Discovery of new potent dual sigma receptor/GluN2b ligands with antioxidant property as neuroprotective agents

Article abstract of DOI:10.1016/j.ejmech.2019.07.012

Among several potential applications, sigma receptors (σRs) can be used as neuroprotective agents, antiamnesic, antipsychotics and against other neurodegenerative disorders. On the other hands, antagonists of the GluN2b-subunit-containing-N-methyl-D-aspar

Full text of DOI:10.1016/j.ejmech.2019.07.012

Accepted Manuscript
Discovery of new potent dual sigma receptor/GluN2b ligands with antioxidant property
as neuroprotective agents
Daniele Zampieri, Sara Fortuna, Antonella Calabretti, Maurizio Romano, Renzo
Menegazzi, Dirk Schepmann, Bernhard Wünsch, Simona Collina, Davide Zanon,
Maria Grazia Mamolo
PII:
S0223-5234(19)30633-6
DOI:
https://doi.org/10.1016/j.ejmech.2019.07.012
EJMECH 11509
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 14 May 2019
Revised Date: 20 June 2019
Accepted Date: 4 July 2019
Please cite this article as: D. Zampieri, S. Fortuna, A. Calabretti, M. Romano, R. Menegazzi, D.
Schepmann, B. Wünsch, S. Collina, D. Zanon, M.G. Mamolo, Discovery of new potent dual sigma
receptor/GluN2b ligands with antioxidant property as neuroprotective agents, European Journal of
Medicinal Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.07.012.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
(a) Docking of compounds 24 (KiGluN2b = 27 nM) and (b) 28 (KiGluN2b = 2.0 nM)
into the binding site of GluN2b subunit.

ACCEPTED MANUSCRIPT
Discovery of new potent dual sigma receptor/GluN2b ligands with antioxidant
property as neuroprotective agents.
a*
a*
a
b
b
Daniele Zampieri , Sara Fortuna , Antonella Calabretti , Maurizio Romano , Renzo Menegazzi ,
c
c
d
e
a
Dirk Schepmann , Bernhard Wünsch , Simona Collina , Davide Zanon , Maria Grazia Mamolo .
a
Department of Chemistry and Pharmaceutical Sciences, Piazzale Europa 1- Via Giorgieri 1,
University of Trieste, 34127 Trieste, Italy.
b
Department of Life Sciences, Via Valerio 28/1, University of Trieste, 34127 Trieste, Italy.
c
Institut für Pharmazeutische und Medizinische Chemie Westfälische Wilhelms-Universität
Münster, Corrensstraße 48, D-48149 Münster, Germany.
d
Department of Drug Sciences, Medicinal Chemistry and Pharmaceutical Technology Section,
Viale Taramelli 6 and 12, University of Pavia, 27100 Pavia, Italy.
e
Pharmacy and Clinical Pharmacology Department Institute for Maternal and Child Health IRCCS
Burlo Garofolo Via dell’Istria 65/1, 34137 Trieste, Italy.
*
Corresponding authors: Prof. Daniele Zampieri; tel: +390405587858; email: dzampieri@units.it;
Prof. Sara Fortuna (molecular simulations); email: s.fortuna@units.it; web: http://sarafortuna.eu/
Keywords
Sigma receptors (σRs), GluN2B receptor, pan-affinity, molecular dynamics, antioxidant activity,
neuroprotective activity.
1

ACCEPTED MANUSCRIPT
Abstract
Among several potential applications, sigma receptors (σRs) can be used as neuroprotective agents,
antiamnesic, antipsychotics and against other neurodegenerative disorders. On the other hands,
antagonists of the GluN2b-subunit-containing-N-methyl-D-aspartate (NMDA) receptors are of
major interest for the same purpose, being this subunit expressed in specific areas of the central
nervous system and responsible for the excitatory regulation of nerve cells. Under these premises,
we have synthesized and biologically tested novel hybrid derivatives obtained from the combination
of phenyloxadiazolone and dihydroquinolinone scaffolds with different amine moieties, peculiar of
σ2R ligands. Most of the new ligands exhibited a pan-affinity towards both σR subtypes and high
affinity against GluN2b subunit. The most promising compounds belong to the dihydroquinolinone
series, with the best affinity profile for the cyclohexylpiperazine derivative 28. Investigation on
their biological activity showed that the new compounds were able to protect SH-SY5Y cells
against oxidative stress induced by hydrogen peroxide treatment. These results proved that our
dual σR/GluN2b ligands have beneficial effects in a model of neuronal oxidative stress and can
represent strong candidate pharmacotherapeutic agents for minimizing oxidative stress-induced
neuronal injuries.
2

ACCEPTED MANUSCRIPT
1
. Introduction
After initial cloning in 1996, the Sigma 1 receptor gene (σ1R) was found to be evolutionarily
conserved. Its encoded protein consists of 223 amino acids with a predicted molecular weight of
5.3 kDa [1,2, 3], and was crystallized in 2016, revealing a trimeric protein organization [4]. The
2
σ1R subtype is particularly enriched in mitochondrion-associated endoplasmic reticulum
membranes (MAM) of neuronal and peripheral cells, such as myocardiocytes and hepatocytes.
They can also translocate to the plasma membrane or ER-membrane and regulate other proteins, as
2
+
+
+
-
well as act as chaperone, modulating different ionic channels (Ca , K , Na , Cl ) via an IP -
3
indipendent mechanism [5,6]. The σ1Rs exert a modulatory role on many transduction systems such
as muscarinic, dopaminergic, serotoninergic and on the NMDA-stimulated neurotransmitter’s
release [7].
Regarding functions, the σ1Rs have neuroprotective and antiamnesic activities [8], and are involved
in modulation of opioid analgesia [9] and drug addiction [10]. Alongside, σ1 antagonists seem to be
effective against the negative manifestations of schizophrenia, without producing extrapyramidal
side effects [11,12].
In addition, several studies suggest a role for σ1R in tumor biology, supported by the observation
that its expression is increased in some cancers and by the growth inhibiting effects of putative
antagonists [13].
In 2017, after 40 years from the discovery of σRs [14], the σ2R subtype has been purified and
identified as the transmembrane protein-97 (TMEM97), also named Meningioma-associated protein
30 (MAC-30) [15]. It is a conserved protein consisting of 176 amino acids with a predicted
molecular weight of 21 kDa, localized to the plasma membrane, endoplasmic reticulum and nuclei.
TMEM97/σ2R is a member of the insulin-like growth factor-binding protein family, and plays a
role in cholesterol homeostasis regulation [16]. This role was then further confirmed by a study
reporting how the interaction of TMEM97 with the progesterone receptor membrane component-1
and the LDL receptor, was critical for the internalization of LDLs [17, 18].
TMEM97/σ2R is implicated in several cellular processes, such as proliferation, signal
transduction, apoptosis and autophagy. In fact, siramesine (SRMS), a specific σ2R agonist, was
shown to trigger apoptosis by caspase activation, autophagy by mTOR kinase inhibition, as well as
cell-cycle alterations [19].
The σ2R is widely distributed in brain and, particularly, in cerebellum, red nucleus and substantia
nigra. Therefore, it is a potential target for the treatment of motor control in movement disorders
3

ACCEPTED MANUSCRIPT
and for counteracting the extrapyramidal side effects produced by neuroleptics drugs [20].
Interestingly, not only σ1R antagonists [21], but also σ2R agonists are involved in neuropathic pain
[
22].
In addition, TMEM97/σ2R seems to have also an impact on tumor growth. Except for pancreatic
and renal cancers [23], it is overexpressed in several types of cancers, and consequently associated
with tumor progression, poor survival and recurrence [24-29].
These observations have suggested the use of σ2 receptor agonists as potential therapeutics for
the treatment of cancer and that TMEM97 could be used as potential prognostic biomarkers of non-
small cell lung carcinoma (NSCLC), squamous cell lung carcinoma (SQCLC), ovarian and breast
cancers [30].
On the other hand, the N-methyl-D-aspartate receptors (NMDARs) are a heterogeneous class of
glutamate-responsive ion-channel receptors located in the post-synaptic membranes of most
excitatory synapses [31]. These proteins, along with AMPA (2-amino-3(3-hydroxy-5-
methylisoxazol-4-yl)propionate) and Kainate, belong to the ionotropic (S)-Glutamate receptors
(iGluRs) family. Structurally, NMDARs are heterotetrameric complexes with four distinct
components derived from three related families, named GluN1, GluN2 and GluN3. In particular,
NMDARs consist of two GluN1 and two GluN2 subunits [32], but one GluN1 can be replaced by
one GluN3 subunit [33].
A further level of structural and functional complexity stems from the observation that up to 8
isoforms can be generated by alternative splicing of the single GluN1 gene (GluN1a-h), while the
GluN2 type is encoded by four genes (GluN2a-d) and the GluN3 by two genes (GluN3a-b) [34].
Overstimulation of NMDARs, as a consequence of (S)-glutamate excess and the subsequent
2+
uncontrolled neuronal influx of Ca ions, causes excitotoxicity and triggers cell death by apoptosis.
Importantly, this phenomenon is responsible for the onset and progression of several
neurodegenerative diseases [35].
Likewise, for σRs, the NMDARs play key roles in synaptic transmission, synaptic plasticity,
neuronal development, learning, memory and other physiological and pathological processes [36,
37]. Hence, antagonists of NMDAR (and, in particular, of the GluN2 subunits) are of interest as
potential neuroprotective drugs to treat several central nervous system (CNS) disorders. While high-
affinity NMDAR-inhibitors include psychotomimetic and neurotoxic agents (such as phencyclidine
and (+)-MK-801), the low-affinity NMDAR- antagonists (such as memantine and amantadine)
represent a class of drugs without such effects already used clinically to treat Alzheimer’s and
Parkinson’s diseases.
4

