New (arylalkyl)azole derivatives showing anticonvulsant effects could have VGSC and/or GABAAR affinity according to molecular modeling studies

Article abstract of DOI:10.1016/j.ejmech.2016.08.032

(Arylalkyl)azoles (AAAs) emerged as a novel class of antiepileptic agents with the invention of nafimidone and denzimol. Several AAA derivatives with potent anticonvulsant activities have been reported so far, however neurotoxicity was usually an issue. We prepared a set of ester derivatives of 1-(2-naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone oxime and evaluated their anticonvulsant and neurotoxic effects in mice. Most of our compounds were protective against maximal electroshock (MES)- and/or subcutaneous metrazol (s.c. MET)-induced seizures whereas none of them showed neurotoxicity. Nafimidone and denzimol have an activity profile similar to that of phenytoin or carbamazepine, both of which are known to inhibit voltage-gated sodium channels (VGSCs) as well as to enhance γ-aminobutiric acid (GABA)-mediated response. In order to get insights into the effects of our compounds on VGSCs and A-type GABA receptors (GABA_ARs) we performed docking studies using homology model of Na⁺channel inner pore and GABA_AR as docking scaffolds. We found that our compounds bind VGSCs in similar ways as phenytoin, carbamazepine, and lamotrigine. They showed strong affinity to benzodiazepine (BZD) binding site and their binding interactions were mainly complied with the experimental data and the reported BZD binding model.

Full text of DOI:10.1016/j.ejmech.2016.08.032

European Journal of Medicinal Chemistry 124 (2016) 407e416
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
New (arylalkyl)azole derivatives showing anticonvulsant effects could
have VGSC and/or GABAAR affinity according to molecular modeling
studies
,
a
Suat Sari a, Arzu Karakurt b *, Harun Uslu b, F. Betül Kaynak c, Ünsal Çalıs¸ ,
Sevim Dalkara a
a Hacettepe University, Faculty of Pharmacy, Department of Pharmaceutical Chemistry, 06100, Ankara, Turkey
Inonü University, Faculty of Pharmacy, Department of Pharmaceutical Chemistry, 44280, Malatya, Turkey
c Hacettepe University, Faculty of Engineering, Department of Physics Engineering, 06532, Ankara, Turkey
b
_
€
a r t i c l e i n f o
a b s t r a c t
Article history:
(Arylalkyl)azoles (AAAs) emerged as a novel class of antiepileptic agents with the invention of nafimi-
done and denzimol. Several AAA derivatives with potent anticonvulsant activities have been reported so
far, however neurotoxicity was usually an issue. We prepared a set of ester derivatives of 1-(2-naphthyl)-
2-(1H-1,2,4-triazol-1-yl)ethanone oxime and evaluated their anticonvulsant and neurotoxic effects in
mice. Most of our compounds were protective against maximal electroshock (MES)- and/or subcu-
taneous metrazol (s.c. MET)-induced seizures whereas none of them showed neurotoxicity. Nafimidone
and denzimol have an activity profile similar to that of phenytoin or carbamazepine, both of which are
Received 23 May 2016
Received in revised form
11 July 2016
Accepted 14 August 2016
Available online 25 August 2016
Keywords:
(Arylalkyl)azole
known to inhibit voltage-gated sodium channels (VGSCs) as well as to enhance
g-aminobutiric acid
(GABA)-mediated response. In order to get insights into the effects of our compounds on VGSCs and A-
type GABA receptors (GABAARs) we performed docking studies using homology model of Naþ channel
inner pore and GABAAR as docking scaffolds. We found that our compounds bind VGSCs in similar ways
as phenytoin, carbamazepine, and lamotrigine. They showed strong affinity to benzodiazepine (BZD)
binding site and their binding interactions were mainly complied with the experimental data and the
reported BZD binding model.
Anticonvulsant
Voltage gated sodium channel
A-type GABA receptor
Molecular docking
X-ray crystallography
Microwave chemistry
© 2016 Elsevier Masson SAS. All rights reserved.
1. Introduction
long term toxicity. Furthermore, currently available antiepileptic
drugs (AEDs) fail to achieve adequate seizure control in one third of
Epilepsy is a chronic neurological disorder, which disrupts social
and labour life, inflicting 1 in every 26 people [1]. The toll epilepsy
takes is even heavier in developing countries [2]. Epilepsy usually
starts in early childhood and lasts for a lifetime. Without a radical
cure available, patients are obliged to take medication for years
thus exposed to severe side effects, drug-drug interactions, and
the patients [3]. That is why the need for new AEDs that are safe,
selective, and more effective is urgent.
(Arylalkyl)azoles (AAAs) emerged as a novel class of antiepi-
leptic agents with nafimidone and denzimol (Fig. 1) [4,5]. They
possess an activity profile similar to that of phenytoin and carba-
mazepine, which act through voltage-gated sodium channels
(VGSCs) inhibition although these drugs were reported to enhance
g-aminobutiric acid (GABA)-mediated response, as well [6,7].
Structure-activity relationship (SAR) studies on AAAs have revealed
that an aromatic ring, an azole (usually imidazole and 1,2,4-
triazole) group, and an alkyl chain of two-carbon long connecting
these two are necessary for high anticonvulsant activity. These
studies point out the essence of a small oxygen functional group
(such as carbonyl, ethylene dioxy, methoxy, acyloxy, hydroxy, and
oxime ether moieties) attached to the alkyl bridge for a better
anticonvulsant activity, as well [8,9].
Abbreviations: AAA, (arylalkyl)azole; MES, maximal electroshock; s. c. MET,
subcutaneous metrazol; VGSC, voltage-gated sodium channel; GABA,
g-amino-
butiric acid; GABAAR, A-type GABA receptor; BZD, benzodiazepine; AED, antiepi-
leptic drug; SAR, structure-activity relationship; MW, microwave; MM, molecular
mass; MP, melting point; DCC, N,N'-dicyclohexylcarbodiimide; DMAP, 4-
dimethylaminopyridine; LA, local anaesthetic.
* Corresponding author.
E-mail addresses: arzu.karakurt@inonu.edu.tr, arzukarakurt@hotmail.com
(A. Karakurt).
http://dx.doi.org/10.1016/j.ejmech.2016.08.032
0223-5234/© 2016 Elsevier Masson SAS. All rights reserved.

408
S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
Table 1
MW irradiation conditions of 3 and 4 in comparison with conventional reaction
conditions.
Comp.
MW irradiation method
Conventional method
Time (min.)
Yield (%)
Time (min.)
Yield (%)
3
4
16
8
98
85
360
180
59
78.
Fig. 1. Chemical structure of nafimidone and denzimol.
this way structural modifications were applied on “R” moiety
(Table 2). They were purified by column chromatography and/or
crystallization. 5a, 5b, 5f-k, and 5m were converted to their HCl
salts with gaseous HCl (gHCl). Their structures were confirmed by
IR, 1H NMR, mass spectra, and elemental analysis.
Previously, Karakurt et al. prepared several oxime ether de-
rivatives of 1-(2-naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone [10],
carboxylic ester derivatives of nafimidone alcohol [11], and oxime
ester derivatives of nafimidone [12], some of which proved highly
protective against convulsions in test animals.
4 was established by X-ray crystallographic analysis and proved
to be in Z configuration (Fig. 2A). It was crystallized under a non-
centrosymmetric orthorhombic structure with space group Pc21n
and its unit cell dimensions were found to be: a ¼ 5.1237(3) Å,
b ¼ 10.9258(8) Å and c ¼ 21.8692(17) Å, Z ¼ 4. The naphthyl moiety
is almost planar and makes an angle 71.09ꢀ with the triazole ring.
The packing of the molecules and the hydrogen bonding in the
structure are shown in Fig. 2B. The X-ray structure contains strong
O/H/N inter-molecular H-bond interaction [O1/H1A 2.6897 Å,
O1-H1A 0.82 Å and O1/H1A-N3 168ꢀ] by Platon program [20]. O/
H/N interaction contributes to the formation of chains in zig-zag
configuration along b direction [symmetry code: ꢁ1/2þx, ꢁ1/2þy,
1/2ꢁz]. There is also a strong C12/H12 … Cg2 (1þx, y, z) interaction
which stabilizes the crystal packing (H … Cg2: of 2.90 Å).
In the light of these facts, we prepared 14 novel 1-(2-naphthyl)-
2-(1H-1,2,4-triazol-1-yl)ethanone oxime ester derivatives through
modifications on the oxime ester function and evaluated their
anticonvulsant and neurotoxic effects in mice using maximal
electroshock (MES), subcutaneous metrazol (s.c. MET), and rotarod
tests. Microwave (MW) irradiation was applied during synthesis of
the synthons, by which the amount of time required was greatly
reduced and the yields were increased [13].
