Design, Synthesis, and Biological Evaluation of Imidazo[1,5-a]quinoline as Highly Potent Ligands of Central Benzodiazepine Receptors

Article abstract of DOI:10.1021/acs.jmedchem.6b00034

A series of imidazo[1,5-a]quinoline derivatives was designed and synthesized as central benzodiazepine receptor (CBR) ligands. Most of the compounds showed high CBR affinity with K_i values within the submicromolar and subnanomolar ranges with interesting modulations in their structure-affinity relationships. In particular, fluoroderivative 7w (K_i = 0.44 nM) resulted in the most potent ligand among the imidazo[1,5-a]quinoline derivatives described so far. Overall, these observations confirmed the assumption concerning the presence of a large though apparently saturable lipophilic pocket in the CBR binding site region interacting with positions 4 and 5 of the imidazo[1,5-a]quinoline nucleus. The in vivo biological characterization revealed that compounds 7a,c,d,l,m,q,r,w show anxiolytic and antiamnestic activities without the unpleasant myorelaxant side effects of the classical 1,4-BDZ. Furthermore, the effect of 7l,q,r, and 8i in lowering lactate dehydrogenase (LDH) release induced by ischemia-like conditions in rat brain slices suggested neuroprotective properties for these imidazo[1,5-a]quinoline derivatives.

Full text of DOI:10.1021/acs.jmedchem.6b00034

Subscriber access provided by KUNGL TEKNISKA HOGSKOLAN
Article
Design, Synthesis, and Biological Evaluation of Imidazo[1,5-a]quinoline
as Highly Potent Ligands of Central Benzodiazepine Receptors
Andrea Cappelli, Maurizio Anzini, Federica Castriconi, Giorgio Grisci, Marco Paolino, Carlo
Braile, Salvatore Valenti, Germano Giuliani, Salvatore Vomero, Angela Di Capua, Laura
Betti, Gino Giannaccini, Antonio Lucacchini, Carla Ghelardini, Lorenzo Di Cesare Mannelli,
Maria Frosini, Lorenzo Ricci, Gianluca Giorgi, Maria Paola Mascia, and Giovanni Biggio
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00034 • Publication Date (Web): 16 Mar 2016
Downloaded from http://pubs.acs.org on March 17, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Design, Synthesis, and Biological Evaluation of Imidazo[1,5ꢀa]quinoline as
Highly Potent Ligands of Central Benzodiazepine Receptors
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
a
a
a
a
Andrea Cappelli, * Maurizio Anzini, * Federica Castriconi, Giorgio Grisci, Marco Paolino,
a
a
a
a
a
Carlo Braile, Salvatore Valenti, Germano Giuliani, Salvatore Vomero, Angela Di Capua, Laura
b
b
b
c
c
Betti, Gino Giannaccini, Antonio Lucacchini, Carla Ghelardini, Lorenzo Di Cesare Mannelli,
d
d
a
e
e
Maria Frosini, Lorenzo Ricci, Gianluca Giorgi, Maria Paola Mascia, and Giovanni Biggio.
a
Dipartimento di Biotecnologie, Chimica e Farmacia and European Research Centre for Drug
Discovery and Development, Università degli Studi di Siena, Via A. Moro 2, 53100 Siena, Italy;
b
Dipartimento di Farmacia, Università di Pisa, Via Bonanno 6, 56126 Pisa, Italy;
c
Dipartimento di Neuroscienze, Psicologia, Area del Farmaco e Salute del Bambino
(NEUROFARBA) ꢀ Sez. Farmacologia e Tossicologia, Università di Firenze, Viale G. Pieraccini 6,
50139 Firenze, Italy;
d
Dipartimento di Scienze della Vita, Università degli Studi di Siena, Via A. Moro 2, 53100 Siena,
Italy;
e
Istituto di Neuroscienze, Consiglio Nazionale delle Ricerche, Cittadella Universitaria, S.S. 554 ꢀ
Km. 4.500, 09042 Monserrato (CA), Italy
RECEIVED DATE
ACS Paragon Plus Environment

Page 3 of 64
Journal of Medicinal Chemistry
ABSTRACT. A series of imidazo[1,5ꢀa]quinoline derivatives was designed and synthesized as
central benzodiazepine receptor (CBR) ligands. Most of the compounds showed high CBR affinity
with Ki values within the submicromolar and subnanomolar ranges with interesting modulations in
their structureꢀaffinity relationships. In particular, fluoroderivative 7w (Ki = 0.44 nM) resulted the
most potent ligand among the imidazo[1,5ꢀa]quinoline derivatives described so far. Overall, these
observations confirmed the assumption concerning the presence of a large though apparently
saturable lipophilic pocket in the CBR binding site region interacting with positions 4 and 5 of the
imidazo[1,5ꢀa]quinoline nucleus. The in vivo biological characterization revealed that compounds
7a,c,d,l,m,q,r,w show anxiolytic and antiamnestic activities without the unpleasant myorelaxant
sideꢀeffects of the classical 1,4ꢀBDZ. Furthermore, the effect of 7l,q,r, and 8i in lowering lactate
dehydrogenase (LDH) release induced by ischemiaꢀlike conditions in rat brain slices suggested
neuroprotective properties for these imidazo[1,5ꢀa]quinoline derivatives.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 64
INTRODUCTION
The neurotransmitting action of γꢀaminobutyric acid (GABA) at the GABA
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A
chloride channel
1
complex modulates the excitability of many central nervous system (CNS) pathways. GABA
A
receptors are ligandꢀgated ion channels (LGICs) belonging to the Cysꢀloop superfamily, the same as
nicotinic acetylcholine, glycine, zincꢀactivated, and 5ꢀHT receptors. Cysꢀloop receptors are
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
characterized by the assembly of five subunits, which form pentameric arrangements around a
1
central ionꢀconducting pore and are the targets of many drugs. The function of GABA
A
receptors is
regulated, in addition to the agonist binding site, by allosteric sites interacting with a large variety of
2
agents. Positive modulators of the GABA
A
receptors, such as the classical 1,4ꢀbenzodiazepine
(BDZ) diazepam (1, Figure 1), are therapeutically employed as sedatives, muscle relaxants,
anxiolytics and anticonvulsants, whereas negative GABA modulators (i. e. BDZ inverse agonists)
A
3
ꢀ6
show anxiogenic and convulsant effects. Finally, neutral modulators, such as the imidazo[1,5ꢀ
a][1,4]benzodiazepines flumazenil (2), bind to GABA receptor but have no intrinsic activity at the
central benzodiazepine receptor (CBR). Thus, flumazenil is recognized to antagonize the activity of
A
both positive and negative GABA
A
modulators acting via the CBR.
N
O
N
R
O
N
N
N
O
N
Y
N
Cl
F
N
3
O
X
diazepam (
1)
flumazenil (2)
R : COOEt, COOtBu, CN, CONH2
X: H, F, Cl, CH , OCH , NO
3
3
2
Y: H, F, Cl
N
N
N
O
O
N
R1
N
N
N
R2
R3
O
O
R6
N
Cl
Cl
R5 R4
O
N
O
N
N
N
4
H
H
PNU-100076 (4a
)
PNU-101017 (5)
Figure 1. Structure of reference compounds.
ACS Paragon Plus Environment

