Identification of a novel, selective GABAa α5 receptor inverse agonist which enhances cognition

Article abstract of DOI:10.1021/jm020582q

In pursuit of a GABA_A α5-subtype-selective inverse agonist to enhance cognition, a series of 6,7-dihydro-2-benzothiophen-4(5H)-ones has been identified as a novel class of GABA_A receptor ligands. These thiophenes have higher binding affinity for the GABA_A α5 receptor subtype compared to the GABA_A α1, α2, and α3 subtypes, and several analogues exhibit high GABA_A α5 receptor inverse agonism. 6,6-Dimethyl-3-(2-hydroxyethyl)thio-1-(thiazol-2-yl)-6,7-dihydro-2- benzothiophen-4(5H)-one (43) has been identified as a full inverse agonist at the GABA_A α5 receptor and is functionally selective over the other major GABA_A receptor subtypes. 43 readily penetrates into the CNS to give selective occupancy of GABA_A α5 receptors. In addition, 43 enhances cognitive performance in rats in the delayed 'matching-to-place' Morris water maze test - a hippocampal-dependent memory task - without the convulsant or proconvulsant activity associated with nonselective, GABA_A receptor inverse agonists.

Full text of DOI:10.1021/jm020582q

J . Med. Chem. 2003, 46, 2227-2240
2227
Id en tifica tion of a Novel, Selective GABAA r5 Recep tor In ver se Agon ist Wh ich
En h a n ces Cogn ition
Mark S. Chambers,* J ohn R. Atack, Howard B. Broughton, Neil Collinson, Susan Cook, Gerard R. Dawson,
Sarah C. Hobbs, George Marshall, Karen A. Maubach, Goplan V. Pillai, Austin J . Reeve, and Angus M. MacLeod
Merck Sharp & Dohme Research Laboratories, The Neuroscience Research Centre, Terlings Park, Eastwick Road,
Harlow, Essex, CM20 2QR, U.K.
Received December 27, 2002
In pursuit of a GABAA R5-subtype-selective inverse agonist to enhance cognition, a series of
6,7-dihydro-2-benzothiophen-4(5H)-ones has been identified as a novel class of GABAA receptor
ligands. These thiophenes have higher binding affinity for the GABAA R5 receptor subtype
compared to the GABAA R1, R2, and R3 subtypes, and several analogues exhibit high GABAA
R5 receptor inverse agonism. 6,6-Dimethyl-3-(2-hydroxyethyl)thio-1-(thiazol-2-yl)-6,7-dihydro-
2-benzothiophen-4(5H)-one (43) has been identified as a full inverse agonist at the GABAA R5
receptor and is functionally selective over the other major GABAA receptor subtypes. 43 readily
penetrates into the CNS to give selective occupancy of GABAA R5 receptors. In addition, 43
enhances cognitive performance in rats in the delayed ‘matching-to-place’ Morris water maze
testsa hippocampal-dependent memory taskswithout the convulsant or proconvulsant activity
associated with nonselective, GABAA receptor inverse agonists.
In tr od u ction
performance in animal models10 but can be anxiogenic,11
convulsant,12 and proconvulsant,13 and they also may
alter attentional processing.14 Finally, a compound
which binds to the BZ site but has no effect on the
frequency of GABA-induced channel openings is termed
a BZ antagonist.
γ-Aminobutyric acid (GABA) is the major inhibitory
neurotransmitter in mammalian brain. There are three
classes of GABA receptors: GABAA and GABAC recep-
tors are ligand-gated ion channels whereas GABAB
receptors are G-protein-coupled receptors.1 Of these, it
is the GABAA receptor family which has been the most
widely studied since this family is the site of action of a
number of clinically important drugs, including benzo-
diazepines (BZs), barbiturates, and anesthetics.2 GABAA
receptors are pentameric assemblies of protein derived
from a family of genes encoding a number of subunits
(R1-6, â1-3, γ1-3, δ, ꢀ, π, and θ). The majority of
GABAA receptors in the brain contain R-, â-, and
γ-subunits,3,4 and those containing R1, R2, R3 or R5 in
conjunction with a â- and γ2-subunit (around 80% of
the total GABAA receptor population)3 contain a BZ
binding site which allosterically modulates the func-
tional effects of GABA. For example, the binding of the
prototypic benzodiazepine diazepam (1) to the BZ site
on the GABAA receptor enhances the action of GABA
by increasing the frequency of GABA-activated channel
openings. Diazepam is a nonselective BZ agonist (or
positive allosteric modulator), in that it has equivalent
affinity and efficacy for R1-, R2-, R3-, and R5-containing
GABAA receptors; nonselective BZ agonists have found
therapeutic use as anxiolytics and anticonvulsants.5
However, they also impair learning and memory pro-
cesses.6,7 In contrast, the â-carboline DMCM (2) pro-
duces a decrease in the frequency of GABA-activated
channel openings, and it is termed an inverse agonist8,9
(or negative allosteric modulator). The opposing effects
of agonists and inverse agonists on GABAA receptor
function are reflected in whole animal pharmacology in
that nonselective BZ inverse agonists enhance cognitive
Using genetically modified (knock-in) mice it has been
demonstrated that GABAA receptors containing an R1-
subunit mediate the sedative/muscle relaxant effects of
benzodiazepines, whereas R2- and/or R3-subunit-con-
taining receptors mediate the anxiolytic and anticon-
vulsant effects.15-17 GABAA R5 receptors have a rela-
tively restricted distribution, being primarily expressed
in the hippocampus, a region of the brain associated
with learning and memory. Although R5 receptors
account for less than 5% of the total GABAA receptor
population in the brain, in the hippocampus they
represent 20% of all GABAA receptors,3,18,19 thereby
implicating this GABAA receptor subtype in learning
and memory processes. Thus, we proposed that a
selective R5 inverse agonist may have therapeutic utility
as a cognition-enhancing agent that may lack the
unwanted side effects associated with activity at other
GABAA receptor subtypes.
Selective GABAA R5 inverse agonism may be achieved
either in terms of binding selectivity (i.e., a high affinity
R5 inverse agonist which has very low affinity at the
other GABAA receptor subtypes) or functional selectivity
(i.e., an inverse agonist at the GABAA R5 receptor but
low efficacy at the other GABAA receptor subtypes). Of
the many different structural classes of GABAA receptor
ligands, it is only the imidazobenzodiazepines,20,21 for
example L-655,708 (3), and some diazepam analogues22
which are reported to exhibit binding selectivity for R5-
containing GABAA receptors compared to the other
receptor subtypes.
In this manuscript, we describe a novel class of
* To whom correspondence should be addressed. Tel: (01144)-1279-
440417. Fax: (01144)-1279-440390. E-mail: mark_chambers@merck.com
GABAA receptor ligands23 which have higher affinity for
10.1021/jm020582q CCC: $25.00 © 2003 American Chemical Society
Published on Web 04/25/2003

2228 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Ch a r t 1. GABAA Benzodiazepine Site Ligands
Chambers et al.
obtained from the cyanothiophene 37,26 via the thioam-
ide (38) and imino ether (40) derivatives, respectively
(Scheme 4).
Resu lts a n d Discu ssion
In Vitr o Bin d in g Affin ity. The in vitro binding
affinity for the thiophenes at the human GABAA recep-
tor subtypes are shown in Tables 1 and 2. In general, it
was found that a variety of thioether substituents are
well tolerated at the C3 position of the tetrahydroben-
zothiophene skeleton. As shown in Table 1, homologa-
tion of the methylthio substituent to ethylthio (7) or
propylthio (8) had no detrimental effect on R5 affinity
or subtype selectivity. Further extension to butylthio (9)
did, however, result in an order of magnitude decrease
in R5 binding affinity compared to 7, although selectivity
for the GABAA R5 subtype over the R1, R2, and R3
receptors was retained. The isopropylthio derivative 10
also had high GABAA R5 affinity but, as indicated by
the tert-butylthio analogue 11, there is a limit to the
extent of steric crowding that can be tolerated adjacent
to sulfur. The introduction of an hydroxyl group onto
the thioalkyl substituent had no detrimental effect on
either GABAA R5 affinity or subtype selectivity (cf. 12
with 7 and 8 with 13). Heteroarylthio substituents also
produced high affinity R5 ligands and, in particular,
thiazolylthio (14) and thienylthio (15) were particularly
well tolerated. In general, replacing the alkylthio sub-
stituent with alkyloxy or alkylamino resulted in a
decrease in R5 affinity, although one exception was the
isopropylamino derivative 20 which has similar R5
affinity to the isopropylthio analogue 10. The ethyl
derivative 22 has similar R5 affinity compared to 5, but,
as a general rule, replacing alkylthio with alkyl proved
detrimental to GABAA R5 binding (data not shown). The
importance of having a substituent at C3 is demon-
strated by the lack of activity demonstrated by 23. The
results in Table 2 show the GABAA receptor subtype
binding affinities for the C1-substituted heterocyclic
derivatives. At the C1 position a heteroaromatic ring
was found to be essential for high R5 affinity as
demonstrated by the lack of GABAA receptor affinity
displayed by the phenyl derivative 30. The position of
the nitrogen atom in the heterocycle was crucial for
optimum R5 binding affinity; it was those isomers in
which the nitrogen is adjacent to the point of attach-
ment of the heterocycle that had the highest R5 affinity.
For example, comparison of the pyridyl analogues 26-
28 reveals that it is the 2-pyridyl isomer 26 which has
the highest R5 affinity, whereas for the thiazolyl isomers
the 2-thiazolyl derivative 39 has 7-fold higher affinity
for the R5-subtype compared to the 5-thiazolyl isomer
29. Indeed, of all the heterocyclic groups we incorporated
at C1, in terms of R5 binding affinity, we found that
2-pyridyl or 2-thiazolyl were the best. It should be noted
that the existence of nitrogen adjacent to the point of
attachment is not, in itself, sufficient to produce high
affinity ligands. For example, the pyrrole 32, which
possesses only a H-bond donor group and no H-bond
acceptor, and the 3-methyloxadiazolyl derivative 36
have an order of magnitude lower GABAA R5 affinity
compared to the analogous pyrazole 5. As was observed
in the pyrazole series, the C3-(2-hydroxyethylthio) sub-
stituent is well tolerated when there is a 2-pyridyl (34)
or 2-thiazolyl group (43) at C1, and, in general, the
the R5-subtype compared to the other GABAA receptor
subtypes. Several examples exhibit high R5 inverse
agonism and, in particular, we describe the synthesis
and biological evaluation of a functionally selective
GABAA R5-subtype receptor full inverse agonist which
enhances cognitive performance in a hippocampal-
dependent memory task, without the convulsant or
proconvulsant liabilities associated with nonselective
GABAA inverse agonists.
Lea d Id en tifica tion . The approach we followed in
order to identify a novel and selective GABAA R5 inverse
agonist was to adopt a similarity searching technique,
using two-dimensional atom pair and topological torsion
descriptors.24 By comparing known BZ ligands such as
the pyridazinone 4 (Ki Bz binding sites: 4.2 nM)25 with
compounds from the Merck database, this directed
screening approach ultimately identified 6,6-dimethyl-
3-methylthio-1-(pyrazol-3-yl)-6,7-dihydro-2-benzothio-
phen-4(5H)-one (5)26 as a high affinity GABAA R5
receptor ligand (Ki 5.2 nM) that has 4-13-fold binding
selectivity over the GABAA R1, R2, and R3 receptor
subtypes. The tetrahydrobenzothiophene 5 represents
a structurally novel class of GABAA receptor ligands and
was used as the starting point in our structure-activity
studies.
Ch em istr y. Synthesis of the C3-substituted ana-
logues was carried out according to Scheme 1. Modifica-
tion of the thioalkyl group was achieved by oxidizing 5
to the sulfone 6 followed by reaction with the desired
thiolate. Similarly, reaction of 6 with sodium isopro-
poxide afforded the isopropyl ether 18. The methoxy
(17), alkylamino (19-21), and ethyl (22) derivatives
were prepared via reaction of the sulfoxide 16 with
sodium methoxide, the appropriate alkylamine, and
ethylmagnesium bromide, respectively. In the case of
the Grignard reaction, the unsubstituted thiophene 23
was also obtained. The introduction of heterocyclic
substituents (and phenyl) at the C1 position was ac-
complished according to Schemes 2-4. As shown in
Scheme 2, the majority of C1-substituents were incor-
porated using either Stille or Suzuki methodology via
the bromothiophene 25. The oxadiazole 36 was synthe-
sized by transforming the commercially available car-
boxylic acid 3526 into the imidazolide which, in turn,
was reacted with acetamide oxime (Scheme 3). The
thiazole (39) and triazole (41) analogues were both

