Discovery of the imidazo[1,5-a][1,2,4]-triazolo[1,5-d][1,4]benzodiazepine scaffold as a novel, potent and selective GABAA α5 inverse agonist series

Article abstract of DOI:10.1016/j.bmcl.2009.07.153

Through iterative design cycles we have discovered a number of novel new classes where the imidazo[1,5-a][1,2,4]-triazolo[1,5-d][1,4]benzodiazepine was deemed the most promising GABA_A α5 inverse agonist class with potential for cognitive enhanc

Full text of DOI:10.1016/j.bmcl.2009.07.153

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5746–5752
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of the imidazo[1,5-a][1,2,4]-triazolo[1,5-d][1,4]benzodiazepine
scaffold as a novel, potent and selective GABAA a5 inverse agonist series
Guido Achermann a, Theresa M. Ballard b, Francesca Blasco c, Pierre-Emmanuel Broutin a,
Bernd Büttelmann a, Holger Fischer a, Martin Graf d, Maria-Clemencia Hernandez b, Peter Hilty a,
Frédéric Knoflach b, Andreas Koblet a, Henner Knust a, Anke Kurt a, James R. Martin b, Raffaello Masciadri a,
b
a
Richard H. P. Porter b, Heinz Stadler a, Andrew W. Thomas a, , Gerhard Trube , Jürgen Wichmann
*
a F. Hoffmann-La Roche Ltd, Pharmaceutical Research Basel, Discovery Chemistry, CH-4070 Basel, Switzerland
b F. Hoffmann-La Roche Ltd, Pharmaceutical Research Basel, CNS Disease Biology Area, CH-4070 Basel, Switzerland
c F. Hoffmann-La Roche Ltd, Pharmaceutical Research Basel, NonClinical Safety, CH-4070 Basel, Switzerland
d F. Hoffmann-La Roche Ltd, Pharmaceutical Research Basel, Discovery Technnologies, CH-4070 Basel, Switzerland
a r t i c l e i n f o
a b s t r a c t
Article history:
Through iterative design cycles we have discovered a number of novel new classes where the imi-
dazo[1,5-a][1,2,4]-triazolo[1,5-d][1,4]benzodiazepine was deemed the most promising GABAA 5 inverse
agonist class with potential for cognitive enhancement. This class combines a modest subtype binding
selectivity with inverse agonism and has the most favourable molecular properties for further lead opti-
misation towards a central nervous system (CNS) acting medicine.
Received 1 July 2009
Revised 30 July 2009
Accepted 31 July 2009
Available online 4 August 2009
a
Ó 2009 Published by Elsevier Ltd.
Keywords:
GABAA receptor
Benzodiazepines
Cognition
Imidazole
Triazole
Medicinal chemistry
Although senile dementia was first recognised by Pythagoras in
the seventh century B.C. as a medical condition, and more recently
diagnosed pathologically in 1906 by Kraepelin and Alzheimer, still
today a significant unmet medical need for the treatment of Alzhei-
mer’s disease (and related dementias) exists.1 Current medicines
based on NMDA receptor antagonism (e.g., Memantine)2 and cho-
linesterase inhibition (e.g., Donepezil)3 suffer from well docu-
mented modest efficacy in patients and mechanism related side
effects for the latter. An alternative pharmacological approach
Since the discovery of these medicines, innovative molecular biol-
ogy has revealed that GABAA receptors comprise of a pentameric
assembly of heteromeric subunits in a 2:2:1 stoichiometry with
the most common protein assemblies being
3b3 2 and 5b3
2.5
Apart from the non-selective antagonist Flumazenil all cur-
a1b2c2, a2b3c2,
a
c
a
c
rently used benzodiazepines are non-selective agonists for these
subtypes and these are in clinical use as sedatives, anxiolytics
and muscle relaxants. The clinical side effects of these compounds
can be explained by interaction with multiple undesired subtypes.
Interestingly, clinical studies with b-carbolines, which are non-
based on selectively targeting the GABAA
a5 receptor subtype of-
fers unique opportunities for the potential treatment of the cogni-
tive deficits in Alzheimer’s disease and related dementias.4
selective inverse agonists for the GABAA
a1b2c2, a2b3c2,
c-Aminobutyric acid (GABA) is the major inhibitory neurotrans-
a
3b3 2 and 5b3 2 subtypes, induce the converse pharmacologi-
c
a
c
mitter in the mammalian brain and the GABAA receptor (ligand
gated chloride channel) has been, historically, one of the most suc-
cessful pharmacological targets delivering a large number of clini-
cally differentiated medicines as anti-epileptics, anxiolytics,
hypnotics and sedatives which predominantly function via in-
creased receptor activation via the benzodiazepine binding site.
cal effects on behaviour including, importantly, enhanced learning
and memory. However, due to poor pharmacological profile, these
compounds induced convulsions and anxiety in animal and
healthy volunteer studies, respectively, so that clinical develop-
ment was not an option.6
Genetic and pharmacological studies have outlined the domi-
nant individual contributions from each GABAA
revealing that agonism at GABAA 1 induces sedation and agonism
at GABAA 2 and/or GABAA 3 imposes the desired anxiolytic effect
axbxc2 subtype
a
* Corresponding author. Tel.: +41 61 688 50 48; fax: +41 61 688 87 14.