ACCEPTED MANUSCRIPT
Ifenprodil (1, Figure 1A) [38] is one of the prototypical allosteric inhibitors which interacts only
with the GluN2b-containing NMDARs [39].
In 2011, the binding site of the selective GluN2b ligands was characterized and it was found to be
at the interface between the GluN1 and GluN2b subunits [40]. In 2014, the full heterotetrameric
NMDAR protein has been co-crystallized with Ifenprodil 1 (4-[2-(4-benzylpiperidin-1-yl)-1-
hydroxypropyl]phenol) [41] and an its partial agonist analogue Ro-25-6981 2 (Figure 1A) [42].
Under these premises, a multi-targets drugs (MTD) approach, such as dual σR/GluN2bR
modulators, can be beneficial for the enhancement of neuroprotection to treat several CNS
disorders.
5

ACCEPTED MANUSCRIPT
In our past work, we have synthesized some new 1-(4-(aryl(methyl)amino)butyl)-heterocyclic
derivatives [43], showing high affinity and selectivity towards the σ1R subtype. Their Ki-values
were in the low nanomolar range (0.95 nM-3 µM), but demonstrated a very low affinity for the σ2R
subtype (Ki = 42 nM-3µM) (Figure 1B).
Aiming to improve the σ2R affinity of our previously synthesized compounds, and concurrently
discover novel GluN2bR ligands, we designed and synthesized new molecules obtained by
retaining two of the most representative scaffolds of the previous series and by jointly replacing the
N-benzylmethylamine (and the 4-chlorobenzyl analogue) fragment with other amine moieties
present in some well-known σ2R ligands.
The preserved structures were the 5-phenyl-1,3,4-oxadiazole-2(3H)-one (3a,b) and the
dihydroquinolin-2(1H)-one (3c,d highlighted in red, Figure 1B) scaffolds, which provided the best
compromise between σ1 profile and ease of synthesis.
The selected amine moieties were the spiro[isobenzofuran-1,4’-piperidine] (4, SRMS), N-
cyclohexylpiperazine (5, PB-28), N-(4-fluorophenyl)piperazine (6, SN-79), N-(pyridin-2-yl)-
piperazine (7), 6,7-dimethoxy-3,4-dihydroisoquinoline (8, RHM-1), 7-nitro-3,4-dihydroisoquinoline
(9) and 2,4-dimethylbenzyl-N-methylamine (10, our SRMS analogue, named DZ-24) [44]
(highlighted in blue, Figure 1C).
O
CH
3
O
R
N
O
N
N
N
R
N
N
OH
4 (SRMS)
Kiσ1 = 17 nM
Kiσ2 = 0.12 nM
5 (PB-28)
Kiσ1 = 0.38 nM
Kiσ2 = 0.68 nM
HO
1 (Ifenprodil)*
3a R = H
3
b R = Cl
N
(AMINES)
F
N
3
H C
N
N
O
HO
N
N
R
R
OH
N
R
N
6
(SN-79)
Kiσ1 = 27 nM
Kiσ2 = 7 nM
7
Kiσ1 = 83 nM
Kiσ2 = 4.9 nM
CH
3
2 (Ro-25-6981)
CH
3
3
c R = H
d R = Cl
N
OCH
OCH
3
3
3
R
3
H C
N
C
N
N
R
H
3
R
NO
2
R
CH
3
9
8
(RHM-1)
Kiσ1 = 10320 nM
10 (DZ-24)
Kiσ1 = 129 nM
Kiσ2 = 5.9 nM
Kiσ1 = 3078 nM
Kiσ2 = 10.3 nM
(A)
(B)
Kiσ2 = 26.8 nM
(C)
O
O
O
N
N
N
20-26
27-33
AMINES
AMINES
6

ACCEPTED MANUSCRIPT
Figure 1. Design of target compounds as dual σR/GluN2b ligands. (A) Structure of two known references GluN2b
ligands. *Ifenprodil represents the racemic erythro diastereomer. (B) Selected scaffold from our previous work. (C)
Selected amine scaffolds from well-known σ2 ligands and their Kiσ1 and Kiσ2 values (nM).
As already pointed out, the rigid piperidine(methyl) central scaffold of ifenprodil, can be replaced
by the more flexible butylamine spacer (highlight in purple, Figure 1A), without altering the affinity
[
[
45]. The same structural central motif is often present in many σR ligands, as previously reported
43, 46]. Moreover, the selectivity σ2/σ1 can be modulated by substitution of one of the two
aromatic side scaffolds. Indeed, the substitution of a simple phenyl ring with the dimethylphenyl
ring causes a marked shift of the selectivity from σ1R to σ2R subtype. Taking this into account, we
planned to synthesize the new hybrid molecules 20-33 (Scheme 1).
Scheme 1. Synthesis of title compounds 20-33. Reagents and conditions: a) K CO , KI (cat) or CsCO , TBAB (cat),
2
3
3
ACN, reflux; b) Amine 13-19, K CO , KI (cat), ACN, reflux.
2
3
2
2
. Results and discussion
.1 Chemistry
On one hand, the synthesis of title compounds 20-26 was carried out from 5-phenyl-1,3,4-
oxadiazole-2(3H)-one, obtained from the cyclization of benzoic acid hydrazide with triphosgene
[
43]. On the other, derivatives 27-33 were obtained from the commercially available 3,4-
dihydroquinolin-2(1H)-one. These two precursors were made react with an excess of 1,4-
7

ACCEPTED MANUSCRIPT
dibromobutane in basic media, in order to obtain the intermediate compounds 11 and 12,
respectively. The latter underwent nucleophilic substitution, in the presence of potassium carbonate
and KI as catalyst, with the corresponding amines 13-19, to afford the final compounds 20-33 in
good yields. The amines 13-19 were commercially available with the exception of 7-nitro-1,2,3,4-
tetrahydroisoquinoline 18 and 1-(2,4-dimethylphenyl)-N-methylmethanamine 19. The first was
prepared by nitration of 1,2,3,4-tetrahydroisoquinoline with H SO /KNO and then converted into
2
4
3
hydrochloride salt. The latter was prepared via a classical indirect reductive amination involving
,4-dimethylbenzaldehyde and methylamine to form the corresponding Schiff base, which was
subsequently reduced with NaBH to afford the desired intermediate 19.
2
4
2.2 Receptor affinity, selectivity and SAR exploration
The synthesized compounds 20-33 were investigated for their potential affinity towards σ1R, σ2R
and GluN2b subunit containing NMDA receptors. Guinea pig brain and rat liver homogenates were
3
used as sources for the σ1R and σ2R assays, respectively. While the [ H]-(+)-pentazocine was
3
utilized as selective σ1R radioligand, the [ H]-1,3-di(o-tolyl)guanidine was utilized as nonselective
σ2R radioligand, in the presence of nontritiated (+)-pentazocine (to selectively occupy the σ1Rs).
The affinity towards GluN2b subunit was performed in a radioligand binding assay developed by
the Authors [47]. Briefly, L(tk-) cells stably transfected with a vector containing the genetic
information for the GluN1a and GluN2B subunits of the NMDA receptor provided the receptor
3
material. Using membrane preparations of these cells and [ H]-labeled ifenprodil as radioligand, the
affinity of compounds was determined in a competition assay. The data are collected in Table 1.
Table 1. GluN2b, σ1R and σ2R affinity and the corresponding selectivity ratios of title compounds 20-33 and reference
compounds. The blood brain barrier (BBB) score of title compounds is related to the ability to diffuse into the CNS
.
a
Ki [nM]
σ 2
Selectivity ratio
GluN2b σ1/σ2 σ2/GluN2b
Cpd
R
BBB score
4.38
σ1
2
2
0
1
19 ± 8
16 ± 5
8.2 ± 3
23 ± 7
1.2
2.0
4.0
35 ± 5
93 ± 35
0.37
4.25
8

ACCEPTED MANUSCRIPT
2
2
2
3
383
944
686
68 ± 7
226
0.57
0.33
10
3.98
4.84
2800
12.4
2
4
1600
864
27 ± 6
1.8
32
4.35
2
2
2
5
6
7
829
383
412
2.2
0.95
0.5
0.9
0.8
6.8
4.07
3.34
4.57
20 ± 9
18 ± 4
21 ± 10
26 ± 2
34 ± 15 5.0 ± 0.9
1.5 ± 1.1 2.0 ± 0.2
2
2
3
8
9
0
21 ± 2
16 ± 7
14
0.76
1.6
0.75
2.6
4.22
3.91
5.44
21 ± 11
8.1 ± 3
78 ± 6
120 ± 19 73 ± 22
116 ± 32 45 ± 22
0.9
3
1
12 ± 5
2.6
3.7
4.91
3
3
2
3
179
80 ± 3
32 ± 9
58 ± 10
2.2
1.9
1.4
6.4
4.60
2.98
60 ± 18
5.0 ± 1.4
Ifenprodil
Haloperidol
DTG
-
-
-
125 ± 24 98 ± 34 10 ± 0.7
1.3
0.08
1.5
9.8
nc
nc
4.41
5.54
nc
6.3 ± 1.6 78 ± 2.3
89 ± 29 57 ± 18
nt
nt
a
b
Only the most potent compounds (Ki < 150 nM) were tested in triplicate; 0-2 don’t cross BBB, 2-5 may cross BBB,
-6 effectively cross BBB; nt: not tested; nc: not calculated.
5
Although the phenyloxadiazolone series 20-26 has provided higher Ki values than the
corresponding dihydroquinolinone series 27-33, some derivatives have proved to be rather
interesting by confirming some known evidences. Among them, compounds 20, 21 and 26 showed
high σ1R affinity, comparable to our previously parent derivatives. In fact, the Kiσ1 values were
ranging from 19 to 35 nM (5.2−18.7 nM for 3a,b), and the Kiσ2 values were 16 nM, 23 nM and 21
nΜ, respectively (versus Kiσ2R values ranging from 110 to 315 nM for 3a-d [43]). Therefore, these
ligands showed an increased affinity towards σ2R subtype.
9

