Synthesis, pharmacology, and structure-activity relationships of novel imidazolones and pyrrolones as modulators of GABAA receptors

Article abstract of DOI:10.1021/jm0509400

New series of imidazolones and pyrrolones were synthesized. The compounds were tested regarding their anxiolytic properties due to modulation of the GABA_A receptor response. Several derivatives exhibit considerable pharmacological activity while lacking the typical side effects of benzodiazepine receptor agonists. 1-(4-chlorophenyl)-4-morpholin-1-yl-1,5- dihydro-imidazol-2-one (2) and 1-(4-chlorophenyl)-4-piperidin-1-yl-1,5-dihydro- imidazol-2-one (3) were protective in the pentylenetetrazole test in rats with oral ED₅₀ of 27.4 and 12.8 mg/kg and TD₅₀ (rotarod) of >500 and 265 mg/kg, respectively. The minimum effective dose in the Vogel conflict test was 3 mg/kg for both compounds. Common structure-activity relationship and comparative molecular field analysis models of the various series of derivatives could be established which are in accordance with a GABA_A mediated pharmacological action. The findings fit well into an established pharmacophore model. This model is refined by an additional steric restriction feature.

Full text of DOI:10.1021/jm0509400

J. Med. Chem. 2006, 49, 1855-1866
1855
Synthesis, Pharmacology, and Structure-Activity Relationships of Novel Imidazolones and
Pyrrolones as Modulators of GABAA Receptors
Christian Grunwald,*,§ Chris Rundfeldt,§ Hans-Joachim Lankau,§ Thomas Arnold,§ Norbert Ho¨fgen,§ Rita Dost,§ Ute Egerland,§
Hans-Jo¨rg Hofmann,‡ and Klaus Unverferth§
elbion AG, Meissner Strasse 191, D-01445 Radebeul, Germany, and Institute of Biochemistry, Faculty of Biosciences,
Pharmacy and Psychology, UniVersity of Leipzig, Bru¨derstrasse 34, D-04103 Leipzig, Germany
ReceiVed September 21, 2005
New series of imidazolones and pyrrolones were synthesized. The compounds were tested regarding their
anxiolytic properties due to modulation of the GABAA receptor response. Several derivatives exhibit
considerable pharmacological activity while lacking the typical side effects of benzodiazepine receptor
agonists. 1-(4-chlorophenyl)-4-morpholin-1-yl-1,5-dihydro-imidazol-2-one (2) and 1-(4-chlorophenyl)-4-
piperidin-1-yl-1,5-dihydro-imidazol-2-one (3) were protective in the pentylenetetrazole test in rats with oral
ED50 of 27.4 and 12.8 mg/kg and TD50 (rotarod) of >500 and 265 mg/kg, respectively. The minimum
effective dose in the Vogel conflict test was 3 mg/kg for both compounds. Common structure-activity
relationship and comparative molecular field analysis models of the various series of derivatives could be
established which are in accordance with a GABAA mediated pharmacological action. The findings fit well
into an established pharmacophore model. This model is refined by an additional steric restriction feature.
Introduction
nists, or inverse agonists. Agonists and partial agonists amplify
GABA-induced currents, thus causing anticonvulsive, anxiolytic,
and sedative effects. On one hand, anxiolytics acting as BzR
ligands are efficient drugs of low toxicity that are relatively
safe against overdosing,8 while, on the other hand, the use of
BzR agonists is limited because of side effects such as sedation,
myorelaxation, ethanol interaction, ataxia and amnesia, psychical
and physical dependence, tolerance, and abuse liability.10,11 The
potential for dependence and drug tolerance constitutes an
impediment to the long-term use of benzodiazepines. There is
a broad consensus that the development of partial or, as
discussed more recently, subtype-specific BzR ligands could
help in the discovery of new drugs without these disadvan-
tages.12-26
Here, we present novel methods for the synthesis of imida-
zolones and pyrrolones aimed at the detection of pharmacologi-
cally active benzodiazepine receptor agonists without the side
effects of sedation, tolerance, and abuse liability. In vitro
receptor binding and in vivo pharmacological screens were
exploited to characterize a wide variety of derivatives obtained
in this way.
Chronic anxiety and epilepsy are common and serious
diseases of the central nervous system. It is assumed that 50
million people worldwide suffer from epilepsy.1 Anxiety is even
more common with a lifetime prevalence of 5% for generalized
anxiety disorders.2 About 17% of men and 21% of women over
55 experience symptoms which need some form of treatment.3,4
Although the genesis and the appearance of these disorders
differ, both can be treated by modulators of GABAA receptors.
GABAA receptors are chloride ion channels that are controlled
by the inhibitory neurotransmitter γ-aminobutyric acid (GABA).
They influence the chloride ion conductivity of the neuronal
membrane. Binding of GABA to the receptor causes an opening
of the channel and a decrease of membrane excitability by the
influx of chloride ions into the cell. GABAA receptors are
symmetric heteropentamers forming a central chloride ion-
selective pore.5 Until now, 21 subunits (6 R, 4 â, 4 γ, 1 δ, 1 ꢀ,
1 π, 1 θ, 3 F) have been identified.6,7 The R, â, and γ subunit
classes are most abundant8 and constitute pentamers that
comprise two R, two â, and one γ subunit. Various mechanisms
are known for the activation and modulation of GABAA
receptors.9 GABA and the competitive ligands muscimol
(agonist) and bicuculline (antagonist) bind to the GABA binding
site. Picrotoxin blocks the ion channel directly. Additionally,
there are modulatory sites for barbiturates, steroids, and ben-
zodiazepines.
Chemistry
The synthesis of the imidazolones 1-19 listed in Table 1 is
depicted in Scheme 1. These compounds are easily available
by heating hydantoins with the appropriate amines and the
corresponding amine hydrochloride. When the reagents are
heated without the acid or when the reaction is performed in a
solvent, the main products are amides.
The pyrrolones 20-27 were prepared by a three-step syn-
thesis (Scheme 2). Methyl-4-chloro-3-methoxybutenoate was
treated with one equivalent of an aniline derivative and
subsequently cyclized in acetic acid to a methoxy pyrrolone.
Substitution with one equivalent of an appropriate amine, with
10 mol % of the amine hydrochloride as catalyst, yielded the
pyrrolone derivatives 20-27. This route of synthesis is more
convenient than a formerly reported procedure.27 It allows the
introduction of different substituents starting with an easily
available material.
The benzodiazepine receptor (BzR), often termed benzodi-
azepine binding site, is a modulatory site of the GABAA
receptor. Beside benzodiazepines, a wide variety of structurally
different compounds are ligands of BzR as for instance
â-carbolines, pyridoindoles, triazolophthalazines, imidazoqui-
noxalines, flavones, and pyrazoloquinolinones. Depending on
their intrinsic activity, they modulate the GABA response as
full agonists, partial agonists, antagonists, partial inverse ago-
* To whom correspondence should be addressed. Phone: (+49) 351 40
43 1552. Fax: (+49) 351 40 43 55 52. E-mail: christian.grunwald@elbion.de.
§ elbion AG.
‡ University of Leipzig.
10.1021/jm0509400 CCC: $33.50 © 2006 American Chemical Society
Published on Web 02/24/2006

1856 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6
Grunwald et al.
Table 1. Physical Properties of Imidazolones, Pyrrolones, and Related Substances
cpd
R1
R2
yield, %
mp, °C
246
264
242
305
158
245
292
172-173
271
180-182
196-197
270
275
268
242-243
217
202
195
236-238
232-234 dec
226
185
211
228
121-122
215-216
165-166
214-215
139
95-96
169
115
112
logPa
formula
1
2
3
4
5
6
7
8
Ph
N(CH2CH2)2O
N(CH2CH2)2O
1-piperidinyl
1-pyrrolidinyl
N(CH2CH2)2O
N(CH2CH2)2NCH3
N(CH3)2
N(CH2CH2)2O
N(CH2CH2)2O
N(CH2CH2)2O
N(CH2CH2)2O
N(CH2CH2)2O
N(CH2CH2)2O
N(CH2CH2)2O
N(CH2CH2)2O
1-azepanyl
69
67
65
55
67
79
32
57
72
71
82
85
81
78
75
87
91
64
57
71
67
41
69
70
68
65
62
48
62
60
55
53
50
45
89
34
27
52
27
55
72
0.6
1.5
2.3
2.0
0.5
2.2
1.6
1.0
1.9
0.8
1.3
1.4
1.7
n.d.f
1.3
n.d.
2.7
1.5
2.7
1.9
2.6
3.4
3.0
n.d.
1.5
1.4
2.2
3.0
2.3
3.4
3.0
1.5
2.6
2.3
4.3
3.1
4.1
3.7
1.7
2.6
2.6
C13H15N3O2
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
benzyl
Ph-4-Cl
Ph-4-Cl
benzyl-2-Cl
Ph-3,4-Cl2
CH2(CH3)-Ph
benzyl-2,6-Cl2
Ph-4-Br
Ph-4-I
C13H14ClN3O2
C14H16ClN3O
C13H14ClN3O
C14H17N3O2
C14H17ClN4O
C11H12ClN3O
C14H16ClN3O2
C13H13Cl2N3O2
C15H19N3O2
C14H15Cl2N3O2
C13H14BrN3O2
C13H14IN3O2
C13H14FN3O2
C13H14ClN3O2
C15H18ClN3O
C16H20ClN3O
C15H19N3O2
C16H20ClN3O
C14H15ClN2O2
C14H15ClN2O
C16H19ClN2O
C15H17ClN2O
C14H15FN2O2
C15H18N2O2
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28b
29c
30
31
32
33
34
35
36
37
38d
39
40e
41
Ph-4-F
Ph-3-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-OCH3
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-F
Ph-3-Me
Ph-4-Me
Ph-3-Cl-4-F
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-F
1-azocanyl
1-piperidinyl
N(CH3)-c-C6H11
N(CH2CH2)2O
1-pyrrolidinyl
1-acepanyl
1-piperidinyl
N(CH2CH2)2O
N(CH2CH2)2O
N(CH2CH2)2O
N(CH2CH2)2O
Ph-4-Cl
N(CH2CH2)2O
1-piperidinyl
1-pyrrolidinyl
N(CH2CH2)2O
1-piperidinyl
1-pyrrolidinyl
2-pyridinyl
C15H18N2O2
C14H14ClFN2O2
C16H9Cl2NO2
C14H13ClN2O3
C15H15ClN2O2
C14H13ClN2O2
C14H13FN2O3
C15H15FN2O2
C14H13FN2O2
C14H10ClN3
C12H12ClN3O2
C15H10ClNO2
C15H11ClN2
C15H18ClN3O4
C15H11ClN2O
C14H9ClN2O2
Ph-4-F
Ph-4-F
102-103
143
125-126
112-113
138
230-233
162
156-157
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
Ph-4-Cl
N(CH2CH2)2O
Ph
Ph
N(CH2CH2)2O
Ph
Ph
a Experimental partition coefficient 1-octanol/water. b mp29 215.5 °C. c mp30 138-139 °C. d mp34 182 °C. e mp35 161-162 °C. f n.d.: not determined.
Scheme 1. Compounds 1-19
The synthesis of compound 39 was carried out by chlorina-
tion of compound 2 using thionyl chloride and treatment of
the resulting product with methanol as described in Scheme
3.
The syntheses of the compounds 28, 29-34, 35, 36, 37, 38,
40, 41, and the reference compound 42 (CGS 9896,28 Chart 1)
were performed using literature methods.28-36
Biology and Pharmacology
Affinities of the Derivatives. The method employed for the
determination of the affinities of the various compounds to the
benzodiazepine receptor is described in detail in the Experi-
mental Section. The affinities were determined as Ki, or
percentage inhibition values, at a given concentration for 37