Selective GABAA R5 Receptor Inverse Agonist
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2229
Sch em e 1a
a
Reagents: (i) DCM, dioxane, m-CPBA (2.0 equiv); (ii) RSH, NaOH, EtOH, 25 °C or 70 °C; (iii) RSNa, THF, 25 °C; (iv) DCM, dioxane,
m-CPBA (1.0 equiv); (v) NaOMe, MeOH, 70 °C; (vi) NaOiPr, PA, 70 °C; (vii) RNH2, MeOH or EtOH, 100 °C; (viii) EtMgBr, THF, -10 °C.
i
Sch em e 2a
a
Reagents: (i) MeSNa, THF; (ii) Pd(PPh3)4, ethylene glycol dimethyl ether, Na2CO3, H2O, 1-(tert-butyloxycarbonyl)pyrrol-2-yl boronic
acid; (iii) TFA; (iv) DCM, m-CPBA (1.0 equiv), -50 °C; (v) NaOH, HS(CH2)2OH, EtOH.
2-hydroxyethylthio derivatives had slightly higher af-
finity at the R1-, R2-, R3-, and R5-subtypes compared
to the thiomethyl analogues. Similarly, the homologous
C3-(2-hydroxypropylthio) group (e.g., 13 and 44) has no
detrimental effect on R5 affinity. All the analogues that
we examined had higher affinity for the R5-subtype than
for any of the other GABAA receptor subtypes. In
particular, the C1-(2-thiazolyl) series maintained at
least 1 order of magnitude selectivity for the R5 receptor
over its next nearest subtype, and in the case of the

2230 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Chambers et al.
Sch em e 3a
R5 binding selectivity over the R1-, R2-, and R3-
subtypes, 43 had relatively low affinity for the R4â3γ2
receptors (Ki 106 nM) and was essentially inactive (Ki
1800 nM) at the R6â3γ2-subtype. Furthermore, when
examined in 96 radioligand binding and enzyme as-
says,32 the only off-target activity which was <1 µM was
at the rat adenosine A1 receptor (Ki 29 nM).
In Vivo Ch a r a cter iza tion of 43. To assess the
occupancy of 43 at the GABAA receptor subtypes in the
rat, an in vivo radioligand binding assay33 was used.
Using [3H]Ro 15-1788 as the radioligand, when admin-
istered at 3 mg/kg ip in the rat the occupancy of GABAA
receptors by 43 was 20% (Figure 2a). Ro 15-1788 is a
nonselective GABAA ligand which binds equally to all
the diazepam-sensitive GABAA receptor subtypes, and
therefore the relative contribution of the R5-subtype to
the total Bz binding site GABAA receptor population
(i.e., combined R1 + R2 + R3 + R5-subtypes) measured
using [3H]Ro 15-1788 is relatively small (∼5%).3 In this
experiment, the major contribution to the GABAA recep-
tor occupancy is from the R1 + R2 + R3-subtypes. When
[3H]L-655,708 is used as the radioligand, the occupancy
in the rat at 3 mg/kg ip was 80% (Figure 2b). L-655,708
is a GABAA R5 binding selective ligand21 and therefore,
assuming that the human and rat in vitro binding data
are comparable, the in vivo binding assay is essentially
measuring occupancy of only R5-containing receptors.
These data indicates that not only is thiazole 43 brain
penetrant in the rat with high occupancy of R5 binding
sites at 3 mg/kg ip, but that the R5 binding selectivity
observed in vitro, is confirmed in vivo.
a
Reagents: (i) Dioxane, 1,1′-carbonyldiimidazole; (ii) acetamide
oxime, 100 °C.
thiazolylthio analogue 45 R5-subtype selectivity was
17-28-fold over the other GABAA receptor subtypes.
In Vitr o Effica cy. Of those compounds which dem-
onstrated good R5 binding affinity, a number was
selected and the efficacy at the human GABAA R5
receptor subtype determined using two-electrode voltage
clamp recording from Xenopus laevis oocytes, which
transiently expressed the GABAA R5â3γ2 receptor
subtype.27,28 The R5 efficacy results obtained for those
compounds tested are shown in Tables 1 and 2. In this
assay, the control compound 2, which is regarded as a
nonselective, full BZ inverse agonist,29,30 produced an
efficacy reading of -34%. All three pyrazoles (5, 14, 20)
are essentially antagonists at GABAA R5 receptors
whereas the pyridines (26-28, 34) are all R5 inverse
agonists. A comparison of the methylthio analogues 26-
28 reveals that the nature of the pyridyl isomer at C1
influences the level of R5 inverse agonism, with efficacy
values ranging from -20% for the 4-pyridyl derivative
28 to -40% for the 3-pyridyl isomer 27. In the thiazole
series the C3 substituent has a significant effect on the
R5 efficacy. Whereas the methylthio (39) and hydroxy-
ethylthio (43) analogues are inverse agonists at the
GABAA R5 receptor (GABAA R5 efficacy: -24% and
-38%, respectively), the hydroxypropylthio (44) and
thiazolylthio (45) derivatives are a partial agonist
(GABAA R5 efficacy: +25%) and an antagonist (GABAA
R5 efficacy: -5%), respectively. It is noteworthy that the
introduction of a single methylene unit into the hy-
droxyethylthio substituent of 43 converts a full BZ
inverse agonist into a partial agonist, thus demonstrat-
ing the difficulty of obtaining a robust structure-
efficacy relationship for this series of compounds.
F u r th er In Vitr o Ch a r a cter iza tion of 43. The
efficacy profile of 43 was examined in more detail using
a whole cell patch clamp technique. This procedure
differs from the two-electrode voltage clamp, in which
compounds were assayed at a single concentration in
transiently infected Xenopus laevis oocytes, in that
human GABAA receptors are stably expressed and
efficacy is determined over a wider concentration range.
As shown in Table 3 and Figure 1, using whole cell
patch-clamp recording from mammalian fibroblast L(tk-)
cells,31 stably expressing the human GABAA Rxâ3γ2
receptors (x ) 1, 2, 3, 5), in the presence of a submaxi-
mal dose (EC20) of GABA, 43 has similar efficacy at the
GABAA R5â3γ2 receptor (-51%) compared to the full
nonselective inverse agonist 2 (-57%). The EC50 value
of 43 at the R5-subtype is 0.7 nM, which complements
its binding affinity of 1.6 nM. It is a low efficacy inverse
agonist at the R1-subtype (-21%) and an antagonist at
R2 (-1%) and R3 (-3%). In addition to being a function-
ally selective, full R5 inverse agonist with 10-13-fold
The proconvulsant potential of 43 was determined in
mice. At 3 mg/kg ip in the mouse, 91% of GABAA R5-
containing receptors ([3H]L-655,708 binding) were oc-
cupied, and, at this dose, 43 had no proconvulsant or
convulsant effects (Figure 3).34 This contrasts with the
significant convulsant effects reported in mice for the
nonselective BZ inverse agonist 2.12
Having found no significant off-target activities with
43 and having shown that 43 is a functionally selective
GABAA R5 inverse agonist with good occupancy of R5-
containing receptors and relatively poor occupancy of
R1, R2 and R3-containing receptors, 43 was selected as
an appropriate probe to test the hypothesis that a
selective GABAA R5 receptor inverse agonist may en-
hance cognition.
To investigate the effect that 43 has on learning and
memory, we assessed the performance of rats dosed with
43 in the water maze, a hippocampal-dependent cogni-
tive test.35 The delayed ‘matching-to-place’ variant of
the water maze was used in which the position of the
submerged platform in a 2 m diameter pool varied each
day, but the location remained constant for the duration
of that day.36 The rat in a given test was required to
find the platform four times each day. On the first trial
of each day the rat must search the pool for the hidden
platform. If it locates the platform more rapidly on trial
2, this indicates that the rat has remembered the
platform position from trial 1. Around the pool are
spatial cues to assist the rat in navigating its way to
the platform. Thus, the improvement in memory can be
quantified by calculating the difference between the
time taken to reach the platform on trial 1 compared
with subsequent trials. As can be seen in Figure 4, rats

Selective GABAA R5 Receptor Inverse Agonist
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2231
Sch em e 4a
a
Reagents: (i) H2S, Et3N, pyridine, (ii) chloroacetaldehyde, EtOH, reflux; (iii) DCM, m-CPBA (2.0 equiv), 25 °C; (iv) RSH, NaOH,
EtOH; (v) EtOH, HCl; (vi) EtOH, MeNHNH2, 50 °C; (vii) HCO2H, 100 °C.
dosed with 43 at 0.3 mg/kg ip showed a significant
improvement in performance between trial 1 and trial
2 compared with vehicle-treated animals (p <0.05). At
this dose the occupancy of GABAA R5 receptors in the
rat ([3H]L-655,708 binding) was 40%. Since 43 was
without effect on swim speed, these data indicate that
the functionally selective GABAA R5 inverse agonist 43
significantly enhanced performance in this hippocam-
pal-dependent memory task.
These results, therefore, clearly provide pharmaco-
logical evidence for the involvement of GABAA R5
receptors in cognitive processes and complements the
recent molecular genetic evidence provided by Collinson
et al.37 and Crestani et al.38 who used R5 ‘knock-out’
and ‘knock-in’ mice, respectively, to demonstrate a role
for the GABAA R5-subtype in cognitive processing.
agonist which improves cognitive performance in a
hippocampal-dependent memory task, without the con-
vulsant and proconvulsant side effects associated with
nonselective full GABAA receptor inverse agonists.
Exp er im en ta l Section
Gen er a l. All reactions were carried out under a nitrogen
atmosphere, using commercially available anhydrous solvents.
Thin-layer chromatography was performed on glass-backed
precoated Merck silica gel (60 F254) plates, and flash chroma-
tography was carried out using 40-63 µm silica gel. Proton
NMR spectra were measured on a Bruker AC250, Bruker
DPX360, or Bruker DPX400 spectrometer in the solvent
specified. Chemical shifts are measured in ppm downfield from
TMS as an internal standard, and coupling constants are
measured in hertz (Hz). Mass spectra were recorded on VG
Quattro spectrometer using positive ionization electrospray
(ES+). High-resolution mass spectra (HRMS) were obtained
at MSD using a micromass Q-Tof spectrometer using electro-
spray (ES+) ionization or by M-Scan Ltd., Sunninghill, Ascot,
Berkshire, UK. Combustion analyses were conducted by
Butterworth Laboratories, Teddington, Middlesex, U.K. Melt-
ing points were determined on a Reichert hot stage apparatus
and are uncorrected. 6,6-Dimethyl-3-methylthio-1-(pyrazol-3-
yl)-6,7-dihydro-2-benzothiophen-4(5H)-one (5), 1-bromo-6,6-
dimethyl-3-methanesulfonyl-6,7-dihydro-2-benzothiophen-4-
(5H)-one (24), 6,6-dimethyl-3-methylthio-6,7-dihydro-2-ben-
zothiophen-4(5H)-one-1-carboxylic acid (35) and 6,6-dimethyl-
3-methylthio-4-oxo-4,5,6,7-tetrahydro-2-benzothiophene-1-car-
bonitrile (37) were obtained from Maybridge Chemical Co.
Ltd., Trevillett, Tintagel, Cornwall, U.K. High performance
Con clu sion
Using a directed screening approach, a series of 6,7-
dihydro-2-benzothiophen-4(5H)-ones has been identified
as a novel class of GABAA receptor ligands which have
binding selectivity for the GABAA R5 receptor subtype
over the R1-, R2-, and R3-subtypes. Several examples
have high inverse agonism at the GABAA R5 receptor.
In particular, 6,6-dimethyl-3-(2-hydroxyethyl)thio-1-
(thiazol-2-yl)-6,7-dihydro-2-benzothiophen-4(5H)-one (43)
has been identified as a high affinity, brain penetrant,
functionally selective, full GABAA R5 receptor inverse

