Ethyl 8-fluoro-6-(3-nitrophenyl)-4H-imidazo[1,5-a][1,4]benzodiazepine-3- carboxylate as novel, highly potent, and safe antianxiety agent

Article abstract of DOI:10.1021/jm8002944

Ethyl 8-fluoro-6-(4-nitrophenyl)- and ethyl 8-fluoro-6-(3-nitrophenyl)-4H- imidazo[1,5-a][1,4]benzodiazepine 3-carboxylate 6 and 7 were synthesized as central benzodiazepine receptor (CBR) ligands and tested for their ability to displace [³H]flumazenil from bovine and human cortical brain membranes. Both compounds showed high affinity for bovine and human CBR. In particular, compound 7 emerged as the most interesting compound, having a partial agonist profile in vitro while possessing useful activity in various animal models of anxiety. In accordance with its partial agonist profile, compound 7 was devoid of typical benzodiazepine side effects. The homology model of the GABA_A receptor developed by Cromer et al. was used to assess the binding modes of ligands 6 and 7. From our docking results, the partial agonist activity elicited by compound 7 is likely to be due to the 3′-nitro substituent, which is in the appropriate position to interact with Thr193 of the γ₂-subunit by means of a hydrogen bond.

Full text of DOI:10.1021/jm8002944

4730
J. Med. Chem. 2008, 51, 4730–4743
Ethyl 8-Fluoro-6-(3-nitrophenyl)-4H-imidazo[1,5-a][1,4]benzodiazepine-3-carboxylate as Novel,
Highly Potent, and Safe Antianxiety Agent
Maurizio Anzini,*,† Carlo Braile,† Salvatore Valenti,† Andrea Cappelli,† Salvatore Vomero,† Luciana Marinelli,‡
Vittorio Limongelli,‡ Ettore Novellino,‡ Laura Betti,§ Gino Giannaccini,§ Antonio Lucacchini,§ Carla Ghelardini,|
,O
Monica Norcini,| Francesco Makovec, Gianluca Giorgi,# and R. Ian Fryer
Dipartimento Farmaco Chimico Tecnologico and European Research Centre for Drug DiscoVery and DeVelopment, UniVersita` degli Studi di
Siena, Via A. Moro, 53100 Siena, Italy, Dipartimento di Chimica Farmaceutica e Tossicologica, UniVersita` di Napoli “Federico II”,
Via D. Montesano 49, 80131 Napoli, Italy, Dipartimento di Psichiatria, Neurobiologia Farmacologia e Biotecnologie, UniVersita` di Pisa,
Via Bonanno 6, 56126 Pisa, Italy, Dipartimento di Farmacologia Preclinica e Clinica “M. Aiazzi Mancini”, UniVersita` degli Studi di Firenze,
Viale G. Pieraccini 6, 50139 Firenze, Italy, Rottapharm S.p.A., Via Valosa di Sopra 7, 20052 Monza, Italy, Dipartimento di Chimica,
UniVersita` degli Studi di Siena, Via A. Moro, 53100 Siena, Italy, Department of Chemistry, Rutgers The State UniVersity of New Jersey,
73 Warren Street, Newark, New Jersey 07102
ReceiVed March 19, 2008
Ethyl 8-fluoro-6-(4-nitrophenyl)- and ethyl 8-fluoro-6-(3-nitrophenyl)-4H-imidazo[1,5-a][1,4]benzodiazepine
3-carboxylate 6 and 7 were synthesized as central benzodiazepine receptor (CBR) ligands and tested for
their ability to displace [3H]flumazenil from bovine and human cortical brain membranes. Both compounds
showed high affinity for bovine and human CBR. In particular, compound 7 emerged as the most interesting
compound, having a partial agonist profile in vitro while possessing useful activity in various animal models
of anxiety. In accordance with its partial agonist profile, compound 7 was devoid of typical benzodiazepine
side effects. The homology model of the GABAA receptor developed by Cromer et al. was used to assess
the binding modes of ligands 6 and 7. From our docking results, the partial agonist activity elicited by
compound 7 is likely to be due to the 3′-nitro substituent, which is in the appropriate position to interact
with Thr193 of the γ2-subunit by means of a hydrogen bond.
Introduction
GABAA receptors are generally hetero-oligomers, the subunits
being selected from four principal families R, ꢀ, γ, and δ,
although others including ε, π, F, and θ have been identified.
At present, a total of 21 subunits (R1-6, ꢀ1-4, γ1-4, δ, ꢀ, π,
θ and F1-3) have been cloned and sequenced. Each of the
subunit isoforms is encoded by a single gene, although additional
heterogeneity is introduced by alternative splicing in a number
of cases.7 This plethora of subunits may suggests that there exists
a vast array of GABAA receptors subtypes, but this does not
appear to be the case.
GABAA-receptor heterogeneity is expected to provide the
basis for flexibility in signal transduction and drug-induced
allosteric modulation. However, the most common form of the
GABAA receptor contains R, ꢀ, and γ subunits8 and recombinant
receptors containing these subunits mimic the biological,
electrophysiological, and pharmacological properties of the
native GABAA receptors. The benzodiazepine (BZ) binding site,
the so-called benzodiazepine receptor (CBR), occurs at the
interface of the R and γ subunits,9 and although the γ2-subunit
is essential to convey classical BDZ sites to recombinant
receptors, the R-subunit variants largely determine their phar-
macological profiles.10 Thus, type I (ω1) and type II (ω2)
benzodiazepine receptors are formed upon coexpression of the
subunits R1ꢀxγ2 and R2- or R3ꢀxγ2, respectively,11 whereas
receptors containing the R4- or R6-subunits have very low
affinities for classical benzodiazepines such as diazepam (7-
chloro-1,3-dihydro-1-methyl-5-phenyl-2H-1,4-benzodiazepin-2-
one).12 In particular, from independent research studies based
on gene “knock in” and “knock out” strategies,7 it has been
GABA (γ-aminobutyric acid) is the most important inhibitory
neurotransmitter in the brain, being able to control the excit-
ability of many central nervous system (CNS)a pathways. GABA
exerts its physiological effects by binding to three different
receptor types in the neuronal membrane: the GABAA, GABAB,
and GABAC receptors. The GABAB receptor belongs to the
G-protein-coupled receptors superfamily,1 while the GABAA
2
3
and GABAC receptors are ligand-gated chloride ion channel
complexes. GABAA receptors, which are responsible for the
majority of neuronal inhibition in the mammalian CNS, mediate
the actions of many pharmacologically useful agents, including
benzodiazepines (BDZs), barbiturates, neuroactive steroids,
anesthetics, and convulsants.4,5
Electron microscopy studies of the native GABAA receptors
have shown that they are composed of five subunits arranged
pseudosymmetrically around the chloride ion channel pathway,
which passes through the cell membrane, and the receptor
appears as a “doughnut” with a diameter of around 8 nm when
viewed from the cell exterior.6
* To whom correspondence should be addressed. Phone: +39 577
234173. Fax: +39 577 234333. E-mail: anzini@unisi.it.
†
Dipartimento Farmaco Chimico Tecnologico, Universita` degli Studi
di Siena.
‡
Dipartimento di Chimica Farmaceutica e Tossicologica, Universita` di
Napoli “Federico II”.
§
Dipartimento di Psichiatria, Neurobiologia Farmacologia e Biotecnolo-
gie, Universita` di Pisa.
|
Dipartimento di Farmacologia Preclinica e Clinica “M. Aiazzi Mancini”,
Universita` degli Studi di Firenze.
a Abbreviations: CNS, central nervous system; CBR, central benzodi-
azepine receptor; BDZ, benzodiazepine; iBDZ, imidazobenzodiazepine;
DPPA, diphenyl phosphorazidate; PTZ, pentylenetetrazole; MD, molecular
dynamics.
Dipartimento di Chimica, Universita` degli Studi di Siena.
Rottapharm S.p.A., Monza.
#
Department of Chemistry, Rutgers State University.
Present address: 421 Morningwood Drive, Olney, MD 20832.
O
10.1021/jm8002944 CCC: $40.75
2008 American Chemical Society
Published on Web 07/19/2008

New Benzodiazepine-3-carboxylate as an Antianxiety Agent
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4731
Chart 1. Title and Reference Compounds
evidenced that the sedative and anticonvulsant actions are
mediated through R1-containing GABAA receptors, while R2-
containing receptors may be involved in anxiolytic-like activity
because these have a preferential brain distribution in circuits
mediating emotional behavior. Furthermore, R2-, R3-, and/or R5-
containing receptors may mediate the muscle relaxant effect,7,13
even if it is now well established that R5-containing receptors
are associated with cognition and memory while the role of R3-
containing receptors remains unclear or, at least, they seem to
be partly involved in mediating anxiety.14,15
carboline-3-carboxylate (ZK 93426),27 PrCC,28 and ꢀ-CCt29)
bind without intrinsic activity.
In particular, among classic 1,4-benzodiazepines, 7-nitro-
derivatives as nitrazepam (1), nimetazepam (2), clonazepam (3),
and flunitrazepam (4) (Chart 1) act as full agonists at central
benzodiazepine receptor (CBR) and are marketed as hypnotics,
anticonvulsants, and sedative/muscle relaxants, respectively,
while in other classes of nitro-substituted ligands at CBR the
different position of the NO2 group present in the scaffold or
in the pendant phenyl ring leads to compounds showing different
intrinsic activity such as Ro 23-2896 (5), an imidazobenzodi-
azepine 4′-nitroderivative which behaves as an antagonist,30
while the indol-3-yl-glioxylamides 8a-e proved to act as partial
agonists, antagonists, or inverse agonists depending on the
substitution pattern considered.31,32
Thus, because little information has been published on
structure-activity relationships (SAR) in the 6-aryl series of
imidazobenzodiazepine esters, in this paper we report on the
synthesis of the unusually substituted imidazo[1,5-a][1,4]ben-
zodiazepine derivative 7 (along with its regioisomer 6), which,
acting as a partial agonist at CBR showed in some animal
models a very interesting anxioselective property in the absence
of the unwanted effects typical of classic 1,4-benzodiazepines.
Moreover, owing to the different behavior showed by com-
pounds 6 and 7, molecular modeling studies were performed
with the double aim to define their binding modes within a
model of GABAA receptor and to account for their different
intrinsic activity.
Ligands acting at the ω modulatory sites can differ by their
intrinsic efficacy.16 At least three classes of compounds have
been identified by their ability to modulate GABA neurotrans-
mission by interacting with this receptor complex. Positive
modulation, which leads to an increase of the GABA-induced
chloride ion flux, is produced by agonists at the ω modulatory
sites such as benzodiazepines [e.g., diazepam, oxazepam (7-
chloro-1,3-dihydro-3-hydroxy-5-phenyl-2H-1,4-benzodiazepin-
2-one), midazolam (8-chloro-6-(2-fluorophenyl)-1-methyl-4H-
imidazo[1,5-a][1,4]benzodiazepine),andtriazolam(8-chloro-6-(o-chlo-
rophenyl)-1-methyl-4H-s-triazolo[4,3-a][1,4]benzodiazepine)],16,17
cyclopyrrolones [e.g., zopiclone [6-(5-chloropyridin-2-yl)-5-oxo-
7H-pyrrolo[3,4-b]pyrazin-7-yl]-4-methylpiperazine-1-carboxy-
late,andsuriclone[6-(7-chloro-1,8-naphthyridin-2-yl)-5-oxo-3,7-di-
hydro-2H-[1,4]dithiino[2,3-c]pyrrol-7-yl]-4-methylpiperazine-1-carbox-
ylate],16,18 triazolopyridazines (e.g., 3-methyl-6-[3-(trifluorom-
rthyl)phenyl]-1,2,4-triazolo[4,3-b]pyridazine (CL 218,872),17
imidazopyridines (e.g., zolpidem (N,N-dimethyl-2-[6-methyl-
2-(4-methylphenyl)imidazo[3,2-a]pyridin-3-yl]acetamide),19 ꢀ-car-
bolines (e.g., ethyl 5-benzyloxy-4-methoxymethyl-ꢀ-carboline-
3-carboxylate (ZK 91296),20 imidazopyrimidines,21 flavones,22
and quinolines.23 Negative modulation, which induces a decrease
of the GABA-induced chloride ion flux, is obtained by inverse
agonists, such as ꢀ-carbolines (e.g., ꢀ-CCM and DMCM),20,24
pyrazoloquinolines (e.g., 2-phenyl-2,5-dihydro-3H-pyrazolo[4,3-
c]quinolin-3-one (CGS 8216),25 whereas competitive antagonists
of the ω modulatory sites (such as the imidazobenzodiazepine
flumazenil,26 the ꢀ-carbolines ethyl 5-isopropoxy-4-methyl-ꢀ-
Chemistry
The synthesis of imidazoester 7 was carried out as depicted
in Scheme 1.
Briefly, the synthetic strategy employed the noncommercially
available 2-amino-5-fluoro-benzophenone 9 as the starting
material. In fact, compound 9 was obtained by means of the
sequential treatment of 5-fluoroanthranilic acid with phenyl
lithium and chlorotrimethysilane.33 Regioselective nitration of