ACCEPTED MANUSCRIPT
These results are supported by the presence of the bulky spiro[isobenzofuran-1,4’-piperidine],
cyclohexylpiperazine and 2,4-dimethylbenzyl-N-methylamine moieties, confirming their
importance for the σ2R affinity/selectivity, with respect to benzyl-N-methylamine or 4-
chlorobenzyl-N-methylamine scaffold of our previously compounds 3a-d. Interestingly, compound
20 showed the best inhibition value towards GluN2b subunit (Ki = 8.2 nM), even better than the
reference drug ifenprodil. On the contrary, the 6,7-dimethoxy-3,4-dihydroisoquinoline derivative 24
exhibited the best selectivity profile, with a 32-fold preference for GluN2b receptor (Ki = 27 nM)
over σ2R (Ki = 864 nM) and could be considered the hit compound to design new selective
GluN2bR ligands.
Regarding the homologous subseries of dihydroquinolinones 27-33, all the compounds showed a
moderate to high σR pan-affinity, with the best result of the entire series against σ1R, for the 4-
fluorophenylpiperazinyl derivative 29 (Kiσ1R = 16 nM). Conversely, the cyclohexylpiperazinyl
derivative 28 showed the best affinity against σ2R (Ki = 1.5 nM) and GluN2bR (Ki = 2 nM), along
with a favourable σR selectivity ratio (σ1/σ2 = 14). Finally, the spiro[isobenzofuran-1,4’-
piperidine] 27 and the 2,4-dimethylbenzyl-N-methylamino 33 derivatives resulted to be twice as
powerful as ifenprodil over GluN2bR (Ki = 5 nM) and demonstrated a comparable selectivity (6.4
fold versus 9.8).
2.3 In-silico properties
In order to predict their drug likeness and the ability to reach the CNS, the compounds 20-33 were
also in silico evaluated for their physiochemical and pharmacokinetics parameter (ADME). For this
latter purpose, we used the following molecular descriptors: molecular weight (MW); acid
dissociation constant (pKa); number of H-bond donors (HBD); calculated partition coefficient
(clogP); calculated distribution coefficient (clogD); and topological polar surface area (TPSA).
These features have been considered following the algorithm of central nervous system
multiparameter optimization (CNS MPO), developed by Wager et al. [48] which attributes a value
between 0 and 1 for each aforementioned parameter. The collective score, ranging from 0 to 6,
gives an indication of the drug’s ability in crossing the BBB and precisely: i) 0-2 the compound
does not cross the BBB; ii) 2-4 the compound can reach the CNS; iii) 5-6 the compound surely
crosses the BBB. We evaluate all the final scores of the entire series 20-33, in comparison with
ifenprodil, siramesine and haloperidol as reference drugs (Table SI1). All the synthetized
compounds showed a good score in the median range (2-5), and the derivative 30 was the best
scored with a value > 5. Furthermore, all the compounds do not violate more than one score of the
extended version of Lipinski’s rule of five (RO5: MW≤ 500; HBA ≤ 10 and HBD ≤ 5, respectively;
1
0

ACCEPTED MANUSCRIPT
logP and logS ≤ 5; TPSA ≤ 140 Å) [49], confirming their good drug likeness and the potential
ability to be orally active in humans. Altogether these results support the hypothesis that all the
derivatives should be able to cross the BBB and reach the CNS.
Table 2. In silico CNS MPO of main pharmacokinetic parameters, and relative scores, of compounds 20-33 to define
their ability in crossing the BBB. (Marvinsketch® Chemaxon).
a
b
c
d
e
f
Cpd
MW
(
Score HBD Score pKa Score clogP Score clogD Score TPSA Score Final RO5
h
KDa) MW
HBD
pKa
logP
logD (Å)
TPSA score viol.
g
BBB
2
2
0
1
405.49 0.47
384.52 0.58
0
0
1
1
9.34 0.33
9.04 0.48
3.84
4.13
0.58
0.43
1.90
2.48
1
0.76
54.37
48.38
1
1
4.38
4.25
0
0
2
2
2
2
2
2
2
2
3
3
2
3
4
5
6
7
8
9
0
1
396.46 0.52
379.46 0.60
409.48 0.45
394.42 0.53
365.47 0.67
390.52 0.55
369.54 0.70
381.49 0.59
364.48 0.77
394.51 0.53
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
1
1
8.28 0.86
8.09 0.95
8.18 0.91
4.36
3.60
3.93
4.19
5.23
3.34
3.62
3.86
3.09
3.43
0.32
0.7
0.53
0.40
0
0.83
0.69
0.57
0.95
0.78
3.43
2.82
3.08
3.65
3.08
1.64
2.00
3.14
2.51
2.79
0.28
0.59
0,46
0.17
0.46
1
48.38
61.27
63.60
90.96
45.14
32.78
26.79
26.79
39.68
42.01
1
1
1
0.97
1
0.64
0.34
0.34
0.98
1
3.98
4.84
4.35
4.07
3.34
4.57
4.22
3.91
5.44
4.91
0
0
0
0
1
0
0
0
0
0
7.79
1
9.57 0.21
9.09 0.45
9.02 0.49
8.03 0.98
1
0.43
0.74
0.60
7.84
7.93
1
1
3
3
2
3
379.45 0.60
350.50 0.75
Ifenp. 325.45 0.87
SRMS 454.59 0.23
Haloo. 375.87 0.62
0
0
2
0
1
1
1
0.50
1
0.75
7.53
1
3.68
4.73
3.57
6.58
3.66
0.66
0.13
0,71
0
3.31
2.87
2.30
4.43
2.93
0.34
0.56
0.85
0
69.37
23.55
43.70
17.40
40.54
1
0.18
1
0
1
4.60
2.98
4.41
1.45
5.54
0
0
0
1
0
9.27 0.36
9.03 0.48
9.56 0.22
8.05 0.97
0.67
0.53
a
b
c
d
Molecular Weight; number of H-bond donors; logarithmic acid dissociation constant of most basic group; calculated
e
f
logarithmic Octanol/Water repartition; calculated logarithmic distribution coefficient (pH = 7.4); Topological Polar
g
h
Surface Area; Blood Brain Barrier permeability score (0-6); Lipinski’s rule of five (RO5) violations (HBA and cLogS
data not shown).
2.4 Molecular modeling
From the obtained results reported in Table 1 we selected two representative compounds: 24, which
showed the best selectivity profile, and 28, with the highest affinity towards GlN2b, for the
molecular dynamic simulations.
We chose the pdb 4PE5 [41] containing the x-ray structure of the rat heterotetrameric NMDA
receptor, where the chains A and C correspond to two GluN1a units, while the chains B and D
correspond to two GluN2b units. The (R, S)-ifenprodil (4-[(1R,2S)-2-(4-benzylpiperidin-1-yl)-1-
hydroxypropyl]phenol) is bound at the interface between chains A/B and C/D. We chose chain A
and B as template to reconstruct one single binding site (Supplementary Figure S1a).
Our compounds are expected to bind to the (R, S)-ifenprodil binding site of GluN2B
(Supplementary Figure S1b). To test the docking method, we first minimised both target and ligand,
then we performed the docking. Autodock led to a first ranked conformation with RMSD 0.902 Å
1
1

ACCEPTED MANUSCRIPT
with respect to the crystal structure and a predicted binding affinity of -11.95 kcal/mol and an
estimated inhibition constant corresponding to 1.73 nM.
We then docked compounds 24 and 28 in the same site, and both wrapped around themselves in the
binding pocket (Supplementary Figure S2a,b and S3a,b). Compound 24 optimum pose was
predicted with binding affinity of -12.16 kcal/mol and an estimated inhibition constant
corresponding to 1.22 nM. Compound 28 was predicted with binding affinity corresponding to -
12.34 kcal/mol and an estimated inhibition constant equal to 894.62 pM. Both values were
comparable to the estimated free binding energy and the estimated inhibition constant towards (R,
S)-ifenprodil, corresponding to -11.95 kcal/mol and to 1.73 nM, respectively.
When both compounds were instead docked to σ1 (PDB 5HK2 [4], Supplementary Figure S4, S5
and S6), Autodock predicted binding affinities and corresponding estimated inhibition constants of
compound 24 (-10.37 kcal/mol and 25.24 nM) as well as those of compound 28 (-10.41 kcal/mol
and 23.34 nM) indicated much weaker interactions involved.
We then ran 160 ns of molecular dynamics (MD) simulation on the complexes in water solvent to
verify whether the complexes were stable upon simulated time. The ligand topologies were built
with ATB [50]. The topologies were validated as the molecular mechanics minimised structure of
compound 24 had root mean squared deviation, RMSD, of 0.01009 nm with respect to the semi-
empirical minimised structure, while, for compound 28, the same RMSD was 0.01111 nm.
Figure 2. MMPBSA analysis: energetic contributions to the GluN receptor binding of (a) (R, S)-ifenprodil, (b)
compound 24, and (c) compound 28 and to the σ1 receptor binding of (d) compound 24, and (e) compound 28. Data
averaged over the last 60 ns of the molecular dynamic trajectory.
1
2

ACCEPTED MANUSCRIPT
As the simulations were run only on a fragment of the receptor ectodomains, minor rearrangements
were expected. Indeed, in all cases the receptor backbone was only slightly rearranged (root mean
squared fluctuation, RMSF, not larger than 0.5 nm). While these rearrangements were minor, both
compounds explored novel poses (Supplementary Figure S7 and S8). The estimated scores and
binding energy values were expected to be not fully comparable with those experimentally
determined, we were thus surprised to find that the enthalpic contribution to the binding affinity of
both compounds to GluN, in terms of total binding energies (Figure 2), resulted still higher
compared to that associated with (R, S)-ifenprodil with an average of 35.7 kcal/mol, while
compound 24 was found with an average of 50.9 kcal/mol and 28 with an average of 42.7 kcal/mol.
The two appear also tightly bound to σ1: -42.1 kcal/mol for compound 24 and -48.1 kcal/mol for
compound 28.
More thoroughly, both compounds were kept bound to their GluN binding pocket through their
neighbouring amino acid and long distance effects were minor (Figure 3a,b). Both compounds
interacted with Phe-91 and Tyr-87 from chain A, and in both cases Arg-93 from the same chain
opposed the binding due to its strongly unfavourable polar desolvation energy (Figure 3c,d). While
in compound 24 also Asp-87 from chain B opposed the binding, Asp-110 from chain A stabilised
the molecule through hydrogen bonding with the compound oxygen atoms (Figure 3e,f and
Supplementary Figure S9). Along the whole simulation, both molecules remained at the interface
between the two chains, as expected (Figure 3e-h).
1
3

