NBQX and DCKA, was evaluated for functional an- drogen bonded with the oxygen of the 2-carboxylate group.
tagonist activity by assessing its ability to inhibit depolariza- The distances of all potential hydrogen bonds (ꢂ3.3 Å) be-
tion induced by 5 mM AMPA in mouse cortical wedge prepa- tween GluR2-S1S2J residues and the ligands are reported in
rations.29) Compound 2b inhibited AMPA responses in a re- Table 3. The (1,2,4-triazol-4-yl) moiety of the lead com-
versible manner with an IC50ꢁ8.0ꢀ1 mM. The electrophysio- pounds and derivatives 2a, 2b occupies, in the receptor, the
logical potency of compound 2b is in accordance with the binding cleft formed by E705, L704, T655, T684 and L650;
binding data at the AMPA receptor, thus confirming that this in this way, the (1,2,4-triazol-4-yl) group, either at position-8
compound is an AMPA receptor antagonist.
or -9 of the TQX framework, blocks the interaction of T686
To rationalize the trend of AMPA binding affinities of the on D2 with E402 on D1. This interaction is known to form
most active derivatives 2a and 2b, a molecular modeling the interdomain lock that stabilizes the closed agonist in-
study was performed by docking these new compounds and
Table 2. Physical Data of the Newly Synthesized Compounds
the leads TQX-197 and TQX-173 in the domain-open state
of the GluR2 ligand-binding core (GluR2-S1S2J/(S)-ATPO
crystal structure, pdb entry 1n0t),37) ((S)-ATPO, namely (S)-
2-amino-3-[5-tert-butyl-3-(phosphonomethoxy)-4-isoxa-
zolyl]propionic acid).
Compd.
mp [°C]
Solventa)
Yield [%]
1a
1b
2a
2b
3a
3b
4a
4b
5a
6
ꢃ300
ꢃ300
ꢃ300
D
E
A
B
A
A
A
B
C
A
F
53
93
20
97
58
55
17
43
6
74
85
80
74
The antagonist binding core of ionotropic glutamate recep-
tor, iGluRs, is formed by two domains, D1 and D2, com-
posed of residues from two discontinuous polypetide seg-
ments S1 and S2. They fold into a clamshell-like structure
that undergoes conformational change upon ligand binding,
switching from an open cleft apo conformation to a domain
closed state when glutamate or agonists bind. Antagonists
prevent the domain closure via a foot-in-the-door mechanism
that involves residues from both domains, namely E402,
T686 and residues from the base of the helix F (S654, T655),
depending on the different volume of the binding cavity oc-
cupied by the ligands.
234—236
260—262
221—223
279—281
189—191
289—291
124—126
172—174
170—172
176—177
7
8
9
F
F
a) Recrystallization solvents: Aꢁethanol; Bꢁthe title compound was dissolved in
the minimal amount of NaOH 1 N, the insoluble material was filtered off and the result-
ing clear solution acidified with HCl 1 N. The resulting solid was collected, treated with
boiling ethanol, collected again and washed with fresh ethanol; Cꢁpurification of the
title compound was performed by silica gel column chromatography as reported in the
The results from QM-polarized docking simulations38—40)
pointed out the similar binding mode of the leads TQX-197 Experimental; Dꢁglacial acetic acid; Eꢁethanol/water; Fꢁcyclohexane/ethyl acetate.
Fig. 2. The Leads TQX-197 and TQX-173 , and Compounds 2a and 2b Docked into the X-Ray GluR2-S1S2J:(S)-ATPO Crystal Structure (1n0t)
Side chains of the domain residues directed towards the binding cavity are shown.

August 2008
1089
Table 3. Distances of Potential Hydrogen Bonds (ꢂ3.3 Å) between tal analyzer for C, H, N, and the results were within ꢀ0.4% of the theoreti-
GluR2-S1S2J Residues and the Ligands
Compd. H-bond
2a
cal values except where otherwise stated (Table 2). The IR spectra were
recorded with a Perkin-Elmer 1420 spectrometer in Nujol mulls and are ex-
1
Distances (Å)
pressed in cmꢄ1. The H-NMR spectra were obtained with a Varian Gemini
200 instrument at 200 MHz. The chemical shifts are reported in d (ppm) and
are relative to the central peak of the solvent. The following abbreviations
are used: sꢁsinglet, dꢁdoublet, ddꢁdouble doublet, tꢁtriplet, qꢁquartet,
mꢁmultiplet, brꢁbroad, and arꢁaromatic protons. The physical data of the
newly synthesized compounds are shown in Table 2.