Page 5 of 64
Journal of Medicinal Chemistry
7
ꢀ10
Since positive modulators show amnesic effects in animal and man,
negative modulators were
1
2
3
4
5
6
7
8
9
11ꢀ12
assumed to possess procognitive properties.
However, the use of nonꢀselective CBR inverse
agonists in the treatment of neurological disorders associated with cognitive impairment was limited
13
by their anxiogenic and convulsant effects.
A large number of subunits (i. e. α1−6, β1−4, γ1−4, δ, ε, π, θ and ρ1−3) has been cloned and sequenced,
but most of GABA receptors are composed of α, , and ꢀsubunits arranged in a 2:2:1
stoichiometry. In fact, among the multitude of theoretically possible combinations deriving from
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A
β
γ
1
4
1
3
5
the coꢀassembly of the subunits, only the receptor subtypes containing a
conjunction with α , α , α , or α appear to bind BDZ ligands with significant affinity. The CBR
binding domain is assumed to be located at the interface between α and subunits, which contribute
to the building of the active site with their amino acid residues. Investigations based on molecular
genetic or pharmacological approaches suggested that α subunit is involved in the sedative and
muscle relaxant effects of the nonꢀselective BDZ agonists, whereas α or α can be responsible
The recognition of the pharmacological and
receptor subtype functions has stimulated new interest
γ
2
or
γ
subunit in
1
2
3
5
γ
1
6
1
2
3
1
7ꢀ19
for the anxiolytic and anticonvulsant effects.
physiological roles of α subunits in GABA
A
in this receptor system as the target for the development of drugs showing fewer sideꢀeffects with
respect to the classical benzodiazepines (e. g. nonꢀsedating anxiolytics) or possessing different
indications with respect to the classical benzodiazepines (e. g. analgesics, cognitionꢀenhancing
20ꢀ27
drugs).
A
A number of compounds has been developed that show GABA receptor subtype
2
0
selectivity. In particular, subtypeꢀselective GABA receptor ligands were obtained either by
A
selective binding (i. e. by forming a receptorꢀligand complex with a particular receptor subtype) or
by selective efficacy (i. e. by eliciting a biological response after binding to the receptor). These two
2
0
properties are both important in defining the potency profile.
Very interestingly, the full range of intrinsic efficacies observed in the series of imidazo[1,5ꢀ
a][1,4]benzodiazepine derivatives 3 was modulated in a rather subtle manner by the substitution
28,29
pattern.
A similar behavior was observed when the sevenꢀmembered ring of the benzodiazepine
system was contracted as in the series of imidazo[1,5ꢀa]quinoxaline derivatives 4, which were
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 64
3
0ꢀ35
developed in the 90’s in Upjohn laboratories.
researchers led to the proposal of the existence of a second low affinityꢀbinding site on GABA
receptors, the occupancy of which (at high drug concentrations) may reverse the positive allosteric
The large body of work performed by the Upjohn
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A
3
6
action on CBR and potentially minimize dependence and abuse liability.
Among the large number of imidazo[1,5ꢀa]quinoxaline derivatives developed, compound 4a (see
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1) was shown to induce a negative allosteric modulation via this second low affinity binding
3
6
site.
Based on the huge amount of structureꢀactivity relationship (SAR) data on imidazo[1,5ꢀ
a]quinoxaline derivatives 4, the structure of 4a was easily translated into the imidazo[1,5ꢀ
a]quinoline one of 5 (see Figure 1), which was considered as a drug candidate for the treatment of
anxiety, but its development was discontinued for safety reason (i. e. centrally mediated respiratory
37
depression having lethal effects). The identification of 5 as a candidate for development studies
was apparently performed among a limited set of imidazo[1,5ꢀa]quinoline derivatives so that the
available SAR data on this class of CBR ligands is limited. In general, the analysis of the available
SAR data suggested that bulky substituents are tolerated by CBR binding site when they are located
in the ligand region corresponding to positions 4 and 5 of the imidazo[1,5ꢀa]quinoline nucleus
3
8,39
consistently with the results obtained with 5ꢀHT receptor ligands based on quinoline structure 6.
3
39
Similarly, these bulky substituents appeared to play a role in modulating the intrinsic efficacy.
These observations, together with the structural analogies existing between GABA and 5ꢀHT
receptors, led us to apply the approach used in studying 5ꢀHT
A
3
3
receptors to the characterization of
CBR binding features by means of the design, synthesis, and pharmacological characterization of
imidazo[1,5ꢀa]quinoline derivatives 7, 8, and 9, in comparison with reference imidazo[1,5ꢀ
a]quinoxalines 10 (Figure 2).
ACS Paragon Plus Environment

Page 7 of 64
Journal of Medicinal Chemistry
N
N
N
O
O
N
1
2
3
4
5
6
7
8
9
X
X
Z
R
R
2
1
R
1
N
N
7
O
Y
8
9
6
N
N
N
R1
8
7
3
4
5
R
2
R2
N
R
3
R3
N
O
O
N
N
N
6
7-9
O
O
R
X
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
10
O
CH
3
Figure 2. Design of imidazo[1,5ꢀa]quinoline 7-9 starting from piperazinylquinoline 5ꢀHT
ligands 6.
3
receptor
RESULTS AND DISCUSSION
Chemistry. The preparation of target imidazo[1,5ꢀa]quinoline derivatives was performed by
imidazoꢀannulation of suitable 2ꢀchloroquinoline derivatives 11ꢀ24 with ethyl isocyanoacetate or
tertꢀbutyl isocyanoacetate in the presence of potassium tertꢀbutoxide providing 7aꢀi,kꢀm,oꢀy
(
Scheme 1) in the yields reported in Table 1. The structure of 7p,s,t was confirmed by
crystallographic studies (see Supporting Information).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Scheme 1. Imidazoꢀannulation of 2ꢀchloroquinoline derivatives 11-24 to target imidazo[1,5ꢀ
a]quinoline derivatives 7a-i,k-m,o-y.
Page 8 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
N
O
N
Cl
R
N
i
X
Z
Z
R
O
Y
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
Y
1
1-
24
7a-i,k-m,o-y
Reagents: (i) CNCH
2
COOC
2
H
5
or CNCH
2
3 3
COOC(CH ) , tertꢀBuOK, DMF. Substituents: see Table
1
.
Table 1. Preparation of target compounds 7a-i,k-m,o-y.
target
X
Y
N(CH
N(CH
Z
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
F
R
H
H
H
H
H
H
H
H
starting
11
11
12
12
13
13
14
14
15
16
17
17
18
18
19
19
20
20
21
21
22
23
24
source
Yield
35%
59%
10%
17%
72%
48%
56%
36%
25%
94%
60%
46%
26%
46%
27%
15%
21%
55%
5%
7a
OC
OC(CH
OC
OC(CH
OC
OC(CH
OC
OC(CH
2
H
5
3
)
)
2
2
Exp Sect
Exp Sect
Exp Sect
Exp Sect
see ref 38
see ref 38
see ref 38
see ref 38
see ref 40
Scheme 2
see ref 38
see ref 38
see ref 38
see ref 38
see ref 38
see ref 38
see ref 38
see ref 38
see ref 38
see ref 38
Scheme 3
Scheme 3
Scheme 3
7b
3
)
)
)
)
3
3
3
3
3
7c
2
H
5
N(CH
N(CH
3
)nꢀC
)nꢀC
4
4
H
H
9
7d
3
3
9
7e
2
H
5
N(nꢀC
N(nꢀC
3
H
7
H
7
)
)
C
C
2
7f
7g
7h
3
3
2
2
H
5
N(CH
N(CH
N(C
OC(CH
3
)CH
2
6
H
H
5
3
3
)CH
2
6
5
7i
OC
OC
OC
2
2
2
H
H
H
5
2
H
4
)
2
NCH
3
H
7k
5
3
)
)
)
C
C
3
CH
CH
CH
CH
CH
3
3
3
3
3
7l
5
3
N(nꢀC
N(nꢀC
3
H
7
H
7
2
2
7m
OC(CH
OC
OC(CH
OC
OC(CH
OC
OC(CH
OC
OC(CH
)
)
)
)
)
3
3
3
3
3
3
7
o
p
q
2
H
5
3
N(CH
N(CH
3
)CH
)CH
2
2
6
6
H
H
5
7
7
3
5
2
H
5
N(nꢀC
N(nꢀC
3
H
7
H
7
)
)
2
C
C
2
H
5
H
5
7r
3
3
2
2
7s
2
H
5
N(C
N(C
2
H
H
5
)
)
2
nꢀC
nꢀC
nꢀC
nꢀC
3
3
3
3
H
7
H
7
H
H
7t
3
2
5
2
7u
2
H
5
N(nꢀC
N(nꢀC
N(nꢀC
3
3
3
H
7
)
)
)
C
C
2
2
2
7
7v
7w
7x
3
H
7
H
7
7
4%
OC
OC
OC
H
H
H
H
H
H
21%
15%
62%
2
2
2
5
5
N(CH
N(CH
3
)CH
)CH
2
6
H
H
5
F
Br
7y
5
3
2
6
5
The starting 2ꢀchloroquinoline derivatives were either compounds known (see Table 1) or prepared
38,40
by standard methodology.
On the other hand, target propargylamide derivative 7n was
synthesized from tertꢀbutyl ester 7k via acid 7j (Scheme 2).
ACS Paragon Plus Environment

