Developing dual functional allosteric modulators of GABAA receptors

Article abstract of DOI:10.1016/j.bmc.2010.09.058

Positive modulators at benzodiazepine sites of α2- and α3-containing GABA_A receptors are believed to be anxiolytic. Negative allosteric modulators of α5-containing GABA_A receptors enhance cognition. By oocyte two-electrode voltage clamp and subsequent structure-activity relationship studies, we discovered cinnoline and quinoline derivatives that were both positive modulators at α2-/α3-containing GABA_A receptors and negative modulators at α5-containing GABA_A receptors. In addition, these compounds showed no functional activity at α1-containing GABA_A receptors. Such dual functional modulators of GABA_A receptors might be useful for treating comorbidity of anxiety and cognitive impairments in neurological and psychiatric illnesses.

Full text of DOI:10.1016/j.bmc.2010.09.058

Bioorganic & Medicinal Chemistry 18 (2010) 8374–8382
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Developing dual functional allosteric modulators of GABA receptors
A
Xiaodong F. Liu a, , Hui-Fang Chang , Richard Jon Schmiesing , Steven S. Wesolowski ,
⇑
b
b
b
a
a
b
Katharine S. Knappenberger , Jeffrey L. Arriza , Marc J. Chapdelaine
Department of Neuroscience Biology, AstraZeneca Pharmaceuticals, Wilmington, DE 19850, USA
a
b
Department of Chemistry, AstraZeneca Pharmaceuticals, Wilmington, DE 19850, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 13 July 2010
Revised 16 September 2010
Accepted 22 September 2010
Available online 29 September 2010
Positive modulators at benzodiazepine sites of
anxiolytic. Negative allosteric modulators of
a
a
2- and
a
3-containing GABAA receptors are believed to be
5-containing GABA receptors enhance cognition. By
A
oocyte two-electrode voltage clamp and subsequent structure–activity relationship studies, we discov-
ered cinnoline and quinoline derivatives that were both positive modulators at 2-/ 3-containing GABA
receptors and negative modulators at 5-containing GABA receptors. In addition, these compounds
showed no functional activity at 1-containing GABA
GABA
neurological and psychiatric illnesses.
a
a
A
a
A
a
A
receptors. Such dual functional modulators of
Keywords:
A
receptors might be useful for treating comorbidity of anxiety and cognitive impairments in
GABAA receptor
Benzodiazepine
Allosteric modulator
Subtype selectivity
Inverse agonist
Cinnoline
Ó 2010 Elsevier Ltd. All rights reserved.
Quinoline
Anxiety
Cognition
TEVC
1
. Introduction
significant adverse effects such as sedation, amnesia, ataxia, and
abuse liability that limit their clinical utility. Neurochemical, ge-
netic and pharmacological discoveries indicate that these effects
7
,8
A
GABA receptors are pentameric ion channels gated by
c
-aminobutyric acid (GABA). There are numerous modulator bind-
are dissociable according to the
mer.
a
subunit subtype in the penta-
The four major benzodiazepine-sensitive GABA receptor
subtypes are defined by incorporation of 1, 2, 3, or 5 subunits.
The anxiolytic effect of diazepam is mainly mediated by the 2- and
3-containing GABA receptors 2-/ 3-GABA Rs), while the
sedation and ataxia effects are primarily mediated by 1-GABA Rs
and the amnesic effect by 1/ 5-GABA Rs (reviewed in Refs. 13
and 14). Compounds made subtype-selective for 2-/ 3-GABA Rs,
such as TPA023, have been successfully demonstrated anxiolytic
9
–12
ing sites on the receptor besides GABA binding sites. Benzodiaze-
pines bind GABA receptors allosterically at the interface of
and subunits exerting diverse effects in the central nervous sys-
tem (CNS). Enhancement of GABA
allosteric modulators (PAMs) at the benzodiazepine site, such as
diazepam (Fig. 1), has been a successful therapy to treat anxiety
disorders. A few structurally distinct nonbenzodiazepines, such
as zolpidem, CGS-9896, TPA023, and
to the benzodiazepine site on GABA receptors and modulate var-
A
A
a
a
a
a
a
c
a
1
A
channel function by positive
a
A
(a
a
A
a
A
a
a
A
2
a
a
A
3
4
5
6
a5IA (Fig. 1), can also bind
5
,15
A
efficacy without sedation in rodents and primates.
Non-selective
ious functions of the channel. Although diazepam and other benzo-
diazepine modulators are often highly efficacious, they also exhibit
negative allosteric modulators (NAMs) of the benzodiazepine site,
such as methyl-6,7-dimethoxy-4-ethyl-b-carboline-3-carboxylate
(
DMCM) (Fig. 1), reportedly have anxiogenic and convulsant liabili-
16
ties as well as some cognitive enhancement properties. NAMs
selective for 5-GABA Rs (traditionally called 5 inverse agonists)
such as 5IA have been shown to be effective in cognitive enhance-
ment in animal models.
Abbreviations: AD, Alzheimer’s disease;
x A x A
a -GABA R, a -containing GABA
a
A
a
receptor; CNS, central nervous system; DIEA, N,N-diisopropylethylamine; DMCM,
methyl-6,7-dimethoxy-4-ethyl-b-carboline-3-carboxylate; DME, dimethoxyeth-
a
6
,17
ane; DMF, dimethylformamide; GABA,
c
-aminobutyric acid; HRMS, high resolution
Recently, a5IA has been reported to
mass spectrometry; NAM, negative allosteric modulator; PAM, positive allosteric
modulator; SAR, Structure–activity relationship; TEVC, two-electrode voltage
clamp.
improve alcohol-induced cognitive impairment without anxiogene-
sis and convulsion in man. Here we report two new series of
nonbenzodiazepine compounds, which are benzodiazepine-site
A A
PAMs at a2-/a3-GABA Rs and NAMs at a5-GABA Rs with little
18
⇑
Corresponding author. Tel./fax: +1 302 998 5126.
E-mail addresses: frankliu8@live.com, frank.liu@astrazeneca.com (X.F. Liu).
0
968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.09.058

X. F. Liu et al. / Bioorg. Med. Chem. 18 (2010) 8374–8382
8375
Cl
Cl
N
N
N
N
N
O
O
N
N
O
N
Diazepam
Zolpidem
CGS-9896
O
F
N
N
N
N
N
O
N
N
N
O
N
O
O
O
O
N
N
N
N
N
N
N
N
TPA023
α5IA
DMCM
Figure 1. Structures of diazepam and various nonbenzodiazepine modulators at the benzodiazepine site of GABA
A
receptors.
functional activity at
useful in treating cognitive dysfunctions in Alzheimer’s disease
AD), with possible added benefit of anxiolysis.
a
1-GABA
A
Rs. Such molecules are potentially
chloride. Azetidine was directly added to the reaction mixture in
presence of triethylamine to produce azetidin-1-yl(5-bromopyri-
din-3-yl)methanone (2b) in 71% yield. Finally, displacement of the
bromide by a trimethylstannyl group resulted in 2a in 83% yield.