4732 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Scheme 1. Synthesis of Compound 7a
Anzini et al.
Scheme 2. Synthesis of Compound 17a
a Reagents and conditions: (i) (CH3)3SiCl, THF, 0 °C; (ii) KNO3, H2SO4;
(iii) BrCOCH2Br, CH2Cl2; (iv) NH3, CH3OH; (v) CIPO(OEt)2, t-BuO-K+,
dry THF; (vi) CNCH2COOEt, t-BuO-K+.
-
a Reagents and conditions: (i) toluene, -40 °C; (ii) (n-Bu)4N+MnO4
,
benzophenone 9 in the presence of pure KNO3 (99.9%) in
concentrated H2SO4 afforded the hitherto unknown 2-amino-
5-fluoro-3′-nitrobenzophenone 10, which, by means of the
sequential treatment with bromoacetyl bromide in dicho-
loromethane and then with liquid ammonia in ethanol at reflux
afforded 1,3-dihydro-7-fluoro-5-(3-nitrophenyl)-2H-1,4-benzo-
pyridine; (iii) KMnO4, pyridine/H2O; (iv) DPPA, TEA, benzene; (v)
BrCH2COBr, CH2Cl2; (vi) NH3, CH3OH, reflux.
of ethyl isocyanoacetate in the presence of potassium tert-
butoxide, provided the imidazoester 7 in satisfactory yield.34
The synthetic route used to prepare 2-amino-5-fluoro-4′-
nitrobenzophenone 16 involved the low-temperature inverse
addition of 4-fluoro-o-tolylmagnesium bromide to p-nitroben-
zaldehyde to give 5-fluoro-2-methyl-4′-nitrobenzhydrol 13.35
Oxidation of this compound with tetrabutyl ammonium per-
manganate in pyridine gave 5-fluoro-2-methyl-4′-nitrobenzophe-
none 14. Reoxidation of this compound in the presence of
potassium permanganate afforded the benzoyl benzoic derivative
15, which was converted in two steps into the aminobenzophe-
none 16 by reaction with diphenyl phosphorazidate (DPPA).36
In this sequence, the resulting acyl azide underwent Curtius
rearrangement to give the corresponding isocyanate derivative,
which was promptly hydrolyzed into the expected 2-amino-5-
fluoro-4′-nitrobenzophenone 16 (Scheme 2). The transformation
of compound 16 into benzodiazepinone 17 and its successive
imidazo-annulation following the above-reported procedure led
to imidazoester 6.
1
diazepin-2-one 11. H NMR spectrum of compound 10 along
with 1H NMR and X-ray diffraction studies performed on
compound 11 (Figure 1), demonstrated the regioselectivity of
the nitration at 3′-position, probably favored by the contempo-
rary presence of the fluorine atom in the 5-position of ben-
zophenone and the steric hindrance which prevents the reaction
at 2′-position.
The imidazo-annulation of 1,4-benzodiazepin-2-one 11 ac-
complished via the intermediate enol phosphonate 12 by means
Results and Discussion
The binding affinity of the imidazo[1,5-a][1,4]benzodiazepine
esters 6 and 7 at the benzodiazepine receptor in bovine and
human cortical membranes was determined by means of
competition experiments against the radiolabeled antagonist
[3H]flumazenil ([3H]ethyl 8-fluoro-5,6-dihydro-5-methyl-6-oxo-
4H-imidazo[1,5-a][1,4]benzodiazepine-3-carboxylate) and ex-
pressed as the Ki because these compounds inhibited radioligand
binding by more than 80% at a fixed concentration of 10 µM.
The in vitro efficacy of both compounds was measured by the
GABA ratio, which predicts the pharmacological profile of a
BzR ligand.16,37,38 The resultant value, expressed as a ratio of
Figure 1. Crystal structure of compound 11 × H2O. Ellipsoids enclose
50% of probability. (The water molecule is omitted for the sake of
clarity).

New Benzodiazepine-3-carboxylate as an Antianxiety Agent
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4733
Table 1. Inhibition of [3H]Flumazenil Specific Binding to Bovine and Human Cortical Membranes and GABA Ratios of Imidazoesters 6 and 7
Ki a(nM) bovine
cortical mambranes (SEM)
Ki a(nM) human
cortical membranes (SEM)
GABA ratiobbovine
cortical membranes
GABA ratiobhuman
cortical membranes
compd
R
R1
5
6
7
H
F
F
4-NO2
4-NO2
3-NO2
1.80 (IC50)c
14.80 ( 1.4
4.40 ( 0.3
5.17 ( 0.20
1.90 ( 0.09
antagonistc
0.89
1.19
1.68
1.03
20.0 ( 1.8
9.70 ( 0.7
6.5 ( 0.5
2.1 ( 0.08
0.86
1.20
1.60
1.01
flunitrazepam
flumazenil
a Ki values are the means ( SEM of three independent determinations. b GABA ratio ) (Ki without GABA/Ki with 50 µM GABA). c See ref 30.
Ki without GABA/Ki with GABA is nearly 2 for a full agonist,
1 for an antagonist; partial agonists are intermediate between 1
and 2, while a GABA ratio value below 1 is typical for inverse
agonists. Table 1 summarizes the biological data, and it appears
that both compounds 6 and 7 are able to bind to CBR with
higher affinity for bovine cortical membranes than for the human
ones, their Ki values being 14.80 and 4.40 nM vs 20.0 and 9.70
nM, respectively.
In particular, compound 7 appears as active as flunitrazepam,
which shows a Ki of 5.15 nM but less potent than compound 5
(1.8 nM). These data suggest that the presence of a nitro
substituent in the pendant phenyl ring of (iBDZ) leads to very
active compounds, the affinity of which appears to be modified
by the contemporary presence of the fluorine atom in position
Figure 2. 36Cl- uptake measured in rat cerebrocortical synaptoneuro-
somes for Imidazoesters 6 and 7.
8 of the iBz scaffold. The influence of 8-substitution is
particularly evident in compound 6 and to a lesser extent in
compound 7, the affinity of which is moderately lower, being
2.5 and 8.2 times lower when compared to compound 5,
suggesting that the fluorine atom causes an ancillary interaction,
which, though not detrimental, seems to be able to decrease to
some extent the affinity of compounds 6 and 7 for CBR binding
site. It is well-known that fluorine can induce significant
stereoelectronic effects in the organic molecules in which it is
inserted. The difference in electronegativity between carbon and
fluorine generates a large dipole moment in this bond that, when
combined with the electrostatic distribution of a specific
Figure 3. Effect on motor coordination 30 min after the treatment.
Each value represents the mean of at least 15-20 mice ^P < 0.05, *P
< 0.01 vs saline/CMC treated mice.
molecule, may contribute to the molecule’s ability to engage
in intermolecular interactions.
This is particularly true in aromatic systems where the
introduction of fluorine changes the electrostatic distribution of
the molecular surface and may also induce new binding sites
localized in the proximity of the fluorine atoms.39
agonist [GABA ratio ) 1.19 (bovine cortical membranes) or
1.20 (human cortical membranes)], respectively, suggesting that:
(a) in accordance with what was previously stated, 4′-nitro
substitution ensures no agonist activity to compound 6 but not
necessarily leads to an antagonist;30 (b) the hitherto unexplored
3′-nitro substitution, leading to a partial agonist endowed with
a promising antianxiety activity, opens new sceneries in the
study of central benzodiazepine receptor-ligand interactions.
A second and more direct measure of in vitro efficacy was
determined by a 36Cl- uptake assay measured in rat cerebro-
cortical synaptoneurosomes.40,41 The synaptic chloride conduc-
tance effected by GABA activating the GABAA receptor
complex is modulated by ligands acting at the BzR. Full agonists
increase the current, and antagonists have no effect, while
inverse agonists decrease ion flow. The test compounds were
compared to flunitrazepam, flumazenil, ethyl ꢀ-carboline, and
to an imidazo[1,5-a]quinoxaline42 (which was resynthesized ad
hoc and named QNX), being a full agonist, an antagonist, an
inverse agonist, and a pure antagonist, respectively (Figure 2).
On the basis of the results obtained with these standards, it
was possible to establish the course of the uptake and to make
a direct comparison with the compounds analyzed. In particular,
compounds 6 and 7 showed a partial decrease and a partial
increase in the 36Cl- influx, respectively, confirming their
efficacies as partial inverse agonist and partial agonist as was
indicated by their respective GABA ratio value.
Pharmacological Results
Compound 7 was examined in mice in vivo for its pharma-
cological effects. Five potential benzodiazepine actions were
considered: the myorelaxant effect with the rota-rod, the
anticonvulsant action evaluated by means of the new compound
against pentilenetetrazole-induced convulsions, potential anxi-
olytic effects screened using light/dark choice test; compound
7 was also tested for its ethanol-potentiating action, and finally
the mouse learning and memory impairment was evaluated by
passive avoidance test.
Effect on Motor Coordination. The effects of compound 7
on animal motor coordination was investigated using the mouse
rota-rod test as the screening method to discover any myore-
laxant effect (Figure 3). Compound 7 was tested in the range
of doses between 0.3 and 12.5 mg kg-1 po, and it statistically
increased the number of falls showing a muscle relaxant effect
As for the different intrinsic activity at CBR elicited by
compounds 6 and 7, it is evident that 4′-nitro or 3′-nitro
substitution is able to modulate the activity, leading to a partial
inverse agonist [GABA ratio ) 0.89 (bovine cortical mem-
branes) or 0.86 (human cortical membranes)] or to a partial