E-mail address: andrew.thomas@roche.com (A.W. Thomas).
a
a
0960-894X/$ - see front matter Ó 2009 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2009.07.153

G. Achermann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5746–5752
5747
Table 2
N
N
N
N
CO2Et
Molecular property profile of selected compounds
O
N
N
N
Compd
Solubilitya
g/mL)
Log D7.4
Permeation coefficient
(10ꢀ6cm sꢀ1) (PAMPA)
Cl (mL/min/mg
(
l
protein)b
O
(rat/human)
N
N
14
20
20
10
2.68
2.68
3.6
2.78
11/4
N
N
"α5IA" 1
2
c
ꢀ /8
N
a
b
c
LYSA measurement.
From incubations with liver microsomes.
Not stable in matrix.
Figure 1. Merck clinical candidate and our selected hit compound.
the alcohol induced memory impairment in a double-blind placebo
controlled cross-over study in 12 healthy volunteers.9 This com-
pound was prevented for further clinical use due to dose limits im-
in rodent models. Conversely, inverse agonism at GABAA
duces pro-convulsant and inverse agonism at GABAA 2 and/or
GABAA 3 is responsible for an anxiogenic response.
a1 in-
a
a
posed on ‘a5IA’ due to the potential for precipitation, crystalluria
Memory impairment is a known undesired clinical result of
benzodiazepine use and is thought to be due to the agonistic
and nephrotoxicity by the main metabolite.10
Additional non-selective GABAA receptor subtype inverse ago-
nists have been studied in the clinics but no information relating
to their safety, tolerability or efficacy is currently publicly
available.11
effects at the GABAA
consideration the clinical knowledge from both benzodiazepine
and b-carboline pharmacology enough evidence is present to
strongly rationalise how selective GABAA
a5 receptor subtype. Therefore, taking into
a5 receptor subtype in-
To the best of our knowledge, very few compounds12 have been
successfully designed that combine some degree of both binding
and functional selectivity and this communication, and subsequent
ones13–15 will reveal our successful efforts to identify a number of
interesting novel series and compounds with best and first in class
pharmacological and pharmacokinetic profiles.
verse agonists, devoid of convulsant or anxiogenic action, could
have a significant clinical impact on the treatment of memory
and learning deficits. The additional evidence7 that the GABAA
a5
receptor subtype is predominant in the hippocampus and the fact
that the density and pharmacology of the GABAA 5 receptor
subtype is preserved in post-mortem AD patients is compelling
a
As our source for starting points we decided to perform a high
throughput screening (HTS) of our corporate compound library
and unsurprisingly we achieved a high hit rate of ‘benzodiazepine
classes’. We decided to pursue further compound 2 (Fig. 1) (an imi-
dazo [1,5-a]pyrimidino[5,4-d]benzazepine) due to its interesting
pharmacological profile and intellectual property opportunities.
enough a reason to design a medicinal chemistry programme
aimed at the discovery of selective GABAA
inverse agonists.
a5 receptor subtype
Researchers at Merck have discovered a plethora of functionally
selective compound series8 culminating in an experimental
medicine study with ‘a5IA’ (Fig. 1) which significantly reversed
Table 1
Binding and efficacy profile of analogues of 2
N
R3
N
R1
N
N
R2
Compd
R1
R2
R3
Affinity GABAA
5b3
2a KI (nM)
Selectivity17
Efficacy GABAA
5b3
2b (%)
a
c
KI (nM)
a1b3
c
c
2
2
a
c
KI (nM)
a5b3
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
H
H
H
H
H
H
H
F
H
Ph
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
H
Cl
Cl
Cl
Cl
CO2Et
Cl
3.7
21
6
+11
+11
+59
+19
170.6
12.5
5.5
3.8
8.9
7.6
4.6
8.0
11.2
7.8
20.2
12.5
410.0
15.6
15.6
136.0
61.0
8.1
PMB
nPr
Me
iPr
cPr
Me
nPr
iPr
cPr
H
Me
H
H
Me
H
Me
Me
Me
16
12
12
11
7
9
5
7
6
31
37
>8
8
14
6
10
70
>24
F
F
F
Br
Br
H
H
H
Br
Br
H
H
ꢀ12
ꢀ19
+15
+21
ꢀ6
ꢀ13
ꢀ29
ꢀ41
133
a
Cloned rat or human receptor subunits were expressed in insect SF9 cells, human HEK293 cells or Xenopus laevis oocytes for 3H-Flumazenil GABAA a1
,a2,a3 and a5 binding
or electrophysiology.
b
Efficacy is determined as the percentage change of a submaximal (EC10) response to GABA.