ACCEPTED MANUSCRIPT
Figure 3. MMPBSA details for GluN bound to compound 24 (left panels) and compound 28 (right panels): (a,b) amino
acids contribution to the total binding energy shaded of red/blue highlight chains A/B ; (c,d) details of the contributions
of each amino acid with binding energy larger than ± 1.5 kcal/mol; (e-h) close ups on compound 24 and 28 with (e,g)
highlighted amino acids with positive (shades of blue) and negative (shades of red) contribution to the binding energy
larger than ± 1.5 kcal/mol, hydrogen bonding groups are highlighted in green; (f-h) the same snapshots with GluN1A
(
chain A, red), GluN2B (chain B, blue).
Rearrangements are also observed for the complex between σ1 and compound 24 (Supplementary
Figure S8), but in this case the molecule tends to slip out of its binding site (Figure 4a). Phe-107
and Tyr-120 keep the molecule in place, while Glu-172 and especially Arg-119 strongly opposed to
the binding (Figure 4b). The same was not true for compound 28 (Figure 4c), kept in the pocket
thanks to Leu-105, Leu-182, and Phe-107 and a hydrogen bond with Thr-181 (Figure 4d, and
Supplementary Figure S10).
Figure 4. MMPBSA details for sigma1 bound to compound 24 (left panels) and compound 28 (right panels): (a,b) close
ups on compound 24 and 28 with highlighted amino acids with positive (shades of blue) and negative (shades of red)
contribution to the binding energy larger than ± 1.5 kcal/mol, hydrogen bonding groups are highlighted in green; (c,d)
details of the contributions of each amino acid with binding energy larger than ± 1.5 kcal/mol.
Overall, the docking results clearly showed that the binding affinity of the explored compounds
towards GluN follows the order (R, S)-ifenprodil < 24 < 28. The trend 24 < 28 was also true
towards σ1. Molecular dynamics simulations on the protein fragment corresponding to the binding
1
4

ACCEPTED MANUSCRIPT
domain further confirmed that the molecules were thermodynamically stable in the pinpointed
binding site, with the exception of compound 24 tending to slip out of the σ1 pocket.
2
2
.5. Antioxidant activity
.5.1 Preliminary in vitro antioxidant activity evaluation
We further evaluated the antioxidant activity of the most interesting compounds of both
phenyloxadiazolone and dihydroquinolinone series (20, 21, 24, 26-29, 31 and 33) by testing the
ability to scavenge ABTS (2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) derived radicals
and H O (hydrogen peroxide) oxidant . Natural antioxidant ascorbic acid and synthetic antioxidant
2
2
Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) were employed as reference
standard antioxidants. Most of the compounds potently inhibited ABTS radicals and H O ,
2
2
compared to the standards (Table 3).
Table 3. Antioxidant activity of compounds 20, 21, 24, 26-29, 31 and 33.
a
Cpd
IC (µg/mL)
50
ABTS
H O
2
2
2
2
2
2
2
2
2
3
3
0
1
4
6
7
8
9
1
3
17.96 ± 0.21
69.25 ± 0.36
20.95 ± 0.12
25.47 ± 0.24
78.80 ± 0.33
35.79 ± 0.31
21.34 ± 0.19
12.35 ± 0.10
26.24 ± 0.19
13.85 ± 0.19
22.15 ± 0.22
21.19 ± 0.52
73.07 ± 0.65
27.63 ± 0.39
31.44 ± 0.48
84.56 ± 0.55
38.29 ± 0.61
24.04 ± 0.45
15.37 ± 0.41
29.16 ± 0.28
17.11 ± 0.25
25.88 ± 0.37
Ascorbic Acid
Trolox
a
All measurements were performed in triplicate
Compound 31 exhibited a significant radical scavenging capacity on the ABTS with a value of
2.35 ± 0,10, lower than ascorbic acid and Trolox (12.35 vs 13.85 and 12.37 vs 22.15 respectively).
Compounds 20, 29 and 31, showed an important radical scavenging activity on the H O (15.37,
1
2
2
2
4.04 and 21.19). These IC values were lower than those of the compared standards.
50
1
5

ACCEPTED MANUSCRIPT
2
.5.2 Effects of ligands on H O -mediated toxicity
2 2
The same compounds tested in the aforementioned antioxidant assay were evaluated for
cytotoxicity in the human SH-SY5Y (neuroblastoma), HEP3B (hepatocarcinoma) and HeLa
(cervical adenocarcinoma) cell lines. As shown in Figure 5, on one hand, the viability of SH-SY5Y
cells increased when they were exposed to at low doses (10 µM) of compounds 20, 26, 27 and 29.
On the other hand, the lowest cell viability was observed following exposure of all cell lines to
compound 20 at 100 µM (Figure 5, Supplementary Table S1). The LD50 values for the selected
dual σR/GluN2b selective ligands were calculated from the dose response curves and are shown in
Table 4.
Figure 5. Cell viability of cells treated with the dual σR/GluN2b selective ligands. The cell viability was tested after
treatment with the titled compounds 20, 21, 24, 26-29, 31 and 33. The human neuroblastoma SH-SY5Y (●), human
hepatocarcina HEP3B (▲) and human cervical adenocarcinoma HeLa (■) cell lines were treated with increasing
concentrations of the dual σR/GluN2b selective ligands for 48h. The cell viability was determined by MTT. The bars
represent mean ± SEM of three experiments in triplicate.
Table 4. Cytotoxicity of the dual σR/GluN2b selective ligands to neuroblastoma, hepatocarcinoma and cervical
adenocarcinoma cell lines.
*
Cpd
LD50 (µM)
SH-SY5Y
Hep3B
HeLa
1
6

ACCEPTED MANUSCRIPT
2
2
2
2
2
2
2
3
0
1
4
6
7
8
9
1
3
53 ± 5
81 ± 7
77 ± 7
77 ± 8
60 ± 5
493 ± 45
63 ± 7
36 ± 5
231 ± 25
51 ± 6
71 ± 8
70 ± 5
70 ± 7
50 ± 6
81 ± 9
77 ± 8
67 ± 7
60 ± 7
55 ± 5
76 ± 7
69 ± 7
69 ± 8
55 ± 6
265 ± 35
128 ± 15
83 ± 9
76 ± 5
3
*
LD values were calculated from the linear regression of the dose-log response curves after 48 h exposure to the
5
0
compounds, determined by the MTT assay. Values are mean ± SEM of three experiments in triplicate.
Finally, neuroprotection tests were carried out in the human neuroblastoma SH-SY5Y cell line
previously used for studies on the sigma receptors [43, 51] and studies on neuroprotective agents
[
52, 53]. In order to test the neuroprotective potential of our ligands, H O was selected as an
2 2
appropriate harmful molecule, given that ROS are normally produced in neurons, and, moreover,
oxidative stress can be considered the major contributor to cell death in several neurodegenerative
disorders [54]. In addition, it has been shown that the loss of cell viability induced by H O in the
2
2
SH-SY5Y cell line is the result of both necrosis and apoptosis [55]. The incubation of SH-SY5Y
cells with the σR/GluN2b ligands at the concentration of 25 µM for 48h did not show significant
toxicity (Figure 6A). On the other hand, cells exposed to 1000 µM H O during 4 h showed a
2
2
significant of decreased (80%) cellular viability (Figure 6B, DMSO). This observation is consistent
with previous studies showing that H O induced a loss of SH-SY5Y cells viability [56]. All the
2
2
tested compounds exhibited a viability > 73% (values ± SEM in Supplementary Tables S2 and S3),
after the treatment with H O . Interestingly, compounds 21 and 28, both bearing the
2
2
cyclohexylpiperazine amine moiety, exhibited the best neuroprotective profile, whereas compounds
4 and 31, both weak σ1R inhibitors, exhibited a slight worsening of cytoprotection. These results,
2
highlight a synergistic effect targeting both σR and GluN2bR, for a neuroprotective action against
harmful agents.
1
7

ACCEPTED MANUSCRIPT
Figure 6. Protective effects of the dual σR/GluN2b selective ligands against H O induced neuronal injury. A) Viability
2
2
of the SH-SY5Y cells treated with 25 µM of the compounds 20, 21, 24, 26-29, 31 and 33; B) Neuroprotective effects of
the selected compounds after treatment of the SH-SY5Y cell line with hydrogen peroxide at 1000 µM. The bars
represent the mean ± SEM from three independent experiments performed in triplicate. The differences of all
compounds vs DMSO (H O ) were statistically significant, with p value < 0.05.
2
2
3. Conclusions
Considering the implication of both σR and GluN2bR in several neurodegenerative disorders, we
designed and synthesized two sets of dual receptor modulators endowed with potential
neuroprotective properties. Such compounds derived from the combination of different well-known
amines moieties with phenyloxadiazolone and dihydroquinolinone scaffolds, previously adopted by
us as selective σ1R ligands.
Within the phenyloxadiazolone series, two (20 and 26) out of seven compounds showed a pan-
affinity towards both σR subtypes but lacked preferential binding. On the other hand, compound 24
1
8