4734 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Anzini et al.
Table 2. Muscle Relaxant, Anticonvulsant, and Anxiolityc-Like Effects of Compound 7a
muscle relaxant effect, rota-rod test anticonvulsant activity against PTZ-induced attacks antianxiety activity, light-dark box
no. of falls in 30 s
treatment
mg/kg
before treatment
30 min
convulsion latency (min)
death latency (min)
no. of crosses
time (s) in light
CMC 1% saline
diazepam
5.0 ( 0.4
4.8 ( 0.4
4.6 ( 0.5
4.7 ( 0.4
5.1 ( 0.4
4.8 ( 0.3
4.7 ( 0.4
1.6 ( 0.0
4.5 ( 0.4c
2.3 ( 0.4
2.2 ( 0.3
2.1 ( 0.4
2.7 ( 0.4
4.4 ( 0.4c
18.0 ( 3.6
22.3 ( 4.4
11.4 ( 3.6
7.5 ( 2.1c
14.6 ( 3.5
8.6 ( 2.2b
5.6 ( 3.2c
5.4 ( 3.1c
5.1 ( 2.3c
119.6 ( 11.5
175.3 ( 10.4c
123.7 ( 8.5
132.0 ( 10.6c
169.3 ( 9.6c
181.2 ( 7.7c
186.8 ( 9.9c
1.0 ip
0.3po
1.0po
3.0po
6.5po
12.5po
no convulsion
no death
7
7
7
7
7
18.7 ( 4.8
19.3 ( 4.5
23.6 ( 5.1
21.9 ( 5.3
22.5 ( 6.7
39.8 ( 7.1c
a Each value represents the mean of at least 15-20 mice. b P < 0.05. c P < 0.01 vs saline/CMC treated mice.
Figure 4. Effect on antianxiety activity. Each value represents the mean of at least 15-20 mice ^P < 0.05, *P < 0.01 vs saline/CMC treated mice.
only at the highest dose employed; this effect is comparable to
that exhibited by reference drug diazepam (1.0 mg kg-1 ip).
Effect on Pentylenetetrazole (PTZ)-Induced Convulsions.
The anticonvulsant activity was studied by means of pentyle-
netetrazole [6,7,8,9-tetrahydro-5H-tetrazoloazepine (PTZ)] as a
chemical convulsant agent. Compound 7 at the same dose, which
showed myorelaxant effect (12.5 mg kg-1 po), was also able
to increase the mouse death latency but not the convulsion
latency. The anticonvulsant efficacy of compound 7 is lower
than that of diazepam, which was able to completely prevent
the attacks (Table 2). In the same experimental conditions, none
of the other doses of compound 7 showed any anticonvulsant
action.
Antianxiety Effect. Effects on mouse anxiety of newly
synthesized molecule and diazepam were studied by a light/
dark box apparatus. In our experiments, compound 7 showed a
statistically significant anxiolytic-like effect starting from the
dose of 1 mg kg-1 po as demonstrated by both the reduction of
transfer number from one compartment to the other and the
increase of the time spent in the light compartment of the light/
dark apparatus (Figure 4). Compound 7, at higher doses 3.0,
6.5, and 12.5 mg kg-1 po, potentiated both parameters in a dose-
dependent manner. Compound 7 showed an efficacy comparable
to that of the reference drug, while the potency was lower than
that of diazepam. The antianxiety action of compound 7
(3.0-12.5 mg kg-1 po) was also confirmed in two other well-
known tests used for the investigation of antianxiety-like effect:
rat lick monitor suppression and rat elevated plus maze tests
(data not shown).
induction time in statistically significant manner (Figure 5). The
examined compound exerted an hypnotic effect at the same dose
that exerted also myorelaxant and antianxiety action. Lower
doses of compound 7 (3.0-6.5 mg kg-1 po) were devoid of
any activity on both induction and endurance time. As expected,
diazepam at the dose of 1.0 mg kg-1 ip exhibited hypnotic action
as shown by the increase in ethanol-induced sleeping time.
Effects on Cognitive Processes. To investigate the effect
that the new compound has on learning and memory, the mice
performance on passive avoidance test was investigated. In this
assay, the difference between the retention latencies of com-
pound 7-treated mice and CMC-treated controls were statistically
significant only at the highest dose (12.5 mg kg-1 po) (Figure
6). Compound 7 showed an amnesic activity similar to that
exerted by the reference drug diazepam (1.0 mg kg-1 ip). Lower
doses of compound 7 did not show any effect on learning and
memory processes.
Molecular Modeling Studies. So far, several models of
ligands interaction with ω modulatory sites have been reported
in the literature.46 These models subsequently started from the
simple “two site hypothesis”,47 made by one of us, according
to which two sites are required for recognition at the ω
modulatory sites: a condensed aromatic ring and an electron-
donating group. The distance between these sites was assumed
to modulate intrinsic activity (agonist and antagonist). An
additional electron-donating site was included for antagonists
and inverse agonists.30 Progressing toward the Shove four-site
model48 surely contributed to a better understanding of ligand/
receptor interaction. But all these models do not take into
account the heterogeneity of the ω modulatory sites.
Effect on Ethanol-Induced Sleeping Time. Compound 7,
at the highest dose employed (12.5 mg kg-1 po), potentiated
the ethanol-endurance sleeping time without modifying the
Cook and co-workers49 very recently published an updated
unified pharmacophore model for the benzodiazepine receptor

New Benzodiazepine-3-carboxylate as an Antianxiety Agent
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4735
Figure 5. Sleeping time effect. Each value represents the mean of at least 15-20 mice ^P < 0.05, *P < 0.01 vs saline/CMC treated mice.
Figure 6. Learning and memory test. Each value represents the mean of at least 15-20 mice ^P < 0.05, *P < 0.01 vs saline/CMC treated mice.
based on structure-activity relationship studies of more than
150 different ligands from 15 structurally different classes of
compounds. This model was developed by assuming that
benzodiazepine receptor agonists, antagonists, and inverse
agonists share the same binding pocket. It identifies receptor
areas of special physicochemical properties that should be
located in a well-defined spatial arrangement in order to be
counterbalanced by the ligands. In the unified pharmacophore
model, H1 and A2 are hydrogen bond donor and acceptor sites,
respectively, whereas H2/A3 is a bifunctional hydrogen donor/
acceptor site, L1, L2, L3, and LDi are lipophilic pockets, and S1,
S2 and S3 denote regions of negative steric interactions.
Through a deep examination of affinity and efficacy SAR, a
ligand-receptor interaction was rationalized for the new imi-
dazoesters 6 and 7, assuming as topological model the above-
cited pharmacophore/model proposed by Cook et al., which
compounds 6 and 7 show to fit almost well. In particular, the
imidazobenzodiazepine system binds the GABA/Bz complex
through N2 and N5 atoms by means of a hydrogen bond
involving H1 and H2 donor sites, respectively, and the lipophilic
region L1 appears to coincide with the benzofused moiety.
However, the Cook’s unified pharmacophore model seems not
to be able to explain how the slight structural differences of
compounds 6 and 7 are responsible for the opposite intrinsic
activity (partial inverse agonist vs partial agonist) elicited,
respectively, by the two compounds.
To better understand the ligand-receptor interactions, and
with the aim to deepen their mode of binding, compounds 6
and 7 were submitted to molecular docking calculations by
means of the homology model of the GABAA receptor, proposed
by Cromer et al. and built starting from the crystal structure of
the nACh receptor.50 It is now well-established that the
benzodiazepine binding site is located between R and γ subunits.
Many residues involved in the benzodiazepine binding site have
been identified through mutagenesis experiments and are shown
in Figure 7.
It is also well-known that the benzodiazepine ring is not
planar and it is able to adopt two preferred conformations that
are in equilibrium, so it is difficult to assess unambiguously
which of the two conformations is the bioactive one because
many factors such as the presence of substituents on the
diazepine ring as well as the receptor environment play a key
role in determining the binding conformation.51
This is the reason why, in this study, both conformations were
considered for compounds 6 and 7 having, respectively, the
pendant phenyl ring flipping above and below the plane formed
by the imidazofused moiety. For the sake of brevity, only the
conformations of compound 7 are reported in Figure 8.
By means of the AutoDock program,52 the two conformations
of compounds 6 and 7, respectively, were subjected to 100
independent runs of docking simulations. Cluster analysis was
then performed with a rms tolerance of 1.0 Å. The lowest energy

4736 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Anzini et al.
His102, Tyr160, Tyr210, and polar interactions established with
Thr207 and Thr142 side chains.
On the other hand, in the alternative binding mode B,
compounds 6 and 7 retain two polar interactions with Thr207
and Thr142, but three important aromatic interactions with
Phe100, Tyr160, and Tyr210 are missing.
Thus, only the binding mode A is in full agreement with the
experimental data according to which residues such as Tyr210,
Phe77, and Thr207 are essential for benzodiazepines selectiv-
ity.53 Furthermore, while it was demonstrated that Thr142 affects
the efficacy of benzodiazepines,54 His102 proved to be deeply
involved in the benzodiazepines binding.55 In addition, the
replacement of Tyr160 by a serine residue resulted in the loss
of [3H]flumazenil binding and it has also been suggested that
Phe77 interacts with the pendant phenyl group of fluni-
trazepam.56
Consequently, taking into account the above-mentioned
energetic differences and the agreement with the main mu-
tagenesis data, the conformers with the pendant phenyl ring
above the imidazofused moiety seem to represent the most
probable receptor binding conformations (binding mode A) of
both 6 and 7 (Figures 9 and 10).
Figure 7. Representation of the structure of the benzodiazepine-binding
site. The R- and γ-subunits are colored in blue and pink, respectively,
while the main interacting residues are represented as atom type colored
stick. Hydrogen atoms are not displayed for the sake of clarity.
According to our docking results, it clearly emerges that a
network of aromatic interactions governs the binding of
compounds 6 and 7. Indeed, the pendant phenyl and the
benzofused ring of the benzodiazepine moiety of both com-
pounds were found to form an off-centered parallel displaced
π-π interaction with Phe77 and His102, respectively, whereas
the diazepine portion and the imidazofused ring forms stacking
interactions with Tyr160 and Tyr210.
In particular, the benzofused ring was found to occupy a
region shaped by His102, Asn103, and Phe100 in perfect
agreement with a recent study highlighting the involvement of
such residues in the imidazobenzodiazepines binding. Indeed,
when His102, Asn103, and Phe100 were individually mutated
into cysteine, they were found to interact covalently with
7-isothiocyano-1,4-benzodiazepines and the substituent at posi-
tion 8 of the imidazobenzodiazepines like flumazenil, 8-azido-
5,6-dihydro-5-methyl-6-oxo-4H-imidazo[1,5-a][1,4]benzodi-
azepine-3-carboxylic acid ethyl ester (Ro 15-4513), and imid-
NCS compound.57
As important as the hydrophobic interactions are the polar
contacts between the carboxyethyl groups of both compounds
and Thr207 and Thr142 side chains by means of two hydrogen
bonds, while in the case of the partial agonist 7, the 3′-nitro
group is also able to form an additional hydrogen bond with
Thr193 side chain (Figure 9).
Figure 8. Two possible conformations of compound 7: (a) with the
pendant phenyl ring above the imidazofused plane (white), (b) with
the pendant phenyl ring below the imidazofused plane (orange).
solutions resulting from the four docking calculations in each
case belong to the most populated cluster and were analyzed
for their consistency with experimental data.
It is worth noting that, in the case of the inverse agonist 4′-
nitro derivative 6, in all the binding conformations proposed
by the docking program, no one showed a favorable interaction
of the nitro group with any receptor counterpart (Figure 10).
Interestingly, the 5-nitro indole derivatives 8a and 8c,
previously reported by Primofiore et al.,58 are both endowed
with a partial agonist activity (GR ) 1.17 and GR ) 1.13,
respectively), while compounds 8d, 8e, and 8b bearing a nitro
group inserted in 4′ position of the pendant phenyl ring, just as
our compound 6, show an inverse agonist activity (8d, GR )
0.75; 8e, GR ) 0.73; 8b, GR ) 0.82).58 Intrigued by a possible
role of the nitro group in the modulation of the intrinsic activity
of ligands, we decided to dock all these derivatives into the
Cromer’s GABAA receptor model. For all the indole derivatives,
two main interactions were observed. The pendant phenyl ring
of compounds 8a, 8b, 8d, 8e, and the indane fragment of 8c
were found to form an aromatic interaction with Tyr160, while
The results obtained from docking simulations of 6 and 7
suggest two alternative binding modes, one for each benzodi-
azepine ring conformation, here referred to as A and B. In the
case of binding mode A, having the pendant phenyl ring above
the imidazofused moiety, the estimated free energy of binding
(BE ) -8.57 kcal/mol for 6 and -10.86 kcal/mol for 7) is in
good accordance with the nanomolar experimental affinity data
(Table 1). In a different way, binding mode B, which is
individuated by the conformations having the pendant phenyl
ring below the imidazofused moiety and endowed with an
estimated free energy of binding, (-5.71 and -6.29 kcal/mol
for 6 and 7, respectively), correspond to a micromolar EK in
i
open contrast with the experimental binding data (Table 1).
In particular, the binding mode A of compounds 6 and 7
highlighted the aromatic interactions of the benzodiazepine
moiety with hydrophobic residues such as Phe77, Phe100,