2232 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Chambers et al.
Ta ble 1. Binding Affinity and Efficacy of the C3-Thio-substituted Thiophenes
a
Displacement of [3H]Ro 15-1788 binding from recombinant human GABAA receptor subtypes. Ki values are the geometric mean (
b
SEM of three independent determinations. Efficacy is determined as the percentage modulation of the submaximal (EC20) response to
GABA. Values given are the arithmetic mean ( SEM of at least three individual cells from the human R5â3γ2s receptor subtype transiently
d
expressed in Xenopus laevis oocytes. c Binding selectivity for GABAA R5 receptors over GABA-A R1, R2, and R3 receptors. Efficacy is the
arithmetic mean ( SEM of eleven individual cells from the human R5â3γ2 receptor subtype transiently expressed in Xenopus laevis
oocytes.
liquid chromatography (HPLC) was carried out on an Agilent
1100 series machine using a Hichrom KR100-5C8 (250 × 4.6
mm i.d.) column or an Ace 3C8 (150 × 4.6 mm i.d.) column. A
flow rate of 1 mL/min was used with an injection volume of 5
µL and detection was by UV at λ254 nm.
MHz, CDCl3) δ 1.07 (6H, s), 1.48 (3H, t, J ) 7.4 Hz), 2.43 (2H,
s), 2.85 (2H, s), 3.10 (2H, q, J ) 7.3 Hz), 6.49 (1H, br s,), 7.65
(1H, d, J ) 2.3 Hz).
6,6-Dim et h yl-3-(1,1-d im et h ylet h yl)t h io-1-(p yr a zol-3-
yl)-6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (11). The title
compound was prepared in a simlar manner to that described
for 7, using sodium 2-methyl-2-propanethiolate. Purification
was achieved by column chromatography on silica gel, eluting
with petrol (60/80):EtOAc (1:1), to afford 11 as a yellow solid.
Yield 48%. mp 220-222 °C. Anal. (C17H22N2OS2) C, H, N.
HRMS (ES+) Observed: 335.1260, Calculated C17H23N2OS2 [M
6,6-Dim et h yl-3-m et h a n esu lfon yl-1-(p yr a zol-3-yl)-6,7-
d ih yd r o-2-ben zoth iop h en -4(5H)-on e (6). To a stirred solu-
tion of 6,6-dimethyl-3-methylthio-1-(pyrazol-3-yl)-6,7-dihydro-
2-benzothiophen-4(5H)-one (5) (3.0 g, 10.2 mmol) in CH2Cl2:
dioxane (250 mL, 4:1) at -78 °C was added m-CPBA (5.1 g
(70% w/w), 20.5 mmol) portionwise. The mixture was stirred
at -78 °C for 30 min and then at room temperature for 1 h.
After this time more m-CPBA (1.4 g (70% w/w), 5.6 mmol) was
added and the mixture stirred at room-temperature overnight.
After this time the mixture was poured into NaHCO3 (sat., 50
mL) and the organic layer separated. The organic phase was
washed with more NaHCO3 (sat., 2 × 50 mL), dried (MgSO4),
and evaporated. The residue was triturated twice with diethyl
ether and the resultant colorless solid (3.1 g, 94%) collected
by filtration. mp 213-216 °C. Anal. (C14H16N2O3S2) C, H, N.
HRMS (ES+) Observed: 325.0665, Calculated C14H17N2O3S2
1
+ H]+: 335.1252. H NMR (360 MHz, DMSO-d6) δ 1.00 (6H,
s), 1.45 (9H, s), 2.39 (2H, s), 2.88 (2H, s), 6.54 (1H, br s), 7.87
(1H, br s), 13.07(1H, br s). MS (ES+) 335 (MH)+.
Gen er a l P r oced u r e for th e Syn th esis of th e 6,7-Dih y-
d r o-2-ben zoth iop h en -4(5H)-on es 8-10 a n d 14-15. To a
suspension of 6 (50 mg, 0.15 mmol) in EtOH (2 mL) was added
NaOH (77 µL of a 4 N solution, 0.31 mmol) followed by the
appropriate thiol (0.31 mmol). The mixture was stirred at room
temperature for 1 h (compounds 8-10) or heated at 70 °C for
72 h (compounds 14 and 15). The mixture was diluted with
MeOH:H2O:1 N HCl (5:5:1) and then poured onto a C-18 Bond
Elut cartridge (prewashed with MeOH followed by water). The
cartridge was eluted with MeOH:H2O (8 mL) (1:1 f 1:0), and
the product fractions were evaporated. The residue was
triturated with diethyl ether to give the thioether as a solid.
6,6-Dim eth yl-3-p r op ylth io-1-(p yr a zol-3-yl)-6,7-d ih yd r o-
2-ben zoth iop h en -4(5H)-on e (8). Yield 20%. mp 181-182 °C.
Anal. (C16H20N2OS2) C, H, N. MS (ES+) 321 [M + H]+. 1H NMR
(360 MHz, CDCl3) δ 1.07 (6H, s), 1.10 (3H, t, J ) 7.3 Hz), 1.86
(2H, sextet, J ) 7.3 Hz), 2.43 (2H, s), 2.85 (2H, s), 3.05 (2H, t,
J ) 7.3 Hz), 6.48 (1H, d, J ) 2.3 Hz), 7.64 (1H, d, J ) 2.3 Hz).
3-Bu tylth io-6,6-d im eth yl-1-(p yr a zol-3-yl)-6,7-d ih yd r o-
2-ben zoth iop h en -4(5H)-on e (9). Yield 49%. mp 149-150 °C.
Anal. (C17H22N2OS2) C, H, N. MS (ES+) 335 [M + H]+. 1H NMR
1
[M + H]+: 325.0665. H NMR (360 MHz, CDCl3) δ 1.12 (6H,
s), 2.55 (2H, s), 2.98 (2H, s), 3.55 (3H, s), 6.58 (1H, d, J ) 2.6
Hz), 7.69 (1H, d, J ) 2.4 Hz).
6,6-Dim eth yl-3-eth ylth io-1-(p yr a zol-3-yl)-6,7-d ih yd r o-
2-ben zoth iop h en -4(5H)-on e (7). To a solution of 6 (100 mg,
3.1 mmol) in THF (10 mL) was added sodium ethanethiolate
(52 mg, 6.2 mmol) and the mixture stirred at room tempera-
ture for 15 min. After this time water (20 mL) and EtOAc (25
mL) were added and the organic layer separated and dried
(Na2SO4). The solvent was evaporated and the resultant solid
tritrated with diethyl ether to afford the title compound (60
mg, 63%) as a pale yellow solid. mp 182-184 °C. Anal.
(C15H18N2OS2) C, H, N. HRMS (ES+) Observed: 307.0938,
Calculated C15H19N2OS2 [M + H]+: 307.0939. 1H NMR (360

Selective GABAA R5 Receptor Inverse Agonist
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2233
Ta ble 2. Binding Affinity and Efficacy of the C1-Heteroaryl-thiophenes
a
Displacement of [3H]Ro 15-1788 binding from recombinant human GABAA receptor subtypes. Ki values are the geometric mean (
b
SEM of three independent determinations. Efficacy is determined as the percentage modulation of the submaximal (EC20) response to
GABA. Values given are the arithmetic mean ( SEM of at least three individual cells from the human R5â3γ2s receptor subtype transiently
expressed in Xenopus laevis oocytes. c Binding selectivity for GABAA R5 receptors over GABAA R1, R2, and R3 receptors.
(360 MHz, CDCl3) δ 0.96 (3H, t, J ) 7.4 Hz), 1.07 (6H, s), 1.52
(2H, sextet, J ) 7.4 Hz), 1.82 (2H, pentet, J ) 7.6 Hz), 2.43
(2H, s), 2.85 (2H, s), 3.08 (2H, t, J ) 7.4 Hz), 6.48 (1H, d, J )
2.3 Hz), 7.64 (1H, d, J ) 2.3 Hz).
6,6-Dim et h yl-1-(p yr a zol-3-yl)-3-(t h ia zol-2-yl)t h io-6,7-
d ih yd r o-2-ben zoth iop h en -4(5H)-on e (14). Yield 13%. mp
195-198 °C. Anal. (C16H15N3OS3) C, H, N. HRMS (ES+)
Observed: 362.0442, Calculated C16H16N3OS3 [M + H]+:
1
6,6-Dim et h yl-3-(1-m et h ylet h yl)t h io-1-(p yr a zol-3-yl)-
6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (10). Yield 50%.
mp 195-198 °C. Anal. (C16H20N2OS2‚0.1(H2O)) C, H, N. HRMS
(ES+) Observed: 321.1086, Calculated C16H21N2OS2 [M + H]+:
362.0456. H NMR (360 MHz, DMSO-d6) δ 1.03 (6H, s), 2.47
(2H, s), 2.87 (2H, s), 6.48-6.50 (1H, m), 7.83-7.85 (1H, m),
8.04-8.08 (2H, m), 13.03 (1H, br s).
6,6-Dim eth yl-1-(p yr a zol-3-yl)-3-(th ien -2-yl)th io-6,7-d i-
h yd r o-2-ben zoth iop h en -4(5H)-on e (15). Yield 47%. mp
226-228 °C. Anal. (C17H16N2OS3‚0.1(H2O)) C, H, N. MS (ES+)
1
321.1095. H NMR (360 MHz, DMSO-d6) δ 0.99 (6H, s), 1.42
(6H, d, J ) 6.7 Hz), 2.37 (2H, s), 2.84 (2H, s), 3.50-3.58 (1H,
m), 6.50 (1H, d, J ) 2.2 Hz), 7.85 (1H, br s), 13.02 (1H, br s).
1
361 [M + H]+. H NMR (360 MHz, CDCl3) δ 1.09 (6H, s), 2.46

