Development of 3-substituted-1H-indole derivatives as NR2B/NMDA receptor antagonists

Article abstract of DOI:10.1016/j.bmc.2008.12.058

A combined ligand-based and structure-based approach has previously allowed us to identify NR2B/NMDA receptor antagonists containing indole scaffold. In order to further explore the main structure activity relationships of this class of derivatives we herein report the design, synthesis and biological evaluation of new analogues. Some derivatives demonstrated to produce significant anticonvulsant properties and NMDA antagonism. The most active of them (3d) showed NR2B binding affinity equipotent to that of ifenprodil. These results were also corroborated by computational studies.

Full text of DOI:10.1016/j.bmc.2008.12.058

Bioorganic & Medicinal Chemistry 17 (2009) 1640–1647
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Development of 3-substituted-1H-indole derivatives as NR2B/NMDA
receptor antagonists
a
a
b
b
Rosaria Gitto a, , Laura De Luca , Stefania Ferro , Rita Citraro , Giovambattista De Sarro ,
*
Lara Costa c, Lucia Ciranna c, Alba Chimirri a
a Dipartimento Farmaco-Chimico, Università di Messina, Viale Annunziata, 98168 Messina, Italy
b Dipartimento di Medicina Sperimentale e Clinica, Università Magna Gr
ꢀ
cia, Viale Europa Località Germaneto, 88100 Catanzaro, Italy
c Dipartimento di Scienze Fisiologiche, Università di Catania, Viale Andrea Doria 6, 95125 Catania, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
A combined ligand-based and structure-based approach has previously allowed us to identify NR2B/
NMDA receptor antagonists containing indole scaffold. In order to further explore the main structure
activity relationships of this class of derivatives we herein report the design, synthesis and biological
evaluation of new analogues. Some derivatives demonstrated to produce significant anticonvulsant prop-
erties and NMDA antagonism. The most active of them (3d) showed NR2B binding affinity equipotent to
that of ifenprodil. These results were also corroborated by computational studies.
Received 8 September 2008
Revised 22 December 2008
Accepted 24 December 2008
Available online 31 December 2008
Keywords:
Ó 2008 Elsevier Ltd. All rights reserved.
NR2B/NMDA antagonist
Molecular modeling
Anticonvulsant activity
Patch clamp
Binding affinity
1. Introduction
typic NR2B subtype-selective antagonists, ifenprodil (1)7–16 and
its more selective analogue Ro 25-6981 (2) (Chart 1), have good po-
The evidence supporting a role of excessive glutamate release or
receptor activation in pathological processes of Central Nervous
System (CNS) led to the development of agents that modulate
the glutamate transmission. NMDA receptors (NMDARs) belonging
to the ionotropic glutamate receptor (iGluR) family are involved in
many physiological processes, such as neuronal development,
learning and memory but also play important roles in pathological
states of CNS, including strokes, seizures, and pain.1–3 Functional
NMDARs are heteromeric complexes containing NR1 and NR2 sub-
units (NR2A-D subtypes), and more rarely NR3 subunit. NR1 and
NR2 subunits share a common membrane topology with other
members of the iGluR family. Each subunit has a large extracellular
amino-terminal domain (ATD), three transmembrane segments
(M1, M3 and M4), a pore lining P-loop region (M2), an extracellular
loop (L3) located between M3 and M4, and an intracellular carboxy
terminal domain (CTD). It was demonstrated that the ATD is
structurally similar to the bacterial periplasmic leucine, isoleucine,
valine binding protein (LIVBP). Using molecular modelling and
site-directed mutagenesis studies, it has been demonstrated that
the LIVBP-like domain of the NR2B subunit contains the binding
pocket for non-competitive NR2B-antagonists.4–6 The two proto-
tential as neuroprotective agents and have been shown to bind at
the so-called ifenprodil site.17,18
Considering it has been established that NR2B/NMDA receptor
may be playing a crucial role in epileptic disorders, in previous
studies we reported a molecular modelling strategy that through
a combined ligand-based and target-based approach led to the
identification of new NR2B subtype-selective ligands containing
an indole scaffold as anticonvulsant agents. In particular, com-
pound 3a (Chart 1), with the best binding data affinity, was able
to prevent audiogenic seizures in DBA/2 mice and reduce NMDA
receptor-mediated current in patch clamp experiments as well as
exert a neuroprotective effect in HCN-1A.19
Starting from these promising results and considering the sig-
nificant structural information obtained from our computational
studies about NR2B subtype-selective ligands, we planned some
modifications of compound 3a as depicted in general formula A
(Chart 1). Our investigations were generally aimed at clarifying
the impact on the pharmacological effects of the following fea-
tures: (i) benzene-fused ring substitution and modification (R
and Y); (ii) the nature of the X spacer; (iii) the presence of another
positive ionizable group (Z) on the benzylpiperidine moiety. All the
synthesized compounds were tested in order to evaluate their
anticonvulsant properties in DBA/2 mice and the binding and func-
tional activity of the most active derivatives were also studied.
* Corresponding author. Tel.: +39 0906766413; fax: +39 0906766402.
E-mail address: rgitto@pharma.unime.it (R. Gitto).
0968-0896/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2008.12.058

R. Gitto et al. / Bioorg. Med. Chem. 17 (2009) 1640–1647
1641
OH
OH
N
N
OH
2, Ro 25-6981
OH
1, ifenprodil
O
N
X
N
R = H, Br, MeO, OH
Y = CH, N
R
z
N
N
H
H
Y
Z = CH, N
3a
A
X = COCH2 , COCO, (CH2)nCO
n = 1, 2
Chart 1. NR2B/NMDA antagonists and planned modifications.