New Benzodiazepine-3-carboxylate as an Antianxiety Agent
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4737
Figure 9. Compounds 7 and 8a docked into the CBR binding site. The R- and γ-subunits are colored in blue and pink, respectively, while the main
interacting residues are represented as atom type colored stick. Hydrogen bonds are indicated as white dashed lines.
Figure 10. Compounds 6 and 8d docked into the CBR binding site. The R- and γ-subunits are colored in blue and pink, respectively, while the
main interacting residues are represented as atom type colored stick. Hydrogen bonds are indicated as white dashed lines.
a hydrogen bond between indole nitrogen and His102 occurred.
Moreover, just for 5-nitro derivatives 8a and 8c, an aromatic
interaction between the indole ring and the Phe77 was observed.
For both compounds 8a and 8c, the 5-nitro group was found to
form a hydrogen bond with Thr193, similarly to imidazoben-
zodiazepine 7 (Figure 9). On the contrary, in none of the inverse
agonists 8d, 8e, and 8b, the 4′-nitro group is capable of making
such hydrogen bond being embedded in a pocket formed by
residues Met130, Tyr141, Arg132, Ala161, and involved in a
polar interaction with the Met130.
The binary complex formed by compound 7 and CBR was
then subjected to an extensive molecular dynamics study (MD)
(Figure 11) aiming at investigating the energy profile, the
stability of the above-proposed binding mode, and the main
intermolecular interactions based on a large time-scale simula-
tion. During 3 ns of MD, the complex showed an energetic and
geometric stable profile (see Supporting Information for more
details), reinforcing the significance of our docking results.
Indeed, the binding mode of 7 does not substantially diverge
from that calculated by AutoDock program, as it keeps all the
above-described main interactions with the protein. In particular,
the benzodiazepine moiety still engages stacking interactions
with His102 and Tyr106 while the imidazofused ring improves
the geometry of its π-π interaction with Tyr210 side chain.
On the other hand, the phenyl pendant branch enforces its
hydrophobic contacts by adding to the previously observed
stacking with Phe77 a supplementary T-shaped interaction with
Tyr58. Furthermore, during the whole MD simulation, the nitro
group lies always near to Thr193 interacting with either the
hydroxyl group or the backbone NH.
It is noteworthy that, during the production run, the polar
interaction between the Thr142 side chain and the ester function
of 7 was completely lost in favor of a new better directed
H-bond of the Thr142 hydroxyl group with the N5 of the
diazepine ring. The ester function of 7, on the contrary, makes
a H-bond with one of the surrounding water molecules occupy-

4738 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Anzini et al.
having minimal benzodiazepine side effects yet is equieffective
to diazepam in a classical animal model of anxiety such as the
light-dark box test. Furthermore, SAR and different pharma-
cological profile of the title compounds were rationalized in
terms of ligand interaction at CBR by means of a modeling
study using the GABAA receptor homology model developed
by Cromer et al. According to docking and molecular dynamics
simulations, it appears that the position of the nitro substituent
in the ligands is considered responsible for the intrinsic activity
although further experimental tests (e.g., synthesis of new
ligands and receptor mutagenesis studies) are required to confirm
such a hypothesis.
Experimental Section
Chemistry. All chemicals used were of reagent grade. Yields
refer to purified products and are not optimized. Melting points
were determined in open capillaries on a Gallenkamp apparatus
and are uncorrected. Microanalyses were carried out by means of
Perkin-Elmer 240C or a Perkin-Elmer series II CHNS/O analyzer,
model 2400. Merck silica gel 60 (230-400 mesh) was used for
column chromatography. Merck TLC plates and silica gel 60 F254
were used for TLC. 1H NMR spectra were recorded with a Bruker
AC 200 spectrometer in the indicated solvent (TMS as internal
standard). The values of the chemical shifts are expressed in ppm,
and the coupling constants (J) are expressed in Hz. Mass spectra
were recorded on a Varian Saturn 3 or a ThermoFinningan LCQ-
deca spectrometer.
Procedures for the Preparation of Substituted 2-Aminoben-
zophenones (9), (10), and (16). 2-Amino-5-fluorobenzophe-
none (9). A stirred solution 1.25 M (titrated according to the
literature procedure)60 n-butyllithium in hexane (24.2 mL, 38.8
mmol) was treated dropwise at -40 °C with 2-bromobenzene (3.36
mL, 35.5 mmol) in dry diethyl ether (40 mL) over 0.5 h under a
nitrogen atmosphere. The temperature was kept at -40 °C during
the addition. The resulting dark-orange solution was stirred for a
further 0.5 h at -40 °C. In a separate flask, a solution of 2-amino-
5-fluorobenzoic acid (7.9 mmol) in dry THF (30 mL) also under a
nitrogen atmosphere and at O °C (ice bath) was added in one portion
to the solution prepared as described above, the mixture was stirred
for 2 h at 0 °C and then treated with freshly distilled Me3SiCl (20
mL, 160 mmol) steel under stirring. The reaction mixture was
allowed to warm to room temperature (ca. 10 min.) and hydrolyzed
with 1N HC1 (60 mL). The resulting two-phase system was
separated. The aqueous phase was neutralized with a 3N NaOH
solution and extracted with diethyl ether (3 × 50 mL). The
combined organic extractions were dried over Na2SO4, and the
solvent was evaporated. The resulting brownish residue was purified
by flash chromatography with dichloromethane as eluent. Recrys-
tallization of the product from a methylene chloride/hexane mixture
gave 2-amino-5-fluorobenzophenone as yellow needle-like crystals.
Analytical and spectroscopic data were consistent with those
reported in literature.61
Figure 11. Snapshot of the MD simulation of the complex between
compound 7 and CBR selected within a stable simulation interval. The
main interacting residues and compound 7 are represented as stick
colored by atom type. Proteins are represented as a cartoon with the
R- and γ-subunits colored in blue and pink, respectively. For the sake
of clarity, only polar hydrogens of the interacting residues are shown.
ing the region in apposition to Ala79, which was found to be
involved in the binding of the substituent at position 3 of the
imidazofused moiety by many experimental data.49,59
The only striking difference between docking and MD results
resides in the location of the imidazofused ring, which does
not support any favorable interaction according to the docking
results, while during MD simulation, it is permanently placed
in such a way as to form a double H-bond with the backbone
NHs of Glu209 and Tyr210.
Nevertheless, because several sources of variability such as
the choice of the GABAA receptor model with all its uncertain-
ties (chosen template, correctness of the alignment, loops
refinement) have been included in this study, the choice of the
docking algorithm and the resulted binding modes of our
compounds have to be interpreted at a qualitative level. As a
general comment, the good coherency between the docking and
the MD results supports our proposed binding mode of
compound 7, suggesting a pivotal role of the nitro group present
in the pendant phenyl ring in determining the intrinsic activity
of such imidazobenzodiazepine compounds.
2-Amino-5-fluoro-3′-nitrobenzophenone (10). A solution of
potassium nitrate 99.9% (4.4 mmol) in 3.5 mL of concentrated
sulfuric acid was added dropwise to a solution of 2-amino-
fluorobenzophenone (11) in concentrated sulfuric acid (4.0 mL) kept
at ice bath temperature. After addition was complete, the reaction
mixture was stirred continuously in the ice bath for 30 min and
then at room temperature for 4 h. The solution was poured into
100 mL of ice water and neutralized with concentrated ammonium
hydroxide, giving a yellow precipitate that was collected by filtration
and washed thoroughly with water until the washing was neutral.
The precipitate was then dissolved into 150 mL of methylene
chloride, and any insoluble material was removed by filtration. The
filtrate was washed with water and brine, dried (magnesium sulfate),
and filtered. The solvent was removed at reduced pressure, and the
residue was recrystallized from methanol to give an analytical
sample melting at 129-130 °C as pale-yellow needles.
Conclusion
Ethyl 8-fluoro-6-(3-nitrophenyl)- and ethyl 8-fluoro-6-(4-
nitrophenyl)-4H-imidazo[1,5-a][1,4]benzodiazepine 3-carboxy-
late showed to be high affinity ligands for CBR. By control of
NO2 substituent at the 3′- and 4′-positions and halogen at the
8-position, analogues show a different efficacy from antagonist
(5) and inverse agonist (6) to partial agonist (7). Most
importantly, the 3′-nitrosubstituted imidazoester (7) with a partial
agonist profile was identified and found active in selected animal
models of anxiety.
So, consistently with the partial agonist hypothesis, side
effects such as physical dependence, ethanol potentiation, and
muscle relaxation were significantly reduced as compared to
benzodiazepine full agonists. Compound 7 stands out as the one
5-Fluoro-2-methyl-4′-nitrobenzhydrol (13). To a suspension
of magnesium turnings (7.2 mmol) in dry ether (10 mL), a solution

