Design and synthesis of anticonvulsive agents as γ-vinyl GABA-based potential dual acting prodrugs and their biological activities

Article abstract of DOI:10.1016/S0960-894X(00)00064-0

For the development of new anticonvulsive agents, γ-vinyl GABA (vigabatrin) and GABA mimetics derivatives were covalently coupled as potential dual acting prodrugs and evaluated for their anticonvulsive activities. Among the prepared compounds, 11 showed the most potent anticonvulsive activity, a shorter onset time and a broader spectrum compared to vigabatrin. (C) 2000 Elsevier Science Ltd. All rights reserved.

Full text of DOI:10.1016/S0960-894X(00)00064-0

Bioorganic & Medicinal Chemistry Letters 10 (2000) 609±613
Design and Synthesis of Anticonvulsive Agents as ꢀ-Vinyl
GABA-Based Potential Dual Acting Prodrugs and their
Biological Activities
Yong-chul Kim, a Long-Xuan Zhao, a Tae-Hyung Kim, a Sun-mi Je, a Eun-kyung Kim, a
Heesung Choi, b Whi-Gun Chae, c Minsoo Park, d Jongwon Choi, d Yurngdong Jahng a
and Eung-Seok Lee a,*
aCollege of Pharmacy, Yeungnam University, Kyongsan 712-749, South Korea
bCentral Research Institute, Dong Kook Pharm. Co., Chungbuk 365-830, South Korea
cSchool of Medicine, Catholic University of Taegu-Hyosung, Taegu 705-034, South Korea
dCollege of Pharmacy, Kyungsung University, Pusan 608-736, South Korea
Received 18 November 1999; accepted 20 January 2000
AbstractÐFor the development of new anticonvulsive agents, g-vinyl GABA (vigabatrin) and GABA mimetics derivatives were
covalently coupled as potential dual acting prodrugs and evaluated for their anticonvulsive activities. Among the prepared com-
pounds, 11 showed the most potent anticonvulsive activity, a shorter onset time and a broader spectrum compared to vigabatrin.
# 2000 Elsevier Science Ltd. All rights reserved.
Introduction
epilepsy.4,7 However, these compounds do not readily
enter into the central nervous system in pharmacologi-
cally e€ective amounts following peripheral administra-
tion, since these compounds can not readily pass the
blood±brain barrier (BBB), presumably due to their
lower lipophilicity.4,8,9
Approximately 1% of the population is reported to be
a€ected by some form of epilepsy and 20±40% of
epileptic patients have experienced failure to control
seizures, and 40±60% of patients have su€ered from
drug resistance to currently available anticonvulsant
drugs.1 Since epilepsy consists of various forms of
seizure,2 combinations of drugs and repeated therapy is
required to control such complex convulsions,3 which
increases the toxicity and adverse side e€ects of clini-
cally using anticonvulsant drugs.
In connection with the studies for the development of
new anticonvulsants having a broader spectrum with
potent anticonvulsant activity and lower toxicity, our
previous results have been published.10
In this study, we designed and synthesized potential
dual acting prodrugs which were covalently coupled
with an amide bond of vigabatrin (1)11 and GABA
mimetic substances such as isonipecotic acid (2)4, nipe-
cotic acid (3)12 and 2-pyrrolidinone (4)10 (Fig. 1). Viga-
batrin (g-vinyl GABA) was recently developed as an
anticonvulsive agent which was reported to be a
mechanism based inhibitor of GABA aminotransferase,
and has been widely used clinically for the treatment of
seizure along with valproic acid. But vigabatrin has a
relatively narrow clinical spectrum and long onset time
(18 h) due to its hydrophilic character, and a relatively
large amount (500±1000 mg) should be administered. 2-
Pyrrolidinone was reported as a prodrug of GABA,
isonipecotic acid as a GABA receptor agonist, nipecotic
Recently, many studies have been carried out for the
development of new types of anticonvulsant agents,
including g-aminobutyric acid (GABA) related com-
pounds,4 derivatives of amino acids,5 and structurally
modi®ed compounds of currently used drugs.6 However,
these compounds cannot provide complete control of
complex seizures. Many GABA mimetic substances such
as GABA receptor agonists, GABA reuptake inhibitors
and GABA metabolism inhibitors have been reported to
be potent anticonvulsant agents, and some of these
agents have been clinically used for the treatment of
*Corresponding author. Tel.: +82-53-810-2801-5; fax: +82-53-811-
3871; e-mail: eslee@yu.ac.kr
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00064-0

610
Y. Kim et al. / Bioorg. Med. Chem. Lett. 10 (2000) 609±613
Figure 1.
