Synthesis and anticonvulsant and neurotoxic properties of substituted N- phenyl derivatives of the phthalimide pharmacophore

Article abstract of DOI:10.1021/jm990068t

A series of compounds including 4-amino (1), 3-amino (2), 4-nitro (3), 2-methyl-3-amino (4), 2-methyl-3-nitro (5), 2-methyl-4-amino (6), 2-methyl-4- nitro (7), 2-methyl-5-amino (8), 2-methyl-5-nitro (9), 2-methyl-6-amino (10), 2-methyl-6-mitro (11), 2,6-dimethyl (12), 2-methyl-3-carboxy (13), 2- methoxycarbonyl (14), 2-methyl-4-methoxy (15), 2,4-dimethoxy (16), 2-chloro- 4-amino (17), and 2-chloro-4-nitro (18) N-phenyl substituents of phthalimide were evaluated along with N-[3-methyl-(2-pyridinyl)]phthalimide (19), N-(3- amino-2-methylphenyl)succinimide (20), and phenytoin for anticonvulsant and neurotoxic properties. Initial screening in the intraperitoneal (ip) maximal electroshock-induced seizure (MES) test and the subcutaneous pentylenetetrazol-induced seizure (scPtz) test in mice led to the selection of 1, 2, 4, 10, 12, 17, and 19 for oral MES evaluation in rats. The resultant ED₅₀ values for 4, 10, 17, and phenytoin were 8.0, 28.3, 5.7 and 29.8 mg/kg, respectively. In the batrachotoxin affinity assay, IC₅₀ values for 17 and phenytoin were 0.15 and 0.93 μM, respectively, and in the recently validated magnesium deficiency-dependent audiogenic seizure test, ED₅₀ values of 5.2 and 23 mg/kg were obtained for 17 and phenytoin, respectively. Electrophysiology studies on compound 17 point out its ability to (i) potentiate GABA-evoked current responses with a failure to directly activate the GABAA receptor and (ii) to affect, at 100 μM excitatory non NMDA, but not NMDA, receptors with a 25% block of kainate-evoked response. Electrophysiology measurements on voltagegated sodium channels in N1E-115 neuroblastoma cells confirm voltage-dependent block of these channels by compound 17. In view of its interaction with multiple ion channels, one would predict that compound 17 might be active in a wide range of seizure models.

Full text of DOI:10.1021/jm990068t

J . Med. Chem. 2000, 43, 1311-1319
1311
Syn th esis a n d An ticon vu lsa n t a n d Neu r otoxic P r op er ties of Su bstitu ted
N-P h en yl Der iva tives of th e P h th a lim id e P h a r m a cop h or e
J oseph Vamecq,*,† Pierre Bac,‡ Christine Herrenknecht,§ Pierre Maurois,† Philippe Delcourt,| and
J ames P. Stables
INSERM/ Neuropaediatrics Department, Hoˆpital Roger Salengro, CHU Lille, 2, Avenue Oscar Lembret, 59037 Lille, France,
INSERM/ Neurosciences, Department of Professor Philippe Evrard, Hoˆpital Robert-Debre´, 48, Boulevard Se´rurier,
75013 Paris, France, Laboratoire de Pharmacologie, Faculte´ de Pharmacie, Chaˆtenay-Malabry, France,
Laboratoire de Chimie Analytique, Faculte´ de Pharmacie, Chaˆtenay-Malabry, France, Laboratoire de Physiologie Cellulaire,
INSERM EPI 9938, SN3 USTL, Villeneuve d’Ascq, France, and Preclinical Pharmacological Section, Epilepsy Branch,
DCDND, NINDS, National Institutes of Health, Neuroscience Center, Rockville, Maryland 20892-9523
Received February 8, 1999
A series of compounds including 4-amino (1), 3-amino (2), 4-nitro (3), 2-methyl-3-amino (4),
2-methyl-3-nitro (5), 2-methyl-4-amino (6), 2-methyl-4-nitro (7), 2-methyl-5-amino (8), 2-methyl-
5-nitro (9), 2-methyl-6-amino (10), 2-methyl-6-nitro (11), 2,6-dimethyl (12), 2-methyl-3-carboxy
(13), 2-methoxycarbonyl (14), 2-methyl-4-methoxy (15), 2,4-dimethoxy (16), 2-chloro-4-amino
(17), and 2-chloro-4-nitro (18) N-phenyl substituents of phthalimide were evaluated along with
N-[3-methyl-(2-pyridinyl)]phthalimide (19), N-(3-amino-2-methylphenyl)succinimide (20), and
phenytoin for anticonvulsant and neurotoxic properties. Initial screening in the intraperitoneal
(ip) maximal electroshock-induced seizure (MES) test and the subcutaneous pentylenetetrazol-
induced seizure (scPtz) test in mice led to the selection of 1, 2, 4, 10, 12, 17, and 19 for oral
MES evaluation in rats. The resultant ED50 values for 4, 10, 17, and phenytoin were 8.0, 28.3,
5.7 and 29.8 mg/kg, respectively. In the batrachotoxin affinity assay, IC50 values for 17 and
phenytoin were 0.15 and 0.93 µM, respectively, and in the recently validated magnesium
deficiency-dependent audiogenic seizure test, ED50 values of 5.2 and 23 mg/kg were obtained
for 17 and phenytoin, respectively. Electrophysiology studies on compound 17 point out its
ability to (i) potentiate GABA-evoked current responses with a failure to directly activate the
GABAA receptor and (ii) to affect, at 100 µM excitatory non NMDA, but not NMDA, receptors
with a 25% block of kainate-evoked response. Electrophysiology measurements on voltage-
gated sodium channels in N1E-115 neuroblastoma cells confirm voltage-dependent block of
these channels by compound 17. In view of its interaction with multiple ion channels, one
would predict that compound 17 might be active in a wide range of seizure models.
In tr od u ction
terpart of ameltolide: the 4-amino-N-(2,6-dimethyl-
phenyl) phthalimide.6-9
Using the 4-aminobenzoylamino template, Clark and
co-workers demonstrated through a series of successive
works the significant anticonvulsant potential in animal
epilepsy models for the amino-substituted benzamides
derived from alkyl-, arylalkyl-, and arylamines.1-4 SAR
studies showed that optimal antiepileptic activity was
found in those amides having a primary amine in the
4-position of the benzamide moiety and an aromatic
N-substitution of the 4-aminobenzamide pharmaco-
phore. From this work emerged ameltolide which ex-
hibits a phenytoin-like profile, i.e., it is quite potent in
the maximal electroshock seizure (MES) test and is
inactive in the subcutaneous pentylenetetrazol (scPTZ)
test.5 More recently, N-phenylphthalimide derivatives
were shown to possess a similar degree of anticonvul-
sant potency also associated with a phenytoin-like
profile, leading to the design of the phthalimide coun-
Our interest in developing phenytoin-like compounds
relies with the need of alternative new drugs because
of well-established side effects of phenytoin which
include in humans neurologic signs (ataxia, nystagmus,
sedation or irritability, oro-facial dyskinesia), along with
hematologic (leucopenia), immunologic (reduction of IgA,
lupus syndrome), endocrinologic (hirsutism), and cell
growth (gingival hypertrophy) dysfunctions. 4-Amino-
N-(2,6-dimethylphenyl)phthalimide was previously de-
signed from the models of ameltolide and thalidomide
(see ref 10 for the anticonvulsant properties of thalido-
mide) (Figure 1, design 1). In a recent work using the
batrachotoxin affinity assay, it was noted that the
4-amino moiety played an important role in the molec-
ular recognition process at the level of the receptor
modulating the voltage-dependent sodium channel sta-
tus.11 This moiety was therefore preserved in most
pharmacomodulations. The question could then be
raised whether this functionality not present in thali-
domide could be omitted in the phthaloyl moiety (Figure
1, design 2). In a previous series, one such compound
was relatively active,11 and in an effort to extend the
SAR of N-phenylphthalimide possessing no substitution
* Corresponding author: Dr J oseph Vamecq, INSERM/Neuropedi-
atrics, Hoˆpital Roger Salengro, CHU Lille, 2, Avenue Oscar Lembret,
59037 Lille, France. Fax: (+33) 3 20 44 53 93.
† INSERM/Neuropaediatrics and INSERM/Neurosciences.
‡ Laboratoire de Pharmacologie.
§ Laboratoire de Chimie Analytique.
| Laboratoire de Physiologie Cellulaire.
Preclinical Pharmacological Section, NIH.
10.1021/jm990068t CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/17/2000

1312 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 7
Vamecq et al.
Ch em istr y
The target compounds (Figure 2) were synthesized by
condensing phthalic anhydride with the appropriate
aniline derivative in acetic acid at reflux temperature
(compounds 3, 5, 7, 9, 11-16, and 18). The amino
derivatives were obtained by reducing the nitro precur-
sor through transfer hydrogenation at reflux tempera-
ture using cyclohexene as hydrogen donor, palladium
on charcoal as catalyst, and 2-propanol as a solvent
(compounds 1, 2, 4, 6, 8, 10, and 17).
Compound 19 was generated by reacting phthalic
anhydride with 2-amino-3-picoline. Compound 20, a
succinimide derivative, was obtained by reducing the
nitro group of the compound resulting from the conden-
sation of succinic anhydride with 2-methyl-3-nitro-
aniline. Most of the evaluations on compound 12 and
phenytoin were previously reported in refs 11 and 13,
respectively.
