Design, synthesis and biological evaluation of N-substituted α-hydroxyimides and 1,2,3-oxathiazolidine-4-one-2,2-dioxides with anticonvulsant activity

Article abstract of DOI:10.1080/14756366.2019.1651722

In this investigation, we studied a family of compounds with an oxathiazolidine-4-one-2,2-dioxide skeleton and their amide synthetic precursors as new anticonvulsant drugs. The cyclic structures were synthesized using a three-step protocol that include so

Full text of DOI:10.1080/14756366.2019.1651722

Journal of Enzyme Inhibition and Medicinal Chemistry
ISSN: 1475-6366 (Print) 1475-6374 (Online) Journal homepage: https://www.tandfonline.com/loi/ienz20
Design, synthesis and biological evaluation
of N-substituted α-hydroxyimides and 1,2,3-
oxathiazolidine-4-one-2,2-dioxides with
anticonvulsant activity
Laureano L. Sabatier, Pablo H. Palestro, Andrea V. Enrique, Valentina
Pastore, María L. Sbaraglini, Pedro Martín & Luciana Gavernet
To cite this article: Laureano L. Sabatier, Pablo H. Palestro, Andrea V. Enrique, Valentina
Pastore, María L. Sbaraglini, Pedro Martín & Luciana Gavernet (2019) Design, synthesis and
biological evaluation of N-substituted α-hydroxyimides and 1,2,3-oxathiazolidine-4-one-2,2-
dioxides with anticonvulsant activity, Journal of Enzyme Inhibition and Medicinal Chemistry, 34:1,
1465-1473, DOI: 10.1080/14756366.2019.1651722
To link to this article: https://doi.org/10.1080/14756366.2019.1651722
© 2019 The Author(s). Published by Informa
UK Limited, trading as Taylor & Francis
Group.
View supplementary material
Published online: 14 Aug 2019.
Submit your article to this journal
View Crossmark data
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=ienz20

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
2019, VOL. 34, NO. 01, 1465–1473
https://doi.org/10.1080/14756366.2019.1651722
RESEARCH PAPER
Design, synthesis and biological evaluation of N-substituted a-hydroxyimides and
1,2,3-oxathiazolidine-4-one-2,2-dioxides with anticonvulsant activity
a
a
b
b
a
ꢀ
Laureano L. Sabatier , Pablo H. Palestro , Andrea V. Enrique , Valentina Pastore , Marıa L. Sbaraglini ,
b
a
ꢀ
Pedro Martın and Luciana Gavernet
aMedicinal Chemistry, Department of Biological Sciences, Faculty of Exact Sciences, National University of La Plata, La Plata, Argentina;
b
ꢀ
ꢀ
Instituto de Estudios Inmunologicos y Fisiopatologicos (IIFP), CONICET—Universidad Nacional de la Plata), Facultad de Ciencias Exactas,
Universidad Nacional de La Plata, La Plata, Argentina
ABSTRACT
ARTICLE HISTORY
Received 15 May 2019
Revised 23 July 2019
Accepted 25 July 2019
In this investigation, we studied a family of compounds with an oxathiazolidine-4-one-2,2-dioxide skeleton
and their amide synthetic precursors as new anticonvulsant drugs. The cyclic structures were synthesized
using a three-step protocol that include solvent-free reactions and microwave-assisted heating. The com-
pounds were tested in vivo through maximal electroshock seizure test in mice. All the structures showed
activity at the lower doses tested (30 mg/Kg) and no signs of neurotoxicity were detected. Compound
KEYWORDS
Epilepsy; 1,2,3-
encoded as 1g displayed strong anticonvulsant effects in comparison with known anticonvulsants (ED50
¼
oxathiazolidine-4-one-2,2-
dioxide; a-hydroxyimides;
docking; sodium channels
blockers; MES test;
microwave-
29 mg/Kg). First approximations about the mechanisms of action of the cyclic structures were proposed
by docking simulations and in vitro assays against sodium channels (patch clamp methods).
assisted synthesis
Introduction
known ACDs7. Even today phenytoin is used in epilepsy treatment
for generalized tonic-clonic and partial seizures8.
The advances in the understanding of epilepsy and its comorbid
conditions have triggered notable improvements of the pharma-
cological options for the treatment of this disorder. The new
developments include innovative processes like the use of 3D
printing technologies for improving the delivery of anticonvulsant
drugs (ACD)1, but mostly involve the introduction of new
approved compounds2. There are more than 20 ACDs currently
available for use in the United States, which means that the thera-
peutic arsenal has been doubled over the past 15 years2. However,
the discovery of more effective compounds continues to be a top
priority for researches in the field3. According to World Health
Organization, there are 50 million people worldwide with epilepsy
and one-third of them are unable to control the disorder with
ACDs4. Therefore, most of the efforts are now focused on lowering
the number of patients that experience resistance to the available
medications. In this investigation, we studied a family of structures
with an oxathiazolidine-4-one-2,2-dioxide skeleton and their syn-
thetic precursors as new candidates of ACDs. The rationale for
selecting the cyclic scaffold was based on its bioisosteric relation-
ship with the classical ACD phenytoin (Figure 1). The anticonvul-
sant action of phenytoin was identified by Putnam and Merrit in
1937 by means of their pioneering model for rapid screening of
ACD, which tested the ability of candidates to protect against
electroshock-induced convulsions in cats6. This phenotypic screen-
ing was the starting point for the development of acute models
In a previous work, on some N-derivatives of oxathiazolidine-4-
one-2,2-dioxides, we synthesized only four new compounds with
the scaffold and they showed activity against the Maximal
Electroshock Seizure (MES) test in mice (Figure 1)5. The MES test
consist of the electrical induction of the convulsions and it is asso-
ciated with the generalized tonic–clonic seizures9. Interestingly,
the synthetic precursors of the cycles, a-hydroxyamides, also
exhibited anticonvulsant action in MES test5. These results encour-
aged us to expand the set with new structures, with the aim of
exploring the anticonvulsant action in both a-hydroxyamides and
their cyclic derivatives.
The synthetic procedure involved the preparation of the
a-hydroxyamides from the ammonolysis reaction between 2-
hydroxyisobutyl methyl ester and the corresponding amine fol-
lowed by a cyclization reaction and the oxidation of the product5.
