Investigation into new anticonvulsant derivatives of α-substituted N-benzylamides of γ-hydroxy- and γ-acetoxybutyric acid. Part 5: Search for new anticonvulsant compounds

Article abstract of DOI:10.1016/j.bmc.2003.10.036

A series of four N-benzylamides of γ-hydroxybutyric acid (GHB), that contain N-(4-phenylpiperazine)-, N-(4-benzylpiperazine)rings, N-benzylamino-, or N-(2-phenylethylamine)-groups in the α-position of GHB were selected as model compounds, for determining the structural elements responsible for their potential anticonvulsant action. Based on the results of pharmacological, physicochemical, and molecular modelling investigations, the pharmacophore model for anticonvulsant N-substituted amides of GHB was defined. In this model, the presence of the N-benzylamide fragment is essential for activity. In addition, all of the amides contained another hydrophobic unit (aryl ring) as a distal binding site and H-bond donor. In consideration of these model parameters, a number of N-substituted amides of GHB, containing a hydrophobic moiety such as: N-benzylamino or N-(4-chlorobenzylamino) group in the α-position of GHB, and a lipophilic substituent in the amide portion, were prepared. It has been shown that the anticonvulsant activities of the newly synthesized compounds might partially be explained on the basis of their lipophilicity (calculated log P values) and the presence of a hydroxyl group in the molecule.

Full text of DOI:10.1016/j.bmc.2003.10.036

Bioorganic & Medicinal Chemistry 12 (2004) 625–632
Investigation into new anticonvulsant derivatives of ꢀ-substituted
N-benzylamides of ꢁ-hydroxy- and ꢁ-acetoxybutyric acid.
Part 5: Search for new anticonvulsant compounds
Barbara Malawska,a,* Katarzyna Kulig,a Agnieszka Spiewaka and James P. Stablesb
a
´
Department of Pharmaceutical Chemistry, Medical College of Jagiellonian University, Medyczna 9, 30-688 Krakow, Poland
bEpilepsy Branch, National Institute of Neurological Disorders and Stroke, National Institute of Health,
Federal Building, Room 114, Bethesda, MD 200892-9020, USA
Received 8 April 2003; accepted 17 October 2003
Abstract—A series of four N-benzylamides of g-hydroxybutyric acid (GHB), that contain N-(4-phenylpiperazine)-, N-(4-benzylpi-
perazine)rings, N-benzylamino-, or N-(2-phenylethylamine)-groups in the a-position of GHB were selected as model compounds, for
determining the structural elements responsible for their potential anticonvulsant action. Based on the results of pharmacological,
physicochemical, and molecular modelling investigations, the pharmacophore model for anticonvulsant N-substituted amides of
GHB was defined. In this model, the presence of the N-benzylamide fragment is essential for activity. In addition, all of the amides
contained another hydrophobic unit (aryl ring) as a distal binding site and H-bond donor. In consideration of these model
parameters, a number of N-substituted amides of GHB, containing a hydrophobic moiety such as: N-benzylamino or N-(4-chlor-
obenzylamino) group in the a-position of GHB, and a lipophilic substituent in the amide portion, were prepared. It has been shown
that the anticonvulsant activities of the newly synthesized compounds might partially be explained on the basis of their lipophilicity
(calculated log P values) and the presence of a hydroxyl group in the molecule.
# 2003 Elsevier Ltd. All rights reserved.
1. Introduction
not been directly linked with any specific binding site
within the brain; therefore, drug identification is
In recent years, several new antiepileptic drugs (AEDs)
such as lamotrigine, oxcarbazepine, felbamate, gaba-
pentin, topiramate, fosphenytoin sodium, tiagabine,
zonisamide and levetiracetam have been approved. Others
like losigamone, pregabaline, remacemide, rufinamide,
retigabine, ganaxolone, soretolide, stiripentol, milace-
mide, valrocemide, carabersat, harkoseride and talam-
panel have been or are currently in human studies.1ꢀ4
Some of the newer AEDs represent structural modi-
fications of pre-existing compounds, others were
developed with the specific objective of modifying
targets such as neurotransmitter function. Some of the
mechanisms of AED action include: potentiation of
GABA-ergic transmission, blockade of voltage-dependent
sodium channels, attenuation of excitatory neuro-
transmission, and/or modulation of voltage-sensitive
calcium channels. Some of the clinically used drugs have
partially conducted via in vivo screening tests using
whole animal systems. The chemical diversity of
compounds, along with multiple mechanism of action of
effective medications, makes it difficult to identify any
one common pharmacophore responsible for prevention
or arrest of seizure activity in all patients.5ꢀ8 This infers
that control of unstable electrical environments in the
brain resulting in seizure activity involves a host of
physiological processes and may require multiple
mechanisms to achieve cessation.
In our search for new anticonvulsants, a study of
GABA and g-hydroxybutyric acid (GHB) analogues
was initiated. Derivatives of a-substituted g-amino-, g-
phthalimido-, and g-hydroxy butyric acid such as: acids,
esters and amides were investigated.9ꢀ13 It was shown
that a-substituted N-benzylamides of g-hydroxybutyric
acid were the most potent compounds and possessed
anticonvulsant activities in the maximal electroshock
(MES) (ip) screens. The most potent anticonvulsant
compounds were a-(benzylamino) - g - hydroxybutyric
Keywords: Anticonvulsant activity; N-Benzylamides of g-hydroxy- or
g-acetoxybutyric acid; Molecular modelling; SAR.