Resu lts
An ticon vu lsa n t a n d Neu r otoxicity Scr een in g
Da ta in Mice Dosed In tr a p er iton ea lly (ip ). Table 1
provides screening data for the activity of compounds
in animal testing of seizures induced by electroshock
(MES test) and pentylenetetrazol (scPtz test) and neu-
rotoxicity determined in the rotorod test. Compounds
were administered to mice by intraperitoneal route 30
min or 4 h before evaluation of their activities in these
tests. Comparison with data recorded under the same
conditions on phenytoin, the reference prototype anti-
epileptic drug, is further provided.
1. MES a n d Rotor od Tests. At doses tested (30, 100,
and 300 mg/kg), compounds 3, 5, 13, 16, and 18 were
found to be devoid of activity in the MES test and
presented no neurotoxicity (neurotoxicity is defined in
the Experimental Section) at any of the doses admin-
istered. Anti-MES activity at 300 mg/kg was recorded
for compounds 7, 11, and 14, with neurotoxicity ex-
pressed only by compound 14. Compounds 1, 2, 4, 8, 9,
10, 12, 15, and 19 were active in the MES test at 30
min using 100 mg/kg in the MES test. Compounds 1, 8,
10, 12, 15, and 19 were active at the same dose at the
4 h time point while compounds 2, 4, and 9 had
protective effects at both time points. The most neuro-
toxic of these compounds were 9 (30 mg/kg) and 2 (100
mg/kg). A lesser degree of neurotoxicity was recorded
for compounds 1, 10, 14, and 19, whereas compounds
4, 8, 12, and 15 were devoid of any neurotoxicity at
doses up to 300 mg/kg. Anti-MES activity at 30 mg/kg
was obtained for compounds 6, 17, and phenytoin.
Regarding their degree of neurotoxicity, the order of
compounds (from none to high neurotoxicity) was 17,
phenytoin, and 6.
F igu r e 1. Anticonvulsant compounds designed from amelt-
olide and thalidomide. Design 1 has previously generated
4-amino-N-phenylphthalimide compounds. The prototype of
this design is 4-amino-N-2,6-dimethylphthalimide and repre-
sents the phthalimide counterpart of the benzamide amelt-
olide; the former may be considered as a rigidified analogue
of the latter. Design 2 is the basis of the present work and, by
reference to thalidomide, deals with the omission of the
4-amino substituent of the phthalimide pharmacophore. Struc-
tures and chemistry of compounds synthesized according to
this design are illustrated by Figure 2.
in the phthaloyl moiety, we undertook to study a series
of 20 derivatives including methyl, amino, chloro, and
nitro substituents of the N-phenyl moiety of N-phenyl-
phthalimide pharmacophore (these substituents were
those giving the most active compounds against animal
seizure models [in MES testing] from an earlier se-
ries2,4,5,8,9,11 on phenylbenzamides and phenylphthal-
imides). The compounds were evaluated for their anti-
epileptic and neurotoxic properties through the Anti-
epileptic Drug Development (ADD) Program developed
by the National Institutes of Health.12-15 These evalu-
ations were completed by testing the anticonvulsant
profile of the most potent compound in the recently
validated16 magnesium deficiency-dependent audiogenic
seizure (MDDAS) model. The selected compound (i.e.,
N-(2-chloro-4-aminophenyl)phthalimide) was found to
be more potent than phenytoin in the batrachotoxin
affinity assay performed on rat brain synaptosomes.
Furthermore, electrophysiology studies point out inter-
actions of this compound with multiple ion channels and
present a promising pharmacological profile for the
treatment of the various types of seizure disorders.
2. scP tz Test. Some activity was recorded in the
scPtz test for compounds 2 and 9 (active at 30 mg/kg at
30 min), 15 and 17 (active at 100 mg/kg at 4 h and 30
min, respectively), and 19 (at 300 mg/kg). But only
compounds 2, 15, and 17 were active in this model at
non-neurotoxic doses.
An ticon vu lsa n t a n d Neu r otoxic P r op er ties of
Com p ou n d s in Ra ts Dosed Or a lly. Compounds 1, 2,
4, 10, 12, 17, 19, and phenytoin were selected for oral
evaluation of anti-MES and neurotoxic activity in rats
(Table 2). Compounds were administered per os, and

Anticonvulsant N-Phenylphthalimides
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 7 1313
F igu r e 2. Chemistry of 20 compounds related to N-phenyl substituents of the phthalimide pharmacophore. Compounds 1-18
represent substituted phenyl N-derivatives of phthalimide. Compounds 19 and 20 are alternatively designed compounds which
correspond to N-piperidinyl and N-phenyl derivatives of phthalimide and succinimide pharmacophores, respectively.
their effects in rats were studied 15, 30, 60, 120, and
240 min after ingestion of either 30 mg/kg (compounds
1, 2, 10, 17, and 19) or 50 mg/kg (4, 12, and phenytoin)
of experimental drug. At doses given orally to rats, none
of the compounds were neurotoxic. The most active
compounds in the MES test were 4, 17, and phenytoin,
giving protection of 100% of animals (4/4) at most time
points. Compound 2 gave 50 to 100% protection, de-
pending on the time point. The other compounds (1, 10,
12, and 19) were active but in a lower percentage of
animals.
were 30 min (compound 10), 1 h (compound 4), 2 h
(compound 17), and 4 h (phenytoin).
E va lu a t ion of Com p ou n d 17 in Ba t r a ch ot oxin
Affin it y Assa y a n d Ma gn esiu m Deficien cy-De-
p en d en t Au d iogen ic Seizu r e Mod el. The ability of
17 to interact with neuronal voltage-dependent sodium
channels was studied in the batrachotoxin affinity
assay. The apparent IC50 (i.e., the concentration inhibit-
ing by 50% the binding to rat synaptosomes of [3H]-
batrachotoxinin-A-20R-benzoate) of this compound was
found to be 0.15 ( 0.02 µM. The apparent IC50 recorded
in the same experimental conditions for phenytoin, as
a standard reference drug interacting with Na channels,
was about 0.93 ( 0.11 µM.
Quantitative evaluation of 17 and phenytoin in the
MES test in mice dosed intraperitoneally indicated ED50
values of 10.5 and 9.5 mg/kg, respectively. Compound
17 provided a TD50 value greater than 500 mg/kg which
exceeded that for phenytoin (65.4 mg/kg), giving rise to
PI exceeding 48 for compound 17 and about 7 for
phenytoin. Determination of ED50 values was also
carried out in adult mice in the MDDAS test. Compound
17 presented an ED50 of 5.2 mg/kg, and phenytoin ED50
amounted to 22 mg/kg. The two drugs were further
tested using their ED50 by recording their effect on the
duration of the four phases of MDDAS. The MDDAS
component modulation profile of 17 was close to that of
phenytoin, reducing the durations of wild running,
Qu an titative Evalu ation on Selected Com pou n ds
in Ra ts Dosed Or a lly. Pharmacological quantitations
were performed for compounds 4, 10, 17, and phenytoin
in the MES and scPtz tests. Table 3 summarizes the
main results of determinations performed in rats dosed
orally, including ED50 (effective dose in 50% of animals
tested), TD50 (neurotoxic dose in 50% of animals tested),
time to peak effect (TPE), and protective index (PI )
ratio between TD50 and ED50 values). None of the
selected compounds were active in the scPtz test at the
doses administered. In the MES test, the best ED50
value was recorded for compound 17 (5.7 mg/kg) which
also offered a good protection index (>88). Compound
4 was also highly active with an ED50 of 8.0 mg/kg and
a PI greater than 50. Compound 10 and phenytoin were
much less active with ED50 values approximating 30 mg/
kg and the PIs exceeding 17 and 100, respectively. TPEs

1314 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 7
Vamecq et al.
Ta ble 1. Anticonvulsant and Neurotoxicity Screening Data in
Ta ble 3. Quantitative Anticonvulsant Data in Rats Dosed
Orallya
Mice Dosed Intraperitoneally with Compoundsa
MES test
30 min 4 h
scPtz test
30 min 4 h
toxicity
30 min
ED50
MES
ED50
scPtz
PI
TPE
MES
test (h)
TD50
MES test
compounds
4 h
compound test (mg/kg) (mg/kg) (mg/kg) (TD50/ED50
)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
++
++
-
-
++
-
-
+++
-
-
-
-
-
-
-
-
-
+
-
-
-
-
-
++
-
+
-
-
-
-
+
++
-
-
+
-
-
-
++
-
-
+
-
-
-
-
-
-
-
-
-
-
-
-
4
8.0 ( 3.0
(6.2-10.6)
28.3 ( 3.8
(26.0-32.0)
5.7 ( 1.6
>200
>250
>250
>800
>400
>500
>50
>17
1
10
17
0.5
2
++
-
++
-
-
-
-
-
>500
>88
+++
-
++
+
-
++
-
(4.0-8.0)
-
phenytoin 29.8 ( 4.3
(21.9-38.9)
>3000
>100
4
++
++
++
-
-
-
-
++
+
+++
-
+++
+
a
Pharmacological values given in this table are as follows: the
ED50, dose of drug required to assure anticonvulsant protection
in 50% of animals; the TD50, dose eliciting minimal neurological
+
-
-
++
-
-
-
-
-
-
-
toxicity in 50% of animals; the PI, protection index (PI ) TD50
/
+
-
-
+
ED50), and the time to peak effect (TPE). ED50 and TD50 values
are expressed as mg/kg, and TPE as hours. Data on phenytoin
are from ref 13.