In order to test whether our compounds could act through
VGSCs and/or A-type GABA receptors (GABAARs), which account for
the principal mechanisms of action for many AEDs [9], and un-
derstand their possible interactions with these targets we per-
formed docking studies on an open (activated) Naþ channel inner
pore [14] and a native GABAAR model [15].
Previously, certain oxime ethers derivatives were reported to be
in the same configuration with their oxime precursors [22].
Accordingly, we expected 5a-n to be in Z configuration however in
the 1H NMR spectra of the most of our compounds we observed
signals of both diastereomers in varying intensities. These signals
were detectable for the protons of the methylene next to the tri-
azole ring (eCH2eN), the triazole protons, and the protons belong
to the oxime ester group. For instance, the eCH2eN protons of 5f
next to the triazole ring were observed as two singlets, an upfield
peak (5.81 ppm) of 31% intensity and a downfield peak (5.90 ppm)
of 69%. These percentages were very close for each proton or set of
protons of the other affected groups on the molecule. The chemical
shifts and intensities of the methylene protons of each diaste-
reomer for compounds that show mixture of diastereomers in their
1H NMR spectra are given in Table 2. One reason for the occurrence
of a mixture of diastereomers in 1H NMR spectra of these com-
pounds could be interconversion of diastereomers due to heating
during NMR run [23]. Previously E and Z isomers of 1-(2-naphthyl)-
2. Results and discussion
2.1. Chemistry
As outlined in Scheme 1, 1-bromo-2-(2-naphthyl)ethanone (2)
was prepared by the reported method [16]. Sodium salt of 1,2,4-1H-
triazole was alkylated with 2 to yield 1-(2-naphthyl)-2-(1H-1,2,4-
triazol-1-yl)ethanone (3) [17]. 3 was then converted to its oxime
derivative (4) using NH2OH.HCl in strong basic medium (pH: 14)
[18]. 3 and 4 were effected via MW irradiation [13], which saved
reasonable time and increased yield regarding conventional
methods (Table 1).
Final compounds (5a-n) were synthesized by the Steglich
esterification [19] of 4 with proper carboxylic acids or anhydrides in
the presence of N,N'-dicyclohexylcarbodiimide (DCC) and 4-
dimethylaminopyridine (DMAP) in dichloromethane (DCM); by
Scheme 1. (1.5 column). Reagents and conditions: (i) Br2, CH3COOH, 0e5 ꢀC; (ii) 1H-1,2,4-triazole, Naꢀ, CH3OH, MW 70 ꢀC, 500 W; (iii) NH2OH.HCl, C2H5OH, pH 14, MW 70 ꢀC,
500 W; (iv) DCC, DMAP, DCM, gHCl.

S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
409
Table 2
Structures, molecular masses, yields, melting points, and clogP values of 5a-n.
Comp.
R
X
MMa (g/mol) Yield (%) MPb (ꢀC) clogP
5a
HCl 308.34
39
138e40 2.89
5b
5c
HCl 322.37
44
134e6
131e2
3.13
3.71
e
336.39
42
5d
5e
e
e
322.37
336.39
39
31
119e21 3.45
109e10 3.64
Fig. 2. The ORTEP 3 [21] (A) and packing diagram (B) of 4. Displacement ellipsoids are
drawn at the 50% probability level. Hydrogen atoms are shown as spheres of arbitrary
radius.
5f
HCl 336.39
29
129e30 3.66
Table 3
5g
5h
HCl 336.39
HCl 350.42
49
79
130e2
127e9
3.95
4.20
Chemical shifts of eCH2eN protons of each diastereomer and their relative in-
tensities for compounds 1H NMR spectra of which show diastereomer mixtures.
Comp.
E
Z
Chemical shift (ppm) Intensity % Chemical shift (ppm) Intensity %
5i
5j
HCl 350.42
HCl 350.42
71
72
122e5
3.91
5a
5b
5d
5f
5g
5h
5i
5.61
5.56
5.50
5.81
5.78
5.80
5.91
5.83
5.90
37
49
5
31
30
35
35
33
44
5.70
5.67
5.65
5.90
5.86
5.88
6.00
5.92
6.01
63
51
95
69
70
65
65
67
56
127e30 3.93
5j
5k
5k
5l
HCl 378.48
82
53
135e8
145e7
5.26
4.30
among 23 algorithms based on principal component and hierar-
chical cluster analysis results to predict drug-likeness [25]. The tool
uses six molecular descriptors (logP, polar surface area, hydrogen
donor count, aliphatic ring count, aromatic ring count, and Balaban
Index) as input, which were previously reported to be the best
among 34 features [26]. 12 compounds were found drug-like ac-
cording to the algorithms provided by MLViS (Table 4). 5m and 5n
were found nondrug-like in two kernel-based methods (lsSVMrbf,
SVMrbf), kNN, an instance-based learning algorithm, and neural
networks (NN) algorithm.
e
362.43
5m
5n
HCl 370.41
35
26
126e9
3.81
4.36
e
382.42
162
a
Molecular mass.
Melting point.
b
2.2. Pharmacology
2-(imidazol-1-yl)ethanone-O-ethyl oxime were separately isolated
and in their 1H NMR spectra, the eCH2-N protons of the Z isomer of
this compound were observed to resonate at higher frequency than
those of the E isomer [24]. Therefore we speculate that the down-
field signals in Table 3, which occur at higher intensities, belong to
the Z isomers.
10 Compounds among 5a-n showed protection in either or both
of the tests at one of the given doses and periods at least. More
compounds were protective against MES-induced seizures than s.c.
MET-induced seizures, which usually is the case for AAAs [10,11,24].
Those active against MES-induced seizures were protective mostly
up to 0.5 h. However those s.c. MET-active were protective up to 4 h
but at 300 mg/kg except for 5f, which was the only derivative to
show protection at 30 mg/kg 5m and 5n, which were inactive,
We tested drug-likeness of our compounds using MLViS, a web
tool which makes use of 10 machine-learning methods selected

410
S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
Table 4
Drug-likeness of 5a-n according to MLViS web tool.
Comp.
FDAa
C5
J48
lsSVMrbfb
SVMrbfc
SVMlind
RFe
Bag SVMf
kNNg
NNh
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
5m
5n
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Nondrug-like
Nondrug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Nondrug-like
Nondrug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Nondrug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Drug-like
Nondrug-like
Nondrug-like
a
Flexible Discriminant Analysis.
b
c
Least Squares Support Vector Machines with Radial Basis Function.
Support Vector Machines with Radial Basis Function.
Support Vector Machines with Linear Kernel Function.
Random Forests.
Bagged Support Vector Machines with Radial Basis Function.
k-Nearest Neighbours.
d
e
f
g
h
Neural Networks.
Table 5
Anticonvulsant and neurotoxicity screening results of 5a-n and reference AEDs (The data indicating observed activity/toxicity is highlighted as bold.).
Screen
Time (h)
Dose (mg/kg)
5a
5b
5c
5d
5e
5f
5g
5h
5i
5j
5k
5l
5m
5n
Phe.a
Carb.b
MESc
0.5
30
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
0/1
0/1
0/1
0/1
1/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
0/1
0/1
0/1
0/1
1/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
1/1
1/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
1/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
0/1
1/1
0/1
1/1
0/1
0/1
0/1
1/1
1/1
1/1
1/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
0/1
1/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
0/1
1/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
1/1
0/1
0/1
0/1
1/1
0/1
0/1
0/1
0/1
0/1
1/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
1/1
1/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
1/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
0/1
0/1
0/1
0/1
1/1
0/1
0/1
0/1
0/1
0/1
1/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
1/1
1/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/2
0/2
0/2
0/2
0/1
0/1
e
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/1
0/2
0/2
0/2
1/1
1/1
1/1
1/1
1/1
1/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
2/2
0/2
2/2
2/2
1/1
1/1
1/1
0/1
1/1
1/1
0/1
0/1
1/1
0/1
1/1
1/1
0/2
0/2
2/2
0/2
0/2
0/2
100
300
30
100
300
30
100
300
30
100
300
30
100
300
30
100
300
4
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/1
0/2
0/2
0/2
0/2
0/2
0/2
s.c. METd
0.5
4
TOXe
0.5
4
a
Phenytoin.
Carbamazepine.
b
c
d
e
Maximal electroshock (number of mice protected/number of mice tested).
Subcutaneous metrazol (number of mice protected/number of mice tested).
Neurotoxicity (number of mice failed rotarod test/number of mice tested).
proved nondrug-like in some of the algorithms of MLViS web tool.