Imidazolone and Pyrrolone GABAA Modulators
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6 1857
Table 2. Affinities for Benzodiazepine Binding Site and in Vivo
Activity and Tolerability in Mice
Scheme 2. Compounds 20-27
Ki (rat)
[µM]
PTZ ED50
[mg/kg i.p.]
rotarod TD50
[mg/kg i.p.]
cpd
1
6.28
4.35
1.04
4.29
n.e.b
n.e.
121
>300
75.6; 175.9a
92; 147a
>300
>300
>100
>300
>300
>300
>100
>300
345
2
17.2a
40.5a
>300
>300
>300
>300
>300
>300
>300
>300
21.8
3
4
5
6
7
n.e.
8
n.a.c
n.e.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
n.a.
n.a.
3.06d
3.32d
n.e.
Scheme 3. Compound 39
43a
>300a
>500a
>300
>300
>300
>100
>300
97.0; 60.9a
>300
>100
ca. 30
>300
<100
ca. 100
>30; <100
>300
>300
>300
>300
>100
>100
>300
>300
>300
<300
>300
>30
35a
n.e.
>300
>300
>300
65.6
>300
10.9
11.6d
n.a.
0.938
n.a.
0.141
0.067
0.44
0.09
0.494
4.45d
0.344
2.05d
n.e.
>300
234
58.2
337
Chart 1. Reference Compounds
62.3
89.3
22.3
>300
8.88d
4.34d
3.26d
n.e.
120
360
>300
128
5.41d
n.e.
289
>200
>300
>300
253
4.48d
26.1d
1.12
3.29d
n.e.
>300
>100
>300
>300
1.36
n.e.
n.e.
0.0024
0.0068
>300
>300
302
42e
diazepam
0.4
2
a Results from NIH. b n.e.: not effective. Percentage inhibition P < 20%
out of 41 imidazolone and pyrrolone derivatives and for 42
(Table 2). For 13 of the 37 compounds and for 42, the
experimental determination of the Ki values was possible. The
Ki values of another 12 compounds were calculated on the basis
of the single percentage inhibition data P, as described in the
Experimental Section (Table 2).
d
at highest available concentration (2 µM). c n.a.: not available. KiEW based
on single inhibition data (see Experimental Section, Calculations). e The
known compound 42 (mp 324-327 °C) was resynthesized, and the physical
and spectral data were found to be consistent with literature values.28
Table 3. Activity and Tolerability of Selected Compounds in Rats after
Oral Administration
Pharmacological Activities. The pharmacological activities
were investigated for all compounds and 42. The rotarod test
was used to record neurotoxicological effects, and the penty-
lenetetrazole (PTZ) test was selected for identifying pharma-
cologically active compounds in vivo. PTZ-induced seizures are
commonly used as a simple seizure model, but PTZ is also well-
known to induce anxiety both in animals and humans. A
suppression of PTZ-induced seizures can thus be used as a
simple screening model to predict anxiolytic activity. The PTZ
model clearly has advantages over anxiety models, which are
experimentally time-consuming and which often lack robust-
ness.37 For selected compounds, the data obtained in mice after
i.p. administration were supported by data obtained in rats after
oral administration using both the PTZ seizure model and the
Vogel conflict test as the specific anxiety model. For both PTZ
models, we selected a PTZ dose that induced seizures in 97%
of control animals. Thus, only compounds with potent activity
were detected. The tests were carried out at elbion and/or at
the NIH (Bethesda, MD). Selected screening results and all ED50
or TD50 values determined are compiled in Tables2 and 3. (For
details of the models used and for methods, see the Experimental
Vogel conflict
PTZ
ED50
[mg/kg]
test
rotarod protective index
TD50 (PI)
[mg/kg] TD50/ED50 PTZ
MEDa
[mg/kg]
cpd
2
27.4, 12.9b, 16.2b
19.3, 12.8b
14.2b
3
998
36.4
3
3
265
13.7
4
3
>500b
473
>35.2b
17.1
12
13
14
20
21
42
27.7
100
n.t.c
n.t.
10
n.e.d
0.3
1
50b
.50b
.60b
>500
n.t.
.1.0b
.1.8b
>38.8
-
34.1b
12.9, 10.5b
>200
2.7
>500
>185.0
2.1
diazepam 6.3
13
a MED minimal effective dose. b Results from NIH. c n.t.: not tested.
d n.e.: not effective at 3 and 10 mg/kg.
Section.) ED50 values of 19 compounds which passed the initial
test were determined in mice. For further evaluation in rats,
compounds with an ED50 below 40 mg/kg in mice and a clear
separation of pharmacological effects from neurotoxicity were
selected. Eight compounds as well as the two reference
compounds diazepam (Chart 1) and 42 were also tested in rats

1858 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6
Grunwald et al.
Table 4. BzR Affinity: Evaluation of CoMFA Fields with SAMPLSa
to determine the ED50 after oral administration in the PTZ test
and to obtain initial insight into the anxiolytic activity. In
addition, compound 21 without activity in mice was selected
due to its high affinity to the benzodiazepine binding site, and
compound 4 was also included in testing in rats due to structural
similarity to 3 (see Tables 2 and 3). Among those candidates,
all except 21 were found also to be active in the PTZ model in
rats and, again, all compounds except 21 also showed at least
some activity in the Vogel conflict model predictive of anxiolytic
activity. This indicates that our screening approach using the
PTZ seizure model was successful. Interestingly, the compounds
identified were better tolerated than diazepam in both mice and
rats as assessed employing the rotarod test in rats. Among these
compounds, 2, 3, 4, and 20 were most interesting due to the
best separation of pharmacological activity and unwanted side
effects in rats after oral administration. The protective index
(PI) calculated as the ratio of effective dose in the PTZ seizure
model versus the dose inducing side effects as observed in the
rotarod test were more than 7-fold better than that for diazepam
(see Table 3). However, the effect size for 4 observed in the
Vogel conflict test was low, indicating a weak anxiolytic
potential which could not be increased at higher doses. For 20,
the duration of effect was very short even at high doses as tested
in the rotarod test. This may be related to low acid stability.
Therefore, 2 and 3 were selected for further development
because of their potent anxiolytic activity, their anticonvulsive
potential, and their excellent tolerability. Compound 2 potently
suppressed PTZ seizures in dogs and reduced the seizure
frequency in epileptic dogs.38 In models of anxiety, a strong
effect could be shown in the elevated maze, the Vogel conflict
test, and in the light and dark box.39 Further studies in monkeys
indicate that 2 has a low abuse liability.40 Compound 3 was
found to be even more potent in models of anxiety, and no
development of tolerance could be seen.41 The potent anxiolytic
activity could be reversed by using the selective antagonist
flumazenil, indicating that the effect was indeed mediated by
the benzodiazepine binding site.41 If the pharmacological results
were compared with the effect of diazepam, it becomes obvious
that the activity profile, both in models of anxiety and in models
of epilepsy, is similar, while these compounds are much better
tolerated. Mechanistic studies revealed that 2 acts as a partial
low-affinity agonist at the benzodiazepine binding site.42 This
partial agonism as well as the low affinity may be related to
the compound’s excellent tolerability. Very recently, we found
3 to be a subtype-selective benzodiazepine agonist preferring
GABAA receptors with the R3 subunit over receptors containing
the R1 subunit.43 Further characterization of this subtype
selectivity is ongoing.
field
training set
only
typeb
all data
s
e
l
se
sl
el
sel
0.572 (5)
0.715 (5)
0.526 (4)
0.687 (5)
0.559 (4)
0.696 (5)
0.672 (5)
0.689 (5)
0.744 (5)
0.532 (4)
0.745 (5)
0.614 (6)
0.716 (6)
0.727 (6)
a Cross-validated coefficients of correlation q2 and number of components
in parentheses. b CoMFA field types: s, steric; e, electrostatic; l, lipophilic
field.
obvious structural similarities between the various derivatives
point to a similar mode of binding.
Three-dimensional quantitative structure-activity relationship
(3D-QSAR) models were derived on the basis of the measured
and estimated Ki values for compounds 1-4, 12, 13, 16, 18,
20-27, 29, 30, 31, 33, and 35-38 in Table 2. From these data,
the compounds 3, 13, 21, and 29 were used as the test set. The
conformational and tautomeric structure of the various com-
pounds was determined employing quantum chemical methods.
In those cases, where a comparison with experimental data is
possible, the agreement is excellent. Generally, planar structures
were found for all compounds. The details of the structure
calculations are given in the Supporting Information. Compara-
tive molecular field analysis (CoMFA) studies were carried out
examining all combinations of steric, electrostatic, and hydro-
phobic fields. The details of the alignment and CoMFA
calculations are given in the Calculation part of the Experimental
Section. The CoMFA results are listed in Table 4. The most
reliable model was derived on the basis of the electrostatic field
(q2 ) 0.72). Less predictive models were computed using steric
(q2 ) 0.57) and lipophilic (q2 ) 0.53) fields. The combination
of the various fields did not improve q2 in comparison with the
model based solely on the electrostatic field. To visualize both
the steric and electrostatic properties, a PLS model without
cross-validation was derived on the basis of the training set (no
column filtering, 5 components, r2 ) 0.970, F ) 91.106, p )
0.000, relative contributions: 0.363 steric, 0.637 electrostatic).
The model is illustrated in Figure 1. The affinities of the test
compounds could be predicted by this model (Figure 2).
PTZ Activity. Next, some relationships between structure
and pharmacological activity will be discussed. It has been stated
above that similar substitution across the three different structural
classes yielded similar effects on affinity. This trend is also
found when one compares the substituent effects on the PTZ
activity for the imidazolone, pyrrolone, and pyrroledione
derivatives. For instance, the replacement of a morpholin-1-yl
group at R2 position by a piperidin-1-yl residue reduces the
activity by a factor of 2-5. Smaller and larger aliphatic rings
lead to inactive compounds or strongly reduce the activity as
in the case of 22. Six-membered aromatic rings generally yield
inactive compounds with exception of compound 37, which is
weakly active. Substituents at the 3-position of R1 reduce the
activity or even result in inactive compounds. Accordingly, 36
should be active. The measured inactivity of this compound
points to the need for a carbonyl group at the 2-position of the
five-membered central ring. There are some other common
patterns of active compounds with respect to this ring. Thus,
the atoms equivalent to N1 of 1 and the atoms at the 4-position
are always nitrogen and carbon atoms, respectively. Besides,
atoms 3 and 4 are always connected by a double bond. These
findings support the assumption of a similar mode of action
for all derivatives.
Structure-Activity Relationships
BzR Affinity. The comparison of substituent effects on the
binding affinity of the imidazolones, pyrrolones, and pyrrole-
diones in Table 2 shows some similarities. Thus, all compounds
with an at least low affinity have a phenyl group at the R1
position. Among these, the 4-chlorophenyl derivatives exhibit
a higher affinity than the 4-fluorophenyl structures. Substituents
at the 3-position of the phenyl ring reduce the affinity exces-
sively in comparison with substituents at the 4-position. The
substituents R2 seem to contact a receptor region with limited
steric accessibility. Morpholin-4-yl, piperidin-1-yl, and pyrro-
lidin-1-yl derivatives are of comparable affinity, whereas
substitution by azepane or other larger rings decreases the
affinity. Regarding the central heterocycles, all derivatives of
the series bear at least one hydrogen acceptor group, but they
have no other common structural elements. Nevertheless, the