++
-
-
+
-
-
-
-
17
18
19
20
+++
-
-
++
-
-
-
-
Ta ble 4. Modulation by Compound 17 and Phenytoin of the
Various Components of the Seizure Episode in the MDDAS
Modela
++
+
-
+
+
-
-
-
phenytoin
+++
++
-
+
latency
for wild
running
duration
of wild
running
convulsion
phase
time-period time-period
recovery
phase
a
The anticonvulsant (MES and scPtz tests) and neurotoxicity
activities were determined 30 min and 4 h after the administration
of compounds. The symbols +++, ++, and + signify activity at
30, 100, and 300 mg/kg, respectively; - denotes no activity
observed at 300 mg/kg. Toxicity was determined by the rotorod
test. Abbreviations are as follows: MES, maximal electroshock
seizures; scPtz, subcutaneous pentylenetetrazol.
administration
of drugs
no drugs
11.2 ( 1.1 20.0 ( 1.6
5.4 ( 0.6
52.4 ( 1.6
6.8 ( 0.9*
7.8 ( 0.8*
PHT (22 mg/kg) 55.7 ( 3.2* 6.7 ( 0.4* 2.3 ( 0.6*
17 (5 mg/kg)
6.4 ( 0.7* 5.1 ( 0.5* 2.9 ( 0.6*
a
The various time period determinations are expressed as
seconds in convulsing animals at drug dosing ensuring an anti-
convulsant protection in 50% and no protection in the other 50%
of tested animals. Adult OF1 mice receiving a magnesium-deficient
(50 ppm for 42 days) were utilized. *p < 0.01 (vs magnesium-
deficient mice having received no drugs).
Ta ble 2. Anticonvulsant [anti-MES] and Toxicity Screening
Data in Rats Dosed Orally with Selected Compounds
compounds 15 min 30 min
1 h
2 h
4 h
toxicity
1
2
4
10
12
17
19
-
++
+
+++
++
+
-
-
-
-
-
-
-
-
++++ +++
+++ ++++
voltage-gated Na channel effects were evaluated using
N1E-115 neuroblastoma cells.
++++ ++++ ++++ ++++ ++++
-
++
-
++
+++
-
+
+
+
++
Compound 17 potentiated GABA-evoked current re-
sponses. At the highest concentration tested (100 µM),
it failed to directly activate the GABAA receptor (Table
5, A). NMDA-evoked potentials were not affected by
compound 17 at 100 µM (Table 5, B). By contrast, the
compound (at 100 µM) was somewhat active at non-
NMDA receptors, producing a 25% block of the kainate-
evoked response (Table 5, C). As expected, compound
17 produced a voltage-dependent block of voltage-gated
sodium channels (Table 5, D). This activity was found
to be of the same order of magnitude as activities
recorded for phenytoin, carbamazepine, and lamotrigine
in similar experimental conditions (see legend to Table
5).
+++ ++++ ++++ ++++ ++++
+
++
++
++
-
+
phenytoin ++++ ++++ ++++ ++++
a
At each time point, rats were given a single dose of either 30
mg (1, 2, 10, 17, and 19) or 50 mg (4, 12, and phenytoin) of
compound per kilogram of body weight, and anticonvulsant
activities were determined in the maximal electroshock seizures
(MES) test. No toxicity was observed at this dose, except for
compound 13 at the 4 h time point. Symbols are as follows: ++++,
activity in 75-100% of administered animals; +++, in 50-75%
of animals; ++, in 25-50% of animals; +, 0-25% of animals; and
-, no activity or toxicity.
convulsion, and recovery phases by factors 4, 2, and 7
(versus factors 3, 2, and 8 for phenytoin) (Table 4). On
the other hand, an effect not induced by other anticon-
vulsants tested in previous studies was the 43% de-
crease by compound 17 of the latency period for wild
running (Table 4). Phenytoin induced a 5-fold increase
of this latency period. Like phenytoin, GABAergic
compounds such as diazepam increase this latency
period but to a lesser extent (by 60-70%).16 Ethosux-
imide does not alter this latency period.16
Electr op h ysiology Mea su r em en ts of Com p ou n d
17. Whole-cell electrophysiology studies were performed
using patch clamp technology17 for compound 17. These
methods were employed in an effort to facilitate a better
understanding of potential interactions with receptors’
gated ion channels.12,13 Anticonvulsant effects are often
associated with either excitatory (NMDA and kainate)
or inhibitory (GABAA) neurotransmission. In addition,
Discu ssion
The present design rests on N-phenyl derivatives of
the phthalimide pharmacophore devoid of the 4-amino
substitution (see design 2, Figure 1) which was utilized
in previous series.6-9 Despite omission of this 4-sub-
stituent on the phthalimide nucleus, the design results
in a series of compounds with highly potent leads active
notably in the MES test. SAR studies disclose that, in
these conditions, the best patterns require amino sub-
stitution of the N-phenyl moiety. Combination of the
amino and methyl substitutions of the N-phenyl group
gives apparently more potent compounds with optimal
anti-MES properties conveyed by compound 4 (combined
2-methyl and 3-amino substitutions of the N-phenyl
ring). When the same combined substitutions of the

Anticonvulsant N-Phenylphthalimides
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 7 1315
Ta ble 5. Electrophysiology Experiments of Compound 17a
current responses recorded during
number of cells
(i.e., number of
experiments on
isolated cells)
incubation with compound 17
(expressed as % of control
current responses
concentration of
compound 17 (µM)
type of current responses
evoked by 1 µM GABA
[mean ( SEM])
Whole-Cell Currents of Mouse Cortical Neurons
A
100
30
10
100
100
10
6
5
6
7
118 ( 4
142 ( 6
174 ( 9
99 ( 2
B
C
evoked by 10 µM NMDA + 1 µM glycine
evoked by 100 µM kainate
74 ( 1
Voltage-Dependent Sodium Channels in N1E-115 Neuroblastoma Cells
evoked by depolarizing voltage steps
D
- from -60 mV
- from -90 mV
100
100
3
5
49 ( 9
62 ( 8
a
Results from several individual cells were averaged, SEM was calculated, and statistical significance was determined by Student’s
t-test. All percentages of control current response values (last vertical column) were significantly different than control at p < 0.001.
Additional data concerning voltage-dependent sodium channels in N1E-115 neuroblastoma cells include the percentage of control current
response values evoked by -90 mV for reference antiepileptic drugs at 100 mM; in these conditions, carbamazepine, lamotrigine, and
phenytoin induced values of 78 ( 2, 67 ( 5, and 78 ( 2% of control current responses, respectively. For extensive protocols, see the
Experimental Section.
N-phenyl ring are associated with succinimide (com-
pound 20) instead of phthalimide (compound 4) as the
pharmacophore, only little anticonvulsant protection
may be offered in the MES test. The most potent
compound of our series might be compound 17. It
combines 2-chloro and 4-amino substitutions of the
N-phenyl ring. This better potency might be accounted
for by a more favorable log P (calculated to be 2.24 for
17 versus 1.17 for 1 and 2.08 for phenytoin).
Compared to phenytoin, a prototype antiepileptic drug
largely utilized in human clinics,18 compound 17 was
found at least as efficient as phenytoin in all determina-
tions performed in mice and rats. In vivo determinations
in rats indicate that compound 17 is severalfold more
potent than phenytoin in the MES test (ED50 of 5.7 mg/
kg versus 29.8 mg/kg for phenytoin). In vitro determina-
tions disclose that 17 is 6-fold more efficacious than
phenytoin in counteracting the binding of batrachotoxi-
nin-A-20R-benzoate to rat brain voltage-dependent so-
dium channels (apparent IC50 of 0.15 µM versus 0.93
µM for phenytoin). In mice submitted to the MES test,
the protection offered by 17 was similar to that recorded
on phenytoin, with ED50 values approximating 10 mg/
kg. PI offered in these conditions was better for 17 (PI
superior to 48 versus 6.9 for phenytoin). Similar activi-
ties of 17 and phenytoin in mice were also noticed
during electrophysiology studies dealing with voltage-
dependent sodium channels.
Activity of 17 in the MES test, batrachotoxin affinity
assay, and electrophysiology evaluations indicates that
reduction of seizure spread by blockade of the neuronal
voltage-dependent Na+ channels contributes to the
anticonvulsant protection given by this compound. This
mechanism is therefore common to 17 and phenytoin.
Results obtained in the MDDAS test confirm this
phenytoin-like mode of action of compound 17. But in
this model, the effects of phenytoin and 17 on the initial
component (i.e., latency period) of the audiogenic seizure
test are, however, opposite, suggesting that underlying
mechanisms for anticonvulsant protection are not strictly
the same for the two drugs. Actually, anticonvulsant
mechanisms identified for 17 further include potentia-
tion of GABA-evoked current responses demonstrated
by electrophysiology studies (this effect may be a coher-
ent explanation for the activity of 17 recorded in the
scPtz test in mice dosed intraperitoneally) and inhibition
of kainate-evoked response. Therefore, compound 17
interacts with multiple ion channels and may be of
interest in a wide range of seizure models and neuro-
logic disorders as explained in the remaining part of this
discussion.