ACCEPTED MANUSCRIPT
displayed the best selectivity profile for GluN2bR. However, the best results were obtained with the
dihydroquinolinones series. Indeed, the cyclohexylpiperazine derivative 28 exhibited the best σR
pan-affinity (Ki σ1 = 21 nM; Ki σ2 = 1.5 nM) and proved to be 5-fold more powerful (Ki = 2.0 nM)
than the reference drug ifenprodil towards the GluN2b-containing NMDAR subunit. All the
compounds possess favourable in silico predicted ADME parameters to reach the BBB.
Importantly, most of the compounds were effective in protecting SH-SY5Y cells from H O -
2
2
mediated cell death, in particular the cyclohexylpiperazine derivatives 21 and 28. Therefore, this
study suggests that our novel dual σR/GluN2b modulators could be useful as neuroprotective agents
that would help to prevent cell death under vulnerability associated with strong ROS production.
From the results obtained, derivatives 24 and 28 could be considered the hit compounds to design
new selective and more potent σR/GluN2bR ligands.
4
4
. Materials and Methods
.1 Chemistry, general remarks
Commercially available chemicals were of reagents grade and used as received. Flash
chromatography was performed on Silica Gel 60 (70-230 mesh, Sigma-Aldrich-Merck); DCVC
(
dry-column-vacuum-chromatography) on Silica Gel 40 (230-400 mesh, Sigma-Aldrich-Merck).
Reaction courses and product mixtures were routinely monitored by thin-layer chromatography
TLC) on silica gel precoated F254 Merck plates. Melting points were determined with a Stuart SMP
00 apparatus and are uncorrected. An Agilent Cary-60 spectrophotometer UV-Vis was employed
(
3
to record the spectra and quantify the absorbance. Infrared spectra were recorded on a Jasco 4700
spectrophotometer in nujol mulls. Nuclear magnetic resonance spectra were determined on a Varian
1
13
4
00 MHz; H-NMR spectra were registered at 400 MHz whereas C-NMR were registered at 101
MHz. Chemical shifts are reported as δ (ppm) in CDCl solution with the solvent reference relative
3
1
to tetramethylsilane (TMS) employed as the internal standard (CDCl , δ = 7.26 ppm for H-NMR
3
1
3
and δ = 77.2 ppm for C-NMR); 1 drop of D O was added to assign NH protons. Coupling
2
constants (J) are reported in Hz and the splitting abbreviations used are: s, singlet; d, doublet; dd,
double doublet; t, triplet; td, triplet of doublets; q, quartet; quint, quintet; m, multiplet; br, broad.
Microanalyses (C, H, N) were carried out with Elementar Vario ELIII apparatus and were in
agreement with theoretical values ± 0.4%. ESI-MS spectra were obtained on a Bruker Daltonics
Esquire 4000 spectrometer by infusion of a solution of the sample in ultrapure MeOH.
1
9

ACCEPTED MANUSCRIPT
4.2 Synthetic procedures
4.2.1 Synthesis of 3-(4-bromobutyl)-5-phenyl-1,3,4-oxadiazol-2(3H)-one 11
This intermediate was synthesized as previously reported by us [43] starting from benzoic acid
hydrazide and triphosgene, to obtain 5-phenyl-1,3,4-oxadiazol-2(3H)-one which was converted into
3
-(4-bromobutyl)-5-phenyl-1,3,4-oxadiazol-2(3H)-one 11 with 1,4-dibromobutane (2.5 eq.), K CO3
2
(2.5 eq.) as base and a catalytic amount of KI, in ACN at reflux temperature.
-1
White needles solid (upon cooling overnight), yield: 53%; m.p. 64-66 °C; I.R. (nujol, cm ): 1763;
1
H-NMR: (CDCl /TMS) δ: 1.97 (m, 4H, N-CH CH CH CH Br), 3.46 (t, 2H, N
3
2
2
2
2
CH CH CH CH Br, J = 8 Hz), 3.83 (t, 2H, N-CH CH CH CH Br, J = 8 Hz), 7.48 (m, 3H, arom.
2
2
2
2
2
2
2
2
Ph), 7.83 (dd, 2H, arom. Ph, J = 8 Hz).
4.2.2 Synthesis of 1-(4-bromobutyl)-3,4-dihydrquinolin-2(1H)-one 12
The procedure for the synthesis of this intermediate is slightly different.
To 0.5 g (3.4 mmol) of the commercially available dihydroquinolin-2(1H)-one dissolved in 50 mL
of ACN, 1.66 g (5.1 mmol) of CsCO , 1.1 g (5.1 mmol) of 1.4-dibromobutane and a catalytic
3
amount of TBAB were added. The mixture was allowed to stir at reflux temperature for 14 hours
and monitored by TLC (DCM/EtOH 95:5). The solvent was evaporated in vacuo and the residue
was taken up with DCM and washed with distilled water (3x50). The collected organic phase was
dried, filtered and evaporated under reduce pressure to afford a pale-yellow oil.
-1
1
Yield: 58%; I.R. (nujol, cm ): 1770; H-NMR (CDCl -TMS) δ: 1.77-1.84 (m, 2H,
3
CH CH CH CH Br), 1.88-1.95 (m, 2H, CH CH CH CH Br), 2.63 (t, 2H, CH CH CO dihydroq. J
2
2
2
2
2
2
2
2
2
2
=
8 Hz), 2.88 (t, 2H, 2H, CH CH CO dihydroq. J = 8 Hz), 3.43 (m, 2H, 2H, CH CH CH CH Br),
2 2 2 2 2 2
3
.96 (t, 2H, 2H, CH CH CH CH Br, J = 8 Hz), 7.00 (m, 2H, arom.), 7.16-7.24 (m, 2H, arom.).
2 2 2 2
4.2.3 Amines 13-19
All amines were purchased from Sigma-Aldrich (Merk) except 7-nitro-1,2,3,4-
tetrahydroisoquinoline 18 and 1-(2,4-dimethylphenyl)-N-methylmethanamine 19, the synthesis of
which is summarized below.
On an ice bath (0°C), a 100-mL round bottom flask with 1.3 g (10 mmol) of 1,2,3,4-
tetrahydroquinoline and 5 mL of conc. H SO was allowed to stir for 10 minutes. To this solution,
2
4
1
.1 g (10 mmol) of KNO were added in small portions, taking care that the temperature of the
3
reaction did not rise above 5°C. The reaction was stirred overnight and monitored by TLC
2
0

ACCEPTED MANUSCRIPT
(
DCM/EtOH 90:10). The brown solution was neutralized with a solution of diluted NH OH until
4
pH = 8 and the basic mixture was extracted with DCM (3x50 mL). The combined organic extracts
were washed with brine (once), dried and filtered. The evaporation of the solvent affords a red oil
which was dissolved in the minimum amount of abs. EtOH and cooled on an ice bath. The alcoholic
solution was treated with 2.5 mL of conc. HCl to affords a yellow precipitate of 7-nitro-1,2,3,4-
tetrahydroisoquinoline 18, which was recrystallized from MeOH.
-1
1
Yellow solid; m.p.: 261-263 (260-262°C [57]); Yield: 32%; I.R. (nujol, cm ): 3173; H-NMR
CDCl -TMS) δ: 1.86 (br s, 1H, NH disapp. on D O), 3.12 (t, 2H, H-4, J = 8 Hz), 3.46 (t, 2H, H-3,
(
3
2
J = 8 Hz), 4.37 (s, 2H, H-1), 7.36 (d, 1H, arom. H-5, J = 12 Hz), 8.00 (m, 2H, arom. H-6 and H-8).
The synthesis of 19 started from 2,4-dimethylbenzaldheyde (5 g, 37.3 mmol) which was dissolved
in 20 mL of abs. EtOH and then added of 2.32 g of methylamine solution (33% in EtOH, 74.6
mmol). The mixture was allowed to heated at 40°C for 1 hours (monitored by TLC) then cooled,
the solvent and the excess of methylamine were eliminated under reduced pressure. The residual
solid of (Z/E)-N-(2,4-dimethylbenzylidene)methanamine (5.45 g, 37.0 mmol), was treated with a
slight excess of NaBH (2.1 g, 55.6 mmol) in abs. EtOH at 0°C; the mixture was stirred at room
4
temperature overnight. The resulting mixture was evaporated in vacuo, poured into distilled water
and extracted with DCM (3x150 mL). The collected organic phase was dried, filtered and finally
evaporated to affords 5.4 g of 19 as a light-yellow semisolid (upon cooling).
-1
1
Yield: 98%; I.R. (nujol, cm ): 3166; H-NMR (CDCl -TMS) δ: 1.52 (br, 1H, NH disapp. on D O),
3
2
2
2
.31 (s, 3H, CH arom.), 2.33 (s, 3H, CH arom.), 2.50 (s, 3H, NCH ), 3.71 (s, 2H, CH ), 7.00 (m,
3
3
3
2
H, arom.), 7.17 (d, 1H, arom. J = 8 Hz).
4.2.4 General synthesis of the final compounds 20-33
To a solution of 0.2 g (0.67 mmol) of 11, 0.19 g (1.35 mmol) of K CO , a catalytic amount of KI
2
3
and 3H-spiro[isobenzofuran-1,4'-piperidine] hydrochloride 13 (0.15 g, 0.67 mmol) in ACN (30 mL)
was heated at reflux temperature for 24 h and monitored by TLC until the reaction was completed.
The hot solution was filtered and concentrated in vacuo to give 0.3 g of 20.
4.2.4.1 3-(4-(3H-spiro[isobenzofuran-1,4'-piperidin]-1'-yl)butyl)-5-phenyl-1,3,4-oxadiazol-2(3H)-
one 20
-1
1
Light-red oil; Yield: 92%; I.R. (nujol, cm ): 1775; H-NMR (CDCl -TMS) δ: 1.58-2.00 (m, 8H,
3
CH CH CH CH and 2xCH pip.), 2.38 (td, 2H, pip. J = 2.4 and 12.0 Hz), 2.46 (t, 2H,
2
2
2
2
2
2
1