2234 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Chambers et al.
Ta ble 3. Efficacy of 2 and 43 in L(tk-) Cells Expressing
Different Human GABAA Receptor Subtypes
mp 196-198 °C. Anal. (C16H20N2O2S‚0.3(H2O)) C, H, N. MS
1
(ES+) 305 [M + H]+. H NMR (360 MHz, CDCl3) δ 1.06 (6H,
s), 1.51 (6H, d, J ) 6.1 Hz). 2.36 (2H, s), 2.80 (2H, s), 4.55-
4.62 (1H, m), 6.45 (1H, br s), 7.62 (1H, d, J ) 2.3 Hz).
efficacy at human GABAA Rxâ3γ2 receptors (%)a
compound
R1
R2
R3
R5
6,6-Dim eth yl-3-m eth yla m in o-1-(p yr a zol-3-yl)-6,7-d ih y-
d r o-2-ben zoth iop h en -4(5H)-on e (19). To 16 (116 mg, 0.38
mmol) was added a solution of methylamine in MeOH (33%
(w/v); 10 mL, 80 mmol). The soution was heated at 100 °C for
2 h in a sealed tube then cooled to room temperature. The
solvent was evaporated and the residue chromatographed on
silica gel, eluting with CH2Cl2:MeOH (95:5) to give the title
compound (92 mg, 88%) as an orange solid. mp 240-243 °C.
Anal. (C14H17N3OS‚0.3(H2O)) C, H, N. MS (ES+) 276 [M + H]+.
1H NMR (360 MHz, CDCl3) δ 1.06 (6H, s), 2.32 (2H, s), 2.72
(2H, s), 3.06 (3H, d, J ) 5.2 Hz), 6.41 (1H, br s), 7.60 (1H, d,
J ) 2.3 Hz), 8.58 (1H, br s).
DMCM(2)
43
-71 ( 2
-53 ( 3
-1 ( 2b
-62 ( 2
-3 ( 1b
-57 ( 1
-51 ( 2
-21 ( 2b
a
Maximum modulation of the current produced by DMCM or
43 relative to a submaximal (EC20) GABA response. Values are
the mean maximum modulation ( SEM from at least five
individually fitted concentration-response curves. b Values are the
arithmetic mean ( SEM of at least five individual cells, produced
using [43] ) 100 nM from human Rxâ3γ2 (x ) 1, 2, or 3) receptor
subtypes stably expressed in mammalian fibroblast L(tk-) cells.
(2H, s), 2.83 (2H, s), 6.40 (1H, d, J ) 2.3 Hz), 7.13 (1H, dd, J
) 5.3 and 3.6 Hz), 7.43 (1H, dd, J ) 3.6 Hz), 7.57-7.61 (2H,
m).
6,6-Dim eth yl-3-isop r op yla m in o-1-(p yr a zol-3-yl)-6,7-d i-
h yd r o-2-ben zoth iop h en -4(5H)-on e (20). In the same way
as described for 19 using isopropylamine in EtOH and heating
for 24 h the title amine (57 mg, 58%) was isolated as an orange
solid. mp 215-218 °C. Anal. (C16H21N3OS) C, H, N. MS (ES+)
6,6-Dim et h yl-3-(2-h yd r oxyet h yl)t h io-1-(p yr a zol-3-yl)-
6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (12). To a suspen-
sion of 6 (100 mg, 0.31 mmol) in EtOH (5 mL) were added
2-mercaptoethanol (27 µL, 0.39 mmol) and NaOH (82 µL, 0.33
mmol). The mixture was stirred for 90 min then partitioned
between water (20 mL), HCl (1 N, 1 mL) and EtOAc (20 mL).
The organic phase was separated, dried (Na2SO4), and evapo-
rated. The residue was triturated with diethyl ether to afford
the title compound (96 mg, 96%) as a pale yellow solid. mp
182-185 °C. Anal. (C15H18N2O2S2‚0.3(H2O)) C, H, N. HRMS
(ES+) Observed: 323.0885, Calculated C15H19N2O2S2 [M + H]+:
1
308 [M + H]+. H NMR (360 MHz, CDCl3) δ 1.06 (6H, s), 1.35
(6H, d, J ) 6.3 Hz), 2.32 (2H, s), 2.71 (2H, s), 3.50-3.60 (1H,
m), 6.40 (1H, d, J ) 2.2 Hz), 7.60 (1H, d, J ) 2.4 Hz), 8.64-
8.70 (1H, m).
6,6-Dim et h yl-3-(2-h yd r oxyet h yl)a m in o-1-(p yr a zol-3-
yl)-6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (21). To a so-
lution of 16 (100 mg, 0.33 mmol) in EtOH (3 mL) was added
ethanolamine (1.0 mL, 16 mmol). The soution was heated at
100 °C for 2.5 h then cooled to room temperature and
partitioned between water (40 mL) and EtOAc (3 × 25 mL).
The combined organic layers were dried (Na2SO4) and evapo-
rated. The residue was chromatographed on silica gel, eluting
with CH2Cl2:MeOH (90:10) to give the title compound (57 mg,
58%) as a pale yellow solid. mp 227-230 °C. Anal. (C15H19N3O2-
S.0.2(H2O)) C, H, N. MS (ES+) 306 [M + H]+. 1H NMR (360
MHz, DMSO-d6) δ 0.99 (6H, s), 2.24 (2H, s), 2.68 (2H, s), 3.28-
3.34 (2H, m), 3.60-3.65 (2H, m), 6.36 (1H, d, J ) 2.2 Hz), 7.74
(1H, d, J ) 2.3 Hz), 8.66-8.76 (1H, m).
1
323.0888. H NMR (360 MHz, DMSO-d6) δ 1.00 (6H, s), 2.37
(2H, s), 2.83 (2H, s), 3.16 (2H, t, J ) 6.4 Hz), 3.70-3.80 (2H,
m), 5.08-5.20(1H, m), 6.49 (1H, d, J ) 3.3 Hz), 7.84 (1H, br
s), 13.04 (1H, br s).
6,6-Dim eth yl-3-(3-h yd r oxyp r op yl)th io-1-(p yr a zol-3-yl)-
6,7-d ih yd r o-2-ben zot h iop h en -4(5H )-on e (13). The title
compound was prepared in a simlar manner to that described
for 12, using 3-mercapto-1-propanol. Yield 91%. mp 172-175
°C. Anal. (C16H20N2O2S2) C, H, N. MS (ES+) 337 [M + H]+. 1H
NMR (360 MHz, DMSO-d6) δ 1.00 (6H, s), 1.88 (2H, pentet, J
) 6.9 Hz), 2.37 (2H, s), 2.84 (2H, s), 3.09 (2H, t, J ) 7.2 Hz),
3.51-3.56 (2H, m), 4.66 (1H, t, J ) 5.3 Hz), 6.49 (1H, br s),
7.84 (1H, br s), 13.00 (1H, br s).
6,6-Dim eth yl-3-m eth a n esu lfin yl-1-(p yr a zol-3-yl)-6,7-d i-
h yd r o-2-ben zoth iop h en -4(5H)-on e (16). To a stirred solu-
tion of 5 (200 mg, 0.69 mmol) in CH2Cl2:dioxane (9 mL, 3:1)
at -78 °C was added m-CPBA (169 mg (70% w/w), 0.69 mmol)
portionwise. After addition the mixture was stirred at -78 °C
for 1 h, then diluted with CH2Cl2 (10 mL) and poured into
NaHCO3 (sat., 10 mL). The organic layer was separated, dried
(Na2SO4), and evaporated. The residue was triturated with
diethyl ether and the resultant colorless solid (106 mg, 50%)
collected by filtration. mp 206-210 °C. HRMS (ES+) Ob-
served: 308.0660, Calculated C14H16N2O2S2 [M]+: 308.0653.
1H NMR (360 MHz, CDCl3) δ 1.09 (3H, s), 1.11 (3H, s), 2.42
(1H, d, J ) 17 Hz), 2.52 (1H, d, J ) 17 Hz), 2.95 (2H, s), 3.01
(3H, s), 6.54 (1H, s), 7.68 (1H, s).
6,6-Dim et h yl-3-m et h oxy-1-(p yr a zol-3-yl)-6,7-d ih yd r o-
2-ben zoth iop h en -4(5H)-on e (17). To a solution of 16 (80 mg,
0.26 mmol) in MeOH (3 mL) was added a solution of sodium
methoxide in MeOH (1.0 mL of a 0.5 M solution, 0.5 mmol).
The soution was heated at 70 °C for 24 h then cooled to room
temperature and partitioned between water (20 mL) and
EtOAc (2 × 20 mL). The combined organic layers were dried
(Na2SO4) and evaporated. The residue was chromatographed
on silica gel, eluting with CH2Cl2:MeOH (95:5) to give the title
compound (11 mg, 15%) as a pale yellow solid. mp 202-204
°C. Anal. (C14H16N2O2S.0.1(H2O)) C, H, N. MS (ES+) 277 [M
+ H]+. 1H NMR (360 MHz, DMSO-d6) δ 0.98 (6H, s), 2.27 (2H,
s), 2.79 (2H, s), 4.03 (3H, s), 6.54 (1H, br s), 7.81 (1H, br s),
12.92 (1H, br s).
6,6-Dim eth yl-3-eth yl-1-(pyr azol-3-yl)-6,7-dih ydr o-2-ben -
zoth iop h en -4(5H)-on e (22) a n d 6,6-Dim eth yl-1-(p yr a zol-
3-yl)-6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (23). To a
stirred solution of 16 (122 mg, 0.4 mmol) in THF at -10 °C
was added ethylmagnesium bromide (0.79 mL of a 1.0 M
solution in THF, 0.79 mmol). The mixture was stirred at -10
°C for 1 h then NH4Cl (sat., 1 mL) was added. The cooling
bath was removed and the mixture stirred at room tempera-
ture for 10 min. The mixture was partitioned between EtOAc
(15 mL) and water (15 mL). The organic layer was separated,
dried (Na2SO4), and evaporated. The residue was chromato-
graphed on silica gel, eluting with EtOAc:CH2Cl2 (6:1). The
fractions containing 6,6-dimethyl-3-ethyl-1-(pyrazol-3-yl)-6,7-
dihydro-2-benzothiophen-4(5H)-one (22) (less polar spot on
TLC) were combined and evaporated, and the residue was
triturated with Et2O. 22 was isolated as a pale yellow solid (6
mg, 6%). mp 152-155 °C. HRMS (ES+) Observed: 274.1141,
1
Calculated C15H18N2OS [M]+: 274.1140. H NMR (360 MHz,
CDCl3) δ 1.07 (6H, s), 1.34 (3H, t, J ) 7.4 Hz), 2.42 (2H, s),
2.87 (2H, s), 3.30 (2H, q, J ) 7.4 Hz), 6.49 (1H, d, J ) 2.2 Hz),
7.65 (1H, d, J ) 2.2 Hz). HPLC (50% CH3CN; 50% 25mM aq
KH2PO4 (containing 0.2% diethylamine)): 99% purity. Reten-
tion time: 4.96 min.
The fractions containing 6,6-dimethyl-1-(pyrazol-3-yl)-6,7-
dihydro-2-benzothiophen-4(5H)-one (23) (more polar spot on
TLC) were combined and evaporated, and the residue was
triturated with Et2O. 23 was isolated as a colorless solid (15
mg, 15%). mp 160-163 °C. HRMS (ES+) Observed: 247.0898,
Calculated C13H15N2OS [M + H]+: 247.0905. 1H NMR (250
MHz, CDCl3) δ 1.09 (6H, s), 2.46 (2H, s), 2.92 (2H, s), 6.54
(1H, d, J ) 2.4 Hz), 7.67 (1H, d, J ) 2.4 Hz), 8.11 (1H, s).
HPLC (40% CH3CN; 60% 25 mM aq KH2PO4 (containing 0.2%
diethylamine)): 99% purity. Retention time: 6.10 min.
6,6-Dim eth yl-3-isopr opoxy-1-(pyr azol-3-yl)-6,7-dih ydr o-
2-ben zoth iop h en -4(5H)-on e (18). In the same way as de-
scribed for 17 using sodium isopropoxide in 2-propanol and 6
the title compound (40 mg, 43%) was isolated as a yellow solid.
1-Br om o-6,6-d im eth yl-3-m eth ylth io-6,7-d ih yd r o-2-ben -
zoth iop h en -4(5H)-on e (25). To a solution of 1-bromo-6,6-