Epilepsy is diagnosed on the basis of clinical and
electroencephalogram criteria. In this respect, epilepsy
covers a wide range of distinct clinical disorders which
share in common episodes of synchronous neuronal
hyperdischarge. The synchronous neuronal hyperdis-
charge episode may result from neuronal recruitment
initiated by a neuronal irritative thorn and leading to
well localized epileptic focus (e.g., symptomatic localiza-
tion-related epilepsies) or from general lowering of the
seizure threshold (e.g., status epilepticus). Simply stated,
presently utilized animal seizure models for the former
and latter situations are MES (the electroshock acts as
a general inducer of irritative thorns which through
recruitment result in multiple epileptic foci) and scPtz
(GABA-directed inhibitory toxins results in lowering
seizure threshold) tests, respectively. The former and
latter cited human disorders/animal models respond
well to neuronal voltage-dependent sodium channel
inhibitors (the drugs reduce seizure spread) and
GABAergic compounds (these drugs inhibit neuronal
firing by enhancing seizure threshold), respectively.
These above cited human/animal epileptic conditions
only account for a part of the epileptic disorders. The
reality is complicated by the fact that in practice (i) both
excessive neuronal firing and seizure spread may concur
in a same patient, (ii) focalised epilepsies may second-
arily generalize (see the importance of long-term po-
tentiation and kindled animal models), and (iii) the
seizure threshold is not governed only by GABA recep-
tors but represents a balance between GABAergic and
excitoaminoacidergic pathways. All these considerations
explain why polytherapy has developed in the field of
human epilepsy, reducing the number of CNS surgery
indications (i.e., the number of pharmacoresistant cases).
Alternatively to (but not ruling out) polytherapy, the
recent development of antiepileptic drugs with multiple
cerebral targets has also been successful in reducing the

1316 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 7
Vamecq et al.
addition of water, filtered, dried, and recrystallized from 95%
ethanol: yield of 3, 79.7%; mp 274.5-246.5 °C; TLC Rf 0.49
(acetone-chloroform, 1/1, v:v); 1H NMR (DMSO-d6) δ ppm 7.80
(d, 2.0 H, 9.1 Hz, H-2 + H-6), 7.98 (m, 4.0 H, H-3′ + H-4′),
8.41 (d, 2.0 H, 8.9 Hz, H-3 + H-5); 13C NMR (DMSO-d6) δ ppm
123.8 (C-3′), 124.3 (C-3 and C-5), 127.9 (C-2 and C-6), 131.6
(C-2′), 135.1 (C-4′), 137.9 (C-1), 146.2 (C-4), 166.5 (C-1′); IR
(ν, cm-1) 1782.7, 1732.1 (CdO phthalimide), 1610.3, 1597.6,
1496.1, 1466.4 (phenyl), 1522.3, 1345.3 (NO2). Anal. (C14H8N2O4)
C, H, N.
Reduction of 3 was performed using cyclohexene as hydro-
gen donor, palladium on charcoal (Pd/C, 10%) as catalyst, and
2-propanol as a solvent. The resulting N-(4-aminophenyl)-
phthalimide compound was obtained pure after being recrys-
tallized from ethanol-water (1/1, v:v): yield of 1: 74.7%; mp
183-185 °C; TLC Rf 0.33 (ethyl acetate-chloroform, 1/1, v:v);
1H NMR (DMSO-d6) δ ppm 5.30 (s, 1.6 H, NH2), 6.4-6.7 (m,
3.0 H, H-2 + H-4 + H-6), 7.12 (t, 1.0 H, 7.8 Hz, H-5), 7.90 (m,
4.0 H, H-3′ + H-4′); 13C NMR (DMSO-d6) δ ppm 112.9, 113.8
and 114.7 (C-2 + C-4 + C-6), 123.5 (C-3′), 129.3 (C-5), 131.7
(C-2′), 132.6 (C-1), 134.8 (C-4′), 149.4 (C-3), 167.3 (C-1′); IR
(ν, cm-1) 3420.7, 3346.1, 1634.7 (NH2), 1765.7, 1704.5 (CdO
F igu r e 3. Number attribution of atom positions in formula
of compounds submitted to NMR studies.
number of pharmacoresistant epilepsies, avoiding a pool
of epileptic patients to undergo CNS surgery. Lamot-
rigine (see references 19 and 20 for the mechanisms of
actions and clinical use, respectively) is a good example
of a compound with multiple modes of action and its
clinical use presents promising activity in some human
epilepsies refractory to other antiepileptic drugs. In this
respect, one of the major contributions of the present
work is the demonstration that the emerging compound
17 acts on both seizure spread (inhibition of neuronal
voltage-dependent sodium channels) and seizure thresh-
old. It is noteworthy that lowering of seizure threshold
by compound 17 is mediated by interaction with both
the GABAergic (potentiation effect) and excitoamino-
acidergic (inhibition effect) pathways. Therefore, com-
pound 17 enters adequately the current strategy de-
voted to treating pharmacoresistant epilepsies and
reducing epilepsy-related CNS surgery indications.
phthalimide), 1605.9, 1497.8, 1468.4 (phenyl). Anal. (C14H10
N2O2) C, H, N.
-
N-(3-Am in op h en yl)p h th a lim id e (2). Using a procedure
similar to that of 1, 3-nitroaniline (6.7 g, 49 mmol) and phthalic
anhydride (6.0 g, 41 mmol) provided the title compound, after
recrystallization from ethanol-water (1/1, v:v): yield 73.7%;
mp 244-246 °C; TLC Rf 0.35 (ethyl acetate-chloroform, 1/1,
1
v:v); H NMR (DMSO-d6) δ ppm 5.35 (s, 1.7 H, NH2), 6.64 (d,
2.0 H, 8.7 Hz, H-3 + H-5), 7.01 (d, 2.0 H, 8.4 Hz, H-2 + H-6),
7.88 (m, 4.0 H, H-3′ + H-4′). 13C NMR (DMSO-d6) δ ppm 113.7
(C-3), 119.8 (C-1), 123.4 (C-3′), 128.4 (C-2), 131.8 (C-2′), 134.7
(C-4′), 149.0 (C-4), 167.8 (C-1′); IR (ν, cm-1) 3462.2, 3369.3,
1625.7 (NH2), 1779.1, 1708.5 (CdO phthalimide), 1611.3,
1516.5, 1462.5 (phenyl). Anal. (C14H10N2O2) C, H, N.
Exp er im en ta l Section
N-(3-Am in o-2-m eth ylp h en yl)p h th a lim id e (4) a n d N-(2-
Meth yl-3-n itr op h en yl)p h th a lim id e (5). Using a procedure
similar to that for 1, 2-methyl-3-nitroaniline (9.0 g, 59 mmol)
and phthalic anhydride (8.0 g, 54 mmol) provided the nitro
intermediate 5 after recrystallization from 95% ethanol: yield
92.5%; mp 179-181 °C; TLC Rf 0.52 (acetone-chloroform, 1/1,
Ma ter ia ls. Reagents were purchased from Sigma-Aldrich
(Saint-Quentin Fallavier, France; Bornem, Belgium). Solvents
were obtained from Merck (Darmstadt, Germany). The [3H]-
batrachotoxinin-A-20R-benzoate was obtained from Amersham
International plc, Buckinghamshire, England.
1
v:v); H NMR (DMSO-d6) δ ppm 2.34 (s, 3.0 H, CH3), 7.48 (t,
An a lytica l P r oced u r es. Melting points (uncorrected) were
determined in open capillary tubes on a Totolli apparatus. IR
spectra were recorded on KBr disks using a Spectrum 2000
Perkin-Elmer FT-IR spectrophotometer (FT-IR spectropho-
tometer, model spectrum 2000). NMR spectra were measured
at 25 °C on compounds in solution in DMSO-d6 at 200 MHz
for 1H and 50 MHz for 13C, on a BRUKER AC 200 apparatus.
Chemical shifts are expressed relative to the resonance of
(CD3)2SO at δ 2.49 for 1H NMR and 39.7 for 13C NMR. The
NMR data presented below for each compound refer to atom
positions for which individual attribution number on formula
of compounds is given in Figure 3. HPLC analyses of com-
pounds were performed on a Lichrocart 125-R Licrospher RP
18 column with a flow of 1 mL/min of a methanol-water (65/
35, v:v) mixture as elution solvent and detection at 260 nm.
Elemental analyses (C, H, N) were performed on a Carlo-Erba
EA 1108 elemental analyzer. All compounds had IR and 1H
and 13C NMR spectra consistent with their assigned structure.
Their microanalytical data were within (0.4% of the calculated
figures. TLC data were obtained using aluminum backed
1.0 H, 8.1 Hz, H-5), 7.53-7.85 (m, 5.0 H, H-3′ + H-4′ + H-6),
7.92 (d, 1.0 H, 7.5 Hz, H-4); 13C NMR (DMSO-d6) δ ppm 13.9
and 14.2 (CH3), 123.9 (C-3′), 126.7 (C-5), 129.9 (C-2′ or C-1),
135.8 (C-4′), 151.1 (C-3); IR (ν, cm-1) 1782.8, 1731.7 (CdO
phthalimide), 1600.8, 1575.8 (phenyl), 1524.5, 1354.0 (NO2).