Moreover, docking simulations were carried out to rationalize the
biological results.
into the corresponding derivatives 3a–c and 6a by reaction with
4-benzylpiperidine and benzylpiperazine; similarly, the 2-(4-ben-
zylpiperidin-1-yl)-1-(1H-pyrrolo[2,3-b]pyridin-3-yl)ethanone 9a
was prepared from chloroacetyl intermediate 8a in turn obtained
from 7-azaindole 7a by treatment with chloroacetyl chloride in
the presence of AlCl3 as catalyst. By using a standard amide cou-
pling reaction with 4-benzylpiperidine the 2-(1H-indol-3-yl)-2-
acetic acids 10a–c and 3-(1H-indol-3-yl)propanoic acid 11a gave
compounds 12a–c and 13a, respectively.
2. Results and discussion
2.1. Synthesis
The synthetic pathways employed to obtain the designed indole
derivatives are depicted in Scheme 1. Following a previously
reported method19 starting from compounds 4a–c, we prepared
2-chloro-1-(1H-indol-3-yl)ethanones 5a–c, which were converted
Due to the high nucleophilic character at the 3-position of in-
dole, the treatment with oxalyl chloride of appropriate substituted
O
Cl
ii
R
R
z
O
i
R
N
N
H
N
H
N
H
4a-c
3 a-d, Z = CH
6 a, Z = N
5a-c
Cl
O
R
N
R
R
O
iii
ii
N
N
H
N
H
N
N
N
H
7a
8a
9a
OH
O
R
n
N
iv
R
n
O
N
H
R
N
H
a
b
c
d
H
10 a-c, n= 1
11 a , n= 2
12 a-d, n= 1
13 a, n= 2
Br
OMe
vii
OH
Cl
O
R
O
O
R
vi
v
N
4 a-c
N
H
O
N
H
15 a-d
14 a-c
Scheme 1. Reagents and conditions: (i) ClCH2COCl, pyridine, dioxane,
D, 2 h; (ii) 4-benzylpiperidine or benzylpiperazine, K2CO3, toluene, D, 3 h; (iii) ClCH2COCl, AlCl3, DCM,
rt, 3 h; (iv) 4-benzylpiperidine, HBTU, TEA, DMF, rt 2 h; (v) ClCOCOCl, diethyl ether, rt, 2 h; (vi) 4-benzylpiperidine, THF, TEA, rt, 2 h; (vii) BBr3 (1.0 M DCM), rt, 10 h.

1642
R. Gitto et al. / Bioorg. Med. Chem. 17 (2009) 1640–1647
indoles 4a–c produced the corresponding 2-(1H-indol-3-yl)-2-oxo-
acetyl chloride intermediates 14a–c which were then coupled with
4-benzylpiperidine in alkaline medium in a one-pot procedure to
give the desired 1-(4-benzylpiperidin-1-yl)-2-(1H-indol-3-yl)eth-
ane-1,2-diones (15a–c). Finally, by treatment with boron tribro-
mide solution (1.0 M in methylene chloride) compounds 3c, 12c,
and 15c were demethylated into the corresponding hydroxyl-
derivatives 3d, 12d, and 15d. The structures of all compounds ob-
tained were supported by elemental analyses and spectroscopic
measurements.
The 4-benzylpiperidine moiety of the reference compound 3a was
replaced with the benzylpiperazine group, but this modification
produced the inactive derivative 6a, probably indicating that the
presence of a supplementary positive charge is not conducive to
pharmacological activity. We also substituted the benzene-fused
ring of compound 3a with a pyridine nucleus and the obtained
analogue 9a showed a significant decrease in the anticonvulsant
properties. As regards the effect of linker modifications, in terms
of both the inversion of the position of carbonyl group and the
length of the spacer, we obtained the following results. Compounds
12a–d, in which the carbonyl function was directly linked to the
nitrogen atom of the 4-benzylpiperidine moiety, generally proved
to be less active than the corresponding compounds 3a–d, with
exception of derivative 12b that protected against seizures show-
ing ED50 values of the same order of magnitude as the ‘hit com-
pound’ 3a. Finally, to confirm the crucial role of the distance
between the positive ionizable nitrogen atom and indole fragment,
the alkyl chain of hit compound 3a was lengthened; this modifica-
tion did not appear to promote activity as it produced to the inac-
tive compound 13a. Moreover, we introduced a further carbonyl
group in the spacer via the replacement of acetyl linker with an
oxalyl fragment for compounds 15a–d; although compound 15a
showed similar activity against audiogenic seizures when com-
pared with parent compound 3a, surprisingly the 15a analogues,
bearing other substituents such as bromine, methoxy, and also a
hydroxyl group (15b–d) did not display in vivo efficacy.
2.2. Anticonvulsant properties
Initially, we examined the impact on anticonvulsant properties
of the planned structural modifications by introducing new sub-
stituents and moieties on the ‘hit compound’ 3a, which had shown
good in vivo activity. The anticonvulsant effects of indole deriva-
tives 3a–d, 9a, 12a–d, 13a, and 15a–d were evaluated against
audiogenic seizures in DBA/2 mice, which has been considered
an excellent animal model for generalized epilepsy and for screen-
ing new anticonvulsant drugs.20 The results were compared with
those of the NMDA/NR2B receptor antagonist Ro 25-6981 (2),
and the ED50 values after intraperitoneal (ip) administration are
presented in Table 1. The pharmacological results obtained high-
lighted some structure–activity relationship considerations for this
class of compounds. The presence of bromine (3b) or methoxy
group (3c) in position 5 of the benzene-fused ring of ‘hit com-
pound’ 3a negatively influences anticonvulsant efficacy. On the
contrary, the introduction of a hydroxyl-substituent in the same
position led to an improvement of anticonvulsant properties so
that compound 3d was two-fold more active than the correspond-
ing unsubstituted derivative 3a and ꢀ4-fold that of Ro 25-6981 (2).
In attempt to find a correlation between the physico-chemical
properties of the tested compounds and the rank order of anticon-
vulsant efficacy, we estimated their relative lipophilicity
(logD7.4).21 However, the different degree of the potency generally
did not seem to be directly related to the logD7.4 values (see Table
1), thus suggesting the influence of other parameters.