(
2
2
. Results and discussion
2
.2. Compound screening
.1. Chemistry
In recent years, we have focused on developing PAMs of a2- and
The chemical synthesis in this report was focused on the forma-
A
a3-GABA Rs for anxiety treatment with minimum functional
tion of biphenyl linkages between heterocyclic groups and the
cinnoline or quinoline cores, which was the final step in the syn-
thesis of all compounds 1–13. Two types of coupling reactions
were used for the C–C bond formation. The synthesis of compound
activity at
a
1- and 5-GABA Rs to reduce side effects such as seda-
a
A
tion and amnesia. In this study, a new assay strategy using two-
electrode voltage clamp (TEVC) was developed to quickly identify
a
5-GABA
quinolines that were mainly
neutral at 1-GABA Rs. During the NAM screening, compound 1
A
R NAMs in our compound collection of cinnolines and
3
was achieved through the Suzuki coupling in 77% yield as out-
a
2-/ 3-GABA R PAMs and also
a
A
lined in Scheme 1. The Stille coupling was used for the rest of
the 12 compounds as described in the Section 4.1.3 and outlined
in Scheme 2. Most of the Stille coupling reactions in this study gave
yields of 30–60%.
a
A
(9-amino-2-cyclobutyl-5-(pyrimidin-2-yl)-2,3-dihydro-1H-pyrrol-
o[3,4-b]quinolin-1-one) (Scheme 2 and Table 1) and compound 2
(4-amino-8-(5-(azetidine-1-carbonyl)pyridin-3-yl)-N-propylcinn-
oline-3-carboxamide) (Table 1) were identified as NAMs of
The8-bromocinnolineand 5-bromoquinolineprecursors usedin
the coupling reactions were either prepared as described previ-
A
a5-GABA Rs. Since they displaced the non-selective benzodiaze-
1
9,20
ously
or as noted in the Section 4.1.2. For the synthesis of
pine-site modulator [3H]flunitrazepam in the radioligand binding
assays, they were most likely benzodiazepine-site modulators.
In case of compound 1, maximum potentiation for the GABA
compound2, theheterocyclicprecursor, azetidin-1-yl(5-(trimethyl-
stannyl)pyridin-3-yl)methanone (2a), was prepared as outlined in
Scheme 3. The synthesis began from 5-bromonicotinic acid, from
which 5-bromonicotinoyl chloride was generated by using oxalyl
current at EC20 of GABA reached about 40% at both
a2- and
a
3-GABA Rs with compound’s EC50 values of 34 and 77 nM,
A
NH2
N
O
N
NH2
O
N
B(OH)
2
a
+
N
N
N
Br
N
N
3
3 4 2 3 2
Scheme 1. Reagents and conditions: (a) Pd(PPh ) , K CO , DME/EtOH/H O, 90 °C.

8
376
X. F. Liu et al. / Bioorg. Med. Chem. 18 (2010) 8374–8382
NH2
N
O
N
NH2
N
O
N
SnBu
3
a
N
N
+
N
N
Br
1
NH2
O
NH2
O
R1
R1
a
N
R2
N
R2
Ar SnBu3
+
N
N
X
N
X
N
Br
Ar
X = H, F
2, 4-13
Scheme 2. Reagents and conditions: (a) Pd(PPh
3 4
) , CuI, DMF, 90–100 °C.
O
N
O
N
N
COOH
COCl
a
b
c
N
N
N
Br
Br
Br
SnMe3
2
b
2a
Scheme 3. Reagents and conditions: (a) (COCl)
2 2
, CH Cl
2
, 0 °C; (b) TEA, azetidine, 0 °C to rt; (c) Pd(PPh
3
4 3 2
) , (SnMe ) , xylene, 140 °C.
Table 1
Comparing selectivity profiles of diazepam and compounds with significant NAM activities at
a
5-GABA
A
Rsa
GABA
subtype
A
1
2
4
DMCM
Diazepam
NH2
N
NH
2
O
NH
N
2
O
Cl
O
O
O
N
H
N
H
N
N
N
N
O
O
N
N
N
N
N
N
N
O
N
N
N
O
b
a1b2
a2b3
a3b3
a5b3
c
c
c
c
2
2
2
2
1 ± 4; (6)(n/a)b
44 ± 5; (4)(34)
40 ± 3; (5)(77)
À14 ± 5; (6)(389)
À5 ± 3; (6)(54)
À5 ± 2; (4)(n/a)
À68 ± 2; (4)(9)
À61 ± 4; (3)(14)
À58 ± 2; (4)(10)
À69 ± 2; (4)(3)
192 ± 14; (6)(85)
177 ± 6; (3)(35)
207 ± 24; (5)(131)
153 ± 9; (5)(45)
22 ± 3; (6)(104)
41 ± 4; (5)(379)
À52 ± 3; (7)(1130)
À11 ± 1; (5)(174)
À45 ± 1; (5)(455)
À46 ± 3; (4)(206)
a
Data in this table were derived from concentration-response curves (Figs. 2–6). Maximum %modulations and EC50 values were obtained by Prism non-linear regression.
Data are represented as the best fit value for Top ± Std. error; (number of individual oocytes)(EC50 in nM).
Data did not converge. Percent modulation at 1-lM point of the curve was used instead.
b
respectively (Fig. 2 and Table 1). At
tion by compound 1 reached À46% with an EC50 value of 206 nM.
Neither potentiation nor attenuation was observed at 1-GABA Rs
for up to 10 M of compound 1. There was no big difference in
terms of binding affinity for compound 1 among different sub-
types. For example, the K values of compound 1 for 2- and 5-
GABA Rs were 8.3 and 14 nM, respectively (Table 2). Compound
turned out to be a weak NAM at all subtypes except at the 5-
a
5-GABA
A
Rs, negative modula-
In parallel experiments, diazepam showed 177% and 153%
potentiation for the GABA response at EC20 of GABA at
5-GABA
and Table 1). A non-selective full ‘inverse agonist’ of GABA
a
2- and
a
A
a
A
Rs with EC50 values of 35 and 45 nM, respectively (Fig. 4
l
A
recep-
tors, DMCM, showed maximum modulation of À61% and À69% with
EC50 values of 14 and 3 nM at 2- and 5-GABA Rs, respectively
(Fig. 5 and Table 1). The maximum attenuation by compound 1 at
5-GABA Rs therefore was 67% of the maximum attenuation
produced by DMCM. The maximum potentiation by compound 1
at 2-GABA Rs was 25% of the maximum potentiation exerted by
i
a
a
a
a
A
A
2
a
a
A
subtype, for which compound 2 showed À45% modulation with
an EC50 value of 455 nM (Fig. 3 and Table 1).
a
A

X. F. Liu et al. / Bioorg. Med. Chem. 18 (2010) 8374–8382
8377
1
00
100
α1β2γ2; n = 6
α2β3γ2; n = 4
α3β3γ2; n = 5
α5β3γ2; n = 4
α1β2γ2; n = 6
8
6
4
2
0
0
0
0
80
60
α2β3γ2; n = 6
α3β3γ2; n = 5
α5β3γ2; n = 5
40
20
0
20
40
60
80
0
-
-
-
-
-20
-40
-60
-80
-9
-8
-7
Log[M]
-6
-5
-4
-9
-8
-7
-6
-5
Log[M]
Figure 2. Functional activities of compound 1 measured separately at four different
Figure 3. Functional activities of compound 2 measured separately at four different
subtypes of human GABA
Mean ± SEM.
A
receptors expressed in oocytes. Data are represented as
subtypes of human GABA
Mean ± SEM.
A
receptors expressed in oocytes. Data are represented as
diazepam. Although it is known that ‘partial agonists’ at the
benzodiazepine site on 2-/ 3-GABA Rs are efficacious anxiolytics
with less side effects, effectiveness of compounds like 1, regarding
either anxiolytic or cognitive effects, can be further corroborated
using relevant animal models and in clinical settings.