Anal. (C15H10N2O4) C, H, N.
Reduction of 5 provided the title compound 4 after recrys-
tallization from ethanol: yield 38%; mp 246-248 °C; TLC Rf
0.39 (acetone-chloroform, 1/1, v:v); 1H NMR (DMSO-d6) δ ppm
1.79 (s, 3.0 H, CH3), 5.12 (s, 1.67 H, NH2), 6.50 (dd, 1.0 H, 7.7
and 1.1 Hz, H-4), 6.73 (dd, 1.0 H, 8.1 and 1.2 Hz, H-6), 6.99 (t,
1.0 H, 7.9 Hz, H-5), 7.92 (m, 4.0 H, H-3′ + H-4′); 13C NMR
(DMSO-d6) δ ppm 12.2 (CH3), 114.8 (C-4), 116.7 (C-6), 119.9
(C-2), 123.6 (C-3′), 126.4 (C-5), 131.4 and 131.8 (C-1 and C-2′),
134.9 (C-4′), 147.9 (C-3), 167.4 (C-1′); IR (ν, cm-1) 3444.9,
3352.8, 1632.9 (NH2), 1784.2, 1769.7, 1705.9 (CdO phthalim-
ide), 1584.8, 1475.8, 1468.4 (phenyl). Anal. (C15H12N2O2) C, H,
N.
It has been observed that compound 5 was not absolutely
stable, and when solubilized in DMSO for 24 h, part of this
compound gave rise to the corresponding 2-carboxy-benzamide
derivative obtained by opening of the phtahlimide ring. For
experimental determinations including anticonvulsant and
neurotoxic evaluation, compound 5 was utilized immediately
after puting the powder in solution, avoiding the formation of
the hydrolyzed product before use. Hydrolyzed product has not
yet been checked.
N-(4-Am in o-2-m eth ylp h en yl)p h th a lim id e (6) a n d N-(2-
Meth yl-4-n itr op h en yl)p h th a lim id e (7). Using a procedure
similar to that for 1, 2-methyl-4-nitroaniline (10.0 g, 66 mmol)
and phthalic anhydride (9.5 g, 64 mmol) provided the nitro
sheets with silica gel 60 F254
.
Ch em istr y. The nitro compounds 3, 5, 7, 9, and 18 produce
the respective amino compounds 1, 4, 6, 8, 10, and 17. The
general procedure for the synthesis of the aminophenyl-
substituted phthalimides 1, 4, 6, 8, 10, and 17 is illustrated
for the preparation of N-(4-aminophenyl)phthalimide (1) using
N-(4-nitrophenyl)phthalimide (3) as the nitro intermediate.
N-(4-Am in op h en yl)p h t h a lim id e (1) a n d N-(4-Nit r o-
p h en yl)p h th a lim id e (3). A mixture of 4-nitroaniline (6.7 g,
49 mmol) and phthalic anhydride (6.0 g, 41 mmol) in acetic
acid (30 mL) was stirred and heated under reflux for 5 h; the
product of this reaction (i.e., compound 3) was precipitated by

Anticonvulsant N-Phenylphthalimides
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 7 1317
intermediate 7 recrystallized from 95% ethanol: yield 77.8%;
mp 193-195 °C; TLC Rf 0.52 (ethyl acetate-chloroform, 1/1,
heated under reflux for 4-5 h. The solvent was evaporated in
vacuo, and the residual material was recrystallized from
methanol-water (1/1, v:v) giving rise to pure compound: yield
65.8%; mp 205-208 °C; TLC Rf 0.68 (acetone-chloroform-
1
v:v); H NMR (DMSO-d6) δ ppm 2.28 (s, 3.0 H, CH3), 7.70 (d,
1.0 H, 8.6 Hz, H-6), 7.96 (m, 4.0 H, H-3′ + H-4′), 8.20 (dd, 1.0
H, 8.6 and 2.4 Hz, H-5), 8.30 (d, 1.0 H, 2.5 Hz, H-3); 13C NMR
(DMSO-d6) δ ppm 17.8 (CH3), 121.8 (C-3), 123.9 (C-3′), 125.6
(C-6), 130.8 (C-5), 131.9 (C-2′), 135.1 (C-4′), 137.2 (C-2), 139.0
(C-1), 147.6 (C-4), 166.5 (C-1′); IR (ν, cm-1) 1788.9, 1728.2 (Cd
O phthalimide), 1615.7, 1587.1, 1489.8 (phenyl), 1518.8, 1340.1
(NO2). Anal. (C15H10N2O4) C, H, N.
Reduction of 7 provided the title compound 6 recrystallized
from ethanol: yield 77.8%; mp 179-181 °C; TLC Rf 0.39 (ethyl
acetate-chloroform, 1/1, v:v); 1H NMR (DMSO-d6) δ ppm 1.92
(s, 3.0 H, CH3), 5.29 (s, 1.4 H, NH2), 6.45 (dd, 1.0 H, 8.4 and
2.4 Hz, H-5), 6.51 (d, 1.0 H, 2.2 Hz, H-3), 6.89 (d, 1.0 H, 8.4
Hz, H-6), 7.91 (m, 4.0, H-3′ + H-4′); 13C NMR (DMSO-d6) δ
ppm 17.7 (CH3), 111.9 (C-5), 115.2 (C-3), 118.8 (C-1), 123.5
(C-3′), 129.7 (C-6), 131.8 (C-2′), 134.8 (C-4′), 136.5 (C-2), 149.6
(C-4), 167.8 (C-1′); IR (ν, cm-1) 3451.2, 3363.9, 1623.4 (NH2),
1776.5, 1756.9, 1722.8, 1697.6 (CdO phthalimide), 1583.2,
1507.2, 1465.8 (phenyl). Anal. (C15H12N2O2) C, H, N.
1
acetic acid, 49/49/2, v:v:v); H NMR (DMSO-d6) δ ppm 2.0 (s,
6.0 H, CH3), 7.28-7.4 (m, 3.0 H, H-3 + H-4 + H-5), 7.97 (m,
4.0 H, H-3′ + H-4′); 13C NMR (DMSO-d6) δ ppm 17.7 (CH3),
123.9 (C-3′), 128.4 (C-3 and C-5), 129.4 (C-4), 130.1 (C-1), 131.4
(C-2′), 135.2 (C-4′), 136.8 (C-2 and C-6), 167.6 (C-1′); IR (ν,
cm-1) 1779.6, 1759.1, 1738.9, 1709.5 (CdO phthalimide),
1595.8, 1469.8 (phenyl). Anal. (C16H13NO2) C, H, N.
N-(3-Ca r boxy-2-m eth ylp h en yl)p h th a lim id e (13). A mix-
ture of 3-amino-2-methylbenzoic acid (5.0 g, 33 mmol) and
phthalic anhydride (4.75 g, 32 mmol) in acetic acid/dimethyl-
formamide [DMF] (50 mL/30 mL)) was stirred and heated
under reflux for 6 h. The resulting intermediate was precipi-
tated by addition of water, filtered, dried, and recrystallized
from 95% ethanol: yield, 80.9%; mp 307-310 °C; TLC Rf 0.43
(ethyl acetate-chloroform-acetic acid, 49/49/2, v:v:v); 1H NMR
(DMSO-d6) δ ppm 2.1 (s, 3.0 H, CH3), 7.72 (t, 1.0 H, 7.6 Hz,
H-5), 7.79 (d, 1.0 H, 7.6 Hz, H-4 or H-6), 7.9 (m, 5.0 H, H-3′ +
H-4′ + [H-6 or H-4]), 13.1 (s, 0.4 H, COOH); 13C NMR (DMSO-
d6) δ ppm 15.4 (CH3), 123.8 (C-3′), 126.6 (C-5), 130.8 (C-4),
131.8 (C-3), 132.4 (C-2′), 132.8 (C-6), 133.1 (C-1), 135.0 (C-4′),
137.2 (C-2), 167.2 and 168.7 (C-1′ + COOH); IR (ν, cm-1) 2500-
3100, 1687.8 (COOH), 1778.9, 1757.8, 1717.3 (CdO phthal-
imide), 1594.6, 1468.9 (phenyl). Anal. (C16H11NO4) C, H, N.
N-(2-Meth yloxyca r bon ylp h en yl)p h th a lim id e (14). A
mixture of 2-methyloxycarbonyl-aniline (6.1 g, 40 mmol) and
phthalic anhydride (5.0 g, 34 mmol) in acetic acid (30 mL) was
stirred and heated under reflux for 5 h. The resulting
intermediate was precipitated by addition of water, filtered,
dried, and recrystallized from 95% ethanol: Yield 72.2%; mp
158-160 °C; TLC Rf 0.52 (ethyl acetate-chloroform, 1/1, v:v);
1H NMR (DMSO-d6) δ ppm 3.65 (s, 3.0 H, CH3-O-[CdO]-),
7.58 and 7.66 (dd and d,t, 1.0 H, 7.6 and 2 Hz, [H-3 or H-4] +
[H-5 or H-2]), 7.81 (t,d, 1.0 H, 7.6 and 2 Hz, H-4 or H-3), 7.96
(m, 4.0 H, H-3′ + H-4′), 8.05 (dd, 1.0 H, 7.6 and 2 Hz, H-2 or
H-5); 13C NMR (DMSO-d6) δ ppm 52.6 (CH3-O-[CdO]-), 123.8
(C-3′), 128.0 (C-6), 129.6 and 131.0 and 131.1 (C-2 + C-4 +
C-5), 131.8 (C-2′ + C-1), 133.7 (C-3), 135.1 (C-4′), 165.1 and
167.2 (C-1′ + [CdO]-O-); IR (ν, cm-1) 1776.6, 1766.6, 1712.9
(CdO phthalimide), 1601.6, 1496.6 (phenyl), 1712.9, 1276.0,
1122.8 (COOCH3). Anal. (C16H11NO4) C, H, N.