Page 9 of 64
Journal of Medicinal Chemistry
Scheme 2. Synthesis of target derivatives 7j,k,n.
1
2
3
4
5
6
7
8
9
N
Cl
N
Cl
i
O
Cl
O
O
ii
25
16
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
N
O
O
O
N
N
iii
O
O
O
O
OH
7
j
7k
iv, v
N
O
N
O
O
N
7
n
Reagents: (i) tertꢀBuOK, THF; (ii) CNCH COOC H , tertꢀBuOK, DMF; (iii) HCOOH; (iv) SOCl ;
2
2
5
2
2 2
(v) Nꢀmethylpropargylamine, CH Cl .
3
8
Acyl chloride 25 was reacted with potassium tertꢀbutoxide in dry THF to obtain ester 16, which
was used in the aboveꢀdescribed imidazoꢀannulation with ethyl isocyanoacetate affording diester 7k.
The cleavage of the tertꢀbutyl ester moiety of the latter with formic acid gave the expected
carboxylic acid 7j, which was in turn transformed via acyl chloride into the expected
propargylamide 7n.
The target imidazo[1,5ꢀa]quinoline derivatives bearing substituents in position 7 of the tricyclic
nucleus was carried out as outlined in Scheme 3.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Scheme 3. Synthesis of the target imidazo[1,5ꢀa]quinoline derivatives bearing substituents in
position 7.
Page 10 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
H
N
O
N
Cl
i
Z
Z
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
OH
O
OH
ii, iii
2
6a,b
27a,b
N
N
Cl
O
N
iv
v
O
Z
Z
O
Y
O
Y
7w,x,y
2
2-24
N
N
O
O
O
N
vi
N
O
Si
O
N
O
N
7aa
7z
Reagents: (i) POCl
3 2 2 2 2 2 5
; (ii) SOCl ; (iii) amine, CH Cl , TEA; (iv) CNCH COOC H , tertꢀBuOK,
DMF; (v) trimethylsilylacetylene, Pd(PPh ) (AcO) , TEA; (vi) Bu NF, THF. Substituents: Z = F in
3
2
2
4
2
2, 23, 26a, and 27a; Z = Br in 24, 26b, and 27b; Y = N(nꢀC
3 7 2
H ) in 7w and 22; Y =
N(CH )CH in 7x,y, 23 and 24.
3
2 6 5
C H
4
1
The appropriately substituted quinolinone derivatives 26a,b were converted into the corresponding
2
ꢀchloroquinoline derivatives 27a,b by reaction with phosphorous oxychloride. The amidation of the
carboxyl group of 27a,b afforded the expected amides 22-24, which were used in the aboveꢀ
described imidazoꢀannulation with ethyl isocyanoacetate or tertꢀbutyl isocyanoacetate to obtain
target compounds 7w,x,y. Bromoderivative 7y was then used in Sonogashira coupling reaction with
trimethylsilylacetylene to afford 7aa, which was promptly desilylated with into 7z.
ACS Paragon Plus Environment

Page 11 of 64
Journal of Medicinal Chemistry
Most of the target imidazo[1,5ꢀa]quinoline derivatives bearing the phenyl substituents in position 5
of the tricyclic nucleus was accomplished as sketched in Scheme 4, while the preparation of
compound 8i is described in Scheme 5.
1
2
3
4
5
6
7
8
9
Scheme 4. Synthesis of the target imidazo[1,5ꢀa]quinoline derivatives bearing the phenyl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
substituents in position 5.
N
O
N
Cl
R
N
i
X
R
28-31
8a,b,e,g,h
ii
N
N
O
O
N
iii, iv
N
OH
X
R
R
3
2, 33
8c,d,f
Reagents: (i) CNCH
,3,5,6ꢀtetrafluorophenol, EDC, Na
for 8d,f), THF. Substituents: R = H in 8aꢀd, 28, and 32; R = CH
in 8g and 30; R = CH COOC in 8h and 31; X = OC in 8a,g,h; X = OC(CH
N(CH in 8c; X = N(CH )CH in 8d,f.
2
COOC
2
H
5
or CNCH
2
COOC(CH
3 3
) , tertꢀBuOK, DMF; (ii) HCOOH; (iii)
2
2
CO , H
3
2
O, CH
3
CN; (iv) HN(CH
3
)
2
for 8c (or HN(CH
3
)CH
2
C
6
H
5
5
3
in 8e,f, 29, and 33; R = COOC
2
H
2
2
H
5
2
H
5
3
)
3
in 8b,e; X =
3
)
2
3
2 6 5
C H
42
Imidazoꢀannulation of suitable 2ꢀchloroꢀ4ꢀphenylquinoline derivatives 28-31
with ethyl
isocyanoacetate or tertꢀbutyl isocyanoacetate in the presence of potassium tertꢀbutoxide gave target
derivatives 8a,b,e,g,h. The cleavage of the tertꢀbutyl ester moiety of esters 8b,e with formic acid
gave carboxylic acid derivatives 32 and 33, which were converted via 2,3,5,6ꢀtetrafluorophenyl
esters into the desired amides 8c,d,f.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 64
Scheme 5. Synthesis of the imidazo[1,5ꢀa]quinoline derivative 8i.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
N
Cl
O
O
3
0
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
Cl
N
Cl
O
ii
OH
34
35
Si
iii
iv
N
N
O
O
O
N
N
O
O
OH
3
6
8i
Si
Reagents: (i) LiAlH
4 2 2 2 2 5
, THF; (ii) TBDMSCl, imidazole, CH Cl ; (iii) CNCH COOC H , tertꢀBuOK,
DMF; (iv) Bu NF, THF.
4
Lithium aluminium hydride reduction of ester 30 gave hydroxymethylquinoline derivative 34, which
was first protected by reaction with tertꢀbutyldimethylsilyl chloride (TBDMSCl) and then submitted
to the conditions of the imidazoꢀannulation with ethyl isocyanoacetate or tertꢀbutyl isocyanoacetate
to afford imidazo[1,5ꢀa]quinoline 36, which was promptly desilylated into target 8i.
4
3
Finally, the imidazoꢀannulation was applied also to chloroderivative 37 to obtain tetracyclic target
31
compound 9, and to quinoxalinone derivatives 38 and 39. However, these latter intermediates
required a more complex reaction pathway consisting of a first activation step with diethyl
chlorophosphate followed by a second step of annulation with ethyl isocyanoacetate or tertꢀbutyl
isocyanoacetate in the presence of potassium tertꢀbutoxide to obtain reference imidazo[1,5ꢀ
3
1
a]quinoxaline 10a-d (Scheme 6). The structure of 10a-c was confirmed by crystallographic studies
(see Supporting Information).
ACS Paragon Plus Environment