4CꢁO···HN(R485)
4CꢁO···HN(T480)
5NH···CO(P478)
2COO···HN(S654)
2COO···HO(S654)
2.0
2.0
1.83
1.73
1.73
Ethyl N1-(4-Chloro-2,6-dinitrophenyl)hydrazono-N2-chloroacetate 6
Commercially available 2,6-dinitrophenylhydrazine (5.06 mmol) was por-
tionwise added, in an overall time of 40 min, to a cooled (ꢄ10 °C) solution
of NaNO2 (5.06 mmol) in concentrated H2SO4 (15 ml). Ice (40 g), and then a
cold solution of NaBF4 (3.0 g) in water (7.5 ml), were carefully added to the
reaction mixture rigorously kept below 5 °C. After elimination of the solid
by filtration under reduced pressure, cold (0—5 °C) methanol (20 ml) and
ethyl 2-chloro-3-oxobutanoate (1.6 ml) were added to the clear solution. The
reaction mixture was kept at room temperature for 20 h. The yellow solid
2b
4CꢁO···HN(R485)
5NH···CO(P478)
2COO···HN(S654)
2.4
2.0
1.9
TQX-197
4CꢁO···HN(R485)
4CꢁO···HN(T480)
5NH···CO(P478)
2COO···HN(S654)
2COO···HO(S654)
1.9
2.0
1.77
1.7
2.3
1
which precipitated was collected by filtration and washed with water. H-
TQX-173
4CꢁO···HN(R485)
4CꢁO···HN(T480)
5NH···CO(P478)
2COO···HN(S654)
2COO···HO(S654)
1.7
2.0
1.78
1.58
2.2
NMR (DMSO-d6) d: 1.27 (t, 3H, CH3, Jꢁ6.96 Hz), 4.26 (q, 2H, CH2,
Jꢁ6.96 Hz), 8.52 (s, 2H, ar), 10.92 (s, 1H, NH); IR. 3290, 3120, 3100, 1750.
Anal. Calcd for C10H8Cl2N4O6: C, 34.21; H, 2.30; N, 15.96. Found: C,
34.00; H, 2.07; N, 16.21.
Ethyl N1-(4-Chloro-2,6-dinitrophenyl)-N2-oxamidrazonate 7 Ammo-
nia was bubbled until saturation into a stirred solution of 6 (3.42 mmol) in
anhydrous dioxane (13 ml). The mixture was stirred at room temperature for
30 min and then was diluted with water (60 ml) to yield a red solid which
1
duced conformation of the iGluR2. In addition, the open
state of the cleft induced by the herein docked compounds
ensued by the direct interactions of the 2-carboxylate group
with the hydroxylic oxygen of S654.
Slight differences in compound location interest the 7-Cl
substituent. In fact, while the position and distance of the 7-
was collected and washed with water. H-NMR (DMSO-d6) d: 1.23 (t, 3H,
CH3, Jꢁ6.95 Hz), 4.17 (q, 2H, CH2, Jꢁ6.95 Hz), 7.00 (s, 2H, NH2), 8.32 (s,
2H, ar), 9.84 (s, 1H, NH); IR 3490, 3380, 3300, 1730, 1670. Anal. Calcd for
C10H10ClN5O6: C, 36.21; H, 3.04; N, 21.12. Found: C, 35.99; H, 3.29; N,
21.35.