These were the only derivatives with an aryl group on the oxime
ester function. Taken together, it is not possible to establish SARs for
these compounds regarding the modifications on the oxime ester
group. On the other hand, relatively high clogP values might be one
of the factors behind the lack of anticonvulsant efficacy at lower
doses because high lipophilicity might cause issues such as meta-
bolism, solubility, deposition in adipose tissues, and plasma protein
binding. We did not observe any sign of neurotoxicity with any of
the compounds unlike previously reported AAAs, which usually
exhibit both protection and neurotoxicity at higher doses
[10e12,24,27]. Anticonvulsant and neurotoxicity screening results
of 5a-n are presented in Table 5.
through cell membrane in order to generate action potentials in
neurons and muscle cells. They have a large pore forming alpha
subunit which consists of four repeated domains (DI to DIV). Each
domain has six transmembrane helices, labelled S1 to S6, and are
assembled around the inner pore. VGSCs exist at three different
states (closed, open, and inactivated). With increasing membrane
potential the activation gate opens and the pore moves from closed
(deactivated) to open (activated) state to conduct Naþ, but at the
highest membrane potential the inactivation gate closes and the
channel becomes inactivated [28].
A member of the cys-loop receptor superfamily, GABAARs are
responsible from the inhibitory postsynaptic potentials in central
nervous system. When activated, a GABAAR conducts Clꢁ selectively
through a central pore formed by its five subunits. Each subunit
consists of an extracellular, transmembrane, and intracellular
domain [29]. VGSCs and GABAARs are primary targets for many
first-line AEDs [9,30]. Previously Lipkind and Fozzard [14] modelled
2.3. Molecular docking studies
VGSCs are transmembrane protein complexes that conduct Naþ

S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
411
an open Naþ channel inner pore composed of S5 and S6 residues.
Through molecular docking they investigated binding modes and
affinities of some local anaesthetics (LAs) and three AEDs that
inhibit VGSCs (phenytoin, carbamazepine, and lamotrigine) which
show higher affinity to the open/inactivated states of neuronal Naþ
channels [14,31]. They found that Phe-1764 and Tyr-1771 of DIV-S6
and Leu-1465 of DIII-S6 were key residues for binding affinities of
these AEDs, in accordance with previous biological studies [32e34].
Bergman et al. [15] reported a unified model of GABAAR of a1b2g2
subunit composition and elucidated agonist and positive allosteric
modulator binding through docking studies in the light of site
directed mutagenesis data. We used the phenytoin bound VGSC
inner pore model and the GABA and diazepam-bound GABAAR
model to perform docking studies of our four active derivatives, 5e,
ligand binding limits suggestions on the effects of “R” groups of our
compounds on possible VGSC inhibition. On the other hand it is
plausible to propose that too large an “R” group would render the
compound unable to infiltrate into the inner pore. 5f and other
derivatives do not have heteroatom-bound hydrogen thus lacked
so-called aromatic-H bond observed between the nitrogen-bound
hydrogens of the three AEDs and Phe-1764 phenyl ring, which
was suggested to be a key feature for voltage dependent VGSC in-
hibition of these drugs [14]. It was also argued that the alkylamine
heads of aminoamide type LAs like lidocaine, since protonated in
physiological pH, created a positive electrostatic barrier for Naþ
passage [31]. This was not the case in our study because our de-
rivatives are non-ionized in physiological pH. Thus we suggest that
5f could sterically occlude the inner pore blocking Naþ influx with
its bulky naphthalene and isobutyl groups (Fig. 4), in a similar way
the three AEDs were suggested to do [14].
€
5f, 5j, 5l, and one inactive, 5a, on Glide, version 6.9, Schrodinger,
LLC, New York, NY, 2015 [35e37].
Docking of 5e, 5j, and 5l yielded similar poses and interactions
to those of 5f (Fig. 5), except that the binding mode of 5l differed
from the others in that it did not have H-bond with Thr-1464, but
had a weak H-bond between one of its aromatic hydrogens on the
naphthalene ring and DIV-S6 Val-1767 carbonyl oxygen and its
triazole ring did not interact with Phe-1468.5e has a chiral centre
thus we took both enantiomers into consideration only to find they
did not exhibit significant difference in binding interactions and
2.3.1. VGSC docking
We docked our compounds into the wide mouth located be-
tween the selectivity filter and the activation gate, which is avail-
able only in the channel's open conformation, where some LAs, and
the three AEDs, phenytoin, carbamazepine, and lamotrigine, were
suggested to bind [14,31,33].
The binding mode of 5f showed striking similarities to those
previously obtained for the three AEDs, and its interactions with
the active site residues were in accordance with the experimental
data [14,31e34]. The naphthyl ring of 5f engaged in strong aromatic
p-p stacking and hydrophobic interactions with Phe-1764 and Tyr-
1771 side chains, while the methylene group between the oxime
and triazole had hydrophobic contacts with Leu-1465 side chain of
DIII-S6 (Fig. 3).
The naphthyl ring occupied the same space as the phenyl of
phenytoin in the proximity of DIV-S6 residues while the bulky
isobutyl moiety faced the pore just like the other phenyl of
phenytoin, suggested to block Naþ conductance in the study of
Lipkind and Fozzard [14]. The triazole lied towards cytosol making
aromatic
p-p interactions with the side chain of DIII-S6 Phe-1468
and the carbonyl oxygen was observed to make H-bond with the
hydroxyl side chain of Thr-1464 of DIII-S6.
Positioned in the centre of channel lumen, the isobutyl moiety
of 5f had weak hydrophobic interactions with the side chains of DII-
S6 residues, however these interactions, along with the strong
hydrophobic and electrostatic interactions with DIV- and DIII-S6
residues, might have stabilizing effect on the molecule in the cen-
tral cavity. Lack of evidence on the importance DII-S6 residues for
Fig. 4. Docking conformation of 5f in the active site of Naþ channel inner pore model
(top view). Molecular surface of ligand and protein are rendered.
Fig. 3. Docking conformation of 5f in the active site of Naþ channel inner pore (top
view). 5f is in yellow and nearby residues are in grey sticks, protein backbone is in
Fig. 5. Superimposition of the docking conformations of 5e (turquoise), 5f (yellow), 5j
(orange), and 5l (green) in the active site of Naþ channel inner pore (top view). Ligands
are in colour, nearby residues are in grey sticks, protein backbone in colour ribbons.
(For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.)
colour ribbons, H-bond and
p-p interactions are shown as dashed lines. (For inter-
pretation of the references to colour in this figure legend, the reader is referred to the
web version of this article.)

412
S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
Table 6
Docking scores of 5e, 5f, 5j, and 5l.
Comp.
Docking score
Ligand efficiency
H-bond score
Internal energy
VDW energy
VGSC
Coulomb energy
VGSC
GABAAR
VGSC
GABAAR
VGSC
GABAAR
VGSC
GABAAR
GABAAR
VGSC
GABAAR
5e
5f
5j
5l
ꢁ4.66
ꢁ4.71
ꢁ4.24
ꢁ3.42
ꢁ6.37
ꢁ6.70
ꢁ6.69
ꢁ6.32
ꢁ0.19
ꢁ0.19
ꢁ0.18
ꢁ0.13
ꢁ0.26
ꢁ0.27
ꢁ0.26
ꢁ0.23
ꢁ0.70
ꢁ0.70
ꢁ0.70
ꢁ0.20
ꢁ0.70
ꢁ0.57
ꢁ0.55
0.00
6.98
2.41
9.29
1.31
1.61
1.10
3.05
6.97
ꢁ30.36
ꢁ29.78
ꢁ30.99
ꢁ29.88
ꢁ24.05
ꢁ29.20
ꢁ29.86
ꢁ29.39
ꢁ3.33
ꢁ3.82
ꢁ3.67
ꢁ0.17
ꢁ3.72
ꢁ4.62
ꢁ4.86
ꢁ0.91
affinity. The inactive derivative 5a, assuming a flipped conforma-
tion, failed to -stack with Phe-1468 aromatic side chain as the
triazole ring positioned towards inner pore instead (see Figure S.1A
of Supplementary Data).