Table 1
Anticonvulsant activity of indole derivatives against audiogenic seizures in DBA/2 mice and prediction of their logD7.4 values
X
N
R
z
N
Y
H
A
mol/kga
LogD7.4
b
R
Y
X
Z
ED50
l
Clonus
Tonus
3ac
3bc
3cc
3d
6a
9a
12a
12b
12c
12d
13a
15a
15b
15c
15d
2
H
Br
OMe
OH
H
H
H
Br
OMe
OH
H
H
Br
OMe
OH
—
CH
CH
CH
CH
CH
N
CH
CH
CH
CH
CH
CH
CH
CH
CH
—
COCH2
COCH2
COCH2
COCH2
COCH2
COCH2
CH2CO
CH2CO
CH2CO
CH2CO
(CH2)2CO
COCO
CH
CH
CH
CH
N
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
—
22.4 (13.9–36.2)
84.9 (64.7–111)
59.9 (38.7–92.7)
11.5 (7.96–16.5)
>100
88.5 (52.5–149)
70.3 (53.3–92.9)
27.3 (15.4–48.2)
89.0 (59.9–132)
45.6 (23.4–89.2)
>100
22.4 (10.9–45.9)
>100
>100
>100
40.0 (21.1–54.7)
8.71 (5.71–13.3)
49.7 (34.8–71.0)
36.5 (26.1–51.0)
6.48 (4.56–9.21)
>100
32.5 (22.8–46.4)
39.2 (20.5–77.8)
5.42 (2.08–14.1)
34.4 (21.3–69.1)
13.9 (11.2–17.1)
57.6 (38.4–86.4)
0.95 (0.26–3.52)
47.9 (28.1–81.7)
78.8 (53.7–116)
>100
4.73
5.86
4.90
3.81
3.08
4.85
4.36
5.21
4.27
3.18
4.87
3.28
4.14
3.20
2.11
—
COCO
COCO
COCO
—
10.6 (5.16–21.8)
a
All data were calculated following the Litchfield and Wilcoxon method. At least 32 animals were used to calculate each ED50 95% confidence limits are given in
parentheses.
b
LogD data at pH 7.4 are predicted from a commercially available programme (ACD/Lab).21
Ref. 19.
c

R. Gitto et al. / Bioorg. Med. Chem. 17 (2009) 1640–1647
1643
2.3. Binding affinity
On the basis of the anticonvulsant efficacy and in order to inves-
tigate the mechanism of action we evaluated the ability of some
selected compounds to interact with the NMDA/NR2B receptor
by testing [3H]ifenprodil binding inhibition (Table 2) in Wistar
rat cerebral cortex following a reported procedure (see Section
4). Compound 3d, for which the highest binding affinity was found
with an IC50 of 0.025 lM (see Table 2), showed a 30-fold improve-
ment over the parent compound 3a and a potency comparable to
that of ifenprodil (1). Figure 1 displays the inhibition of [3H]ifen-
prodil in the binding assay; the response curve for the most potent
ligand 3d (IC50 = 0.025
the reference compound ifenprodil (1) (IC50 = 0.0205
Ki = 0.019 M) at different doses. This result suggests that the
lM, Ki = 0.023
lM) is very similar to that for
lM,
l
introduction of the 5-hydroxyl group onto the benzene-fused ring
positively influences the binding affinity. When testing the binding
affinity of the oxalyl-analogues 15a and 15d, we observed a signif-
icant decrease in NR2B binding affinity when compared with that
of 3a and 3d; as expected, a moderate affinity re-emerged in the
5-hydroxyl substituted derivative 15d, although it was inactive
in the in vivo pharmacological evaluation (see Table 1). From these
results we speculated that the presence of a 5-hydroxyl group cou-
pled with a basic amine centre might optimize the NR2B binding
affinity for this class of compounds.
Figure 1. Inhibition of [3H]ifenprodil binding to Wistar rat cerebral cortex22 by 3d
(blue circles) and ifenprodil (1) (red squares).
2.4. Electrophysiological studies
Subsequently, we focused our attention on the 2-(4-benzylpi-
peridin-1-yl)-1-(5-hydroxy-1H-indol-3-yl)ethanone (3d), which
demonstrated the highest anticonvulsant activity and NR2B ifen-
prodil-site receptor binding affinity. With the aim of studying its
mechanism of action, a patch clamp study was also performed on
brain slices from rat hippocampus, prepared as described previ-
ously.22 Membrane currents were recorded from single CA1 pyra-
midal neurons at a holding potential of ꢁ60 mV; application of
Table 2
Structure and binding affinitya of ifenprodil and selected indoles
IC50% at 10
l
M
IC50 (lM)
NMDA (50 lM, 5 s) together with glycine (0.5 lM) evoked an in-
ward current, the amplitude of which ranged between 110 and
O
998 pA in different neurons (Fig. 2A and B). Following continuous
N
application of compound 3d (10
system, the amplitude of NMDA-mediated current was reduced
(Fig. 2A and B). The effect of 3d was dose dependent: at 1 M con-
centration the amount of inhibition was statistically significant in
1 out of 4 cases whereas higher concentrations (10 and 100 M) in-
lM, 5 min) through the perfusion
3a
89
0.769
N
H
l
l
duced a significant inhibition in the majority of neurons tested (5
out of 7, Fig. 2B). Percent reduction values of NMDA-mediated cur-
rent by compound 3d at 1, 10 and 100 lM concentrations were,
O
N
HO
respectively, 23 9, 52 10 and 75 (mean SEM, n = 4, 5 and 1;
Fig. 2C). The effect of 3d was reversible; the amplitude of NMDA-
mediated current returned to control values within 10 min after
the end of 3d application (Fig. 2D).
3d
92
0.0250
N
H
These results indicate that compound 3d behaved as an antag-
onist of NMDA receptors, exerting a dose-dependent and reversible
inhibition of NMDA-mediated current.
O
O
N
2.5. Docking studies
15a
36
N.D.