5748
G. Achermann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5746–5752
H
N
a)
b)
R2
NH2
NH2
CO2Et
N
N
O
CO2H
Br
CO2Et
Br
CO2Et
Br
R1
N
*
*
*
*
O
H
N
O
H
O
H
N
Figure 2. Imidazo benz-‘diaza’-series explored.
c)
f)
d)
e)
N
Br
Br
Br
CO2Et
HO
erties, 14 (Table 2) was selected to be tested in an in vivo passive
avoidance test of cognition in rats where activity was demon-
strated at 100 mg/kg ip (Supplementary data).
O
O
Me2N
N
CO2Et
O
H
N
Since it was not possible to optimally combine both favourable
pharmacological and pharmacokinetic profiles within this series
we decided to terminate our efforts and begin to explore a number
of additional pyrimidine replacements (including pyrimidine iso-
mers, and pyridine and phenyl isomers). All these changes were
met with no success, so we terminated these benzazepine-classes.
A number of synthetic routes were developed to access the imi-
dazo [1,5-a]pyrimidino[5,4-d]benzazepines series. The route de-
picted in Scheme 116 was found to be suitable for both small and
large scale preparation of derivatives in Table 1. The synthetic
strategy relied on an optimised base (KH is far superior to all other
bases examined) mediated intramolecular Dieckmann cyclisation
to the key benzazepinedione building block. A complimentary
route using zinc homoenolate methodology proved to be an effi-
cient alternative reductive cyclisation method (Scheme 2). Trans-
formation of the benzazepinediones to the final products
proceeded without event through standard processes.
N
g)
Br
Br
N
N
N
N
14
Scheme 1. Synthesis of 14. Reagents and conditions: (a) EtOH, HCl (concd), reflux,
13 h, 80%; (b) ethyl succinyl chloride, CaCO3, toluene, reflux, 1 h, 79%; (c) KH,
toluene, 70 °C, 2 h, 71%; (d) DMSO, H2O, 150 °C, 5 h, 65%; (e) Me2NCH(OMe)2,
120 °C, 1 h, 90%; (f) acetamidine HCl, NaOMe, rt, 4 h, 87%; (g) POCl3, N,N-dimethyl-
p-toluidine, LiHMDS, (E)-(dimethylaminomethyleneamino)-acetic acid ethyl ester,
AcOH, 50%.
Initial SAR exploration revealed that 2 demonstrated very good
potency (3.7 nM) at the GABAA
binding selectivity (21-fold) versus the GABAA
(Table 1).16 However, in functional testing 2 was found to be ago-
nistic at the GABAA 5 receptor subtype and this trend was shown
for compounds 3–5. Introduction of a F-substituent resulted in a
reduction of binding selectivity. Pleasingly the introduction of a
Br substituent offered the best compromise in potency, binding
selectivity17 and inverse agonism as shown in compound 14.
We then sought out to identify ester replacements but were
surprised to learn that all typical ester isosteres investigated re-
a
5 receptor subtype and also high
a
1 receptor subtype
a
We then sequentially and thoroughly examined the effect of
five-membered rings and designed indoles, indazoles, pyrazoles,
imidazoles, triazoles and tetrazole isomers keeping the privileged
imidazobenzodiazepine scaffold constant (Fig. 2). The most inter-
esting from the many sub-series examined are shown in Tables
3–8.
Pleasingly, the imidazo[1,5-a]pyrazolo[1,5-d][1,4]benzodiaze-
sulted in complete loss of affinity at the GABAA
a5 receptor sub-
pine class delivered very potent GABAA
a5 receptor subtype in-
type and additionally any relatively potent examples (ꢁ100 nM)
were all agonistic in functional response. No substituent (e.g., com-
pound 15) was not tolerated and, as will be shown in more detail in
the subsequent paper, a Cl substituent was found to be a suitable
verse agonists and in this case ester replacements were even
more selective and more efficacious (Table 3). The molecular prop-
erties (Table 4) of these classes were also favourable, but since we
were seeking higher levels of binding selectivity we decided to
study further the effect of additional structural diversity (Tables
5–7).