N-(4-Meth oxy-2-m eth ylph en yl)ph th a lim id e (15). A mix-
ture of 4-methoxy-2-methylaniline (6.0 g, 44 mmol) and
phthalic anhydride (6.5 g, 44 mmol) in acetic acid (75 mL) was
stirred and heated under reflux for 5 h. The resulting
intermediate was precipitated by addition of water, filtered,
dried, and recrystallized from 95% ethanol: yield 44.5%; mp
189-190 °C; TLC Rf 0.56 (ethyl acetate-chloroform, 1/1, v:v);
1H NMR (DMSO-d6) δ ppm 2.07 (s, 3.0 H, CH3), 3.79 (s, 3.0 H,
CH3-O-), 6.87 (dd, 1.0 H, 8.6 and 2.5 Hz, H-5), 6.95 (d, 1.0 H,
2.2 Hz, H-3), 7.25 (d, 1.0 H, 8.6 Hz, H-6), 7.92 (m, 4.0 H, H-3′
+ H-4′); 13C NMR (DMSO-d6) δ ppm 17.8 (CH3), 55.5 (CH3-O-
), 112.2 (C-5), 115.8 (C-3), 123.6 (C-3′ + C-1), 130.4 (C-6), 131.8
(C-2′), 134.8 (C-4′), 137.8 (C-2), 159.7 (C-4), 167.4 (C-1′); IR
(ν, cm-1) 1780.8, 1717.2, 1697.2 (CdO phthalimide), 1617.5,
1583.6, 1507.1 (phenyl), 1253.2 (O-CH3). Anal. (C16H13NO3)
C, H, N.
N-(5-Am in o-2-m eth ylp h en yl)p h th a lim id e (8) a n d N-(2-
Meth yl-5-n itr op h en yl)p h th a lim id e (9). Using a procedure
similar to that for 1, 2-methyl-5-nitroaniline (9.6 g, 65 mmol)
and phthalic anhydride (10.0 g, 66 mmol) provided the nitro
intermediate 9 recrystallized from 95% ethanol: yield 71.9%;
mp 232-235 °C; TLC Rf 0.52 (ethyl acetate-chloroform, 1/1,
1
v:v); H NMR (DMSO-d6) δ ppm 2.28 (s, 3.0 H, CH3), 7.71 (d,
1.0 H, 8.6 Hz, H-3), 7.96 (m, 4.0 H, H-3′ + H-4′), 8.26 (dd, 1.0
H, 8.6 and 2.5 Hz, H-4), 8.41 (d, 1.0 H, 2.5 Hz, H-6); 13C NMR
(DMSO-d6) δ ppm 18.0 (CH3), 123.8 (C-3′), 124.0 and 124.7
(C-4 and C-6), 131.9 (C-3′), 132.1 (C-3), 135.0 (C-4′), 145.4 (C-
2), 146.2 (C-5), 166.8 (C-1′); IR (ν, cm-1) 1780.6, 1717.8 (CdO
phthalimide), 1595.0, 1519.7, 1468.2 (phenyl), 1519.7, 1343.4
(NO2). Anal. (C15H10N2O4) C, H, N.
Reduction of 9 provided the title compound 8: 70.6% after
recrystallization from ethanol; mp 166-169 °C; TLC Rf 0.32
(ethyl acetate-chloroform, 1/1, v:v); 1H NMR (DMSO-d6) δ ppm
1.90 (s, 3.0 H, CH3), 5.12 (s, 1.6 H, NH2), 6.46 (d, 1.0, 2.2 Hz,
H-6), 6.60 (dd, 1.0 H, 8.1 and 2.2 Hz, H-4), 7.00 (d, 1.0 H, 8.2
Hz, H-3), 7.92 (m, 4.0 H, H-3′ + H-4′); 13C NMR (DMSO-d6) δ
ppm 16.5 (CH3), 114.4 and 115.1 (C-4 and C-6), 122.5 (C-2),
123.6 (C-3′), 130.9 (C-3), 131.3 (C-1), 131.7 (C-2′), 134.9 (C-
4′), 147.5 (C-5), 167.2 (C-1′); IR (ν, cm-1) 3442.3, 3362.3, 1622.4
(NH2), 1782.7, 1765.8, 1714.1 (CdO phthalimide), 1510.6,
1464.9 (phenyl). Anal. (C15H12N2O2) C, H, N.
N-(6-Am in o-2-m eth ylph en yl)ph th alim ide (10) an d N-(2-
Meth yl-6-n itr op h en yl)p h th a lim id e (11). Using a procedure
similar to that for 1, 2-methyl-6-nitroaniline (10.0 g, 66 mmol)
and phthalic anhydride (9.6 g, 65 mmol) provided nitro
intermediate 11: 31% after recrystallization from 95% ethanol;
mp 171-173 °C; TLC Rf 0.53 (ethyl acetate-chloroform, 1/1,
1
v:v); H NMR (DMSO-d6) δ ppm 2.29 (s, 3.0 H, CH3), 7.71 (t,
1.0 H, 8.0 Hz, H-4), 7.87 (d, 1.0 H, 7.2 Hz, H-3 or H-5), 8.0 (m,
5.0 H, H-3′ + H-4′ + [H-5 or H-3]). 13C NMR (DMSO-d6) δ ppm
18.4 (CH3), 124.3 (C-5), 125.0 (C-3′), 131.3 (C-4), 132.2 (C-2′),
136.4 (C-4′), 137.3 (C-3), 141.3 (C-2), 147.5 (C-6), 167.5 (C-1′);
IR (ν, cm-1) 1780.8, 1742.4, 1713.8 (CdO phthalimide), 1605.0,
1579.3, 1466.9 (phenyl), 1529.5, 1369.2 (NO2). Anal. (C15H10
N2O4) C, H, N.
-
Reduction of 11 provided the title compound 10: 53.7% after
recrystallization from ethanol; mp 189-190 °C; TLC Rf 0.44
(ethyl acetate-chloroform, 1/1, v:v); 1H NMR (DMSO-d6) δ ppm
1.88 (s, 3.0 H, CH3), 5.30 (s, 1.2 H, NH2), 6.46 and 6.59 (d, 1.0
H, 7.3 and 7.2 Hz, H-3 and H-5), 7.01 (t, 1.0 H, 7.6 Hz, H-4),
7.91 (m, 4.0 H, H-3′ + H-4′); 13C NMR (DMSO-d6) δ ppm 17.6
(CH3), 113.3 (C-5), 115.2 (C-1), 117.0 (C-3), 123.4 (C-3′), 129.6
(C-4), 132.3 (C-2′), 134.5 (C-4′), 137.0 (C-2), 146.9 (C-6), 167.6
(C-1′); IR (ν, cm-1) 3435.7, 3345.4, 1623.2 (NH2), 1780.5, 1760.4,
1736.6, 1708.9 (CdO phthalimide), 1597.5, 1487.3, 1473.9
(phenyl). Anal. (C15H12N2O2) C, H, N.
N-(2,4-Dim eth oxyp h en yl)p h th a lim id e (16). A mixture
of 2,4-dimethoxyaniline (9.5 g, 62 mmol) and phthalic anhy-
dride (9.1 g, 61 mmol) in acetic acid (50 mL) was stirred and
heated under reflux for 5 h. The resulting intermediate was
precipitated by addition of water, filtered, dried, and recrystal-
lized from 95% ethanol: yield 75.7%; mp 228-231 °C; TLC Rf
0.52 (ethyl acetate-chloroform, 1/1, v:v); 1H NMR (DMSO-d6)
δ ppm 3.74 (s, 3.0 H, CH3-O-), 3.80 (s, 3.0 H, CH3-O-), 6.62
(dd, 1.0 H, 8.2 and 2.5 Hz, H-5), 6.73 (d, 1.0 H, 2.2 Hz, H-3),
7.25 (d, 1.0 H, 8.0 Hz, H-6), 7.92 (m, 4.0 H, H-3′ + H-4′); 13C
NMR (DMSO-d6) δ ppm 55.6 (CH3-O-), 55.9 (CH3-O-), 99.3
(C-3), 105.2 (C-5), 112.9 (C-1), 123.5 (C-3′), 130.9 (C-6), 131.6
(C-2′), 134.8 (C-4′), 156.3 (C-2), 161.3 (C-4), 167.3 (C-1′); IR
(ν, cm-1) 1781.6, 1761.9, 1730.9, 1710.5 (CdO phthalimide),
N-(2,6-Dim eth ylp h en yl)p h th a lim id e (12). A mixture of
2,6-dimethylaniline (5.0 g, 41 mmol) and phthalic anhydride
(18.3 g, 123 mmol) in acetic acid (30 mL) was stirred and

1318 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 7
Vamecq et al.