acid as a GABA uptake inhibitor. They show potent
anticonvulsive activities in vitro, but are barely potent
in vivo due to their hydrophilic properties.4,8,12 The
biological activity of the prepared prodrugs were eval-
uated for the development of potent anticonvulsive
agents having a broader clinical spectrum and higher
hydrophobic properties, which might exhibit dual acting
anticonvulsive activities with di€erent mechanisms.13
We considered that these prodrugs had higher hydro-
phobic properties sucient to pass through the BBB,
and expected them to enter readily into the CNS and
transform into active compounds after metabolic enzy-
matic hydrolysis in the CNS. Then the hydrolyzed
compounds could display their anticonvulsive activity
with di€erent mechanisms. Therefore, these prodrugs
are expected to have increased potency through a
synergistic e€ect and a broader spectrum. Also we
expected the hydrolyzed compounds to stay longer in
CNS due to their hydrophilic properties, which would
increase the duration of their anticonvulsant e€ects.
We designed and prepared 5 compounds based on the
vigabatrin moiety as follows. First, acid moiety of viga-
batrin was coupled by an amide bond with the amine
moiety of isonipecotic acid methyl ester, nipecotic acid
methyl ester and 2-pyrrolidinone (compounds 10, 11,
14, respectively) as shown in Scheme 1. Second, the
amine moiety of vigabatrin methyl ester was coupled by
an amide bond with the acid moiety of isonipecotic acid
and nipecotic acid (compounds 18, 19 respectively) as
shown in Scheme 2. The prepared compounds were
evaluated for their anticonvulsive activity.
Chemistry
Compounds 10 and 11 could be prepared from pro-
tected vigabatrin 5 with isonipecotic acid methyl ester
(6) and nipecotic acid methyl ester (7), respectively, in
moderate yield by known general reactions as outlined
in Scheme 1.
Scheme 1.

Y. Kim et al. / Bioorg. Med. Chem. Lett. 10 (2000) 609±613
611
Scheme 2.
Boc protected vigabatrin (5) was prepared from viga-
batrin (1) by treatment with BOC-ON and sodium
bicarbonate in water/dioxane in 92% yield as a colorless
crystal. Intermediates 8 and 9 were prepared from 5 and
isonipecotic acid methyl ester (6) and nipecotic acid
methyl ester (7), respectively, with addition of EDCI
and HOBT in 68 and 73% yield as colorless oils. Inter-
mediates 8 and 9 were treated with 33% TFA and 5%
DMS in CH2Cl2 to yield oils as TFA salt, which was
treated with diisopropyl ethyl amine (DIEA) to give
®nal products 10 and 11, respectively in 96 and 94%
yield as a colorless oil.
(PTZ), bicuculline (BIC) and picrotoxin (PCR) induced
seizure tests were carried out according to the protocol
of the Antiepileptic Drug Development Program of the
National Institute of Neurological Disorders and
Stroke.11,14 It has been reported that the MES test cor-
relates to both generalized tonic-clonic and psycho-
motor seizure, and the PTZ test to generalized absence
seizure.14 BIC and PCR were reported to be selective to
GABAA receptor, and vigabatrin was e€ective only
against BIC and PCR induced seizure.11 These seizure
tests are meaningful for clinical prediction of anti-
convulsant agents and for recognition of broader clinical
spectrum and dual action for prepared compounds. The
anticonvulsant activities of the compounds are shown in
Table 1, in comparison with the parent compounds.