* Corresponding author. E-mail: mfmalaws@cyf-kr.edu.pl
0968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.10.036

626
B. Malawska et al. / Bioorg. Med. Chem. 12 (2004) 625–632
acid N-benzylamide 1a and N-(2-chlorobenzylamide)
1b (Fig. 1), their median effective dose (ED50) were
respectively 63.0 and 54.0 mg/kg (MES).
In the MES Screen, these compounds were less active
than the commonly used anticonvulsants: carbamaze-
pine and phenytoin, but possess higher activity than
sodium valproate. The biochemical tests indicate that
the active amides act as allosteric modulators of the g-
aminobutyric acid, GABAA complex, and have an
affinity to voltage sensitive calcium channels (VSCC)
receptors. The N-(4-methylbenzyl)amide of a-(4-phe-
nylpiperazin-1-yl)-g-hydroxybutyric acid 12 displaced
the binding of [35S] TBPS ([35S] tert-buthylbicyclophos-
phorothionate) to the chloride channel of the GABAA
receptor complex (IC50=95 mM).14,15 This may be the
possible mechanism mediating the anticonvulsant effect
of these compounds.
The aim of our study was determination of the structural
elements and/or physicochemical parameters necessary
for anticonvulsant activity. GHB derivatives were
investigated, and based on these findings a series of new
compounds were synthesized.
2. Investigations
2.1. Results and discussion
As a part of our search for a potential anticonvulsant
agent, we have explored compounds structurally related
to GHB. A series of four N-benzylamides of GHB (Fig.
2) that contain the N-(4-phenylpiperazine)- (series A),
N-(4-benzylpiperazine)- (series B), N-benzylamino-(ser-
ies C), and N-(2-phenylethylamine)- (series D), all
grouped in the a-position of GHB were selected as
models to determine the structural elements and/or
physicochemical properties responsible for their
anticonvulsant action.
Scheme 1. Reagents and conditions: (i) relevant substituted
benzylamine, toluene, reflux, 12 h. *The classifications are as follows;
1: anticonvulsant activity at 100 mg/kg or less; 2: anticonvulsant
activity at doses greater than 100 mg/kg; 3: compound inactive at
300 mg/kg; 4: compound inactive at 300 mg/kg and toxic at 30 mg/kg
or less.
common to all evaluated molecules. Their similarity was
calculated as RMS fit. The RMS routine provided
estimates of how closely molecules fit to each other. The
lower the RMS value, the better similarity observed.
The RMS deviations for each series are as follows:
0.330 A (series A), 0.328 A (series B), 0.197 A (series C)
(Fig. 3), and 0.599 A (series D).
To study the flexibility of synthesized compounds,
molecular modelling investigations and their conforma-
tional analysis were performed. The conformational
analyses were carried out by systematic stepwise rota-
tion of 10ꢁ four torsion angles marked in Figure 2 as
f1–f4, common to all molecules. Energy criterion, set to
3 kcal/mol above the lowest energy conformation found,
was used to accept new conformation. Although this
approach explores only a part of all possible conforma-
tions, the results give an estimate of the range of energy
changes in geometry of isolated molecules. The calculated
structures were optimized and compared within a
particular series, taking into account five atoms
In order to obtain more general information concerning
the shape of investigated molecules, a representative
compound from each series (R=H) was selected to
superimpose. Two nitrogen atoms and the first carbon
atom from the hydroxyalkyl chain were selected for the
fitting procedures. The RMS deviations for those four
compounds were 0.087 A (Fig. 4). The molecular
Figure 1. Anticonvulsant active a-(benzylamino)-g-hydroxybutyric acid N-benzylamides.

B. Malawska et al. / Bioorg. Med. Chem. 12 (2004) 625–632
627
modelling investigations indicated that features of
molecules from each series were similar. Comparison of
the representative compounds (Fig. 4) showed similarity
in the orientation of both the GHB chain and the
benzylamide moiety, while a substituent in the a-position
of GHB possessed a different orientation. These results
indicate that the N-benzylamide moiety may be an
important structural unit for their activity.
activity. Introduction of the acetyl substituent in the
hydroxyl group allowed us to examine the role of a
hydroxyl group in the proposed model.
The N-substituted amides of a-(benzylamino)- or a-(4-
chlorobenzylamino)-GHB (3–11) were synthesized
according to the procedures shown in Scheme 1. 3-Sub-
stituted derivatives of tetrahydrofuran-2-one were
synthesized from a-bromo-g-butyrolactone and benzyl-
or 4-chlorobenzylamine by method which was
previously described.9 The aminolysis of lactones with
various 2- or 4-substituted benzylamine yielded target
compounds (3–11).
Based on the results of pharmacological, physicochem-
ical, roentgenostructural and molecular modelling
investigations
a
pharmacophore-model for anti-
convulsant activity of N-substituted amides of GHB
was defined. In this model, the presence of the N-
benzylamide fragment is essential for that activity. All
active amides should contain a hydrophobic unit (aryl
ring) as a distal binding site and a group which could
act as a H-bond donor (Fig. 5).
Acylation of the active anticonvulsants, derivatives of
a-(4-phenylpiperazine)-GHB with acetic anhydride led
to g-acetoxy-GHB derivatives (13–16) (Scheme 2).
Preliminary anticonvulsant evaluation (phase I) of all
the synthesized compounds 3–11, 13–16 was provided
by testing procedures which have been described
earlier.16 Phase I studies of the investigated compounds
involved three tests: maximal electroshock seizure
(MES), subcutaneous metrazole (scM), and rotorod test
for neurological toxicity. Phase I is a qualitative assay
involving a small number of mice (total 16). For the
MES assay, experimental compound is administered to
one, three and one animals at three respective doses (30,
100, and 300 mg/kg) at both 0.5- and 4-h time periods.