Selective GABAA R5 Receptor Inverse Agonist
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2235
F igu r e 1. Efficacy of 43 and DMCM in mammalian fibroblast L(tk-) cells expressing human GABAA Rxâ3γ2 (x ) 1, 2, 3, 5)
receptors
F igu r e 2. Occupancy of 43 at rat brain Bz sites. (a) Occupancy of the combined (i.e., R1 + R2 + R3 + R5-containing) GABAA
receptor population measured using [3H]Ro 15-1788. (b) Occupancy at only R5-containing GABAA receptors measured using
[3H]L-655,708. Rats (n ) 5-7/group) were dosed ip with either vehicle (70% PEG 300) and 0.5 h later receptor occupancy measured.
Values shown are mean ( SEM.
F igu r e 3. Lack of effect of 43 on the dose of pentylenetetrazole (PTZ) required to produce (a) clonic or (b) tonic seizures in mice.
Mice (n ) 11-12/group) were dosed ip with either vehicle (70% PEG 300) or 1 or 3 mg/kg of 43. Thirty minutes later PTZ (15
mg/kg) was infused at a rate of 0.2 mg/min, and the time (and therefore by calculation the dose) at which clonic and tonic seizure
activity occurred was recorded. Values shown are mean ( SEM. In a separate experiment (data not shown), a dose of 3 mg/kg of
43 corresponds to an occupancy of R5-containing GABAA receptors (measured using [3H]L-655,708) of 91 ( 4%.
dimethyl-3-methanesulfonyl-6,7-dihydro-2-benzothiophen-4(5H)-
one (24) (2 g, 5.9 mmol) in THF (70 mL) was added sodium
methanethiolate (0.9 g, 13 mmol) portionwise. The mixture
was stirred at room temperature for 72 h then partitioned
between EtOAc (100 mL) and water (100 mL). The organic
layer was separated, dried (MgSO4), and evaporated. The
residue was triturated with isohexane and the title compound
(1.1 g, 58%) isolated as a yellow solid. mp 114-116 °C. Anal.
(C11H13BrOS2) C, H. HRMS (ES+) Observed (79Br): 304.9673,
1
Calculated C11H14BrOS2 [M (79Br) + H]+: 304.9669. H NMR
(360 MHz, CDCl3) δ 1.00 (6H, s), 2.37 (2H, s), 2.54 (2H, s),
2.56 (3H, s).
6,6-Dim eth yl-3-m eth ylth io-1-(pyr idin -2-yl)-6,7-dih ydr o-
2-ben zoth iop h en -4(5H)-on e (26). A solution of 25 (700 mg,
2.3 mmol) and 2-(tri-n-butylstannyl)pyridine (1.27 g, 3.45
mmol) in dioxane (70 mL) was degassed with nitrogen for 15

2236 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Chambers et al.
°C. Anal. (C16H17NOS2‚0.1(H2O)) C, H, N. HRMS (ES+) Ob-
served: 304.0830, Calculated C16H18NOS2 [M + H]+: 304.0830.
1H NMR (360 MHz, CDCl3) δ 1.05 (6H, s), 2.45 (2H, s), 2.64
(3H, s), 2.83 (2H, s), 7.33 (2H, dd, J ) 4.6, 1.6 Hz), 8.83 (2H,
dd, J ) 4.6, 1.6 Hz).
6,6-Dim eth yl-3-m eth ylth io-1-(th ia zol-5-yl)-6,7-d ih yd r o-
2-ben zoth iop h en -4(5H)-on e (29). In the same way as de-
scribed for 26 using 5-(tri-n-butylstannyl)thiazole and dichlo-
robis(triphenylphosphine)palladium, the title compound (20
mg, 20%) was isolated as a yellow solid. mp 138-140 °C.
HRMS (ES+) Observed: 309.0318, Calculated C14H15NOS3
[M]+: 309.0316. 1H NMR (250 MHz, CDCl3) δ 1.07 (6H, s),
2.44 (2H, s), 2.62 (3H, s), 2.75 (2H, s), 7.90 (1H, s), 8.80 (1H,
s). HPLC (55% CH3CN; 45% 25mM aq KH2PO4 (containing
0.2% diethylamine)): >99% purity. Retention time: 7.71 min.
6,6-Dim eth yl-3-m eth ylth io-1-ph en yl-6,7-dih yd r o-2-ben -
zoth iop h en -4(5H)-on e (30). A solution of 25 (400 mg, 1.3
mmol), benzeneboronic acid (190 mg, 1.5 mmol), Na2CO3 (1
N, 20 mL), and tetrakis(triphenylphosphine)palladium(0) (300
mg, 0.26 mmol) in toluene (20 mL) and EtOH (10 mL) was
heated at reflux for 18 h. The mixture was cooled to room
temperature, evaporated, and partitioned between HCl (1 N)
and EtOAc. The organic phase was separated, washed with
water and K2CO3 (sat.), and then dried (MgSO4). The solvent
was evaporated then chromatographed on silica gel, eluting
with hexane:EtOAc (9:1 f 4:1) to afford the title compound
(130 mg, 33%) as a colorless solid. mp 143-145 °C. Anal.
(C17H18OS2) C, H. HRMS (ES+) Observed: 303.0882, Calcu-
lated C17H19OS2 [M + H]+: 303.0877. 1H NMR (250 MHz,
CDCl3) δ 1.02 (6H, s), 2.43 (2H, s), 2.62 (3H, s), 2.77 (2H, s),
7.42-7.44 (5H, m).
6,6-Dim eth yl-3-m eth ylth io-1-(p yr r ol-2-yl)-6,7-d ih yd r o-
2-ben zoth iop h en -4(5H)-on e (32). In the same way as de-
scribed for 28 using 1-(tert-butyloxycarbonyl)pyrrol-2-yl boronic
acid, 6,6-dimethyl-3-methylthio-1-(1-(tert-butyloxycarbonyl)-
pyrrol-2-yl)-6,7-dihydro-2-benzothiophen-4(5H)-one (31) (80
mg, 20%) was isolated as a pale yellow solid. mp 165-167 °C.
Anal. (C20H25NO3S2) C, H, N. MS (ES+) 392 [M + H]+. 1H NMR
(360 MHz, CDCl3) δ 1.00 (6H, s), 1.42 (9H, s), 2.37 (2H, s),
2.45 (2H, s), 2.57 (3H, s), 6.23-6.26 (2H, m), 7.41 (1H, s), 8.80
(1H, t, J ) 2.2 Hz).
31 (80 mg, 0.2 mmol) was dissolved in trifluoroacetic acid
(5 mL) and allowed to stand at 20 °C for 30 min. The solvent
was removed in vacuo and the residue dissolved in EtOAc. The
solution was washed with 2 N Na2CO3, water, and brine. The
organic layer was separated, dried (Na2SO4), and evaporated.
The residue was chromatographed on silica gel, eluting with
EtOAc:hexane (1:4), to afford the title pyrrole (35 mg, 59%)
as a yellow solid. mp 195-197 °C. Anal. (C15H17NOS2) C, H,
N. MS (ES+) 292 [M + H]+. 1H NMR (360 MHz, CDCl3) δ 1.04
(6H, s), 2.40 (2H, s), 2.59 (3H, s), 2.73 (2H, s), 6.30-6.33 (2H,
m), 6.85-6.88 (1H, m), 8.23 (1H, br s).
Figu r e 4. Enhanced performance of rats dosed with thiophene
43 in the delayed ‘matching-to-place’ water maze test. 43 was
administered intraperitoneally (ip) at a dose of 0.3 mg/kg in a
70% poly(ethylene glycol) 300 (PEG 300) aqueous solution.
Vehicle-treated animals were dosed ip with a 70% PEG 300
aqueous solution. Ten drug-treated and 10 vehicle-treated
animals were used in this test. The difference in time taken
between trial 1 and trial 2 (savings) to find the hidden platform
over a 10 day test period. Rats dosed with 43 made signifi-
cantly higher savings (*) compared with vehicle-treated ani-
mals.
min. Tetrakis(triphenylphosphine)palladium (0) (250 mg, 0.2
mmol) was added and the mixture heated at 100 °C for 18 h.
After this time the solvent was evaporated and the residue
partitioned between CH2Cl2 (20 mL) and water (20 mL). The
organic phase was separated and the aqueous layer washed
once more with CH2Cl2 (20 mL). The combined organic layers
were dried (MgSO4) and evaporated. The residue was tritu-
rated with diethyl ether and the resultant pale yellow solid
collected by filtration. This solid was then triturated with CH2-
Cl2 to afford the title compound (266 mg) as a colorless solid.
The filtrate was chromatographed on silica gel, eluting with
isohexane:EtOAc (2:1), to afford the title thiophene (137 mg)
as a colorless solid. Both batches of 26 were identical by 1H
NMR. Total mass ) 403 mg; 58% yield. mp 217-220 °C. Anal.
(C16H17NOS2) C, H, N. HRMS (ES+) Observed: 304.0831,
Calculated C16H18NOS2 [M + H]+: 304.0830. 1H NMR (360
MHz, CDCl3) δ 1.09 (6H, s), 2.45 (2H, s), 2.65 (3H, s), 2.93
(2H, s), 7.16-7.19 (1H, m), 7.47 (1H, d, J ) 8 Hz), 7.72-7.76
(1H, m), 8.60-8.61 (1H, m).
6,6-Dim eth yl-3-m eth ylth io-1-(pyr idin -3-yl)-6,7-dih ydr o-
2-ben zoth iop h en -4(5H)-on e (27). A solution of 25 (410 mg,
1.3 mmol) and 3-(tri-n-butylstannyl)pyridine (740 mg, 2.0
mmol) in dioxane (15 mL) was degassed with nitrogen for 30
min. Tetrakis(triphenylphosphine)palladium(0) (350 mg, 0.3
mmol) was added and the mixture heated at 100 °C for 10 h.
After this time the solvent was evaporated and the residue
partitioned between EtOAc (20 mL) and water (20 mL). The
organic phase was separated, dried (Na2SO4), and evaporated.
The residue was chromatographed on silica gel, eluting with
hexane:EtOAc (1:1), to afford the title compound (115 mg, 28%)
as a colorless solid. mp 227-230 °C. Anal. (C16H17NOS2‚0.1-
6,6-Dim eth yl-3-m eth a n esu lfin yl-1-(p yr id in -2-yl)-6,7-d i-
h yd r o-2-ben zoth iop h en -4(5H)-on e (33). 26 (250 mg, 0.83
mmol) was dissolved in CH2Cl2 (25 mL) and cooled to -78 °C.
A solution of m-CPBA (286 mg of 50-55% technical grade, 0.83
mmol) in CH2Cl2 (5 mL) was added and the mixture warmed
to -50 °C. After stirring for 1 h at -50 °C more CH2Cl2 (20
mL) was added and the mixture washed with NaHCO3 (sat.,
20 mL) and brine (20 mL). The organic layer was separated,
dried (Na2SO4), and evaporated. The residue was chromato-
graphed on silica gel, eluting with isohexane:EtOAc (1:1) f
CH2Cl2:MeOH (95:5), to afford the sulfoxide (209 mg, 79%) as
a colorless solid. mp 165-167 °C. HRMS (ES+) Observed:
320.0769, Calculated C16H18NO2S2 [M + H]+: 320.0779. 1H
NMR (400 MHz, CDCl3) δ 1.09 (3H, s), 1.12 (3H, s), 2.44 (1H,
d, J ) 18 Hz), 2.53 (1H, d, J ) 18 Hz), 3.00 (3H, s), 3.03-3.04
(2H, m), 7.25-7.28 (1H, m), 7.53-7.55 (1H, m), 7.77-7.81 (1H,
m), 8.66-8.68 (1H, m). HPLC (70% CH3CN; 30% H2O (con-
taining 0.1% TFA)): >99% purity. Retention time: 3.82 min.
(H2O)) C, H, N. HRMS (ES+) Observed: 304.0831, Calculated
1
C
16H18NOS2 [M + H]+: 304.0830. H NMR (250 MHz, CDCl3)
δ 1.04 (6H, s), 2.44 (2H, s), 2.63 (3H, s), 2.76 (2H, s), 7.34-
7.40 (1H, m), 7.70-7.75 (1H, m), 7.47 (1H, d, J ) 8 Hz), 8.56-
8.60 (1H, m), 8.69-8.70 (1H, m).
6,6-Dim eth yl-3-m eth ylth io-1-(pyr idin -4-yl)-6,7-dih ydr o-
2-ben zoth iop h en -4(5H)-on e (28). A solution of 25 (374 mg,
1.2 mmol), 4-pyridylboronic acid (605 mg, 4.9 mmol) and Na2-
CO3 (782 mg, 7.4 mmol) in ethylene glycol dimethyl ether (25
mL) and water (10 mL) was degassed with nitrogen for 40 min.
After this time tetrakis(triphenylphosphine)palladium (0) (300
mg, 0.26 mmol) was added and the mixture heated at reflux
for 5 h. More 4-pyridylboronic acid (100 mg, 0.8 mmol) was
added and heating continued for a further 3 h. The solution
was cooled to room temperature, and K2CO3 (sat., 10 mL) was
added to the solution. The mixture was extracted with EtOAc
(2 × 30 mL), and the combined organic layers were dried (Na2-
SO4) and evaporated. The residue was chromatographed on
silica gel, eluting with hexane:EtOAc (1:1) to afford the title
compound (164 mg, 44%) as a pale yellow solid. mp 230-233
6,6-Dim et h yl-3-(2-h yd r oxyet h yl)t h io-1-(p yr id in -2-yl)-
6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (34). To a suspen-
sion of 33 (150 mg, 0.47 mmol) in EtOH (5 mL) were added
2-mercaptoethanol (37 µL, 0.52 mmol) and NaOH (130 µL, 0.52