Page 13 of 64
Journal of Medicinal Chemistry
Scheme 6. Synthesis of target compounds 9 and 10a-d.
1
2
3
4
5
6
7
8
9
N
O
N
Cl
N
i
O
37
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
H
N
O
O
N
N
ii, iii
O
R
X
N
X
O
CH
3
O
CH
3
3
8, 39
10a-d
Reagents: (i) CNCH
2
2 5 2 5 2
COOC H , tertꢀBuOK, DMF; (ii) tertꢀBuOK, (C H O) POCl, THF; (iii)
CNCH COOC H or CNCH COOC(CH ) , tertꢀBuOK, THF. Substituents: X = H in 10a,b and 38;
2
2
5
2
3 3
X = F in 10c,d and 39; R = C
2
H
5
3 3
in 10a,c; R = C(CH ) in 10b,d.
In vitro binding. The affinity of the imidazo[1,5ꢀa]quinoline derivatives 7a-aa, 8a-g, 9 and 10a-c
for CBR in bovine cortical membranes was measured by means of competition experiments against
3
the radiolabeled antagonist [ H]flumazenil. The results of the binding studies are expressed as K
i
values in Tables 2 and 3. The in vitro efficacy of the target compounds was tentatively estimated by
measuring the GABA ratio (GR, expressed as a ratio of K without GABA/K with GABA), which is
i
i
44ꢀ47
considered reasonably predictive of the pharmacological profile of a CBR ligand.
Usually, this
value approximates 2 for full agonists and 1 for antagonists, while partial agonists show intermediate
values between 1 and 2; finally, GABA ratio values below 1 are typical of inverse agonists.
Most of the compounds were found to inhibit the specific binding of radiolabeled flumazenil at the
bovine CBR with Ki values within the submicromolar and the subnanomolar ranges with the full
range of the intrinsic efficacy as predicted by GABA ratio values (0.52ꢀ1.6).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 64
3
Table 2. Inhibition of [ H]flumazenil specific binding to CBR in cortical membranes and GABA
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ratio values of compounds 7a-aa.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Compd
X
Y
Z
R
K
i
± SEM
nM)
GABA
ratio
b
(
7
a
OC
OC(CH
OC
OC(CH
OC
OC(CH
OC
OC(CH
2
H
5
3
5
3
5
3
5
3
5
5
5
5
3
5
5
3
5
3
5
3
5
3
5
5
5
N(CH
N(CH
3
)
)
2
2
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
F
H
H
H
H
H
H
H
H
H
CH
CH
CH
CH
CH
CH
CH
8.2 ± 2.3
11 ± 1.9
1.3
1.6
0.76
0.62
1.0
1.1
0.52
0.98
1.0
1.2
1.4
1.3
1.1
1.0
1.1
1.0
1.5
1.4
0.9
1.1
1.1
1.0
0.8
0.8
7b
)
)
)
)
3
3
3
3
3
7
c
2
H
N(CH
N(CH
3
)nꢀC
)nꢀC
4
4
H
H
9
1.9 ± 0.7
1.8 ± 0.5
0.91 ± 0.01
1.2 ± 0.5
20 ± 2.9
7d
3
9
7
e
2
H
N(nꢀC
N(nꢀC
3
H
H
7
)
)
C
C
2
7f
7g
7h
7
7
3
7
2
2
H
N(CH
N(CH
N(C
3
)CH
)CH
2
6
H
H
5
3
2
6
5
14 ± 0.35
55 ± 4.5
i
j
OC
OC
OC
OC
2
2
2
2
H
H
H
H
2
H
4
)
2
NCH
OH
OC(CH
3
3
3
3
3
3
3
3
105 ± 26
1.0 ± 0.37
1.5 ± 0.08
1.0 ± 0.03
1.2 ± 0.02
6.5 ± 0.4
2.6 ± 0.1
2.5 ± 0.3
5.6 ± 0.4
169 ± 14
2870 ± 250
34 ± 4.5
7k
)
3 3
7
l
N(nꢀC
N(nꢀC
3
H
H
7
)
)
2
7
m
OC(CH
)
3
3
7
2
7n
OC
OC
2
H
H
N(CH
N(CH
N(CH
3 2
)CH CCH
7
o
p
q
2
3
)CH
)CH
2
C
C
)
)
6
H
H
5
7
7
OC(CH
OC
OC(CH
OC
OC(CH
OC
OC(CH
)
)
)
)
3
3
3
3
3
2
6
5
2
H
N(nꢀC
N(nꢀC
3
H
H
7
2
C
C
2
H
H
5
7
7
7t
7u
7v
r
s
3
7
2
2
5
2
H
N(C
N(C
2
H
H
5
)
)
2
nꢀC
nꢀC
nꢀC
nꢀC
3
3
3
3
H
7
7
7
7
2
5
2
H
H
H
H
N(nꢀC
N(nꢀC
N(nꢀC
H
H
H
)
)
)
C
C
2
3
3
3
7
2
2
2
7
1437 ± 120
0.44 ± 0.2
5.7 ± 2.6
7w
OC
OC
OC
2
2
2
H
H
H
7
H
H
H
7
7
x
y
N(CH
N(CH
N(CH
3
)CH
)CH
2
6
H
H
5
F
Br
c
c
3
2
6
5
24 ± 1.7
0.90
c
c
7
z
OC
2
H
5
3
)CH
2
C
6
H
5
CCH
H
161 ± 24
0.95
7aa
OC
2
H
5
N(CH
3
)CH
2
C
6
H
5
3 3
CCSi(CH )
H
510 ± 48c
0.99c
flunitrazepam
flumazenil
5.2 ± 0.2
1.9 ± 0.09
1.68
1.03
a
K
i
values are means ± SEM of three independent determinations in bovine cortical membranes.
b
c
GABA ratio = (K
i
without GABA/K
i
with 50 ꢀM GABA). The values were obtained in rat cortical
membranes.
ACS Paragon Plus Environment