The same similar paradigm for dosing and time is used
For verification of this model, a number of N-sub-
stituted amides of GHB, containing a hydrophobic
moiety such as N-benzylamino or N-(4-chloro-
benzylamino) group in the a-position of GHB, and
lipophylic substituent in the amide group, were synthe-
sized. In our research we were particularly interested in
the influence of arylalkylamino substituent in the a
position and lipophylic substituent such as methyl-,
trifluoromethyl-, or chloro- in the ortho-position in the
phenyl ring in benzylamide fragment on anticonvulsant
Figure 2. Schematic structure of N-benzylamides derivatives of
a-substituted GHB.
Figure 4. Superimposition of the four anticonvulsant active
representative compounds of each series.
Figure 3. Superimposition of the seven anticonvulsant active
compounds of series C.
Figure 5. Pharmacophore model for anticonvulsant of a-substituted
N-benzylamides of g-hydroxybutyric acid.

628
B. Malawska et al. / Bioorg. Med. Chem. 12 (2004) 625–632
for the scM testing. All animals were observed for
potential neurological deficit at each dose and time
period.
administering drug, however, toxicity was also observed
at these doses. The compounds 9 and 10 that displayed
a marked MES activity in mice, were evaluated for oral
activity in rats (Table 2). Amides 9 and 10 demonstrated
only weak activity in the MES screen at 30 mg/kg (po
dose) (amide 9: 1=4 animals protected at 0.5 and 2 h;
amide 10: 1=4 animals protected both at 0.25 and 4 h).
Amides 9 and 10 used in the tests in rats were not toxic
at the doses used. Compounds 4, 11, 14–16 were inac-
tive in the MES and scMet mouse screens at these doses
and time points. The investigated N-benzylamides of 2-
(4-phenylpiperazine) - g - acetoxy-GHB (14–16) were
inactive and amide 17 was marginally active in the MES
screen in mice (not shown). Compounds in which the
hydroxyl group was replaced by acetoxy substituent
were not active 13–16 (not shown). The results obtained
showed that the presence of a hydroxy group in the
molecule was necessary for MES activity.
According to the results compounds were classified into
four classes. The classification is as follows; (1) anti-
convulsant activity at 100 mg/kg or less; (2) anti-
convulsant activity at doses higher than 100 mg/kg; (3)
compound inactive at any dose up to 300 mg/kg; (4)
compound inactive at 300 mg/kg and toxic at 30 mg/kg
or less. For the identification of anticonvulsant activity
in mice, test compounds were administered intraperi-
toneally and challenged by maximal electroshock and
subcutaneous metrazol. Compounds found to be
effective in these seizure challenges are generally regarded
to be potentially useful candidates in treatment of
partial, generalized and even absence seizures. Neuro-
toxicity of the test compounds was determined using the
rotorod toxicity test (TOX). The results of in the vivo
tests are summarized in Table 1.
Taking into account that physicochemical properties of
chemical compounds have influence on their pharma-
cological properties, the partition coefficients (clog P) of
the amides 3–16 were calculated17 (Schemes 1 and 2).
The predicted clog P values were in the range 1.58–3.47
for amides 3–16, and for most compounds this value
was near 2. Among the investigated compounds, amides
5 and 7 displayed the highest lipophilicity.
N-Substituted amides of a-(benzylamino)- and a-(4-
chlorobenzylamino)-GHB showed some protection
against MES seizures in mice. The most potent anti-
convulsant compounds were 5, 7, 9 and 10. Compounds
3, 9, 10 and 13 were active at doses of 100 mg//kg
(1=3; 2=3; 3=3 and 1=3 animals protected) 0.25 h after
administration, inactive at later time points. The amides
3, 5, 7–10 and 13 were active at 300 mg/kg 0.5 h after
Table 1. Anticonvulsant screening project (ASP): phase 1—test
results in mice
Compd Dose (mg/kg) Activity MESa time (h)
TOXb time (h)
0.25 0.5
1/4 — 0/2
0.25
0.5
1
4
1
4
3
30
100
300
—
1/3
—
0/1
0/3
1/1
—
0/3
—
0/1
—
0/3 0/3 5/8 0/3 1/4
0/1
—
4/4* — 1/2
5
30
100
300
—
—
—
0/1
1/3
1/1
—
—
—
0/1
0/3
0/1
—
—
—
0/4 — 0/2
2/8 — 0/4
2/4 — 0/2
6
30
100
300
—
—
—
0/1
1/6
0/4
—
—
—
0/1
0/3
0/1
—
—
—
1/4 — 0/2
3/8 — 0/4
3/4 — 1/2
7
30
100
300
—
—
—
0/1
1/3
1/1
—
—
—
0/1
0/3
1/1
—
—
—
0/4 — 0/2
3/8 — 0/4
4/4* — 0/2
8
30
100
300
—
0/3
—
0/1
0/3
1/1
—
0/3
—
0/1
0/3 0/3 1/8 0/3 0/4
—
0/4 — 0/2
0/1
—
2/4* — 0/2
9
30
100
300
—
2/3
—
0/1
0/3
1/1
—
0/3
—
0/1
0/3 0/3 0/8 0/3 0/4
—
0/4 — 0/2
0/1
—
2/4 — 0/2
10
13
30
100
300
—
3/3
—
0/1
0/3
1/1
—
0/3
—
0/1
0/3 0/3 2/8 0/3 0/4
—
0/4 — 0/2
0/1
—
4/4* — 1/2
30
100
300
—
1/3
—
0/1
0/3
1/1
—
0/3
—
0/1
—
0/4 — 0/2
0/3 0/3 0/8 0/3 0/4
0/1 4/4* — 0/2
—
a Maximal electroshock test (number of animal protected/number of
animals tested).
b Rotorod toxicity (number of animals exhibiting toxicity/number of
animals tested) endpoint: unable to maintain balance on rotorod.