2
50
a
a
A
2
1
α1β2γ2; n = 6
α2β3γ2; n = 3
α3β3γ2; n = 5
α5β3γ2; n = 5
200
2
.3. SAR studies for the NAM activity at
A
a5-GABA Rs
150
The discovery of compound 1 prompted further investigation.
Herein, efforts to improve the overall profile of the modulators
by analyzing structure–activity relationships (SAR) are described.
Compound 3 (9-amino-2-cyclobutyl-5-(pyrimidin-5-yl)-2,3-dihy-
dro-1H-pyrrolo[3,4-b]quinolin-1-one) (Scheme 1) differs from
compound 1 with respect to the nitrogen positions on the pyrimi-
dine ring. This difference significantly reduced the NAM activity of
1
00
50
0
the compound at
ity. One micromolar of compound 3 had only À2% modulation for
5-GABA Rs with a K value of 12 nM while compound 1 showed
À32% modulation under the same experimental conditions with
a K value of 14 nM (Table 2), suggesting that the pyrimidine ring
might be the most important side group of compound 1 for its
A
a5-GABA Rs without affecting the binding affin-
-9
-8
-7
-6
a
A
i
Log[M]
i
Figure 4. Functional activities of diazepam measured separately at four different
subtypes of human GABA
Mean ± SEM.
A
receptors expressed in oocytes. Data are represented as
NAM activity at
A
a5-GABA Rs and the position of the two nitrogen
Table 2
a
Impact of the 2-pyrimidyl group on the dual functional activities
Compound
Structure
a5b3c
2
a2b3c2
%
modulation
K
i
(nM)
%modulation
38 ± 4; (4)
i
K (nM)
NH2
O
N
1
À32 ± 2; (4)
14
8.3
N
N
N
NH2
O
N
3
À2 ± 3; (4)
12
18 ± 2; (4)
1.5
N
N
N
a
Compounds were tested at 1 lM with GABA concentration at EC10. Data are represented as Mean ± SEM; (number of individual oocytes).

8
378
X. F. Liu et al. / Bioorg. Med. Chem. 18 (2010) 8374–8382
1
00
sional profiles of the pyrimidine ring. Quantum mechanical assess-
ments of conformations of 4 and 9 further illustrate the effect of
the fluoro group on the geometries and torsional energy barriers
(Section 4.3 and Fig. 7). Compound 4 had lower energy barrier to pla-
narity (3.4 kcal/mol) than compound 9 (9.8 kcal/mol), suggesting
that a low energy barrier to planarity might be helpful to the NAM
activity. Based on the differences between compounds 1 and 3 and
α1β2γ2; n = 4
α2β3γ2; n = 3
α3β3γ2; n = 4
α5β3γ2; n = 4
6
2
0
0
-
10
-9
-8
-7
-6
the effects of 7-fluorination on the
A
a5-GABA R NAM activity, we
postulate that the position of the two nitrogen atoms in the pyrimi-
dine ring, approximal to the cinnoline core structure, must play a
pivotal role within the receptor in achieving the compound’s and/
-
-
20
60
or receptor’s active conformation needed for the
A
a5-GABA R NAM
efficacy.
The 7-fluorination also affected the
ties of compounds 4–13 but not as much, percentage wise, as
observed for the 5-GABA R NAM activities (Table 3) and both in-
creased and decreased 2/ 3-GABA R PAM activities were seen
upon 7-fluorination. Comparing compounds 4 and 9, fluorination
reduced the 2-GABA R PAM activity by 47%, indicating that the
conformation needed for the 5-GABA NAM activity of
compound 4 also helped its 2-GABA R PAM activity. Interestingly,
the affinity of compound (K = 6 nM) was higher than
= 380 nM) at 2-GABA Rs. So far, little correlation between
a2/a3-GABA R PAM activi-
A
-
100
a
A
a
a
A
Log[M]
Figure 5. Functional activities of DMCM measured separately at four different
subtypes of human GABA
Mean ± SEM.
a
A
A
receptors expressed in oocytes. Data are represented as
a
A
R
a
A
9
i
4
atoms proximal to the quinoline core structure could be critical. In
addition, since the PAM activity of compound 3 at 2-GABA Rs
was 50% less than compound 1 (Table 2), the 2-pyrimidyl group
of compound 1 favored the 2-GABA R PAM efficacy as well.
These findings were further confirmed when a 2-pyrimidyl
group was placed on a cinnoline core structure. By replacing the
side group at the 8-position of compound 2 with a 2-pyrimidyl
group, we maintained the NAM activity of compound 2 at
BA Rs, eliminated the NAM activity at 1-GABA Rs, and enhanced
the PAM activities at 2/ 3-GABA Rs. The resulting cinnoline,
-amino-N-propyl-8-(pyrimidin-2-yl)cinnoline-3-carboxamide
4)(Table 1), showed 22% potentiation at 2-GABA Rs with an EC50
value of 104 nM, 41% potentiation at 3-GABA Rs with an EC50 va-
lue of 379 nM, and À52% modulation at 5-GABA Rs with an EC50
value of 1 M. Little activity was observed at 1-GABA Rs for up to
00 M of compound 4 (Fig. 6).
Attaching a fluoro group at the 7-position of the cinnoline core
(K
i
a
A
a
A
binding affinities and functional activities has been observed for
these compounds.
a
A
The
A A
a5-GABA R NAM and a2-/a3-GABA R PAM activities of
cinnolines 4–13 also depended on R1/R2 side groups at the 3-
position (Table 3). With 3-methylazetidin-1-yl, n-propyl, and azeti-
din-1-yl groups at R1/R2, compounds 6, 4, and 7 showed À35%,
a5-GA-
À26%, and À22% of
trast, high 2-GABA
with R1/R2 of n-propyl and cyclopropyl groups. For example, com-
pounds 4 and 5 showed more than 50% potentiation at 2-GABA Rs.
A
a5-GABA R NAM activities, respectively. In con-
A
a
A
a
A
R PAM activities were found in the compounds
a
a
A
4
(
a
A
a
A
In this study, selection of the side groups at R1/R2 was based on
their contribution to the overall selectivity profile. Most of the R1/
a
A
a
A
R2 side groups used in 4–13 tend to result in high
A
a2/a3-GABA R
l
a
A
and low 1/ 5-GABA R PAM activities. However, since the side
a
a
A
1
l
groups at R1/R2 and at the 8-position of cinnolines are both
structure was originally intended for maintaining metabolic stabil-
ity of the cinnolines. Interestingly, the 7-fluorination reduced the
negative modulation and even, in some cases, made compounds
Table 3
Fluorination reduced the NAM activity at
a5-GABA
A
Rsa
positive modulators of
negative modulation of
a
a
5-GABA
A
Rs (Table 3), indicating that the
NH2
O
5-GABA Rs may be affected by the tor-
A
R1
N
N
N
R2
1
00
X
X
a5b3
c
2
a2b3c2
α1β2γ2; n = 4
8
6
4
2
0
0
0
0
0
N
N
α2β3γ2; n = 6
α3β3γ2; n = 5
α5β3γ2; n = 7
1 2
NR R
H
F
4: À26 ± 2; (4)
9: À3 ± 4; (4)
4: 58 ± 10; (5)
9: 31 ± 1; (4)
nPr
cPr
N
H
H
F
5: À5 ± 2; (4)
10: 2 ± 5; (3)
5: 54 ± 6; (4)
10: 39 ± 4; (4)
N
H
-
-
-
-
20
40
60
80
H
F
6: À35 ± 3; (4)
11: 0 ± 2; (4)
6: 14 ± 2; (4)
11: 18 ± 2; (5)
N
H
F
7: À22 ± 1; (4)
12: 10 ± 3; (4)
7: À5 ± 2; (5)
12: 8 ± 1; (4)
N
N
H
F
8: À4 ± 4; (4)
13: 13 ± 4; (3)
8: 6 ± 1; (4)
13: À7 ± 5; (5)
-
9
-8
-7
-6
-5
-4
F
Log[M]
F
a
Compounds were tested at 1 lM with GABA concentration at EC10. Data are
represented as compound number: Mean of %modulation ± SEM (number of indi-
vidual oocytes).