ACCEPTED MANUSCRIPT
CH CH CH CH , J = 7.6 Hz), 2.84 (br d, 2H, pip. J = 10.4 Hz), 3.83 (t, 2H, CH CH CH CH , J =
2
2
2
2
2
2
2
2
7
.2 Hz), 5.05 (s, 2H, OCH fur.), 7.10-7.27 (m, 4H, arom. isobenzofuran), 7.47 (m, 3H, arom.
2
1
3
phenyl), 7.83 (m, 2H, arom. phenyl). C-NMR (CDCl -TMS) δ: 24.0, 26.3, 36.6, 45.9, 50.2, 58.2,
3
70.7, 84.7, 120.8, 121.0, 123.9, 125.6, 127.3, 127.5, 128.9, 131.5, 138.9, 145.65, 153.2, 153.6. MS:
+
m/z 406 [MH ]. Anal. calcd for C H N O (%): C, 71.09; H, 6.71; N, 10.36. Found: C, 71.11; H,
2
4
27
3
3
6.74; N, 10.18.
Following the same procedure described above (only 1 equivalent of K CO when amines are as
2
3
free-bases), compounds 21-33 were synthesized.
4.2.4.2 3-(4-(4-Cyclohexylpiperazin-1-yl)butyl)-5-phenyl-1,3,4-oxadiazol-2(3H)-one 21.
-1
1
Pale-yellow Oil; Yield: 94%; I.R. (nujol, cm ): 1778; H-NMR (CDCl -TMS) δ: 1.03-126 (m, 5H,
3
(
CHH) ) cyclohex.), 1.56-1.86 (m, 9H, CH CH CH CH and (CHH) ) cyclohex.), 2.18 (m, 1H, CH
5 2 2 2 2 5
cyclohex.), 2.35 (t, 2H, CH CH CH CH , J =7.6 Hz), 2.45 (br m, 4H, pip.), 2.56 (br m, 4H pip.),
2
2
2
2
3
.78 (t, 2H, CH CH CH CH J = 7.2 Hz), 7.45 (m, 3H, arom. phenyl), 7.80 (m, 2H, arom. phenyl).
2 2 2 2
1
3
C-NMR (CDCl -TMS) δ: 23.7, 25.9, 26.2, 26.3, 28.9, 45.8, 48.9, 53.7, 57.9, 63.4, 123.9, 125.6,
3
+
1
28.9, 131.4, 153.1, 153.6. MS: m/z 385 [MH ]. Anal. calcd for C H N O (%): C, 68.72; H, 8.39;
2
2
32
4
2
N, 14.57. Found: C, 68.76; H, 8.40; N, 14.58.
4.2.4.3 3-(4-(4-(4-Fluorophenyl)piperazin-1-yl)butyl)-5-phenyl-1,3,4-oxadiazol-2(3H)-one 22.
-1
Pale-brown oil (purified after trituration with petroleum ether); Yield: 69%; I.R. (nujol, cm ): 1782;
1
H-NMR (CDCl -TMS) δ: 1.63 (quint, 2H, CH CH CH CH , J = 7.2 Hz), 1.87 (quint, 2H,
3
2
2
2
2
CH CH CH CH , J = 7.2 Hz), 2.44 (t, 2H, CH CH CH CH , J = 7.6 Hz), 2.59 (t, 4H, pip. J = 4.8
2
2
2
2
2
2
2
2
Hz), 3.10 (t, 4H, pip. J = 4.8 Hz), 3.83 (t, 2H, CH CH CH CH , J = 7.6 Hz), 6.86 (m, 2H, arom. p-
2
2
2
2
F-phenyl), 6.94 (m, 2H, arom. p-F-phenyl), 7.47 (m, 3H, arom. phenyl), 7.83 (m, 2H, arom.
1
3
phenyl). C-NMR (CDCl -TMS) δ: 23.7, 26.2, 45.8, 50.1, 53.2, 57.7, 115.3 (2), 115.6 (2), 117.7,
3
117.8, 117.9, 123.9, 125.6, 128.9, 131.5, 147.9, 148.0, 153.2, 153.6, 155.9, 158.3. MS: m/z 397
+
[
MH ]. Anal. calcd for C H FN O (%): C, 66.65; H, 6.36; N, 14.13. Found: C, 66.69; H, 6.34; N,
2
2
25
4
2
14.11.
4.2.4.4 5-Phenyl-3-(4-(4-(pyridin-2-yl)piperazin-1-yl)butyl)-1,3,4-oxadiazol-2(3H)-one 23.
-1
1
Yellow Oil; Yield: 96%; I.R. (nujol, cm ): 1776; H-NMR (CDCl -TMS) δ: 1.59 (quint, 2H,
3
CH CH CH CH , J = 7.2 Hz), 1.85 (quint, 2H, CH CH CH CH , J = 7.2 Hz), 2.41 (td, 2H,
2
2
2
2
2
2
2
2
CH CH CH CH , J = 1.2 and 7.2 Hz), 2.51 (t, 4H, pip. J = 4.8 Hz), 3.51 (t, 4H, pip. J = 4.8 Hz),
2
2
2
2
2
2

ACCEPTED MANUSCRIPT
3
.81 (t, 2H, CH CH CH CH , J = 1.2 and 7.2 Hz), 6.59 (m, 2H, arom. pyr.), 7.80 (d, 2H, arom.
2
2
2
2
1
3
phenyl, J = 8.0 Hz), 8.15 (m, 2H, arom. phenyl). C-NMR (CDCl -TMS) δ: 23.7, 26.2, 45.2, 45.8,
3
53.0, 57.8,107.0, 113.2, 123.9, 125.6, 128.9, 131.5, 137.4, 147.9, 153.1, 153.6, 159.5. MS: m/z 380
+
[
MH ]. Anal. calcd for C H N O (%): C, 66.47; H, 6.64; N, 18.46. Found: C, 66.50; H, 6.44; N,
2
1
25
5
2
18.55.
4.2.4.5
3-(4-(6,7-Dimethoxy-3,4-dihydroisoquinolin-2(1H)-yl)butyl)-5-phenyl-1,3,4-oxadiazol-
2(3H)-one 24.
-
Brown oil (purified by dry-flash chromatography with DCM as eluant); Yield: 57%; I.R. (nujol, cm
1
1
)
: 1767; H-NMR (CDCl -TMS) δ: 1.66 (quint, 2H, CH CH CH CH , J = 7.2 Hz), 1.89 (quint, 2H,
3 2 2 2 2
CH CH CH CH , J = 7.2 Hz), 2.54 (t, 2H, CH CH CH CH , J = 7.2 Hz), 2.68 (t, 2H, H-4,
2
2
2
2
2
2
2
2
dihydroisoquin. J = 5.6 Hz), 2.79 (t, 2H, H-3, dihydroisoquin. J = 5.6 Hz), 3.52 (s, 2H, H-9,
dihydroisoquin.), 3.81 (m, 8H, 2x OCH and CH CH CH CH ), 6.49 (s, 1H, H-5, arom.
3
2
2
2
2
dihydroisoquin.), 6.56 (s, 1H, H-8, arom. dihydroisoquin.), 7.45 (m, 3H, arom. phenyl), 7.81 (m,
1
3
2
H, arom. phenyl). C-NMR (CDCl -TMS) δ: 24.0, 26.2, 28.7, 45.8, 51.0, 55.7, 55.9, 57.4, 109.1,
3
111.3, 123.9, 125.6 (2), 126.2, 126.5 (2), 128.9 (2), 131.4, 131.5, 147.1, 147.4, 153.1, 153.6. MS:
+
m/z 410 [MH ]. Anal. calcd for C H N O (%): C, 67.46; H, 6.65; N, 10.26. Found: C, 67.50; H,
2
3
27
3
4
6.67; N, 10.29.
4.2.4.6 3-(4-(7-Nitro-3,4-dihydroisoquinolin-2(1H)-yl)butyl)-5-phenyl-1,3,4-oxadiazol-2(3H)-one
25.
-1
1
Brown oil; Yield: 99%; I.R. (nujol, cm ): 1777; H-NMR (CDCl -TMS) δ: 1.68 (quint, 2H,
3
CH CH CH CH , J = 7.2 Hz), 1.91 (quint, 2H, CH CH CH CH , J = 7.2 Hz), 2.59 (t, 2H,
2
2
2
2
2
2
2
2
CH CH CH CH , J = 7.2 Hz), 2.75 (t, 2H, H-4, dihydroisoquin. J = 5.6 Hz), 2.97 (t, 2H, H-3,
2
2
2
2
dihydroisoquin. J = 5.6 Hz), 3.67 (s, 2H, H-9, dihydroisoquin.), 3.85 (t, 2H, CH CH CH CH , J =
2
2
2
2
7.6 Hz), 7.22 (d, 1H, H-5, arom. dihydroisoquin. J = 8.4 Hz), 7.46 (m, 3H, H-8, arom. phenyl), 7.81
(m, 2H, arom. phenyl), 7.88 (d, 1H, H-8 arom. dihydroisoquin. J = 2.4 Hz), 7.95 (dd, 1H, arom.
13
phenyl, J = 2.4 and 8.4 Hz). C-NMR (CDCl -TMS) δ: 23.8, 26.0, 29.4, 45.7, 50.1, 55.7, 57.2,
3
+
1
21.1, 121.7, 123.8, 125.6, 128.9, 129.5, 131.5, 136.4, 142.4, 153.2, 153.6. MS: m/z 395 [MH ].
Anal. calcd for C H N O (%): C, 63.95; H, 5.62; N, 14.20. Found: C, 63.97; H, 5.64; N, 14.24.
2
1
22
4
4
4.2.4.7 3-(4-((2,4-Dimethylbenzyl)(methyl)amino)butyl)-5-phenyl-1,3,4-oxadiazol-2(3H)-one 26.
2
3