N
H
Docking simulations were carried out in order to rationalize
the obtained findings for this class of compounds. In our previous
studies19 we modelled the structure of NR2B/ifenprodil-like bind-
ing domain and highlighted the binding mode of ‘hit compound’
3a in the binding pocket ( Fig. 3). The main contact observed
was the electrostatic interaction between the protonated nitrogen
atom of the ligand and the carboxyl group of aspartate D101,
which previous site-directed mutagenesis studies5 had high-
lighted as being crucial for NR2B interaction. Moreover, com-
pound 3a was involved not only in an H bonding interaction
between the indole ring NH and residue E235, but also in hydro-
phobic interactions with some specific residues (T233, V42, L261,
A263, H127, A298, I297 and F176). Notably, the binding pose of
the new identified ‘hit compound’ 3d was very similar to the
one previously reported for reference compound 3a but, as shown
O
O
N
OH
15d
77
2.66
N
H
1
Ifenprodil
—
0.0205
Displacement of [3H]ifenprodil.25 N.D. = not detectable.
a

1644
R. Gitto et al. / Bioorg. Med. Chem. 17 (2009) 1640–1647
control
3d
A
C
B
1 µM
10 µM
100 µM
NMDA
50 µM
*
1200
800
400
0
*
3d 10 µ M
*
**
***
*
500 pA
20 s
control
1
2
3
4
5
6
7
8
9
10 11
Tested neurons
D
3d 10 µ M
75
50
25
0
1.5
1.0
0.5
0.0
µ
M
3d 1 µM
10
100 µM
0
2
4
6
8
10
12
14
16
18
Time (min)
Figure 2. Inhibition of NMDA-mediated current by compound 3d. (A) Patch clamp recording of membrane ion currents from a single CA1 pyramidal neuron (HP ꢁ60 mV) of
rat hippocampus: application of NMDA (50 M, 5 s) induced an inward current, the amplitude of which was strongly reduced during bath perfusion of compound 3d (10 M,
l
l
5 min). (B) The amplitude of NMDA-mediated current (mean SEM of at least three consecutive responses) in the presence of 3d (filled columns) was significantly decreased
with respect to control conditions (empty columns) in most of the neurons examined (***P < 0.0001; **P < 0.001; *P < 0.01; student’s unpaired t-test, two-tailed). (C) The
amount of inhibition of NMDA-mediated current increased with concentration of compound 3d (1, 10 and 100 lM). (D) Normalized amplitude values of NMDA-mediated
currents in different neurons were averaged (mean SEM, n = 6) and represented as a function of time. Inhibition of NMDA-mediated current by 3d (10 lM, 5 min) was
reversible: recovery was observed within 10 min after washout.
tions for 3d and 15d (65.77 and 60.86, respectively) also corre-
lated with their IC50 binding affinity. These findings could justify
the data binding affinity of compound 15d and simultaneously
confirmed that our modelled binding pocket would be able to pre-
dict the ranking order of ligand potency of to binding recognition
sites for this class of compounds.
3. Conclusions
In an attempt to carefully investigate the structure–activity
relationships of new NR2B/antagonists containing indole nucleus,
a set of 3-substituted derivatives were synthesized and tested.
Some of them demonstrated anticonvulsant properties against
audiogenic seizures in DBA2/mice. In particular, the compound
3d with the best in vivo pharmacological profile also showed the
highest binding affinity to NR2B ifenprodil-like receptor subunits.
Notably, we found that the activity of compound 3d was better
than 3a and that it was equipotent to ifenprodil (1) in ‘in vitro’ test.
These results were supported by our docking studies, which clari-
fied its good interaction into the binding pocket.
Moreover, by electrophysiology studies we found that com-
pound 3d was able to reduce NMDA receptor-mediated current
in CA1 pyramidal neurons from rat hippocampus, suggesting that
it could be considered a new potent antagonist of NR2B/NMDA
receptors and might be useful as a neuroprotective agent.
Figure 3. Docking model of compound 3a (cyan) at the modelled binding pocket.
The most important residues identified by mutagenesis-site directed studies5 are
depicted in yellow.
in Figure 4A, the presence of a 5-hydroxyl group on the indole
scaffold of 3d generates the possibility of a new contact as hydro-
gen bonding with the important residue D104. Considering the re-
ported mutations results5 this new contact could lead to more
favourable interactions within the binding pocket and provide a
possible explanation for the improvement of the binding affinity
observed for compound 3d, which resulted 30-fold more potent
with respect to parent compound 3a (see Table 2). In an attempt
to explain the poor in vitro efficacy of compound 15d we studied
its binding pose (Fig. 4B) in comparison with that shown by ana-
logue 3d. We mainly observed that the less active compound 15d,
lacking positive ionizable nitrogen atom, did not show any elec-
trostatic interaction with the carboxyl group of residue D101.
However, it did maintain the hydrophobic contacts and the
favourable H-bonding interaction with residue E235 observed also
for derivative 3d. The GOLD fitness score of the best docking solu-
4. Experimental
4.1. Chemical experiments
Melting points were determined on a Stuart SMP10 apparatus
and are uncorrected. Elemental analyses (C, H, N) were carried

R. Gitto et al. / Bioorg. Med. Chem. 17 (2009) 1640–1647
1645
Figure 4. The best docked conformations of compounds 3d (4A) and 15d (4B) at the modelled binding pocket. The most important residues identified by mutagenesis-site
directed studies5 are depicted in yellow.
out on a Carlo Erba Model 1106 Elemental Analyzer and the results
are within 0.4% of the theoretical values. Merck silica gel 60 F254
plates were used for analytical TLC; column chromatography was
performed on Merck silica gel 60 (70–230 mesh). Flash chromatog-
raphy (FC) was carried out on a Biotage SP4 EXP. 1H NMR spectra
were measured in CDCl3 or dimethylsulphoxide-d6 (DMSO-d6)
with a Varian Gemini 300 spectrometer; chemical shifts are ex-
pressed in d (ppm) and coupling constants (J) in Hz. All exchange-
able protons were confirmed by addition of D2O. GC–MS spectra
for selected compounds were recorded on a Shimadzu QP500 EI
151 mass spectrometer.