Figure 6. Functional activities of compound 4 measured separately at four different
subtypes of human GABA
Mean ± SEM.
A
receptors expressed in oocytes. Data are represented as

X. F. Liu et al. / Bioorg. Med. Chem. 18 (2010) 8374–8382
8379
Figure 7. Quantum mechanical simulations of the torsional profiles of compounds 4 (dashed line) and 9 (solid line). Top and side views of the overlays of 4 and 9 are shown
for the minima and transition states. The minimum energy structure for 4 has a dihedral angle of ca. 45° whereas the fluorinated compound 9 has a dihedral angle of ca. 75°.
In both cases the energy barrier corresponding to orthogonal rings is small (<1 kcal/mol) whereas the barriers to planarity are estimated to be 3.4 and 9.8 kcal/mol for
A
compounds 4 and 9, respectively. The shapes of these torsional profiles may impact the positive versus negative modulation of a5-GABA Rs.
important for the functional selectivity, caution must be exercised
to make any conclusion for the SAR based solely on R1/R2 variation.
3. Conclusions
Although dual functional allosteric modulators of GABA
A
In summary, we discovered, in the cinnoline and quinoline ser-
ies, 5-GABA R NAMs that also act on 2-/ 3-GABA Rs as PAMs.
The subtype selectivity of these dual functional modulators was
achieved through differentiation in their functional efficacy rather
than their binding affinity. They were benzodiazepine-site ‘neutral
receptors are technically challenging to develop, the strategy
can be an attractive option for treating certain indications in
neurological and psychiatric disorders and may offer certain
advantages over single functional modulators. Compounds with
a subtype selectivity profile like 1 or 4 might be useful when
both cognitive enhancement and anxiolysis are needed. For
example, clinically relevant levels of anxiety were recorded in
a
a
a
A
A
antagonists’ at
a
1-GABA Rs and were ‘partial agonists’ at
A
a
2-/a3-
GABA Rs and ‘partial inverse agonists’ at
A
a
5-GABA Rs in compari-
A
son to diazepam and DMCM, respectively. SAR studies revealed
that the pyrimidine side group at the 5-position of quinoline and
8-position of cinnoline was important for the dual functional activ-
ities, especially for the NAM activity at a5-GABA Rs. The alignment
A
of the two nitrogen atoms in the 2-pyrimidyl ring with the cinno-
3
0% of AD patients in addition to their memory loss.22,23 Besides
anxiolytic effects, 2-/ 3-GABA R PAMs have been shown to im-
prove performance in treating cognitive deficits of schizophre-
a
a
A
2
4
nia. It remains to be seen the extent to which improvement
of both prefrontal cortical function and hippocampus-dependent
line or quinoline core structures may impact the compound’s effi-
memory by a dual functional modulator of GABA
efits AD or schizophrenia patients.
A
receptors ben-
A
cacy as a5-GABA R NAMs.
Both
sensorimotor gating. Yee and co-workers discuss weighing the
potential use of 5-GABA R NAMs to treat hippocampal-related
a
5-GABA
A
Rs and
a
3-GABA
A
Rs are believed to impact
4. Experimental
4.1. Chemistry
a
A
mnemonic dysfunction against the possibility that such com-
pounds may be adversely associated with deficient sensorimotor
4.1.1. General
2
5
gating. They also discuss treating sensorimotor-gating deficits
with 3-GABA
R PAMs.26 This and other remaining concerns with
5-GABA R NAMs such as anxiogenic and pro-convulsion might be
addressed by incorporating 2-/ 3-GABA R PAM activities into the
same molecule with 5-GABA R NAMs.
All chemicals and reagents were purchased from Sigma-Aldrich,
Fisher Scientific Inc., Strem Chemicals, Matrix Scientific, ACE
Synthesis, and Frontier Scientific Inc. All reactions were carried
out under a nitrogen atmosphere. Silicycle SillaFlash cartridges
were used for flash chromatography. H NMRs were obtained on
a
A
a
A
Ò
a
a
A
1
a
A

8
380
X. F. Liu et al. / Bioorg. Med. Chem. 18 (2010) 8374–8382
a Bruker Avance DPX300 NMR or on a Bruker Avance II 500 NMR,
and were consistent with the assigned structures. All NMRs were
recorded in CDCl or DMSO-d unless otherwise specified. High
3 6
at 140 °C overnight. Upon completion, the reaction mixture was
cooled to room temperature, washed with water (100 mL Â 3),
dried through magnesium sulfate, and evaporated to dryness.
The crude product was loaded onto a silica gel column, and eluted
resolution mass spectrometry (HRMS) were determined on an
Agilent Technologies 6210 Time-of-Flight LC/MS; its HPLC was
conducted using an Agilent SB C-8 analytical column (2.1 Â 30
with 0–100% ethyl acetate in hexane to give 2a in 83% overall yield
1
(846 mg). H NMR (300 MHz, DMSO-d
6
) d ppm 0.33 (s, 9H), 2.26
mm, 1.8
l
m) and eluting from 5% to 95% methanol (containing
(quin, J = 7.7 Hz, 2H), 4.06 (t, J = 7.5 Hz, 2H), 4.31 (t, J = 7.6 Hz,
2H), 7.94–8.12 (m, 1H), 8.68 (dd, J = 3.2, 1.9 Hz, 2H).
0
.1% formic acid) on a 2 min run (1.2 mL/min).
4
5
.1.5.2. Azetidin-1-yl(5-bromopyridin-3-yl)methanone (2b).
-Bromonicotinic acid (1.0 g, 4.95 mmol) was suspended in
4
.1.2. Typical preparation of bromo azetidine precursor
To the corresponding 8-bromo-3-carboxylic acid (1 equiv) in
methylene chloride (15 mL) at 0 °C. Oxalyl chloride (817 mg,
.44 mmol) was added and stirred at 0 °C for 30 min. Triethyl-
amine (1.252 g, 12.38 mmol) was added, and then azetidine
565 mg, 9.9 mmol) was added at 0 °C. The reaction was warmed
dimethylformamide (DMF) at 0 °C was added in one portion
carbonyldiimidazole (1.6 equiv), mixture stirred at 0 °C for
1
6
5 min, ice bath removed and stirred at room temperature for
h (becomes turbid). To this turbid mixture at 0 °C was added in
(
2
to room temperature for 2 h, and then quenched with water. The
reaction mixture was diluted with methylene chloride (200 mL),
washed with 10% potassium carbonate, dried with magnesium sul-
fate, and evaporated to dryness. The crude product was loaded to a
silica gel column, and eluted with 0–10% methanol in methylene
one portion a preformed solution of the azetidine (1.1 equiv) and
N,N-diisopropylethylamine (DIEA) (1.1 equiv) in DMF. The reaction
concentration was 0.3 M. The pale yellow clearing solution was
allowed to warm to room temperature, monitoring by LC/MS. Upon
completion, the volatiles were removed in vacuo to give a crude
product residue which was subjected to a silica gel column eluting
with a 0–50% ethyl acetate in methylene chloride gradient to give a
solid as the desired product.