Imidazolone and Pyrrolone GABAA Modulators
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6 1859
Figure 1. Stereo representation of the CoMFA model of BzR affinity. Top: steric field, bottom: electrostatic field. The color coding is as follows:
green/yellow regions, more steric bulk increases/decreases the affinity; red/blue regions, more negative charge increases/decreases affinity.
Table 5. PTZ Activity: Evaluation of CoMFA Fields with SAMPLSa
all
without
22
all
without
24
all
without
22 + 24
field
all
data
typeb
s
e
l
se
sl
el
sel
0.215 (1)
0.258 (1)
0.231 (1)
0.254 (1)
0.261 (1)
0.288 (1)
0.281 (1)
0.288 (1)
0.287 (1)
0.293 (1)
0.299 (1)
0.324 (1)
0.329 (1)
0.328 (1)
0.347 (1)
0.481 (2)
0.337 (1)
0.458 (4)
0.393 (1)
0.497 (5)
0.458 (5)
0.506 (2)
0.707 (3)
0.446 (2)
0.677 (4)
0.509 (1)
0.677 (4)
0.654 (5)
a Cross-validated coefficients of correlation q2 and number of components
in parentheses. b CoMFA field types: s, steric; e, electrostatic; l, lipophilic
field.
Figure 2. Observed Ki values of BzR affinity vs predicted data using
the CoMFA model (“+” training set, “∆” test set).
To establish 3D-QSAR models based on the activity data,
the 19 compounds with ED50 data (1-3, 12-14, 18, 20, 22-
27, 29, 30, 32, 33, and 37) were used (Table 2). Because of the
relatively small number of compounds, it was not possible to
establish a test set. As in the case of the affinity data, all
combinations of standard steric and electrostatic CoMFA fields
and the hydrophobic CLIP field were studied. Using the entire
training set, no predictive CoMFA model could be derived. A
systematic search identified compounds 22 and 24 as outliers.
Excluding 22 and 24 from the data set, models with q2 up to
0.7 were obtained. For 24, an ED50 of 337 mg/kg was found,
which corresponds to a decrease in activity by a factor of 31
compared with the 4-chlorophenyl derivative 20. In contrast to
this, there are only small differences of the BzR affinities
between these two compounds. Therefore, differences between
24 and the other compounds could be assumed with respect to
receptor activation or pharmacokinetics. The results of the
various PLS calculations are listed in Table 5. As in the case
of the affinity models, the best model was derived on the basis
of the electrostatic field. Consideration of the other fields
decreased q2. Therefore, a PLS model without cross-validation,
taking account of all data except those for 22 and 24, was
Figure 3. Observed ED50 values of PTZ activity vs predicted data
using the CoMFA model (“+” training set, “∆” excluded data).
derived (no column filtering, 4 components, r2 ) 0.930, F )
40.021, p < 0.001, only electrostatic field). As expected from
the initial SAMPLS calculations, a good correlation between
measured and predicted EC50 values of all compounds but 22
and 24 was obtained (Figure 3), which confirms a uniform
mechanism of the action of all substances.
Affinity vs Activity. A comparison between receptor affinity
and PTZ activity could provide hints of parameters beyond
pharmacodynamics that may influence activity. On the basis of

1860 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6
Grunwald et al.
a good agreement of Ki values from experiments with rat and
mice receptor preparations from preliminary studies, we tried
to correlate rat Ki and mice PTZ values. Assuming BzR
interaction of our compounds as the basic mechanism for PTZ
activity, active compounds must be affine and common features
of structure-affinity and structure-activity relationships should
be found. Actually, nearly all affine compounds are also active.
The two compounds 14 and 32, which are active in the PTZ
test but do not have a significant receptor affinity, have a
relatively low solubility in aqueous media. As expected,
similarities between the SARs can be found. Thus, derivatives
with 4-chlorophenyl substitution are both affine and active.
However, affine compounds need not be active due to poor
pharmacokinetics/ADME properties or due to antagonistic or
inverse agonistic receptor activation. Accordingly, seven com-
pounds 4, 16, 21, 31, 35, 36, 38 with affinity to the BzR do not
elicit effects in the PTZ test. With respect to the central rings
in the derivatives, it seems to be sufficient for affinity to have
one hydrogen acceptor group while active compounds show a
closer correspondence in structure and have more atom and
bonding patterns in common. The replacement of the morpholin-
1-yl moiety by pyrrolidin-1-yl or aromatic groups does not
change the affinity very much, but it leads to distinctly less
active or inactive compounds. The compounds 4, 21, and 31,
bearing a pyrrolidin-1-yl substituent, are all affine, but are not
active.
Figure 4. Alignment of diazepam (thin dark gray sticks), 42 (thick
light gray sticks), and 2 (black sticks) according to the pharmacophore/
receptor model of Cook et al.46 and alignment calculations. H, A, L,
and S are hydrogen-bond donor sites, hydrogen-bond acceptor sites,
lipophilic receptor regions, and sterically limited regions, respectively.
features of the Cook model together with the structures of
diazepam and 42.46 Being a representative example of our
compounds, 2 was added to Figure 4 using the alignment with
42 mentioned above.
The features depicted in Figure 4 are suggested attributes of
the BzR receptor which the ligands have to correspond to. In
detail, there are two hydrogen-bond donor sites H1 and H2, a
hydrogen-bond acceptor site A2, four lipophilic regions L1, L2,
L3, and LDI, and regions S of steric repulsion. L3 is located
out of the plane of the other regions. Inverse agonists interact
with L1, H1, and A2. Agonists additionally interact with H2
and L2 or L3. Occupation of the out-of-plane region L3 is a
prerequisite of full agonism.55 The DI binding site is smaller
than the diazepam-sensitive (DS) site. Ligands with substituents
capable of filling L3 do not fit into DI binding sites. The initial
model has been extended to describe the differences between
the pharmacophores of recombinant BzR Rxâ3γ2 subtypes56-59
with “x” ranging from 1 to 6.
Hints on an excellent bioavailability of some of our com-
pounds can be found by analyzing affinity/activity ratios. Thus,
compound 2 has an affinity which is weaker by 3 orders of
magnitude than that of 42. In contrast to this, the PTZ activities
differs by just 1 order of magnitude pointing to differences
regarding pharmacokinetics. We used own data and data from
literature44,45 to compare these compounds. Owing to the
heterogeneous character of these data, only qualitative or
semiquantitative conclusions can be drawn. There are consider-
able differences between the two drugs regarding their plasma
concentrations and biodegradation. Using comparable oral doses
of 30 mg/kg for 2 and 5 mg/kg for 42,44 an approximately 1000-
fold higher plasma concentration of 2 was found (39 µM for 2
and 0.027 µM for 42). Furthermore, high concentrations of 2
in the brain were detected represented by a brain/plasma ratio
(BP) of 1:1.5. In contrast to this finding, the BP of 43 (CGS
8216,28 Chart 1, dechloro-derivative of 42) is 1:20.45 The BP
of 42 is not known, but it may be assumed to be similar. On
the basis of these values, it can be estimated that 2 reaches peak
brain concentrations of about six times the Ki value at an oral
application of 30 mg/kg, whereas with a comparable oral dose
of 42 only very small brain concentrations could be reached.
Nevertheless, these concentrations also correspond to the Ki
value of this compound. Thus, similar pharmacological effects
of both compounds can be expected despite their very different
affinities.
The planar structure of our compounds is in agreement with
the generally flat geometry of BzR ligands of different structure
classes. It can be seen from the alignment results that the phenyl
ring and its 4-chloro substituent occupy the hydrophobic pockets
L1 and L2, respectively, of the BzR binding site. Our com-
pounds do not have substituents that fit into the out-of-plane
lipophilic pocket L3. Therefore, full agonism should not be
possible according to Cook’s pharmacophore model. This is
consistent with the results from electrophysiological studies on
2,42 which show a partial agonistic behavior. Furthermore, in
agreement with the pharmacophore model, our compounds with
affinity to the BzR binding site possess a hydrogen-bonding
acceptor structure that can interact with the H1 donor site of
the model. In most cases, this acceptor is a carbonyl group.
However, the imidazolones and pyrrolones do not have an
acceptor structure for the H2 donor size. This could be one
reason for their reduced affinity compared with 42, but it cannot
be the only one, because the oxadiazolone 41 has a nitrogen
atom in the corresponding position although it is not affine.
This may hint at an influence of the quinoline nitrogen of 42
on the affinity as a hydrogen bond donor. Furthermore, the
existence of an H2 counterpart is considered to be a prerequisite
of compounds with affinity to diazepam-insensitive (DI) receptor
subtypes. In agreement with this, no affinity of 2 to DI subtypes
was found up to a concentration of 20 µM, whereas 42 binds
to these subtypes with Ki ) 0.77 µM. A result of the SAR
studies is that substituents in the 3-position of R1 reduce the
affinity. This suggests steric restrictions within this region and
corresponds to the S1 and S2 features of the Cook model, which
are regions of negative steric repulsion. The reduced affinity of
Pharmacophore. The structural basis of our data can well
be referred to the comprehensive pharmacophore/receptor model
of Cook et al.,46 which is widely accepted. It is based on 136
ligands from 10 different structural families and includes
previous models for BzR agonists, antagonists, inverse agonists,
and ligands on the diazepam-insensitive (DI) site.47-54 The
assumption of a unique binding site for all types of ligands is
based on the biochemical and pharmacological competition
between all ligands, the continuum of pharmacological effects
ranging from agonists over antagonists to inverse agonists, and
on the observation that small structural modifications cause shifts
between these types. Figure 4 illustrates the most important