Scheme 2. Reagents and conditions: (i) (CH3CO)2O, pyridine, 90 ꢁC,
4 h. *Classification as for Scheme 1.

B. Malawska et al. / Bioorg. Med. Chem. 12 (2004) 625–632
629
Table 2. Anticonvulsant screening project; test results in rats; po
identification (dose 30 mg/kg)
was performed on silica gel plates (5ꢂ10 cm, 0.25 mm)
Kiselgel 60 F254 (Merck) using solvent systems: S1,
chloroform/acetone (1:1); S2, chloroform/methanol/
acetic acid (60:10:5); S3, methanol/ammonium (25%)
Compd.
Test
Time (h)
1.00
(98:2) with visualisation by UV light. 1H NMR and 13
C
0.25
0.5
2.00
4.00
NMR spectra were recorded on a Gemini spectrometer
at 200 MHz in CDCl3 with tetramethylsilane as an
internal standard. The mass spectra at 70 eV were taken
with an LKB 2091 GCMS spectrometer.
9
9
10
10
MESa
TOXb
MESa
TOXb
0/4
0/4
1/4
0/4
1/4
0/4
0/4
0/4
0/4
0/4
0/4
0/4
1/4
0/4
0/4
0/4
0/4
0/4
1/4
0/4
a Maximal electroshock test (number of animal protected/number of
animals tested).
4.1.1. 3-(4 - Chlorobenzylamino) - tetrahydrofuran-2-one
hydrochloride (2). To a solution of 4-chlorobenzylamine
(0.10 mol, 14.1 g, 12.2 mL) in acetonitrile (170 mL)
anhydrous K2CO3 (0.2 mol, 27.6 g) and tetrabutyl-
ammonium bromide (0.01 mol, 3.2 g) were added. The
mixture was stirred at room temperature for 0.5 h. Then
a solution of 3-bromobutyrolactone (0.1 mol, 16.5 g,
8.2 mL) in acetonitrile (30 mL), the stirring was
continued for 48 h. The resultant mixture was filtered
and the solvent evaporated. The oily residue was
dissolved in anhydrous methanol and acidified with
25% hydrochloric acid. The solvent was evaporated, the
product was washed with anhydrous methanol and
acetone. The obtained salt was dried in vacuo.
b Rotorod toxicity (number of animals exhibiting toxicity/number of
animals tested) endpoint: unable to maintain balance on rotorod.
The most active compounds 5 and 7 were derivatives
with 4-chlorobenzyl substituent in the a position of
GHB and the 2-chloro- or 2-trifluoromethyl- substituent
in the benzylamide fragment. These amides had the
highest lipophilicity, clog P>3. This indicated that the
difference in their efficacy might be partially explained
on the basis of lipophilicity.
3. Conclusion
Based on the results of pharmacological, physicochem-
ical, and molecular modelling investigations the phar-
macophore model for anticonvulsant N-substituted
amides of GHB was defined. In this model, the presence
of the N-benzylamide fragment is essential for activity.
Beside this, all of the amides contained another hydro-
phobic unit (aryl ring) as a distal binding site and H-
bond donor. To support the suggested pharmacophore
model we have synthesized a series of N-benzylamides
of GHB containing a hydrophobic moiety such as N-
benzylamino or a N-(4-chlorobenzylamino)- group in
the a-position of GHB and a lipophylic substituent in the
ortho-position in the phenyl ring in benzylamide frag-
ment. The 2-chloro- and 2-trifluoromethylbenzylamides
of a-(4-chlorobenzylamine)-GHB (5, 7) were the most
potent in the MES test. Among derivatives, the intro-
duction of the lipophilic substituent at the ortho-position
of phenyl ring was sufficient for their anticonvulsant
activity. We have also synthesized a series of (13–16) g-
acetoxy analogues of corresponding anticonvulsant
active g-hydroxy derivatives. In comparison with
hydroxy derivatives, g-acetoxy derivatives of a-sub-
stitutes butyric acid did not exhibit anticonvulsant activity.
These findings led us to conclude that hydroxy group
was necessary for MES activity, and more lipophylic
compounds showed better anticonvulsant properties.
Yield 14.38 g (53.5%); mp 246–249 ꢁC; TLC; Rf
(S1)=0.47, Rf (S2)=0.28. Anal. calcd for: C11H12O2NCl
HCl; Mr=262.1, C: 50.40; H: 5.00; N: 5.34 found: C:
50.54, H: 4.99, N: 5.43.
4.1.2. Synthesis of N-substituted amides of ꢁ-hydroxybu-
tyric acid (3–11). 3-(Substituted benzylamin)-tetra-
hydrofuran-2-one hydrochloride (0.005 mol) was
dissolved in water (10 mL) and a saturated solution of
NaHCO3 was added till the mixture become basic. The
product was extracted in AcOEt (3 ꢂ 25 mL). The
organic layers were combined, dried over Na2SO4 and
evaporated. The obtained oil was dissolved in toluene
(30 mL) and an appropriate amine (0.007 mol) was
added. The mixture was refluxed for 12 h. After
evaporation of solvent, the crude oil which crystallized
at room temperature was purified by crystallization.