1613.3, 1588.9, 1514.6 (phenyl), 1216.4 (-O-CH3). Anal.
(C16H13NO4) C, H, N.
KCl, 1 mM CaCl2, 10 mM HEPES, 10 mM glucose, and 20
mM sucrose (320 mOsm, pH 7.4).17 The bathing solution also
contained 200 nM strychnine to block the glycine receptor and
200-500 nM tetrodotoxin to block voltage-gated sodium chan-
nels. For experiments in which GABA receptor currents were
examined, the bathing solution also contained 1 mM MgCl2
to block NMDA receptor currents.
For experiments on glutamate receptor current, the bathing
solution contained 1 µM glycine (a required co-agonist at the
NMDA receptor), and 10 µM picrotoxin was added to block
GABAA channels. Whole-cell voltage clamp recordings were
obtained with an Axopatch 200 amplifier (Axon instruments)
using patch electrodes (2-3 Ω) filled with an intracellular
solution containing 153 mM CsCl, 10 mM EGTA, 10 mM
HEPES, and 4 mM MgCl2 (290 mOsm, pH 7.4). Currents were
filtered at 1-2 kHz, digitally sampled at 1 kHz, and acquired
on computer using Axotape software (Axon instruments).
Currents were also acquired on a chart recorder.
Cells were voltage clamped at -60 mV. Agonist and test
compound were applied using a rapid perfusion system that
consisted of a gravity fed multibarreled microperfusion pipet
that was positioned 200-400 cm from the cell. Solution
exchange was controlled by solenoid valves with a valvebank
8 controller (Automate Scientific). GABAA receptor currents
were evoked using 1 µM GABA, whereas ionotropic glutamate
currents were evoked using 10 µM NMDA and 100 µM kainate.
Agonists were applied for 1-5 s and separated by a 20-30 s
wash period. Within this protocol, GABA and glutamate
receptor currents were relatively stable for the duration of the
recording period.
The effects of compound 17 on GABA currents were evalu-
ated using a pretreatment paradigm which allows measure-
ment of the maximal effects of the compound without the
confounding effects of GABA receptor desensitization. Follow-
ing at least two control applications of GABA, the cell was
perfused with compound 17 (100 µM) for 20-30 s prior to
application of GABA plus the drug. Following the combined
GABA and drug application, the perfusate was then switched
back to drug alone. This sequence (drug/drug + agonist/drug)
was repeated until at least two consistent currents were
observed. The perfusate was then returned to the control
bathing solution, and additional GABA responses were ob-
tained to monitor return of the GABA response to control
levels. This protocol allowed measurement of both the direct
and modulatory actions of the test compound at GABAA
receptors. An identical protocol to that described for GABA
was employed to assess the interaction of compound 17 with
the ionotropic glutamate receptors gated by NMDA and
kainate.
N-(4-Am in o-2-ch lor op h en yl)p h th a lim id e (17) a n d N-(2-
Ch lor o-4-n itr op h en yl)p h th a lim id e (18). Using a procedure
similar to that for 1, 2-chloro-4-nitroaniline (12.0 g, 70 mmol)
and phthalic anhydride (10.2 g, 69 mmol) provided nitro
intermediate 18 recrystallized from 95% ethanol: yield 52.8%;
mp 180-182 °C; TLC Rf 0.52 (ethyl acetate-chloroform, 1/1,
v:v); 1H NMR (DMSO-d6) δ ppm 7.8-8.2 (m, 5.0 H, H-6 + H-3′
+ H-4′), 8.39 (dd, 1.0 H, 8.7 and 2.5 Hz, H-5), 8.54 (d, 1.0 H,
2.5 Hz, H-3); 13C NMR (DMSO-d6) δ ppm 123.4 (C-5), 124.2
(C-3′), 125.2 (C-3), 131.6 (C-2′), 132.6 (C-6), 133.6 (C-2), 135.4
(C-4′), 135.7 (C-1), 148.5 (C-4), 165.9 (C-1′); IR (ν, cm-1) 1785.7,
1733.0, 1714.2 (CdO phthalimide), 1590.6, 1597.1, 1481.5,
1466.5 (phenyl), 1524.0, 1338.9 (NO2). Anal. (C14H7ClN2O4) C,
H, N.
Reduction of 18 provided the title compound 17, after
recrystallization from acetic acid: yield 17.3%; mp 205-207
°C; TLC Rf 0.44 (ethyl acetate-chloroform, 1/1, v:v); 1H NMR
(DMSO-d6) δ ppm 5.70 (s, 1.24 H, NH2), 6.60 (dd, 1.0 H, 8.6
and 1.5 Hz, H-5), 6.76 (d, 1.0 H, 1.4 Hz, H-3), 7.14 (d, 1.0 H,
8.6 Hz, H-6), 7.93 (m, 4.0 H, H-3′ + H-4′); 13C NMR (DMSO-
d6) δ ppm 112.9 and 113.5 (C-3 + C-5), 116.35 (C-1), 123.7
(C-3′), 131.6 (C-6), 132.5 (C-2′), 135.1 (C-4′), 151.2 (C-4), 167.3
(C-1′); IR (ν, cm-1) 3433.9, 3352.5, 1646.6 (NH2), 1778.8, 1758.8,
1725.6, 1703.8 (CdO phthalimide), 1600.0, 1502.0, 1466.2
(phenyl). Anal. (C14H9ClN2O2) C, H, N.
N-[3-Meth yl-(2-p yr id in yl)]p h th a lim id e (19). A mixture
of 2-amino-3-picoline (5.1 g, 47 mmol) and phthalic anhydride
(7.6 g, 51 mmol) in acetic acid (30 mL) was stirred and heated
under reflux for 5-6 h. The resulting intermediate was
precipitated by addition of water, filtered, dried, and recrystal-
lized from 95% ethanol: yield 36.4%; mp 169-171 °C; TLC Rf
0.42 (ethyl acetate-chloroform, 1/1, v:v); 1H NMR (DMSO-d6)
δ ppm 2.18 (s, 3.0 H, CH3), 7.48 (dd, 1.0 H, 7.6 and 4.7 Hz,
H-4), 7.96 (m, 5.0 H, H-3′ + H-4′ + H-5), 8.46 (dd, 1.0 H, 4.7
and 1.5 Hz, H-3); 13C NMR (DMSO-d6) δ ppm 17.4 (CH3), 124.8
(C-3′), 125.9 (C-5 or C-4), 132.2 and 133.3 (C-2′ + [C-1 or C-6]),
136.1 (C-4′), 141.1 (C-4 or C-5), 145.7 (C-6 or C1), 148.2 (C-3),
167.4 (C-1′); IR (ν, cm-1) 1786.3, 1763.8, 1714.5 (CdO phthal-
imide). Anal. (C14H10N2O2) C, H, N.
N-(3-Am in o-2-m eth ylp h en yl)su ccin im id e (20). A mix-
ture of 2-methyl-3-nitroaniline (9.0 g, 59 mmol) and succinic
anhydride (6.3 g, 63 mmol) in acetic acid (50 mL) was stirred
and heated under reflux for 5 h. The resulting intermediate
was precipitated by addition of water, filtered, dried, and
recrystallized from 95% ethanol. Reduction of the nitro group
was performed using cyclohexene (hydrogen donor) and pal-
ladium on charcoal (Pd/C, 10%) (catalyst) in 2-propanol
(solvent). The resulting N-(3-amino-2methylphenyl)succin-
imide compound was obtained pure after being recrystallized
from ethanol: yield 50.6%; mp 178-180 °C; TLC Rf 0.15 (ethyl
acetate-chloroform, 1/1, v:v); 1H NMR (DMSO-d6) δ ppm 1.75
(s, 3.0 H, CH3), 2.79 (m, 4.0 H, CH2-CH2), 5.04 (s, 1.6 H, NH2),
6.30 (d, 1.0 H, 7.6 Hz, H-4), 6.67 (d, 1.0 H, 8.0 Hz, H-6), 6.95
(t, 1.0 H, 7.8 Hz, H-5); 13C NMR (DMSO-d6) δ ppm 12.0 (CH3),
28.7 (CH2-CH2), 114.5 (C-4), 116.0 (C-6), 119.3 (C-2), 126.25
The agonist-evoked current values were measured using the
axotape software package or obtained directly from the chart-
paper records. The effect of compound 17 on agonist-evoked
currents was determined by comparing the agonist currents
in the presence of drug with control agonist responses in the
absence of drug.