Imidazolone and Pyrrolone GABAA Modulators
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6 1861
was heated under reflux. After 2 h the solution was diluted with
20 mL of ethanol and allowed to cool to room temperature. The
precipitate was isolated by filtration and washed with 150 mL of
cold ethanol. The 1-aryl-2,4-dihydro-4-methoxy-pyrrol-2-ones ob-
tained by this method were used without further purification.
Synthesis of 1-Ar(alk)yl-4-amino-1,5-dihydro-imidazol-2-ones
1-19. General Procedure. A mixture of 30 mmol of 1-ar(alk)-
yl-hydantoin and 60 mmol of the appropriate amine hydrochloride
in 50 mL of the corresponding amine was stirred and heated at a
bath temperature of 100-140 °C for 2-12 h. After the mixture
had been cooled to room temperature, a solid precipitated, which
was collected by filtration, washed with water, and purified by
recrystallization from alcohol. Compounds 3, 4, and 18 were
synthesized in a high-pressure reaction vessel.
the 4-fluorophenyl derivatives may be caused by unfavorable
interactions of the fluoro group with the hydrophobic region
L2. There are differences between fluoro groups and the other
halogen substituents with respect to the hydrophobic properties
indicated by a hydrophobic constant for a fluoro substituent of
π ) -0.17, whereas the corresponding constants of the other
halogen substituents are > 0.60 Furthermore, a fluoro substituent
is a hydrogen-bond acceptor61 and interacts with water mol-
ecules. It has already been mentioned that R2 substituents with
more than six atoms reduce the affinity. This indicates the
possibility of steric restrictions in this region, which is denoted
by S4 in Figure 4. There is no corresponding feature in the
Cook model, but similar results were also found in a study on
imidazo[1,2-b]pyridazines62 and in a study on flavones.63 As
mentioned above, the occupation of the lipophilic region L2 or
L3 is a prerequisite for agonists according to Cook’s pharma-
cophore model. L2 is occupied by the chloro atom of the partial
agonist 2. However, compound 1 also shows weak anticonvul-
sive activity in the PTZ test and, therefore, must be an at least
partial agonist at the BzR binding site.
1-Phenyl-4-morpholin-4-yl-1,5-dihydro-imidazol-2-one (1). IR
(KBr) 2862, 1703, 1592, 1504;13C NMR (DMSO-d6) 41.4 (CH2-
N), 51.3 (CH2-N), 66.5 (CH2-O), 119.7, 125.1, 125.9, 129.4,
136.2 (CAr), 165.8 (CdO), 174.1 (NdC-N). Anal. C, H, N.
1-(4-Chlorophenyl)-4-morpholin-4-yl-1,5-dihydro-imidazol-
1
2-one (2). IR (KBr) 2863, 1707, 1592, 1498; H NMR (DMSO-
d6) 4.7 (s, CH2-N), 7.3, 7.7 (Ar-H); 13C NMR (DMSO-d6) 40.3
(CH2-N), 49.71 (CH2-N), 65.5 (CH2-O), 118.4, 125.5, 128.4,
138.2 (CAr), 165.4 (CdO), 173.3 (NdC-N). Anal. C, H, N.
1-(4-Chlorophenyl)-4-piperidin-1-yl-1,5-dihydro-imidazol-2-
one (3). IR (KBr) 2943, 2858, 1702, 1587;1H NMR (DMSO-d6)
2.2 (s, N-CH3), 4.7 (s, CH2-N), 7.4, 7.7 (Ar-H); 13C NMR
(DMSO-d6) 23.7, 21.8 (CH2), 45.9 (N-CH2), 48.4 (CH2-N), 116.9,
123.8, 126.9, 136.9 (CAr), 164.2 (CdO), 171.1 (NdC-N). Anal.
C, H, N.
In summary, our derivatives fit well into the pharmacophore
model of Cook but show some features not yet considered in
this model.
Conclusions
Based on novel syntheses of imidazolones and pyrrolones,
various compounds have been obtained for the first time. Among
these, we found anxiolytic compounds lacking the side effects
of full benzodiazepine receptor agonists. Qualitative structure-
activity relationships and CoMFA models were derived on the
basis of compounds from different classes. The results can be
interpreted within the pharmacophore model of Cook et al.
which could be refined. Compounds 2 and 3 were selected for
further development because of their excellent pharmacological
profiles. The partial agonistic activity, which is consistent with
the pharmacophore model, is a possible reason for the excellent
tolerability. Recently, we obtained hints on subtype selectivity
of compound 3. Future work will have to focus on the
identification of the structural elements responsible for subtype
selectivity.
1-(4-Chlorophenyl)-4-pyrrolidin-1-yl-1,5-dihydro-imidazol-2-
1
one (4). IR (KBr) 2979, 2881, 1703, 1607; H NMR (DMSO-d6)
1.92 (mCH2), 3.49 (m, CH2), 4.61 (s, NCH2), 7.41 (m, Ar-H),
7.76 (m, Ar-H); 13C NMR (DMSO-d6) 120.7, 127.6, 130.2, 140.5
(CAr), Anal. C, H, Cl.
1-Benzyl-4-morpholin-4-yl-1,5-dihydro-imidazol-2-one (5). IR
(KBr) 2971, 2844, 1695, 1592;1H NMR (DMSO-d6) 3.6 (m, CH2),
4.2 (s, CH2-N), 4.5 (s, CH2-N), 7.2, 7.4 (Ar-H); 13C NMR
(DMSO-d6) 44.8 (CH2-N), 46.1 (CH2-N), 49.7 (CH2-N), 65.7
(CH2-O), 127.5, 127.9, 129.0, 138.4 (CAr), 168.7 (CdO), 175.1
(NdC-N). Anal. C, H, N.
1-(4-Chlorophenyl)-4-(4-methyl piperazin-1-yl)-1,5-dihydro-
1
imidazol-2-one (6). IR (KBr) 2944, 2789, 1707, 1583; H NMR
(DMSO-d6) 2.2 (s, N-CH3), 4.7 (s, CH2-N), 7.4, 7.7 (Ar-H);
13C NMR (DMSO-d6) 45.9 (N-CH3), 54.1 (CH2-N), 118.9, 125.9,
128.9, 138.7 (CAr), 165.4 (CdO), 173.1 (NdC-N). Anal. C, H,
N.
Experimental Section
Synthesis of 1-(4-Chlorophenyl)-4-(dimethylamino)-1,5-dihy-
dro-imidazol-2-ones (7). Instead of dimethylamine hydrochloride,
dimethylammonium dimethyl carbamate (DIMCARB) was used.
The reaction time was 48 h. IR (KBr) 2992, 1707, 1620, 1498;13C
NMR (DMSO-d6) 39.8 (CH2-N), 58.6 (N-CH), 118.8, 119.8,
129.0, 136.0 (CAr), 166.6 (CdO), 174.7 (NdC-N). Anal. C, H,
Cl, N.
1-(2-Chlorobenzyl)-4-morpholin-4-yl-1,5-dihydro-imidazol-2-
one (8). IR (KBr) 2990, 2865, 1692, 1592;1H NMR (DMSO-d6)
3.7 (m, CH2), 4.1 (s, CH2-N), 4.5 (s, CH2-N), 7.2, 7.4 (Ar-H);
13C NMR (DMSO-d6) 44.8 (CH2-N), 46.1 (CH2-N), 49.7 (CH2-
N), 65.7 (CH2-O), 127.5, 128.9, 129.1, 129.4, 132.1, 135.2 (CAr),
168.3 (CdO), 174.7 (NdC-N). Anal. C, H, N.
1-(3,4-Dichlorophenyl)-4-morpholin-4-yl-1,5-dihydro-imida-
zol-2-one (9). IR (KBr) 3106, 2919, 1704, 1592, 1484, 1435, 1370,
805. Anal. C, H, Cl, N.
1-(1-Phenylethyl)-4-morpholin-4-yl-1,5-dihydro-imidazol-2-
one (10). IR (KBr) 3423, 2913, 1685, 1587, 1449, 1290, 1108,
699. Anal. C, H, Cl, N.
Chemistry. All melting points were determined on a Boetius
melting-point apparatus PHMK 05 and are uncorrected. The IR
spectra were registered on a Perkin-Elmer 1725x spectrometer. All
absorption values are expressed in wavenumbers (cm-1). Proton
(1H NMR) and carbon (13C NMR) nuclear magnetic resonance
spectra were recorded on a Bruker ARX 300 NMR spectrometer.
Chemical shifts (δ) are in parts per million (ppm) relative to Si-
(CH3)4, and coupling constants (J) are in hertz. The partition
coefficients were determined according to Yamagami and Takao64
and are given as log P values for 1-octanol/water.
Key Intermediates. 1-Aryl-hydantoins and 1-Aralkyl-hydan-
toins. The synthesis of the used hydantoins is described by Biltz
and Slotta.65
Methyl 4-(Arylamino)-3-methoxybut-2-enoates. A mixture of
60 mmol of a substituted aniline derivative, 60 mmol of methyl
4-chloro-3-methoxybutenoate (easily prepared from methyl 4-chlo-
roacetoactetate66), 22.8 g of sodium phosphate dodecahydrate, and
0.9 g of sodium iodide in 300 mL acetonitrile was heated under
reflux for 24 h. The mixture was allowed to cool to room
temperature, and the remaining salts were removed by filtration.
The filtrate was evaporated to yield a dark brown oil, which was
used without further purification.
1-(2,6-Dichlorobenzyl)-4-morpholin-4-yl-1,5-dihydro-imida-
zol-2-one (11). IR (KBr) 3068, 2966, 1698, 1581, 1451, 1289, 1114,
766. Anal. C, H, Cl, N.
1-(4-Bromophenyl)-4-morpholin-4-yl-1,5-dihydro-imidazol-2-
one (12). IR (KBr) 2862, 1704, 1503; 1H NMR (DMSO-d6) 4.7 (s,
CH2-N), 7.4, 7.7 (Ar-H); 13C NMR (DMSO-d6) 45.7 (CH2-N),
1-Aryl-2,4-dihydro-4-methoxy-pyrrol-2-ones. A solution of 40
g of methyl-4-aryl-3-methoxybut-2-enoate in 10 mL of acetic acid