Docking scores and other scoring terms are summarized in
Table 6. The docking scores of the four compounds were pretty
close to each other. 5e and 5j had relatively high internal energy,
indicative of high strain on their docked conformations in Naþ
channel inner pore. As 5l did not have H-bond with Thr-1464 and
with a1His-101 and g2Tyr-58. This ring also formed a hydrogen
bond via its N2 with a1Ser-204. Among these residues a1His-101
and g2Phe-77 are necessary for high-affinity binding of ligands to
this site, according to the site directed mutagenesis studies
[39e44]. The naphthalene ring of 5f situated in the hydrophobic
cleft surrounded by the aromatic side chains of g2Phe-77, a1Phe-99,
and a1Tyr-159, exactly where the phenyl ring of diazepam (L3) lied
p
in the model, making
p-p stacking with them. This ring also
interacted with g2Asn-128 and g2Arg-144. a1Phe-99 and a1Tyr-159
were previously reported to be involved in BZD-like ligand binding
and part of the binding site [45,46]. Positioned in the gap between
two subunits and close to loop C, the isobutanoyl moiety of 5f was
in close contacts with a1Thr-206, g2Leu-140, and g2Thr-142. This
gap is not deep enough for larger moieties like those of 5m and 5n
to fit and this may account for their lack of activity. Mutations on
a1Thr-206 reportedly decreased binding affinity of ligands to this
site significantly [44,47e50]. Diazepam, instead, reportedly
engaged in two hydrogen bonds with a1Thr-206 and g2Thr-142 as
to build a bridge between two subunits [15]. The interactions of N-
methyl group of diazepam was not satisfied by our compounds.
Almost the same interactions were observed for 5e, 5j, and 5l
p-p stacking with Phe-1468 either, its H-bond score and Coulomb
energy was higher.
According to the per-residue interactions scores DIII- and DIV-
S6 residues were responsible for most of the interactions with 5f.
Phe-1764, Thr-1771, and Leu-1465 were among those residues with
the lowest interaction scores. Apart from these residues Thr-1464
and Phe-1468 of DIII-S6 and Val-1768 and Val-1767 of DIV-S6
were important contributors of ligand-receptor complex, howev-
er the corresponding residues of Val-1768 and Thr-1464 in NaV1.4,
namely Val-1583 and Thr-1279, respectively, were showed by
alanine substitution to have no effect on binding affinities of LAs
[31].
except that 5l did not form H-bond with a1Ser-204 and
p-p
stacking with g2Phe-77 either (Fig. 7). Their docking scores were
quite close to each other and indicated strong binding to the BZD
binding site (Table 5). 5a, however, failed to interact Phe-99, which
could account for its inefficacy (see Fig. S.1B of Supplementary
Data). The internal energy of 5l was relatively high and its elec-
trostatic interactions were low, showing that the conformation this
compound assumed to fit in the narrow gap between a1 and g2
subunits caused high strain. According to the per-residue in-
teractions scores of 5f g2Phe-77, a1Ser-204, g2Tyr-58, g2Thr-142,
a1His-101, and g2Asn-128 were the highest-contributing residues
to GABAAR binding. The docking poses mainly complied with the
experimental and theoretical data regarding the interactions of
GABAAR positive allosteric modulators with the BZD binding site
residues. Therefore, our compounds could activate GABAARs in a
similar way the BZD-like compounds do and this activation could
2.3.2. GABAAR docking
We docked our compounds into the putative GABA and benzo-
diazepine (BZD) binding sites between extracellular domains (ECD)
of b2 and a1 and a1 and g2 subunits, respectively. These sites include
loop A-F, where most of the residues that affect ligand binding and
constitute part of binding sites are present [38]. No docking pose
was obtained for the GABA binding site, thus we suggest that our
compounds are too bulky for this site and unlikely to be GABAAR
agonists.
On the other hand the compounds fit well into the BZD binding
site (Fig. 6). The triazole ring of 5f and the ethylene chain attached
to it corresponded to the fused ring system (so called L1 pharma-
cophore) of diazepam present in the model. The triazole ring
engaged in vertical p-p stacking with g2Phe-77 and close contacts
Fig. 6. Docking conformation of 5f in the BZD-binding site of GABAAR. 5f is in yellow
and nearby residues are in grey sticks, protein backbone is in colour ribbons, H-bond
Fig. 7. Superimposition of docking conformations of 5e (turquoise), 5f (yellow), 5j
(orange), and 5l (green) in the BZD-binding site of GABAAR. Ligands are in colour sticks,
protein backbone in colour ribbons. (For interpretation of the references to colour in
this figure legend, the reader is referred to the web version of this article.)
and
p-p interactions are shown as dashed lines. (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article.)

S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
413
partly or wholly account for their efficacy.
300 MHz Ultrashield™ (Germany) NMR spectrometer and mass
spectra with Micromass ZQ (USA) using electro spray ionization
(ESIþ) method and MassLynx 4.1 software. All chemical shifts are
3. Conclusion
reported as
d (ppm) values. Splitting patterns are designated as
Aiming to find safe, selective, and more efficacious novel AEDs,
we prepared 14 new oxime ester derivatives and screened them
in vivo for their anticonvulsant and neurotoxic effects. Among them
10 compounds showed protection against MES- and/or s.c. MET-
induced seizures without causing neurotoxicity. One of them (5f)
was active at 30 mg/kg in s.c. MET test at 4 h.1-(2-Naphthyl)-2-(1H-
1,2,4-triazol-1-yl)ethanone oxime (4) was previously found inac-
tive in the same animal models of seizure [24] which makes it is
less likely that our active compounds act as prodrugs. Two of the
inactive derivatives, 5m and 5n, were predicted nondrug-like ac-
cording to some of the algorithms calculated by MLViS web tool,
which shows the importance of dug-likeness prediction for anti-
convulsant activity. However it was not possible to establish SARs
for these compounds regarding anticonvulsant activity and
neurotoxicity. Relatively high clogP values, indicative of high lip-
ophilicity, could be one of the reasons for the absence of anticon-
vulsant activity at lower doses.
follows: s: singlet; d: doublet; t: triplet; q: quartet; and m: multi-
plet. Elemental analyses were performed using an LECO 932 CHNS
elemental analysis apparatus (USA).
4.1.1. Synthesis of the compounds
4.1.1.1. Preparation of 2-bromo-1-(2-naphthyl)ethanone (2), 1-(2-
naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone (3), and 1-(2-
naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone
oxime
(4).
2-Bromo-1-(2-naphthyl)ethanone (2), 1-(2-naphthyl)-2-(1H-1,2,4-
triazol-1-yl)ethanone (3), and 1-(2-naphthyl)-2-(1H-1,2,4-triazol-
1-yl)ethanone oxime (4) were prepared according to the methods
elucidated in the relative literature [16e18]. MW irradiation was
applied during synthesis of 3 and 4 instead of refluxing. The reac-
tion media were kept at room temperature for 3 min then heated
up to 70 ꢀC and irradiated for 8 min under maximum power of
500 W. Completion of the reactions was determined via thin-layer
chromatography.
VGSC inhibition and positive allosteric modulation of GABAAR
are possible anticonvulsant mechanisms of action for our active
compounds therefore we performed molecular docking studies of
5e, 5f, 5j, and 5l using homology models of Naþ channel inner pore
and GABAAR. The docking solutions obtained in VGSC active site
were in accordance with those previously found for phenytoin,
carbamazepine, and lamotrigine and with the experimental find-
ings thereof [14,31e34]. Phe-1764, Tyr-1771, and Leu-1465 were
among the residues with the highest affinity according to the per-
residue interaction scores from Glide. On the other hand docking
poses obtained in the BZD binding site of GABAAR were somewhat
similar to the diazepam binding model [15] and mostly in accor-
dance with the mutagenesis data [39e50]. However, docking re-
sults of 5a, one of the inactive derivatives, showed that this
compound failed certain key interactions with both receptors. We
suggest that our compounds might inhibit VGSCs by sterically
occluding the channel inner pore and are stabilized mainly by the
key hydrophobic and electrostatic interactions with DIV- and DIII-
S6 residues in a similar manner suggested for phenytoin, carba-
mazepine, and lamotrigine [14]. They might as well modulate
GABAARs by binding with high affinity to the BZD binding site but
not to the orthosteric binding site. We found that these interactions
together were only possible for the Z isomers, which supports our
hypothesis considering the X-ray crystallographic data obtained for
4. Docking studies also revealed that increasing the size of the “R”
group would probably hinder binding of our compounds to the
putative binding sites, which could underlie the inefficacy of 5m
and 5n.
4.1.1.2. Oxime esters (5a-n). 5a-c, 5e-n: A mixture of DCC (2 mmol)
and DMAP (0.17 mmol) in dry DCM was added dropwise to a
mixture of 1-(2-naphtyl)-2-(1H-1,2,4-triazol-1-yl)ethanone oxime
(1 mmol) and proper carboxylic acid (2 mmol) in dry DCM at
0e5 ꢀC. The resulting mixture was stirred for 0.5 h at 0e5 ꢀC and for
an additional 3e6 h at room temperature. The precipitate was
filtered off, the filtrate was evaporated to dryness, and the resulting
residue was purified via column chromatography. HCl salts of the
compounds were obtained by addition of ethereal solution of HCl to
their ethereal solution and crystallized from proper solvents. 5c, 5e,
5l, and 5n were not converted to their HCl salts.