Selective GABAA R5 Receptor Inverse Agonist
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2237
mmol). The mixture was stirred at room temperature for 1 h
then the solvent evaporated. The residue was partitioned
between water (20 mL) and EtOAc (20 mL). The organic phase
was separated, dried (MgSO4), and evaporated. The residue
was chromatographed on silica gel, eluting with isohexane:
EtOAc (1:1 f 0:1), to afford the title compound (96 mg, 61%)
as a colorless solid. mp. 175-177 °C. Anal. (C17H19NO2S2) C,
rated K2CO3 solution. The organic layer was separated, dried
(MgSO4), and evaporated. The crude residue was chromato-
graphed on silica gel, eluting with EtOAc, to afford the triazole
(150 mg, 15%) as a yellow solid. mp 146-148 °C. Anal.
(C14H17N3OS2) C, H, N. MS (ES+) 308 [M + H]+. 1H NMR (250
MHz, CDCl3) δ 1.05 (6H, s), 2.44 (2H, s), 2.63 (3H, s), 2.74
(2H, s), 3.96 (3H, s), 7.98 (1H, s).
H, N. HRMS (ES+) Observed: 334.0919, Calculated C17H20
-
6,6-Dim eth yl-3-m eth a n esu lfon yl-1-(th ia zol-2-yl)-6,7-d i-
h yd r o-2-ben zoth iop h en -4(5H)-on e (42). To a stirred solu-
tion of 39 (1.44 g, 4.7 mmol) in CH2Cl2:dioxane (60 mL, 5:1)
was added m-CPBA (1.7 g (95% w/w), 9.3 mmol) at room
temperature. The mixture was stirred for 2 h then more
m-CPBA (0.42 g (95% w/w), 2.3 mmol) was added. The mixture
was stirred for a further 2 h then washed with Na2CO3 (sat.,
50 mL). The organic phase was separated, dried (Na2SO4), and
evaporated. The residue was chromatographed on silica gel,
eluting with isohexane:EtOAc (3:1 f 1:1) to afford the sulfone
NO2S2 [M + H]+: 334.0935. 1H NMR (400 MHz, CDCl3) δ 1.09
(6H, s), 2.46 (2H, s), 2.93 (2H, s), 3.34 (2H, t, J ) 6.0 Hz), 4.01
(2H, t, J ) 6.0 Hz), 7.17-7.20 (1H, m), 7.47-7.49 (1H, m),
7.75 (1H, dt, J ) 7.8, 1.9 Hz), 8.60-8.61 (1H, m).
6,6-Dim eth yl-3-m eth ylth io-1-(3-m eth yl-1,2,4-oxa d ia zol-
5-yl)-6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (36). To a
solution of 6,6-dimethyl-3-methylthio-4-oxo-4,5,6,7-tetrahydro-
2-benzothiophene-1-carboxylic acid (35) (1.0 g, 3.7 mmol) in
dioxane (90 mL) was added 1,1′-carbonyldiimidazole (0.6 g, 3.7
mmol) and the mixture stirred at 20 °C for 30 min. Acetamide
oxime (0.41 g, 5.5 mmol) was added and the mixture heated
at 100 °C for 48 h. The solution was evaporated to dryness
and the residue chromatigraphed on silica gel, eluting with
EtOAc:hexane (2:3), to yield the oxadiazole (0.1 g, 9%) as a
pale yellow solid. mp 177-180 °C. Anal. (C14H16N2O2S2) C, H,
N. MS (ES+) 309 [M + H]+. 1H NMR (250 MHz, CDCl3) δ 1.11
(6H, s), 2.45-2.47 (5H, s), 2.65 (3H, s), 3.11 (2H, s).
6,6-Dim eth yl-3-m eth ylth io-4-oxo-4,5,6,7-tetr a h yd r o-2-
ben zoth iop h en e-1-ca r both ioa m id e (38). H2S was bubbled
through a solution of 6,6-dimethyl-3-methylthio-4-oxo-4,5,6,7-
tetrahydro-2-benzothiophene-1-carbonitrile (37) (5.0 g, 20
mmol) and Et3N (3 mL) in pyridine (50 mL) for 30 min. After
this time the solution was allowed to stand at room temper-
ature overnight. Nitrogen was bubbled through the solution
which was subsequently diluted with CH2Cl2. The mixture was
washed with water (2×), HCl (2×) and brine. The organic layer
was separated, dried (Na2SO4), and evaporated. The title
compound (3.9 g, 68%) was isolated as a yellow solid and used
in the following reaction without further purification. mp 238-
240 °C. Anal. (C12H15NOS3) C, H, N. HRMS (ES+) Observed:
286.0412, Calculated C12H16NOS3 [M + H]+: 286.0394. 1H
NMR (360 MHz, DMSO-d6) δ 0.99 (6H, s), 2.34 (2H, s), 2.61
(3H, s), 2.93 (2H, s), 8.68 (1H, br s), 9.82 (1H, br s).
1
(1.33 g, 84%) as a colorless solid. H NMR (400 MHz, CDCl3)
δ 1.16 (6H, s), 2.59 (2H, s), 3.02 (2H, s), 3.56 (3H, s), 7.51 (1H,
d, J ) 3.3 Hz), 7.95 (1H, d, J ) 3.3 Hz).
6,6-Dim et h yl-3-(2-h yd r oxyet h yl)t h io-1-(t h ia zol-2-yl)-
6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (43). In the same
way as described for 12 using 42. Yield 91%. mp 175-177 °C.
Anal. (C15H17NO2S3‚0.2(H2O)) C, H, N. HRMS (ES+) Ob-
served: 340.0510, Calculated C15H18NO2S2 [M + H]+: 340.0500.
1H NMR (360 MHz, CDCl3) δ 1.12 (6H, s), 2.47 (2H, s), 2.90
(2H, s), 3.32 (2H, t, J ) 5.9 Hz), 4.01 (2H, t, J ) 5.9 Hz), 7.36
(1H, d, J ) 3.2 Hz), 7.82 (1H, d, J ) 3.2 Hz).
6,6-Dim eth yl-3-(3-h yd r oxyp r op yl)th io-1-(th ia zol-2-yl)-
6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (44). In the same
way as described for 43 using 3-mercapto-1-propanol. Yield
75%. mp 136-139 °C. Anal. (C16H19NO2S3) C, H, N. MS (ES+)
1
354 [M + H]+. H NMR (360 MHz, DMSO-d6) δ 1.05 (6H, s),
1.89 (2H, pentet, J ) 6.8 Hz), 2.44 (2H, s), 2.91 (2H, s), 3.15
(2H, t, J ) 7.2 Hz), 3.54 (2H, q, J ) 5.8 Hz), 4.69 (1H, t, J )
5.2 Hz), 7.82 (1H, d, J ) 3.3 Hz), 7.88 (1H, d, J ) 3.3 Hz).
6,6-Dim eth yl-1-(th ia zol-2-yl)-3-(th ia zol-2-yl)th io-6,7-d i-
h yd r o-2-ben zoth iop h en -4(5H)-on e (45). In the same way
as described for 14 using 42 and stirring at room temperature
for 20 h. Yield 88%. mp 200-202 °C. Anal. (C16H14N2OS4.0.3
1
(H2O)) C, H, N. MS (ES+) 379 [M + H]+. H NMR (360 MHz,
CDCl3) δ 1.14 (6H, s), 2.50 (2H, s), 2.91 (2H, s), 7.34 (1H, d, J
) 3.2 Hz), 7.57 (1H, d, J ) 3.5 Hz), 7.79 (1H, d, J ) 3.2 Hz),
8.00 (1H, d, J ) 3.3 Hz).
6,6-Dim eth yl-3-m eth ylth io-1-(th ia zol-2-yl)-6,7-d ih yd r o-
2-ben zoth iop h en -4(5H)-on e (39). A solution of 38 (500 mg,
1.7 mmol) in EtOH (5 mL) was heated with chloroacetaldehyde
(0.5 mL of a 50% (w/v) aqueous solution, 3.2 mmol) at reflux
for 18 h. After this time the solvent was evaporated and the
residue partitioned between between EtOAc (20 mL) and water
(20 mL). The organic layer was separated, dried (Na2SO4), and
evaporated. The residue was chromatographed on silica gel,
eluting with hexane:EtOAc (3:1) to afford the title thiazole
(350 mg, 65%) as a pale yellow solid. mp 149-151 °C. HRMS
(ES+) Observed: 310.0394, Calculated C14H16NOS3 [M + H]+:
310.0394. 1H NMR (360 MHz, CDCl3) δ 1.12 (6H, s), 2.46 (2H,
s), 2.64 (3H, s), 2.90 (2H, s), 7.35 (1H, d, J ) 3.6 Hz), 7.81
(1H, d, J ) 3.6 Hz). HPLC (70% CH3CN; 30% H2O (containing
0.1% TFA)): >99% purity. Retention time: 6.66 min.
6,6-Dim eth yl-3-m eth ylth io-1-(1-m eth yl-1,2,4-tr ia zol-3-
yl)-6,7-d ih yd r o-2-ben zoth iop h en -4(5H)-on e (41). A solu-
tion of 37 (5.0 g, 20 mmol) in saturated ethanolic hydrogen
chloride solution (500 mL) was stirred at 20 °C for 15 h. The
solvent was removed in vacuo and the residue triturated with
EtOAc. The resultant ethyl 6,6-dimethyl-3-methylthio-4-oxo-
4,5,6,7-tetrahydro-2-benzothiophene-1-carboximidoate hydro-
chloride (40) (2.5 g, 38%), which was isolated as a colorless
solid, was collected by filtration and dried under vacuum. mp
134-136 °C. MS (ES+) 298 [M + H]+. 1H NMR (250 MHz,
DMSO-d6) δ 1.00 (6H, s), 1.44 (3H, t, J ) 7.0 Hz), 2.45 (2H, s),
2.67 (3H, s), 3.00 (2H, s), 4.52 (2H, q, J ) 7.0 Hz).
Biologica l Meth od s
Ra d ioliga n d Bin d in g Stu d ies. L(tk-) cells expressing
human recombinant GABAA receptors containing â3 and γ2
subunits in combination with various R subunits were har-
vested and binding performed as described by Hadingham et
al.39 The displacement of [3H]Ro 15-1788 binding by the test
compounds was measured in GABAA receptors containing
either an R1-, R2-, R3-, or R5-subunit and from the IC50 the Ki
was calculated assuming respective KD values of [3H]Ro 15-
1788 binding of 0.92, 1.05, 0.58, and 0.45 nM at the R1-, R2-,
R3-, and R5-subtypes. Nonspecific binding was defined by the
inclusion of 10 µM flunitrazepam for the R1-, R2-, R3-, and
R5-subtypes. The percentage inhibition of [3H]Ro 15-1788, the
IC50 and the Ki values were calculated using ActivityBase
(IDBS).
E lect r op h ysiology. (a ) Volt a ge Cla m p R ecor d in g of
Xen opu s la evis Oocytes Tr a n sien tly Tr a n sfected w ith
th e Hu m a n GABA-A r5 Recep tor Su btyp e. Oocytes were
isolated from adult female Xenopus laevis frogs which had been
anaesthetized by immersion for 30 to 45 min in a solution of
0.4% ethyl m-aminobenzoate (MS 222 or Tricaine) in ac-
cordance with the UK Animals (Scientific Procedures) Act
1986. Oocytes were then transfected with human GABAA
receptor cDNA encoding one R5 plus â3 and γ2s subunits
engineered into the expression vector pCDM8 or pcDNAI/Amp
injected directly into the nucleus using a Drummond Scientific
Instruments 100 µL syringe attached to a micromanipulator.
Injected oocytes were then transferred into incubation medium
(1 L of regular MBS, 2 mL of penicillin (10,000 units/mL) and
40 (1.0 g, 3.4 mmol) was dissolved in EtOH (60 mL), and
methylhydrazine (160 mg, 3.4 mmol) was added. The solution
was heated at 50 °C for 7 h then evaporated to dryness. The
residue was dissolved in formic acid (20 mL) and the solution
heated at 100 °C for 16 h. The solvent was then removed in
vacuo and the residue partitioned between CH2Cl2 and satu-