(0.005 mol) in the same solvent (2 mL) was added dropwise.23
The reaction mixture was stirred for 2 h, then methanol (5 mL)
was carefully added at 0 °C and the solvent removed under re-
duced pressure. A saturated aqueous NaHCO3 solution (5 mL)
was added and the reaction mixture was extracted with ethyl ace-
tate. The organic phase was dried over Na2SO4 and the solvent re-
moved under reduced pressure to give a crude residue; by
treatment of residue with diethyl ether the intermediate 2-
chloro-1-(1H-pyrrolo[2,3-b]pyridin-3-yl)ethanone (8a) was ob-
tained. Subsequently derivative 8a (0.001 mol), benzylpiperidine
(175 mg, 0.001 mol) and K2CO3 (0.0005 mol) were refluxed in tol-
uene (5 mL) for 3 h. The organic phase was filtered and, after re-
moval of the solvent under reduced pressure, the residue was
powdered by treatment with diethyl ether and recrystallized from
ethanol to give the desired compound 9a. 1H NMR spectral data
confirmed the structure of intermediate 8a and were in accordance
with the literature.23 Yield 42%. Mp 151 °C dec. 1H NMR (DMSO-d6)
1.37–3.00 (m, 9H), 2.52 (d, 2H, J = 6.84, CH2Ph), 3.59 (s, 2H, CH2N),
7.13–8.72 (m, 9H, ArH), 10.95 (br s, 1H, NH).). GC–MS (EI) m/z (%):
333 (M+, 0.5), 189 (17), 188 (100), 174 (17), 145 (18), 91 (22). Anal.
Calcd for C21H23N3O: C, 75.65; H, 6.95; N, 12.60. Found: C, 75.44; H,
6.70; N, 12.88.
4.1.1. Synthesis of 2-(4-benzylpiperidin-1-yl)-1-(5-hydroxy-1H-
indol-3-yl)ethanone (3d)
Derivative 3c (0.001 mol), which was synthesized from com-
pound 5c as previously reported,19 was dissolved in methylene
chloride (DCM) (5 mL), treated with BBr3 (1 M in DCM),(6 mL,
0.006 mol) and stirred overnight. Successively, methanol (7 mL)
was carefully added at 0 °C and the solvents removed under re-
duced pressure. The residue was dissolved in ethyl acetate
(10 mL) and washed with water (10 mL ꢂ 3). The organic layer
was dried (Na2SO4), combined and concentrated in vacuo. The
crude product was crystallized from ethanol to give the desired
compound 3d. Yield 68%. Mp 214 °C dec. 1H NMR (DMSO-d6)
1.14–3.24 (m, 11H), 3.33 (s, 2H, CH2N), 6.68–7.56 (m, 8H, ArH),
8.33 (s, 1H, H-2), 9.00 (s, 1H, OH), 11.66 (br s, 1H, NH), Anal. Calcd
for C22H24N2O2: C, 75.83; H, 6.94; N, 8.04. Found: C, 75.89; H, 6.85;
N, 8.15.
4.1.4. General procedure for the synthesis of 1-(4-benzylpiperi-
din-1-yl)-2-(1H-indol-3-yl)ethanones (12a–c) and 1-(4-benzyl-
piperidin-1-yl)-3-(1H-indol-3-yl)propan-1-one (13a)
A mixture of 2-(1H-indol-3-yl)acetic acid (10a–c) or 3-(1H-in-
dol-3-yl)propanoic acid (11a) (0.001 mol) with N,N,N0,N0-tetramethyl-
O-(1H-benzotriazol-1-yl)uronium hexafluorophosphate (HBTU) in
dimethylformamide (2 mL) was stirred for 30 min at room tempera-
ture. Successively, 4-benzylpiperidine (0.001 mol) and catalytic
amount of triethylamine were added, the mixture was stirred for
2 h at room temperature. The reaction mixture was then quenched
with water (10 mL) and extracted with ethyl acetate. The combined
extracts were dried with dry Na2SO4 and concentrated in vacuo.
The crude compound was crystallized by treatment with ethanol to
give the desired final products (12a–c and 13a).
4.1.2. Synthesis of 2-(4-benzylpiperazin-1-yl)-1-(1H-indol-3-
yl)ethanone (6a)
A solution of 3-chloroacetylindole derivative (5a, 0.001 mol),
previously reported by us,19 benzylpiperazine (0.001 mol) and
K2CO3 (0.0005 mol) was refluxed in toluene (5 mL) for 3 h. The or-
ganic phase was filtered and, after removal of the solvent under re-
duced pressure, the residue was powdered by treatment with
diethyl ether and recrystallized from EtOH to give the desired com-
pound (6a). Yield 49%. Mp 192 °C dec. 1H NMR (CDCl3) 2.55–2.65
(m, 8H), 3.53 and 3.64 (2s, 4H, CH2), 7.25–7.43 (m, 8H, ArH), 8.31
(s, 1H, H-2), 8.40 (br s, 1H, H-4), 8.66 (br s, 1H, NH). GC–MS (EI)
m/z (%): 333 (M+, 12), 190 (14), 189 (100), 175 (13), 146 (9), 91
(34). Anal. Calcd for C21H23N3O: C, 75.65; H, 6.95; N, 12.60. Found:
C, 75.72; H, 6.68; N, 12.40.
4.1.4.1. 1-(4-Benzylpiperidin-1-yl)-2-(1H-indol-3-yl)ethanone
(12a). Yield 82%. Mp 54–55 °C 1H NMR (CDCl3) 0.94–4.66 (m, 9H),
2.49 (mc, 2H, CH2Ph), 3.83 (s, 2H, CH2), 7.07–7.64 (m, 10H, ArH),
8.07 (br s, 1H, NH). GC–MS (EI) m/z (%): 332 (M+, 53), 215 (8), 202
(13), 187 (6), 174 (3), 130 (100), 91 (20). Anal. Calcd for C22H24N2O:
C, 79.48; H, 7.28; N, 8.43. Found: C, 79.55; H, 7.40; N, 8.21.