Imidazo[1,5-a]pyrazolo[1,5-d][1,4]benzodiazepines were pre-
pared by an efficient sequence of reactions with a key step being
a reductive ring cyclisation with subsequent base mediated ring
expansion reaction to give the desired pyrazolo[1,5-d][1,4]benzo-
ester replacement where some potency at the GABAA a5 receptor
subtype was retained but it was not able to combine with an in-
verse agonistic effect within this class. Introduction of a terminal
acetylene moiety in compound 20 offered the best compromise
in potency, selectivity and inverse agonist effects but replacement
of the ester functional group by many groups including a Cl atom
proved to be detrimental. Overall compounds 14 and 20 were per-
haps the most promising from the imidazo [1,5-a]pyrimidino[5,4-
d]benzazepine series. Characterised by acceptable molecular prop-
Table 3
Binding and efficacy profile of analogues of 22
N
R2
NO2
H
N
NO2
O
R1
NO2
b)
a)
N
O
N
CO2Et
+
CO2Et
EtO
OTMS
OX
Compd R1 R2
Affinity GABAA
5b3 2f KI (nM) KI (nM)
Selectivity17
Efficacy GABAA
5b3
2a (%)
X = H and TMS
a
c
a1b3
c
c
2
2
a
c
KI (nM)
a5b3
O
H
N
NH2
c)
d)
22
23
24
25
26
27
H
CO2Et 0.2
4
1
14
7
8
ꢀ18
ꢀ7
CO2Et
Cl CO2Et 1.4
O
H
H
H
Me
Cl
3.0
2.0
ꢀ26
ꢀ39
ꢀ38
ꢀ24
O
COcPr 1.0
49.1
Scheme 2. Alternative route to benzazepinediones. Reagents and conditions: (a)
ZnI2, DCM, rt, 5 min, 77%; (b) PCC, DCM, rt, 20 h, 92%; (c) H2, Pd/C, EtOAC, rt, 3 h,
80%; (d) NaH, ꢀ40 °C to rt, 3 h, 81%.
Br Me
5
a
See Table 1.

G. Achermann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5746–5752
5749
Table 4
deficiency of relatively high microsomal clearance even with polar
substituents such as 33.
Molecular property profile of compound 26 and 27
The key building block can be conveniently prepared in three
synthetic steps and upon reaction with hydroxylamine gives the
seven-membered ring selectively. Selective ‘O’ 1,4-addition to
ethyl propiolate followed by [3,3]-sigmatropic rearrangement
and dehydrative cyclisation to the imidazole forms the ester inter-
mediate which could be converted to 33 in a straightforward man-
ner (Scheme 4).
26
27
Solubility (
Log D7.4
l
g/mL)a
46
9
2.39
3.77
26/9
54/63
2.81
3.27
8/2
Permeation coefficient (10ꢀ6cm sꢀ1) (PAMPA)
Cl (mL/min/mg protein)a (rat/human)
MAB (%)
78/88
a
See Table 2.
We next turned our attention to a novel class of imidazo[1,5-
a][1,2,4]triazolo[4,3-d][1,4]benzodiazepines where we were able
to develop an extensive data set from more than 180 derivatives
(selected examples shown in Table 6).19 As expected small substit-
diazepine which was transformed to the final product under stan-
dard conditions (Scheme 3).18
The diimidazo[1,5-a:10,20-d][1,4]benzodiazepine class allowed
us for the first time to achieve very high levels of binding selectiv-
ity with concomitant high potency and inverse agonism at the GA-
uents at R1 resulted in very potent GABAA
a
5 receptor subtype
binding and increasing lipophilicity or polarity resulted in in-
creased binding selectivity versus the GABAA 1,2,3 receptor sub-
1 shown but similar selectivity is
a
BAA
a5 receptor subtype but unfortunately this class had the
types (for brevity only GABAA
a
Table 5
Binding and efficacy profile of analogues of 28
N
N
N
N
Br
R1
R2
Compd
R1
R2
Affinity GABAA
Selectivity17
Efficacy GABAA
5b3
2 (%)a
a5b3c
2f KI (nM)
KI (nM)
a1b3
c
c
2
2
a
c
KI (nM)
a5b3
28
29
30
31
32
33
H
H
Me
H
H
H
Me
H
CO2Et
CH2OH
CH2N–pyrolidinone
6.6
35
27
75
43
101
118
ꢀ35
ꢀ8
119.0
11.7
5.4
8.1
1.6
ꢀ19
ꢀ34
ꢀ15
ꢀ25
H
a
See Table 1.