5d: The mixture of 1-(2-naphtyl)-2-(1H-1,2,4-triazol-1-yl)
ethanone (1 mmol), DMAP (0.17 mmol), and butyric anhydride
(2 mmol) in dry DCM was stirred at room temperature for 6 h. The
precipitate which occurred with addition of diethyl ether was
filtered off and the filtrate was evaporated till dryness. The result-
ing residue was purified via column chromatography.
4.1.1.2.1. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone
propanoyloxime hydrochloride (5a). IR (
cmꢁ1) 3142, 3009 (aro-
matic CeH), 2983 (aliphatic CeH), 1769 (C]O), 1599 (C]N). 1H
NMR (CDCl3, 300 MHz)
1.06 (39%), 1.21 (% 61) (t, J ¼ 7.5 Hz, 3H,
O-
n
d
CH3), 2.21e2.32 (39%), 2.52e2.60 (61%) (q, 2H, CH2CO), 5.61 (37%),
5.70 (63%) (s, 2H, CH2N), 7.44e8.17 (m, 7H, naphthalene), 8.15
(37%), 8.26 (63%) (s, 1H, H3, triazole), 8.78 (37%), 9.12 (63%) (s, 1H,
H5, triazole). ESIþ: m/e 332 ([MþNaþH]þ), 331 ([MþNa]þ), 309
([MþH]þ), 235 (100%). Anal. Calcd. for C17H17ClN4O2.1/3H2O
(350.81): C, 58.21; H, 5.08; N, 15.97; Found: C, 58.10; H: 5.06; N,
16.37%.
4. Experimental part
4.1.1.2.2. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-(2-
methyl)propanoyloxime hydrochloride (5b). IR (
(aromatic CeH), 2975 (aliphatic CeH), 1722 (C]O), 1561 (C]N). 1H
NMR (CDCl3-d, 300 MHz)
n
cmꢁ1) 3136, 3045
4.1. Chemistry
d
1.06 (49%), 1.09 (51%) (d, J ¼ 7 Hz, 6H,
All the chemicals used in this study were purchased from E.
Merck, Fluka AG, and Aldrich. Milestone SartSYNTH (USA) micro-
wave reactor was used for microwave-assisted synthesis. Merck
Kieselgel 60 F254 plates were used for thin layer chromatography.
Melting points were determined using a Barnstead Electrothermal
9100 (USA) capillary melting point apparatus and uncorrected. The
column chromatography was performed using Kieselgel 60
(0.040e0.063 mm) (230e400 mesh ASTM) silica gel (Merck, Ger-
many). IR spectra were recorded with a Perkin Elmer Spectrum One
FTIR spectrometer using Attenuated Total Reflectance apparatus. 1H
NMR spectra of the compounds were recorded with Bruker Avonce
CH3), 2.37e2.50 (49%), 2.69e2.78 (51%) (m, 1H, CHCO), 5.56 (49%),
5.67 (51%) (s, 2H, CH2N), 6.65e8.08 (m, 7H, naphthalene), 8.12
(49%), 8.22 (% 51) (s, 1H, H3, triazole), 8.54 (49%), 8.90 (51%) (s, 1H,
H5, triazole). ESIþ: m/e 323 ([MþH]þ), 235,123 (100%). Anal. Calcld.
for C18H19ClN4O2.1/3H2O (364.83): C, 59.26; H, 5.43; N, 15.36;
Found: C, 58.79; H, 5.33; N, 15.82%.
4.1.1.2.3. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-tri-
methylacetyloxime (5c). IR (
(aliphatic CeH), 1755 (C]O), 1500 (C]N). 1H NMR (CDCl3-d,
300 MHz) 1.35 (s, 9H, CH3), 5.65 (s, 2H, CH2N), 7.53e7.96 (m, 7H,
naphthalene), 8.13 (s, 1H, H3, triazole), 8.25 (s, 1H, H5, triazole).
n
cmꢁ1) 3116 (aromatic CeH), 2975
d

414
S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
ESIþ: m/e 359 ([MþNa]þ, 100%), 322, 235. Anal. Calcld. for
(35%), 6.00 (65%) (s, 2H, CH2N), 7.28e8.40 (m, 8H, H3, triazole,
naphthalene), 10.36 (35%), 10.43 (65%) (s, 1H, H5, triazole). ESIþ: m/
e 373 ([MþNa]þ), 351 ([MþH]þ), 123 (100%). Anal. Calcld. for
C
19H20N4O2.1/4H2O (340.90): C, 66.94; H, 6.06; N, 16.44; Found: C,
67.45; H, 6.15; N, 16.24%.
4.1.1.2.4. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone
O-
C20H23ClN4O2.H2O (404.89): C, 59.33; H, 6.22; N, 13.84; Found: C,
butanoyloxime (5d). IR (
(aliphatic CeH), 1769 (C]O), 1506 (C]N). 1H NMR (CDCl3-d,
300 MHz)
0.91 (5%), 1.06 (95%) (t, J ¼ 7.5 Hz, 3H, CH3), 1.60e1.66
n
cmꢁ1) 3108 (aromatic CeH), 2968
59.91; H, 5.93; N, 14.16%.
4.1.1.2.10. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-
d
(4-methyl)pentanoyloxime hydrochloride (5j). IR (
3009 (aromatic CeH), 2954 (aliphatic CeH), 1765 (C]O), 1561 (C]
N). 1H NMR (CDCl3-d, 300 MHz)
0.82 (34%), 0.95 (% 66) (d,
n
cmꢁ1) 3133,
(5%), 1.77e1.84 (95%) (m, 2H, CH2CH3), 2.30 (5%), 2.57 (95%) (t,
J ¼ 7.5 Hz, 2H, CH2CO), 5.50 (5%), 5.65 (95%) (s, 2H, CH2N), 7.52e7.98
(m, 6H, H3ꢁ8, naphthalene), 7.95 (s, 1H, H3, triazole), 8.20 (s, 1H, H5,
triazole), 8.31 (m, 1H, H1, naphthalene). ESIþ: m/e 345 ([MþNa]þ),
323 ([MþH]þ), 235 (100%). Anal. Calcld. for C18H18N4O2.1/2H2O
(331.37): C, 65.24; H, 5.78; N, 16.91; Found: C, 64.95; H, 5.50; N,
16.98%.
d
J ¼ 6.2 Hz, 6H, CH3), 0.90e0.92 (m, 1H, CH), 1.45e1.53 (39%),
1.58e1.69 (61%) (m, 2H, CH2CH), 2.30 (39%), 2.61 (61%) (t, J ¼ 7.5 Hz,
2H, CH2CO), 5.83 (33%), 5.92 (67%) (s, 2H, CH2N), 7.50e8.24 (m, 7H,
naphthalene), 8.05 (33%), 8.38 (67%) (1H, s, H3, triazole), 9.83 (32%),
10.00 (68%) (s, 1H, H5, triazole). ESIþ: m/e 373 ([MþNa]þ), 351
([MþH]þ), 123 (100%). Anal. Calcld. for C20H23ClN4O2.1/2H2O
(395.89): C, 60.68; H, 6.11; N, 14.15; Found: C, 61.03; H, 5.84; N,
14.56%.
4.1.1.2.5. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-(2-
methyl)butanoyloxime (5e). IR (
2971, 235 (aliphatic CeH), 1751 (C]O), 1502 (C]N). 1H NMR
(CDCl3-d, 300 MHz)
n
cmꢁ1) 3115 (aromatic CeH),
d
1.00 (t, J ¼ 7.5 Hz, 3H, CH2CH3), 1.30 (d,
4.1.1.2.11. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone
(3-propyl)pentanoyloxime hydrochloride (5k). IR (
cmꢁ1) 3116,
3009 (aromatic CeH), 2959 (aliphatic CeH), 1757 (C]O), 1541 (C]
N). 1H NMR (CDCl3-d, 300 MHz)
0.80 (43%), 0.95 (57%) (t,
O-
J ¼ 7.2 Hz, 3H, CHCH3), 1.59e1.69 (m, 1H, Ha, CHCH2), 1.75e1.91 (m,
1H, Hb, CHCH2), 2.58 (m, 1H, CH), 5.65 (s, 2H, CH2N), 7.51e7.98 (m,
7H, naphthalene), 8.17 (s, 1H, H3, triazole), 8.30 (s, 1H, H5, triazole).
ESIþ: m/e 360 ([MþNaþH]þ, 100%), 359 ([MþNa]þ), 235. Anal.