4.1.3. Synthesis of 2-(4-benzylpiperidin-1-yl)-1-(1H-pyrrolo[2,3-b]
pyridin-3-yl)ethanone (9a)
4.1.4.2. 1-(4-Benzylpiperidin-1-yl)-2-(5-bromo-1H-indol-3-yl)
ethanone (12b). Yield 75%. Mp 72–74 °C 1H NMR (DMSO-d6) 0.90–
4.34 (m, 9H), 2.42 (d, 2H, J = 7.26, CH2Ph), 3.71 (s, 2H, CH2), 7.10–
A
suspension of 7-azaindole (7a) (0.001 mol) and AlCl3
(0.005 mol) in methylene chloride (5 mL) was stirred at room tem-
perature for 1 h, then solution of chloroacetyl chloride
a

1646
R. Gitto et al. / Bioorg. Med. Chem. 17 (2009) 1640–1647
7.72 (m, 9H, ArH), 11.09 (br s, 1H, NH). Anal. Calcd for
C22H23BrN2O: C, 64.24; H, 5.64; N, 6.81. Found: C, 64.20; H, 5.66;
N, 6.83.
4.1.7. Synthesis of 1-(4-benzylpiperidin-1-yl)-2-(5-hydroxy-1H-
indol-3-yl)ethane-1,2-dione (15d)
Compound 15d was prepared following the same procedure
employed for compound 3d using 1-(4-benzylpiperidin-1-yl)-2-
(5-methoxy-1H-indol-3-yl)ethane-1,2-dione 15c as starting mate-
rial. Yield 59%. Mp 140–142 °C 1H NMR (DMSO-d6) 1.02–4.39 (m,
9H), 2.52 (s, 2H, CH2Ph), 6.71–7.47 (m, 8H, ArH), 7.94 (s, 1H, H-
2), 9.13 (s, 1H, OH), 12.01 (br s, 1H, NH). Anal. Calcd for
C22H22N2O3: C, 72.91; H, 6.12; N, 7.73. Found: C, 72.88; H, 6.21;
N, 7.50.
4.1.4.3. 1-(4-Benzylpiperidin-1-yl)-2-(5-methoxy-1H-indol-3-
yl)ethanone (12c). Yield 96%. Mp 60–62 °C. 1H NMR (CDCl3) 0.96–
4.70 (m, 9H), 2.55 (mc, 2H, CH2Ph), 3.91 (s, 3H, MeO), 3.84 (s, 2H,
CH2), 6.89–7.38 (m, 9H, ArH), 8.02 (br s, 1H, NH). Anal. Calcd for
C23H26N2O2: C, 76.21; H, 7.23; N, 7.73. Found: C, 76.44; H, 7.00;
N, 7.55.
4.1.4.4. 1-(4-Benzylpiperidin-1-yl)-3-(1H-indol-3-yl)propan-1-
one (13a). Yield 35%. Mp 40–42 °C. 1H NMR (DMSO-d6) 0.89–4.70
(m, 15H), 7.02–7. 62 (m, 10H, ArH), 8.00 (br s, 1H, NH). GC–MS (EI)
m/z (%): 346 (M+, 100), 216 (27), 174 (42), 143 (64), 13 (79), 117
(19), 91 (17). Anal. Calcd for C23H26N2O: C, 79.73; H, 7.56; N,
8.09. Found: C, 79.80; H, 7.66; N, 8.20.
4.2. Pharmacology testing of anticonvulsant activity
All experiments were performed with DBA/2 mice which are
genetically susceptible to sound-induced seizures. DBA/2 mice
(8–12 g; 22–25-day-old) were purchased from Harlan Italy (Core-
zzano, Italy). Groups of 10 mice of either sex were exposed to audi-
tory stimulation 30 min following administration of either the
vehicle or each dose of the drugs studied.20 The compounds were
administered intraperitoneally (ip) (0.1 mL/10 g of mouse body
weight) as a freshly-prepared solution in 50% dimethylsulphoxide
(DMSO) and 50% sterile saline (0.9% NaCl). Individual mice were
placed under a hemispheric perspex dome (diameter 58 cm), and
60 s were allowed for habituation and assessment of locomotor
activity. Auditory stimulation (12–16 kHz, 109 dB) was applied
for 60 s or until tonic extension occurred, and induced a sequential
seizure response in control DBA/2 mice, consisting of an early wild
running phase, followed by generalized myoclonus and tonic flex-
ion and extension sometimes followed by respiratory arrest. The
control and drug-treated mice were scored for latency to and inci-
dence of the different phases of the seizures. The experimental pro-
tocol and all the procedures involving animals and their care were
conducted in conformity with the institutional guidelines and the
European Council Directive of laws and policies.
4.1.5. Synthesis of 1-(4-benzylpiperidin-1-yl)-2-(5-hydroxy-1H-
indol-3-yl)ethanone (12d)
Compound 12d was synthesized starting from derivative 12c
using the same procedure employed to prepare derivative 3d. Yield
36%. Mp 187–88 °C. 1H NMR (DMSO-d6) 0.89–4.35 (m, 9H), 2.42 (d,
2H, J = 7.26, CH2Ph), 3.61 (s, 2H, CH2CO), 6.55–7. 27 (m, 9H, ArH),
8.57 (s, 1H, OH), 10.54 (br s, 1H, NH). Anal. Calcd for C22H24N2O2:
C, 75.83; H, 6.94; N, 8.04. Found: C, 75.85; H, 6.99; N, 8.11.
4.1.6. General procedure for the synthesis of 1-(4-benzylpiperi-
din-1-yl)-2-(1H-indol-3-yl)ethane-1,2-diones (15a–c)
Oxalyl chloride (0.175 mL, 0.002 mol) was added dropwise at
0 °C to a solution of appropriate indole (0.001 mol) in diethyl
ether (5 mL), under N2 atmosphere. The reaction mixture was
stirred at the same temperature for 2 h; this was followed by con-
centration in vacuo to remove the diethyl ether. To a solution in
tetrahydrofuran (5 mL) of the crude material obtained in previous
4-benzylpiperidine (175 mg, 0.001 mol) and catalytic amount of
triethylamine were added, the mixture was stirred for 2 h at
room temperature. A saturated aqueous NaHCO3 solution (5 mL)
was added to quench the reaction and the mixture was extracted
with ethyl acetate. The combined extracts were dried with
dry Na2SO4 and concentrated in vacuo. The crude compound
was purified by flash chromatography (FC) (cyclohexane/ethyl
acetate, 1:1).