Intermediate 13 was prepared with Boc vigabatrin (5) and
isobutyl chloroformate in the presence of TEA to form
a mixed anhydride, followed by the addition of an anion
of 2-pyrrolidinone (12) which was prepared from 2-pyr-
rolidinone (4) and n-butyllithium in 50% yield as a
colorless oil. Intermediate 13 was treated with 33% TFA
and 5% DMS in methylene chloride to yield a colorless
oil as TFA salt, followed by treatment with DIEA to
a€ord product 14 in 95% yield as a colorless oil.
Compound 11 showed the most potent anticonvulsive
activity, and the other compounds showed moderate
activity against the PTZ, BIC and PCR induced seizure
tests, compared to those of the parent compounds.
Especially, compound 11 showed the twice potency of
vigabatrin in the BIC and PCR tests and had potent
activity in the PTZ test, whereas the parent compound,
nipecotic acid, showed no activity. This indicates three
conclusions. First, the potent activity in the PTZ test
of compound 11 would be derived from the parent
compound, nipecotic acid, which was reported to be
e€ective in vitro12 but not e€ective in vivo due to its
Boc protected isonipecotic acid 15 was prepared from
isonipecotic acid (2) by treatment with BOC-ON and
sodium bicarbonate in water/dioxane in 90% yield as a
colorless crystal (Scheme 2). Intermediate 17 was pre-
pared from the methyl ester of vigabatrin 16 with the
addition of EDCI and HOBT in 64% yield as a color-
less oil. Intermediate 17 was treated with 33% TFA and
5% DMS in CH2Cl2 to yield an oil as TFA salt, which
was treated with DIEA to give the ®nal product 18 in
96% yield as a colorless oil. Final product 19 was pre-
pared from nipecotic acid (3) in total 48% yield as a
colorless oil by the same method as described above.
Since both vigabatrin and nipecotic acid are available as
its racemic form, the compounds 11 and 19 were syn-
thesized as its diastereomeric mixture.
Table 1. The anticonvulsive activities of prepared compounds
Compound
ED50 (mmol/kg)
MES
PTZ
BIC
PCR
10
11
14
18
>0.54
>0.54
>0.54
>0.54
>0.54
na
n
n
n
0.32
0.15
>0.54
>0.54
0.41
n
n
n
0.54
0.15
0.14
0.20
0.46
0.31
0.22
n
0.26
0.10
0.26
0.39
0.26
0.21
n
19
Vigabatrin
Nipecotic acid
Isonipecotic acid
2-Pyrrolidinone
Pharmacological Results and Discussion
n
n
n
n
The anticonvulsant activity of the prepared compounds in
maximal electroshock seizure (MES), pentylenetetrazole
an: Not e€ective (>1.0 mmol/kg).

612
Y. Kim et al. / Bioorg. Med. Chem. Lett. 10 (2000) 609±613
Table 2. Anticonvulsive activities over time
Good Health R&D Project, Ministry of Health and
Welfare, R.O.K.
Compound
Administered amounta
Time Convulsiveb/tested
(h)
mice
10
100 mg/kg (0.27 mmol/kg)
3
6
9
18
2/4
2/4
0/4
0/4
References and Notes
1. Begley, C. E.; Annegers, J. F.; Lairson, D. R.; Reynolds, T.
F.; Hauser, W. A. Epilepsia 1994, 35, 1230.
11
100 mg/kg (0.27 mmol/kg)
3
6
9
18
2/4
0/4
0/4
0/4
2. Lindhart, D.; Hopperner, R. J. E. A. Epilepsia 1984, 25, 77.
3. Vida, J. A. In Principles of Medicinal Chemistry; Foye, W.
O., Ed.; Lea & Febiger: Philadelphia, 1995; Chapter 11.