Page 15 of 64
Journal of Medicinal Chemistry
3
Table 3. Inhibition of [ H]flumazenil specific binding to CBR in cortical membranes and GABA
1
2
3
4
5
6
7
8
9
ratio values of compounds 8a-i, 9 and 10a-d.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
bovine a
human a
Compd
X
R
K
i
± SEM
GABA
ratio
K
i
± SEM
(nM)
GABA
ratio
b
b
(nM)
8a
OC
OC(CH
N(CH
N(CH )CH
OC(CH
N(CH )CH
2
H
5
H
H
H
H
CH
CH
42 ± 9.6
55 ± 32
0.9
0.8
1.44
0.92
1.0
8b
8c
8d
8e
3
)
2
3
3
)
3515 ± 360
449 ± 10
10 ± 2.8
3
2
3
2
C
)
C
6
H
5
5
3
3
c
8f
3
6
H
3
> 1000
8g
8h
OC
OC
OC
2
2
2
H
H
H
5
5
5
COOC
COOC
2
H
5
2193 ± 633
3.8 ± 2.1
1.8 ± 0.1
18 ± 3.5
25 ± 9.6
7.3 ± 1.5
0.66
0.72
1.18
1.3
0.90
0.98
CH
H
2
2 5
8i
CH
2
OH
2.0 ± 0.2
9
1
0a
H
H
C
2
H
5
33 ± 4.2
7.7 ± 1.3
d
C(CH
3
)
)
3
1.04
10b
10c
F
F
C
2
H
5
2.2 ± 0.42
1.0 ± 0.46
5.2 ± 0.2
0.75
1.0
1.68
1.03
2.7 ± 0.3
2.0 ± 0.3
6.4 ± 0.5
2.0 ± 0.08
0.95
1.0
1.61
1.02
10d
C(CH
3
3
flunitrazepam
flumazenil
1.9 ± 0.09
a
b
i i
K values are means ± SEM of three independent determinations. GABA ratio = (K without
c
d
GABA/K with 50 ꢀM GABA). 19% displacement at 1000 nM. See ref 31.
i
The analysis of the structureꢀactivity relationships confirmed the importance of the substituents in
positions 4 and 5 of the imidazo[1,5ꢀa]quinoline scaffold in modulating both the CBR affinity and
the intrinsic efficacy (Figure 3). The most potent CBR ligands were obtained in the series of
imidazo[1,5ꢀa]quinoline derivatives bearing a N,Nꢀdipropylaminocarbonyl group at position 5 of the
tricyclic nucleus. However, also other lipophilic substituents are tolerated in the binding site region
interacting with position 5 of the imidazo[1,5ꢀa]quinolone, namely N,Nꢀdimethylaminocarbonyl, Nꢀ
methybutylaminocarbonyl, Nꢀmethypropargylaminocarbonyl, Nꢀmethybenzylaminocarbonyl, tertꢀ
butoxycarbonyl, and phenyl groups (Figure 3).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 64
N
N
O
N
N
O
O
O
N
N
N
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
N
O
O
O
O
CH
3
CH
3
O
Y
O
Y
O
Y
O
Y
Y
Y
CH
N
CH
CH
3
(
7a) 8.2 nM PartAGO
(7b) 11 nM PartAGO
OH
(7j) 105 nM PartAGO
3
3
(
7k) 1.0 nM PartAGO
O
N
N
(
(
(
7c) 1.9 nM InvAGO
7e) 0.91 nM ANTAGO
7g) 20 nM InvAGO
(
7d) 1.8 nM InvAGO
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
7l) 1.5 nM PartAGO
(7m) 1.0 nM ANTAGO
(
7f) 1.2 nM ANTAGO
N
CH
N
3
(
7n) 1.2 nM ANTAGO
CH
N
3
(7h) 14 nM ANTAGO
CH
N
3
(
7o) 6.5 nM ANTAGO
(7p) 2.6 nM ANTAGO
N
N CH
3
(7i) 55 nM ANTAGO
Figure 3. SAR in imidazo[1,5ꢀa]quinoline derivatives 7a-i and 7j-p. Effects of the lipophilic
substituents in position 5.
However, the tolerance to large substituents in position 5 appeared to be conditioned by the presence
of additional steric bulk in position 4 of the tricyclic scaffold (Figure 4).
N
N
N
N
O
O
O
O
N
N
N
N
O
O
O
O
R
R
R
R
O
N
O
N
R
R
(
7e) 0.91 nM ANTAGO
(7f) 1.2 nM ANTAGO
H
(8a) 42 nM ANTAGO
(
8b) 55 nM InvAGO
H
CH
3
(8e) 10 nM ANTAGO
CH
3
(7l) 1.5 nM PartAGO
(7m) 1.0 nM ANTAGO
CH
2
OH
(
8i) 1.8 nM PartAGO
CH
2
CH
3
(7q) 2.5 nM PartAGO
(7r) 5.6 nM PartAGO
COOCH
2
CH
3
(8g) 2193 nM InvAGO
(7u) 34 nM ANTAGO
(7v) 1437 nM ANTAGO
CH
2
COOCH
2
CH
3
(
8h) 3.8 nM InvAGO
CH
2
CH
2
CH
3
Figure 4. SAR in imidazo[1,5ꢀa]quinoline derivatives 7e,f,l,m,q,r,u,v and 8a,b,e,g,h,i. Effects of
the alkyl substituents in position 4.
In fact, in the dipropylamido subꢀseries (Figure 4), the increase of the alkyl side chain length led to a
stepwise decrease in CBR affinity with a rapid acceleration in propyl derivatives 7u,v. This result
can be rationalized in terms of interactions between the ester group and the alkyl moiety in position
ACS Paragon Plus Environment

Page 17 of 64
Journal of Medicinal Chemistry
4
, but also with by assuming that the lipophilic pocket can be saturated as previously observed in 5ꢀ
1
2
3
4
5
6
7
8
9
38
3
HT receptors.
On the other hand, in the imidazo[1,5ꢀa]quinoline subꢀseries bearing a phenyl group in position 5
Figure 4), the effects of the substituents in position 4 were highly variable and appeared to depend
on the stereoelectronic features of the substituent itself.
In general, small substituents such as H, CH and CH
(
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
2
OH are tolerated better than the bulkier
carbethoxy group of compound 8g, but the spacing of the ester group by a methylene bridge as in 8h
restored nanomolar CBR affinity suggesting the involvement of specific interactions.
As already observed in related CBR ligands, the presence of small atoms such as H or F in position 7
is required for nanomolar CBR affinity, whereas the presence of the bulkier bromine atom produces
a significant drop in CBR affinity that became dramatic (i. e. almost two orders of magnitude) when
alkyne substituents are present in this position as in compounds 7zꢀaa (Figure 5).
N
N
O
O
N
N
O
O
Z
Z
CH
3
O
N
O
N
Z
H
(
7e) 0.91 nM ANTAGO
7w) 0.44 nM InvAGO
(
7g) 20 nM InvAGO
F
(
(7x) 5.7 nM InvAGO
Br
(
7y) 24 nM ANTAGO
7z) 161 nM ANTAGO
H
(
CH
Si CH
CH
3
(7aa) 510 nM ANTAGO
3
3
N
N
O
O
N
N
N
O
N
O
X
X
O
CH
3
O
CH3
X
H
(
10a) 25 nM ANTAGO
(
10b) 7.3 nM ANTAGO
F
(
10c) 2.2 nM InvAGO
(10d) 1.0 nM ANTAGO
Figure 5. SAR in imidazo[1,5ꢀa]quinolines 7e,g,w,x,y,z,aa and reference imidazo[1,5ꢀ
a]quinoxalines 10aꢀd. Effects of the substituents in position 7.
It is noteworthy that the beneficial effect of the fluorine substituent was less evident in imidazo[1,5ꢀ
a]quinoline derivatives 7w,x than in imidazo[1,5ꢀa]quinoxaline derivatives 10c,d. However, by
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 64
virtue of its subnanomolar CBR affinity (Ki = 0.44 nM) fluoroderivative 7w resulted more potent
than the corresponding imidazo[1,5ꢀa]quinoxaline derivative 10c and is the most potent ligand
among the imidazo[1,5ꢀa]quinoline derivatives described so far.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Finally, the replacement of ester groups in position 3 of imidazo[1,5ꢀa]quinoline derivatives 8a,b,e
with the amide ones of 8c,d,f (Figure 6) was deleterious from the point of view of the interaction
with CBR binding site producing low affinity ligands.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
N
O
O
N
N
X
X
CH
3
X
OCH
2
CH
3
(8a) 42 nM ANTAGO
(
8b) 55 nM InvAGO
(8e) 10 nM ANTAGO
O
CH
CH
3
N
(8c) 3515 nM PartAGO
3
CH3
N
(
8d) 449 nM ANTAGO
(8f) IC50 >1000 nM
Figure 6. SAR in imidazo[1,5ꢀa]quinolines 8aꢀf. Effects of the substituents in position 3.
36
-
In vitro efficacy in Cl uptake assay in rat cerebrocortical synaptoneurosomes. The comparison
of the GABA ratio values showed a rather complex pattern in the structureꢀactivity relationships of
imidazo[1,5ꢀa]quinoline derivatives 7a-aa, 8a-g, 9 and 10a-c. In general, the agonistꢀlike properties
appeared to be linked to the presence of relatively small substituents (i. e. 7a,b), whereas the
presence of large lipophilic substituents appeared to be associated with antagonistꢀlike features.
However, ligands showing apparently different intrinsic efficacy were present in the same subꢀseries
without evidencing a clear trend. This could be due both to the complexity of the interaction and to
possible experimental errors.
Thus, the predictive capability of GABA ratio values was evaluated in the limited set of reference
imidazo[1,5ꢀa]quinoxaline derivatives 10b-d by means of a more direct measure of in vitro efficacy
ACS Paragon Plus Environment