ACCEPTED MANUSCRIPT
-1
1
Pale-yellow oil; Yield: 98%; I.R. (nujol, cm ): 1760; H-NMR (CDCl -TMS) δ: 1.58 (m, 2H,
3
CH CH CH CH ), 1.84 (m, 2H, CH CH CH CH ), 2.15 (s, 3H, NCH ), 2.29 (s, 3H, CH arom.),
2
2
2
2
2
2
2
2
3
3
2
.32 (s, 3H, CH arom.), 2.42 (t, 2H, CH CH CH CH , J = 8 Hz), 3.39 (s, 2H, CH -Ph), 3.78 (t, 2H,
3 2 2 2 2 2
CH CH CH CH , J = 8 Hz), 6.95 (m, 2H, arom.), 7.13 (d, 1H, arom. J = 8 Hz), 7.46 (m, 3H, arom),
2
2
2
2
1
3
7
.84 (dd, 2H, arom. phenyl, J = 4 and 8 Hz). C-NMR (CDCl -TMS) δ: 19.1, 21.0, 24.3, 26.0,
3
41.9, 45.9, 56.8, 60.4, 76.7, 77.4, 124.0, 125.6 (2), 126.1, 128.9, 129.0, 129.8, 131.0, 131.4, 134.2,
+
1
36.4, 137.1, 153.1, 153.6. MS: m/z 366 [MH ]. Anal. calcd for C H N O (%): C, 72.30; H, 7.45;
2
2
27
3
2
N, 11.50. Found: C, 72.33; H, 7.45; N, 11.47.
4.2.4.8
1-(4-(3H-spiro[isobenzofuran-1,4'-piperidin]-1'-yl)butyl)-3,4-dihydroquinolin-2(1H)-one
27.
-1
1
Pale-brown oil; Yield: 95%; I.R. (nujol, cm ): 1687; H-NMR (CDCl -TMS) δ: 1.60-1.85 (m, 6H,
3
CH CH CH CH and CH pip.), 2.00 (m, 2H, CH pip.), 2.38 (br td, 2H, pip. J = 4 and 12 Hz),
2
2
2
2
2
2
2
.46 (t, 2H, CH CH CH CH J = 7.2 Hz), 2.61 (m, 2H, CH (CHH ) dihydroquin.), 2.86 (m, 3H,
2 2 2 2 2 3,3’
(
CHH ) dihydroquin. and pip.), 3.15 (br td, 1H, pip. J = 2.4 and 8.8 Hz), 3.96 (t, 2H,
4,4’
1
3
CH CH CH CH , J = 7.2 Hz), 5.05 (s, 2H, OCH fur.), 6.95-7.27 (m, 8H, arom.). C-NMR
2
2
2
2
2
(
CDCl -TMS) δ: 24.2, 25.1, 25.6, 31.9, 36.6, 41.9, 50.1, 58.2, 70.7, 84.7, 114.9, 120.8, 121.0,
3
+
1
22.7, 126.6, 127.3, 127.4, 127.5, 128.0, 138.9, 145.6, 170.1. MS: m/z 391 [MH ]. Anal. calcd for
C H N O (%): C, 76.89; H, 7.74; N, 7.17. Found: C, 76.93; H, 7.55; N, 7.47.
25
30
2
2
4.2.4.9 1-(4-(4-Cyclohexylpiperazin-1-yl)butyl)-3,4-dihydroquinolin-2(1H)-one 28.
Reddish oil (purified by column chromatography using DCM/EtOH 95-5 then DCM/EtOH 90-10 as
-1
1
eluant); Yield: 56%; I.R. (nujol, cm ): 1670; H-NMR (CDCl -TMS) δ: 1.19 (m, 5H, (CHH)
3
5
cyclohex.), 1.55-1.83 (m, 6H, CH CH CH CH and (CHH) cyclohex.), 2.24 (m, 1H, N-CH
2
2
2
2
5
cyclohex.), 2.35 (t, 2H, CH CH CH CH , J = 7.2 Hz), 2.48 (br m, 4H, pip.), 2.59 (m, 6H, pip.,
2
2
2
2
cyclohex. and (CHH ) dihydroquin.), 2.85 (t, 2H, (CHH ) dihydroquin. J = 7.2 and 8.0 Hz),
3
,3’
4,4’
3
7
=
.91 (t, 2H, CH CH CH CH J = 7.2 and 8.0 Hz), 6.96 (td, 1H, H-6 arom. dihydroquin. J = 1.2 and
2 2 2 2
.2 Hz), 7.02 (d, 2H, H-5 arom. dihydroquin. J = 8.0 Hz), 7.12 (br d, 2H, H-8 arom. dihydroquin. J
13
7.6 Hz), 7.20 (td, 2H, H-7 arom. dihydroquin. J = 1.6 and 8.4 Hz). C-NMR (CDCl -TMS) δ:
3
2
1
1
3.9, 24.9, 25.6, 25.8, 26.2, 28.8, 31.9, 41.8, 48.8, 53.3, 53.4, 57.8, 63.6, 114.9, 122.6, 126.5, 127.3,
+
27.9, 139.5, 170.0. MS: m/z 370 [MH ]. Anal. calcd for C H N O (%): C, 74.75; H, 9.55; N,
2
3
35
3
1.37. Found: C, 74.73; H, 9.55; N, 11.36.
2
4

ACCEPTED MANUSCRIPT
4.2.4.10 1-(4-(4-(4-fluorophenyl)piperazin-1-yl)butyl)-3,4-dihydroquinolin-2(1H)-one 29.
Pale-yellow oil (purified column chromatography using DCM 100 then DCM/EtOH 94-6 as
-1
1
eluant); Yield: 35%; I.R. (nujol, cm ): 1675; H-NMR (CDCl -TMS) δ: 1.61 (quint, 2H,
3
CH CH CH CH , J = 7.2), 1.71 (m, 2H, CH CH CH CH J = 7.2), 2.44 (t, 2H, CH CH CH CH , J
2
2
2
2
2
2
2
2
2
2
2
2
=
7.6 Hz), 2.60 (t, 4H, pip. J = 5.2 Hz), 2.64 (m, 2H, (CHH ) dihydroquin.), 2.88 (td, 2H, pip. J =
3,3’
6.8 Hz), 3.12 (t, 2H, (CHH ) dihydroquin. J = 5.2 Hz), 3.97 (t, 2H, CH CH CH CH , J = 7.2 Hz),
4,4’ 2 2 2 2
6.88 (m, 2H, arom. p-F-phenyl), 6.94 (m, 2H, arom. p-F-phenyl), 6.99 (m, 1H, H-6 arom.
dihydroquin.), 7.06 (dd, 1H, H-5 arom. dihydroquin., J = 1.2 and 7.6 Hz), 7.16 (dd, 1H, H-8 arom.
13
dihydroquin., J = 1.2 and 7.6 Hz), 7.23 (td, 1H, H-7 arom. dihydroquin., J = 1.6 and 7.6 Hz). C-
NMR (CDCl -TMS) δ: 24.0, 24.9, 25.6, 31.9, 41.8, 50.1, 57.8, 114.9, 115.3, 117.7, 117.8, 122.7,
3
+
1
27.4, 127.8, 139.5, 148.0 (2), 155.9, 158.3, 170.1. MS: m/z 382 [MH ]. Anal. calcd for
C H FN O (%): C, 72.41; H, 7.40; N, 11.01. Found: C, 72.43; H, 7.40; N, 11.03.
23
28
3
4.2.4.11 1-(4-(4-(pyridine-2-yl)piperazin-1-yl)butyl)-3,4-dihydroquinolin-2(1H)-one 30.
Pale-yellow oil (purified by column chromatography using DCM 100 then DCM/EtOH 95-5 as
-1
1
eluant); Light-yellow Oil; Yield: 36%; I.R. (nujol, cm ): 1685; H-NMR (CDCl -TMS) δ: 1.58-
3
1
.75 (m, 4H, CH CH CH CH ), 2.43 (t, 2H, CH CH CH CH , J = 7.2 Hz), 2.51 (t, 4H, pip. J = 4.8
2 2 2 2 2 2 2 2
Hz), 2.63 (m, 2H, (CHH ) dihydroquin.), 2.89 (td, 2H, (CHH ) dihydroquin. J = 6.8 Hz), 3.54
3
,3’
4,4’
(t, 4H, pip. J = 5.2 Hz), 3.97 (t, 2H, CH CH CH CH , J = 7.2 Hz), 6.63 (m, 2H, arom. pyr.), 6.99
2 2 2 2
(
td, 1H, arom. J = 1.2 and 7.2 Hz), 7.06 (d, 1H, arom. J = 8.4 Hz), 7.16 (dd, 1H, arom. J = 1.6 and
.6 Hz), 7.23 (td, 1H, arom. pyr. J = 1.6 and 7.6 Hz), 7.47 (m, 1H, arom. pyr.), 8.18 (m, 1H, arom.
7
1
3
pyr.). C-NMR (CDCl -TMS) δ: 23.9, 24.9, 25.5, 31.9, 41.8, 45.2, 53.0, 57.9, 107.0, 113.2, 114.9,
3
+
1
22.6, 126.5, 127.4, 128.0, 137.4, 139.5, 147.9, 159.5, 170.1. MS: m/z 365 [MH ]. Anal. calcd for
C H N O (%): C, 72.50; H, 7.74; N, 15.37. Found: C, 72.53; H, 7.54; N, 15.37.
22
28
4
4.2.4.12 1-(4-(6,7-dimethoxy-3,4-dihydroisoquinolin-2(1H)-yl)butyl)-3,4-dihydroquinolin-2(1H)-
one 31.
Reddish oil (purified by dry-flash chromatography with DCM then DCM/EtOH 94-6 as eluant);
-1
1
Yield: 65%; I.R. (nujol, cm ): 1660; H-NMR (CDCl -TMS) δ: 1.68 (quint, 2H, CH CH CH CH ,
3
2
2
2
2
J = 7.2), 1.78 (quint, 2H, CH CH CH CH , J = 7.2), 2.56 (m, 2H, CH CH CH CH ), 2.81 (m, 4H,
2
2
2
2
2
2
2
2
(
CHH ) dihydroisoquin. and (CHH ) dihydroquin.), 2.94 (m, 4H, (CHH ) dihydroisoquin.
4,4’ 3,3’ 3,3’
and (CHH ) dihydroquin.), 3.77 (ds, 6H, 2x OCH ), 3.81 (s, 2H, (CHH ) dihydroisoquin.), 3.91
4
,4’
3
9,9’
(
t, 2H, CH CH CH CH , J = 7.2 Hz), 6.48 (s, 1H, H-5, arom. dihydroisoquin.), 6.54 (s, 1H, H-8,
2 2 2 2
2
5