Table 6
Binding and efficacy profile of analogues of 34
N
N
R2
R3
R1
N
N
N
Compd
R1
R2
R3
Affinity GABAA
5b3 2f KI (nM)
Selectivity17
Efficacy GABAA
5b3
2a (%)
a
c
KI (nM)
a
1b3
c
c
2
2
a
c
KI (nM)
a
5b3
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
H
F
Cl
Br
Me
MeO
HO
H
H
NC
Br
Br
Br
Br
Br
MeO
MeO
MeO
Br
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
Cl
Cl
Cl
I
Me
H
COiPr
CO2Et
CO2Et
CO2Et
CO2Et
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
0.8
1.7
1.2
0.6
2.1
1.4
0.6
2.3
33.6
25.3
27.4
7.8
150.0
7.8
55.5
14.5
53.8
22.9
19.1
8
7
29
65
43
26
28
59
94
11
33
32
18
60
38
33
59
89
47
ꢀ2
ꢀ5
+20
ꢀ12
ꢀ13
ꢀ19
ꢀ24
ꢀ22
ꢀ5
ꢀ10
ꢀ20
0
ꢀ11
+55
ꢀ31
+75
+40
Me
Ph
Bn
Bn
a
see Table 1

5750
G. Achermann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5746–5752
Table 7
Binding and efficacy profile of analogues of 53
N
R2
N
R1
N
N
N
Compd
R1
R2
Affinity GABAA
5b3 2f KI (nM)
Selectivity: KI (nM)
KI (nM) 5b3
a
xb3
c
2
Efficacy GABAA
5b3
2 (%)a
a
c
a
c
a
c2
vs
a
1
vs
a
2
vs
a3
1
5
6
7
3
5
7
18
2
1
1
6
6
6
1
5
19
11
6
ꢀ16
ꢀ35
ꢀ25
ꢀ22
ꢀ56
ꢀ39
ꢀ39
ꢀ51
ꢀ29
54
55
56
57
58
59
60
61
62
63
64
Cl
Br
Me
cPr
OMe
OCF3
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
H
0.2
0.3
0.5
0.6
0.2
0.8
0.6
6.4
0.2
6.4
0.5
15
13
7
1
9
16
7
11
2
H
iPrCOHN
H
H
H
2
2
16
15
0
16
0
19
ꢀ37
ꢀ52
Et
a
See Table 1.
also not tolerated. As a result of the relatively low solubility, high
lipophilicity of this class coupled to the fact that it was not possible
to identify single compounds which met our overall target profile
we abandoned this class even though 39 was tested in an in vivo
passive avoidance test of cognition in rats where activity was dem-
onstrated at 3 mg/kg ip (Supplementary data).
Our preferred route to imidazo[1,5-a][1,2,4]triazolo[4,3-
d][1,4]benzodiazepines is outlined in Scheme 5 where the efficient
condensation of 1H-benzo[d][1,3]oxazine-2,4-diones with N-pro-
tected glycine derivatives proceeds efficiently and is convenient
for large scale preparation. This allowed imidazole formation to
proceed first followed by acid (cat. TfOH in TFA) deprotection of
the preferred 2,4-dimethoxybenzyl protecting group to proceed
smoothly. Amide activation via formation and isolation of the imi-
doyl chloride followed by cyclisation to the 1,2,4-triazole with
formylhydrazine completed an overall efficient reaction sequence
(Scheme 5).19
Table 8
Molecular property profile of selected compounds
Compd
Solubilitya
g/mL)
Log D7.4
Permeation coefficient
(10ꢀ6cm sꢀ1) (PAMPA)
Cl (mL/min/mg
protein)a
(rat/human)
(
l
54
63
64
15
231
100
1.83
0.73
1.35
5.59
2.87
6.79
19/50
4/8
3/0
a
See Table 2.
normally achieved for all subtypes within this class). The introduc-
tion of the terminal acetylene moiety 41 again proved to give a bal-
anced profile but replacement of the ester to typical isosteres or
amides resulted in functional agonism at the GABAA
a5 receptor
subtype for all R1 substituents. Pleasingly, we were able to replace
the ester with a Cl substituent but this was met with a 15-fold drop
in potency although the selectivity versus the GABAA
a
1 receptor
From the possible triazole isomers the imidazo[1,5-a][1,2,4]-
triazolo[1,5-d][1,4]benzodiazepine class proved to have overall
the most favoured balanced profile and Table 7 details the initial
development of this class.20 Initially we were excited to observe
subtype was enhanced. This trend was also observed with our pre-
ferred Br substituent as R1. Examination of the opportunities for
the R3 substituent variation revealed that all substituents except
phenyl resulted in an agonistic functional response at the GABAA
that all derivatives had excellent potency at the GABAA a5 receptor
a5 receptor subtype. Attempts to introduce F and/or OMe substit-
subtype and concomitant good levels of inverse agonism—nearly
uents onto the phenyl ring (not shown in Table 6) were met with
all full inverse agonists. In this class we were able to successfully
reduced affinity (>300 nM) and pyridine isomer replacements were
N
N
a)
c)
b)
N
N
N+
NO2
N
NO2
H
N
a-b)
Cl
c-d)
Cl
I
CN
Br
N
N
Br
Cl
Cl
N
N
OH
O
O
HN
N
N
N
f)
N
O
CO2Et
H
N
N
N
e)
f)
N
Br
N
X
Br
N
Cl
N
Cl
N
N
N
N
X = CO2Et
d)
e)
23
N
33
X = CH2OH
X = CH2Cl
Scheme 3. Synthesis of 23. Reagents and conditions: (a) Me2NCH(OMe)2, toluene,
100 °C, 3 h, 84%; (b) hydrazine, EtOH, rt, 16 h, 87%; (c) H2, Pd/C, EtOAC, rt, 3 h, 82%;
(d) chloroacetyl chloride, pyridine, THF, 5–10 °C, 1 h then HCl. Dioxane, 90 °C, 2 h,
69%; (e) NaOH, dioxane, rt, 2 h, 80%; (f) POCl3, N,N-dimethyl-p-toluidine, 1,2-
dichloroethane, 85 °C, then add KOBu-t, ethyl isocyanoacetate, NMP, ꢀ20 °C to rt,
30 min, 67%.