New Benzodiazepine-3-carboxylate as an Antianxiety Agent
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4739
of 2-bromo-4-fluorotoluene (5.2 mmol) in dry ether (5.0 mL) was
added dropwise. The mixture was refluxed for 30 min and, after
cooling at room temperature, was cooled at -40 °C. After the
addition of a solution of 4-nitrobenzaldehyde (5.0 mmol) in dry
toluene (35 mL) was completed, the mixture was stirred at -40
°C for 1 h and then allowed to warm slowly to room temperature.
3N HCl (10 mL) was then added to decompose any remaining
Grignard reagent. The organic phase layer was removed, washed
with bicarbonate, and dried over magnesium sulfate. Evaporation
of the solvent gave a brownish oily residue, which was purified by
means of flash-chromatography, eluting with CHCl3/EtOAc (9:1
v/v) to give pure 13 as yellowish oil.
5-Fluoro-2-methyl-4′-nitrobenzophenone (14). To a solution
of compound 13 (0.03 mmol) in dry pyridine (4.0 mL), tetrabutyl
ammonium permanganate (0.08 mmol), dissolved in the same
solvent (5.0 mL), was slowly added and the mixture, after keeping
at room temperature for 20 min while stirring, was poured onto an
ice-cooled solution of 2N HCl (20.0 mL) containing NaHSO3 (0.8
mmol). The organic phase was removed and extracted with CHCl3,
washed with water to neutrality, then with brine, and dried over
sodium sulfate. After filtration and evaporation of the solvent the
yellowish residue was flash-chromatographed using n-hexane/
EtOAc 8:2 v/v as eluent to give pure 14 as a light-yellow solid.
Recrystallization from methanol gave an analytical sample melting
at 93-99 °C as colorless needles.
2-(4-Nitrobenzoyl)-4-fluorobenzoic Acid (15). To a solution of
5-fluoro-2-methyl-4′-nitrobenzophenone (14) (3.31 mmol) in py-
ridine (15 mL), H2O (15 mL), and potassium permanganate (3.31
mmol) were added in sequence. The mixture was refluxed under
stirring for 6 days, adding KMnO4 (0.38 mmol) and a solution of
H2O/pyridine (1:1 v/v) (5.0 mL) every 24 h. When the reaction
was completed, the mixture was cooled and MnO2 formed was
filtered off through celite. The acidification of the filtrate by means
of 3N HCl gave a white precipitate, which was separated by
filtration, washed to neutrality with water, and air-dried to give the
expected acid fairly pure which was used as such in the next step.
2-Amino-5-fluoro-4′-nitrobenzophenone (16). To a mixture of
compound 17 (0.34 mmol) in dry benzene (5.0 mL) in the presence
of triethylamine (0.34 mmol), cooled at 0-5 °C, DPPA (0.34 mmol)
was added. The resulting mixture was stirred at room temperature
for 3 h, heated to reflux for 4 h, and cooled, then 7 mL of 50%
sulfuric acid were added. After overnight stirring at room temper-
ature, the reaction mixture was poured into ice-water, made
alkaline with concentrated NH4OH and extracted with dichlo-
romethane (3 × 20 mL). The combined extracts were washed to
neutrality with water, dried over sodium sulfate, and concentrated
under reduced pressure to give a residue as an orange solid.
Purification of the residue by flash chromatography with dichlo-
romethane as eluent gave pure 16 (0.19 mmol, 56%) as a yellow
solid, which after recrystallization from methanol afforded an
analytical sample melting at 146-147 °C as yellow prisms.
General Procedure for the Synthesis of Substituted 1,4-
Benzodiazepinones (11) and (17). A solution of the suitable
substituted 2-aminobenzophenone 10 or 16 (5.2 mmol) in dichlo-
romethane (12 mL) was cooled in an ice bath. The mixture was
stirred while a solution of bromoacetyl bromide (6.17 mmol) and
an aqueous sodium carbonate solution (20%) were added alterna-
tively, keeping the solution slightly basic. After the addition was
completed, the reaction mixture was stirred at room temperature
for 1 h. The precipitate formed was collected by filtration to afford
the respective bromoacetanilide as an almost pure fine yellow
powder which was used without any further purification in the next
step. Approximately 100 mL of anhydrous liquid ammonia were
condensed from a dry ice condenser in a stirred solution of the
proper acetanilide (3.0 mmol) in dichloromethane (50 mL). After
4 h of reflux, the condenser was removed in order to allow the
ammonia to evaporate. The solution was then washed with water
until the washings were neutral, then washed with brine, dried over
magnesium sulfate, and filtered. After the solvent was removed
under reduced pressure, the correspondent 2-amino acetamide was
obtained as a yellow oil. Without further purification, this compound
was heated to reflux in 50 mL of ethanol for 3 h in order to effect
ring closure. On cooling, a yellow precipitate was formed and
collected by filtration to give the expected substituted 1,4-
benzodiazepinone 11 or 17, which after recrystallization from the
suitable solvent gave the corresponding analytical sample.
7-Fluoro-1,3-dihydro-5-(3-nitrophenyl)-2H-1,4-benzodiazepin-
2-one (11). Pale-yellow needles from cyclohexane/EtOAc, mp
207-208 °C (yield 85%). 1H NMR (CDCl3) δ ppm: 4.35 (s, 2H),
6.92-6.97 (dd, 1H, J ) 2.23, 8.72), 7.27-7.32 (m, 2H), 7.54-7.62
(t, 1H), 7.85-7.89 (t, 1H), 8.29-8.32 (d, 1H, J ) 6.55), 8.42-8.44
(t, 1H), 9.85 (bs, 1H, exchangeable with D2O).
7-Fluoro-1,3-dihydro-5-(4-nitrophenyl)-2H-1,4-benzodiazepin-
2-one (17). Pale-yellow needles from cyclohexane/EtOAc, mp
211-212 °C (yield 78%). 1H NMR (CDCl3) δ ppm: 4.23 (s, 2H),
6.90-6.96 (dd, 1H, J ) 2.45-8.31), 7.14-7.32 (m, 2H), 7.71-7.76
(d, 2H, J ) 8.76), 8.22-8.26 (d, 2H, J ) 8.79), 8.82 (bs, 1H,
exchangeable with D2O).
General Procedure for the Synthesis of Substituted 4H-
Imidazo[1,5-a][1,4]benzodiazepine-3-ethoxycarbonyl Derivatives
(6) and (7). A solution of the suitable 1,4-benzodiazepin-2-one (11
or 17) (1.21 mmol) and potassium t-butoxide (2.74 mmol) in dry
THF (30 mL) was stirred in an ice bath for 10 min under nitrogen
and was then treated with diethyl chlorophosphate (4.63 mmol).
After stirring for 30 min, ethyl isocyanoacetate (6.1 mmol) and
potassium t-butoxide (6.6 mmol) were added (the solution turned
dark brown). The mixture was continuosly stirred in the ice bath
for 1 h, was allowed to stir at room temperature overnight under
nitrogen and after addition of acetic acid (1.8 mL), was stirred for
an additional 20 min, and was then poured onto crushed ice to give
a brownish solid. After filtration, the solid was dissolved into
dichloromethane and the organic layer washed with brine, dried
over Na2SO4, filtered, and concentrated in vacuo. The residue
purified by flash-chromatography eluting with the suitable solvent
afforded the expected imidazoester 6 or 7 as a yellow solid, which
after a recrystallization from the suitable solvent gave an analytical
sample.
Ethyl 8-fluoro-6-(4-nitrophenyl)-4H-imidazo[1,5-a][1,4]benzo-
diazepine-3-carboxylate (6). Pale-yellow needles from benzene/
1
cyclohexane mp 217-218 °C (yield 30%). H NMR (CDCl3) δ
ppm: 1.38-1.45 (t, 3H, J ) 7.08), 4.13 (bs, 1H), 4.40-4.43 (q,
2H, J ) 7), 6.12 (bs, 1H), 7.03-7.09 (dd, 1H, J ) 2.62, 8.12),
7.38-7.48 (m, 1H), 7.60-7.67 (m, 1H), 7.69-7.73 (d, 2H, J )
8.72), 7.91 (s, 1H), 8.21-8.25 (d, 2H, J ) 8.65). MS: m/z 394
(M+ 70).
Ethyl8-Fluoro-6-(3-nitrophenyl)-4H-imidazo[1,5-a][1,4]benzo-
diazepine-3-carboxylate (7). Pale-yellow platelets from EtOAc/
1
MeOH, mp 222-223 °C (yield 30%). H NMR (CDCl3) δ ppm:
1.38-1.45 (t, 3H, J ) 7.38), 4.09 (bs, 1H), 4.36-4.47 (q, 2H, J )
7.08), 6.11 (bs, 1H), 7.06-7.12 (dd, 1H, J ) 2.8, 8.63), 7.39-7.49
(m, 1H), 7.53-7.68 (m, 2H), 7.89-7.93 (d, 2H, J ) 7.89),
8.23-8.28 (d, 1H, J ) 7.99), 8.36 (s,1H). MS: m/z 394 (M+ 70).
X-Ray Crystallography. A single crystal of 11 × H2O was
submitted to X-ray data collection by means of a Siemens P4 four-
circle diffractometer at 293 K equipped with a graphite monochro-
mated Mo KR radiation (λ ) 0.71069 Å). The structure was solved
by direct methods implemented in the SHELXS-97 program.62 The
refinements were carried out by full-matrix anisotropic least-squares
on F2 for all reflections for non-H atoms by means of the SHELXL-
97 program.63
Crystallographic data (excluding structure factors) of this crystal
structure have been deposited with the Cambridge Crystallographic
Data Centre as supplementary publication numbers CCDC 675538
(11 × H2O). Copies of the data can be obtained, free of charge, on
application to CCDC, 12 Union Road, Cambridge CB2 1EZ, UK
[fax: 144-(0)1223-336033 or E-mail: deposit@ccdc.cam.ac.uk].
Radioligand Binding Studies. [3H]flumazenil (specific activity
70.8 Ci/mmol) was obtained from Perkin-Elmer Life Science
(Milan, Italy). All other chemicals were reagent grade and were
obtained from commercial suppliers.
Bovine cortex was obtained from the local slaughterhouse.
Human cortex samples were taken postmortem at the Department