Ethyl N1-(4-Chloro-2,6-dinitrophenyl)-N3-ethoxalyl-N2-oxamidra-
zonate 8 A solution of ethyloxalyl chloride (5.72 mmol) in anhydrous di-
chloro atom of the lead compounds (TQX-173 and TQX- ethyl ether (14 ml) was dropwise added to a suspension of compound 7
(3.02 mmol) in anhydrous toluene (11 ml). Then, the reaction mixture was
heated at reflux for 1.5 h. After evaporation of the solvent under reduced
pressure, the resulting solid was worked up with diethyl ether (7 ml) and col-
197) are optimal for a Van der Waals contact with M708,
compounds 2a and 2b are not stabilized by this interaction
with the same residue. This, at least partly, might account for
the slightly lower AMPA binding affinities of compounds 2a
and 2b with respect to the leads TQX-197 and TQX-173.
In conclusion, the study on the herein reported compounds
has allowed us to further explore the SAR of the TQX de-
rivatives, affording some new AMPA receptor antagonists.
Moving the crucial heteroaryl substituent from position-8 of
TQX-173 and TQX-197 to position-9 gave rise, respectively,
to 2a and 2b which are slightly less potent than the lead com-
pounds. The binding mode of the 9-substituted heterocyclic
compounds, as it results from docking experiments, does not
substantially differ from that of the 8-substituted analogs be-
longing to the TQX series. Thus, the lower AMPA binding
affinity of the 9-heteroaryl-TQX derivatives with respect to
the 8-substituted ones, does not seem to be correlated with
the movement of the substituent to position-9. In fact, the
lected by filtration. 1H-NMR (DMSO-d6) d: 1.20—1.35 (m, 6H, 2CH3), 4.18
(q, 2H, CH2), 4.33 (q, 2H, CH2), 8.44 (s, 2H, ar), 10.75 (s, 1H, N1H), 11.25
(s, 1H, N3H); IR 3340, 3200, 3110, 1745, 1730, 1720. Anal. Calcd for
C14H14ClN5O9: C, 38.95; H, 3.27; N, 16.22. Found: C, 39.11; H, 3.47; N,
16.47.
Diethyl 1-(4-Chloro-2,6-dinitrophenyl)-1,2,4-triazolo-3,5-dicarboxyl-
ate 9 A suspension of compound 8 (2.1 mmol) in concentrated H2SO4
(8 ml) was kept at room temperature for 4.5 h. Then, the resulting solution
was diluted with ice/water (180 ml) and extracted with ethyl acetate
(120 mlꢅ2). The two portions of ethyl acetate, put together, were washed
with an aqueous solution of NaOH (0.5%, 120 mlꢅ2) and then with water
(100 mlꢅ3). Evaporation of the dried (Na2SO4) organic layers to small vol-
ume (2.5 ml) gave a solid which was collected by filtration and washed
1
with a little petroleum ether. H-NMR (DMSO-d6) d: 1.19 (t, 3H, CH3,
Jꢁ7.0 Hz), 1.33 (t, 3H, CH3, Jꢁ7.0 Hz), 4.26—4.45 (m, 4H, 2CH2), 8.96 (s,
2H, ar). IR 3100, 1735. Anal. Calcd for C14H12ClN5O8: C, 40.64; H, 2.92; N,
16.93. Found: C, 40.40; H, 3.09; N, 17.11.
Ethyl 9-Amino-7-chloro-4,5-dihydro-4-oxo-1,2,4-triazolo[1,5-a]-
quinoxaline-2-carboxylate 1a Iron powder (1.52 g) was added to a solu-
task which accommodates the 9-substituent seems to be large tion of 9 (1.52 mmol) in glacial acetic acid (1 ml). The mixture was heated at
reflux for 5 h. Evaporation of the solvent at reduced pressure yielded a
residue which was dried under vacuum and extracted in soxhlet with acetone
enough to tolerate the herein studied 9-heteroaryl moieties.
In contrast, the weak interaction between the 7-chloro group
and the receptor protein might explain the small difference in
(250 ml). Most of the solvent was evaporated to yield a solid which was col-
1
lected by filtration. H-NMR (DMSO-d6) d: 1.36 (t, 3H, CH3, Jꢁ6.96 Hz),
binding affinities of the herein reported compounds and
leads. Nevertheless, the modification performed on TQX de-
rivatives was not so lucky and does not warrant the synthesis
of any other compound belonging to this series with a flat
heteroaryl substituent at position-9.