Calcld. for C19H20N4O2 (336.40): C, 67.84; H, 5.99; N, 16.66; Found:
C, 67.65; H, 5.85; N, 16.44%.
n
d
J ¼ 7.4 Hz, 6H, CH3), 1.16e1.82 (m, 8H, 2CH2CH2), 2.30e2.37 (42%),
2.63e2.71 (58%) (m, 1H, CH), 5.90 (44%), 6.01 (56%) (s, 2H, CH2N),
7.50e8.43 (m, 9H, naphthalene, triazole). ESIþ: m/e 401
([MþNa]þ), 379 ([MþH]þ), 123 (100%). Anal. Calcld. for
4.1.1.2.6. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-(3-
methyl)butanoyloxime hydrochloride (5f). IR:
(aromatic CeH), 2969 (aliphatic CeH), 1759 (C]O), 1562 (C]N). 1H
NMR (CDCl3-d, 300 MHz)
n
(cmꢁ1) 3116, 3006
C22H27ClN4O2 (414.93): C, 63.40; H, 6.51; N, 13.45; Found: C, 63.40;
H, 6.51; N, 13.45%.
d
0.92 (31%), 1.07 (69%) (d, J ¼ 6.6 Hz, 6H,
4.1.1.2.12. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-
CH3), 2.02e2.13 (31%), 2.16e2.31 (69%) (m, 1H, CH), 2.18 (31%), 2.47
(69%) (d, J ¼ 6.9 Hz, 2H, CH2CO), 5.81 (31%), 5.90 (69%) (s, 2H, CH2N),
6.74e8.24 (m, 7H, naphthalene), 8.03 (31%), 8.38 (69%) (s, 1H, H3,
triazole), 9.60 (31%), 9.77 (69%) (s, 1H, H5, triazole). ESIþ: m/e 359
([MþNa]þ), 337 ([MþH]þ), 123 (100%). Anal. Calcld. for
cyclohexylcarbonyloxime (5l). IR (
2932 (aliphatic CeH)1748 (C]O), 1500 (C]N). 1H NMR (CDCl3-d,
300 MHz) 1.13e2.00 (m, 10H, 5CH2, cyclohexane), 2.53e2.63 (m,
n
cmꢁ1) 3120 (aromatic CeH),
d
1H, CH, cyclohexane), 5.65 (s, 2H, CH2N), 7.51e7.97 (m, 7H, naph-
thalene), 8.17 (s,1H, H3, triazole), 8.29 (s,1H, H5, triazole). ESIþ: m/e
385 ([MþNa]þ,100%), 236, 235. Anal. Calcld. for C21H22N4O2.1/3H2O
(368.44): C, 68.46; H, 6.20; N, 15.21; Found: C, 69.02; H, 6.34; N,
14.73%.
C
19H21ClN4O2.1/2H2O (381.15): C, 59.76; H, 5.81; N, 14.67; Found: C,
60.18; H, 5.63; N, 14.79%.
4.1.1.2.7. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone
O-
pentanoyloxime hydrochloride (5g). IR (
matic CeH), 2959 (aliphatic CeH), 1762 (C]O), 1601 (C]N). 1H
NMR (CDCl3-d, 300 MHz)
0.84 (30%), 0.97 (69%) (t, J ¼ 7.3 Hz, 3H,
n
cmꢁ1) 3124, 3009 (aro-
4.1.1.2.13. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-
phenylacetyloxime hydrochloride (5m). IR (
CeH), 2958 (aliphatic CeH), 1774 (C]O), 1536 (C]N). 1H NMR
(CDCl3-d, 300 MHz) 3.51 (s, 2H, CH2CO), 5.53 (s, 2H, CH2N),
n
cmꢁ1) 3133 (aromatic
d
CH3), 1.23e1.38 (31%), 1.40e1.48 (69%) (m, 2H, CH2CH3), 1.50e1.60
(31%), 1.65e1.79 (70%) (m, 2H, CH2CH2CH3), 2.31 (34%), 2.61 (66%)
(t, J ¼ 7.5 Hz, 2H, CH2CO), 5.78 (30%), 5.86 (70%) (s, 2H, CH2N),
7.32e8.15 (m, 7H, naphthalene), 8.17 (30%), 8.36 (70%) (s, 1H, H3,
triazole), 9.49 (30%),9.69 (70%) (s, 1H, H5, triazole). ESIþ: m/e 359
([MþNa]þ), 337 (MþH), 235 (100%). Anal. Calcld. for
d
6.65e9.01 (m, 14H, benzene, naphthalene, triazole). ESIþ: m/e 393
([MþNa]þ), 371 ([MþH]þ), 123 (100%). Anal. Calcld. for
C22H19ClN4O2.H2O (424.88): C, 62.19; H, 4.98; N, 13.19; Found:
C,61.91; H, 4.97; N, 13.40%.
4.1.1.2.14. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-
C
19H21ClN4O2.1/5H2O (376.45): C, 60.62; H, 5.73; N, 14.88; Found: C,
(E)-3-phenylprop-2-enoyloxime (5n). IR (
CeH), 1746 (C]O), 1633 (alkenyl C]C), 1505 (C]N). 1H NMR
(CDCl3-d, 300 MHz)
n
cmꢁ1) 3104 (aromatic
60.34; H, 5.49; N, 15.52%.
4.1.1.2.8. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-(2-
d
5.73 (s, 2H, CH2N), 6.64 (d, J ¼ 15.9 Hz, 1H,
methyl)pentanoyloxime hydrochloride (5h). IR (
(aromatic CeH), 2958 (aliphatic CeH), 1758 (C]O), 1538 (C]N). 1H
NMR (CDCl3-d, 300 MHz)
0.80 (35%), 0.96 (65%) (t, J ¼ 7.2 Hz, 3H,
n
cmꢁ1) 3119, 3009
COCH), 7.41e8.02 (m, 12H, benzene, naphthalene, CH-Ph), 8.25 (s,
1H, H3, triazole), 8.33 (s, 1H, H5, triazole). ESIþ: m/e 405 ([MþNa]þ,
100%), 236, 235. Anal. Calcld. for C23H18N4O2.1/3H2O (388.43): C,
71.12; H, 4.84; N, 14.4; Found: C, 71.41; H, 4.62; N, 14.4%.
d
CH2CH3), 1.11 (38%), 1.30 (62%) (d, J ¼ 7 Hz, 3H, CHCH3), 1.16e1.26
(29%), 1.33e1.45 (71%) (m, 2H, CH2CH3), 1.47e1.61 (37%), 1.65e1.84
(63%) (m, 2H, CH2CH), 2.37e2.49 (35%), 2.71e2.81 (65%) (m, 1H,
CH), 5.80 (35%),5.88 (65%) (s, 2H, CH2N), 7.28e8.19 (m, 7H, naph-
thalene), 8.02 (35%), 8.37 (65%) (s, 1H, H3, triazole), 9.58 (35%), 9.86
(65%) (s, 1H, H5, triazole). ESIþ: m/e 373 ([MþNa]þ), 351 ([MþH]þ),
235 (100%). Anal. Calcld. for C20H23ClN4O2 (386.88): C, 62.09; H,
5.99; N, 14.48; Found: C, 62.41; H, 5.59; N, 14.45%.
4.2. Pharmacology
Male Swiss albino mice of 25 2 g weight provided by the
Institution of Veterinary Control and Research Institute, Elazig,
Turkey, and Dual Impedance Research Stimulator (Harvard),
corneal electrodes, PEG 400 (Merck), and rotarod were used for the
anticonvulsant and neurotoxicity evaluation of the compounds.
Screenings were fulfilled in the laboratories of Laboratory Animal
Production Centre of Inonu University and Hacettepe University
Faculty of Pharmacy Department of Pharmaceutical Chemistry
laboratories according to the protocol of National Institute of
Neurological Disorder and Stroke (NINDS) Epilepsy Therapy
4.1.1.2.9. 1-(2-Naphthyl)-2-(1H-1,2,4-triazol-1-yl)ethanone O-(3-
methyl)pentanoyloxime hydrochloride (5i). IR (
(aromatic CeH), 2964 (aliphatic CeH), 1758 (C]O), 1562 (C]N). 1H
NMR (CDCl3-d, 300 MHz)
n
cmꢁ1) 3117, 3008
d
0.80 (35%), 0.93 (65%) (t, J ¼ 7.4 Hz, 3H,
CH3), 0.78 (35%), 1.02 (65%) (d, J ¼ 6.7 Hz, 3H, CHCH3), 1.13e1.51 (m,
2H, CH2CH3),1.77e2.04 (m,1H, CH), 2.05e2.64 (m, 2H, CH2CO), 5.91

S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
415
Screening Program (ETSP) (formerly known as Anticonvulsant
Acknowledgements
Screening Program) [51] with the permission granted by the Lab-
oratory Animals Ethic Committee of Inonu University, Malatya,
Turkey (permission date and number: 29.11.2011, 2011/A92).