1862 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6
Grunwald et al.
50.8 (CH2-N), 66.5 (CH2-O), 114.4, 119.8, 132.4, 139.7 (CAr),
166.4 (CdO), 174.3 (NdC-N). Anal. C, H, N.
H2), 4.61 (s, C3-H), 7.28 (m, 2Ar-H), 7.75 (m, 2Ar-H); 13C
NMR (DMSO-d6) 25.8 (CH2), 26.8 (CH2), 27.5 (CH2), 29.0 (CH2),
49.0 (-CH2-N-CH2-), 49.7 (C5), 87.1 (C3), 118.6 (Ar-C3,
Ar-C3′), 125.1 (Ar-C4), 128.7 (Ar-C2, Ar-C2′), 139.9 (Ar-
C1), 161.9 (C4), 172.2 (C2). Anal. C, H, N.
1-(4-Chlorophenyl)-4-piperidin-1-yl-2,4-dihydro-pyrrol-2-
one (23). IR (KBr) 3061, 2939, 1673, 1615, 1494; 1H NMR
(DMSO-d6) 1.68 (m, -CH2-CH2-CH2-), 3.18 (m, -CH2-N-
CH2-), 4.40 (s, C5-H2), 4.79 (s, C3-H), 7.28 (m, 2Ar-H), 7.77
(m, 2Ar-H); 13C NMR (DMSO-d6) 23.8 (-CH2-CH2-CH2-),
25.2 (-CH2-CH2-CH2-), 47.7 (-CH2-N-CH2-), 50.0 (C5),
88.5 (C3), 118.6 (Ar-C2, Ar-C2′), 125.2 (Ar-C4), 128.7 (Ar-
C3, Ar-C3′), 140.0 (Ar-C1), 162.0 (C4), 172.2 (C2). Anal. C,
H, N.
1-(4-Fluorophenyl)-4-morpholin-4-yl-2,4-dihydro-pyrrol-2-
one (24). IR (KBr) 2978, 1668, 1606, 1510; 1H NMR (DMSO-d6)
3.25 (m, -CH2-N-CH2-), 3.69 (m, -CH2-O-CH2-), 4.49 (s,
C5-H2), 4.88 (s, C3-H), 7.14 (m, 2Ar-H), 7.73 (m, 2Ar-H);
13C NMR (DMSO-d6) 46.0 (-CH2-N-CH2-), 49.3 (C5), 64.9
(-CH2-O-CH2-), 89.4 (C3), 114.4 (2d, J ) 22 Hz, Ar-C3,
Ar-C3′), 118.7(2d, J ) 7.5 Hz, Ar-C2, Ar-C2′), 136.4 (d, J )
2.3 Hz, Ar-C1), 156.9 (d, J ) 239 Hz, Ar-C4), 161.3 (C4), 170.8
(C2). Anal. C, H, N.
1-(4-Iodophenyl)-4-morpholin-4-yl-1,5-dihydro-imidazol-2-
1
one (13). IR (KBr) 2920, 1704; H NMR (DMSO-d6) 3.33-3.87
(m, NCH2CH2O), 4.65 (s, CH2), 7.32 (d, Ph-H), 7.65 (d, Ph-H);
13C NMR (DMSO-d6) 45.11 (CH2-N), 50.01 (CH2), 65.75 (CH2O),
85.34 (C-I), 119.57 (CPh), 137.67 (CPh), 139.56 (CN), 165.87
(NCN), 173.88 (CdO). Anal. C, H, Cl, N.
1-(4-Fluorophenyl)-4-morpholin-4-yl-1,5-dihydro-imidazol-2-
one (14). 13C NMR (DMSO-d6) 45.61 (N-CH2), 50.09 (CH2),
66.25 (CH2O), 118.27 (CPh), 128.47 (CPh), 138.21 (CN), 156.98
(d, J ) 234.0 Hz, C-F), 164.38 (NCN), 173.48 (CdO). Anal. C,
H, Cl, N.
1-(3-Chlorophenyl)-4-morpholin-4-yl-1,5-dihydro-imidazol-
2-one (15). IR (KBr) 1723, 1598; 13C NMR (DMSO-d6) 45.12
(NCH2), 50.26 (CH2), 65.85 (CH2O), 116.30 (CPh), 121.77 (CPh),
130.72 (CPh), 133.78 (C-Cl), 141.21 (CN), 165.87 (NCN), 173.88
(CdO). Anal. C, H, Cl, N.
1-(4-Chlorophenyl)-4-azepan-1-yl-1,5-dihydro-imidazol-2-
one (16). IR (KBr) 3063, 2932, 1699, 1582; 1H NMR (DMSO-d6)
1.6, 1.8 (m, CH2-CH2), 3.6 (m, N-CH2), 4.7 (s, CH2-N), 7.4,
7.7 (Ar-H); 13C NMR (DMSO-d6) 26.8, 28.0 (CH2-CH2), 48.9
(CH2-N), 50.4 (CH2-N), 119.0, 125.9, 128.9, 138.8 (CAr), 166.2
(CdO), 174.1 (NdC-N). Anal. C, H, N.
1-(3-Methylphenyl)-4-morpholin-1-yl-2,4-dihydro-pyrrol-2-
one (25). IR (KBr) 2966, 1673, 1613, 1594; 1H NMR (DMSO-d6)
2,22 (Ar-CH3), 3.25 (m, -CH2-N-CH2-), 3.65 (m, -CH2-
O-CH2-), 4.47 (s, C5-H2), 4.89 (s, C3-H), 6.75 (m, Ar-H),
7.18 (m, Ar-H), 7.78 (m, 2Ar-H); 13C NMR (DMSO-d6) 21.7
(Ar-CH3), 46.7 (-CH2-N-CH2-), 49.9 (C5), 65.8 (-CH2-O-
CH2-), 90.2 (C3), 114.7, 118.0, 122.6, 128.8 (4Ar-C), 138.1,
140.7 (Ar-C1, Ar-C3), 162.3 (C4), 171.9 (C2). Anal. C, H, N.
1-(4-Methylphenyl)-4-morpholin-4-yl-2,4-dihydro-pyrrol-2-
one (26). IR (KBr) 2861, 1666, 1608, 1515; 1H NMR (DMSO-d6)
2,22 (Ar-CH3), 3.23 (m, -CH2-N-CH2-), 3.66 (m, -CH2-
O-CH2-), 4.42 (s, C5-H2), 4.80 (s, C3-H), 7.09 (m, 2Ar-H),
7.57 (m, 2Ar-H); 13C NMR (DMSO-d6) 18.7 (Ar-CH3), 44.8
(-CH2-N-CH2-), 47.9 (C5), 63.8 (-CH2-O-CH2-), 88.3
(C3), 115.6 (Ar-C3, Ar-C3′), 127.4, (Ar-C2, Ar-C2′), 128.8
(Ar-C4), 136.4 (Ar-C1), 160.3 (C4), 169.8 (C2). Anal. C, H, N.
1-(3-Chloro-4-fluorophenyl)-4-(morpholin-4-yl)-1,5-dihydro-
pyrrole-2-one (27). IR (KBr) 2850, 1673, 1613; 1H NMR (DMSO-
d6) 3.19 (m, -CH2-N-CH2-), 3.62 (m, -CH2-O-CH2-), 4.39
(s, C5-H2), 4.79 (s, C3-H), 7.23 (m, Ar-H), 7.60 (m, Ar-H),
7.90 (m, Ar-H); 13C NMR (DMSO-d6) 45.94 (-CH2-N-CH2-
), 49.10 (C5), 64.90 (-CH2-O-CH2-), 88.62 (C4), 116.10 (d, J
) 22.2 Hz, Ar), 117.62 (Ar), 118.70 (d, J ) 18.0 Hz, Ar), 136.40
(d, J ) 2.3 Hz, Ph), 156.92 (d, J ) 239.1 Hz, Ar), 161.35 (s, C3),
170.78 (s, C2). Anal. C, H, N.
1-(4-Chlorophenyl)-4-azocan-1-yl-1,5-dihydro-imidazol-2-
one (17). IR (KBr) 2930, 1701; 13C NMR (DMSO-d6) 24.72 (CH2),
25.41 (CH2), 25.71 (CH2), 25.91 (CH2), 26.22 (CH2), 49.15
(NCH2), 118.44 (CPh), 127.02 (C-Cl), 128.74 (CPh), 137.56 (CN),
165.89 (NCN), 172.33 (CdO). Anal. C, H, Cl, N.
1-(4-Methoxyphenyl)-4-piperidin-1-yl-1,5-dihydro-imidazol-
2-one (18). 1H NMR (DMSO-d6) 1.53 (m, CH2), 3.46 (m, NCH2),
3.55 (s, OCH3), 4.65 (s, CH2), 6.88 (d, Ph-H), 7.59 (d, Ph-H);
13C NMR (DMSO-d6) 23.32 (CH2), 24.13 (CH2), 45.33 (NCH2),
50.33 (CH2), 55.12 (OCH2), 113.11 (CPh), 118.02 (C-Ph), 132.78
(CPh), 138.96 (CN), 154.32 (C-O), 165.09 (NCN), 172.27 (Cd
O). Anal. C, H, Cl, N.
1-(4-Chlorophenyl)-4-(1-cyclohexyl-1-methyl)-1,5-dihydro-
imidazol-2-one (19). IR (KBr) 2926, 2854, 1693, 1595;1H NMR
(DMSO-d6) 1.7 (m, CH2), 3.4 (m, CH2-N), 4.6 (s, N-CH2), 6.9,
7.4 (Ar-H); 13C NMR (DMSO-d6) 23.3 (CH2), 25.5 (CH2), 46.7
(CH2-N), 50.0 (CH2-N), 55.1 (CH3-O), 113.8, 118.6, 132.8,
154.3 (CAr), 165.7 (CdO), 172.7 (NdC-N). Anal. C, H, Cl, N.
Synthesis of 4-Substituted 1-Aryl-2,4-dihydro-pyrrol-2-ones
20-27. General Procedure. A solution of 22.4 mmol of 1-aryl-
2,4-dihydro-4-methoxy-pyrrole-2-one and 0.25 g of the correspond-
ing amine hydrochloride in 22 mL of the amine was heated to reflux
or, in the case of higher-boiling amines, to a temperature of 120-
125 °C for 7 h. The solution was allowed to cool to room
temperature, and the precipitate was separated by filtration. To
obtain the pure compound, the crude product was washed with 5
mL of cold ethanol and 50 mL of water. Recrystallization from
n-butanol afforded the analytically pure pyrrolones.
1-(4-Chlorophenyl)-4-morpholin-4-yl-2,4-dihydro-pyrrol-2-
one (20). IR (KBr) 2968, 1672, 1616, 1495; 1H NMR (DMSO-d6)
3.21 (m, -CH2-N-CH2-), 3.72 (m, -CH2-O-CH2-), 4.43 (s,
C5-H2), 4.81 (s, C3-H), 7.28 (m, 2Ar-H), 7.70 (m, 2Ar-H);
13C NMR (DMSO-d6) 46.4 (-CH2-N-CH2-), 49.8 (C5), 65.8
(-CH2-O-CH2-), 89.8 (C3), 118.8 (Ar-C2, Ar-C2′), 125.2
(Ar-C4), 128.8 (Ar-C3, Ar-C3′), 139.7 (Ar-C1), 162.5 (C4),
172.9 (C2). Anal. C, H, N.
Synthesis of 1-Aryl-3-amino-pyrrole-2,5-diones (29-34).30 The
compound 29 is known. The physical and spectral data of the
resynthesized compound are consistent with literature values. The
compounds 30-34 were synthesized in the same manner.
1-(4-Chlorophenyl)-3-piperidin-1-yl-pyrrole-2,5-dione (30). 1H
NMR (DMSO-d6) 1.61 (s, CH2), 3.70 (s, CH2-N), 5.31 (m, Ar-
H), 7.28 (m, Ar-H), 7.51 (m, Ar-H); 13C NMR (DMSO-d6) 22.87
(CH2), 25.75 (CH2), 48.11 (CH2N), 88.68 (C-Ph), 127.66, 128.08
(C-Ph); 130.32 (C-Ph); 131.01 (C-Ph), 149.40 (C-Ph); 164.77,
168.04 (CO 2,5). Anal. C, H, N.
1-(4-Chlorophenyl)-3-pyrrolidin-1-yl-pyrrole-2,5-dione (31).
1H NMR (DMSO-d6) 1.95 (m, CH2) 3.30 and 3.82 (t, CH2), 5.31
(m, Ar-H), 7.28 (m, Ar-H), 7.51 (m, Ar-H); 13C NMR (DMSO-
d6) 24.00 and 26.21 (CH2), 49.21 and 50.79 (CH2-N); 88.68 (C-
Ph), 127.63, 128.06 (C-Ph), 130.30 (C-Ph), 131.03 (C-Ph),
149.38 (C-Ph), 164.74, 168.02 (CO 2,5). Anal. C, H, N.
1-(4-Fluorophenyl)-3-morpholin-4-yl-pyrrole-2,5-dione (32).
1H NMR (DMSO-d6) 3.7 (s, CH2), 5.30 (m, Ar-CH), 7.31 (m,
Ar-H); 13C NMR (DMSO-d6) 46.69 (CH2-N), 66.09 (CH2-O),
1-(4-Chlorophenyl)-4-pyrrolidin-1-yl-2,4-dihydro-pyrrol-2-
1
one (21). IR (KBr) 2862, 1667, 1617; H NMR (DMSO-d6) 1.92
(m, -CH2-CH2-), 3.25 (m, CH2-N-CH2), 4.38 (s, C5-H2), 4.50
(s, C3-H), 7.25 (m, 2Ar-H), 7.70 (m, 2Ar-H); 13C NMR
(DMSO-d6) 24.2 (-CH2-CH2-), 47.2 (-CH2-N-CH2-), 49.2
(C5), 86.8 (C3), 118.0 (Ar-C2, Ar-C2′), 124.6 (Ar-C4), 127.7
(Ar-C3, Ar-C3′), 139.2 (Ar-C1), 159.1 (C4), 171.1 (C2). Anal.
C, H, N.
2
88.18 (C-Ph), 115.92 (C-Ph, H-, JCF ) 23.15), 128.50 (C-
4
3
1-(4-Chlorophenyl)-4-azepan-1-yl-2,4-dihydro-pyrrol-2-one (22).
Ph, JCF ) 1.69), 129.82 (C-Ph, JCF ) 8.72), 150.51 (C-Ph),
161.24 (C-Ph, 1JCF ) 243.18), 166.06 and 169.38 (CO, 2,5). Anal.
C, H, N.
1
IR (KBr) 2920, 1662, 1607, 1494; H NMR (DMSO-d6) 1.49 (m,
2-CH2), 1.68 (m, 2-CH2), 3.29 (m, -CH2-N-CH2-), 4.45 (s, C5-