4.43 (q, 2H, CH2, Jꢁ6.96 Hz), 6.61 (s, 1H, ar), 6.71 (s, 1H, ar), 7.00 (s, 2H,
NH2), 12.34 (s 1H, NH); IR 3460, 3340, 1750, 1740, 1710, 1650, 1625.
Anal. Calcd for C12H10ClN5O3: C, 46.84; H, 3.28; N, 22.76. Found: C,
46.99; H, 3.45; N, 22.61.
Ethyl 7-Chloro-4,5-dihydro-4-oxo-9-(1,2,4-triazol-4-yl)-1,2,4-triazolo-
[1,5-a]quinoxaline-2-carboxylate 2a Diformylhydrazine (4.86 mmol) and
then, drop by drop, trimethylsilyl chloride (24.3 mmol) and triethylamine
(11.3 mmol), were added to a suspension of 1a (1.62 mmol) in anhydrous
pyridine (8.5 ml). The mixture was heated at 100 °C for 11 h. Evaporation of
the solvent at reduced pressure yielded a solid which was treated with water
(10 ml), collected by filtration and purified by silica gel column chromatog-
raphy, eluting system CHCl3/MeOH (9 : 1). Further purification was per-
Experimental
Chemistry Silica gel plates (Merck F254) and silica gel 60 (Merck; 70—
230 mesh) were used for analytical and column chromatography, respec-
tively. All melting points were determined on a Gallenkamp melting point
apparatus. Microanalyses were performed with a Perkin-Elmer 260 elemen-

1090
Vol. 56, No. 8
formed by crystallization. 1H-NMR (DMSO-d6) d: 1.30 (t, 3H, CH3,
Jꢁ6.96 Hz), 4.32 (q, 2H, CH2, Jꢁ6.96 Hz), 7,65 (d, 1H, ar, Jꢁ2.20 Hz),
7.71 (d, 1H, ar, Jꢁ2.20 Hz), 8.70 (s, 2H, triazole H-2, H-5), 12.9 (br s, 1H,
NH); IR 1750, 1710, 1610. Anal. Calcd for C14H10ClN7O3: C, 46.74; H,
2.80; N, 27.26. Found: C, 46.58; H, 2.56; N, 27.41.
Ethyl 7-Chloro-4,5-dihydro-4-oxo-9-(pyrrol-1-yl)-1,2,4-triazolo[1,5-a]-
quinoxaline-2-carboxylate 3a A solution of 2,5-diethoxytetrahydrofuran
(2.4 mmol) in glacial acetic acid (4.5 ml) was dropwise added to a suspen-
sion of 1a (8.13 mmol) in glacial acetic acid (9 ml). The reaction mixture
was heated at 90 °C for 20 min. Dilution with water (50 ml) afforded a solid
which was collected by filtration and washed with water. 1H-NMR (DMSO-
d6) d: 1.27 (t, 3H, CH3, Jꢁ6.96 Hz), 4.28 (q, 2H, CH2, Jꢁ6.96 Hz), 6.23 (t,
2H, pyrrole H-3, H-4, Jꢁ2.2 Hz), 6.89 (t, 2H, pyrrole H-2, H-5, Jꢁ2.2 Hz),
7.39 (d, 1H, ar, Jꢁ2.2 Hz), 7.51 (d, 1H, ar, Jꢁ2.2 Hz), 12.7 (br s, 1H, NH);
IR 1720, 1700. Anal. Calcd for C16H12ClN5O3: C, 53.72; H, 3.38; N, 19.58.
Found: C, 53.91; H, 3.10; N, 19.71.