Suspensions of each compound in PEG 400 at 30, 100, and
300 mg/kg concentrations were intraperitoneally administered to
the mice 0.5 or 4 h prior to seizure induction. In MES test the mice
were applied 0.9% saline in the corneas and electrocuted for 0.2 s
with 60 Hz alternative current using a pair of electrodes and the
mice with a hind limb tonic extension of less than 90ꢀ were
considered to be protected by the administered compound. In s.c.
MET test 85 mg/kg of metrazol was injected subcutaneous in the
midline of the neck to induce seizure then the mice were observed
for the next 30 min. The mice without an episode of clonic spasms,
approximately 3e5 s, of the fore and/or hind limbs, jaws, or
vibrissae were considered to be protected by the administered
compound. A negative control was performed by evaluation of the
responses from placebo (PEG)-administered mice to seizure in-
duction and mice were observed to undergo the episodes specific
for each model [51].
We are thankful to Prof. Dr. Erhan Palaska for providing mass
spectra of some of the compounds.
Appendix A. Supplementary data
Supplementary data associated with this article can be found in
the online version, at http://dx.doi.org/10.1016/j.ejmech.2016.08.
032. These data include MOL files and InChiKeys of the most
important compounds described in this article.
References
[1] M.J. England, C.T. Liverman, A.M. Schultz, L.M. Strawbridge, Epilepsy across
the spectrum: promoting health and understanding, Epilepsy Behav. 25
(2012) 266e276.
[2] R.A. Scott, S.D. Lhatoo, J.W. Sander, The treatment of epilepsy in developing
countries: where do we go from here? B. World Health Organ. 79 (2001)
344e351.
[3] D. Schmidt, Drug treatment of epilepsy: options and limitations, Epilepsy
Behav. 15 (2009) 56e65.
[4] D. Nardi, A. Tajana, A. Leonardi, R. Pennini, F. Portioli, M.J. Magistretti,
A. Subissi, Synthesis and anticonvulsant activity of N-(benzoylalkyl) imidaz-
oles and N-(.omega.-phenyl-.omega.-hydroxyalkyl) imidazoles, J. Med. Chem.
24 (1981) 727e731.
Prior to seizure induction the mice were put on a rod revolving
at 6 rpm. Those which fell off the rod three times in 1 min were
considered toxic and the compounds administered to them were
deemed to have neurotoxicity. Placebo-administered mice were
observed to succeed the test [51].
[5] K.A. Walker, M.B. Wallach, D.R. Hirschfeld, 1-(Naphthylalkyl)-1H-imidazole
derivatives, a new class of anticonvulsant agents, J. Med. Chem. 24 (1981)
67e74.
[6] P. Granger, B. Biton, C. Faure, X. Vige, H. Depoortere, D. Graham, S.Z. Langer,
B. Scatton, P. Avenet, Modulation of the gamma-aminobutyric acid type A
receptor by the antiepileptic drugs carbamazepine and phenytoin, Mol.
Pharmacol. 47 (1995) 1189e1196.
[7] R.L. Macdonald, K.M. Kelly, Antiepileptic drug mechanisms of action, Epilepsia
36 (1995) S2eS12.
[8] D.W. Robertson, J.H. Krushinski, E. Beedle, J.D. Leander, D.T. Wong, R. Rathbun,
Structure-activity relationships of (arylalkyl) imidazole anticonvulsants:
comparison of the (fluorenylalkyl) imidazoles with nafimidone and denzimol,
J. Med. Chem. 29 (1986) 1577e1586.
4.3. Molecular docking studies
Ligands were sketched on MarvinSketch and their physico-
chemical properties were calculated using Calculator Plugins,
Marvin 15.5.25.0, 2015, ChemAxon (http://www.chemaxon.com).
Ligands were prepared for docking using LigPrep, version 3.6,
€
Schrodinger, LLC, New York, NY, 2015. OPLS 2005 force field pa-
[9] S. Dalkara, A. Karakurt, Recent progress in anticonvulsant drug research:
strategies for anticonvulsant drug development and applications of antiepi-
leptic drugs for non-epileptic central nervous system disorders, Curr. Top.
Med. Chem. 12 (2012) 1033e1071.
rameters were used for their minimization [52]. The open/activated
Naþ channel inner pore and GABAAR structure models were ob-
tained from the related literature [14,15]. The coordinates of the
docked ligands (phenytoin for VGSC, GABA and diazepam for
GABAAR) were set as centres of search space for grid generation and
docking. Following settings were used on Glide, version 6.9,
€
€
[10] A. Karakurt, M.D. Aytemir, J.P. Stables, M. Ozalp, F. Betül Kaynak, S. Ozbey,
S. Dalkara, Synthesis of some oxime ether derivatives of 1-(2-naphthyl)-2-(1,
2, 4-triazol-1-yl)ethanone and their anticonvulsant and antimicrobial activ-
ities, Arch. Pharm. 339 (2006) 513e520.
€
€
[11] A. Karakurt, M. Ozalp, S¸ . Is¸ ık, J.P. Stables, S. Dalkara, Synthesis, anticonvulsant
Schrodinger, LLC, New York, NY, 2015: Extra precision was set, li-
and antimicrobial activities of some new 2-acetylnaphthalene derivatives,
gands were kept flexible, 10 poses were written at most per ligand,
post-minimization was enabled and per-residue interaction scores
for residues within 12 Å of grid centre were written. Maestro,
Bioorg. Med. Chem. 18 (2010) 2902e2911.
€
ꢀ
[12] A. Karakurt, M.A. Alagoz, B. Sayoglu, Ü. Çalıs¸ , S. Dalkara, Synthesis of some
novel 1-(2-naphthyl)-2-(imidazol-1-yl)ethanone oxime ester derivatives and
evaluation of their anticonvulsant activity, Eur. J. Med. Chem. 57 (2012)
275e282.
€
version 10.4, Schrodinger, LLC, New York, NY, 2015, was used as the
user interface during all procedures. Phenytoin and diazepam were
re-docked into their binding sites and their original conformations
were reproduced (RMSD: 1.28 Å for phenytoin and 0.84 Å for
diazepam). Docking solutions were visually inspected using pose
viewer of Maestro.
[13] C.O. Kappe, Controlled microwave heating in modern organic synthesis,
Angew. Chem. Int. Ed. 43 (2004) 6250e6284.
[14] G.M. Lipkind, H.A. Fozzard, Molecular model of anticonvulsant drug binding to
the voltage-gated sodium channel inner pore, Mol. Pharmacol. 78 (2010)
631e638.
[15] R. Bergmann, K. Kongsbak, P.L. Sørensen, T. Sander, T. Balle, A Unified model of
the GABAA receptor comprising agonist and benzodiazepine binding sites,
PloS One 8 (2013) e52323.
[16] T. Immediata, A.R. Day,
b-Naphthyl derivatives of ethanolamine and n-
4.4. X-ray crystallography
substituted ethanolamines, J. Org. Chem. 5 (1940) 512e527.
[17] Ü. Çalis¸ , S. Dalkara, M. Ertan, R. Sunal, The significance of the imidazole ring in
anticonvulsant activity of (arylalkyl) imidazoles, Arch. Pharm. 321 (1988)
841e846.
[18] H. Baji, M. Flammang, T. Kimny, F. Gasquez, P. Compagnon, A. Delcourt,
Synthesis and antifungal activity of novel (1-aryl-2-heterocyclyl)ethyl-
ideneaminooxymethyl-substituted dioxolanes, Eur. J. Med. Chem. 30 (1995)
617e626.
[19] B. Neises, W. Steglich, Simple method for the esterification of carboxylic acids,
Angew. Chem. Int. Ed. 17 (1978) 522e524.
[20] A.L. Spek, Structure validation in chemical crystallography, Acta Crystallogr.
Sect. D. 65 (2009) 148e155.
Single crystal data was collected with RIGAKU R-Axis Rapid II
DW with Dual Wavelength Micro Max 007DW XG and VariMax DW
optics single crystal X-ray diffractometer system with curved im-
aging plate detector. The data collection, cell refinement and data
reduction were performed by CrystalClear-SM Expert 2.0 r16 [53]
program. The crystal structure was solved by using SIR-2011 [54]
and refined by using the SHELX [55].
[21] L.J. Farrugia, ORTEP-3 for windows e a version of ORTEP-III with a graphical
user interface (GUI), J. Appl. Crystallogr. 30 (1997) 565.
[22] A. Balsamo, B. Macchia, A. Martinelli, E. Orlandini, A. Rossello, F. Macchia,
G. Brocalli, P. Domiano, Synthesis and antimicrobial properties of substituted
3-aminoxy-(E)-2-methoxyiminopropionyl penicillins and cephalosporins, Eur.
Conflict of interest
Authors declare no conflict of interest.

416
S. Sari et al. / European Journal of Medicinal Chemistry 124 (2016) 407e416
Pharmacol. Sci. 16 (1995) 162e168.