1
chloride to give 2b in 71% overall yield (846 mg). H NMR
(
300 MHz, DMSO-d
J = 7.8 Hz, 2H), 4.34 (t, J = 7.6 Hz, 2H), 8.20 (t, J = 2.0 Hz, 1H), 8.76
d, J = 1.8 Hz, 1H), 8.83 (d, J = 2.3 Hz, 1H).
6
) d ppm 2.27 (dt, J = 15.6, 7.8 Hz, 2H), 4.07 (t,
(
4
.1.3. General procedure A for the Stille coupling
The corresponding bromide (1 equiv), the corresponding
4
.1.6. 9-Amino-2-cyclobutyl-5-(pyrimidin-5-yl)-2,3-dihydro-1H-
pyrrolo[3,4-b]quinolin-1-one (3)
-Amino-5-bromo-2-cyclobutyl-2,3-dihydro-1H-pyrrolo[3,4-b]
tributylstannyl reagent (2.2 equiv), tetrakis(triphenylphosphine)
palladium(0) (0.1 equiv) and copper(I) iodide (0.2 equiv) in DMF
were heated at 90–100 °C, monitoring by LC/MS. The reaction con-
centration was 0.3 M. Upon completion, the reaction mixture was
cooled to room temperature, diluted with chloroform (200 mL)
and washed with a minimal volume of ammonium hydroxide and
water (100 mL Â 3). The chloroform layer was dried with magne-
sium sulfate, and evaporated to constant mass. The crude product
was loaded to a silica gel column, and eluted with 0–10% methanol
in methylene chloride to give a yellow or tan solid. The solid was fur-
ther purified by HPLC, and eluted with 5–100% acetonitrile in water
to give a white or yellow solid as the desired product.
9
19
quinolin-1-one (250 mg, 0.75 mmol), pyrimidine-5-boronic acid
186 mg, 1.51 mmol), potassium carbonate (624 mg, 4.52 mmol)
(
and tetrakis(triphenylphosphine)palladium(0) (43.5 mg, 0.04
mmol) were added to dimethoxyethane (DME) (3 mL)/water
(
1.286 mL)/ethanol (0.857 mL) to give a yellow suspension. The
reaction mixture was heated to 90 °C as a yellow solution for 2 h.
The reaction mixture was cooled to room temperature, evaporated,
all of the solvent removed and the residue dissolved in methylene
chloride (200 mL). The methylene chloride solution was washed
with saturated sodium bicarbonate (20 mL Â 2) and brine (20
mL), dried with magnesium sulfate, filtered and concentrated.
The crude product was loaded to a silica gel column and eluted
with 30–100% ethyl acetate in hexane to give a brown solid
4
1
.1.4. 9-Amino-2-cyclobutyl-5-(pyrimidin-2-yl)-2,3-dihydro-
H-pyrrolo[3,4-b]quinolin-1-one (1)
(
210 mg). The brown solid was further crystallized from diethyl
Compound1wassynthesizedin9%yieldfrom9-amino-5-bromo-
-cyclobutyl-2,3-dihydro-1H-pyrrolo[3,4-b]quinolin-1-one and
ether/methylene chloride (4 mL/1 mL) at 0 °C overnight to give a
1
9
2
2
white solid as the desired product, which was dried at 90 °C under
-(tributylstannyl)pyrimidine according to the general procedure
high vacuum overnight to give 3 (193 mg; 77% yield). 1H NMR
1
A. H NMR (500 MHz, chloroform-d) d ppm 1.65–1.83 (m, 2H),
(
300 MHz, chloroform-d) d ppm 1.70–1.92 (m, 2H), 2.17–2.39 (m,
H), 4.44 (s, 2H), 4.91 (quin, J = 8.5 Hz, 1H), 6.45 (br s, 2H), 7.58
dd, J = 8.3, 7.3 Hz, 1H), 7.76 (dd, J = 7.2, 1.3 Hz, 1H), 7.93 (dd,
J = 8.3, 1.4 Hz, 1H), 9.09 (s, 2H), 9.23 (s, 1H); HRMS (ESI, Pos) m/z
2
2
.17–2.33 (m, 4H), 4.46 (s, 2H), 4.90 (quin, J = 8.7 Hz, 1H), 6.41 (br s,
H), 7.33 (t, J = 5.0 Hz, 1H), 7.56 (dd, J = 8.2, 7.3 Hz, 1H), 7.96 (t,
4
(
J = 7.2 Hz, 2H), 8.93 (d, J = 4.9 Hz, 2H); HRMS (ESI, Pos) m/z
3
32.1515 (MH+); HPLC 0.74 min.
3
32.1515 (MH+); HPLC 0.81 min.
4
.1.5. 4-Amino-8-(5-(azetidine-1-carbonyl)pyridin-3-yl)-N-pro-
4
.1.7. 4-Amino-N-propyl-8-(pyrimidin-2-yl)cinnoline-3-carb-
pylcinnoline-3-carboxamide (2)
oxamide (4)
Compound 2 was synthesized in 44% yield from 4-amino-8-bro-
mo-N-propylcinnoline-3-carboxamide20 and azetidin-1-yl(5-(trib-
utylstannyl)pyridin-3-yl)methanone (2a) according to the general
procedure A. H NMR (300 MHz, chloroform-d) d ppm 1.02 (t,
J = 7.4 Hz, 3H), 1.63–1.75 (m, 2H), 2.40 (quin, J = 7.8 Hz, 2H), 3.48
Compound 4 was synthesized in 49% yield from 4-amino-8-bromo-
20
N-propylcinnoline-3-carboxamide
dine according to the general procedure A. H NMR (300 MHz, chloro-
and 2-(tributylstannyl)pyrimi-
1
1
form-d) d ppm 8.96 (d, J = 4.8 Hz, 2H), 8.55 (br s, 1H), 8.09 (dd, J = 7.2,
.1 Hz, 1H), 7.98 (dd, J = 8.4, 1.3 Hz, 1H), 7.74 (dd, J = 8.4, 7.2 Hz, 1H),
.35 (t, J = 5.0 Hz, 1H), 3.45 (q, J = 6.9 Hz, 2H), 1.58–1.72 (m, 2H), 0.99
1
7
(
(
2
(
q, J = 6.7 Hz, 2H), 4.28 (br s, 2H), 4.46 (br s, 2H), 7.69–7.89 (m,
H), 7.95 (d, J = 8.1 Hz, 1H), 8.40 (s, 1H), 8.50 (br s, 1H), 8.89
br s, 1H), 9.00 (br s, 1H); HRMS (ESI, Pos) m/z 391.19 (MH+); HPLC
t, J = 7.4 Hz, 3H); HRMS (ESI, Pos) m/z 309.15 (MH+); HPLC 0.70 min.
0
.88 min.