Ethyl 7-Chloro-4,5-dihydro-9-(3-formylpyrrol-1-yl)-4-oxo-1,2,4-tria-
zolo[1,5-a]quinoxaline-2-carboxylate 4a A solution of 2,5-dimetoxytet-
rahydrofuran-3-carboxaldehyde (2.4 mmol) in glacial acetic acid (10 ml) was
dropwise added to a suspension of 1a (1.63 mmol) in glacial acetic acid
(10 ml). The reaction mixture was heated at 90 °C for 15 min. After distilla-
tion of the solvent under reduced pressure, the resulting solid was worked up
with a little ethanol, collected by filtration and purified by silica gel column
chromatography, eluting system acetone. Evaporation at small volume of the
first eluates gave a white solid which was collected by filtration, washed with
petroleum ether and recrystallized. 1H-NMR (DMSO-d6) d: 1.21 (t, 3H,
CH3, Jꢁ6.95 Hz), 4.24 (q, 2H, CH2, Jꢁ6.95 Hz), 6.64 (s, 1H, pyrrole pro-
ton), 7.00 (s, 1H, pyrrole proton), 7.61 (s, 2H, ar), 7.78 (s, 1H, pyrrole pro-
ton), 9.77 (s, 1H, CHO), 11.9 (br s, 1H, NH); IR 1740, 1715, 1700, 1675.
Anal. Calcd for C17H12ClN5O4: C, 52.93; H, 3.14; N, 18.15. Found: C,
53.09; H, 3.38; N, 17.95.
Ethyl 7-Chloro-4,5-dihydro-8-(3-carboxypyrrol-1-yl)-4-oxo-1,2,4-tria-
zolo[1,5-a]quinoxaline-2-carboxylate 5a Solid potassium permanganate
(0.85 mmol) was portionwise added to a cooled (0 °C) suspension of 4a
(0.83 mmol) in a 1 : 1 acetone/water mixture (10 ml). The reaction mixture
was stirred at 0—3 °C for 48 h and then diluted with ice/water (30 g); the
small excess of potassium permanganate was quenched with a 40% solution
of sodium bisulfite and the resulting suspension was extracted with ethyl ac-
etate (15 mlꢅ4). Evaporation of the dried (Na2SO4) organic layers at small
volume afforded a solid which was collected by filtration and purified by sil-
ica gel column chromatography (eluting system CHCl3/MeOH 9 : 1). 1H-
NMR (DMSO-d6) d: 1.23 (t, 3H, CH3, Jꢁ6.96 Hz), 4.24 (q, 2H, CH2,
Jꢁ6.96 Hz), 6.53 (s, 1H, pyrrole proton), 6.90 (s, 1H, pyrrole proton), 7.45
(s, 1H, pyrrole proton), 7.51 (d, 1H, ar, Jꢁ2.2 Hz), 7.55 (d, 1H, ar,
Jꢁ2.2 Hz), 11.8 (br s, 1H, NH or OH); IR 1735, 1690. Anal. Calcd for
C17H12ClN5O5: C, 50.82; H, 3.01; N, 17.43. Found: C, 51.03; H, 2.89; N,
17.27.
7-Chloro-4,5-dihydro-8-(3-carboxypyrrol-1-yl)-4-oxo-1,2,4-triazolo-
[1,5-a]quinoxaline-2-carboxylic Acid 4b An aqueous solution of NaOH
(3%, 7.0 ml) was added to a suspension of 4a (0.28 mmol) in EtOH (7.0 ml).
The mixture was stirred at room temperature for 1 h, and then diluted with
water (40 ml). The small amount of insoluble solid was filtered off and the
resulting solution was further fined with charcoal and filtered again. The
cold (5 °C) alkaline phase was acidified with HCl 6 N and then extracted with
ethyl acetate (30 mlꢅ4). Evaporation of the dried (Na2SO4) organic layers
yielded a solid which was collected by filtration and washed with water. 1H-
NMR (DMSO-d6) d: 6.68 (s, 1H, pyrrole proton), 7.10 (s, 1H, pyrrole pro-
ton), 7.63 (s, 2H, ar), 7.86 (s, 1H, pyrrole proton), 9.80 (s, 1H, CHO); IR
3700—3300, 1700. Anal. Calcd for C15H8ClN5O4: C, 50.37; H, 2.25; N,
19.58. Found: C, 50.58; H, 2.41; N, 19.35.
Pharmacology. Binding Assay Rat cortical synaptic membrane prepa-
ration, [3H]glycine and [3H]AMPA binding experiments were performed fol-
lowing the procedures described in refs. 23 and 41, respectively. High-affin-
ity [3H]kainate binding assays were performed on rat cortical membranes ac-
cording to previously reported methods.31)
Electrophysiological Assay The mouse cortical wedge preparation de-
scribed by Mannaioni et al.42) was used, while the electrophysiological as-
says were performed following the procedures described in ref. 29.