Both sets of a-hidroxyamides and oxathiazolidine-4-one-2,2-diox-
ide derivatives were evaluated against MES test in mice9. Toxicity
was also tested by the standardized Rotorod test, which is also
included in the primary phase of anticonvulsant screening pro-
gram9. To get inside into the possible mode of action of the com-
pounds, we analysed the capacity of the structures of blocking
the Nav1.2 isoform of the voltage-gated sodium channels (VGSC).
The selection of the target was supported on preliminary studies
about one compound of the cyclic family, the 3-butyl-5,5-dimethyl
of seizures in animals that successfully identified the majority of oxathiazolidine-4-one-2,2-dioxide (compound D, Figure 1), which
CONTACT Luciana Gavernet
lgavernet@biol.unlp.edu.ar, lgavernet@gmail.com
Medicinal Chemistry, Department of Biological Sciences, Faculty of Exact
Sciences, National University of La Plata, La Plata, Argentina
Supplemental data for this article can be accessed here.
ß 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.

1466
L. L. SABATIER ET AL.
Figure 1. Chemical structure of phenytoin (A), and the oxathiazolidine-4-one-2,2-dioxides previously synthesized (B–E).5
has sodium channels (VGSC) blocking properties10. Additionally, it (–CH3)2), 0.90 (t, J ¼ 6.7 Hz, 3H, –CH3). 13C NMR (126 MHz, CDCl3) d
is well known that Nav1.2 isoform is the molecular target of many
ACDs like phenytoin, lamotrigine, carbamazepine, oxcarbazepine,
eslicarbazepine, zonisamide and lacosamide11,12. First, we simu-
lated the interaction between the compounds synthesized and a
3D model of the Nav1.2 isoform by means of docking protocols.
Then, the docking candidates and other structures predicted as
inactive were tested on Nav1.2 currents using the patch
clamp technique.
175.91 (CO), 73.29 (–C–OH), 39.48 (–CH2–NH–), 31.16
(–CH2–(CH2)3–NH–), 29.51(–CH2–CH2–NH–), 27.61 ((–CH3)2), 26.54
(–CH2–(CH2)2–NH–), 22.76 (–CH2–(CH2)4–NH–), 13.8 (–CH3).
N-isopropyl-2-hydroxyisopropylamide (1b). Yield with MW
heating: 24%, yield with conventional heating 32% (mp: 66–67,
hexane). 1H NMR (500 MHz, CDCl3) d 6.71 (s, 1H, –NH), 4.05–3.88
(m, 1H, –CH), 3.40 (s, 1H, –OH), 1.43 (s, 6H, (–CH3)2), 1.16 (d,
J ¼ 6.6 Hz, 6H, –CH–(CH3)2). 13C NMR (126 MHz, CDCl3) d 175.96
(CO), 72.52 (–C–OH), 40.64 (–NH–CH–), 27.54((–CH3)2),
22.24(–CH–(CH3)2).
Materials and methods
N-isobutyl-2-hydroxyisobutylamide (1c). Yield 57% (mp:
104–105, hexane). 1H NMR (500 MHz, CDCl3) d 6.17 (s, 1H, –NH),
3.08 (t, J ¼ 6.5 Hz, 2H, –CH2), 1.84–1.71 (m, 1H, –CH), 1.45 (s, 6H,
(–CH3)2), 0.93 (d, J ¼ 6.7 Hz, 6H, –CH–(CH3)2) 13C NMR (126 MHz,
Chemistry
Microwave reactions were carried out in an Anton-Paar-mono-
wave-300 reactor (monowave, maximum power 850 W, tempera-
ture control via IR-sensor, vial volume: 10–30 mL). Melting points
were determined using capillary tubes with an electrothermal
melting point apparatus and are uncorrected. Thin-layer chroma-
tography (TLC) was performed with aluminium backed sheets with
silica gel 60 F254 (Merck, ref 1.05554), and the spots were visual-
ized with 254 nm UV light and 5% aqueous solution of ammonium
molybdate (VI) tetrahydrate. Column chromatography was per-
CDCl3)
d
176.23 (CO), 73.33(–C–OH), 46.37(–NH–CH2–),
28.55(–NH–CH2–CH–), 27.95 ((–CH3)2), 20.02(–CH–(CH3)2).
N-(1-ethyl) benzyl-2-hydroxyisobutylamide (1f). Yield with
MW heating: 18%, Yield with conventional heating: 47% (mp:
1
63–64, hexane). H NMR (500 MHz, CDCl3) d 7.38–7.27 (m, 5H, Ar),
7.02 (1, 1H, –NH), 4.86 (q, 1H, –NH–CH–Ar), 2.13 (s, 1H, –OH), 1.86
(p, J ¼ 7.3 Hz, 2H, –CH2–), 1.49 (s, 3H, –CH3), 1.44 (s, 3H, –CH3),
0.93 (t, J ¼ 7.4 Hz, 3H, –CH2–CH3). 13C NMR (126 MHz, CDCl3) d
175.50 (CO), 142.15 – 126.49 (Ar), 73.68 (–C–OH), 54.54 (–NH–CH–),
29.38–28.06 ((–CH3)2), 27.86 (–CH2–CH3),10.69 (–CH2–CH3).
N-(p-fluoro) benzyl-2-hydroxyisobutylamide (1g). Yield 28%
(mp: 73–74.5, hexane). 1H NMR (500 MHz, CDCl3) d 7.27–6.97 (m,
5H, Ar þ NH), 4.41(d, J ¼ 6.0 Hz, 2H, –NH–CH2–Ar), 2.22 (s, 1H,
–OH), 1.48 (s, 6H, (–CH3)2). 13C NMR (126 MHz, CDCl3) d 176.02
(CO), 163.16, 161.20 (C–F), 134.07 –115.48 (Ar), 73.78 (–C–OH),
42.57(–NH–CH2–), 28.04 ((–CH3)2).
1
formed on silica gel 60 (70–230 mesh, Merck, ref 1.07734.2500). H
and 13C NMR spectra were recorded on a Bruker Avance 500 MHz
spectrometer. The chemical shifts were reported in ppm (d scale)
relative to internal TMS, and coupling constants were reported in
Hertz (Hz).