4.1.3. N-Benzylamide of ꢀ-(4-chlorobenzylamine)-ꢁ-hy-
droxybutyric acid (3). Yield: 1.50 g (92.8%); mp 89–
91 ꢁC (from toluene); TLC: Rf (S1)=0.42, Rf (S3)=0.89;
1H NMR (CDCl3): 1.80–1.86 (2H, m, CH2), 3.14 (2H, s
wide, OH, NH), 3.27 (1H, t, CH), 3.60–3.75 (4H, m,
2CH2), 4.36–4.39 (2H, m, CH2), 7.10–7.31 (9H, m,
arom.), 7.61 (1H, t, NH amid); 13C NMR (CDCl3):
174.14 (C¼O), 127.38–137.99 (arom), 61.36 (C-2), 60.66
(C-4), 51.67 (C-12), 43.02 (C-5), 35.36 (C-3). Anal. calcd
for: C18H21O2N2Cl, Mr=332.8, C: 64.96; H: 6.36; N:
8.42 found: C: 64.43, H: 6.18, N: 8.22.
From these data, ideas for future molecular modifi-
cations leading to compounds with greater favorable
pharmacological properties may be derived.
4.1.4. N-4 - Methylbenzylamide of ꢀ - (4 - chlorobenzyl-
amine)-ꢁ-hydroxybutyric acid (4). Yield: 0.99 g
(57.22%); mp 104–106 ꢁC (from toluene); TLC: Rf
4. Experimental
4.1. Chemistry
1
(S2)=0.39, Rf (S3)=0.83; H NMR (CDCl3): 1.79–1.88
(2H, m, CH2), 2.35 (3H, s, CH3), 3.26 (2H, s wide, OH,
NH), 3.32 (1H, t, CH), 3.58–3.66 (2H, m, CH2), 3.69–
3.75 (2H, m, CH2), 4.36–4.39 (2H, m, CH2), 7.13–7.28
(8H, m, arom), 7.53 (1H, t, NH amid); 13C NMR
Melting points were determinated with a Buchi 535 and
were uncorrected. Thin layer-chromatography (TLC)

630
B. Malawska et al. / Bioorg. Med. Chem. 12 (2004) 625–632
1
(CDCl3): 174.04 (C¼O), 127.64–137.70 (arom), 61.43
(C-2), 60.76 (C-4), 51.71 (C-12), 42.83 (C-5), 35.62 (C-3),
20.95 (C-19). Anal. calcd for C19H23O2N2Cl;
Mr=346.9, C: 65.79, H: 6.68, N: 8.07 found: C: 66.66,
H: 6.51, N: 7.88.
(S3)=0.88; H NMR (CDCl3): 1.81–1.90 (2H, m, CH2),
2.32 (3H, s, CH3), 3.10 (2H, s wide, OH, NH), 3.33 (1H,
t, CH), 3.70–3.76 (4H, m, 2CH2), 4.41–4.44 (2H, CH2),
7.19–7.30 (9H, m, arom), 7.45 (1H, t, NH amid); 13C
NMR (CDCl3): 174.04 (C¼O), 126.10–139.16 (arom),
61.52 (C-2), 60.86 (C-4), 52.56 (C-12), 41.19 (C-5), 35.73
(C-3), 18.86 (C-19). Anal. calcd for C19H24O2N2;
Mr=312.4, C: 73.04, H: 7.74, N: 8.96, found: C: 70.95,
H: 6.08, N: 8.80.
4.1.5. N-2 - Chlorobenzylamide of ꢀ - (4 - chlorobenzyl-
amine)-ꢁ-hydroxybutyric acid (5). Yield: 0.75 g
(40.98%); mp 76–78 ꢁC (from n-hexane and ethyl ace-
tate); TLC: Rf (S1)=0.97, Rf (S3)=0.85; 1H NMR
(CDCl3): 1.80–1.89 (2H, m, CH2), 2.84 (2H, s wide, OH,
NH), 3.27 (1H, t, CH), 3.65–3.96 (4H, m, 2CH2), 4.51–
4.54- (2H, m, CH2), 7.13–7.39 (8H, m, arom), 7.55 (1H,
t, NH amid); 13C NMR (CDCl3): 174.06 (C¼O),
127.11–137.66 (arom), 61.52 (C-2), 60.96 (C-4), 51.86
(C-12), 41.29 (C-5), 35.67 (C-3). Anal. calcd for
C18H20O2N2Cl; Mr=367.3, C: 58.86, H: 5.48, N: 7.63,
found: C: 58.35, H: 5.38, N: 7.28.
4.1.10. N-2 - Trifluoromethylbenzylamide of ꢀ - benzyl-
amine-ꢁ-hydroxybutyric acid (10). Yield: 1.14 g
(39.89%); mp 83–85 ꢁC (from cyclohexane and ethyl
1
acetate); TLC: Rf (S2)=0.31, Rf (S3)=0.83; H NMR
(CDCl3): 1.79–1.88 (2H, m, CH2), 2.90 (2H, s wide, OH,
NH), 3.32 (1H, t, CH), 3.67–3.73 (4H, m, 2CH2), 4.59–
4.62 (2H, m, CH2), 7.23–7.48 (9H, m, arom), 7.67 (1H,
t, NH amid); 13C NMR (CDCl3): 174.43 (C¼O),
121.62–139.12 (arom), 61.37 (C-2), 60.72 (C-4), 52.61
(C-12), 39.73 (C-5), 39.68 (C-19), 35.62 (C-3). Anal.
calcd for C19H21O2N2F3; Mr=366.4, C: 62.28, H: 5.77,
N: 7.64, found: C: 62.99, H: 5.55, N: 7.39.