Scheme 4. Synthesis of 33. Reagents and conditions: (a) see Ref. X; (b) hydroxyl-
amine HCl, triethylamine, DMF, 3 h, 120 °C, 68%; (c) ethyl propiolate, K2CO3, EtOH,
80 °C, 5 h, 35%; (d) LiAlH4, THF, rt, 1 h, 54%; (e) SOCl2, rt, on, 92%; (f) NaH,
pyrolidinone, DMF, rt, 3 h, 32%.

G. Achermann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5746–5752
5751
O
H
N
OMe
a)
MeO
b)
N
Cl
N
H
CO2H
CO2H
H2N
O
OMe
67
MeO
N
N
CO2Et
CO2Et
e)
N
c)
CO2Et
Me2N
68
N
N
d)
N
Cl
N
H
Cl
O
O
OMe
N
MeO
CO2Et
f)
N
CO2Et
N
N
N
Cl
N
Cl
Cl
N
N
36
Scheme 5. Synthesis of 36. Reagents and conditions: (a) 2,4-dimethoxybenzaldehyde, NaOH, MeOH then H2, Pd/C, MeOH, rt, 24 h, 85%; (b) 6-chloro-1H-benzo[d][1,3]oxa-
zine-2,4-dione, xylene, 140 °C, 2 h, 70%; (c) 67, POCl3, N,N-dimethyl-p-toluidine, toluene, reflux, 2.5 h then 68, LiHMDS, THF, ꢀ70 °C to rt, 1 h then AcOH, H2O, ꢀ30 °C to reflux,
30 min., 70%; (d) cat. TfOH, TFA, rt, 2 h, 85%; (e) POCl3, N,N-dimethyl-p-toluidine, chlorobenzene, reflux, 2.5 h, 87%; (f) formylhydrazine, Hünigs base, chlorobenzene, reflux,
1 h, 80%.
Pharmacol., Biochem. Behav. 2008, 90, 58; (e) Möhler, H.; Rudolph, U. Drug Disc.
Today: Therapeutic Strategies 2004, 1, 117.
5. (a) Da Settimo, F.; Taliani, S.; Trincavelli, M. L.; Montali, M.; Martini, C. Curr.
replace the ester substituent by H or ethyl and maintain full in-
verse agonism although with low and modest levels of binding
selectivity, respectively. Additional successful ester replacements
and the synthetic routes to imidazo[1,5-a][1,2,4]-triazolo[1,5-
d][1,4]benzodiazepines will be discussed in more detail.13
Overall the class showed favourable molecular properties (Table
8) for a CNS lead series and we therefore decided to focus our lead
optimisation efforts on the imidazo[1,5-a][1,2,4]-triazolo[1,5-
d][1,4]benzodiazepine class and the subsequent papers will de-
scribe in detail our successful endeavours to identify further potent
Med. Chem. 2007, 14, 2680; (b) McKernan, R. M.; Whiting, P. J. Trends Neurosci.
1996, 19, 139; (c) Olsen, R. W.; Sieghart, W. Pharmacol. Rev. 2008, 60, 243.
6. (a) Jensen, L. H.; Stephens, D. N.; Sarter, M.; Petersen, E. N. Brain Res. Bull. 1987,
19, 359; (b) Duka, T.; Edelmann, V.; Schutt, B.; Dorrow, R. Psychopharmacology
1988, 6, 246; (c) Venault, P.; Chapouthier, G.; de Carvalho, L. P.; Simiand, J.;
Morre, M.; Dodd, R. H.; Rossier, J. Nature 1986, 321, 864; (d) Dorow, D.;
Horowski, R.; Paeschelke, G.; Amin, M.; Braestrup, C. Lancet 1983, July 9, 98; (e)
Evans, A. K.; Lowry, C. A. CNS Drug Rev. 2007, 13, 475.
7. Howell, O.; Atack, J. R.; Dewar, D.; McKernan, R. M.; Sur, C. Neuroscience 2000,
98, 669.
8. (a) Romanelli, M. N.; Gualtieri, F. Exp. Opin. Ther. Patents 2007, 17, 1365; (b)
Carling, R. W.; Moore, K. W.; Street, L. J.; Wild, D.; Isted, C.; Leeson, P. D.;
Thomas, S.; O’Connor, D.; McKernan, R. M.; Quirk, K.; Cook, S. M.; Atack, J. R.;
Wafford, K. A.; Thompson, S. A.; Dawson, G. R.; Ferris, P.; Castro, J. L. J. Med.