Synthesis of a-hydroxyamides 1(a–g). Compounds 1d and 1e
were previously synthesized5,13–17. MW-assisted free solvent syn-
thesis was performed by means of the previously developed pro-
cedures:5 a-hydroxyisobutylmethyl ester (20 mmol) was put into a
dry vessel with the corresponding amine (24 mmol) and a Teflon-
coated magnetic stirring bar. The reactor was set at 150 ꢀC, the
reaction time was 30 min and the mixtures were monitored by
TLC. Amides 1b, 1d and 1f, were also synthesized by conventional
heating and 30% w/w of 1,5,7-triazabicyclo [4.4.0] dec-5-ene (TBD)
as catalyst to improve the yield of the products achieved with
MW heating. These reactions were conducted at 70 ꢀC during 20 h
of reaction and the yields with this alternative route increased
from 24 to 32% for 1b, 5 to 56% for 1d and 18 to 47% for 1f.
The isolation steps were similar in both conventional and
microwave assisted heating. After the reaction was concluded,
dichloromethane (30 mL) was added and the solution was washed
with 10 mL of 5% v/v hydrochloric acid (2ꢁ) and brine (1ꢁ). The
combined organic layers were dried over Na2SO4 and concen-
trated under reduced pressure. The residue was then purified by
silica column chromatography and/or crystallization. The pure
products were obtained as white solids in all cases except for
compound 1a.
Synthesis of N-derivative-1,2,3-oxathiazolidine-4-one-2,2-dioxides
3(a, c-e, g). First, we synthesized the monoxides as previously
described for cyclic sulfamates5. To a mixture of N-substituted
amide-2-hydroxyisobutylamide
(3 mmol)
and
triethylamine
(13 mmol) in CH2Cl2 anhydrous (8 mL) at 0 ꢀC, thionyl chloride
(4 mmol) in CH2Cl2 anhydrous (4 mL) was added dropwise. The
mixture was stirred under N2 overnight and concentrated to dry-
ness under reduced pressure to give N-derivative-1,2,3 oxatiazoli-
dine-4-one-2-oxides as yellow oils. Then, the dioxides were
obtained after oxidation with NaIO4 and RuCl3.6H2O:18–20 An
aqueous solution of NaIO4 (7 mmol) and ruthenium chloride
was added dropwise to a solution of N-derivative-1,2,3-oxathiazoli-
dine-4-one-2-oxides (4.5 mmol) in acetonitrile (6 mL) and dichloro-
methane (6 mL) at 0 ꢀC. The mixture was warmed to room
temperature and an hour later extracted twice with CH2Cl2. The
organic phases were combined, washed with water and brine,
dried (Na2SO4) and concentrated to dryness. The residue obtained
was flash chromatography on silica-gel 60 (70–230 mesh, Merck)
and mixtures of CH2Cl2/hexane to give a white solid for all the
N-hexyl-2-hydroxyisobutylamide (1a). Yield 40% (oil). 1H
NMR (500 MHz, CDCl3) d 6.75 (s, 1H, –NH), 3.26 (m, 2H, –CH2–NH),
1.53 (m, 2H, –CH2–CH2–NH–), 1.47 (s, 6H, –CH2–), 1.32 (s, 6H, compounds excluding 3a.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1467
Scheme 1. Synthetic route used in this investigation: (I) Ammonolysis reaction: MW radiation or catalyst, N2 atmosphere, (II) Cyclization reaction: SOCl2, triethylamine,
(III) Oxidation reaction: NaIO4 RuCl3.6H2O. Code for compounds: (1) a-hydroxyamide, (2) N-derivative-1,2,3-oxathiazolidine-4-one-2-oxide, (3) N-derivative-1,2,3-oxathia-
zolidine-4-one-2,2—dioxide. And a ¼ hexyl, b ¼ isopropyl, c ¼ isobutyl, d ¼ cyclohexyl, e ¼ 1-methylbenzyl, f ¼ 1-ethylbenzyl, g ¼ p-fluorbenzyl.
3-hexyl-5,5-dimethyl-1,2,3-oxathiazolidine-4-one-2,2-diox-
ide (3a). Yield 25% (oil). 1H NMR (500 MHz, CDCl3) d 3.67 (m, 2H,
–CH2–N), 1.79 (m, 2H, –CH2), 1.76 (s, 6H, (–CH3)2), 1.34–1.29 (m,
6H, –CH2), 0.91 (t, J ¼ 6.8 Hz, 3H, –CH3). 13C NMR (126 MHz, CDCl3)
electrodes by means of a UGO Basile equipment. In these condi-
tions, normal mice experience maximal seizures, characterized by
a short period of tonic flexion followed by a longer period of tonic
extension of the hind limbs and a final clonic episode9. Three
minutes before induction of convulsion, all animals were eval-
uated in the Rotorod test. Rotorod equipment consists of a fluted
roll divided by opaque discs rotating at a speed of 6 rpm. Animals
were arranged on the cylinder and the ability to maintain balance
on the rotarod for 1 min, in three consecutive tests, was evaluated.
The inability of animals to maintain balance during the three tests,
showing ataxia and sedation, was considered a sign of neurotox-
icity21. Quantitative studies in MES test were conducted for 1g at
the previously determined time of peak effect (TPE). The ED50 was
determined by treating groups of six albino mice. Different doses
were used for each drug at TPE. The method of Litchfield and
d
168.38 (CO), 93.26 ((CH3)2–C–O–), 42.03 (–N–CH–), 30.98
(–CH2–CH2–N–), 27.46 (–CH2–(CH2)3–N–), 26.09 (–CH2–(CH2)2–N–),
24.35 (–CH2–(CH2)4–N–), 22.49 ((–CH3)2), 13.85 (–CH3).
3–(1-isobutyl) 5,5-dimethyl-1,2,3-oxathiazolidine-4-one-2,2-
dioxide (3c). Yield 55% (mp: 40–41, hexane). 1H NMR (500 MHz,
CDCl3) d 3.49 (d, J ¼ 7.7 Hz, 2H, –N–CH2–), 2.30 – 2.19 (m, 1H,
–CH–CH3), 1.77 (s, 6H,(–CH3)2), 1.00 (d, J ¼ 6.7 Hz, 6H, –CH–CH3).
13C NMR (126 MHz, CDCl3) d 168.90 (CO), 93.21 ((CH3)2–C–O–),
48.99 (–N–CH2–), 27.13 (–N–CH2–CH–), 24.44 ((–CH3)2),
19.72 (–N–CH2–CH–(CH3)2).