Volta ge-Ga ted Na + Ch a n n el. The effect of compound 17
on voltage-gated Na+ channels was assessed using N1E-115
neuroblastoma cells and whole-cell voltage clamp recording
techniques. The N1E-115 neuroblastoma cell line was main-
tained at 35 °C in Dulbecco’s modified Eagles Medium
supplemented with 5% fetal calf serum, 20 mM HEPES, 80
µg/mL gentamycine, and 4 mM glutamine. Prior to electro-
physiological studies, cells were plated and incubated for 3-5
days in a differentiation medium similar to above with reduced
(2.5%) fetal calf serum and 2% DMSO. Recordings were carried
out at room temperature in a bathing solution containing 130
mM NaCl, 5 mM KCl, 1.5 mM CaCl2, 1 mM MgCl2, 5 mM
glucose, and 5 mM HEPES. The buffer also contained 0.1 mM
CdCl2 and 25 mM tetraethylamonium chloride to block voltage-
gated Ca2+ and K+ channels, respectively. Whole-cell record-
ings were obtained using patch electrodes (1-2 MΩ) filled with
the intracellular solution described above. The currents were
filtered at 5 kHz and acquired on computer using PClamp 6
(Axon Instruments). Series resistance and capacitive currents
(C-5), 132.4 (C-1), 147.8 (C-3), 177.2 (-[CdO]-N-); IR (ν, cm-1
)
3433.3, 3355.9, 1624.9 (NH2), 2991.4, 2938.2, 2911.2, 2863.0
(CH3, CH2), 1781.0, 1703.3 (CdO succinimide), 1587.6, 1491.5,
1478.5 (phenyl). Anal. (C11H12N2O2) C, H, N.
In Vivo a n d in Vitr o Exp er im en ts. Anticonvulsant
evaluations in the MES and scPtz tests, determination of
neurotoxicity (i.e., minimal neurotoxicity evaluated by the
rotarod test), and batrachotoxin affinity assay procedure were
performed as described previously.11 MDDAS test was also
described elsewhere.16
Whole-cell electrophysiology studies were performed using
patch clamp technology.17 Whole-cell voltage clamp data
recordings were obtained using cortical cells cultured from 15-
gestational-day-old mouse fetuses used 2-3 weeks after plat-
ing.21,22
Recordings were carried out at room temperature (23 °C)
in a control bathing solution containing 142 mM NaCl, 1.5 mM

Anticonvulsant N-Phenylphthalimides
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 7 1319
(10) Friderichs, E. Studies on the CNS activity profile of thalidomide
analogues. Arzneimittelforschung 1982, 32, 613-620
were compensated using the internal clamp circuitry. The
series resistance was 3-5 MΩ, and 80-90% of the series
resistance was compensated. Cells were voltage clamped at
-70 mV, and compound 17 was applied using the perfusion
system described above. To activate voltage-gated sodium
channels, cells were hyperpolarized to -90 mV for 90 ms and
then depolarized to potentials from -80 to +60 mV in 10 mV
increments in control and at 1 and 2 min following incubation
with test compound. The process was repeated, hyperpolariz-
ing at -60 mV.
In the electrophysiology experiments, the results from
several individual cells were averaged, the SEM was calcu-
lated, and statistical significance was determined using the
Student’s t-test. Significance was taken to be in the range of
p values p < 0.01.23
(11) Vamecq, J .; Lambert, D.; Poupaert, J . H.; Masereel, B.; Stables,
J . P. Anticonvulsant activity and interactions with neuronal
voltage-dependent sodium channel of 15 analogues of ameltolide,
a novel antiepileptic drug. J . Med. Chem. 1998, 41, 3307-3313.
(12) White, H. S.; J ohnson, M.; Wolf, H. H.; Kupferberg, H. J . The
early identification of anticonvulsant activity: role of the
maximal electroshock and subcutaneous pentylenetetrazol sei-
zure models. Ital. J . Sci. 1995, 16, 73-77.
(13) White, H. S.; Woodhead, J . H.; Franklin, M. R. General
Principles: Experimental Selection, Quantification, and Evalu-
ation of Anticonvulsant Drugs. In Antiepileptic drugs, 4th ed.;
Levy, R., Mattson, R., Meldrum, B. S., Eds.; Raven Press: New
York, 1995; pp 99-110.
(14) Stables, J . P.; Kupferberg, H. J . The NIH Anticonvulsant Drug
Development (ADD) Program: Preclinical Anticonvulsant Screen-
ing Project. In: Molecular and Cellular Targets for Antiepileptic
drugs; Avanzini, G., Tanganelli, P., Avoli, M., Eds.; J ohn Libbey
& Company Ltd.: London, 1997; pp 191-198. .
Ack n ow led gm en t. The present work was supported
by grants from the Comite´ Directeur de la Recherche
(Faculty of Medicine and University Hospital of Lille,
France). J .V. and P.M. are sponsored by the French
INSERM. The authors thank Dr. H. Wolfe, Dr. S. White,
and Ms. J . Woodhead at the University of Utah for their
critical role in anticonvulsant screening.
(15) Kupferberg, H. K., Stables, J . P. Mechanism of Action Revis-
ited: Drug Discovery, Testing and Clinical Prediction. In Chal-
lenge Epilepsy - New Anticonvulsant Drugs; Stefan, H., Kramer,
G., Mamoli, B., Eds.; Blackwell Science Ltd.: Boston, 1998; pp
7-29.
(16) Bac, P.; Maurois, P.; Dupont, C.; Pages, N.; Stables, J . P.;
Gressens, P.; Evrard, Ph.; Vamecq, J . Magnesium deficiency-
dependent audiogenic seizures (MDDASs) in adult mice:
a
Refer en ces
nutritional model for discriminatory screening of anticonvulsant
drugs and original assessment of neuroprotection properties. J .
Neurosci. 1998, 18, 4363-4373.
(1) Clark, C. R.; Wells, M. J . M.; Sansom, R. T.; Norris, G. N.;
Dockens, R. C.; Ravis, W. R. Anticonvulsant activity of some
4-aminobenzamides. J . Med. Chem. 1984, 27, 779-782.
(2) Clark, C. R.; Sansom, R. T.; Lin, C. M.; Norris, G. N. Anticon-
vulsant activity of some 4-aminobenzanilides. J . Med. Chem.
1985, 28, 1259-1262.
(3) Clark, C. R.; Lin, C. M.; Sansom, R. T. Anticonvulsant activity
of 2- and 3-aminobenzanilides. J . Med. Chem. 1986, 29, 1534-
1537.
(4) Clark, C. R.; Davenport, T. W. Synthesis and anticonvulsant
activity of analogues of 4-amino-N-(1-phenylethyl)benzamide. J .
Med. Chem. 1987, 30, 1214-1218.
(5) Clark C. R. Comparative anticonvulsant activity and neurotox-
icity of 4-amino-N-(2,6-dimethylphenyl)benzamide and prototype
antiepileptic drugs in mice and rats. Epilepsia 1988, 29, 198-
203.
(6) Bailleux, V.; Valle´e, L.; Nuyts, J . P.; Vamecq, J . Anticonvulsant
activity of some 4-amino-N-phenylphthalimides and N-(3-amino-
2-methylphenyl)phthalimides. Biomed. Pharmacother. 1994, 48,
95-101.
(7) Bailleux, V.; Valle´e, L.; Nuyts, J . P.; Vamecq, J . Synthesis and
anticonvulsant activity of some N-phenylphthalimides. Chem.
Pharm. Bull. 1994, 42, 1817-1821.
(8) Bailleux, V.; Valle´e, L.; Nuyts, J . P.; Hamoir, G.; Poupaert, J .
H.; Stables, J . P.; Vamecq, J . Comparative anticonvulsant
activity and neurotoxicity of 4-amino-N-(2,6-dimethylphenyl)-
phthalimide and prototype antiepileptic drugs in mice and rats.
Epilepsia 1995, 36, 559-565.
(9) Poupaert, J .; Hamoir, G.; Barbeaux, P.; Lambert, D.; He´nichart,
J . P. Anticonvulsant activity of some N-phenylphthalimide
derivatives in rats and mice. J . Pharm. Pharmacol. 1995, 47,
89-91.
(17) Hamil, O. P.; Marty, A.; Neher, E.; Sakmann, B.; Sigworth, F.
Improved patch clamp technique for high-resolution current
recordings from cells and cell free membranes. Pflugers Arch.
1981, 391, 85-100.
(18) Macdonald, R. L. Antiepileptic drug actions. Epilepsia 1989, 30,
S19-S28.
(19) Leach, M. J .; Lees, G.; Riddall, D. R. Lamotrigine - Mechanisms
of Action. In Antiepileptic Drugs, 4th ed.; Levy, H. L., Mattson,
R. H., Meldrum, B. S., Eds.; Raven Press: New York, 1995; pp
861-869.
(20) Leach, J . P.; Brodie, M. J . Lamotrigine - Clinical Use. In
Antiepileptic Drugs, 4th ed.; Levy, H. L., Mattson, R. H.,
Meldrum, B. S., Eds.; Raven Press: New York, 1995; pp 889-
895.
(21) Hertz, E.; Yu, A. C. H.; Hertz, L.; J uurlink, B. H. J .; Schousboe,
A. Preparation of primary cultures of mouse cortical neurons.
In A Dissection and Tissue Culture Manual of the Nervous
System; Shahar, A., Devillis, J ., Vernadakis, A., Haber, B., Eds.;
Alan R. Liss, Inc.: New York, 1990; pp 183-186.
(22) Skeen, G. A.; White, H. S.; Twyman, R. E. The dihydropyridine
reduces N-methyl-D-aspartate (NMDA)-evoked currents of ro-
dent cortical neurons through a direct interaction with the
NMDA receptor-associated ion channel. J . Pharmacol. Exp. Ther.
1994, 271, 30-38.
(23) Tallarida, R. J .; Murray, R. B. Manual of Pharmacological
Calculations with computer programs, 2nd ed.; Raven Press
Ltd.: New York, 1986.
J M990068T