Chem. 2004, 47, 1807; (c) Sternfeld, F.; Carling, R. W.; Jelley, R. A.;
Ludduwahetty, T.; Merchant, K. J.; Moore, K. W.; Reeve, A. J.; Street, L. J.;
O’Connor, D.; Sohal, B.; Atack, J. R.; Cook, S.; Seabrook, G.; Wafford, K.;
Tattersall, F. D.; Collinson, N.; Dawson, G. R.; Castro, J. L.; MacLeod, A. M. J. Med.
Chem. 2004, 47, 2176; (d) Chambers, M. S.; Atack, J. R.; Carling, R. W.; Collinson,
N.; Cook, S. M.; Dawson, G. R.; Ferris, P.; Hobbs, S. C.; O’Connor, D.; Marshall, G.;
Rycroft, W.; MacLeod, A. M. J. Med. Chem. 2004, 47, 5829; (e) van Neil, M. B.;
Wilson, K.; Adkins, C. H.; Atack, J. R.; Castro, J. L.; Clarke, D. E.; Fletcher, S.;
Gerhard, U.; Mackey, M. M.; Malpas, S.; Maubach, K.; Newman, R.; O’Connor,
D.; Pillai, G. V.; Simpson, P. B.; Thomas, S. R.; MacLeod, A. M. J. Med. Chem. 2005,
48, 6004; (f) Jones, P.; Atack, J. R.; Braun, M. P.; Cato, B. P.; Chambers, M. S.;
O’Connor, D.; Cook, S. M.; Hobbs, S. C.; Maxey, R.; Szekeres, H. J.; Szeto, N.;
Wafford, K. A.; MacLeod, A. M. Bioorg. Med. Chem. Lett. 2006, 16, 872; (g)
O’Connor, D.; Jones, P.; Chambers, M. S.; Maxey, R.; Szekeres, H. J.; Szeto, N.;
Scott-Stevens, P.; MacLeod, A. M.; Braun, M.; Cato, B. Xenobiotica 2006, 36, 315.
9. (a) 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; (b) Nutt, D. J.; Besson, M.; Wilson, S. J.;
Dawson, G. R.; Lingford-Hughes, A. R. Neuropharmacology 2007, 53, 810.
10. (a) Merschman, S. A.; Rose, M. J.; Pearce, G. E. S.; Woolf, E. J.; Schaefer, B. H.;
Huber, A. C.; Musson, D. G.; Petty, K.; Rush, D. J.; Varsolona, R. J.; Matuszewski,
B. K. Pharmazie 2005, 60, 359; (b) Atack, J. R. CNS Neurosci. Ther. 2008, 14, 25; (c)
Xue, L.; Lin, L.; Hsieh, J. Y.-K.; Matuszewski, B. K. J. Liq. Chromatogr. Related
Technol. 2004, 27, 689.
and binding selective GABAA
from this class.
In summary we have identified through iteratively designed
screening cycles a number of novel binding and selective inverse
agonist series for the GABAA
dazo[1,5-a][1,2,4]-triazolo[1,5-d][1,4]benzodiazepine class being
a5 receptor subtype inverse agonists
a5 receptor subtype with the imi-
the most preferred for further lead optimisation.
Acknowledgements
Special thanks to Brigitte Algeyer, Béatrice David, Cécile Guizi-
ani, Rachel Haab, Marie-Laurence Harle-Yge, Maria Karg, Marie
Claire Pflimlin, Pascal Pflimlim, Regina Wolf for technical excel-
lence in their work.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.07.153.
References and notes
1. Berchtold, N. C.; Cotman, C. W. Neurobiol. Aging 1998, 19, 173.
2. McShane, R.; Areosa Sastre, A.; Minakaran, N. Cochrane Database Syst. Rev. 2006,
CD003154.
3. Birks, J.; Harvey, R. J. Cochrane Database Syst. Rev. 2006, CD001190.
4. (a) Maubach, K. Current Drug Targets – CNS & Neurological Disorders 2003, 2,
233; (b) Maubach, K. A. Drugs Future 2006, 31, 151. and references cited within;
(c) Rissmann, R. A.; De Blas, A. L.; Armstrong, D. M. J. Neurochem. 2007, 103,
1285; (d) Möhler, H.; Rudolph, U.; Boison, D.; Singer, P.; Feldon, J.; Yee, B. K.
11. S-8510 (a) Kawasaki, K.; Eigyo, M.; Ikeda, M.; Kihara, T.; Koike, K.; Matsuchita,
A.; Murata, S.; Shiomi, T.; Takada, S.; Yasui, M. Prog. Neuropsychopharmacol.