4.1.6. N - 2 - Methylbenzylamide of ꢀ - (4 - chlorobenzyl-
amine)-ꢁ-hydroxybutyric acid (6). Yield: 1.21g (69.94%);
mp 100–104 ꢁC (from toluene); TLC: Rf (S2)=0.26, Rf
1
(S3)=0.90; H NMR (CDCl3): 1.81–1.90 (2H, m, CH2),
2.31 (3H, s, CH3), 2.90 (2H, s wide, OH, NH), 3.28 (1H,
t, CH), 3.67–3.77 (4H, m, 2CH2), 4.41–4.44 (2H, m
CH2), 7.11–7.30 (9H, m, arom, NH); 13C NMR
(CDCl3): 173.86 (C¼O), 126.20–137.67 (arom), 61.55
(C-2), 60.94 (C-4), 51.86 (C-12), 41.29 (C-5), 35.72 (C-3),
18.89 (C-19). Anal. calcd for C19H23O2N2Cl;
Mr=346.9, C: 65.79, H: 6.68, N: 8.07 found: C: 65.09,
H: 6.41, N: 7.71.
4.1.11. N-2-Methoxybenzylamide of ꢀ-benzylamine-ꢁ-
hydroxybutyric acid (11). Yield: 1.14 g (64.02%); mp
61–64 ꢁC (from cyclohexane and ethyl acetate); TLC: Rf
1
(S2)=0.30, Rf (S3)=0.80; H NMR (CDCl3): 1.78–1.86
(2H, m, CH2), 3.07 (2H, s wide, OH, NH), 3.28 (1H, t,
CH), 3.65–3.73 (4H, m, 2CH2), 3.79 (3H, s, OCH3),
4.43–4.46 (2H, m, CH2), 6.84–6.94 (3H, m, arom), 7.22–
7.25 (6H, m, arom), 7.65 (1H, t, NH amid); 13C NMR
(CDCl3): 173.88 (C¼O), 110.15–157.37 (arom), 61.55
(C-2), 60.77 (C-4), 55.09 (C-19), 52.31 (C-12), 38.97
(C-5), 35.67 (C-3). Anal. calcd for: C19H24O2N2;
Mr=328.4, C: 69.48, H: 7.36, N: 8.53, found: C: 69.13,
H: 7.30, N: 8.66.
4.1.7. N-2-Trifluoromethylbenzylamide of ꢀ-(4-chloro-
benzylamine) - ꢁ - hydroxybutyric acid (7). Yield: 0.70 g
(34.90%); mp 72–76 ꢁC (from n-hexane and ethyl ace-
tate); TLC: Rf (S2)=0.63, Rf (S3)=0.86; 1H NMR
(CDCl3): 1.79–1.88 (2H, m, CH2), 2.90 (2H, s wide,
OH, NH), 3.28 (1H, t, CH), 3.64–3.75 (4H, m, 2CH2),
4.59–4.62 (2H, m, CH2), 7.11–7.51 (8H, m, arom),
7.63 (1H, t, NH amid); 13C NMR (CDCl3): 174.20
(C¼O), 121.63–137.63 (arom), 61.41 (C-2), 60.80 (C-4),
51.86 (C-12), 39.83 (C-5), 39.79 (C-12), 35.57 (C-3).
Anal. calcd for: C19H20O2N2ClF3; Mr=400.826, C:
56.93, H: 5.03, N: 6.98, found: C: 56.13, H: 5.05, N:
7.38.
4.1.12. Synthesis of N-substituted amides of ꢁ-acetoxy-
butyric acid (13–16). N-Substituted benzylamide of a-
phenylpiperazin-1-yl-g-hydroxybutyric acid (0.002 mol),
acetic anhydride (2 mL) and piridine (2 mL) was stirred
in 100 ꢁC for 15 min. The obtained mixture was cooled
down, dissolved in water (30 mL) and a 10% solution of
HCl was added in drops till the mixture become acidic.
The product was extracted in methylene chloride
(3ꢂ25 mL). The organic layers were combined, dried over
Na2SO4 and evaporated. After evaporation of solvent, the
crude oil which crystallized at room temperature was
purified by crystallization or by column chromatography.
4.1.8. N-2-Methoxybenzylamide of ꢀ-(4-chlorobenzyl-
amine)-ꢁ-hydroxybutyric acid (8). Yield: 0.73 g (40.33%);
mp 63–68 ꢁC (from cyclohexane and ethyl acetate);
1
TLC: Rf (S2)=0.46, Rf (S3)=0.87; H NMR (CDCl3):
1.78–1.87 (2H, m, CH2), 3.02 (2H, s wide, OH, NH),
3.25 (1H, t, CH), 3.62–3.72 (4H, m, 2CH2), 3.79 (3H, s,
OCH3), 4.42–4.46 (2H, m, CH2), 6.84–6.94 (2H, m,
arom), 7.12–7.31 (6H, m, arom), 7.51 (1H, s, NH amid);
13C NMR (CDCl3): 173.70 (C¼O), 110.25–157.43
(arom), 61.60 (C-2), 60.89 (C-4), 55.20 (C-19), 51.62
(C-12), 39.11 (C-5), 35.68 (C-3). Anal. calcd for:
C19H23O3N2Cl; Mr=362.9, C: 62.89, H: 6.38, N: 7.72,
found: C: 63.03, H: 6.51, N: 7.81.