Page 19 of 64
Journal of Medicinal Chemistry
36
ꢀ
48,49
consisting in Cl uptake assay in rat cerebrocortical synaptoneurosomes.
The synaptic chloride
1
2
3
4
5
6
7
8
9
conductance effected by GABA activating the GABA receptor complex is modulated by ligands
A
acting at the CBR. In particular, agonists increase the current, antagonists are ineffective, and
inverse agonists decrease the ion flow. The results shown in Figure 7 suggested that reference 10b
behaved as a CBR antagonist in agreement with its GABA ratio values (0.98ꢀ1.04) and with the data
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
1
described in the literature.
flunitrazepam
flumazenil
ethyl βꢀcarboline
10b
1
20
8
0
1
0c
10d
40
0
ꢀ
10
ꢀ9
ꢀ8
ꢀ7
ꢀ6
Log[M]
36
ꢀ
Figure 7. Cl uptake measured in rat cerebrocortical synaptoneurosomes for compounds 10b
(magenta), 10c (blue), 10d (cyan), flunitrazepam (empty circles), flumazenil (empty diamonds), and
ethyl βꢀcarboline (empty squares).
36
ꢀ
Compound 10d (GABA ratio = 1.0) showed a very slight increase of Cl influx, behaving as a
partial agonist characterized by a very low intrinsic efficacy. On the other hand, an even slight
decrease in the ion flow was observed with compound 10c (GABA ratio in bovine CBR = 0.75, in
human CBR = 0.95), which could be therefore classified either as a partial inverse agonist showing a
3
6
ꢀ
very low intrinsic efficacy or as an antagonist. No massive Cl influx was promoted by these
reference compounds in agreement with their antagonistꢀlike or partial inverse agonist properties as
predicted by GABA ratio values. On the whole, these apparent discrepancies emphasized the
importance of a suitable biological characterization of the newly synthesized CBR ligands in order
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 64
to appreciate their pharmacological profile.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In vitro efficacy in excytotoxic-mediate injury. The disruptions in GABA signaling is involved in
many acute and chronic neurodegenerative disorders such as temporal lobe epilepsy, Parkinson's
disease (PD), Huntington's disease (HD) and brain ischemia. The GABAergic system is indeed
indispensable to keep the balance between the excitation and the inhibition required for normal
neuronal function. An imbalance between these systems contribute to excitotoxicity and neuronal
cell death. Consequently, modulation of the GABAergic system can successfully reverse
excitotoxicꢀinduced injury in disease models, suggesting that therapeutic strategies targeting the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0ꢀ52
GABAergic system could be effective in treating neurodegenerative disorders.
modulators of the GABA receptors, such as diazepam and the partial agonist imidazo[1,5ꢀ
a]quinoline derivative 5 have proven to show neuroprotective properties in different models of
Positive
A
5
3ꢀ57
excytotoxicꢀmediate injury.
selected on the basis of their K
Therefore, imidazo[1,5ꢀa]quinoline derivatives 7l,q,r, and 8i were
and GABA ratio values (i. e. nanomolar CBR affinity and partial
i
agonist profile) and tested for their potential neuroprotective activity. Rat cortical brain slices were
subjected to excitotoxicꢀmediated damage (i. e. oxygenꢀglucose deprivation and reoxygenation,
OGD/R) and neuronal injury/neuroprotection was assessed by measuring the release of lactate
dehydrogenase (LDH). All drug molecules were added during reperfusion and their effects were
compared to those exerted by diazepam. The results demonstrated that diazepam exerted
neuroprotective effects according to a "Uꢀshaped", hormeticꢀlike, concentrationꢀresponse curve, with
an efficacy window of 0.5ꢀ10 µM (Figure 8). In this range, the maximum recovery in LDH release
was 55.9%, which was observed at both 1 and 5 µM concentrations (Table 4).
ACS Paragon Plus Environment

Page 21 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 8. Effects of 7l,q,r, 8i, and diazepam on oxygen–glucose deprivation and reoxygenation
(OGD/R)ꢀinduced release of LDH of rat brain cortical slices. Slices were incubated in artificial
cerebrospinal fluid (ACSF) for 120 min (control conditions, CTRL) or subjected for 30 min to
oxygen/glucose deprivation followed by 90 min incubation in normally oxygenated ACSF
(
reperfusion). Increasing concentrations of the compounds (0.1–100 µM) were added to ACSF
during the 90 min reperfusion phase. Data are means ± S.E.M. of at least 4 different experiments.
°°P < 0.01 vs CTRL; *P < 0.05, **P < 0.01, ***P < 0.001 vs OGD/R (ANOVA followed by
Dunnet post test).
°
54
The hormetic effect of diazepam was already observed and might be explained by considering that
elevated diazepam levels at the synaptic cleft might cause an excessive activation of GABA
A
−
receptors, which increases the overload of Cl and causes GABA
A
desensitisation, thus resulting in
5
8,59
depolarization and neuron damages.
In the same way, imidazo[1,5ꢀa]quinoline derivatives 7l,q,r,
and 8i showed neuroprotective properties. All the compounds, in fact, reduced OGD/Rꢀinduced
LDH release in an hormeticꢀlike fashion although with different efficacy windows. In particular, 8i
and 7r were the most interesting compounds (Figure 8) since they exerted their effects in a wider
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 64
concentration range than diazepam (i. e. 1ꢀ25 µM 8i or 1ꢀ50 µM 7r) but the maximum recovery in
LDH release was lower (39.5% and 44.4%, respectively, at 5 µM, see Table 4). Also 7l,q reverted
the release of the endocellular enzyme caused by the ischemiaꢀlike conditions but the maximum
effect was observed at 50 µM concentration and the recovery was 44%. Taken together, these results
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
suggests that 7l,q,r, and 8i could behave as partial agonists at GABA
A
receptors in the present
modulator
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
experimental model since they exhibited lower efficacy than the positive GABA
A
diazepam.
Table 4. Effects of 7l,q,r, 8i, and diazepam on oxygen–glucose deprivation and reoxygenation
OGD/R)ꢀinduced release of LDH of rat brain cortical slices.
(
a
b
c
Compd
Efficacy Window
µM)
EC
Recovery
(
(µM)
50
50
5
(% )
7l
5ꢀ50
5ꢀ100
1ꢀ50
44.0 ± 8.5*
7q
44.3 ± 4.8***
44.4 ± 2.4***
39.5 ± 5.6***
55.9 ± 7.1***
55.9 ± 5.9***
7r
8i
1ꢀ25
5
1
0.5ꢀ10
1
5
Rat cortical brain slices were subjected to oxygenꢀglucose deprivation and reoxygenation and
neuronal injury/neuroprotection was assessed by measuring the release of lactate dehydrogenase
a
(
LDH). All drugs were added during reperfusion. The efficacy windows represent the range of
concentrations at which a significant reduction of OGD and reperfusionꢀinduced LDH release was
b
observed. EC (Effective Concentration) is the µM concentration at which the highest reduction was
c
observed. The Recovery value represents the % of reversion exerted at EC concentration; 100% was
taken as the return to basal values (CTRL). Recovery data are reported as mean ± esm and the
comparison between values was performed by using ANOVA followed by Dunnet post hoc test.*P <
0
.05, ***P < 0.001 vs OGD/R.
ACS Paragon Plus Environment