4.1.8. 4-Amino-N-cyclopropyl-8-(pyrimidin-2-yl)cinnoline-3-
carboxamide (5)
4
.1.5.1. Azetidin-1-yl(5-(trimethylstannyl)pyridin-3-yl)metha-
Compound 5 was synthesized in 58% yield from 4-amino-8-
20
none (2a). Compound 2b (670 mg, 2.50 mmol), hexamethyldist-
annane (1.58 g, 4.50 mmol) and tetrakis(triphenylphosphine)
palladium(0) (140 mg, 0.25 mmol) in xylene (40 mL) were refluxed
bromo-N-cyclopropylcinnoline-3-carboxamide and 2-(tributylstannyl)
1
pyrimidine according to the general procedure A.
H NMR
(300 MHz, chloroform-d) d ppm 8.96 (d, J = 4.8 Hz, 2H), 8.55 (br s,

X. F. Liu et al. / Bioorg. Med. Chem. 18 (2010) 8374–8382
8381
1
(
7
H), 8.10 (dd, J = 7.1, 1.2 Hz, 1H), 7.98 (dd, J = 8.5, 1.2 Hz, 1H), 7.78
dd, J = 8.4, 7.2 Hz, 1H), 7.36 (t, J = 5.0 Hz, 1H), 2.91–3.03 (m, J = 7.3,
.3, 3.8, 3.8, 3.7 Hz, 1H), 0.83–0.92 (m, 2H), 0.58–0.67 (m, 2H);
HRMS (ESI, Pos) m/z 306.12 (MH+); HPLC 0.63 min.
2H), 7.96 (dd, J = 9.2, 5.3 Hz, 1H), 7.51 (t, J = 8.9 Hz, 1H), 7.39 (t,
J = 4.9 Hz, 1H), 4.89 (dd, J = 10.0, 8.5 Hz, 1H), 4.28–4.48 (m, 2H),
3.79 (dd, J = 9.9, 5.6 Hz, 1H), 2.62–2.85 (m, 1H), 1.27 (d, J = 7.0 Hz,
3H); HRMS (ESI, Pos) m/z 338.13 (MH+); HPLC 0.78 min.
4
.1.9. (4-Amino-8-(pyrimidin-2-yl)cinnolin-3-yl)(3-methylazeti-
4.1.15. (4-Amino-7-fluoro-8-(pyrimidin-2-yl)cinnolin-3-yl)(azeti-
din-1-yl)methanone (12)
din-1-yl)methanone (6)
Compound 6 was synthesized in 31% yield from (4-amino-8-
bromocinnolin-3-yl)(3-methylazetidin-1-yl)methanone and 2-(tri-
butylstannyl)pyrimidine according to the general procedure A. H
Compound 12 was synthesized in 20% yield from (4-amino-8-
bromo-7-fluorocinnolin-3-yl)(azetidin-1-yl)methanone and 2-(trib-
1
utylstannyl)pyrimidine according to the general procedure A. 1
H
NMR (300 MHz, chloroform-d) d ppm 1.29 (d, J = 6.9 Hz, 3H),
NMR (300 MHz, chloroform-d) d ppm 8.96 (d, J = 4.9 Hz, 2H), 7.97
(dd, J = 9.3, 5.2 Hz, 1H), 7.51 (t, J = 8.9 Hz, 1H), 7.39 (t, J = 5.0 Hz,
1H), 4.82 (t, J = 7.7 Hz, 2H), 4.27 (t, J = 7.7 Hz, 2H), 2.34 (quin,
J = 7.8 Hz, 2H); HRMS (ESI, Pos) m/z 324.11 (MH+); HPLC 0.55 min.
2.70–2.84 (m, J = 13.7, 6.8, 1.3, 1.3 Hz, 1H), 3.81 (dd, J = 10.2,
5.4 Hz, 1H), 4.38 (td, J = 9.3, 0.8 Hz, 1H), 4.44–4.51 (m, 1H), 4.91–
5.02 (m, 1H), 7.35 (t, J = 4.9 Hz, 1H), 7.74 (dd, J = 8.4, 7.2 Hz, 1H),
8.00–8.07 (m, 1H), 8.09–8.15 (m, 1H), 8.95 (d, J = 4.9 Hz, 2H);
HRMS (ESI, Pos) m/z 321.15 (MH+); HPLC 0.74 min.
4.1.16. (4-Amino-7-fluoro-8-(pyrimidin-2-yl)cinnolin-3-yl)
(
3,3-difluoroazetidin-1-yl)methanone (13)
4
.1.10. (4-Amino-8-(pyrimidin-2-yl)cinnolin-3-yl)(azetidin-1-yl)
Compound 13 was synthesized in 37% yield from (4-amino-8-
methanone (7)
bromo-7-fluorocinnolin-3-yl)(3,3-difluoroazetidin-1-yl)methanone
Compound 7 was synthesized in 62% yield from (4-amino-8-
bromocinnolin-3-yl)(azetidin-1-yl)methanone and 2-(tributylstan-
nyl)pyrimidine according to the general procedure A. H NMR
and 2-(tributylstannyl)pyrimidine according to the general proce-
1
6
dure A. H NMR (300 MHz, DMSO-d ) d ppm 4.52 (t, J = 12.3 Hz,
1
2H), 4.96 (t, J = 12.3 Hz, 2H), 7.63 (t, J = 5.0 Hz, 1H), 7.81 (t,
J = 9.1 Hz, 1H), 8.65 (dd, J = 9.4, 5.6 Hz, 1H), 9.00 (d, J = 5.0 Hz, 2H);
HRMS (ESI, Pos) m/z 361.10 (MH+); HPLC 0.64 min.
(
300 MHz, chloroform-d) d ppm 2.29–2.42 (m, 2H), 4.25–4.33 (m,
2
7
H), 4.84–4.93 (m, 2H), 7.34 (t, J = 4.9 Hz, 1H), 7.73 (dd, J = 8.4,
.2 Hz, 1H), 7.98 (d, J = 8.1 Hz, 1H), 8.08 (d, J = 7.1 Hz, 1H), 8.95
(
d, J = 4.9 Hz, 2H); HRMS (ESI, Pos) m/z 306.12 (MH+); HPLC
4.2. Biology
0
.64 min.
4
.2.1. Radioligand binding assays
4
.1.11. (4-Amino-8-(pyrimidin-2-yl)cinnolin-3-yl)(3,3-difluoro-
Competitive binding assays were performed using the non-
azetidin-1-yl)methanone (8)
selective benzodiazepine-site modulator [3H]flunitrazepam on re-
Compound 8 was synthesized in 42% yield from (4-amino-8-bro-
combinant Sf9 membranes co-expressing the
of GABA receptor in combination of 1b2
or 5b3 2, which are major benzodiazepine-sensitive
a
, b, and
c
2,
subunits
3b3 2,
mocinnolin-3-yl)(3,3-difluoroazetidin-1-yl)methanone and 2-(trib-
A
a
c
2, 2b3
a
c
a
c
1
utylstannyl)pyrimidine according to the general procedure A.
NMR (300 MHz, DMSO-d
H
a
c
a-subtypes
2
7
6
) d ppm 4.46–4.61 (m, 2H), 4.91–5.09 (m,
found in CNS. The test compound was incubated for 1 h at 4 °C
2
1
H), 7.59 (t, J = 4.9 Hz, 1H), 7.85 (t, J = 7.7 Hz, 1H), 7.97–8.09 (m,
H), 8.57 (d, J = 8.2 Hz, 1H), 8.74 (br s, 2H), 8.98 (d, J = 4.8 Hz, 2H);
in assay buffer (20 mM Tris–Citrate and 200 mM NaCl, pH 7.8) con-
taining 20–100
l
g
of membrane protein and 3 nM of [3H]
l. The reaction was termi-
HRMS (ESI, Pos) m/z 342.10 (MH+); HPLC 0.68 min.
flunitrazepam in a total volume of 250
l
nated by rapid filtration through PerkinElmer GF/B UniFilters.