4.1.13. N-Benzylamide of ꢀ-phenyl-piperazin-1-yl-ꢁ-
acetoxybutyric acid (13). Yield: 0.28 g (25.04%); mp
82–83 ꢁC (from n-hexane and ethyl acetate); TLC: Rf
1
(S1)=0.87, Rf (S2)=0.95; H NMR (CDCl3): 2.05–2.15
(2H, m, CH2), 2.06 (3H, s, CH3), 2.68–2.76 (4H, m,
piper.), 3.13–3.24 (5H, m, piper., CH), 4.23–4.31 (2H,
m, CH2), 4.46–4.50 (2H, d-d, CH2), 6.88–6.93 (3H, m,
arom.), 7.24–7.42 (8H, m, arom., NH); MS (70 eV), m/z
(%): 395 (2.47) [M+], 261(25.47), 202(15.16), 201(100),
132(13.68), 105(8.28), 104(11.53), 96(5.03), 91(13.30),
77(6.84), 43(5.42). Anal. calcd for: C23H29O3N3;
Mr=395.5, C: 69.85, H: 7.39, N: 10.62, found: C: 70.64,
H: 7.26, N: 10.82.
4.1.9. N-2-Methylbenzylamide of ꢀ-benzylamine-ꢁ-hy-
droxybutyric acid (9). Yield: 0.77 g (49.35%); mp 70–
75 ꢁC (from ethyl acetate); TLC: Rf (S2)=0.44, Rf

B. Malawska et al. / Bioorg. Med. Chem. 12 (2004) 625–632
631
sffiffiffiffiffiffiffiffiffiffiffi
P
4.1.14. N-4-Chlorobenzylamide of ꢀ-phenyl-piperazin-1-
yl-ꢁ-acetoxybutyric acid (14). Yield: 0.28 g (25.05%);
mp 101–103 ꢁC (from n-hexane and ethyl acetate); TLC:
d2
n
RMS ¼
1
Rf (S1)=0.80, Rf (S2)=0.92; H NMR (CDCl3): 2.04
(2H, m, CH2), 2.06 (3H, s, CH3), 2.67–2.75 (4H, m,
piper.), 3.13–3.19 (4H, m, piper.), 3.21–3.24 (1H, m,
CH), 4.23–4.31 (2H, m, CH2), 4.41–4.45 (2H, d-d, CH2),
6.89–6.93 (3H, m, arom.), 7.20–7.34 (6H, m, arom.),
7.43 (1H, t, NH amid); MS (70 eV), m/z (%): 429 (2.61)
[M+], 261(37.76), 201(100), 159(5.03), 132(10.28),
104(6.83), 77(4.29). Anal. calcd for: C23H28O3N3;
Mr=429.9, C: 64.25, H: 6.56, N: 9.77, found: C: 64.31,
H: 6.11, N: 10.10.
where d is the distance between two paired atoms, n is
the number of pairs that are fitted.
The lower RMS fit value is, the better similarity is.
4.3. Pharmacology
4.3.1. Anticonvulsant assays. Phase I of the evaluation
included three tests: maximal electroshock (MES),
subcutaneus pentylenetrazole (scMet), and rotorod test
for neurological toxicity (Tox). Male albino, CF No. 1
mice (18–25 g, Charles River, Willimington, MA, USA)
and male albino, Sprague–Dawley rats (100–150 g,
Charles River, Willimington, MA, USA) were used in
the experiment. Compounds were either dissolved
saline or suspended in 0.5% methylcellulose, and
were administered by ip injection at three dosage
levels (30, 100, and 300 mg/kg) with anticonvulsant
activity, and neurotoxicity noted 0.5 h and 4 h after
administration.
4.1.15. N-4-Fluorobenzylamide of ꢀ-phenyl-piperazin-1-
yl-ꢁ-acetoxybutyric acid (15). Yield: 0.64 g (57.48%);
mp 90–92 ꢁC (from cyclohexane and ethyl acetate);
1
TLC: Rf (S1)=0.88, Rf (S2)=0.82; H NMR (CDCl3):
1.98–2.14 (2H, m, CH2), 2.05 (3H, s, CH3), 2.62–2.77
(4H, m, piper.), 3.08–3.22 (5H, m, piper. CH), 4.16–4.24
(2H, m, CH2), 4.29–4.46 (2H, m, CH2), 6.85–7.05 (4H,
m, arom.), 7.34–7.27 (5H, m, arom.), 7.39 (1H, t, NH
amid.); MS (70 eV), m/z (%): 413 (2.84) [M+],
261(28.79), 202(15.01), 201(100), 159(6.93), 132(12.65),
109(9.24), 104(10.21), 77(6.30). Anal. calcd for:
C23H28O3N3F; Mr=413.5, C: 66.81, H: 6.82, N: 10.16,
found: 66.51, H: 6.71, N: 10.05.
4.4. Maximal electroshock seizure (MES) test18
Maximal electroshock seizures were elicited with a 60-
cycle alternating current of 50 mA intensity (5–7 times
that necessary to elicit minimal electroshock seizures)
delivered for 0.2 s via corneal electrodes. A drop of
0.5% tetracaine hydrochloride in 0.9% saline was
instilled in the eye prior to application of the electrodes.
Protective endpoints were defined as abolition of the
hind limb tonic extension component of the seizure.
Results are expressed as a ratio of the number of
animals protected/number of animals tested.
4.1.16. N-4-Methylbenzylamide of ꢀ-phenyl-piperazin-1-
yl-ꢁ-acetoxybutyric acid (16). Yield: 0.52 g (41.90%);
mp 113–114 ꢁC (from cyclohexane and ethyl acetate);
TLC: Rf (S1)=0.80, Rf (S2)=0.89; 1H NMR
(CDCl3): 1.96–2.16 (2H, m, CH2), 2.05 (3H, s,
CH3), 2.33 (3H, s, CH3), 2.63–2.78 (4H, m, piper.),
3.09–3.21 (5H, m, piper., CH), 4.11–4.25 (2H, m,
CH2), 4.30–4.43 (2H, m, CH2), 6.84–6.91 (3H, m,
arom), 7.12–7.26 (6H, m, arom), 7.33 (1H, t, NH
amid); MS (70 eV), m/z (%): 409 (2.20) [M+],
261(28.13), 201(100), 160(4.14), 132(11.83), 105(15.6),
96(4.29), 91(3.93) 77(6.91). Anal. calcd for:
C24H31O3N3; Mr=409.5, C: 70.39, H: 7.63, N: 10.26,
found: C: 70.03, H:7.80, N: 10.09.