Imidazolone and Pyrrolone GABAA Modulators
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6 1863
1-(4-Fluorophenyl)-3-piperidin-1-yl-pyrrole-2,5-dione (33). 1H
NMR (DMSO-d6) 1.62 (s, CH2,), 3.71 (s, CH2-N), 5.30 (m, Ar-
H), 7.31 (m, Ar-H); 13C NMR (DMSO-d6) 22.94 (CH2), 25.78
(CH2), 48.20 (CH2N), 88.18 (C-Ph), 115.92 (C-Ph, 2JCF ) 23.15),
128.50 (C-Ph, 4JCF ) 1.69), 129.82 (C-Ph, 3JCF ) 8.72), 150.51
(C-Ph), 161.24 (C-Ph, 1JCF ) 243.18), 166.06 and 169.38 (CO).
Anal. C, H, N.
after filtration it was carefully washed with 0.05 mol Tris/HCl buffer
at pH ) 7.7 to separate free and bound radioactivity. The filters
were counted in a scintillation counter (Betaplate 1205, Berthold,
Wildbad, Germany) in order to determine the specific binding of
[3H]-flunitrazepam. With the help of preliminary saturation experi-
ments, the dissociation constant, Kd, of [3H]-flunitrazepam of 1.5
nM, the maximum number of binding sites Bmax of 0.38 nM, and
the specific binding in the assay of 90% were determined. Generally,
test compounds were screened at 6-10 increasing concentrations
for the determination of IC50 values, and the Ki values were
calculated according to the Cheng-Prusoff relationship67 (eq 1)
with [L] being the concentration of the radioactive ligand and Kd
its dissociation constant. If at the highest used concentration an
inhibition of less than 50% was measured, then this percentage
inhibition P was used. Determinations of IC50 and P values,
respectively, were repeated at least four times and arithmetic means
were used.
1-(4-Fluorophenyl)-3-pyrrolidin-1-yl-pyrrole-2,5-dione (34).
1H NMR (DMSO-d6) 1.95 (m, CH2), 3.30 and 3.82 (m, CH2-N),
5.00 (s, Ar-H), 7.30 (m, Ar-H); 13C NMR (DMSO-d6) 24.02 and
26.25 (CH2), 49.25 and 50.82 (CH2), 85.69 (CPh), 115.90 (CPh,
4
3
2JCF ) 23.05); 128.96 (CPh, JCF ) 1.64), 129.26 (CPh, JCF
)
1
8.78) 148.79 (CPh), 161.16 (CPh, JCF ) 243.16), 165.37 and
160.79 (CO, 2,5). Anal. C, H, N.
4-(4-Chlorophenyl)-1-pyrid-2-yl-pyrazole (35). The 1-phenyl
derivative is known.31 Compound 35 was synthesized in the same
manner. IR (KBr) 1596, 1459; 13C NMR (DMSO-d6) 112.34 (CPh),
122.39 (CPh), 123.45 (CPh), 124.05 (CPh), 127.56 (CPh), 129.10
(CPh), 130.78 (CPh), 131.61 (CPh), 139.75 (CPh), 140.22 (CPh),
148.57 (CPh), 151.04 (CPh), Anal. C, H, Cl, N.
3-(4-Chlorophenyl)-5-morpholin-4-yl-1,2,4-oxadiazole (36).
This compound was synthesized by the route described by Weber
et al.32 Instead of N-(morpholinyl-thiocarbonyl) benzimide chloride
the 4-chlorphenyl derivative was used. IR (KBr) 1639, 1416; 13C
NMR (DMSO-d6) 45.04 (NCH2CH2O), 65.15 (NCH2CH2O), 125.93
(CPh), 122.41 (C-Cl), 128.36 (CPh), 129.24 (CPh), 166.70 (NCN),
170.77 (OCN). Anal. C, H, Cl, N.
IC50
Ki )
(1)
1 + [L]/Kd
Estimation of Ki Values. As mentioned above, in some cases
percentage inhibition data P instead of Ki values were determined.
To broaden the data basis of CoMFA modeling, IC50E values were
estimated by using these data according to eq 2. Equation 2 results
from the standard sigmoidal curve assuming a Hill coefficient of
1. The Ki values were calculated according to eq 1.
Synthesis of 2-(4-Chlorophenyl)-5-phenyl-4-isoxazolin-3-one
(37). The synthesis is described for 2,5-diphenyl-4-isoxazolin-3-
one.33 The cyclization of phenyl propionic acid N-hydroxy-4-
chloroanilide in alkaline media leads to the desired compound. 1H
NMR (DMSO-d6) 6.7 (s, dCH-), 7.7, 7.9 (Ar-H); 13C NMR
(DMSO-d6) 95.2 (dCH), 117.8, 125.3, 125.8, 129.0, 129.4, 131.9,
135.2 (CAr), 163.6 (dC-O), 167.6 (CdO). Anal. C, H, N.
1-(4-Chlorophenyl)-5,5-dimethoxy-4-morpholin-4-yl-1,5-dihy-
dro-imidazol-2-one (39). Ten grams of 1-(4-chloro-phenyl)-4-
morpholin-4-yl-1,5-dihydro-imidazol-2-one (2), 0.1 mL of pyridine,
and 100 mL of thionyl chloride were stirred and heated under reflux
for 2 h. The solution was concentrated in vacuo, and the residue
was dissolved in 50 mL of methanol. After heating for 0.5 h, the
methanol was evaporated off and the crude product was purified
by chromatography (silica gel 60, ethyl acetate). 1H NMR (DMSO-
d6) 3.15 (s, OCH3), 3.80 (m, CH2), 7.40 (d, Ph-H), 7.60 (d, Ph-
H); 13C NMR (DMSO-d6) 45.55 (NCH2), 51.42 (OCH3), 65.785
(CH2O), 110.50 (CPh), 122.48 (C-Cl), 128.44 (CPh), 128.53
(CPh), 133.85 (CN), 161.46 (NCN), 168.02 (CdO). Anal. C, H,
Cl, N.
100 - P
IC50E ) [I]‚
(2)
P
To validate this approach, all raw data (393 percentage inhibition
values) from receptor-binding studies of the 13 available compounds
with regularly measured IC50 values were used to calculate IC50E
values according to eq 2. The results were related to the experi-
mental IC50 values. The ratios of estimated to measured values were
grouped according to the height of percentage inhibition, and the
95% confidence limits of these groups were calculated. For ratios
from percentage inhibition data lower than 10%, lower and upper
confidence interval limits of 0.17 and 5.94 were found. This
indicates that ratios of estimated to experimental IC50 values of up
to 6 may be found. Using percentage inhibition values between
10% and 20%, ratios of up to 4 (0.38, 3.78) can be expected. For
data > 20%, ratios < 3 were found. This seems to be acceptable
in view of a ratio of 2 generally found due to biological variance
and other experimental effects. According to these results, IC50E
values were calculated on the basis of percentage inhibition values
if P was greater than 20%. It should be noted that such an approach
is valid only if the Hill coefficient is close to 1.
Experimental Procedure for Screening in Mice. For screening
of compounds, we usually use the MES (maximal electroshock),
the PTZ, and the rotarod test. Data described here only come from
the PTZ and rotarod tests. At elbion, the substance to be screened
or vehicle was administered to three male mice (Crl:NMRI BR,
Charles River, Sulzfeld, Germany) each per dose, pre-treatment
time, and per experimental model. At the NIH,68 up to five male
mice (Crl:CF-1, Charles River, Wilmington, MA) were used. The
doses applied i.p. were 30, 100, and 300 mg/kg. The pre-treatment
time chosen was 30 min and 4 h. Shortly before the seizure tests
were carried out, possible side effects of the screening compounds
were assessed in the rotarod test. For selected compounds the
median effective doses (ED50, TD50) in the tests were determined.
Furthermore, selected compounds were also tested in rats after oral
administration.
3-(4-Chlorophenyl)-5-phenyl-3H-1,3,4-oxadiazol-2-one (41).36
The synthesis of 3,5-diphenyl-3H-1,3,4-oxadiazol-2-one has been
described.36 The 4-chloro derivative 41 was synthesized in the same
manner.
Receptor Preparation. Male Wistar rats (180 to 200 g) were
killed by suffocation in a CO2 chamber for 2 min. Whole brains
without cerebellum were removed and dissected on ice, placed in
closed vials, and stored at -70 °C. To isolate membrane fractions
with receptors, 1 g of the brains was placed into 10 mL of 0.05 M
Tris/HCl buffer, pH 7.7, at 4 °C and homogenized for 30 s at 20 000
rpm with an Ultraturrax T25 (Jahnke & Kunkel, IKA-Labortechnik,
Staufen, Germany). The homogenate was centrifuged at 4 °C for
30 min at 48 000g (OPTIMA XL-70, Beckman, Palo Alto, CA).
The resulting pellet was resuspended and homogenized in 100 mL
of 0.05 M Tris/HCl buffer, pH ) 7.7, 4 °C, at 20 000 rpm with an
Ultraturrax and used for binding assays.
Receptor Binding Screening. During preparation of the assays,
a volume of 150 µL of the membrane preparation (1.5 mg original
tissue) was incubated with 0.5 nM [3H]-flunitrazepam for 30 min
at 4 °C. Nonspecific binding was estimated in the presence of 10
µM diazepam. Binding was terminated by filtration of the incubated
membrane preparations using Filtermat A (Pharmacia, Uppsala,
Sweden) and a Micro Cell Harvester (Skatron, Lier, Norway). The
Filtermat A had been presoaked with 1% polyethylene imine, and
PTZ Test. PTZ (pentylenetetrazole 85 mg/kg b.w. mouse) was
injected s.c. (0.2 mL/20 g mouse) into the back of the neck of the
mouse. In rats, a dose of 70 mg/kg of PTZ was administered s.c.
(0.2 mL/100 g rat). After injection, the animals were observed for
30 min. The first generalized clonic seizure with loss of righting
reflex (convulsions of fore and hind legs, lateral position) was used