3-cyclohexyl-5,5-dimethyl-1,2,3-oxathiazolidine-4-one-2,2-
dioxide (3d). Yield 71% (mp: 59–60, hexane). 1H NMR (500 MHz,
CDCl3) d 4.00–3.92 (m, 1H, –CH–), 2.06 – 1.17 (m, 10 H, cyclohexyl),
1.73 (s, 6H, (–CH3)2) 13C NMR (126 MHz, CDCl3) d 168.46 (CO),
91.81 ((CH3)2–C–O–), 56.05 (–N–CH–), 29.77, 25.71, 24.82 (cyclo-
hexyl), 24.28 ((–CH3)2).
Wilcoxon was used to compute the ED50 value22
.
Electrophysiology. The electrophysiological recordings were per-
formed using the patch-clamp technique in HEK293 cell line stably
expressing the hNav1.2 channel isoform (a kind gift from
GlaxoSmithKline, Stevenage, UK). The standard tight-seal whole-
cell configurations of the patch-clamp technique was used to
record macroscopic currents23. Whole-cell currents were filtered
with a 4-pole lowpass Bessel filter (Axopatch 200 A amplifier) at
2 kHz and digitized (Digidata 1440, Molecular devices) at a sample
frequency of 200 kHz (5 ms). Once the whole-cell configuration was
obtained, current stability was evaluated with a 15 ms-voltage-
clamp step from a holding potential of ꢂ80 mV to a test potential
of ꢂ20 mV repeated each 10 s. The time needed for the stabiliza-
tion was variable (approximately 10 min). The same voltage-clamp
step protocol was applied in the control (vehicle) or in the pres-
ence of compounds 1e, 1g, 3e and 3g, dissolved in 0.1% dime-
thylsulphoxide. After current stabilization on each condition, the
voltage dependence of the steady-state inactivation of sodium
channels was evaluated using a double voltage step protocol,
where the same depolarization to ꢂ10 mV followed different pre-
conditioning 2.5 s steps (from ꢂ130 to ꢂ40 mV). The available
fraction of sodium channels at each membrane potential (IVc/
Imax) was calculated as the ratio of peak sodium current meas-
ured at ꢂ10 mV, at each pre-conditioning voltage test pulse (IVc)
and the maximum peak current observed (Imax). The relationship
between the available fraction of sodium channels and the pre-
conditioning (named h curve) was plotted and fitted with a
Boltzmann equation (Equation (1)):
3–(1-methyl)
benzyl-5,5-dimethyl-1,2,3-oxathiazolidine-4-
one-2,2 dioxide (3e). Yield 58% (mp:66–67, hexane). 1H NMR
(500 MHz, CDCl3) d 7.53 – 7.27 (m, 5H, Ar), 5.32 (q, 1H, –CH), 1.94
(d, J ¼ 7.3 Hz, 3H, CH–CH3), 1.81 (s, 3H, –CH3), 1.51 (s, 3H, –CH3)
13C NMR (126 MHz, CDCl3) d 163.06 (CO), 133.81–115.44 (Ar),
73.24 ((CH3)2–C–O–), 41.68 (–N–CH–), 27.55 ((–CH3)2),
27.20 (–CH–CH3).
3-p-fluorbenzyl-5,5-dimethyl-1,2,3-oxathiazolidine-4-one-
2,2-dioxide (3g). Yield 52% (mp: 86–87, dichloromethane/hexane).
1H NMR (500 MHz, CDCl3) d 7.47 – 7.39 (m, 2H, Ar), 7.08 (m, 2H,
Ar), 4.77 (s, 2H, –CH2–Ar), 1.76 (s, 6H, (–CH3)2). 13C NMR (126 MHz,
CDCl3) d 168.44 (CO), 163.89, 161.92 (C–F), 130.87 –115.82 (Ar),
93.56 ((CH3)2–C–O–), 44.72 (–N–CH2–), 24.30 ((–CH3)2).
Biological assays
In vivo test. We used male albino mice (18–23 g) provided by the
Faculty of Veterinary, of the National University of La Plata. They
were maintained under a regime of 12-h light/dark cycle and
allowed free access to food and water. The animal care for the
experimental protocols was conducted in accordance with the
National Institutes of Health (NIH) guidelines for the Care and Use
of Laboratory Animals and it was approved by the Ethical
Committee of Exact Sciences Faculty of University of La Plata.
Mice were randomized to different treatments. Candidate’s solu-
tions were performed in polyethyleneglycol 400 (PEG 400) at a
rate of 3 mL/kg body weight and physiological solution was added
up to a maximum volume of 7 mL/kg. Mice were i. p. administered
with the synthetized compounds at doses of 30 or 100 mg/kg and
they were evaluated at 0.5 or 4 h.
IVc=Imax ¼ 1=ð1 þ expððVh ꢂ VÞ=kÞÞ
(1)
where the available fraction is given as IVc/Imax, Vh is the poten-
tial of half-maximal inactivation and k is the slope parameter.
Statistical significance of the changes in the Vh parameter induced
by the compounds was tested with F method (GraphPad Prism).
More details of the experiment are given in Supporting
online material.
Maximal electroshock seizures were provoked in mice by deliv-
Docking simulations. The molecules were docked into the
ering a 60 Hz/50 mA electrical stimulus for 0.2 s via ear clip model of the 3D structure of the a-subunit of the human Nav1.2

1468
L. L. SABATIER ET AL.
(open-pore conformation) previously constructed by us24. The
“docking active site” was defined based on the experimental data,
and includes residues numbered from Phe1764 to Tyr1771 as
important for the interactions of the channel with known ACDs25.
Previous analyses of the performance of different docking proto-
cols allowed us to select Autodock Vina26 as the best method to
discriminate known binders from non-binders of Nav1.2 through
the docking score, so the conditions of the docking are already
reported. We set the cut-off value as ꢂ8.1 Kcal/mol to differentiate
active from inactive compounds, which has associated specificity
and sensibility rates of 80% and 88%, respectively, in the receiver
operating characteristic (ROC) curve of the set of compounds used
for the validation of the protocol24.
Table 1. Pharmacological profile (phase I) of the synthesized compounds.