Sample Preparation and Result Calculation A stock 1 mM solution of
the test compound was prepared in 50% DMSO. Subsequent dilutions were
accomplished in buffer. The IC50 values were calculated from three to four
displacement curves based on four to six scalar concentrations of the test
compound in triplicate using the ALLFIT computer program43) and, in the
case of tritiated glycine and AMPA binding, converted to Ki values by appli-
cation of the Cheng–Prusoff equation.44) In our experimental conditions the
dissociation constants (KD) for [3H]glycine (10 nM) and [3H]-DL-AMPA
(8 nM) were 75ꢀ6 and 28ꢀ3 nM, respectively.
Molecular Modeling Compounds were modeled using the LigPrep
Schrodinger ligands preparation procedure (pH 7.4) and minimized with
Macromodel8.5.45)
The experimental crystal structure of GluR2S1S2J/ATPO complex (PDB
entry: 1n0t) was used in docking computation. Docking calculations were
performed using the Schrodinger QM-Polarized Ligand docking protocol38)
that overcomes the assumption of standard molecular mechanics force field
by assuming the charge distribution to be invariant to electrostatic fields of
the surrounding environment. In the QM-polarized docking protocol,38) the
ligands are docked with Glide v4.0,46) then charges are derived from quan-
tum mechanical calculation on the ligand in the field of the receptor using
QSitev4.0.47) This procedure allows the polarization of the charges on the
ligand by the receptor to be accounted for.39,40) Redocking the ligands with
these new charges can result in improved docking accuracy.
The crystal structure was prepared according to the protein preparation
procedure recommended. Default input parameters were used in all compu-
tations. Upon completion of the docking calculation, three poses per ligand
were saved. The best-docked structure was chosen using a model energy
score (Emodel) derived from a combination of the Glide Score (Gscore, a
modified and extended version of the empirically based ChemScore
function48)), Coulombic and the van der Waals energies and the internal
strain energy of the ligands. All computations were performed on an
Intel®Pentium®4 3 GHz processor running Linux.
General Procedure for the Hydrolysis of the 2-Carboxylate Ethyl Es-
ters 1a—3a An aqueous solution of NaOH (3%, 7.0 ml) was added to a
suspension of the ethyl 2-carboxylate esters 1a—3a (0.63 mmol) in EtOH
(7.0 ml). The mixture was stirred at room temperature until the disappear-
ance of the starting material (TLC monitoring, eluting system CHCl3/MeOH
9 : 1). The solid was collected by filtration and dissolved in the minimal
amount of water; the clear cold (5 °C) solution was acidified with HCl 6 N,
and then kept at room temperature for 30 min. The solid which precipitated
was collected by filtration and washed with water. Compounds 1b—3b dis-
played the following spectral data:
References
1) “Excitatory Amino Acids and Synaptic Transmission,” ed. by Wheal
H. V., Thomson A. M., Academic Press, London, 1995.
2) “The Ionotropic Glutamate Receptors,” ed. by Monaghan D. T., Wen-
thold R. J., Humana Press, Totowa, New Jersey, 1997.
3) Watkins J. C., Biochem. Soc. Trans., 28, 297—310 (2000).
4) Dingledine R., Borges K., Bowie D., Traynelis S. F., Pharmacol. Rev.,
51, 7—61 (1999).
9-Amino-7-chloro-4,5-dihydro-4-oxo-1,2,4-triazolo[1,5-a]quinoxaline-2-
1
carboxylic Acid 1b: H-NMR (DMSO-d6) d: 6.56 (d, 1H, ar, Jꢁ2.21 Hz),
6.20 (d, 1H, ar, Jꢁ2.21 Hz), 7.17 (s, 2H, NH2). IR 3500, 3365, 3095, 2800—
2100, 200—1900, 1720. Anal. Calcd for C10H6ClN5O3: C, 42.95; H, 2.16; N,
25.04. Found: C, 43.16; H, 2.00; N, 25.16.
5) Danysz W., Parsons C. G., Pharmacol. Rev., 50, 597—664 (1998).