Page 23 of 64
Journal of Medicinal Chemistry
In vivo efficacy. The newly synthesized compounds 7a,c,d,l,m,p,q,r,w were evaluated in mice as
modulators of central nervous system functionalities after per os administration. In particular, four
pharmacological actions were taken into consideration. The lightꢀdark box test was used to ascertain
the potential anxiolytic effect, while the rotaꢀrod test measured the myorelaxant effect, the holeꢀ
board test was performed to assess the effects on mouse spontaneous motility and explorative
activity, and the passive avoidance test was finally used to evaluate the mouse learning and memory
impairment.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The anxiolytic effect of the molecules is shown in Figure 9. The lightꢀdark box test is based on the
innate aversion of rodents to the brightly lit and open areas, and on the spontaneous noveltyꢀinduced
exploratory behavior allowing the evaluation of potential anxiolytic compounds. All compounds,
ꢀ1
with the exception of 7p, increased the time spent in the light box after dosing at 10 mg kg .
Compounds 7d (167.2 ± 6.5 s), 7r (172.5 ± 7.2 s) and 7w (181.4 ± 8.3 s) were the most effective
ꢀ1
showing a comparable effect with diazepam (1 mg kg subcutaneously) (Figure 9). Compound 7d
ꢀ1
presented a doseꢀdependent anxiolytic effect, which was significant starting from 3 mg kg (144.8 ±
.3 s; data not shown). A similar potency was showed by compound 7l, which was active starting
6
ꢀ1
ꢀ1
from 3 mg kg (153.4 ± 6.1 s; data not shown) peaking at 10 mg kg (Figure 9). Compound 7m was
ꢀ1
the most potent since it was effective at 1 mg kg (Figure 9), whereas 7p was ineffective when
ꢀ1
ꢀ1
administered at 1 and 10 mg kg (Figure 9), but by increasing the dose to 30 mg kg the time was
enhanced up to 139.2 ± 5.1 s (data not shown).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 9. Lightꢀdark box test. Anxiolytic activity. The new compounds were administered per os,
ꢀ1
diazepam (1 mg kg ) was administered subcutaneously. All treatments were performed 30 min
before the test. Each value represents the mean ± SEM of at least 10 mice. **P < 0.01 in comparison
to vehicleꢀtreated mice.
In order to validate the behavioral measurements, possible neurological or muscular side effects of
the tested compounds were excluded by the holeꢀboard and the rotaꢀrod tests. All the compounds (10
ꢀ1
mg kg ) did not alter the neurological and muscular abilities of the mice as evaluated by the holeꢀ
board test 30 min after treatment. The number of movements (motor activity) and the number of
inspections (exploratory activity) were comparable to vehicleꢀtreated animals (Figure 10).
Figure 10. Holeꢀboard test. Effects on neurological and muscular abilities. All compounds were
administered per os 30 min before the test. Each value represents the mean ± SEM of at least 10
mice.
ACS Paragon Plus Environment

Page 25 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Similarly, no negative effects on motor coordination emerged in the rotaꢀrod test (Figure 11). The
treated animals showed a progressive ability to maintain the balance on a rotating rod.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 11. Rotaꢀrod test. Effects on motor coordination. All compounds were administered per os
30 min before the test. Each value represents the mean ± SEM of at least 10 mice.
The nootropic effects were assessed in the passive avoidance test measuring the prevention of
scopolamineꢀinduced amnesia. The muscarinic antagonist drastically reduced the time spent in the
light box of the apparatus during the retention session (44.8 ± 8.1 s vs 101.4 ± 7.0 s of vehicleꢀ
nd
treated animals; 2 experimental day) highlighting a lack of memory of the punishment received in
ꢀ1
the dark box (Figure 12). Compounds 7c,d,l,m,p,q,r,w (10 mg kg ) were able to significantly
prevent scopolamineꢀinduced amnesia (Figure 12). Compound 7w was the most effective (99.6 ± 9.2
ꢀ1
s), both 7r and 7w were effective also at 1 mg kg . Compound 7d was effective starting from 3 mg
ꢀ1
ꢀ1
kg (73.5 ± 8.8 s; data not shown). Ten mg kg compound 7p was not effective in the passive
ꢀ1
avoidance test (Figure 12). However the dose of 30 mg kg increased the time of the retention
session up to 77.9 ± 7.8 s (data not shown), while 7a was ineffective (Figure 12).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 64
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 12. Passive avoidance test. Effects on learning and memory. All compounds were
ꢀ1
administered per os 30 min before the test. Scopolamine (1.5 mg kg intraperitoneally) was
administered immediately after the punishment. The time recorded during the retention session is
reported. Each value represents the mean ± SEM of at least 10 mice. *P < 0.05 and **P < 0.01 in
comparison to vehicleꢀtreated mice.
CONCLUSION
The structural analogies existing between GABA
A 3
and 5ꢀHT receptors suggested the application of
the approach we used in studying 5ꢀHT receptors to the characterization of CBR interaction
3
features. Thus, a series of imidazo[1,5ꢀa]quinoline derivatives related to 5 (a previously described
drug candidate for the treatment of anxiety) was designed, synthesized, and biologically
characterized in comparison with reference imidazo[1,5ꢀa]quinoxalines 10a-d. Most of the newlyꢀ
synthesized compounds showed high CBR affinity with Ki values within the submicromolar and the
subnanomolar ranges and interesting SAR trends, which suggested the existence of a large though
apparently saturable lipophilic pocket in the CBR binding site region interacting with positions 4 and
5
. From another perspective, this result could be interpreted as the evidence of a certain degree of
conformational freedom of the amino acid residues interacting with the substituents in positions 4
ACS Paragon Plus Environment