4.3. Statistical analysis
Statistical comparisons between groups of control and drug-
treated animals were made using Fisher’s exact probability test
(incidence of the seizure phases). The ED50 values of each phase
of audiogenic seizures was determined for each dose of the com-
pound administered, and dose–response curves were fitted using
a computer programme by Litchfield and Wilcoxon’s method.24
4.4. Receptor binding studies
4.1.6.1. 1-(4-Benzylpiperidin-1-yl)-2-(1H-indol-3-yl)ethane-
1,2-dione (15a). Yield 45%. Mp 187–188 °C 1H NMR (CDCl3) 1.25–
4.63 (m, 9H), 2.55 (d, 2H, J = 6.87), 7.11–7.40 (m, 8H, ArH), 7.78 (s,
1H, H-2), 8.31 (br s, 1H, H-4), 9.50 (br s, 1H, NH). GC–MS (EI) m/z
(%): 346 (M+, 10), 174 (16), 144 (100), 91 (10). Anal. Calcd for
C22H22N2O2: C, 76.28; H, 6.40; N, 8.09. Found: C, 76.30; H, 6.41;
N, 8.19.
The radioligand binding assays were carried out using [3H]ifen-
prodil (Custom Screen by MDS pharma services, Taiwan).25
4.5. Electrophysiology
Young Wistar rats (age 11–17 days) were decapitated under
ether anaesthesia; brains were rapidly removed, placed in oxygen-
ated ice-cold artificial cerebrospinal fluid (ACSF; composition in
mM: NaCl 124; KCl 3.0; NaH2PO4 1.2; MgSO4 1.2; CaCl2 2.0; NaH-
4.1.6.2. 1-(4-Benzylpiperidin-1-yl)-2-(5-bromo-1H-indol-3-
yl)ethane-1,2-dione (15b). Yield 48%. Mp 73–75 °C 1H NMR
(CDCl3) 1.25–4.60 (m, 9H), 2.56 (d, 2H, J = 6.87), 7.11–7.35 (m,
7H, ArH), 7.66 (s, 1H, H-2), 8.44 (br s, 1H, H-4), 10.05 (br s, 1H,
NH). Anal. Calcd for C22H21BrN2O2: C, 62.13; H, 4.98; N, 6.59.
Found: C, 62.15; H, 4.77; N, 6.66.
CO3 26; D-glucose 10, pH 7.3) and cut into 300 lm slices with a
vibratome. One slice containing the hippocampus was then trans-
ferred to the recording chamber of a patch-clamp set-up, continu-
ally perfused with oxygenated ACSF and viewed under an infrared
differential interference contrast microscope (Leica DMLFS). CA1
pyramidal neurons were visually identified by their location and
by their typical shape and dimension. Membrane current was re-
corded from single CA1 pyramidal neurons with the patch clamp
technique in the whole-cell configuration, using an L/M-EPC7
amplifier (List Electronic, Germany). The recording electrode was
4.1.6.3. 1-(4-Benzylpiperidin-1-yl)-2-(5-methoxy-1H-indol-3-
yl)ethane-1,2-dione (15c). Yield 58%. Mp 92–94 °C 1H NMR
(CDCl3) 1.25–4.63 (m, 9H), 2.56 (d, 2H, J = 6.87), 3.89 (s, 3H,
MeO), 6.90–7.30 (m, 7H, ArH), 7.76 (s, 1H, H-2), 7.81 (br s, 1H, H-
4), 9.20 (br s, 1H, NH). Anal. Calcd for C23H24N2O3: C, 73.38; H,
6.43; N, 7.44. Found: C, 73.43; H, 6.58; N, 7.60.
a glass micropipette (resistance 1.5–3 MX) filled with intracellular

R. Gitto et al. / Bioorg. Med. Chem. 17 (2009) 1640–1647
1647
solution (composition in mM: K-gluconate 170; HEPES 10; NaCl
10; MgCl2 2; EGTA 0.2; Mg-ATP 3.5; Na-GTP 1; pH 7.3). Capacitive
currents were electronically cancelled and series resistance was
compensated by 60–80%. Data were acquired and analysed using
EPC and Signal softwares (Cambridge Electronic Design, England).
Immediately after the beginning of recording, the slice was per-
hydrogen bonds and 4.0 Å for VdW were employed. Results dif-
fering by less than 1.5 Å in ligand-all atom RMSD were clustered
together.
Acknowledgements
fused with a Mg++-free ACSF containing tetrodotoxin (TTX, 1
Tocris) to prevent synaptic activity. N-methyl- -aspartic acid
M, Sigma) was dissolved in Mg++-free ACSF containing
M, Sigma) and TTX (1 m) and was applied during 5 s
lm,
We gratefully acknowledge expert technical assistance at Salva-
tore Todaro of Dipartimento Farmaco-Chimico—Università di Mes-
sina. Financial support for this research by MiUR is gratefully
acknowledged.
D
(NMDA, 50
l
glycine (0.5
l
l
using a computer-controlled electrovalve solution changer (MSC-
200, Bio-Logic). Compound 3d was dissolved in Mg++-free ACSF
References and notes
(containing TTX, 1 lm) and applied by bath perfusion at a flow rate
of 1 mL/min. The holding potential of CA1 pyramidal neurons was
set at ꢁ60 mV and membrane current was continually recorded at
an acquisition rate of 4 KHz. Several pulses of NMDA were applied
for each neuron respectively in control conditions and during con-
1. Paoletti, P.; Neyton, J. Curr. Opin. Pharmacol. 2007, 7, 39.
2. Mayer, M. L. Curr. Opin. Neurobiol. 2005, 15, 282.
3. Chazot, P. L. Curr. Med. Chem. 2004, 11, 389.
4. Marinelli, L.; Cosconati, S.; Steinbrecher, T.; Limongelli, V.; Bertamino, A.;
Novellino, E.; Case, D. A. Chem. Med. Chem. 2007, 2, 1498.