2238 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Chambers et al.
streptomycin (10 mg/mL), 50 mg of gentamycin, and 2 mM
Na+pyruvate).
administration the animals were killed, as described above,
and the brains rapidly removed. Whole brain ([3H]Ro 15-1788)
or forebrain ([3H]L-655,708) were removed, weighed, and
rapidly homogenized in 10 times the volume of ice cold buffer
(10 mM KH2PO4, 100 mM KCl pH 7.4) using a Polytron
PT2100 homogenizer, and 300 µL of the homogenate was
filtered though presoaked Whatman GF/B filters using a
filtration manifold (Hoefer Scientific Instruments) and washed
twice with 5 mL of the ice cold buffer. Filters were then placed
in 10 mL of scintillation fluid (Hydrofluor, National Diagnos-
tics), and the level of radioactivity present was determined
using a liquid scinitillation counter (Beckman LS6500).
Specific binding was defined as radioactivity in vehicle or
compound-treated animals minus radioactivity in bretazenil-
treated animals.
Percentage occupancy ) 100 - (binding in drug group/
binding in vehicle group) × 100.
Electrophysiological recordings were made 2 to 4 days after
cDNA injection since the efficacy values determined for
standard compounds on day 1 were variable. If an oocyte
expressed the receptor subtype with an atypically high EC20
value then this cell was rejected from analysis. The ap-
proximate ideal EC20 value for R5 receptor subunits in oocytes
was 1-6 µM. Cells were superfused with MBS in a chamber
of approximately 50 µL volume. Glass recording microelec-
trodes were fabricated on a Brown Flaming electrode puller
and the tips were backfilled to approximately 10 mm with an
Agar solution containing KCl (2 M) solution. The electrode tips
were broken back to give a resistance of 1 to 2 MΩ in MBS
solution. Oocytes were impaled with these microelectrodes
using micromanipulators and then voltage-clamped at -60 mV
using an Axoclamp 2A/2B or Geneclamp amplifier. The change
in current following drug application was then plotted on a
chart recorder.
For each oocyte the expression of receptors was first checked
by a 30 s to 1 min application of 3 mM GABA in MBS. GABA
currents ranged typically from 0.3 to 3 µA, and cells with
currents <0.3 µA were rejected. Following a high concentration
of GABA the oocyte was perfused with MBS for 8 to 10 min to
facilitate recovery from desensitization. The concentration of
GABA required to activate a response which was 20% of the
maximum amplitude (EC20) was then determined. The test
P r ocon vu lsa n t Activity. The proconvulsant potential of
43 was determined using the method described by Roberts and
Keith.34 The convulsant used in this study was pentylenetet-
razole which was dosed at 15 mg/mL with an infusion rate of
0.2 mL/min. This rate was chosen to ensure that drug-naive
mice reached the terminal convulsion sign within 1 min.
Compound 43 was dosed ip at a concentration of 3 mg/kg in
70% PEG 300.
Wa ter Ma ze Test. The Morris water maze consisted of a
white, circular fiber-glass pool, of diameter 2 m, filled with
an opaque mixture of water and white dye (E308 opacifier,
Rohm and Haas (UK) Limited) maintained at 26-28 °C. It
was located in the center of a sound-attenuated room and was
illuminated by overhead fluorescent strip lights. A light gray
curtain was hung around the pool from above, such that it
almost touched the sides upon which eight laminated high
contrast black and white patterned pictures (42 cm × 30 cm)
were attached to serve as spatial ‘extra-maze’ cues. A direction
‘north’ was arbitrarily determined and the pool was divided
into four equal quadrants, ‘northeast’, ‘southeast’, ‘southwest’,
and ‘northwest’. The curtain was split and overlapped at points
north, south, east, and west to allow access to the pool in order
to place rats into the water. A round platform of diameter 10
cm was submerged 2 cm below the surface of the water
rendering it invisible at surface level. A closed-circuit CCD
video camera fitted with a wide-angle lens was mounted
directly above the center of the pool and was connected (via a
VCR) to an image analyzer (VP 200, HVS Image Ltd, UK)
which digitized the image. The dark heads of the hooded lister
rats provided a high contrast image against the white colored
water, which could be tracked by the image analyzer. The
digital information was relayed to a PC running ‘HVS Water
2020’ a software package supplied by HVS Image, UK. This
software provided a multitude of measures including latency
to reach the platform, length of path taken, swimming speed,
and time spent in areas of the pool.
Before the experiments began the male hooded Lister rats
were trained in the water maze task by being given four trials
each day for 8-10 days. During these trials the submerged
platform was placed in a different location each day, but
remained constant for the duration of each day. During this
training phase, the rats learned the procedural rules of the
water maze delayed ‘matching-to-place’ task: (i) there is an
escape platform hidden in the pool, (ii) the location of the
hidden platform changes each day, but it is consistent within
the day, and (iii) the extra maze spatial cues can be used to
locate the platform position. On trial 1 the latency to find the
platform is relatively long, but is shorter on trials 2, 3, and 4,
demonstrating that the animals’ memory for the platform
location improves each time it escapes to the hidden platform.
The improvement in the animals’ memory was quantified by
subtracting the trial 2 latency from the trial 1 latency to give
the savings score. For each daily trial the rat was taken from
the home cage and placed into the water maze at one of the
four quasi-randomly determined locations (‘north’, ‘east’, ‘south’,
or ‘west’) with its head facing, and almost touching, the pool
wall. Trials began when the rat was released by the experi-
compound was applied at
a concentration approximately
equivalent to 100× the Ki value. Efficacy of the test compounds
was then determined as the percentage change in the current
activated by the EC20 concentration of GABA according to the
equation: Efficacy (%) ) ((current with test compound/control
current) × 100) - 100.
(b) Wh ole Cell P a tch -Cla m p of L(tk -) Cells Sta bly
Tr a n sfected w ith Hu m a n GABA-A Recep tor s. Experi-
ments were performed on L(tk-) cells expressing human cDNA
combinations R1â3 γ2s, R2â3 γ2s, R3â3 γ2s, and R5â3γ2s.
Glass cover-slips containing the cells in a monolayer culture
were transferred to a Perspex chamber on the stage of Nikon
Diaphot inverted microscope. Cells were continuously perfused
with a solution containing 124 mM NaCl, 2 mM KCl, 2 mM
CaCl2, 1 mM MgCl2, 1.25 mM KH2PO4, 25 mM NaHCO3, 11
mM D-glucose, at pH 7.2, and observed using phase-contrast
optics. Patch-pipets were pulled with an approximate tip
diameter of 2 µm and a resistance of 4 MΩ with borosilicate
glass and filled with 130 mM CsCl, 10 mM HEPES, 10 mM
EGTA, 3 mM Mg+-ATP, pH adjusted to 7.3 with CsOH. Cells
were patch-clamped in whole-cell mode using an Axopatch-
200B patch-clamp amplifier. Drug solutions were applied by
a double-barreled pipet assembly, controlled by a stepping
motor attached to a Prior manipulator, enabling rapid equili-
bration around the cell. Increasing GABA concentrations were
applied for 5 s pulses with a 30 s interval between applications.
Curves were fitted using a nonlinear square-fitting program
to the equation f(x) ) Bmax/[1 + (EC50/x)n] where x is the drug
concentration, EC50 is the concentration of drug eliciting a half-
maximal response, and n is the Hill coefficient. Allosteric
modulation of GABA receptors was measured relative to a
GABA EC20, individually determined for each cell to account
for differences in GABA affinity.
In Vivo Bin d in g. In vivo occupancy of the GABAA receptor
subtypes was determined using the method described by Atack
et al.33 Rats or mice were dosed intraperitoneally at 1 and 10
mL/kg, respectively, with either 43 in 70% PEG 300, 70% PEG
300 (to determine total radioligand binding) or 5 mg/kg
bretazenil in PEG 300 (to determine nonspecific radioligand
binding). To determine the occupancy at GABAA R1-, R2-, R3-,
and R5-subunit-containing receptors, rats were given an
intravenous (iv) dose of a 15 µCi/mL solution of [3H]Ro 15-
1788 (DuPont NEN, 1 mL/kg) 3 min prior to killing by
stunning with decapitation. Occupancy at GABAA R5-subunit-
containing receptors was determined using [3H]L-655,708 as
a 11 µCi/mL solution dosed at 1 mL/kg, which was adminis-
tered iv 1 min prior to killing. Thirty minutes after drug