1864 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6
Grunwald et al.
as the endpoint. Animals with no seizure with a duration of at least
5 s were considered to be protected.
the chlorophenyl ring of 42 with the aliphatic rings of our
compounds. To check this, an ASP calculation was carried out
without a systematic search, but with Simplex optimization using
the assumption as an initial alignment. We found a persistently high
steric similarity, with a mean Carbo´ index of 0.85 ( 0.09, but a
very low electrostatic similarity with a Carbo´ index of only 0.27
( 0.13. This indicates that there is only one well-defined alignment.
CoMFA Calculations. CoMFA steric and electrostatic fields
were calculated employing the SYBYL71 defaults (Tripos standard
field class, distance-dependent electrostatic energy calculation, no
smoothing of field values, cutoff 30 kcal/mol, smoothed cutoff
transition). A 1 Å grid spacing and an sp3 carbon probe atom with
the charge +1 was used. In addition to the steric and electrostatic
fields, hydrophobic fields were calculated using the program CLIP.73
Models were derived based on all possible combinations of these
three fields. The calculations with cross-validation were performed
employing SAMPLS. Only models with cross-validated r2 (q2) >
0.5 were considered to be predictive.
Rotarod Test. Motor impairment was identified in male mice
and rats by the rotarod procedure. Inability of an animal to maintain
its equilibrium for 1 min in at least one of three trials on the rotating
rod (6 rpm and diameter 0.9 cm for mice; 8 rpm and diameter 6
cm for rats) was used as an indication of the impairment. The
animals were trained before drug experiments.
Anxiety Model. To obtain more detailed insight into the possible
anxiolytc activity, a modification of the method of Vogel et al.69
was used. The method has been recently described in detail.41 In
the Vogel conflict test, thirsty rats have to decide whether to drink
and take the risk of being punished with a brief electric shock or
whether to stay thirsty. Anxiolytics increase the number of licks
during punishment. The test required three consecutive days in
which drinking sessions were performed once a day, always at the
same time. On the first day, rats were allowed to drink water for
15 min without being punished, to become accustomed to the
operant box. They were put back into their home cage then and
left without water for 24 h. On the second day, water was replaced
by 5.0% glucose solution, and the rats were allowed to drink for
five minutes during the training session. Then, they were again
left without water for the next 24 h in their home cage. On the last
day, the rats were offered drinking water again. The whole test
session lasted for 210 s. For the first 30 s, the rats were allowed to
drink water without punishment. In the remaining 180 s, rats
received a mild electric shock when touching the drinking tube.
The number of unpunished and punished licks were counted by a
computer program “Graphic state notation” (Coulbourn Instruments,
Allentown, USA).
Calculation of ED50, TD50, and Protective Index. In the PTZ
test an approximation of the time peak effect was achieved by
administrating the test substance using different groups of animals
(n ) 3-4) at different pre-treatment times and determining the
percentage of protection. The time at which the highest protection
was measured was used for the dose-response curve and deter-
mination of the median effective dose (ED50). The dose-response
curve was based upon at least three different active dose groups
and one vehicle-treated group (n ) 8/group). In the rotarod test,
an approximation of the time of peak effect was achieved by testing
the same animals after administration of the test substance (at least
three dose groups, n ) 8/group) at different time points and
determining the percentage of animals with disturbance of motor
coordination. The time at which the highest neurotoxic effect was
measured was used for the dose-response curve and determination
of the median effective dose (TD50). The ED50 or TD50 were
calculated by probit analysis.70 The protective index (PI) in rats
was calculated as ratio between TD50 and the ED50 in the PTZ
seizure model. Only data generated in the same lab were used.
Calculations of Molecular Structures. Initial molecular models
of all compounds were built using the molecular sketcher and
energy-minimizing tools within SYBYL.71 Subsequently, geometries
were fully optimized by means of various methods of ab initio and
semiempirical MO theory. Details can be found in the Supporting
Information.
Acknowledgment. We thank Dr. Kupferberg for his scien-
tific cooperation over many years and Dr. Stables and his group
at the NIH in Bethesda for their permanent efforts and their
support of our work with novel imidazolones and pyrrolones.
The presented investigations were funded by EFRE grants from
the EU and by grants from the Free State of Saxony (project
no. 8093).
Supporting Information Available: Results from elemental
analysis, conformation analysis of 2, tautomeric structure of 42,
and additional references. This material is available free of charge
via the Internet at http://pubs.acs.org.
References
(1) Lo¨scher, W. New visions in the pharmacology of anticonvulsion.
Eur. J. Pharmacol. 1998, 342, 1-13.
(2) Wittchen, H. U.; Hoyer, J. Generalized anxiety disorder: nature and
course. J. Clin. Psychiatry 2001, 62 (Suppl. 11), 15-19.
(3) Himmelfarb, S.; Murrell, S. A. The prevalence and correlates of
anxiety symptoms in older adults. J. Psychol. 1984, 116, 159-167.
(4) Burke, W. J.; Folks, D. G.; McNeilly, D. P. Effective use of
anxiolytics in older adults. Clin. Geriatr. Med. 1998, 14, 47-65.
(5) Sigel, E.; Buhr, A. The benzodiazepine binding site of GABAA
receptors. Trends Pharmacol. Sci. 1997, 18, 425-429.
(6) Mehta, A. K.; Ticku, M. K. An update on GABAA receptors. Brain
Res. Brain Res. ReV. 1999, 29, 196-217.
(7) Bonnert, T. P.; Mckernan, R. M.; Farrar, S.; le Bourdelles, B.;
Heavens, R. P.; Smith, D. W.; Hewson, L.; Rigby, M. R.; Siri-
nathsinghji, D. J.; Brown, N.; Wafford, K. A.; Whiting, P. J. Theta,
a novel γ-aminobutyric acid type A receptor subunit. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 9891-9896.
(8) Korpi, E. R.; Mattila, M. J.; Wisden, W.; Lu¨ddens, H. GABA(A)-
receptor subtypes: clinical efficacy and selectivity of benzodiazepine
site ligands. Ann. Med. 1997, 29, 275-282.
(9) Macdonald, R. L.; Olsen, R. W. GABAA receptor channels. Annu.
ReV. Neurosci. 1994, 17, 569-602.
(10) Albaugh, P. A.; Marshall, L.; Gregory, J.; White, G.; Hutchison, A.;
Ross, P. C.; Gallagher, D. W.; Tallman, J. F.; Crago, M.; Cassella,
J. V. Synthesis and biological evaluation of 7,8,9,10-tetrahydroimi-
dazo[1,2-c]pyrido[3,4-e]pyrimdin-5(6H)-ones as functionally selective
ligands of the benzodiazepine receptor site on the GABA(A) receptor.
J. Med. Chem. 2002, 45, 5043-5051.
(11) Basile, A. S.; Lippa, A. S.; Skolnick, P. Anxioselective anxiolytics:
can less be more? Eur. J. Pharmacol. 2004, 500, 441-451.
(12) Haefely, W.; Martin, J. R.; Schoch, P. Novel anxiolytics that act as
partial agonists at benzodiazepine receptors. Trends Pharmacol. Sci.
1990, 11, 452-456.
(13) Costa, E.; Guidotti, A. Benzodiazepines on trial: a research strategy
for their rehabilitation. Trends Pharmacol. Sci. 1996, 17, 192-200.
(14) Richards, J. G.; Martin, J. R. Binding profiles and physical
dependence liabilities of selected benzodiazepine receptor ligands.
Brain Res. Bull. 1998, 45, 381-387.
(15) Medina, J. H.; Viola, H.; Wolfman, C.; Marder, M.; Wasowski, C.;
Calvo, D.; Paladini, A. C. Overview - Flavonoids: A new family
of benzodiazepine receptor ligands. Neurochem. Res. 1997, 22, 419-
425.
Alignment of Structures. We used the program ASP within
TSAR72 to align our compounds with 42 as the target structure. A
systematic search was carried out with an increment of 15°. The
Carbo´ molecular similarity index was used as the optimization
criterion on the basis of three equally weighted Gaussian ap-
proximations of electrostatic potential and molecular shape. Carbo´
indices range between 0 and 1 (steric fields) and -1 and 1
(electrostatic fields). The results were optimized by a Simplex
procedure without simulated annealing using default parameters.
We found an alignment of the halogen-substituted phenyl groups,
the central rings, and the (mostly) aliphatic moieties, respectively.
There was not only a high steric (0.91 ( 0.08), but also a good
electrostatic similarity (0.74 ( 0.07). The similarities of substances
with measurable affinities and pharmacological activities are still
slightly higher with values of 0.94 ( 0.03 (steric) and 0.76 ( 0.05
(electrostatic). An alternative alignment would result from matching