4
3
.1.12. 4-Amino-7-fluoro-N-propyl-8-(pyrimidin-2-yl)cinnoline-
-carboxamide (9)
Compound 9 was synthesized in 54% yield from 4-amino-8-bro-
Bound radioligands were measured by liquid scintillation counting.
The assay was run in parallel in presence of 10
determine nonspecific binding.
lM of flumazenil to
mo-7-fluoro-N-propylcinnoline-3-carboxamide20 and 2-(tributyl-
stannyl)pyrimidine according to the general procedure A.
After sigmoidal non-linear curve fitting to the data of specific
binding, compound’s IC50 values were obtained at the concentra-
1
H
NMR (300 MHz, chloroform-d) d ppm 8.98 (d, J = 5.1 Hz, 2H), 8.44
t, J = 4.7 Hz, 1H), 7.98 (dd, J = 9.3, 5.3 Hz, 1H), 7.52 (d, J = 9.1 Hz,
tion replacing 50% of radioligand bound to the receptor. The K
ues were calculated using the equation K = IC50/(1 + L/K ), where L
is the equilibrium dissoci-
i
val-
(
i
d
1H), 7.41 (t, J = 5.0 Hz, 1H), 3.36–3.49 (m, 2H), 1.58–1.68 (m, 2H),
0.98 (t, J = 7.4 Hz, 3H); HRMS (ESI, Pos) m/z 326.13 (MH+); HPLC
0.66 min.
is the radioligand concentration and K
d
ation constant of the radioligand.
4
.2.2. Electrophysiological assays to measure functional activity
4
.1.13. 4-Amino-N-cyclopropyl-7-fluoro-8-(pyrimidin-2-yl)cinno-
A
of allosteric modulators of GABA receptors
line-3-carboxamide (10)
Compound 10 was synthesized in 47% yield from 4-amino-8-bro-
mo-N-cyclopropyl-7-fluorocinnoline-3-carboxamide and 2-(trib-
4.2.2.1. Preparation of Xenopus oocytes and injection of
cRNA. Ovarian lobes of Xenopus laevis frogs were purchased from
NASCO (Fort Atkinson, WI) and were torn open carefully in OR2
20
1
utylstannyl)pyrimidine according to the general procedure A.
NMR (300 MHz, chloroform-d) d ppm 8.97 (d, J = 5.1 Hz, 2H), 8.42
d, J = 3.0 Hz, 1H), 7.98 (dd, J = 9.4, 5.2 Hz, 1H), 7.35–7.52 (m, 2H),
H
2 4 2
solution (82 NaCl, 2.5 HEPES, 1.5 NaH PO , 1 MgCl , 0.1 EDTA, in
mM, pH 7.4). The oocytes were defolliculated by incubating twice
in a total of 25 mL of OR2 solution containing 0.2% collagenase IA
(Sigma C5894) for a total of 60 min on a test tube rocker rocking
at 0.5 Hz. After several washes with OR2, oocytes were stored in
0.5Â Leibovitz’s L-15 medium (Sigma L1518) containing 0.1
mg/mL gentamicin (Sigma G1914), 100 units/mL of penicillin,
and 0.1 mg/mL of streptomycin (Sigma P4333). Stage V or VI oo-
cytes were selected and injected with cRNA on the same or follow-
ing day. Each oocyte was injected 20–40 nL of capped and
(
2.81–3.02 (m, J = 7.3, 7.3, 3.9, 3.8, 3.7 Hz, 1H), 0.81–0.89 (m, 2H),
0.56–0.66 (m, 2H); HRMS (ESI, Pos) m/z 324.11 (MH+); HPLC
0.57 min.
4
.1.14. (4-Amino-7-fluoro-8-(pyrimidin-2-yl)cinnolin-3-yl)(3-
methylazetidin-1-yl)methanone (11)
Compound 11 was synthesized in 28% yield from (4-amino-8-
bromo-7-fluorocinnolin-3-yl)(3-methylazetidin-1-yl)methanone
and 2-(tributylstannyl)pyrimidine according to the general proce-
linearized cRNA containing 0.2–2 ng of
human GABA receptor genes. The cRNA ratio of
1:1:10, 1:1:2, 1:1:10, and 1:1:5 by weight for (
a
, b, or
c
subunits of the
:b: were
1) (b2) 2,
A
a
a
c
1
dure A. H NMR (300 MHz, chloroform-d) d ppm 8.96 (d, J = 4.9 Hz,
2
2
c

8
382
X. F. Liu et al. / Bioorg. Med. Chem. 18 (2010) 8374–8382
(a2)
2
(b3)
2
c2, (a
3)
2
(b3)
2
c
2, and (
a5)
2
(b3)
2
c2, respectively. Oocytes
Supplementary data
were used for TEVC experiments 2–10 days after injection.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmc.2010.09.058. These data
include MOL files and InChiKeys of the most important compounds
described in this article.
4
.2.2.2. Evaluate modulators by two-electrode voltage clamp
recording. TEVC recording was carried out using OpusXpress
000A (Molecular Devices, Sunnyvale, CA), which allows simulta-
neous recording from eight oocytes. Oocytes were perfused with
ND96 (96 NaCl, 2 KCl, 1.8 CaCl , 1 MgCl , 5 HEPES, in mM, pH
.5) at a flow rate of 2 ml/min. Oocytes were impaled with two
micro-glass electrodes filled with 3 M KCl. Tip resistances of
.5–2 M and 0.5–10 M were used for the current and voltage
6
2
2
References and notes
7
1. Johnston, G. A. R. Pharmacol. Ther. 1996, 69, 173.
2
3
.
.
Kaplan, E. M.; DuPont, R. L. Curr. Med. Res. Opin. 2005, 21, 941.
Depoortere, H.; Zivkovic, B.; Lloyd, K. G.; Sanger, D. J.; Perrault, G.; Langer, S. Z.;
Bartholini, G. J. Pharmacol. Exp. Ther. 1986, 237, 649.
0
X
X
electrodes, respectively. Membrane potential was held at
À60 mV. Oocytes with leak current above 50 nA at the holding
potential were discarded.
4. Leidenheimer, N. J.; Schechter, M. D. Pharmacol. Biochem. Behav. 1988, 31,
49.
2
5
.
.
Atack, J. R.; Wafford, K. A.; Tye, S. J.; Cook, S. M.; Sohal, B.; Pike, A.; Sur, C.;
Melillo, D.; Bristow, L.; Bromidge, F.; Ragan, I.; Kerby, J.; Street, L.; Carling, R.;
Castro, J. L.; Whiting, P.; Dawson, G. R.; McKernan, R. M. J. Pharmacol. Exp. Ther.
2006, 316, 410.
Dawson, G. R.; Maubach, K. A.; Collinson, N.; Cobain, M.; Everitt, B. J.; MacLeod,
A. M.; Choudhury, H. I.; McDonald, L. M.; Pillai, G.; Rycroft, W.; Smith, A. J.;
Sternfeld, F.; Tattersall, F. D.; Wafford, K. A.; Reynolds, D. S.; Seabrook, G. R.;
Atack, J. R. J. Pharmacol. Exp. Ther. 2006, 316, 1335.