Compound
Dose (mg/Kg)
MESa
Rotorodb
Class
1
1a
30
100
30
100
30
100
30
100
30
100
30
100
30
100
30
100
30
100
30
100
30
100
30
100
1/2
2/2
0/2
1/2
1/3
1/3
1/3
3/3
0/3
3/3
2/5
3/5
3/4
4/4
0/2
2/4
0/2
1/2
4/4
(–)
2/3
2/2
3/4
2/4
1/3
2/3
3/3
2/3
1/3
1/3
0/2
0/2
0/2
0/2
0/2
2/4
0/2
0/2
1/4
(–)
0/2
0/3
0/2
0/4
0/4
0/3
0/3
0/3
0/3
0/3
0/3
0/2
0/2
0/2
0/2
0/2
0/4
0/2
0/2
0/4
(–)
0/2
0/2
0/2
0/3
0/3
0/3
0/3
0/3
0/3
0/5
0/5
0/4
0/4
0/2
0/4
0/2
0/2
0/4
(–)
1b
1c
1
1
1
1
1
1
1
1
1
1
1
1d(8)
1e
1f
1g
3a
3c
Results and discussion
Chemistry
3d
3e
3g
The synthetic route followed in this investigation is shown in
Scheme 1.
1/2
0/2
1/5
0/2
1/2
1/2
1/4
0/2
0/2
0/2
0/3
0/2
0/2
0/2
0/2
0/2
Synthesis of N-substituted 2-hydroxyisobutyl methyl amides. The
a-hydroxyamides 1(a–g) were synthesized by ammonolysis of 2-
hydroxyisobutyl methyl ester (Scheme 1). Compounds 1a, 1c, 1e
and 1g were achieved under a solvent-free environment and
microwave (MW)-assisted heating5. In these conditions, com-
pounds 1b, 1d and 1f were obtained with low yields. To improve
the synthetic method for these amides, we conducted the reac-
tions with a conventional heating system, and we used TBD as
catalyst. The selection of this compound was supported by previ-
ous studies about the synthesis of other secondary and tertiary
amides under solvent-free conditions27.
Synthesis of N-derivatives-5,5-dimethyl oxathiazolidine-4-one-2,2-
dioxides. The cyclization reaction of the resulting a-hydroxyamides
to get the corresponding sulphamates was carried out following
the conditions previously described for cyclic sulphamates
(Scheme 1)5,18–20. Initially we obtained the cyclic monoxides
through the reaction of amides with thionyl chloride and triethyl-
amine. Then, the compounds were oxidized with NaIO4 and RuCl3
to yield the final dioxide derivatives. N-isopropyl (2b and 3b) and
(1-ethyl) benzyl (2f and 3f) derivatives were not obtained in these
conditions, so they were not evaluated. The yields of the reactions
are in the range between 25% and 71% for pure compounds.
In vivo tests. We followed the standard procedures proposed
by the Epilepsy Therapy Screening Program (ETSP) of the NIH9,
which is described in the Materials and Methods section. In
Table 1, we report the results of MES test after the administration
of the compounds in terms of the fraction of animals that did not
aNumber of protected animals relative to the total number of mice tested in
MES at each time and concentration.
bNumber of animals with sedative effects relative to the total number of mice
tested at each time and concentration. Compound 3d was not tested at dose of
100 mg/kg due to solubility problems, referenced as (–). (8) See reference in
Bibliographic section.
lower doses tested (30 mg/kg). These results are encouraging since
class 1 includes the most promising candidates for the next step
of the biological testing28. In addition to the anticonvulsant activ-
ity found for the compounds, no signs of sedation and/or ataxia
were detected in the Rotorod test, which is important in terms of
the safety profile of the active structures. Compound 3d had solu-
bility problems, so it was not possible to evaluate them at the
highest dose with the recommended vehicles of the
ETSP Program6.
ED50 value was calculated for compound 1g, since it showed
strong anticonvulsant action at both doses evaluated in MES test.
Again, we followed the standard procedures for the calculation,
described in the Materials and Methods section. Essentially, ED50
measures the dose of drug that is effective in 50% of the tested
animals9. This value is calculated at the time of peak effect (TPE),
which has to be previously identified9. The final ED50 value of 1g
was 29 mg/kg, which is equivalent to 0,106 mmol/kg (TPE ¼ 0.5 h).
This result is interesting in terms of potency, since it is in the
show hind limbs tonic extension for each dose/time group. The range of ED50 values measured for classical ACDs in the same test.
For example, phenytoin showed ED50 values of 0.0218 mmol/kg;
primary toxicity of the drugs (signs of sedation and/or ataxia) was
measured by the standardized Rotorod test, also included in the
primary phase of the ETSP9.
whereas valproic acid, a representative ACD of broad spectrum,
showed an ED50 value of 1.962 mmol/kg. Also important is the
fact that all the structures passed the Rotorod test, which detects
neurotoxicity in terms of sedation or ataxia; and none of the mice
treated with the compounds died during the assays.
The compounds were administered to mice intraperitoneally at
the lower doses of the program (30, 100 mg/kg), and all the assays
were performed at 0.5 and 4 h. Table 1 includes the classification
of the structures according to the following criteria28. Class (1):
compound with anticonvulsant activity at 100 mg/kg or less, Class
(2): compound with anticonvulsant activity at doses higher than
100 mg/kg, Class (3): inactive compound at any doses up to
300 mg/kg, and Class (4): inactive compound at 300 mg/kg and
toxic at 30 mg/kg or less. The results achieved from the assays
showed that all the structures were active against MES test, which
allowed us to classify them as class 1. In fact, most of the struc-
In silico studies
To explore one of the possible mechanisms of action of the syn-
thesized compounds we analyse them as sodium channel block-
ers. The blockage of VGSC is a validated mechanism of action of
many ACD with probed clinical efficacy7,11,12. Among the four dif-
tures exhibited protection against the induced seizure at the ferent VGSCs isoforms recognized in CNS (Nav1.1, Nav1.2, Nav1.3

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1469
and Nav1.6), Nav1.2 is the molecular target with confirmed inter- with ACD. A docking score of ꢂ8.1 kcal/mol was defined as the
action with many ACDs11,12
threshold value to differentiate active from inactive compounds.
.
Initially, the molecules were docked into the Nav1.2 channel This cut-off value was previously selected for the screening of a
with Autodock Vina software26. As the 3D structure of human virtual database, since it shows a good balance between specifi-
Nav1.2 is not available, we employed a 3D model of this macro- city and sensibility for the test set that validated the docking
molecule previously constructed in our laboratory24. It includes model (80% and 88%, respectively)24. Table 2 shows the docking
the 3D architecture of the a-subunit of the Nav1.2 channel, which scores of the synthesized compounds.
is functional on its own and comprises the region of interactions
Our docking results suggest that the amides are poorer candi-
dates to block the Nav1.2 channel since none of them were able
to pass the threshold value defined for active compounds accord-
ing to the docking classificatory model (ꢂ8.1 Kcal/mol). Among
cyclic structures, compounds 3e and 3g were predicted as active.