ACCEPTED MANUSCRIPT
arom. dihydroisoquin.), 6.91 (td, 1H, H-6, arom. dihydroquin. J = 1.2 and 7.2 Hz), 6.97 (td, 1H, H-
13
5
, arom. dihydroquin. J = 1.2 and 7.2 Hz), 7.09 (t, 1H, H-5, arom. dihydroquin. J = 8 Hz). C-
NMR (CDCl -TMS) δ: 22.0, 23.7, 24.5, 25.5, 30.9, 40.3, 49.1, 53.2, 54.9, 55.0, 55.1, 108.4, 110.2,
3
+
1
13.9, 121.8, 123.5, 125.5, 126.5, 127.0, 138.2, 146.7, 147.2, 169.3. MS: m/z 395 [MH ]. Anal.
calcd for C H N O (%): C, 73.07; H, 7.10; N, 15.37. Found: C, 73.11; H, 7.14; N, 15.39.
2
4
30
2
3
4.2.4.13 1-(4-(7-nitro-3,4-dihydroisoquinolin-2(1H)-yl)butyl)-3,4-dihydroquinolin-2(1H)-one 32.
Brownish oil (purified by dry-flash chromatography with DCM then DCM/EtOH 98-2 as eluant);
-1
1
Yield: 30%; I.R. (nujol, cm ): 1674; H-NMR (CDCl -TMS) δ: 1.70 (m,4H, CH CH CH CH ),
3
2
2
2
2
2
.60 (m, 4H, CH CH CH CH and (CHH ) dihydroisoquin.), 2.75 (t, 2H, (CHH ) dihydroquin.
2 2 2 2 4,4’ 3,3’
J = 6.0 Hz), 2.87 (m, 2H, (CHH ) dihydroquin), 2.97 (t, 2H, (CHH ) dihydroisoquin. J = 6.0
4,4’
3,3’
Hz), 3.67 (s, 2H, H-9, dihydroisoquin.), 3.98 (t, 2H, CH CH CH CH , J = 7.2 Hz), 6.97 (td, 1H, H-
2
2
2
2
6, arom. dihydroquin. J = 1.2 and 7.2 Hz), 7.04 (d, 1H, H-5, arom. dihydroisoquin. J = 7.6 Hz), 7.13
(m, 2H, H-5,8, arom. dihydroquin.), 7.23 (d, 2H, H-7, arom. dihydroquin. J = 7.6 Hz), 7.90 (d, 1H,
1
3
H-8 dihydroisoq. J = 2.4 Hz), 7.96 (dd, 1H, H-6 dihydroisoq. J = 2.4 and 7.6 Hz). C-NMR
(
CDCl -TMS) δ: 24.1, 24.8, 25.6, 29.4, 31.9, 41.7, 50.0, 55.7, 57.2, 114.8, 121.1, 121.8, 122.7,
3
+
1
26.6, 127.3, 128.0, 129.5, 136.4, 139.4, 142.5, 146.0, 170.1. MS: m/z 380 [MH ]. Anal. calcd for
C H N O (%): C, 69.64; H, 6.64; N, 11.07. Found: C, 69.63; H, 6.62; N, 11.07.
22
25
3
3
4.2.4.14 1-(4-((2,4-dimethylbenzyl)(methyl)amino)butyl)-3,4-dihydroquinolin-2(1H)-one 33.
-1
1
Pale-red oil; Yield: 96%; I.R. (nujol, cm ): 1689; H-NMR (CDCl -TMS) δ: 1.58 (quint, 2H,
3
CH CH CH CH , J = 7.2 Hz), 1.67 (quint, 2H, CH CH CH CH , J = 7.2 Hz), 2.14 (s, 3H, NCH ),
2
2
2
2
2
2
2
2
3
2
.29 (s, 3H, CH arom.), 2.30 (s, 3H, CH arom.), 2.41 (t, 2H, CH CH CH CH , J = 7.2 Hz), 2.63
3 3 2 2 2 2
(
m, (CHH ) dihydroquin.), 2.88 (m, (CHH ) dihydroquin.), 3.38 (s, 2H, CH -Ph), 3.93 (t, 2H,
3,3’ 4,4’ 2
CH CH CH CH , J = 7.6 Hz), 6.97 (m, 4H, arom.), 7.14 (m, 1H, arom.), 7.23 (td, 1H, arom. J = 2.4
2
2
2
2
1
3
and 7.6 Hz). C-NMR (CDCl -TMS) δ: 19.2, 21.0, 24.8, 25.1, 25.6, 31.9, 42.0, 42.1, 57.4, 60.1,
3
114.9 (2), 122.6, 126.1, 126.6, 127.4, 127.9, 128.0, 129.8, 131.0, 134.2, 136.3, 137.1, 139.6, 170.1.
+
MS: m/z 351 [MH ]. Anal. calcd for C H N O (%): C, 78.82; H, 8.63; N, 7.99. Found: C, 78.85;
2
3
30
2
H, 8.66; N, 7.87.
4
4
.3 Computational studies
.3.1 Docking
2
6

ACCEPTED MANUSCRIPT
We chose for GluN chains A and B from structure 4PE5 [41] as template, while for σ1 we chose
chain C from structure 5HK2 [4]. We reconstructed the missing atoms and residues with Swiss-
Model [58]. Molecules initial conformation were minimized with AM1 method as implemented in
MOPAC [59]. Each system to be docked was then prepared with AutoDock tools, and docked with
AutoDock [60]. We used Lamarckian Genetic Algorithm with docking box centered on the ligand
and 70x40x30 grid points for GlunN, and docking box centered on Tyr 103 hydroxylic oxygen and
40x42x40 grid points for σ1, in both cases with spacing 0.375Å. The docking was performed with
10 runs and 25,000,000 maximum numbers of evaluations and standard parameters. 2D ligand-
protein interaction diagrams were generated with LigPlot+[61].
4.3.2 Molecular Dynamics Simulations
We consider complexes formed by the ligands with the first 370 amino acids of both chain A and
chain B for GluN, and with residues between 40 and 210 for σ1. We minimized each complex by
first minimizing the protein side chains alone, then whole protein and finally the whole system by
constraining selected portions of the system. We placed each complex in a cubic box with a water
layer of 0.7 nm and performed a second minimization. We used GROMOS force field and Simple
Point Charge water. Ligand topologies were built with ATB [50]. We performed NVP and NPT
equilibrations for 100 ps, followed by 160 ns NPT production run at 300 K. The iteration time step
was set to 2 fs with the Verlet integrator and LINCS [62] constraint. We used periodic boundary
conditions. All the simulations and their analysis were run as implemented in the Gromacs package
v. 2016.1 [63]. RMSDs and RMSF have been calculated from configurations sampled every 10ps
and as running averages over 100 sampled points. VINA scorings were calculated over
configurations sampled every 100 ps and as running averages over 10 points. 2D ligand-protein
interaction diagrams were generated with LigPlot+ [61]. The binding free energy was estimated
with the MM/PBSA method, with the apolar solvation energy calculated as solvent accessible
surface area (SASA) and default parameters, as implemented in the g_mmpbsa tool [64].
Simulations were run on Marconi (CINECA, Italy).
4.4 Physiochemical and Pharmacokinetics parameters
The physiochemical and pharmacokinetics parameter of compounds 20-33 was calculated using
Marvinsketch® (Chemaxon) software and the BBB scores were predicted with algorithm of central
nervous system multiparameter optimization (CNS MPO) developed by Wager et al. [48].
2
7

ACCEPTED MANUSCRIPT
4
4
.5 Binding studies
.5.1 Materials
Guinea pig brains, rat brains and rat livers were commercially available (Harlan-Winkelmann,
Borchen, Germany). Pig brains were a donation of the local slaughterhouse (Coesfeld, Germany).
The recombinant L(tk-) cells stably expressing the GluN2B receptor were obtained from Prof. Dr.
Dieter Steinhilber (Frankfurt, Germany). Homogenizers: Elvehjem Potter (B. Braun Biotech
®
International, Melsungen, Germany) and Soniprep 150 (MSE, London, UK). Centrifuges: Cooling
centrifuge model Eppendorf 5427R (Eppendorf, Hamburg, Germany) and High-speed cooling
®
centrifuge model Sorvall RC-5C plus (Thermo Fisher Scientific, Langenselbold, Germany).
Multiplates: standard 96 well multiplates (Diagonal, Muenster, Germany). Shaker: self-made device
with adjustable temperature and tumbling speed (scientific workshop of the institute). Harvester:
®
®
MicroBeta FilterMate 96 Harvester. Filter: Printed Filtermat Typ A and B. Scintillator: Meltilex
®
(
Typ A or B) solid state scintillator. Scintillation analyzer: MicroBeta Trilux (all Perkin Elmer
LAS, Rodgau-Jügesheim, Germany).
4
.5.2 Preparation of membrane homogenates from pig brain cortex
Fresh pig brain cortex was homogenized with the potter (500-800 rpm, 10 up and down strokes) in
volumes of cold 0.32 M sucrose. The suspension was centrifuged at 1,200 x g for 10 min at 4 °C.
6
The supernatant was separated and centrifuged at 31,000 x g for 20 min at 4 °C. The pellet was
resuspended in 5-6 volumes of TRIS/EDTA buffer (5 mM TRIS/1 mM EDTA, pH 7.5) and
centrifuged again at 31,000 x g (20 min, 4 °C). The final pellet was resuspended in 5-6 volumes of
buffer and frozen (-80 °C) in 1.5 mL portions containing about 0.8 mg protein/mL.
4.5.3 Preparation of membrane homogenates from rat liver
Two rat livers were cut into small pieces and homogenized with the potter (500-800 rpm, 10 up and
down strokes) in 6 volumes of cold 0.32 M sucrose. The suspension was centrifuged at 1,200 x g
for 10 min at 4 °C. The supernatant was separated and centrifuged at 31,000 x g for 20 min at 4°C.
The pellet was resuspended in 5-6 volumes of buffer (50 mM TRIS, pH 8.0) and incubated at rt for
30 min. After the incubation, the suspension was centrifuged again at 31,000 x g for 20 min at 4 °C.
The final pellet was resuspended in 5-6 volumes of buffer and stored at -80 °C in 1.5 mL portions
containing about 2 mg protein/mL.
4.5.4 Cell culture and preparation of membrane homogenates for the GluN2B [47].
2
8