4.5. Neurotoxicity19,20
The rotorod test was used to evaluate neurotoxicity in
mice. The animal was placed on a 1-inch diameter
knurled plastic rod rotating at 6 rpm. Non-toxic
(normal animals) mice can remain on a rod rotating at
this speed indefinitely. Neurologic toxicity is defined as
the failure of the animal to remain on the rod for 1 min
and is expressed as number of animals exhibiting toxi-
city/number of animals tested. Animals are considered
toxic if they fail this test on three successive attempts.
Rat toxicity was determined using overt evidence of
ataxia, abnormal gait or the positional sense test.
4.2. Molecular modelling study
The compounds were modelled and minimized using
PM3 (MOPAC) method of the Alchemy 2000
programme (Alchemy 2000 ver. 2.0, Tripos Inc., St.
Louis, USA, 1997). The conformational analyses were
carried out by systematic stepwise rotation of 100 four
torsion angles marked in Figure 2 as f1–f4, common to
all molecules. The energy was calculated for all
conformations taking into account both the molecular
mechanics and Van der Waals interactions. Energy
criterion, set to 3 kcal/mol above the lowest energy
conformation found, was used to accept new
conformation. Distances between atoms, which are not
involved in building atomic bonds were above 76% of
sum their Van der Waal’s radius. The RMS routine is
Acknowledgements
This study was supported by the Medical College of the
Jagiellonian University Research Programme 501/P/
156/F. Acknowledgement to Drs. S. White, H. Wolfe,
K. Wilcox and Ms. J. Woodhead at the University
of Utah for anticonvulsant testing and to the
Anticonvulsant Screening Project (ASP), National
Institute of Neurological Disorders and Stroke, NIH,
Bethesda, MD, USA.
a
procedure which provided an estimation how
closely molecules fit to each other. The quality of the
fit is determined via an RMS calculation using the
equation:

632
B. Malawska et al. / Bioorg. Med. Chem. 12 (2004) 625–632
12. Malawska, B.; Antkiewicz-Michaluk, L. Die Pharmazie
1999, 54, 239.
References and notes
13. Malawska, B.; Kulig, K.; Antkiewicz-Michaluk, L.;
Porter, R.; Misra, A.; Cliffe, A. Arch. Pharm. Pharm.
Med. Chem. 1999, 332, 167.
1. Cosford, N. D. P.; McDonald, I. A.; Schweiger, E. J.
Annu. Rep. Med. Chem. 1998, 33, 61.
2. Nicolson, A.; Leach, J. P. CNS Drugs 2001, 15, 955.
3. Gatti, G.; Bonomi, I.; Jannuzzi, G.; Perucca, E. Curr.
Pharm. Des. 2000, 6, 839.
14. Cros, A. J.; Stirling, J. M.; Robinson, T. N.; Bowen,
D. M.; Francis, P. T.; Green, A. R. Br. J. Pharmacol.
1989, 98, 284.
4. Lopes Lima, J. M. Curr. Pharm. Des. 2000, 6, 873.
5. Wong, M. G.; Defina, J. A.; Andrews, P. R. J. Med.
Chem. 1986, 29, 562.
6. Unverferth, K.; Engel, J.; Hofgen, N.; Rostock, A.;
Gunther, R.; Lankau, H.-J.; Menzer, M.; Rolfs, A.;
Liebscher, J.; Muller, B.; Hofmann, H.-J. J. Med. Chem.
1998, 41, 63.
7. Dimmock, J. R.; Vashishtha, S. C.; Stables, J. P. Eur. J.
Med. Chem. 2000, 35, 241.
8. Pandeya, S. N.; Yogeeswari, P.; Stables, J. P. Eur. J. Med.
Chem. 2000, 35, 879.
15. Green, A. R.; Misra, A.; Murray, T. K.; Snape, M. F.;
Cross, A. J. Neuropharmacology 1996, 35, 1243.
16. Stables, J.P.; Kupferberg, H.J. In Molecular and Cellular
Targets for Antiepileptic Drugs; Avanzini, G., Tanganelli,
P., Avoli, M., Eds.; John Libbey: London, 1997; p 191.
17. Prolog P module of the PALLAS system (version for
Windows 1.2) distributed by Compu Drug Chemistry Ltd.,
1995.
18. Gladding, G. D.; Kupferberg, H. J.; Swinyard, E. A.
Handbook of Experimental Pharmacology, Antiepileptic
Drugs; Springer: Berlin, Tokyo, 1985; p 74.
19. Dunham, N. W.; Mitiya, T. A. J. Am. Pharm. Assoc. Sci.
Ed. 1957, 46, 208.
9. Malawska, B.; Gobaille, S. Pharmazie 1995, 50, 390.
10. Malawska, B.; Zejc, A. Pharmazie 1995, 50, 722.
11. Malawska, B.; Kulig, K.; Ciechanowicz-Rutkowska, M.
Arch. Pharm. Pharm. Med. Chem. 1997, 330, 91.
20. Racine, R. J. Electroenceph. Clin. Neurophysiol. 1972, 32,
281.