Biol. Psychiatry 1996, 20, 1413; (b) AC-3933/SX-3933/Resequinil: Furukawa, K.;
Kurumiya, S.; Okimoto, K., Ohno, K. WO Patent 2001098300A1.
12. (a) Atack, J. R.; Pike, A.; Clarke, A.; Cook, S. M.; Sohal, B.; McKernan, R. M.;
Dawson, G. R. Drug Metab. Dispos. 2006, 34, 887; (b) Quirk, K.; Blurton, P.;
Fletcher, S.; Leeson, P.; Tang, F.; Mellilo, D.; Ragan, C. I.; McKernan, R. M.
Neuropharmacology 1996, 35, 1331; (c) Skolnick, P.; Hu, R. J.; Cook, C. M.; Hurt,

5752
G. Achermann et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5746–5752
S. D.; Trometer, J. D.; Liu, R.; Huang, Q.; Cook, J. M. J. Pharmacol. Exp. Ther. 1997,
14. Büttelmann, B.; Ballard, T.; Fischer, H.; Gasser, R.; Hernandez, M.-C.; Knoflach,
F.; Knust, H.; Stalder, H.; Thomas, A. W.; Trube, G. Bioorg. Med. Chem. Lett.,
doi:10.1016/j.bmcl.2009.08.027.
15. Ballard, T. M.; Knoflach, F.; Prinssen, E.; Borroni, E.; Vivian, J. A.; Basile, J.;
Gasser, R.; Moreau, J.-L.; Wettstein, J. G.; Büttelmann, B.; Knust, H.; Thomas, A.
W.; Trube, G.; Hernandez, M.-C. Psychopharmacology 2009, 202, 207.
16. Masciadri, R.; Thomas, A. W.; Wichmann, J. WO Patent 2002094834A1: Chem.
Abstr. 2002, 137, 384862.
283, 488; (d) Bailey, D. J.; Tetzlaff, J. E.; Cook, S. M.; Smith, A. J.; He, X.;
Helmstetter, F. J. Neurobiol. Learn Membr. 2002, 78, 1; (e) Street, L. J.; Sternfield,
F.; Jelley, R. A.; Reeve, A. J.; Carling, R. W.; Moore, K. W.; McKernan, R. M.; Sohal,
B.; Cook, S.; Pike, A.; Dawson, G. R.; Bromidge, F. A.; Wafford, K. A.; Seabrook, G.
R.; Thompson, S. A.; Marshall, G.; Pillai, G. V.; Castro, J. L.; Atack, J. R.; MacLeod,
A. M. J. Med. Chem. 2004, 47, 3642; (f) 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. J. Med. Chem. 2002, 45,
1176; (g) Chambers, M. S.; Atack, J. R.; Broughton, H. B.; Collinson, N.; Cook, S.;
Dawson, G. R.; Hobbs, S. C.; Marshall, G.; Maubach, K. A.; Pillai, G. V.; Reeve, A.
J.; MacLeod, A. M. J. Med. Chem. 2003, 46, 2227; (h) Szekeres, H. J.; Atack, J. R.;
Chambers, M. S.; Cook, S. M.; Macaulay, A. J.; Pillai, G. V.; MacLeod, A. M. Bioorg.
Med. Chem. Lett. 2004, 14, 2871; (i) Savic, M. M.; Clayton, T.; Furtmüller, R.;
Gavrilovic, I.; Samardzic, J.; Savic, S.; Huck, S.; Sieghart, W.; Cook, J. M. Brain
Res. 2008, 1208, 150.
17. Typically in this class binding selectivity versus GABAA
similar.
axb3c2 (x = 1, 2, 3) is
18. Gerecke, M.; Kyburz, E.; Borer, R.; Gassner, G. Heterocycles 1994, 39, 693.
19. Masciadri, R.; Thomas, A. W.; Wichmann, J. WO Patent 200240487A1: Chem.
Abstr. 2002, 136, 386146.
20. (a) Knust, H.; Stadler, H.; Thomas, A. W. U.S. Patent 20060079507A1: Chem.
Abstr. 2006, 144, 370129.; (b) Knust, H.; Thomas, A. W. U.S. Patent
20060084642A1: Chem. Abstr. 2006, 144, 390949.; (c) Knust, H.; Thomas, A.
W. U.S. Patent 20060084801A1: Chem. Abstr. 2006, 144, 412549.; (d)
Büttelmann, B.; Knust, H.; Thomas, A. W. U.S. Patent 2006128691A1: Chem.
Abstr. 2006, 145, 46099.
13. Knust, H.; Büttelmann, B.; Ballard, T.; Fischer, H.; Gasser, R.; Hernandez, M.-C.;
Knoflach, F.; Stalder, H.; Thomas, A. W.; Trube, G.; Bioorg. Med. Chem. Lett.,
doi:10.1016/j.bmcl.2009.08.053.