Page 27 of 64
Journal of Medicinal Chemistry
and 5 of the imidazo[1,5ꢀa]quinoline nucleus. Thus, this promising evidence paves the way to the
1
2
3
4
5
6
7
8
9
application of our approach to the study of the 5ꢀHT
interaction of CBR with divalent and more in general multivalent ligands. Fluoroderivative 7w (Ki
0.44 nM) resulted the most potent ligand and despite its inverse agonistꢀantagonist profile
3
receptor to the characterization of the
6
0
=
suggested by its GABA ratio value of 0.8, acted as an agonist in the lightꢀdark box test, the classical
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
animal model of anxiety, and was devoid of the undesired myorelaxant side effects. In addition,
ꢀ1
compound 7w (at 1 mg kg ) was found to significantly prevent scopolamineꢀinduced amnesia
showing the best efficacy among the compounds evaluated in the in vivo studies (7a,c,d,l,m,p,q,r).
Furthermore, imidazo[1,5ꢀa]quinoline derivatives 7l,q,r, and 8i showed neuroprotective properties
since they reduced LDH release induced by ischemiaꢀlike condition in an hormeticꢀlike fashion
although with different efficacy windows.
EXPERIMENTAL SECTION
Chemistry. All chemicals used were of reagent grade. Yields refer to purified products and are not
optimized. Melting points were determined in open capillaries on a Gallenkamp apparatus and are
uncorrected. Merck silica gel 60 (230ꢀ400 mesh) was used for column chromatography. Merck TLC
plates, silica gel 60 F254 were used for TLC. NMR spectra were recorded by means of either a
Bruker AC 200 or a Bruker DRX 400 AVANCE spectrometers in the indicated solvents (TMS as
internal standard); the values of the chemical shifts are expressed in ppm and the coupling constants
(
J) in Hz.
The purity of compounds 7a-aa, 8a-i, 9, and 10a-d was assessed by RPꢀHPLC and was found to
be higher than 95%. An Agilent 1100 Series system equipped with a Phenomenex C18 (3.9 x 150
mm, 10 m) column or a Zorbax Eclipse XBDꢀC8 (4.6 x 150 mm, 5 m) column was used in the
HPLC analysis with acetonitrileꢀmethanolꢀwater (10:20:70) or (10:40:50) or (10:50:40) or
10:70:20) as the mobile phases at a flow rate of 2.0 mL/min. UV detection was achieved at 254 nm.
ꢀ
ꢀ
(
Mass spectra were recorded on either a Thermo LCQꢀDeca or an Agilent 1100 LC/MSD.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 64
General procedure for the synthesis of target imidazo[1,5-a]quinoline derivatives 7a-i,k-m,o-y.
A mixture of the suitable 2ꢀchloroquinoline derivative (11-24, 1 equivalent) was cooled at 0ꢀ5 °C in
dry DMF (typically, 10 mL for 1 mmol) under argon and then treated with the suitable
isocyanoacetate (3 equivalents) and potassium tertꢀbutoxide (3 equivalents). The resulting mixture
was stirred for 30 min at 0ꢀ5 °C, then allowed to stir at room temperature for 1 h, and finally heated
at 80 °C for 1ꢀ20 h (following the reaction progress by TLC). After cooling to room temperature,
acetic acid (typically, 1.0 mL for 1 mmol) was added and the mixture was stirred for additional 20
min and then poured onto crushed ice. The precipitate was collected by filtration, washed with
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2 4
water, dissolved into chloroform and the organic layer washed with brine, dried over Na SO , and
concentrated under reduced pressure. The residue was purified by flashꢀchromatography with the
suitable eluent to afford the expected imidazo[1,5ꢀa]quinoline derivative (7a-i,k-m,o-y), which after
reꢀcrystallization from the suitable solvent gave an analytical sample.
Ethyl 5-(dimethylcarbamoyl)imidazo[1,5-a]quinoline-3-carboxylate (7a).
The title compound was obtained as a white solid from 11 according to the above general procedure
and purified by flash chromatography with ethyl acetate as the eluent (yield 35%). An analytical
sample was obtained by recrystallization from ethyl acetateꢀchloroform by slow evaporation (white
1
crystals, mp 209ꢀ210 °C). H NMR (400 MHz, CDCl
3
): 1.45 (t, J = 7.1, 3H), 2.96 (s, 3H), 3.23 (s,
3
1
1
H), 4.46 (q, J = 7.1, 2H), 7.52 (t, J = 7.6, 1H), 7.63ꢀ7.77 (m, 2H), 8.03 (s, 1H), 8.09 (d, J = 8.3,
13
H), 8.70 (s, 1H). C NMR (100 MHz, CDCl ): 14.5, 35.0, 38.9, 60.8, 115.3, 121.2, 125.6, 126.9,
3
17 3 3
27.0, 128.4, 130.4, 130.6, 131.2, 132.7, 163.0, 168.0. MS (ESI) m/z calcd for C17H N O Na [M +
+
Na] , 334.1; found, 334.2.
tert-Butyl 5-(dimethylcarbamoyl)imidazo[1,5-a]quinoline-3-carboxylate (7b).
This compound was prepared from 11 according to the above general procedure and purified by
flash chromatography with ethyl acetate as the eluent to obtain 7b as a white solid (yield 59%, mp
1
222ꢀ223 °C). H NMR (400 MHz, CDCl
3
): 1.66 (s, 9H), 2.97 (s, 3H), 3.22 (s, 3H), 7.51 (t, J = 7.6,
ACS Paragon Plus Environment

Page 29 of 64
Journal of Medicinal Chemistry
1
3
1
H), 7.67 (t, J = 7.3, 1H), 7.71 (d, J = 8.0, 1H), 8.01 (s, 1H), 8.08 (d, J = 8.3, 1H), 8.66 (s, 1H). C
1
2
3
4
5
6
7
8
9
NMR (100 MHz, CDCl
3
): 28.4, 34.9, 38.9, 81.5, 115.3, 115.6, 121.2, 126.8, 127.0, 128.1, 130.2,
+
1
30.6, 130.7, 132.2, 162.4, 168.1. MS (ESI) m/z calcd for C19
62.1.
H
21
N
3
O
3
Na [M + Na] , 362.2; found,
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ethyl 5-[butyl(methyl)carbamoyl]imidazo[1,5-a]quinoline-3-carboxylate (7c).
This compound was synthesized from 12 according to the above general procedure and purified by
flash chromatography with ethyl acetate/nꢀhexane (9:1) as the eluent to obtain 7c as a white solid
(yield 10%, mp 148ꢀ149 °C). Since the amide nitrogen of the compound bears two different
1
substituents, its H NMR spectrum (CDCl
3
) shows the presence of a (ca. 1:1) mixture of two
1
rotamers in equilibrium; for the sake of simplification, the integral values have not been reported. H
NMR (400 MHz, CDCl ): 0.73 (t, J = 7.3), 1.03 (t, J = 7.3), 1.07ꢀ1.19 (m), 1.42ꢀ1.58 (m), 1.67ꢀ1.78
m), 2.92 (s), 3.18ꢀ3.25 (m), 3.61ꢀ3.69 (m), 4.47 (q, J = 7.1), 7.50ꢀ7.57 (m), 7.66ꢀ7.77 (m), 8.02 (s),
3
(
+
8.03 (s), 8.10 (d, J = 8.3), 8.71 (s), 8.72 (s). MS (ESI) m/z calcd for C20
76.2; found, 376.0.
23 3 3
H N O Na [M + Na] ,
3
tert-Butyl 5-[butyl(methyl)carbamoyl]imidazo[1,5-a]quinoline-3-carboxylate (7d).
This compound was prepared from 12 according to the above general procedure and purified by
flash chromatography with ethyl acetate/nꢀhexane (9:1) as the eluent to obtain 7d as a creamy solid
(yield 17%, mp 148ꢀ150 °C). Since the amide nitrogen of the compound bears two different
1
substituents, its H NMR spectrum (CDCl
3
) shows the presence of a (ca. 1:1) mixture of two
1
rotamers in equilibrium; for the sake of simplification, the integral values have not been reported. H
NMR (400 MHz, CDCl ): 0.73 (t, J = 7.3), 1.01 (t, J = 7.3), 1.07ꢀ1.18 (m), 1.38ꢀ1.77 (m), 2.14 (s),
.92 (s), 3.18 (s), 3.19ꢀ3.24 (m), 3.64 (t, J = 7.5), 7.47ꢀ7.54 (m), 7.63ꢀ7.73 (m), 7.97 (s), 8.08 (d, J =
3
2
8
+
.3), 8.68 (s), 8.69 (s). MS (ESI) m/z calcd for C H N O Na [M + Na] , 404.2; found, 404.2.
22
27
3
3
Ethyl 5-(dipropylcarbamoyl)imidazo[1,5-a]quinoline-3-carboxylate (7e).
ACS Paragon Plus Environment