4. Krogsgaard-Larsen, P. J. Med. Chem. 1981, 24, 1377.
5. (a) Kohn, H.; Sawhney, K. N.; LeGall, P.; Robertson, D.
W.; Leander, J. D. J. Med. Chem. 1991, 34, 2444. (b) Kohn,
H.; Sawhney, K. N.; LeGall, P.; Conley, J. D.; Robertson, D.
W.; Leander, J. D. J. Med. Chem. 1990, 33, 919.
6. (a) Kwon, C. H.; Iqbal, M. T.; Wurpel, J. N. D. J. Med.
Chem. 1991, 34, 1845. (b) Eda®ogho, I. O.; Scott, K. R.;
Moore, J. A.; Farrar, V. A.; Nicholson, J. M. J. Med. Chem.
1991, 34, 387.
Vigabatrin 100 mg/kg
3
6
9
18
4/4
4/4
3/4
0/4
aEach mouse was administered a dose of 100 mg/kg (0.27 mmol/kg) of
each tested compound, and protection from convulsion was observed
over time.
bPicrotoxin induced seizure test was performed.
7. Matsuyama, K.; Gomita, Y.; Yamashita, C. Chem. Pharm.
Bull. 1984, 32, 4089.
8. Sasaki, H.; Mori, Y.; Nakamura, J.; Shibasaki, J. J. Med.
Chem. 1991, 34, 628.
hydrophilicity, since vigabatrin was reported not to be
e€ective in PTZ test. That indicated that compound 11
readily entered into the CNS and was transformed back
into its parent compounds by metabolizing enzymes.
Second, the twice potent activity in the BIC and PCR
test of compound 11 compared to vigabatrin indicated
that compound 11 easily entered into the brain, was
readily hydrolyzed into its parent compounds and
stayed long enough in the brain to display better activity.
Finally, the hydrolyzed compounds exhibited dual
activity with di€erent mechanisms, to provide increased
anticonvulsive potency through a synergistic e€ect and
broader clinical spectrum as expected.
9. Smith, M. B.; Kwon, T. W. Synth. Commu. 1992, 22, 2865.
10. (a) Park, M.; Lee, J.; Choi, J. Bioorg. Med. Chem. Lett.
1996, 6, 1297. (b) Jung, K.; Son, K.; Kim, M.; Lee, J.; Choi, J.;
Lee, E.-S.; Park, M. Arch. Pharm. Res. 1998, 21, 759. (c) Son,
K.; Jung, K.; Kim, M.; Lee, J.; Choi, J.; Lee, E.-S.; Park, M.
Arch. Pharm. Res. 1998, 21, 764. (d) Lee, J.; Son, K.; Kim, M.;
Jung, G.; Choi, J.; Lee, E-S.; Park, M. Arch. Pharm. Res.
1999, 22, 491
11. Holland, K. D.; McKeon, A. C.; Canney, D. J.; Covey, D.
F.; Ferrendelli, J. A. Epilepsia 1992, 33, 981.
12. N'Goka, V.; Schlewer, G.; Linget, J-M.; Chambon, J-P.;
Wermuth, C-G. J. Med. Chem. 1991, 34, 2547.
13. (a) Maynert, E. W.; Kaji, H. K. J. Pharmacol. Exp. Ther.
1962, 137, 114. (b) Frey, H. H.; Loscher, W. Neuropharma-
cology 1980, 19, 217. (c) Shashoua, V. E.; Jacob, J. N.; Ridge,
R.; Campbell, A.; Baldessarini, R. J. J. Med. Chem. 1984, 27,
659. (d) Jacob, J. N.; Shashoua, V. E.; Campbell, A.; Baldes-
sarini, R. J. J. Med. Chem. 1985, 28, 106.