4740 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Anzini et al.
of Pathological Anatomy, University of Pisa, during autopsy
sessions. The subjects had died from causes not primarily involving
the brain and had not suffered from any psychiatric or neurological
disorders. The time between death and tissue dissection/freezing
ranged from 18 to 36 h. The samples were immediately packed in
dry ice and stored in a -80° freezer. The study was approved by
the Ethics Committee of the University of Pisa, Italy.
soaked with 0.05% polyethylenimine to reduce nonspecific binding
of
36Cl-. The filters were washed three more times with 5 mL of
ice-cold buffer and placed into scintillation vials containing 4 mL
of Ready Protein Beckman scintillation cocktail and radioactivity
was counted in a Packard LS 1600 liquid-phase scintillation ꢀ
counter.
Data are expressed as percent stimulation of 36Cl- uptake above
basal level.
Bovine and human cerebral cortex membranes were prepared in
accordance with Martini et al.64 Briefly, cerebral cortex were
homogenized in 10 volumes of ice cold 0.32 M sucrose containing
protease inhibitors. The homogenate was centrifuged at 1000g for
10 min at 4 °C. the resulting pellet was discarded and the
supernatant was recentrifuged at 48000g for 15 min at 4 °C. Then
the pellet was osmotically shocked by suspension in 10 volumes
of 50 mM Tris-citrate buffer at pH 7.4 containing protease inhibitors
and recentrifuged at 48000g for 15 min at 4 °C. The resulting
membranes were frozen and washed by means of a procedure
previously described for removing endogenous GABA from cerebral
cortex.65 Finally, the pellet was suspended in 10 volumes of 50
mM Tris-citrate buffer pH 7.4 and used in the binding assay. Protein
concentration was assayed by the method of Lowry et al.66 using
bovine serum albumin as the standard. [3H]flumazenil binding
studies were performed as previously reported.67 The [3H]flumazenil
binding was performed in triplicate by incubating aliquots of the
membrane fractions (0.2-0.3 mg of protein) at 0 °C for 90 min in
0.5 mL of 50 mM Tris-citrate buffer, pH 7.4, with approximately
0.2 nM [3H]flumazenil. Nonspecific binding was defined in the
presence of 10 µM diazepam. After incubation, the samples were
diluted at 0 °C with 5 mL of the assay buffer and immediately
harvested onto GF/B filters (Brandel) by means of a harvester and
washing with ice-cold assay buffer. The filters were washed twice
with 5 mL of the buffer, dried, and 4 mL of Ready Protein Beckman
scintillation cocktail added; radioactivity was counted in an Packard
LS 1600 liquid-phase scintillation ꢀ counter.
Compounds were routinely dissolved into DMSO and added to
the assay mixture to make a final volume of 0.5 mL. Blank
experiments were carried out to determine the effect of the solvent
(2%) on binding. At least six different concentrations spanning 3
orders of magnitude, adjusted approximately for the IC50 of each
compound, were used. IC50 values, computer-generated by a
nonlinear formula on a computer program (GraphPad, San Diego,
CA), were converted to Ki values, knowing the Kd values of
radioligand in these different tissues calculated by the Cheng and
Prusoff equation.68 The Kd of [3H]flumazenil binding to cortex
membrane from bovine and human was 0.85 and 0.91 nM, res-
pectively. The GABA ratio was determined by calculating Ki
without GABA/Ki with GABA 50 µM for each compounds.
Functional Efficacy Studies. (36Cl- Uptake Studies). 36Cl-
(specific activity 9.69 µCi/g) was obtained from Perkin-Elmer Life
Science (Milan, Italy). All other chemicals were reagent grade and
were obtained from commercial suppliers.
Pharmacological Methods. The experiments were carried out
in accordance with the Animal Protection Law of the Republic of
Italy, DL no. 116/1992, based on the European Communities
Council Directive of 24 November 1986 (86/609/EEC). All efforts
were made to minimize animal suffering and to reduce the number
of animals involved. Male CD-1 albino mice (22-24 g) and male
Swiss Webster (20-26 g) (Morini Italy) were used. Twelve mice
were housed per cage and fed a standard laboratory diet, with tap
water ad libitum for 12 h/12 h light/dark cycles (lights on at 7:00).
The cages were brought into the experimental room the day before
the experiment, for acclimatization purposes. All experiments were
performed between 10:00 and 15:00.
Rota-Rod Test. The integrity of the animals’ motor coordination
was assessed using a rota-rod apparatus (Ugo Basile, Varese, Italy)
at a rotating speed of 16 rpm The treatment was performed before
the test. The numbers of falls from the rod were counted for 30 s,
30 min after drug administration, and the test was performed
according to the method described by Vaught et al.70
Light/Dark Box Test. The apparatus (50 cm long, 20 cm wide,
and 20 cm high) consisted of two equal acrylic compartments, one
dark and one light, illuminated by a 60 W bulb lamp and separated
by a divider with a 10 cm × 3 cm opening at floor level. Each
mouse was tested by placing it in the center of the lighted area,
away from the dark one, and allowing it to explore the novel
environment for 5 min. The number of transfers from one
compartment to the other and the time spent in the illuminated side
were measured. This test exploited the conflict between the animal’s
tendency to explore a new environment and its fear of bright light.71
Pentylenetetrazole (PTZ)-Induced Seizure. PTZ (90 mg/kg sc)
was injected 30 min after the administration of drugs. The frequency
of the occurrence of clonic generalized convulsions was noted over
a period of 30 min.72
Passive-Avoidance Test. The test was performed according to
the step-through method described by Jarvik et al.73 The apparatus
consisted of a two-compartment acrylic box with a lighted compart-
ment connected to a darkened one by a guillotine door. As soon as
the mouse entered the dark compartment, it received a thermal shock
punishment. The latency times for entering the dark compartment
were measured in the training test and after 24 h in the retention
test. The maximum entry latency allowed in the training and
retention sessions was, respectively, 60 and 180 s.
Ethanol-Induced Sleeping Time Test. Ethanol (4 g/kg ip) was
injected 30 min after drug administration. The duration of a loss
of the righting reflex was measured as the sleep time. The end-
point was recorded as 210 min.
36Cl- uptake was measured in rat cerebrocortical synaptoneuro-
somes as described by Schwartz et al.69 with minor modifications.
Briefly, cerebral cortex was dissected from Sprague-Dawley male
rats suspended 1:10 with ice-cold solution containing 145 mM NaCl,
5 mM KCl, 5 mM MgC2, 1 mM CaCl2, 10 mM HEPES, pH 7 (T1
buffer), and 10 mM D-glucose; they were homogenized with a
glass-glass homogenizer (five strokes) and filtered through three
layers of nylon mesh (160 µm) and a 10 µm Millipore filter. The
filtrates were centrifuged at 1000g for 15 min. After discarding
the supernatant, the pellet was gently resuspended in T1 buffer and
washed once more by centrifugation (1000g for 15 min). The final
pellet containing the synaptoneurosomes was suspended 1:2 in T1
buffer was kept on ice until ready for assay (no longer than 30
min).
Aliquots of synaptoneurosome suspensions (1.5-2 mg of protein)
were preincubated at 30 °C for 10 min prior to the addition of 0.2
µCi of 36Cl-. Drugs were added simultaneously with the 36Cl-
(0.35 mL total assay volume). 36Cl- uptake was stopped 10 s later
by the addition of 5 mL of ice-cold HEPES, followed by vacuum
filtration through glass fiber filters (Whatman GF/B) that had been
Drugs. Diazepam (Valium 10) (Roche) and Pentylenetetrazole
(PTZ) (Sigma) were used. All drugs were dissolved into isotonic
(NaCl 0.9%) saline solution and injected sc/ip. The new compound
was administered by the po route and was suspended in 1%
carboxymethylcellulose sodium salt and sonicated immediately
before use. Drug concentrations were prepared in such a way that
the necessary dose could be administered in a 10 mL/kg volume
of carboxymethylcellulose (CMC) 1% by the po, ip, or sc routes.
Statistical Analysis. All experimental result are given as the
mean ( SEM An analysis of variance, ANOVA, followed by
Fisher’s protected least significant difference procedure for post
hoc comparison, were used to verify significance between two
means of behavioral results. The data were analyzed with the
StataView software for Macintosh (1992). P values of less than
0.05 were considered significant.
Docking Calculations. All the ligands were docked into the CBR
model by means of AUTODOCK 3.0 program.

New Benzodiazepine-3-carboxylate as an Antianxiety Agent
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4741
Docking simulations of the compounds were carried out by
means of the Lamarckian Genetic Algorithm52 and applying a
protocol with an initial population of 50 randomly placed individu-
als, a maximum number of 1.0 × 106 energy evaluations, a mutation
rate of 0.02, a crossover rate of 0.80, and an elitism value of 1.
The pseudo-Solis and Wets algorithm with a maximum of 300
interactions was applied for the local search. One hundred
independent docking runs were carried out for each ligand, and
the resulting conformations that differ by less than 1.0 Å in
positional root-mean-square deviation (rmsd) were clustered to-
gether. The conformer with the lowest free energy of binding was
taken as the representative of each cluster.
Ligand Setup. The structures of the ligands were generated by
means of Discovery Studio software version 1.5. Minimizations
energies were achieved by means of the CHARMm force field74
as implemented in Discovery Studio. Minimizations were carried
out by means of 50 steps of steepest descent and 10000 steps of
conjugate gradient as minimization algorithms, with a rms conver-
gence criterion of 0.01 Å. Partial atomic charges were assigned by
means of the Gasteiger-Marsili formalism.75 All the relevant
torsion angles were treated as rotatable during the docking process,
allowing thus a search of the conformational space.
Protein Setup. The GABAA receptor model was set up for
docking as follows: polar hydrogens were added by means of
Discovery Studio software version 1.5, Kollman united-atom partial
charges were assigned. ADDSOL utility of the AutoDock program
was used to add salvation parameters to the protein structures, and
the grid maps representing the proteins in the docking process were
calculated by means of AutoGrid. The grids, one for each atom
type in the ligand, plus one for electrostatic interactions, were
chosen to be large enough to include not only the hypothetical
benzodiazepine binding site but also a significant part of the protein
around it. As a consequence, for all docking calculations, the
dimensions of grid map was 46 Å × 50 Å × 56 Å with a grid-
point spacing of 0.375 Å.
Energy Refinement of the CBR/Ligand Complexes. The com-
plexes obtained were subjected to an energy minimization by means
of CHARMm force field as implemented in Discovery Studio.
Energy optimizations were carried out by means of 200 steps of
steepest descent followed by 10000 steps of conjugated gradient
with an rms of 0.01 as gradient value.
by means of the Langevin piston Nose-Hoover method.81 Eventually
3 ns of production run in NPT ensemble were carried out.
Van der Waals and short-range electrostatic interactions were
estimated within a 10 Å cutoff with the switch value set at 8 Å
and the pair list distance extended to 13.5 Å. The long-range
electrostatic interactions were assessed by using the particle mesh
Ewald (PME) method,82 with a grid size set to 100, 80, and 90 for
x, y, and z axes, respectively, interpolated with a fourth-order
function and by setting the direct sum tolerance to 10-5
.
All the simulations of the solvated complex were performed with
a time step of 2 fs in combination with the SHAKE83 algorithm to
constrain bond lengths involving hydrogen atoms. Trajectory
coordinates were saved every 2500 steps.
All calculations were performed on a Linux dual-core machine.
The VMD program was used for visualization and data analy-
sis.84
Acknowledgment. We are grateful to Dr Laura Salvini
(C.I.A.D.S., University of Siena, Italy) for the recording of the
mass spectra, to Prof. Stefania D’Agata D’Ottavi for the careful
reading of the manuscript, and Consorzio Interuniversitario
Nazionale per la Scienza e la Tecnologia dei Materiali (INSTM)
for the access to Accelrys software. We are also grateful to Prof.
Cromer for kindly providing us with the homology model of
the GABAA receptor. This work was financially supported by
Consiglio Nazionale delle Ricerche (CNR) and Ministero
dell’Universita` e della Ricerca (MIUR)sProgrammi di Ricerca
di Rilevante Interesse Nazionale (PRIN).
Supporting Information Available: Analytical data of com-
pounds 6 and 7, pharmacological data, docking, and molecular
dynamics calculations. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
(1) Kaupmann, K.; Huggel, K.; Heid, J.; Flor, P. J.; Bischoff, S.; Mickel,
S. J.; McMaster, G.; Angst, C.; Bittiger, H.; Froestl, W; Bettler, B.
Expression Cloning of GABA-B Receptors Uncovers Similarity to
Metabotropic Glutamate Receptors. Nature 1997, 386, 239–246.
(2) Rabow, L. E.; Russek, S.H.; Farb, D. H. From Ion Currents to Genomic
Analysis: Recent Advances in GABAA Receptor Research. Synapse
1995, 21, 189–274.
(3) Johnston, G.A R. GABAC Receptors: Relatively Simple Transmitter-
Gated Ion Channels. Trends Pharmacol. Sci. 1996, 17, 319–323.
(4) Skolnick, P.; Paul, S. The Benzodiazepine/GABA Receptor Chloride
Channel Complex. ISI Atlas Pharmacol. 1988, 2, 19–22.
(5) Doble, A.; Martin, I. L. Multiple Benzodiazepine Receptors: No Reason
for Anxiety. Trends Pharmacol. Sci. 1992, 13, 76–81.
(6) Nayeem, N.; Green, T. P.; Martin, J. L.; Barnard, E. A. Quaternary
Structure of the Native GABAA Receptor Determined by Electron
Microscopic Image Analysis. J. Neurochem. 1994, 62, 815–818.
(7) Rudolph, U.; Crestani, F.; Mo¨hler, H. GABAA Receptor Subtypes.
Dissecting their Pharmacological Functions. Trends Pharmacol. Sci.
1999, 22, 188–194.
(8) Sieghart, W.; Sperk, G. Subunit Composition, Distribution and
Function of GABA-A Receptor Subtypes. Curr. Top. Med. Chem.
2002, 2, 795–816.
Molecular Dynamics Simulations. The pose representative of
the most populated cluster for compound 7 was used as the starting
point for further molecular dynamics simulations.
The binary complex between the benzodiazepine receptor and
(7) was subjected to a molecular dynamics study (MD) by means
of NAMD2 molecular dynamics simulation code.76 Three ns MD
simulations on the complex were carried out in explicit solvent
and periodic boundary conditions. Protein and ligand were param-
etrized by using Xleap module of AMBER977 and the parm99
version78 of the all-atom Amber force field.79 The complex was
subjected to the following procedure prior to molecular dynamics
data collection: the complex was surrounded by a box of 6 Å layer
of TIP3P31 pre-equilibrated water molecules80 and neutralized by
randomly placing 2 Cl- ions by means of the solvate and add ions
routine of Xleap, respectively. A two-step preliminary minimization
was carried out by means of a sophisticated conjugate gradient and
line search algorithm, minimizing in the first 2000 steps water,
counterions, and hydrogens, and in the following 2000 steps, the
loops and all the side chains of the protein. A convergence criterion
on the gradient of 0.0001 kcal mol-1 Å-1 was established. The
microcanonical NVE ensemble (constant volume and energy) was
used to raise the system temperature to 300 K during 43 ps of
Langevin dynamics simulations with starting 5 kcal/mol Å2
restraints on protein backbone atoms and on the ligand heavy atoms,
which were gradually decreased during the heating. The system
was then equilibrated with respect to volume running for 70 ps in
NPT conditions (constant pressure and temperature) until all the
restraints were left off. Langevin dynamics was used to keep the
temperature constant (300 K) while the pressure (1 atm) was controlled
(9) Ernst, M.; Brauchart, D.; Boresch, S.; Sieghart, W. Comparative
Modeling of GABA-A Receptor: Limits, Insights, Future Develop-
ments. Neuroscience 2003, 119, 933–943.
(10) Barnard, E. A.; Skolnick, P.; Olsen, R. W.; Mo¨lhler, H.; Sieghart,
W.; Biggio, G.; Braestrup, C.; Bateson, A. N.; Langer, S. Z.
International Union of Pharmacology. XV. Subtypes of γ-Aminobu-
tyric AcidA Receptors: Classification on the Basis of Subunit Structure
and Receptor Function. Pharmcol. ReV. 1998, 50, 291–313.
(11) Pritchett, D. B.; Luddens, H.; Seeburg, P. H. Type I and type II
GABAA-Benzodiazepine Receptors Produced in Transfected Cells.
Science 1989, 245, 1389–1392.
(12) Luddens, H.; Pritchett, D. B.; Kohler, M.; Killisch, I.; Keinanen, L.;
Monyer, H.; Sprengel, R.; Seeburg, P. H. Cerebellar GABAA-Receptor
Selective for a Behavioural Alcohol Antagonist. Nature 1990, 346,
648–651.
(13) Mo¨lher, H.; Crestani, F.; Rudolph, U. GABAA-Receptor Subtypes: A
New Pharmacology. Curr. Opin. Pharmacol. 2001, 1, 22–25.
(14) Whiting, P. J. Receptor Subtypes in the Brain: a Paradigm for CNS
Drug Discovery. Drug DiscoVery Today 2003, 8, 445–450.