6) Chase T. N., Oh J. O., Konitsiotis S., J. Neurol., 247 (Suppl. 2), II/36—
II/42 (2000).
7) Kew J. N. C., Kemp J. A.. Psycopharmacology, 179, 4—29 (2005).
8) Millan M. J., Psycopharmacology, 179, 30—53 (2005).
9) Bleakman D., Lodge D., Neuropharmacology, 37, 1187—1204 (1998).
10) Bigge C. F., Nikam S. S., Expert Opin. Ther. Patents, 7, 1099—1114
(1997).
7-Chloro-4,5-dihydro-4-oxo-9-(1,2,4-triazol-4-yl)-1,2,4-triazolo[1,5-
a]quinoxaline-2-carboxylic Acid 2b: H-NMR (DMSO-d6) d: 7.66 (d, 1H,
ar, Jꢁ2.2 Hz), 7.72 (d, 1H, ar, Jꢁ2.2 Hz), 8.71 (s, 2H, triazole H-2, H-5),
12.86 (s, 1H, NH), 12.8 (br s, 1H, COOH); IR 3100—2000, 2000—1800,
1710. Anal. Calcd for C12H6ClN7O3: C, 43.45; H, 1.82; N, 29.56. Found: C,
43.68; H, 2.09; N, 29.39.
7-Chloro-4,5-dihydro-4-oxo-9-(pyrrol-1-yl)-1,2,4-triazolo[1,5-a]quinoxa-
line-2-carboxylic Acid 3b: 1H-NMR (DMSO-d6) d: 6.23 (t, 2H, pyrrole H-3,
H-4, Jꢁ2.2 Hz), 6.92 (t, 2H, pyrrole H-2, H-5, Jꢁ2.2 Hz), 7.37 (s, 1H, ar),
7.50 (s, 1H, ar), 12.70 (s, 1H, NH); IR 3500—2600, 1720, 1465, 1375. Anal.
Calcd for C14H8ClN5O3: C, 51.00; H, 2.45; N, 21.24. Found: C, 51.31; H,
2.29; N, 21.00.
1
11) Bräuner-Osborne H., Egebjerg J., Nielsen E. Ø., Madsen U., Krogs-
gaard-Larsen P., J. Med. Chem., 43, 2609—2645 (2000).
12) Lees G. J., Drugs, 59, 33—78 (2000).
13) Auberson Y. P., Drugs Future, 26, 463—471 (2001).

August 2008
1091
14) De Sarro G., Gitto R., Russo E., Ferreri Ibbadu G., Barreca M. L., De
Luca L., Chimirri A., Curr. Top. Med. Chem., 5, 31—32 (2005).
15) Madsen U., Stensbøl T. B., Krogsgaard-Larsen P., Curr. Med. Chem.,
8, 1291—1301 (2001).
16) Bigge C. F., Boxer P. A., Ortwine D. F., Curr. Pharm. Des., 2, 397—
412 (1996).
33) Varano F., Catarzi D., Colotta V., Calabri F. R., Lenzi O., Filacchioni
G., Galli A., Costagli C., Deflorian F., Moro S., Bioorg. Med. Chem.,
13, 5536—5549 (2005).
34) Calabri F. R., Colotta V., Catarzi D., Varano F., Lenzi O., Filacchioni
G., Costagli C., Galli A., Eur. J. Med. Chem., 40, 897—907 (2005).
35) Colotta V., Catarzi D., Varano F., Lenzi O., Filacchioni G., Costagli C.,
Galli A., Ghelardini C., Galeotti N., Gratteri P., Sgrignani J., Deflorian
F., Moro S., J. Med. Chem., 49, 6015—6026 (2006).
17) Chimirri A., Gitto R., Zappalà M., Expert Opin. Ther. Patents, 9,
557—570 (1999).
18) Nikam S. S., Kornberg B. E., Curr. Med. Chem., 8, 155—170 (2001).
19) Stensbøl T. B., Madsen U., Krogsgaard-Larsen P., Curr. Pharm. Des.,
8, 857—872 (2002).