It is worth mentioning that the docking protocol has been already
validated and used to identify new Nav1.2 blockers in a previous
investigation24. Figure 2(A) shows the binding geometries of struc-
tures 3e and 3g into the docking active site, which includes resi-
dues of the ion conducting pore of the channel that interact with
Table 2. Docking scores of the synthesized compounds.
Compound
Score
Compound
Score
1a
1b
1c
1d
1e
1f
ꢂ6.4
ꢂ5.5
ꢂ6.3
ꢂ6.9
ꢂ7.7
ꢂ7.8
1g
3a
3c
3d
3e
3g
ꢂ7.5
ꢂ6.7
ꢂ6.5
ꢂ7.2
ꢂ8.2
ꢂ8.3
Figure 2. (A): Binding geometry of compounds 3e and 3g into the Nav1.2 active site predicted by the Autodock Vina docking algorithms. (B) binding geometry pre-
dicted for phenytoin. Carbon atoms are highlighted in pink, yellow and green for compounds 3e, 3g and phenytoin, respectively. Nitrogen, oxygen and fluorine atoms
of the compounds are highlighted in blue, red and light blue, respectively. Residues that surround the ligand within 5 angstroms of distance are highlighted as sticks.
Important aromatic residues are numbered in the figure. Hydrogen atoms are omitted for simplicity except for the non-aromatic hydrogen atoms of phenytoin.
Figure 3. Whole cell configuration with a voltage-clamp step protocol from ꢂ80 to ꢂ20 mV. (A) Typical traces of Na þ currents in Nav 1.2 isoform, control and com-
pound 1e or 1g, 100mM. (B) p > 0.05. No statistical differences were found for both compounds (n ¼ 4).

1470
L. L. SABATIER ET AL.
known ACDs25. The simulations suggest that the compounds dir-
ect their aromatic ring towards the region delimited by aromatic evaluated in the Naþ currents carried through the Nav1.2 channel
The in vitro activity of compounds 1e, 1g, 3e and 3g was
side chains of Phe1462 and Phe1754. In the docking conditions,
the ACD phenytoin orientates its heterocycle and one of its aro-
matic rings to the same regions occupied by 3e and 3g, with a
docking score of ꢂ9.9 kcal/mmol (Figure 2(B))26. The overall orien-
tation of 3e, 3g and phenytoin into the Nav1.2 channel is given
as Supplemental online material (Figure S1).
isoform stably expressed in HEK 293 cells by using the whole-cell
configuration. The current stability, before and after perfusion of
the tested compounds, was monitored every 10 s by a 10 ms volt-
age step to ꢂ20 mV from
a
holding potential of ꢂ80 mV.
The inhibitory activity was observed as decay in the peak
amplitude of the Naþ currents. After current stabilization, and
with the aim to evaluate the state-dependent inhibition, we tested
the ability of these compounds to modify the steady-state inacti-
vation curves (h curves) of the Nav1.2 channels. The h curve
shows the available fraction of Na þ channels in condition to be
activated (in resting state) as a function of the resting membrane
potential (see Material and Methods section). The state-dependent
inhibition, due to the stabilization of the inactivated state, is
observed as a left-shift in the h curves, which means a reduction
of available Naþ channel which is higher at more depolar-
ized potential.
Electrophysiology
The virtual strategy proposed only two compounds, 3e and 3g, as
Nav1.2 blockers. Both structures showed anticonvulsant action in
mice at the lower dose tested. On the other hand, the rest of
compounds (which were predicted as inactive in the Nav1.2
model) showed protection against MES test.
To analyse if the contrast between the in silico predictions and
the in vivo assays can be explained as a fail in the sensibility of
the docking protocol, we selected two pairs of compounds
(1e–3e and 1g–3g) to test their effect in Nav1.2 currents using
the patch-clamp technique. The rationale for compounds selection
was to test the two heterocycles predicted as active, and their cor-
Amides 1e and 1g were not able to block Naþ current (Figure
3(A,B)) and did not change the steady state inactivation curves
(Figure 4(A,B,C)), while the corresponding cyclic compounds 3e
and 3 g inhibited Na þ current (Figure 5(A,B)) and shifted the h
responding a-hydroxyamides, which were classified as not active curve to more hyperpolarized potentials, (Figure 6 (A,B,C)). These
in the docking simulations (Table 2).
results are consistent with the docking model predictions.
Figure 4. Compounds 1e and 1g did not affect the Nav1.2 channel steady state inactivation curves. (A) Typical traces of the steady-state inactivation protocol, where
Nav1.2 currents were activated by a depolarizing voltage step to ꢂ10 mV for 15 ms following 2500 ms hyperpolarizing pre-pulses ranging from ꢂ130 to ꢂ40 mV in
the absence and in the presence of 100 m of 1e and 1g. (B) Mean available fraction (I/Imax) values for compound 1e or 1g and their respect control plotted as func-
tion of membrane voltage (mV) and fitted to the Boltzmann function. (C) Mean Vh values obtained from plots showed in (B). No statically difference was observed
(p > 0.05, n ¼ 4 for each group). Vhcontrol (1e) ¼ ꢂ62.77 2.5, Vh(1e) ¼ ꢂ60.63 1.40, n ¼ 4. Vhcontrol (1g) ¼ ꢂ58.09 0.43, Vh (1 g) ¼ ꢂ59.96 1.19, n ¼ 4.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1471
Figure 5. Compounds 3e and 3g inhibited the Nav1.2 current. (A) Typical traces of Nav1.2 currents recorded in HEK 293 cells in the whole cell configuration with a
voltage-clamp step protocol from ꢂ80 to ꢂ20 mV before and after 100 mM 3e (up) and 3g (down) extracellular perfusion. (B) Mean Nav1.2 currents values obtained
from current showed in (A). p < 0.05 indicates significant difference between control and compound 3e (%Current Inhibition ¼ 29.6%); ##p < 0.005 indicates signifi-
ꢃ
cant differences between control and compound 3g group. (%Current Inhibition ¼ 33%). Paired t–test n ¼ 4, for each group.