4742 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15
Anzini et al.
(15) Collins, I.; Moyes, C.; Davey, W. B.; Rowley, M.; Bromidge, F. A.;
Quirk, K.; Atack, J. R.; McKernan, R. M.; Thompson, S. A.; Wafford,
K.; Dawson, G. R.; Pike, A.; Sohal, B.; Tsou, N. N.; Ball, R. G.;
Castro, J. L. 3-Heteroaryl-2-pyridones: Benzodiazepine Site Ligands
with Functional Selectivity for R2/R3-Subtypes of Human GABAA
Receptor-Ion Channels. J. Med. Chem. 2002, 45, 1887–1900.
(16) Braestrup, C.; Nielsen, M.; Honore´, T.; Jensen, L. H.; Petersen, E. N.
Benzodiazepine Receptor Ligands with Positive and Negative Efficacy.
Neuropharmacology 1983, 22, 1451–1457.
(17) Haefely, W.; Kyburz, E.; Gerecke, M.; Mo¨hler, H. Recent Advances
in the Molecular Pharmacology of Benzodiazepine Receptors and in
the Structur-Activity Relationships of their Agonists and Antagonists.
AdV. Drug Res. 1985, 14, 165–322.
Arylpiperazine Derivatives at the 5-HT3 Receptors. J. Med. Chem. 1999,
42, 1556–1575.
(37) Braestrup, C.; Nielsen, M. Benzodiazepine Receptors. In Handbook
of Psychopharmacology; Iversen, L. L., Iversen, S. D., Snyder, S. H,
Eds.; Plenum Press: New York, 1983, Vol. 17, pp 285-384.
(38) Fryer, R. I.; Rios, R.; Zhang, P.; Gu, Z. Q.; Basile, A. S.; Skolnick,
P. Structure-Activity relationships of 2-Pyrazolo[4,3-c]quinoline-3-
ones and their N- and O-Methyl Analogues at Benzodiazepine
Receptors. Med. Chem. Res. 1993, 3, 122–130.
(39) Razgulin, A. V.; Mecozzi, S. Binding Properties of Aromatic Carbon-
Bound Fluorine. J. Med. Chem. 2006, 49, 7902–7906.
(40) Schwartz, R. D.; Suzdak, P. D.; Paul, S. M. R-Aminobutyric Acid
(GABA)- and Barbiturate-Mediated 36Cl- Uptake in Rat Brain
Synaptoneurosomes: Evidence for Rapid Desensitization of the GABA
Receptor-Coupled Chloride Ion Channel. Mol. Pharmacol. 1986, 30,
419–426.
(18) Julou, L.; Blanchard, J. C.; Dreyfus, J. F. Pharmacological and Clinical
Studies of Cyclopyrrolones: Zopiclone and Suriclone. Pharmacol.,
Biochem. BehaV. 1985, 23, 653–659.
(19) Arbilla, S.; Depoortere, H.; George, P.; Langer, S. Z. Pharmacological
Profile of the Imidazopyridine Zolpidem at Benzodiazepine Receptors
and Electrocorticogram in Rats. Naunyn-Schmiedeberg’s Arch. Phar-
macol. 1985, 330, 248–251.
(20) Malatynska, E.; Serra, M.; Ikeda, M.; Biggio, G. Yamamura, H.
Modulation of GABA-Stimulated Chloride Influx by ꢀ-Carbolines in
Rat Brain Membrane Vesicles. I. Brain Res. 1988, 443, 395–397.
(21) Gardner, C. R.; Tully, W. R.; Hedgecock, C. J. R. The Rapidly
Expanding Range of Neuronal Benzodiazepine Receptor Ligands.
Prog. Neurobiol. 1993, 40, 1–61.
(22) Kahnberg, P.; Lager, E.; Rosenberg, C.; Schougaard, J.; Camet, L.;
Sterner, O.; Østergaard Nielsen, E.; Nielsen, M.; Liljefors, T. Refine-
ment and Evaluation of a Pharmacophore Model for Flavone Deriva-
tives Binding to the Benzodiazepine Site of the GABAA Receptor.
J. Med. Chem. 2002, 45, 4188–4201.
(23) Lager, E.; Andersson, P.; Nilsson, J.; Pettersson, I.; Østergaard Nielsen,
E.; Nielsen, M.; Sterner, O.; Liljefors, T. 4-Quinolone Derivatives:
High-Affinity Ligands at the Benzodiazepine Site of Brain GABAA
Receptors. Synthesis, Pharmacology, and Pharmacophore Modelling.
J. Med. Chem. 2006, 49, 2526–2533.
(24) Jensen, M. S.; Lambert, J. D. C. The Interaction of the beta-Carboline
Derivative DMCM with Inhibitory Amino Acid Responses on Cultured
Mouse Neurones. Neurosci. Lett. 1983, 40, 175–179.
(25) Gardner, C. R. Pharmacological Profiles in vivo of Benzodiazepine
Receptor Ligands. Drug DeV. Res. 1988, 12, 1–28.
(26) Hunkeler, W.; Mo¨hler, H.; Pieri, L.; Polc, P.; Bonetti, E. P.; Cumin,
R.; Schaffner, R.; Haefely, W. Selective Antagonists of Benzodiaz-
epines. Nature 1981, 290, 514–515.
(41) Im, H. K.; Im, W. B.; Hamilton, B. J.; Carter, D. B.; VonVoigtlander,
P. F. Potentiation of GABA-Induced Chloride Currents by Various
Benzodiazepine Site Agonists with R1γ2ꢀ2γ2 and R1ꢀ2γ2 Subtypes of
Cloned GABAA Receptors. Mol. Pharmacol. 1993, 44, 866–870.
(42) tert-Butyl 5-Acetyl-4,5-dihydroimidazo[1,5-a]quinoxaline-3-carboxy-
late is reported as compound 13h in the paper by TenBrink, R. E.;
Im, W. B.; Sethy, V. H.; Tang, A. H.; Carter, D. B. Antagonist, Partial
Agonist, and Full Agonist Imidazo[1,5-a]quinoxaline Amides and
Carbamates Acting through GABAA/Benzodiazepine Receptor. J. Med.
Chem. 1994, 37, 758-768. This compound was re-synthesized
according to the published procedure and its spectral and analytical
data were consistent with those reported therein.
(43) Kloda, J. H.; Czajkowski, C. Agonist-, Antagonist-, and Benzodiaz-
epine-Induced Structural Changes in the {alpha}1 Met113-Leu132
Region of the GABAA Receptor. Mol. Pharmacol. 2007, 71, 483–
493.
(44) Campo-Soria, C.; Chang, Y.; Weiss, D. S. Mechanism of Action of
Benzodiazepines on GABAA Receptors. Br. J. Pharmacol. 2006, 148,
984–990.
(45) Jones-Davis, D. M.; Song, L.; Gallagher, M. J.; MacDonald, R. L.
Structural Determinants of Benzodiazepine Allosteric Regulation of
GABAA Receptor Currents. J. Neurosci. 2005, 25, 8056–8065.
(46) Lopez-Romero, B.; Evrard, G.; Durant, F.; Sevrin, M.; P. George
Molecular Structure and Stereoelectronic Properties of Sarmazenil a
Weak Inverse Agonist at the Omega Modulatory Sites (Benzodiazepine
Receptors): Comparison with Bretazenil and Flumazenil. Bioorg. Med.
Chem. 1998, 6, 1745–1757.
(47) Fryer, R. I.; Cook, C.; Gilman, N. W.; Walser, A. Conformational
Shift at Benzodiazepine Receptor Related to the Binding of Agonist,
Antagonists and Inverse Agonists. Life Sci. 1986, 39, 1947–1957.
(48) Shove, L. T.; Perez, J. J.; Loew, G. H. Molecular Determinants of
Recognition and Activation at Cerebellar Benzodiazepine Receptor
Site. Bioorg. Med. Chem. 1994, 2, 1029–1049.
(49) Clayton, T.; Chen, J. L.; Ernst, M.; Richter, L.; Cromer, B. A.; Morton,
C. J.; Ng, H.; Kaczorowski, C. C.; Helmstetter, F. J.; Furtmuller, R.;
Ecker, G.; Parker, M. W.; Sieghart, W.; Cook, J. M. An updated unified
pharmacophore model of the benzodiazepine binding site on γ-ami-
nobutyric acida receptors: correlation with comparative models. Curr.
Med. Chem. 2007, 14, 2755–2775.
(27) Jensen, L. H.; Petersen, E. N.; Braestrup, C.; Honore´, T.; Kehr, W.;
Stephens, D. N.; Schneider, H.; Seidelmann, D.; Schmiechen, R.
Evaluation of the beta-Carboline ZK93426 as a Benzodiazepine
Receptor Antagonist. Psychopharmacology 1984, 83, 249–256.
(28) Valin, A.; Dodd, R. H.; Liston, D. R.; Potier, P.; Rossier, J. Methyl-
ꢀ-Carboline-Induced Convulsions are Antagonized by Ro 15-1788 and
by Propyl-ꢀ-Carboline. Eur. J. Pharmacol. 1982, 85, 93–97.
(29) Cox, E. D.; Hagen, T. J.; Mckernan, R. M.; Cook, J. M. BZ1 Receptor
Subtype Specific Ligands. Synthesis and Biological Properties of BCCt,
a BZ1 Subtype Specific Antagonist. Med. Chem. Res. 1995, 5, 710–
718.
(50) Cromer, B. A.; Morton, J. C.; Parker, M. W. Anxiety over GABAA
Receptor Structure Relieved by AChBP. Trends Biochem. Sci. 2002,
27, 280–287.
(30) Fryer, R. I. In ComprehensiVe Medicinal Chemistry; Hansch, C. ,
Sammes, P. G. , Taylor, J. B., Eds; Pergamon:Oxford,1990, Vol. 3,
pp 539-566.
(51) Lam, P. C.-H.; Carlier, P. R. Experimental and Computational Studies
of Ring Inversion of 1,4-Benzodiazepin-2-ones: Implications for
Memory of Chirality Transformations. J. Org. Chem. 2005, 70, 1530–
1538.
(52) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart, W. E.;
Belew, R. K.; Olson, A. J. Automated Docking Using a Lamarckian
Genetic Algorithm and Empirical Binding Free Energy Function.
J. Comput. Chem. 1998, 19, 1639–1662.
(53) Sigel, E.; Buhr, A. The Benzodiazepines Binding Site of GABAA
Receptor. Trends Pharmacol. Sci. 1997, 18, 425–429.
(54) Mihic, S. J.; Withing, P. J.; Klein, R. L.; Wafford, K. A.; Harris, R. A.
A Single Aminoacid of the Human γ-Aminobutyric Acid Type A
Receptor γ2 Subunit Determines Benzodiazepine Efficacy. J. Biol.
Chem. 1994, 269, 32768–32773.
(31) Da Settimo, A.; Primofiore, G.; Da Settimo, F.; Marini, A. M.;
Novellino, E.; Greco, G.; Martini, C.; Giannaccini, G.; Lucacchini,
A. Synthesis, Structure-Activity Relationships and Molecular Model-
ing Studies of N-(Indol-3-ylglyoxylyl))benzylamine Derivatives Acting
at the Benzodiazepine Receptor. J. Med. Chem. 1996, 39, 5083–5091.
(32) Primofiore, G.; Da Settimo, F.; Taliani, S.; Marini, A. M.; Novellino,
E.; Greco, G.; La Vecchia, A.; Besnard, F.; Trincavelli, L.; Costa, B.;
Martini, C. Novel N-(Arylalkyl)indol-3-ylglyoxylamides Targeted as
Ligands of the Benzodiazepine Receptor: Synthesis, Biological Evalu-
ation and Molecular Modeling Analysis of the Structure-Activity
Relationships. J. Med. Chem. 2001, 44, 2286–2297.
(33) Fryer, R. I.; Zhang, P.; Rios, R. The Synthesis of Substituted
2-Aminophenyl Heterocyclic Ketones. Synth. Commun. 1993, 23, 985–
992.
(34) Fryer, R. I.; Gu, Z. Q.; Wang, C. G. Synthesis of Novel, Substituted
4H-Imidazo[1,5-a][1,4]benzodiazepine. J. Heterocycl. Chem. 1991, 28,
1661–1669.
(55) Wieland, H. A.; Luddens, H.; Seeburg, P. H. A Single Histidine in
GABAA Receptors is Essential for Benzodiazepine Agonist Binding.
J. Biol. Chem. 1992, 267, 1426–1429.
(56) Sigel, E.; Schaerer, M. T.; Buhr, A.; Baur, R. The Benzodiazepine
Binding Pocket of Recombinant R1ꢀ2γ2 γ-Aminobutyric AcidA
receptors: Relative Orientation of Ligands and Amino Acid Side
Chains. Mol. Pharmacol. 1998, 54, 1097–1105.
(57) Tan, K. R.; Baur, R.; Gonthier, A.; Goleldner, M.; Siegel, E. Two
neighboring residues of loop A of the R1 subunit point towards the
benzodiapine binding site of GABAA receptor. FEBS Lett. 2007, 581,
4718–4722.
(35) Hall, J. H.; Behr, F. E.; Reed, R. L. Cyclization of 2-Azidobenzophe-
nones to 3-Phenylanthranyls. An Example of an Intramolecular 1,3-
Dipolar Addition. J. Am. Chem. Soc. 1972, 94, 4952–4958.
(36) Cappelli, A.; Anzini, M.; Vomero, S.; Canullo, L.; Mennuni, L.;
Makovec, F.; Doucet, E.; Hamon, M.; Menziani, M. C.; De
Benedetti, P. G.; Bruni, G.; Romeo, M. R.; Donati, S. Novel Potent
and Selective 5-HT3 Receptor Ligands Provided with Different
Intrinsic Efficacy. 2. Molecular Basis of the Intrinsic Efficacy of