Imidazolone and Pyrrolone GABAA Modulators
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6 1865
(16) Yasumatsu, H.; Morimoto, Y.; Yamamoto, Y.; Takehara, S.; Fukuda,
T.; Nakao, T.; Setoguchi, M. The pharmacological properties of
Y-23684, a benzodiazepine receptor partial agonist. Br. J. Pharmacol.
1994, 111, 1170-1178.
(17) Schoch, P.; Moreau, J. L.; Martin, J. R.; Haefely, W. E. Aspects of
benzodiazepine receptor structure and function with relevance to drug
tolerance and dependence. Biochem. Soc. Symp. 1993, 59, 121-134.
(18) Tang, A. H.; Franklin, S. R.; Carter, D. B.; Sethy, V. H.; Needham,
L. M.; Jacobsen, E. J.; Von Voigtlander, P. F. Anxiolytic like effects
of PNU-101017, a partial agonist at the benzodiazepine receptor.
Psychopharmacology (Berl.) 1997, 131, 255-263.
(38) Lo¨scher, W.; Potschka, H.; Rieck, S.; Tipold, A.; Rundfeldt, C.
Anticonvulsant efficacy of the low-affinity partial benzodiazepine
receptor agonist ELB 138 in a dog seizure model and in epileptic
dogs with spontaneously recurrent seizures. Epilepsia 2004, 45,
1228-1239.
(39) Rostock, A.; Tober, C.; Dost, R.; Bartsch, R. AWD 131-138 is a
potential novel anxiolytic without sedation and amnesia: A com-
parison with diazepam and buspirone. Naunyn-Schmiedeberg’s Arch.
Pharmacol. 1998, 358, P35.
(40) Yasar, S.; Bergman, J.; Munzar, P.; Redhi, G.; Tober, C.; Knebel,
N.; Zschiesche, M.; Paronis, C. Evaluation of the novel antiepileptic
drug, AWD 131-138, for benzodiazepine-like discriminative stimulus
and reinforcing effects in squirrel monkeys. Eur. J. Pharmacol. 2003,
465, 257-265.
(19) Rowlett, J. K.; Cook, J. M.; Duke, A. N.; Platt, D. M. Selective
antagonism of GABAA receptor subtypes: an in vivo approach to
exploring the therapeutic and side effects of benzodiazepine-type
drugs. CNS Spectr. 2005, 10, 40-48.
(41) Langen, B.; Egerland, U.; Berno¨ster, K.; Dost, R.; Unverferth, K.;
Rundfeldt, C. Characterization in rats of the anxiolytic potential
of ELB139 [1-(4-chlorophenyl)-4-piperidin-1-yl-1,5-dihydro-imid-
azol-2-on], a new agonist at the benzodiazepine binding site of
the GABAA receptor. J. Pharmacol. Exp. Ther. 2005, 314, 717-
724.
(42) Sigel, E.; Baur, R.; Netzer, R.; Rundfeldt, C. The antiepileptic drug
AWD 131-138 stimulates different recombinant isoforms of the rat
GABAA receptor through the benzodiazepine binding site. Neurosci.
Lett. 1998, 245, 85-88.
(43) Rabe, H.; Rundfeldt, C.; Kronbach, C.; Lu¨ddens, H. GABAA receptor
subtype selectivity of the benzodiazepine receptor ligand ELB139.
Abstract No. 678.2. Society for Neuroscience: Washington, DC,
2005.
(44) Kochak, G. M.; Rakhit, A.; Honc, F.; Mahoney, C. J. Pharmacokinetic
disposition of the novel atypical anxiolytic CGS 9896 in the
cynomolgus monkey. Psychopharmacology (Berl.) 1988, 94, 332-
335.
(45) Lister, R. G.; Greenblatt, D. J.; File, S. E. A pharmacokinetic study
of CGS-8216, a benzodiazepine receptor ligand, in the rat. Psychop-
harmacology (Berl.) 1984, 84, 420-422.
(46) Zhang, W.; Koehler, K. F.; Zhang, P.; Cook, J. M. Development of
a comprehensive pharmacophore model for the benzodiazepine
receptor. Drug Des. DiscoV. 1995, 12, 193-248.
(47) Diaz-Arauzo, H.; Koehler, K. F.; Hagen, T.; Cook, J. M. Synthetic
and computer assisted analysis of the pharmacophore for agonists at
benzodiazepine receptors. Life Sci. 1991, 49, 216.
(48) Zhang, W.; Koehler, K. F.; Harris, B.; Skolnick, P.; Cook, J. M.
Synthesis of benzo-fused benzodiazepines employed as probes of
the agonist pharmacophore of benzodiazepine receptors. J. Med.
Chem. 1994, 37, 745-757.
(49) Martin, M. J.; Trudell, M. L.; Diaz-Arauzo, H.; Allen, M. S.;
LaLoggia, A. J.; Deng, L.; Schultz, C. A.; Tan, Y. C.; Bi, Y.;
Narayanan, K.; et al. Molecular yardsticks. Rigid probes to define
the spatial dimensions of the benzodiazepine receptor binding site.
J. Med. Chem. 1992, 35, 4105-4117.
(50) Allen, M. S.; Skolnick, P.; Cook, J. M. Synthesis of novel 2-phenyl-
2H-pyrazolo[4,3-c]isoquinolin-3-ols: topological comparisons with
analogues of 2-phenyl-2,5-dihydropyrazolo[4,3-c]quinolin-3(3H)-ones
at benzodiazepine receptors. J. Med. Chem. 1992, 35, 368-374.
(51) Zhang, P.; Zhang, W.; Liu, R.; Harris, B.; Skolnick, P.; Cook, J. M.
Synthesis of novel imidazobenzodiazepines as probes of the phar-
macophore for “diazepam-insensitive” GABAA receptors. J. Med.
Chem. 1995, 38, 1679-1688.
(52) Ananthan, S.; Clayton, S. D.; Ealick, S. E.; Wong, G.; Evoniuk, G.
E.; Skolnick, P. Synthesis and structure-activity relationships of 3,5-
disubstituted 4,5-dihydro-6H-imidazo[1,5-a][1,4]benzodiazepin-6-
ones at diazepam-sensitive and diazepam-insensitive benzodiazepine
receptors. J. Med. Chem. 1993, 36, 479-490.
(53) Gu, Z. Q.; Wong, G.; Dominguez, C.; de Costa, B. R.; Rice, K. C.;
Skolnick, P. Synthesis and evaluation of imidazo[1,5-a][1,4]-
benzodiazepine esters with high affinities and selectivities at “diaz-
epam-insensitive” benzodiazepine receptors. J. Med. Chem. 1993,
36, 1001-1006.
(20) Dawson, G. R.; Collinson, N.; Atack, J. R. Development of subtype
selective GABAA modulators. CNS Spectr. 2005, 10, 21-27.
(21) Rudolph, U.; Crestani, F.; Benke, D.; Bru¨nig, I.; Benson, J. A.;
Fritschy, J. M.; Martin, J. R.; Bluethmann, H.; Mo¨hler, H. Benzo-
diazepine actions mediated by specific γ-aminobutyric acidA receptor
subtypes. Nature (London) 1999, 401, 796-800.
(22) 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 GABA(A)
receptor-ion channels. J. Med. Chem. 2002, 45, 1887-1900.
(23) Atack, J. R. Anxioselective compounds acting at the GABA(A)
receptor benzodiazepine binding site. Curr. Drug Targets: CNS
Neurol. Disord. 2003, 2, 213-232.
(24) Mitchinson, A.; Atack, J. R.; Blurton, P.; Carling, R. W.; Castro, J.
L.; Curley, K. S.; Russell, M. G.; Marshall, G.; Mckernan, R. M.;
Moore, K. W.; Narquizian, R.; Smith, A.; Street, L. J.; Thompson,
S. A.; Wafford, K. 2,5-Dihydropyrazolo[4,3-c]pyridin-3-ones: func-
tionally selective benzodiazepine binding site ligands on the GABAA
receptor. Bioorg. Med. Chem. Lett. 2004, 14, 3441-3444.
(25) Carling, R. W.; Moore, K. W.; Street, L. J.; Wild, D.; Isted, C.;
Leeson, P. D.; Thomas, S.; O’Connor, D.; Mckernan, R. M.; Quirk,
K.; Cook, S. M.; Atack, J. R.; Wafford, K. A.; Thompson, S. A.;
Dawson, G. R.; Ferris, P.; Castro, J. L. 3-phenyl-6-(2-pyridyl)-
methyloxy-1,2,4-triazolo[3,4-a]phthalazines and analogues: high-
affinity γ-aminobutyric acid-A benzodiazepine receptor ligands with
R2, R3, and R5-subtype binding selectivity over R1. J. Med. Chem.
2004, 47, 1807-1822.
(26) Crawforth, J.; Atack, J. R.; Cook, S. M.; Gibson, K. R.; Nadin, A.;
Owens, A. P.; Pike, A.; Rowley, M.; Smith, A. J.; Sohal, B.; Sternfeld,
F.; Wafford, K.; Street, L. J. Tricyclic pyridones as functionally
selective human GABAA R2/3 receptor-ion channel ligands. Bioorg.
Med. Chem. Lett. 2004, 14, 1679-1682.
(27) Moulin, F.; Boosen, K.-J. Procedure for the synthesis of 1-substituted
4-aminopyrrolin-3-ones. DE 22 14 488, 1972.
(28) Yokoyama, N.; Ritter, B.; Neubert, A. D. 2-Arylpyrazolo[4,3-c]-
quinolin-3-ones: novel agonist, partial agonist, and antagonist of
benzodiazepines. J. Med. Chem. 1982, 25, 337-339.
(29) Yuki, H.; Yamamoto, T.; Tohira, Y.; Aoki, B.; Kano, T. Studies on
antiviral agents. II. Synthesis and biological activity of maleimide
derivatives. Chem. Pharm. Bull. (Tokyo) 1967, 15, 1101-1106.
(30) Augustin, M.; Ko¨hler, M. Synthesis and spectra of substitution
products of 2-chloro-maleinimides. J. Prakt. Chem. 1974, 316, 113-
138.
(31) Kano, S.; Yuasa, Y.; Shibuya, S.; Hibino, S. A new and facile
synthesis of 5-arylpyrimidines and 4-arylpyrazoles. Heterocycles
1982, 19, 1079-1082.
(32) Weber, G.; Hartung, J.; Beyer, L. Reactions of N-(morpholino-
thiocarbonyl)-benzimidchloride with dinucleophiles. Z. Chem. 1986,
26, 70-71.
(33) Nakamura, N.; Kishida, Y.; Ishida, N. Studies of acetylenic com-
pounds. LVI. Syntheses of 4-isoxazolin-3-ones and 1-phenyl-3-
hydroxypyrazoles. Chem. Pharm. Bull. 1971, 19, 1389-1394.
(34) Ali, A.; Saksena, R. K. Synthesis of some 1-substituted-4-phenyl-
2-(3H)-imidazolethiones and their desulphurised products. J. Indian
Chem. Soc. 1981, 58, 624-625.
(35) Gagnon, P. E.; Boivin, J. L.; Laflamme, Y. Syntheses and absorption
spectra of 1-chlorophenyl-3-phenyl-4-alkyl-5-pyrazolones and pyra-
zolones-4-C.14 Can. J. Chem. 1956, 34, 530-540.
(36) Barnish, I. T.; Fung, S. C.; Gibson, M. S.; Khan, S. R.; Pawalchak,
G. A.; Tse, K. M. Benzohydrazides, benzothiohydrazides, and
benzamidrazones as sources of 1,3,4(2H)-oxadiazolenones, 1,3,4(2H)-
thiadiazolenones, and 1,2,4(5H)-triazolenones. J. Heterocycl. Chem.
1986, 23, 417-419.
(37) File, S. E. What can be learned from the effects of benzodiazepines
on exploratory behavior? Neurosci. BiobehaV. ReV. 1985, 9, 45-54.
(54) Wong, G.; Koehler, K. F.; Skolnick, P.; Gu, Z. Q.; Ananthan, S.;
Schonholzer, P.; Hunkeler, W.; Zhang, W.; Cook, J. M. Synthetic
and computer-assisted analysis of the structural requirements for
selective, high-affinity ligand binding to diazepam-insensitive ben-
zodiazepine receptors. J. Med. Chem. 1993, 36, 1820-1830.
(55) Tebib, S.; Bourguignon, J. J.; Wermuth, C. G. The active analog
approach applied to the pharmacophore identification of benzodiaz-
epine receptor ligands. J. Comput.-Aided Mol. Des. 1987, 1, 153-
170.

1866 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 6
Grunwald et al.
(56) Huang, Q.; Cox, E. D.; Gan, T.; Ma, C.; Bennett, D. W.; Mckernan,
R. M.; Cook, J. M. Studies of molecular pharmacophore/receptor
models for GABAA/benzodiazepine receptor subtypes: binding
affinities of substituted â-carbolines at recombinant R x â 3 γ 2
subtypes and quantitative structure-activity relationship studies via
a comparative molecular field analysis. Drug Des. DiscoV. 1999, 16,
55-76.
(57) Harris, D. L.; DeLorey, T. M.; He, X.; Cook, J. M.; Loew, G. H.
Determinants of recognition of ligands binding to benzodiazepine
receptor/GABA(A) receptors initiating sedation. Eur. J. Pharmacol.
2000, 401, 271-287.
(58) Huang, Q.; He, X.; Ma, C.; Liu, R.; Yu, S.; Dayer, C. A.; Wenger,
G. R.; McKernan, R.; Cook, J. M. Pharmacophore/receptor models
for GABA(A)/BzR subtypes (R1â3γ2, R5â3γ2, and R6â3γ2) via a
comprehensive ligand-mapping approach. J. Med. Chem. 2000, 43,
71-95.
derivatives binding to the benzodiazepine site of the GABA(A)
receptor. J. Med. Chem. 2002, 45, 4188-4201.
(64) Yamagami, C.; Takao, N. Hydrophobicity parameters determined by
reversal-phase liquid chromatography. V. Relationship between the
capacity factor and the octanol-water partition coefficient for simple
heteroaromatic compounds and their ester derivatives. Chem. Pharm.
Bull. 1992, 40, 925-929.
(65) Biltz, H.; Slotta, K. On the synthesis of hydantoines. J. Prakt. Chem.
1926, 113, 233-267.
(66) Duc, L.; McGarrity, J. F.; Meul, T.; Warm, A. Methyl (E)-4-Chloro-
3-methoxy-2-butenoate: An extremely versatile four carbon building
block. Synthesis 1992, 1992, 391-394.
(67) Cheng, Y.-C.; Prusoff, W. H. Relationship between the inhibition
constant (KI) and the concentration of inhibitor which causes 50
percent inhibition (I50) of an enzymatic reaction. Biochem. Phar-
macol. 1973, 22, 3099-3108.
(59) He, X.; Huang, Q.; Ma, C.; Yu, S.; McKernan, R.; Cook, J. M.
Pharmacophore/receptor models for GABA(A)/BzR R2â3γ2, R3â3γ2
and R4â3γ2 recombinant subtypes. Included volume analysis and
comparison to R1â3γ2, R5â3γ2, and R6â3γ2 subtypes. Drug Des.
DiscoV. 2000, 17, 131-171.
(68) Krall, R. L.; Penry, J. K.; White, B. G.; Kupferberg, H. J.; Swinyard,
E. A. Antiepileptic drug development: II. Anticonvulsant drug
screening. Epilepsia 1978, 19, 409-428.
(69) Vogel, J. R.; Beer, B.; Clody, D. E. A simple reliable conflict
procedure for testing anti-anxiety agents. Psychopharmacology (Berl.)
1971, 21, 1-7.
(60) Hansch, C.; Leo, A. Substituent constants for correlation analysis
in chemistry and biology; Wiley: New York, 1979.
(70) Litchfield, J. T.; Wilcoxon, F. A simplified method of evaluating
dose-effect experiments. J. Pharmacol. Exp. Ther. 1949, 96, 99-
113.
(61) Goodford, P. J. GRID version 14, 1995. West Way House, Elms
Parade, Oxford OX2 9LL, U.K., Molecular Discovery Ltd.
(62) Harrison, P. W.; Barlin, G. B.; Davies, L. P.; Ireland, S. J.; Matyus,
P.; Wong, M. G. Syntheses, pharmacological evaluation and molec-
ular modelling of substituted 6-alkoxyimidazo[1,2-b]pyridazines as
new ligands for the benzodiazepine receptor. Eur. J. Med. Chem.
1996, 31, 651-662.
(63) Kahnberg, P.; Lager, E.; Rosenberg, C.; Schougaard, J.; Camet, L.;
Sterner, O.; Ostergaard Nielsen, E.; Nielsen, M.; Liljefors, T.
Refinement and evaluation of a pharmacophore model for flavone
(71) SYBYL version 6.5, 1998. 1699 S. Hanley Road, St. Louis, MO
63144-2913, USA, Tripos Inc.
(72) TSAR version 3.1, 1997. The Medawar Centre, Oxford Science Park,
Oxford OX4 4GA, UK, Oxford Molecular Group.
(73) Gaillard, P.; Carrupt, P. A. CLIP, 1997. BEP-Dorigny, 1015
Lausanne, CH, University of Lausanne.
JM0509400