Figure 6. Compounds 3e and 3g produced a state-dependent inhibition in Nav1.2 channel. (A) Typical traces of the steady-state inactivation protocol where Nav1.2
currents were activated by a depolarizing voltage step to ꢂ10 mV for 15ms following 2500 ms hyperpolarizing pre-pulses ranging from ꢂ130 to ꢂ40 mV in the
absence and in the presence of 100 m of 3e and 3g. (B) Mean available fraction (I/Imax) values for compound 3e or 3g and their respect control plotted as function of
ꢃꢃ
membrane voltage (mV) and fitted to the Boltzmann function (Equation (1)). (C) Vh values obtained from plots showed in (B). p < 0.05. , ### Significant differences
between 3e, 3g and their respective control groups, Vhcontrol (3e) ¼ ꢂ60.18 1.68, Vh(e) ¼ ꢂ68.19 2.19, n ¼ 5, Vhcontrol (3 g) =-74.76 1.26 Vh 3 g¼
ꢂ80.14 3.08, n ¼ 4.

1472
L. L. SABATIER ET AL.
As mentioned before, the electrophysiological assays confirmed channel blocking effect of these drugs is greater in neurons
our docking predictions, showing that the oxathiazolidine-4-one-
2,2-dioxide derivatives 3e and 3g inhibit Na þ currents by
left-shifting the steady-state inactivation curve, suggesting the
interaction of these compounds with Nav1.2 channel.
involved in the seizure, reducing the probability of side effect
induced by the channel block in normal neurons36. Regarding
future investigations on VGSC blockers, more studies will be com-
pleted in the future for N-derivatives of 5,5-dimethyl-1,2,3-oxa-
thiazolidine-4-one-2,2-dioxide and their synthetic precursors, to
propose their possible mechanisms of action and to optimize
their anticonvulsant action. Finally, the discovery of new VGSC
blockers will allow us to evaluate them in other biological assays
related with CNS pathologies where the VGSCs are also impli-
cated. Good examples would be depression and anxiety, two
pathologies well recognized as comorbid with epilepsy.
Conclusions
In this investigation, we synthesized and studied the anticonvul-
sant action of new a-hydroxyimides and N-derivatives-1,2,3-oxa-
thiazolidine-4-one-2,2-dioxides, which scaffold is a bioisosteric
partner of the known ACD phenytoin. The compounds were
tested in mice and they evidenced a promising anticonvulsant
activity as well a good safety profile at this pre-clinical stage.
Particularly, structure 1g shows an ED50 comparable with clas-
sical ACDs.
Acknowledgements
L. Gavernet, P. Martin and V. Pastore are members of Consejo
Docking simulations suggested the interaction of two com-
pounds of the set with the Nav1.2 target and patch clamp experi-
ments were performed to corroborate the in silico predictions.
These two structures are N-aromatic derivatives of 1,2,3-oxathiazo-
lidine-4-one-2,2-dioxides. Conversely, N-aliphatic-derivatives and
all the a-hydroxyamides were classified as non-active. However,
they showed anticonvulsant action in mice. The differences
observed between in vitro (and in silico) assays and in vivo experi-
ments can be generated by multiple factors. First, post-transcrip-
tional modifications such as alternative splicing, channel protein
phosphorylation and/or glycosylation are capable to modify the
pharmacological sensitivity of ion channels. Moreover, the activity
of the channel blockers can be modified by the expression of aux-
iliary beta subunits and co-localization with other membrane pro-
teins in native cells29. A second factor is related with the
selectivity of the compounds, since most ACDs can elicit the antie-
pileptic effect by acting in multiple targets. They are nonselective
among neuronal Nav channels isoforms, and they can modify
other isoforms (such Nav 1.1 and 1.6), other ions channels (such
as Kþ and Ca2þ channels), ionotropic receptors (like GABA-A and
NMDA) and enzymes (GABA transaminase and glutamic acid
decarboxylase)7,30. Finally, the drug metabolism can produce
active metabolites which were not tested in the docking protocol
and in vitro experiments31.
Our electrophysiological experiments allowed us to explore
the inhibitory mechanism of action of 3e and 3g. Both com-
pounds showed inhibitory activity on Nav 1.2 isoform in the two
protocols tested. Interestingly, the left shift produced in the h
curve indicates that these compounds can stabilize the steady-
state inactivation of the channel. This state-dependent mechan-
ism in Nav channels is shared with ACDs such as phenytoin,
carbamazepine and lamotrigine, among others32,33. Nav channels
are voltage dependent and have at least three well recognized
states34,35. When the cell membrane potential is hyperpolarized,
the channel is in the resting closed state, then when the mem-
brane potential rise, the channel turns to an open state and sub-
sequently fall into a not conducting inactive state. The Nav
channels are not able to open from this inactivated state and
hyperpolarization is needed to return the channels to the resting
state. In this way, the membrane potential value determines the
fraction of Nav channels that are able to open after a depolariz-
ing stimulus and trigger an action potential. So, drugs that stabil-
ize the inactivated state show a higher inhibitory effect in cells
where the membrane is more depolarized (are voltage-depend-
ent). This is particularly interesting since, in epilepsy, the neurons
responsible for seizure are hyper-excitable and this is, in part,
due to a more positive membrane potential. Thus, the Nav
ꢀ
ꢀ
ꢀ
Nacional de Investigaciones Cientıficas y Tecnicas de la Republica
Argentina (CONICET). P. Palestro, L. Sabatier, and A. Enrique are
fellowship holders of CONICET.
Disclosure statement
No potential conflict of interest was reported by the authors.
Funding
ꢀ
ꢀ
This work was supported by the Agencia de Promocion Cientıfica
ꢀ
Tecnologica under Grants [PICT 3175–2013] and [PICT
y
0165–2016], CONICET, and Universidad Nacional de La Plata,
Argentina. We gratefully acknowledge the support of NVIDIA
Corporation with the donation of the Quadro M5000 GPU used
for this research.
References
1. Prasad LK, Smyth H. 3D Printing technologies for drug deliv-
ery: a review. Drug Dev Ind Pharm 2016;42:1019–31.
2. Sirven JI, Noe K, Hoerth M, Drazkowski J. Antiepileptic drugs
2012: recent advances and trends. Mayo Clin Proc 2012;87:
879–89.
3. Bialer M, Johannessen SI, Levy RH, et al. Progress report on
new antiepileptic drugs: a summary of the Thirteenth Eilat
Conference on new antiepileptic drugs and devices (EILAT
XIII). Epilepsia 2017;58:181–221.