New Benzodiazepine-3-carboxylate as an Antianxiety Agent
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 15 4743
(58) Primofiore, G.; Da Settimo, F.; Taliani, S.; Marini, A. M.; Novellino,
E.; Greco, G.; Lavecchia, A.; Besnard, F.; Trincavelli, L.; Costa, B.;
Martini, C. Novel N-(Arylalkyl)indol-3-ylglyoxylylamides Targeted
as Ligand of the Benzodiazepine Receptor: Synthesis, Biological
Evaluation, and Molecular Modeling Analysis of the Structure-
Activity Relationships. J. Med. Chem. 2001, 44, 2286–2297.
(59) Kucken, A. M.; Teisse´re, J. A.; Seffinga-Clark, J.; Wagner, D. A.;
Czajkowski, C. Structural Requirements for Imidazobenzodiazepine
Binding to GABAA Receptors. Mol. Pharmacol. 2003, 63, 289–296.
(60) Kofron, W. G.; Baclawski, L. A Convenient Method for Estimation
of Alkyl Lithium Concentrations. J. Org. Chem. 1976, 41, 1879–1880.
(61) Walsh, D. A. The Synthesis of 2-Aminobenzophenones. Synthesis
1980, 32, 677–688.
(62) Sheldrick, G. M. SHELXS-97, Release 97-2, A Program for Automatic
Solution of Crystal Structures; Go¨ttingen University: Go¨ttingen,
Germany, 1997.
(72) Malcangio, M; Bartolini, A; Ghelardini, C; Bernardini, F; Malmberg-
Aiello, P; Franconi, F; Giotti, A. Effect of ICV Taurine on the
Impairment of Learning, Convulsions and Death Caused by Hypoxia.
Psychopharmacology 1989, 98, 316–320.
(73) Jarvik, M. E.; Kopp, R. An Improved One-Trial Passive Avoidance
Learning Situation. Psychol. Rep., 1967, 21, 221–224.
(74) Brooks, B. R.; Bruccoleri, R. E.; Olafson, D.; States, D. J.; Swami-
nathan, S.; Karplus, M. CHARMM: A Program for Macromolecular
Energy, Minimization and Dynamics Calculations. J. Comput. Chem.
1993, 4, 187–217.
(75) Gasteiger, J.; Marsili, M. Iterative Partial Equalization of Orbital
ElectronegativitysA Rapid Access to Atomic Charges. Tetrahedron
1980, 36, 3219–3228.
(76) Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid, E.;
Villa, E.; Chipot, C.; Skeel, R. D.; Kale, L.; Schulten, K. Scalable
Molecular Dynamics with NAMD. J. Comput. Chem. 2005, 26, 1781–
1802.
(63) Sheldrick, G. M. SHELXL-97, Release 97-2, A Program for Crystal
Structure Refinement;Go¨ttingen University: Go¨ttingen, Germany, 1997.
(64) Martini, C.; Lucacchini, A.; Ronca, G.; Hrelia, S.; Rossi, C. A. Isolation
of Putative Benzodiazepine Receptors from Rat Brain Membranes by
Affinity Chromatography. J. Neurochem. 1982, 38, 15–19.
(65) Martini, C.; Rigacci, T.; Lucacchini, A. [3H]muscimol Binding Site on
(77) Case, D. A.; Darden, T. A.; Cheatham, T. E.; Simmerling, C. L.; Wang,
J.; Duke, R. E.; Luo, R.; Merz, K. M.; Pearlman, D. A.; Crowley, M.;
Walker, R. C.; Zhang, W.; Wang, B.; Hayik, S.; Roitberg, A.; Seabra,
G.; Wong, K. F.; Paesani, F.; Wu, X.; Brozell, S.; Tsui, V.; Gohlke,
H.; Yang, L.; Tan, C.; Mongan, J.; Hornak, V.; Cui, G.; Beroza, P.;
Mathews, D. H.; Schafmeister, C.; Ross, W. S.; Kollman, P. A.
AMBER 9; University of California: San Francisco, 2006.
(78) Wang, J.; Cieplak, P.; P.A. Kollman. How Well Does a Restrained
Electrostatic Potential (RESP) Model Perform in Calculating Confor-
mational Energies of Organic and Biological Molecules. J. Comput.
Chem. 2000, 21, 1049–1074.
(79) 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. A Second Generation Force Field for the Simulation of Proteins,
Nucleic Acids, and Organic Molecules. J. Am. Chem. Soc. 1995, 117,
5179–5197.
Purified Benzodiazepine Receptor. J. Neurochem. 1983, 41, 1183–1185.
(66) Lowry, O. H.; Rosenbrouh, N. J.; Farr, A. L.; Randall, R. J. Protein
Measurement with the Folin Phenol Reagent. J. Biol. Chem. 1951,
193, 265–275.
(67) Bertelli, L.; Biagi, G.; Giorgi, I.; Manera, C.; Livi, O.; Scartoni, V.;
Betti, L.; Giannaccini, G.; Trincavelli, L.; Barili, P. L. 1,2,3-
Triazolo[1,5-a]quinoxalines: Synthesis and Binding to Benzodiazepine
and Adenosine Receptors. Eur. J. Med. Chem. 1998, 33, 113–122.
(68) Cheng, Y. C.; Prusoff, W. H. Relationship Between the Inhibition
Constant (Ki) and the Concentration of Inhibition which Causes 50%
Inhibition (IC50) of an Enzyme Reaction. Biochem. Pharmacol. 1973,
22, 3099–3108.
(80) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.;
Klein, L. M. Comparison of Simple Potential Functions for Simulating
Liquid Water. J. Chem. Phys. 1983, 79, 926–935.
(81) Martyna, G. J.; Tobias, D. J.; Klein, M. L. Constant Pressure Molecular
Dynamics Algorithms. J. Chem. Phys. 1994, 101, 4177–4189.
(82) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.;
Pedersen, L. G. A Smooth Particle Mesh Ewald Method. J. Chem.
Phys. 1995, 103, 8577–8593.
(83) Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. J. C. Numerical Integration
of the Cartesian Equations of Motion of a System with Constraints:
Molecular Dynamics of n-Alkanes. J. Comput. Phys. 1977, 23, 327–
341.
(69) Schwartz, R. D.; Suzdak, P. D.; Paul, S. M. γ-Aminobutyric Acid
(GABA)- and Barbiturate-Mediated 36Cl- Uptake in Rat Brain
Synaptoneurosomes: Evidence for Rapid Desensitization of the GABA
Receptor-Coupled Chloride Ion Channel. Mol. Pharmacol. 1986, 30,
419–426.
(70) Vaught, J.; Pelley, K.; Costa, L. G.; Setler, P.; Enna, S. J. A
Comparison of the Antinociceptive Responses to GABA-receptor
agonist THIP and Baclofen. Neuropharmacology 1985, 4, 211–216.
(71) Walsh, D. M.; Stratton, S. C.; Harvey, F. J.; Beresford, I. J.; Hagan,
R. M. The Anxiolytic-Like Activity of GR159897, a Non-Peptide NK2
Receptor Antagonist, in Rodent and Primate Models of Anxiety.
Psychopharmacology 1995, 122, 186–191.
(84) Humphrey, W.; Dalke, A.; Schulten, K. VMD:Visual Molecular
Dynamics. J. Mol. Graphics 1996, 14, 33–38.
JM8002944