To determine EC20 concentration of GABA for each oocyte, a series
of 30 s-pulses with increasing concentrations of GABA was applied
to oocytes with 4.5 min-intervals. To test a modulator, oocytes were
pre-treated with the compound for 100 s before they were co-trea-
ted with GABA at EC20 for 30 s. A set of three pulses of GABA alone at
EC20 prior to modulator testing was conducted and the average of
the three readings was used to establish thebaseline GABA response.
For concentration-response curves, increasing concentrations of the
modulator were applied on the same oocyte at 5 min-intervals. The
GABA EC50 values averaged by geometric mean from eight individual
6
7
8
.
.
Lader, M. H. Eur. Neuropsychopharmacol. 1999, 9, S399.
Ashton, H. In Adverse Syndromes and Psychiatric Drugs: A Clinical Guide; Haddad,
P., Dursun, S., Deakin, B., Eds.; Oxford University Press: Oxford, 2004; pp 239–
260.
9.
Rudolph, U.; Crestani, F.; Benke, D.; Brünig, I.; Benson, J. A.; Fritschy, J.; Martin,
J. R.; Bluethmann, H.; Möhler, H. Nature 1999, 401, 796.
oocytes for (
b3) 2 were typically 34, 27, 63, and 17
Modulation was calculated according to the amplitude change
a1)
2 2
(b2) c
2, (a
2)
2
(b3)
2
c
2 2 2
2, (a3) (b3) c2, and (a5)
1
0. McKernan, R. M.; Rosahl, T. W.; Reynolds, D. S.; Sur, C.; Wafford, K. A.; Atack,
J. R.; Farrar, S.; Myers, J.; Cook, G.; Ferris, P.; Garrett, L.; Bristow, L.; Marshall,
G.; Macaulay, A.; Brown, N.; Howell, O.; Moore, K. W.; Carling, R. W.; Street, L.
J.; Castro, J. L.; Ragan, C. I.; Dawson, G. R.; Whiting, P. J. Nat. Neurosci. 2000, 3,
(
2
c
lM, respectively.
of GABA current caused by the modulator [(current evoked by
modulator + GABA/current evoked by GABA alone) À1]. The cur-
rent amplitude was measured from baseline to peak using Clampfit
587.
11. Löw, K.; Crestani, F.; Keist, R.; Benke, D.; Brünig, I.; Benson, J. A.; Fritschy, J. M.;
Rülicke, T.; Bluethmann, H.; Möhler, H.; Rudolph, U. Science 2000, 290, 131.
12. Crestani, F.; Keist, R.; Fritschy, J. M.; Benke, D.; Vogt, K.; Prut, L.; Bluthmann, H.;
(
Molecular Devices, Sunnyvale, CA). Concentration-response
Mohler, H.; Rudolph, U. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 8980.
curves were plotted and fitted with Prism GraphPad using sigmoi-
dal non-linear regression with variable slopes (GraphPad Software,
Inc. San Diego, CA).
13. Rudolph, U.; Möhler, H. Curr. Opin. Pharmacol. 2006, 6, 18.
14. Settimo, F. D.; Taliani, S.; Trincavelli, M. L.; Montali, M.; Martini, C. Curr. Med.
Chem. 2007, 14, 2680.
1
1
1
5. Atack, J. R. CNS Neurosci. Ther. 2008, 14, 25.
6. Maubach, K. Curr. Drug Targets—CNS Neurol. Disord. 2003, 2, 233.
7. Collinson, N.; Atack, J. R.; Laughton, P.; Dawson, G. R.; Stephens, D. N.
Psychopharmacol.(Ber.) 2006, 188, 619.
4
.2.3. Screening for
For negative modulator screening, oocytes were treated with
GABA at EC20 for 30 s to establish a baseline GABA response fol-
lowed by a 30 s treatment with 10 M of the test compound and
A
a5-GABA R NAMs
18. Atack, J. R. Pharmacol. Ther. 2010, 125, 11.
19. Chang, H. F.; Chapdelaine, M.; Dembofsky, B. T.; Herzog, K. J.; Horchler, C.;
Schmiesing, R. J. WO2008155572 A2, 2008.
l
20. Chapdelaine, M. J.; Ohnmacht, C. J.; Becker, C.; Chang, H. F.; Dembofsky, B. T.
GABA at EC20. Eight compounds were tested on the same oocyte
sequentially with 4 min intervals. For n = 2, one 96-well plate
was assayed from A to H and another assayed from H to A to
monitor interference among compounds with a throughput of
US2007142328 A1 and WO2007073283 A1, 2007.
21. Ator, N. A.; Atack, J. R.; Hargreaves, R. J.; Burns, H. D.; Dawson, G. R. J.
Pharmacol. Exp. Ther. 2010, 332, 4.
2
2. Porter, V. R.; Buxton, W. G.; Fairbanks, L. A.; Strickland, T.; O’Connor, S. M.;
Rosenberg-Thompson, S.; Cummings, J. L. J. Clin. Neuropsychiatry Clin. Neurosci.
2003, 15, 180.
6
4 data points/32 compounds per run on OpusXpress 6000A.
2
2
3. Jost, B. C.; Grossberg, G. T. J. Am. Geriatr. Soc. 1996, 44, 1078.
4. Lewis, D. A.; Cho, R. Y.; Carter, C. S.; Erklund, K.; Forster, S.; Kelly, M. A.;
Montrose, D. Am. J. Psychiatry 2008, 165, 1585.
Positives were re-tested at EC20 of GABA to establish concentra-
tion-response curves of the test compound. Before the NAM
screening, compounds were pre-screened by preset criteria for
other physical properties and biological activities such as their
aqueous solubility at pH7.4 and their stability in human and
rat liver microsomes.
25. Hauser, J.; Rudolph, U.; Keist, R.; Mohler, H.; Feldon, J.; Yee, B. K. Mol. Psychiatry
2005, 10, 201.
2
6. Yee, B. K.; Keist, R.; von Boehmer, L.; Studer, R.; Benke, D.; Hagenbuch, N.;
Dong, Y.; Malenka, R. C.; Fritschy, J. M.; Bluethmann, H.; Feldon, J.; Möhler, H.;
Rudolph, U. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 17154.
2
2
7. Whiting, P. J. Drug Discovery Today 2003, 8, 445.
8. Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
4
.3. Quantum mechanical simulations
29. Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
3
3
3
0. Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys. 1980, 72, 650.
1. McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639.
Absolute energies, equilibrium structures and harmonic vibra-
2. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K. N.; Burant, J. C.;
Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.;
Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.;
Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken,
V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A.
J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G.
A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe,
M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A.
GAUSSIAN03, Revision C.02, Gaussian, Inc., Wallingford CT, 2004.
tional frequencies were determined for each stationary point
described. The B3LYP generalized gradient approximation ex-
change-correlation density functional was used throughout.
The 6-311G basis set was used for all computations.
and transition states were optimized via analytic gradients until
2
8,29
**
30,31
Minima
À4
the residual root mean square gradient was less than 10
hartree/bohr. The mass-weighed Hessian matrix, and hence the
harmonic vibrational frequencies, were determined analytically.
Minima and transition states were confirmed by the presence of
zero or one imaginary frequency, respectively. All computations
were carried out with the GAUSSIAN03 program package.32