5. Malherbe, P.; Mutel, V.; Broger, C.; Perin-Dureau, F.; Kemp, J. A.; Neyton, J.;
Paoletti, P.; Kew, J. N. J. Pharmacol. Exp. Ther. 2003, 307, 897.
6. Loftis, J. M.; Janowsky, A. Pharmacol. Ther. 2003, 97, 55.
7. Nikam, S. S.; Meltzer, L. T. Curr. Pharm. Des. 2002, 8, 845.
8. Borza, I.; Kolok, S.; Gere, A.; Nagy, J.; Fodor, L.; Galgoczy, K.; Fetter, J.; Bertha, F.;
Agai, B.; Horvath, C.; Farkas, S.; Domany, G. Bioorg. Med. Chem. Lett. 2006, 16,
4638.
9. Borza, I.; Greiner, I.; Kolok, S.; Galgoczy, K.; Ignacz-Szendrei, G.; Horvath, C.;
Farkas, S.; Gati, T.; Hada, V.; Domany, G. Pharmazie 2006, 61, 799.
10. Borza, I.; Domany, G. Curr. Top. Med. Chem. 2006, 6, 687.
11. McCauley, J. A. Exp. Opin. Ther. Patent 2005, 15, 389.
12. McCauley, J. A.; Theberge, C. R.; Romano, J. J.; Billings, S. B.; Anderson, K. D.;
Claremon, D. A.; Freidinger, R. M.; Bednar, R. A.; Mosser, S. D.; Gaul, S. L.;
Connolly, T. M.; Condra, C. L.; Xia, M. H.; Cunningham, M. E.; Bednar, B.; Stump,
G. L.; Lynch, J. J.; Macaulay, A.; Wafford, K. A.; Koblan, K. S.; Liverton, N. J. J. Med.
Chem. 2004, 47, 2089.
13. Barta-Szalai, G.; Borza, I.; Bozo, E.; Kiss, C.; Agai, B.; Proszenyak, A.; Keseru, G.
M.; Gere, A.; Kolok, S.; Galgoczy, K.; Horvath, C.; Farkas, S.; Domany, G. Bioorg.
Med. Chem. Lett. 2004, 14, 3953.
tinuous application of 3d at different doses (1, 10, and100 lM,
5 min) through the bath perfusion system. NMDA-induced current
amplitude was measured as the difference between peak current
amplitude and baseline. For each neuron studied, the amplitude
of at least three consecutive NMDA-induced currents were mea-
sured respectively in control conditions and in the presence of
compound 3d and the two sets of raw values were compared using
Student’s unpaired t-test (Graph Pad software). To illustrate the
time course of NMDA-mediated current blockade by 3d (10 lM),
current amplitude values were normalized (dividing by mean
amplitude value, calculated from at least three NMDA responses
in control conditions); normalized NMDA-mediated currents from
all the responsive neurons were pooled (mean SEM) and repre-
sented on a graphic as a function of time.
14. Claiborne, C. F.; McCauley, J. A.; Libby, B. E.; Curtis, N. R.; Diggle, H. J.;
Kulagowski, J. J.; Michelson, S. R.; Anderson, K. D.; Claremon, D. A.; Freidinger,
R. M.; Bednar, R. A.; Mosser, S. D.; Gaul, S. L.; Connolly, T. M.; Condra, C. L.;
Bednar, B.; Stump, G. L.; Lynch, J. J.; Macaulay, A.; Wafford, K. A.; Koblan, K. S.;
Liverton, N. J. Bioorg. Med. Chem. Lett. 2003, 13, 697.
15. Menniti, F. S.; Pagnozzi, M. J.; Butler, P.; Chenard, B. L.; Jaw-Tsai, S. S.; White, W.
F. Neuropharmacology 2000, 39, 1147.
16. Chenard, B. L.; Menniti, F. S. Curr. Pharm. Des. 1999, 5, 381.
4.6. Automated molecular docking experiments
Prior to docking, the ligand structures were built using stan-
dard bond lengths and angles from the Sybyl 8.2 fragment library
and fully optimized by the semiempirical quantum mechanical
method AM1. The structure of the NR2B ifenprodil-like recogni-
tion site was recently reported by us.19 Automated docking stud-
ies were then performed using CCDC GOLD (Genetic Optimization
for Ligand Docking) software version 3.1.1.26 A 10.0 radius active
site was defined centred on the position of ligand Ro-25,6981, as
previously reported.19 For all calculations GOLD score was chosen
as fitness function and the standard default settings were used in
all the calculations. For each of the 100 independent genetic algo-
rithm runs, a default maximum of 100,000 genetic operations
was performed, using the default operator weights and a popula-
tion size of 100 chromosomes. Default cutoff values of 2.5 Å for
17. Gogas, K. R. Curr. Opin. Pharmacol. 2006, 6, 68.
18. Kosowski, A. R.; Liljequist, S. J. Pharmacol. Exp. Ther. 2004, 311, 560.
19. Gitto, R.; De Luca, L.; Ferro, S.; Occhiuto, F.; Samperi, S.; De Sarro, G.; Russo, E.;
Ciranna, L.; Costa, L.; Chimirri, A. Chem. Med. Chem. 2008, 3, 1539.
20. Chapman, A. G.; Croucher, M. J.; Meldrum, B. S. Arzneimittelforschung 1984, 34,
1261.
21. Advanced; Chemistry; Development. www.acdlabs.com/.
22. Ciranna, L.; Cavallaro, S. Exp. Neurol. 2003, 184, 778.
23. Galvez, C.; Viladoms, P. J. Heterocycl. Chem. 1982, 19, 665.
24. Litchfield, J. T., Jr. J. Pharmacol. Exp. Ther. 1949, 97, 399.
25. MDS pharma services, T. http://www.mdsps.com/.
26. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. J. Mol. Biol. 1997, 267,
727.