20) Catarzi D., Colotta V., Varano F., Med. Res. Rev., 27, 239—278 (2007).
21) Loscher W., Wlaz P., Szabo L., Neurosci. Lett., 240, 33—36 (1998).
36) Homori J., Shimizu-Sasamata M., Okada M., Sakamoto S., J. Med.
Chem., 40, 2053—2063 (1997).
37) Hogner A., Greenwood J. R., Liljefors T., Lunn M. L., Egebjerg J.,
Larsen I. K., Gouaux E., Kastrup J. S., J. Med. Chem., 46, 214—221
(2003).
22) Varano F., Catarzi D., Colotta V., Cecchi L., Filacchioni G., Galli A., 38) Schrodinger Suite 2006 QM-Polarized Ligand Docking protocol.
Costagli C., Il Farmaco, 52, 179—182 (1997). L.L.C., New York (http://www.schrodinger.com).
23) Colotta V., Catarzi D., Varano F., Cecchi L., Filacchioni G., Galli A., 39) Cho A. E., Guallar V., Berne B. J., Friesner R., J Comput. Chem., 26,
Costagli C., Arch. Pharm. Pharm. Med. Chem., 330, 129—134 (1997). 915—931 (2005).
24) Varano F., Catarzi D., Colotta V., Filacchioni G., Cecchi L., Galli A., 40) Friesner R., Adv. Protein Chem., 72, 79—104 (2005).
Costagli C., Il Farmaco, 53, 752—757 (1998).
41) Nielsen E. O., Madsen U., Schaumburg K., Brehm L., Krogsgaard-
Larsen P., Eur. J. Chem.-Chim. Ther., 21, 433—437 (1986).
42) Mannaioni G., Carlà V., Moroni F., Br. J. Pharmacol., 118, 1530—
1536 (1996).
43) De Lean A., Munson P. J., Rodbard D., Am. J. Physiol., 235, E97—
E102 (1978).
44) Cheng Y.-C., Prusoff W. H., Biochem. Pharmacol., 22, 3099—3108
(1973).
45) Macromodel v8.5, Schrödinger. L.L.C., New York (http://www.
schrodinger.com).
46) Glide v.4.0, Schrödinger. L.L.C., New York, NY, 2005 (http://www.
schrodinger.com).
47) Qsite v4.0, Schrödinger. L.L.C., New York, NY, 2005. Jagurar v6.5,
Schrödinger. L.L.C., New York, NY, 2005 (http://www.schrodinger.
com).
25) Varano F., Catarzi D., Colotta V., Cecchi L., Filacchioni G., Galli A.,
Costagli C., Arch. Pharm. Pharm. Med. Chem., 332, 201—207 (1999).
26) Catarzi D., Colotta V., Varano F., Cecchi L., Filacchioni G., Galli A.,
Costagli C., J. Med. Chem., 42, 2478—2484 (1999).
27) Catarzi D., Colotta V., Varano F., Cecchi L., Filacchioni G., Galli A.,
Costagli C., Carlà V., J. Med. Chem., 43, 3824—3826 (2000).
28) Varano F., Catarzi D., Colotta V., Cecchi L., Filacchioni G., Galli A.,
Costagli C., Eur. J. Med. Chem., 36, 203—209 (2001).
29) Catarzi D., Calotta V., Varano F., Filacchioni G., Galli A., Costagli C.,
Carlà V., J. Med. Chem., 44, 3157—3165 (2001).
30) Varano F., Catarzi D., Colotta V., Filacchioni G., Galli A., Costagli C.,
Carlà V., J. Med. Chem., 45, 1035—1044 (2002).
31) Catarzi D., Colotta V., Varano F., Calabri F. R., Filacchioni G., Galli
A., Costagli C., Carlà V., J. Med. Chem., 47, 262—272 (2004).
32) Colotta V., Catarzi D., Varano F., Calabri F. R., Filacchioni G., Costagli 48) Jorgensen W. L., Maxwell D. S., Tirado-Rives J., J. Am. Chem. Soc.,
C., Galli A., Bioorg. Med. Chem. Lett., 14, 2345—2349 (2004). 118, 11225—11236 (1996).