14. (a) Swinyard, E. A.; Woodhead, J. H.; White, H. S.;
Franklin, M. R. In General Principles, Experimental Section,
Quanti®cation and Evaluation of Anticonvulsants in Anti-
epileptic Drugs, 3rd Ed. Levy, R. Ed.; Raven Press: New York,
1988; p. 88. (b) Krall, R. L.; Penry, J. K.; White, B. G.; Kup-
ferberg, H. J.; Swinyard, C. A. Epilepsia 1978, 19, 409. The
pharmacological tests were carried out as follows: all tested
compounds were dissolved in polyethylene glycol 400 and
administered intraperitonealy to ICR male mice at doses of 25,
50, 75 and 100 mg/kg. The anticonvulsant tests were per-
formed 30 min after administration in groups of four mice. We
had determined the lowest dose that would induce seizures in
all tested animals during preliminary screening. Seizures were
then arti®cially induced by either electric shock or chemicals.
The MES tests were carried out with a 60-cycle a.c. of 50 mA
intensity delivered for 0.2 seconds via corneal electrodes with
an ECT unit (UGO Basline, Italy). A drop of 0.9% saline was
instilled in the eye prior to application of electrodes. Protection
in this test was de®ned as the abolition of the hind limb tonic
extension component of a seizure. The pentylenetetrazole
(PTZ), bicuculline (BIC) and picrotoxin (PCR) induced sei-
zure test entailed the administration of 80 mg/kg of PTZ
(CD97), 3.2 mg/kg of BIC (CD97) and 5 mg/kg of PCR (CD97)
as a 0.5% solution subcutaneously in the posterior midline of
the mice, respectively, and observation lasted for 30 min. Pro-
tection was de®ned as the failure to observe even a threshold
The anticonvulsant activities over time of compounds
10 and 11 in the PCR induced seizure test is shown in
Table 2. While vigabatrin exhibited maximum activity
18 h after administration, compound 11 displayed
moderate activity 3 h and maximum activity 6 h
after administration. The shorter onset time also indi-
cated that compound 11 more readily passed BBB
compared to vigabatrin, and easily hydrolyzed to its
parent compounds.
In conclusion, we have designed and prepared ®ve
potential anticonvulsant agents as dual acting prodrugs
to increase potency and to broaden the clinical spec-
trum. Compound 1115 showed the most potent anti-
convulsant activity in the PTZ, BIC and PCR tests, and
displayed shorter onset time, which are believed to be a
promising drug candidate having higher potency and a
broader clinical spectrum. Further studies are in pro-
gress to evaluate and determine the rate of penetration
into the brain and the amount of transformation of
compound 11 into parent compounds in the brain.
Acknowledgements
This study was supported in part by a grant from
Yeungnam University and in part by a grant from the

Y. Kim et al. / Bioorg. Med. Chem. Lett. 10 (2000) 609±613
613
seizure, a single episode of chronic spasms that persisted for at
least 5 sec. Quantitative evaluation of the anticonvulsant
activity of ED50 was estimated from dose-response data.
15. The spectral data of compound 11: H NMR (300 MHz,
CDCl3) ppm 5.82 (m, 1H), 5.38 (dd, 1H), 5.34 (dd, 1H), 3.87
(m, 1H), 3.71 (dt, 1H), 3.69 (s, 3H), 3.45 (dt, 1H), 3.11 (m, 1H),
2.95 (dt, 1H), 2.51 (m, 1H), 2.43 (m, 2H), 2.11±1.47 (m, 6H)
13C NMR (75 MHz, CDCl3) ppm 173.2, 171.7, 132.8, 120.9,
54.3, 52.0, 47.2, 46.3, 44.0, 42.7, 29.39, 27.9, 26.7 FABMS
(MH+) 255 TLC: Rf=0.63 (10%MeOH/CH2Cl2 in NH3 vapor).
1