J. Med. Chem. 25 (1990) 227e233.
[23] R.J. Stoodley, Molecular rearrangements, in: B. Capon, C.W. Rees (Eds.),
Organic Reaction Mechanisms, John Wiley & Sons Ltd., London, 1972, pp.
265e267.
[24] A. Karakurt, S. Dalkara, M. Ozalp, S. Ozbey, E. Kendi, J.P. Stables, Synthesis of
some 1-(2-naphthyl)-2-(imidazole-1-yl) ethanone oxime and oxime ether
derivatives and their anticonvulsant and antimicrobial activities, Eur. J. Med.
Chem. 36 (2001) 421e433.
[39] D. Berezhnoy, Y. Nyfeler, A. Gonthier, H. Schwob, M. Goeldner, E. Sigel, On the
benzodiazepine binding pocket in GABAA receptors, J. Biol. Chem. 279 (2004)
3160e3168.
[40] M. Davies, A.N. Bateson, S.M. Dunn, Structural requirements for ligand in-
teractions at the benzodiazepine recognition site of the GABAA receptor,
J. Neurochem. 70 (1998) 2188e2194.
€
€
[41] P.B. Wingrove, P. Safo, L. Wheat, S.A. Thompson, K.A. Wafford, P.J. Whiting,
[25] S. Korkmaz, G. Zararsiz, D. Goksuluk, MLViS: a Web tool for machine learning-
based virtual screening in early-phase of drug discovery and development,
PloS One 10 (2015) e0124600.
[26] S. Korkmaz, G. Zararsiz, D. Goksuluk, Drug/nondrug classification using sup-
port vector machines with various feature selection strategies, Comput. Meth.
Prog. Bio. 117 (2014) 51e60.
Mechanism of
aminobutyric acid type A receptors, Eur. J. Pharmacol. 437 (2002) 31e39.
[42] A. Buhr, R. Baur, E. Sigel, Subtle changes in residue 77 of the subunit of
GABAA receptors drastically alter the affinity for ligands of the
benzodiazepine binding site, J. Biol. Chem. 272 (1997) 11799e11804.
[43] P.B. Wingrove, S.A. Thompson, K.A. Wafford, P.J. Whiting, Key amino acids in
a-subunit selectivity of benzodiazepine pharmacology at g-
g
a1b2g2
[27] Z. Ozdemir, A. Karakurt, U. Calis, S. Günal, S. Isik, Z. Sibel Sahin, S. Dalkara,
Synthesis, anticonvulsant and antimicrobial activities of some new [1-(2-
naphthyl)-2-(pyrazol-1-yl) ethanone] oxime ethers, Med. Chem. 11 (2015)
41e49.
the g subunit of the g-aminobutyric acidA receptor that determine ligand
binding and modulation at the benzodiazepine site, Mol. Pharmacol. 52
(1997) 874e881.
[44] E. Sigel, M.T. Schaerer, A. Buhr, R. Baur, The benzodiazepine binding pocket of
[28] W.A. Catterall, From ionic currents to molecular mechanisms: the structure
and function of voltage-gated sodium channels, Neuron 26 (2000) 13e25.
[29] E. Sigel, M.E. Steinmann, Structure, function, and modulation of GABAA re-
ceptors, J. Biol. Chem. 287 (2012) 40224e40231.
[30] M. Mantegazza, G. Curia, G. Biagini, D.S. Ragsdale, M. Avoli, Voltage-gated
sodium channels as therapeutic targets in epilepsy and other neurological
disorders, Lancet Neurol. 9 (2010) 413e424.
recombinant a1b2g2 g-aminobutyric acidA receptors: relative orientation of
ligands and amino acid side chains, Mol. Pharmacol. 54 (1998) 1097e1105.
[45] K.R. Tan, R. Baur, A. Gonthier, M. Goeldner, E. Sigel, Two neighboring residues
of loop A of the a1 subunit point towards the benzodiazepine binding site of
GABAA receptors, FEBS Lett. 581 (2007) 4718e4722.
[46] S.M. Hanson, E.V. Morlock, K.A. Satyshur, C. Czajkowski, Structural re-
quirements for eszopiclone and zolpidem binding to the g-aminobutyric acid
[31] G.M. Lipkind, H.A. Fozzard, Molecular modeling of local anesthetic drug
binding by voltage-gated sodium channels, Mol. Pharmacol. 68 (2005)
1611e1622.
type-A (GABAA) receptor are different, J. Med. Chem. 51 (2008) 7243e7252.
[47] K.R. Tan, A. Gonthier, R. Baur, M. Ernst, M. Goeldner, E. Sigel, Proximity-
accelerated chemical coupling reaction in the benzodiazepine-binding site of
[32] D.S. Ragsdale, J.C. McPhee, T. Scheuer, W.A. Catterall, Molecular determinants
of state-dependent block of Naþ channels by local anesthetics, Science 265
(1994) 1724e1728.
[33] D.S. Ragsdale, J.C. McPhee, T. Scheuer, W.A. Catterall, Common molecular
determinants of local anesthetic, antiarrhythmic, and anticonvulsant block of
voltage-gated Naþ channels, Proc. Natl. Acad. Sci. U. S. A. 93 (1996)
9270e9275.
g-aminobutyric acid type A receptors superposition of different allosteric
modulators, J. Biol. Chem. 282 (2007) 26316e26325.
[48] K.R. Tan, R. Baur, S. Charon, M. Goeldner, E. Sigel, Relative positioning of
diazepam in the benzodiazepine-binding-pocket of GABAA receptors,
J. Neurochem. 111 (2009) 1264e1273.
[49] A. Buhr, M.T. Schaerer, R. Baur, E. Sigel, Residues at positions 206 and 209 of
the
a1 subunit of g-aminobutyric acidA receptors influence affinities for
[34] V. Yarov-Yarovoy, J. Brown, E.M. Sharp, J.J. Clare, T. Scheuer, W.A. Catterall,
Molecular determinants of voltage-dependent gating and binding of pore-
benzodiazepine binding site ligands, Mol. Pharmacol. 52 (1997) 676e682.
[50] M.T. Schaerer, A. Buhr, R. Baur, E. Sigel, Amino acid residue 200 on the a1
subunit of GABAA receptors affects the interaction with selected benzodiaz-
epine binding site ligands, Eur. J. Pharmacol. 354 (1998) 283e287.
[51] R. Krall, J. Penry, B. White, H. Kupferberg, E. Swinyard, Antiepileptic drug
development: II. Anticonvulsant drug screening, Epilepsia 19 (1978) 409e428.
[52] J.L. Banks, H.S. Beard, Y. Cao, A.E. Cho, W. Damm, R. Farid, A.K. Felts,
T.A. Halgren, D.T. Mainz, J.R. Maple, Integrated modeling program, applied
chemical theory (IMPACT), J. Comput. Chem. 26 (2005) 1752e1780.
[53] Rigaku, CrystalClear, Rigaku Corporation, Tokyo, Japan, 2014.
[54] M.C. Burla, R. Caliandro, M. Camalli, B. Carrozzini, G.L. Cascarano,
C. Giacovazzo, M. Mallamo, A. Mazzone, G. Polidori, R. Spagna, SIR2011: a new
package for crystal structure determination and refinement, J. Appl. Crys-
tallogr. 45 (2012) 357e361.
blocking drugs in transmembrane segment IIIS6 of the Naþ channel
a sub-
unit, J. Biol. Chem. 276 (2001) 20e27.
[35] R.A. Friesner, J.L. Banks, R.B. Murphy, T.A. Halgren, J.J. Klicic, D.T. Mainz,
M.P. Repasky, E.H. Knoll, M. Shelley, J.K. Perry, Glide: a new approach for
rapid, accurate docking and scoring. 1. Method and assessment of docking
accuracy, J. Med. Chem. 47 (2004) 1739e1749.
[36] R.A. Friesner, R.B. Murphy, M.P. Repasky, L.L. Frye, J.R. Greenwood,
T.A. Halgren, P.C. Sanschagrin, D.T. Mainz, Extra precision glide: docking and
scoring incorporating a model of hydrophobic enclosure for protein-ligand
complexes, J. Med. Chem. 49 (2006) 6177e6196.
[37] T.A. Halgren, R.B. Murphy, R.A. Friesner, H.S. Beard, L.L. Frye, W.T. Pollard,
J.L. Banks, Glide: a new approach for rapid, accurate docking and scoring. 2.
Enrichment factors in database screening, J. Med. Chem. 47 (2004)
1750e1759.
[55] G.M. Sheldrick, A short history of SHELX, Acta Crystallogr. Sect. A 64 (2008)
112e122.
[38] G.B. Smith, R.W. Olsen, Functional domains of GABAA receptors, Trends