Selective GABAA R5 Receptor Inverse Agonist
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11 2239
(16) Low, K.; Crestani, F.; Keist, R.; Benke, D.; Brunig, I.; Benson,
J . A.; Fritschy, J .-M.; Rulicke, T.; Bluethmann, H.; Mohler, H.;
Rudolph, U. Molecular and neuronal substrate for the selective
attenuation of anxiety. Science 2000, 290, 131-134.
(17) 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 .
Sedative but not anxiolytic properties of benzodiazepines are
mediated by the GABAA receptor R1 subtype. Nat. Neurosci.
2000, 3, 587-592.
menter and ended when the rat climbed onto the platform and
the mean escape latency was recorded. The maximum trial
length was 60 s. If by that time the rat had not climbed on to
the platform, the trial ended automatically; the experimenter
intervened and placed the rat on the platform, and an escape
latency of 60 s was recorded. The rat remained on the platform
for 30 s (inter-trial interval, ITI). At the end of the ITI, the
rat was placed into the pool again, but at a different location,
and upon release the next trial began. This procedure was
repeated until four trials had been completed. Before com-
pound treatment commenced, animals were assigned to treat-
ment groups in such a way so as to ensure that the level of
performance (using mean savings) during the training phase
was not significantly different. Rats were dosed ip with either
vehicle (70%/30% PEG 300/water) or 43 (0.3 mg/kg), 30 min
before commencing the trials. The compound treatment phase
lasted for 5 days. It was identical to the methodology described
above with the exception that a 4 h delay period was inserted
between trial 1 and trial 2. Statistically significant differences
were determined using two-way ANOVA.
(18) Sur, C.; Quirk, K.; Dewar, D.; Atack, J . R.; McKernan, R. Rat
and human hippocampal R5 subunit-containing γ-aminobutyric
acidA receptors have R5â3γ2 pharmacological characteristics.
Mol. Pharmacol. 1998, 54, 928-933.
(19) Sieghart, W. Structure and pharmacology of γ-aminobutyric
acidA receptor subtypes. Pharmacol. Rev. 1995, 47, 181-234.
(20) (a) 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 GABAA/BzR subtypes (R1â3γ2, R5â3γ2 and
R6â3γ2) via a comprehensive ligand-mapping approach. J . Med.
Chem. 2000, 43, 71-95. (b) Masciadri, R.; Thomas, A. W.;
Wichmann, J . Preparation of 9H-imidazo[1,5-a][1,2,4]triazolo-
[4,3-d][1,4]benzodiazepines as alpha 5 subunit selective GABA
A receptor inverse agonists. In PCT Int. Appl.; (F. Hoffmann-
La Roche AG, Switz.). WO 02/040487. (c) Masciadri, R.; Thomas,
A. W.; Wichmann, J . Imidazo[1,5-a]pyrimido[5,4-d] benzodiaz-
epine derivatives as GABA A receptor modulators. In PCT Int.
Appl.; (F. Hoffmann-La Roche AG, Switz.). WO 02/094834.
(21) Quirk, K.; Blurton, P.; Fletcher, S.; Leeson, P.; Tang, F.; Mellilo,
D.; Ragan, C. I.; McKernan, R. M. [3H]L-655, 708, a novel ligand
selective for the benzodiazepine site of GABAA receptors which
contain the R5 subunit. Neuropharmacology 1996, 35, 1331-
1335.
Ack n ow led gm en t . The authors thank Steve
Thomas for obtaining the high resolution mass spectra.
Su p p or tin g In for m a tion Ava ila ble: Data showing the
mean swim speeds of rats dosed with 43 (0.3 mg/kg ip) and
vehicle (70% PEG 300) in the delayed ‘matching-to-place’ water
maze test. This material is available free of charge via the
Internet at http://pubs.acs.org.
(22) Yu, S.; Ma, C.; He, X.; McKernan, R.; Cook, J . M. Studies in the
search for R5 subtype selective agonists for GABAA/BzR sites.
Med. Chem. Res. 1999, 9, 71-88.
Refer en ces
(1) Bormann, J . The ‘ABC’ of GABA receptors. Trends Pharmacol.
Sci. 2000, 21, 16-19.
(23) Chambers, M. S.; Atack, J . R.; Bromidge, F. A.; Broughton, H.
B.; Cook, S.; Dawson, G. R.; Hobbs, S. C.; Maubach, K. A.; Reeve,
A. J .; Seabrook, G. R.; Wafford, K.; MacLeod, A. M. 6,7-Dihydro-
2-benzothiophen-4(5H)-ones: A novel class of GABA-A R5 recep-
tor inverse agonists. J . Med. Chem. 2002, 45, 1176-1179.
(24) Kearsley, S. K.; Sallamack, S.; Fluder, E. M.; Andose, J . D.;
Mosley, R. T.; Sheridan, R. P. Chemical similarity using phys-
iochemical property descriptors. J . Chem. Inf. Comput. Sci. 1996,
36, 118-127.
(25) Patent EP 0601184. Inventor, Yoshitomi Pharmaceutical Ind.
Ltd.
(26) Available from Maybridge Chemical Co. Ltd., Trevillett, Corn-
wall, U.K.
(27) Casula, M. A.; Bromidge, F. A.; Pillai, G. V.; Wingrove, P. B.;
Martin, K.; Maubach, K.; Seabrook, G. R.; Whiting, P. J .;
Hadingham, K. L. Identification of amino acid residues respon-
sible for the R5 subunit binding selectivity of L-655, 708, a
benzodiazepine binding site ligand at the GABAA receptor. J .
Neurochem. 2001, 77, 445-451.
(28) Hadingham, K. L.; Garrett, E. M.; Wafford, K. A.; Bain, C.;
Heavens, R. P.; Sirinathsinghji, D. J . S.; Whiting, P. J . Cloning
of cDNAs encoding the human γ-aminobutyric acid type A
receptor R6 subunit and characterization of the pharmacology
of R6-containing receptors. Mol. Pharmacol. 1996, 49, 253-259.
(29) Wafford, K. A.; Whiting, P. J .; Kemp, J . A. Differences in affinity
and efficacy of benzodiazepine receptor ligands at recombinant
γ-aminobutyric acidA receptor subtypes. Mol. Pharmacol. 1993,
43, 240-244.
(2) Whiting, P. J .; Wafford, K. A.; McKernan, R. M. Pharmacologic
subtypes of GABAA receptors based on subunit composition. In
GABA in the nervous system: The view at fifty years; Martin, D.
L.; Olsen, R. W., Eds.; Lippincott, Williams & Wilkins: New
York, 2001; pp 113-126.
(3) McKernan, R. M.; Whiting, P. J . Which GABAA-receptor sub-
types really occur in the brain? Trends Neurosci. 1996, 19, 139-
143.
(4) Barnard, E. A.; Skolnick, P.; Olsen, R. W.; Mohler, H.; Sieghart,
W.; Biggio, G.; Braestrup, C.; Bateson, A. N.; Langer, S. Z.
Subtypes of γ-aminobutyric acidA receptors: Classification on
the basis of subunit structure and receptor function. Pharmacol.
Rev. 1998, 50, 291-313.
(5) Olsen, R. W.; Venter, J . C. Benzodiazepine/ GABA receptors and
chloride channels: Structural and functional properties; New
York: Alan R. Liss Inc., 1986.
(6) Cole, S. O. Effects of benzodiazepines on acquisition and
performance: A critical assessment. Neurosci. Biobehav. Rev.
1986, 10, 265-272.
(7) Ghoneim, M. M.; Mewaldt, S. P. Benzodiazepines and human
memory: A review Anaesthesiology 1990, 72, 926-938.
(8) Hagen, T. J .; Guzman, F.; Schultz, C.; Cook, J . M.; Skolnick, P.;
Shannon, H. E. Synthesis of 3,6-disubstituted â-carbolines which
possess either benzodiazepine antagonist or agonist activity.
Heterocycles 1986, 24, 2845-2855.
(9) Liu, R.; Zhang, P.; McKernan, R. M.; Wafford, K.; Cook, J . M.
Synthesis of novel imidazobenzodiazepines selective for the
R5â2γ2 (Bz5) GABAA/benzodiazepine receptor subtype. Med.
Chem. Res. 1995, 5, 700-709.
(10) McNamara, R. K.; Skelton, R. W. Benzodiazepine receptor
antagonists flumazenil and CGS 8216 and inverse agonist
â-CCM enhance spatial learning in the rat: Dissociation from
anxiogenic actions. Psychobiology 1993, 21 (2), 101-108.
(11) Dorow, R.; Horowski, R.; Paschelke, G.; Amin, M.; Braestrup,
C. Severe anxiety induced by FG 7142, a â-carboline ligand for
benzodiazepine receptors. Lancet 1983, 2, 98-99.
(12) Petersen, E. N. DMCM: A potent convulsive benzodiazepine
receptor ligand. Eur. J . Pharmacol. 1983, 94, 117-124.
(13) Little, H. J .; Nutt, D. J .; Taylor, S. C. Acute and chronic effects
of the benzodiazepine receptor ligand FG 7142: proconvulsant
properties and kindling. Br. J . Pharmacol. 1984, 83, 951-958.
(14) Sarter, M.; Bruno, J . P.; Berntson, G. G. Psychotogenic properties
of benzodiazepine receptor inverse agonists. Psychopharmacol-
ogy 2001, 156, 1-13.
(15) Rudolph, U.; Crestani, F.; Benke, D.; Brunig, I.; Benson, J . A.;
Fritschy, J .-M.; Martin, J . R.; Bluethmann, H.; Mohler, H.
Benzodiazepine actions mediated by specific γ-aminobutyric
acidA receptor subtypes. Nature 1999, 401, 796-800.
(30) Hadingham, K. L.; Wafford, K. A.; Thompson, S. A.; Palmer, K.
J .; Whiting, P. J . Expression and pharmacology of human
GABAA receptors containing γ3 subunits. Eur. J . Pharmacol.
1995, 291, 301-309.
(31) Horne, A. L.; Hadingham, K. L.; Macaulay, A. J .; Whiting, P.;
Kemp, J . A. The pharmacology of recombinant GABAA receptors
containing bovine R1, â1, γ2L subunits stably transfected into
mouse fibroblast L-cells. Br. J . Pharmacol. 1992, 107, 732-737.
(32) Radioligand binding and enzyme assays performed by MDS
Pharma Services, PO Box 26-127, Taipei, Taiwan, R.O.C.
(33) Atack, J . R.; Smith, A. J .; Emms, F.; McKernan, R. M. Regional
differences in the inhibition of mouse in vivo [3H]Ro 15-1788
binding reflect selectivity for R1 versus R2 and R3 subunit-
containing GABAA receptors. Neuropsychopharmacology 1999,
20, 255-262.
(34) Roberts, A. J .; Keith, L. D. Mineralocorticoid receptors mediate
the enhancing effects of corticosterone on convulsion susceptibil-
ity in mice. J . Pharm. Exp. Ther. 1994, 270, 505-511.
(35) Morris, R. Developments of a water-maze procedure for studying
spatial learning in the rat. J . Neurosci. Methods 1984, 11, 47-
60.

2240 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 11
Chambers et al.
(36) Steele, R. J .; Morris, R. G. M. Delay-dependent impairment of a
matching-to-place task with chronic and intrahippocampal infu-
sion of the NMDA-antagonist D-AP5. Hippocampus 1999, 9,
118-136.
(37) Collinson, N.; Kuenzi, F. M.; J arolimek, W.; Maubach, K. A.;
Cothliff, R.; Sur, C.; Smith, A.; Otu, F. M.; Howell, O.; Atack, J .
R.; McKernan, R. M.; Seabrook, G. R.; Dawson, G. R.; Whiting,
P. J .; Rosahl, T. W. Enhanced learning and memory and altered
GABAergic synaptic transmission in mice lacking the R5 subunit
of the GABAA receptor. J . Neurosci. 2002, 22, 5572-5580.
(38) Crestani, F.; Keist, R.; Fritschy, J .-M.; Benke, D.; Vogt, K.; Prut,
L.; Bluthmann, H.; Mohler, H.; Rudolph, U. Trace fear condi-
tioning involves hippocampal R5 GABAA receptors. Proc. Natl.
Am. Sci. 2002, 99, 8980-8985.
(39) Hadingham, K. L.; Wingrove, P.; Le Bourdelles, B.; Palmer, K.
J .; Ragan, C. I.; Whiting, P. J . Cloning of cDNA sequences
encoding human R2 and R3 γ-aminobutyric acidA receptor
subunits and characterization of the benzodiazepine pharmacol-
ogy of recombinant R1-, R2-, R3 and R5-containing human
γ-aminobutyric acidA receptors. Mol. Pharmacol. 1993, 43,
970-975.
J M020582Q