4. W.H.O. Epilepsy. Fact sheet. No. 999; 2015.
5. Pastore V, Sabatier L, Enrique AV, et al. Synthesis and
anticonvulsant activity of bioisosteres of trimethadione,
N-derivative-1,2,3-oxathiazolidine-4-one-2,2-dioxides from a-
hydroxyamides. Bioorg Med Chem 2013;21:841–6.
6. Putnam TJ, Merritt HH. Experimental determination of the
anticonvulsant properties of some phenyl derivatives.
Science 1937;85:525–6.
7. Meldrum BS, Rogawski MA. Molecular targets for antiepilep-
tic drug development. Neurotherapeutics 2007;4:18–61.
8. Abou-Khalil BW. Update on antiepileptic drugs 2019.
Continuum (Minneap Minn) 2019;25:508–36.
9. Kehne JH, Klein BD, Raeissi S, Sharma S. The National
Institute of Neurological Disorders and Stroke (NINDS)
Epilepsy Therapy Screening Program (ETSP). Neurochem. Res
2017;42:1894–903.

JOURNAL OF ENZYME INHIBITION AND MEDICINAL CHEMISTRY
1473
10. Pastore V, Wasowski C, Higgs J, et al. A synthetic bioisoster
of trimethadione and phenytoin elicits anticonvulsant effect,
protects the brain oxidative damage produced by seizures
and exerts antidepressant action in mice. Eur
Neuropsychopharmacol 2014;24:1405–14.
11. Vohora D, Saraogi P, Yazdani M, et al. Recent advances in
adjunctive therapy for epilepsy: focus on sodium channel
blockers as third-generation antiepileptic drugs. Drugs
Today 2010;46:265–77.
12. Brodie MJ. Antiepileptic drug therapy the story so far.
Seizure 2010;19:650–5.
13. Banfi L, Riva R. The passerini reaction. Org React 2005;65:
1–140.
14. Saegusa T, Taka-ishi N, Ito Y. The thermal rearrangement
and degradation of 2,3-Bis(alkylimino)oxetane. Bull Chem
Soc Japan 1971;44:1121–5.
15. Hagedorn I, Eholzer U. Single-step synthesis of a-hydroxy
acid amides by a modification of the Passerini reaction.
Chemische Berichte 1965;98:936–40.
16. Hagedorn I, Eholze U. N-Monosubstituted a-hydroxycarboxa-
mides. DE 1964;1173082:19640702.
17. Shapiro L, Rose IM, Testa FC, et al. N-Substituted oxazolidi-
for the discovery of anticonvulsant agents. J Chem Inf
Model 2018;58:1331–42.
25. Boiteux C, Vorobyov I, French RJ, et al. Local anesthetic and
antiepileptic drug access and binding to a bacterial voltage-
gated sodium channel. Proc Natl Acad Sci USA 2014;111:
13057–62.
26. Trott O, Olson AJ. AutoDock Vina: improving the speed and
accuracy of docking with a new scoring function, efficient
optimization, and multithreading. J Comp Chem 2010;31:
455–61.
27. Sabot KC, Kumar A, Meunier S, Mioskowski C. A convenient
aminolysis of esters catalyzed by 1,5,7-triazabicyclo[4.4.0]-
dec-5-ene (TBD) under solvent-free conditions. Tetrahedron
Lett 2007;48:3863–6.
ꢀ
28. Malawska B, Kulig K, Spiewak A, Stables JP. Investigation
into new anticonvulsant derivatives of alpha-substituted
N-benzylamides of gamma-hydroxy- and gamma-acetoxybu-
tyric acid. Part 5: search for new anticonvulsant compounds.
Bioorg Med Chem 2004;12:625–32.
29. Uebachs M, Albus C, Opitz T, et al. Loss of b1 accessory
Na þ channel subunits cause failure of carbamazepine, but
not of lacosamide, in blocking high-frequency firing via dif-
ferential effects on persistent Na þ currents. Epilepsia 2012;
53:1959–67.
30. Younus I, Reddy DS. A resurging boom in new drugs for epi-
lepsy and brain disorders. Expert Rev Clin Pharmacol 2018;
11:27–45.
nediones. J Am Chem Soc 1959;81:6498–504.
ꢀ
18. Melendez RE, Lubell WD. Synthesis and reactivity of cyclic sul-
famidites and sulfamidates. Tetrahedron 2003;59:2581–616.
19. Posakony JJ, Grierson JR, Tewson TJ. New routes to N-alky-
lated cyclic sulfamidates. J Org Chem 2002;67:5164–9.
20. White GJ, Garst ME. Cyclic sulfamate from N-substituted
2-amino-3-phenyl-1-propanol and its nucleophilic reactions.
J Org Chem 1991;56:3177–8.
21. Jones BJ, Roberts DJ. The quantitative measurement of
motor incoordination in naive mice using an accelerating
rotarod. J Pharm Pharmacol 1968;20:302–4.
22. Litchfield JT, Wilcoxon F. A simplified method of evaluating
dose-effect experiments. J Pharmacol Exp Ther 1949;96:
99–113.
23. Hamill O, Marty A, Neher E, et al. Improved patch-clamp
techniques for high-resolution current recording from cells
and cell-free membrane patches. Pflugers Arch 1981;391:
85–100.
31. LaPenna P, Tormoehlen LM. The pharmacology and toxicol-
ogy of third-generation anticonvulsant drugs. J Med Toxicol
2017;13:329–42.
€
32. Rogawski MA, Loscher W. The neurobiology of antiepileptic
drugs. Nat Rev Neurosci 2004;5:553–64.
33. Nardi A, Damann N, Hertrampf T, Kless A. Advances in tar-
geting voltage-gated sodium channels with small molecules.
ChemMedChem 2012;7:1712–40.
34. Denac H, Mevissen M, Scholtysik G. Structure, function and
pharmacology of voltage-gated sodium channels. Naunyn-
Schmiedeberg’s Arch Pharmacol 2000;362:453–79.
35. Bertrand D, Biton B, Licher T, et al. Functional studies of
sodium channels: from target to compound identification.
Curr Protoc Pharmacol 2016;7:9.21.1–.21.35.
24. Palestro P, Enrique AV, Goicoechea S, et al. Searching for
new leads to treat epilepsy: target-based virtual screening
36. Goldin AL. Mechanisms of sodium channel inactivation. Curr
Opin Neurobiol 2003;13:284–90.