Synthesis, characterization and anticonvulsant activity of enaminones. Part 6: Synthesis of substituted vinylic benzamides as potential anticonvulsants

Article abstract of DOI:10.1016/S0968-0896(99)00185-6

A comparison of enaminones from various unsubstituted and p-substituted benzamides to the analogous benzylamines has been undertaken with the aim of elucidating the essential structural parameters necessary for anticonvulsant activity. Initial studies on methyl 4-N-(benzylamino)-6-methyl-2-oxocyclohex- 3-en-1-oate, 3a, 3-N-(benzylamino)cyclohex-2-en-1-one, 3p, and 5,5-dimethyl- 3-N-(benzylamino)-cyclohex-2-en-1-one, 3r indicated that benzylamines possessed significant anti-maximal electroshock seizure (MES) activity. Evaluation of the analogous benzamides revealed significant differences in anticonvulsant activity, these differences were most probably related to the differences in their three-dimensional structures. (C) 1999 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(99)00185-6

Bioorganic & Medicinal Chemistry 7 (1999) 2415±2425
Synthesis, Characterization and Anticonvulsant Activity of
Enaminones. Part 6: Synthesis of Substituted Vinylic Benzamides
as Potential Anticonvulsants
James E. Foster, a Jesse M. Nicholson, a Raymond Butcher, a James P. Stables, b
Ivan O. Eda®ogho, c Angela M. Goodwin, c Michael C. Henson, c Carlynn A. Smith c
and K. R. Scott c,*
aDepartment of Chemistry, Graduate School of Arts and Sciences, Howard University, Washington, DC 20059, USA
bEpilepsy Branch, Division of Convulsive, Developmental and Neuromuscular Disorders,
National Institute of Neurological Disorders and Stroke, Bethesda, MD 20892, USA
cDepartment of Pharmaceutical Sciences, College of Pharmacy, Nursing and Allied Health Sciences, Howard University,
Washington, DC 20059, USA
Received 9 March 1999; accepted 18 May 1999
AbstractÐA comparison of enaminones from various unsubstituted and p-substituted benzamides to the analogous benzylamines
has been undertaken with the aim of elucidating the essential structural parameters necessary for anticonvulsant activity. Initial
studies on methyl 4-N-(benzylamino)-6-methyl-2-oxocyclohex-3-en-1-oate, 3a, 3-N-(benzylamino)cyclohex-2-en-1-one, 3p, and 5,5-
dimethyl-3-N-(benzylamino)-cyclohex-2-en-1-one, 3r indicated that benzylamines possessed signi®cant anti-maximal electroshock
seizure (MES) activity. Evaluation of the analogous benzamides revealed signi®cant di€erences in anticonvulsant activity, these
di€erences were most probably related to the di€erences in their three-dimensional structures. # 1999 Elsevier Science Ltd. All
rights reserved.
Introduction
Despite optimal use of available antiepileptic drugs
marketed in the United States, many patients with epi-
lepsy fail to experience seizure control and others do so
only at the expense of signi®cant toxic side e€ects. Esti-
mates suggest that available medication controls the
seizures in only 50% of patients or decreases the inci-
dence in only 75% of patients.1 A new group of unique
drugs have entered the antiepileptic armamentarium.
These include felbamate, lamotrigine, oxacarbazine,
dezinamide, gabapentin, topiramate, vigabatrin, and
zonisamide.2 In addition, a new series of compounds
have been reported, most notably the aroyl(amino-
acyl)pyrroles, 1,3 and the 3-aminopyrroles, 2.4 The most
active compound in each series is shown below.
Carson et al.3 indicated in their report that the struc-
tural similarity of the aroyl(aminoacyl)-pyrroles to the
N-benzyl enaminones, 3 (Table 1), synthesized in our
laboratories5±10 was pervasive in their synthetic e€orts.
Further, Carson indicated the extensive charge deloca-
lizaton involving the nitrogen atom and the carbonyl
oxygen was analogous in both the acyl pyrroles11 and
enaminones.12 Further molecular modeling studies of 1
with 3a (Fig. 1) by Carson et al. showed a congruent
spatial relationship of the phenyl ring, nitrogen atoms
and carbonyl groups, indicating RMS deviation=0.38
A; volume 1=320 A; volume 3b=254 A; common
volume=168 A.3
Key words: Anticonvulsants; NMR; X-ray crystal structures; mole-
cular modeling/mechanics.
* Corresponding author. Tel.: +1-202-806-7288; fax: +1-202-806-
4636; e-mail: kscott@fac.howard.edu
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00185-6

2416
J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
The Carson report provided an impetus to reinvestigate
the N-benzylamine enaminone series and to synthesize
the analogous benzamides, 4, the latter functionality
which had previously been shown to possess signi®cant
anticonvulsant activity in their own right in other
laboratories.13±16 We had reported an initial study of
the enaminones of the benzylamines,5,6 however, this
series was considerably abbreviated due to signi®cant
activity shown with the aniline derivatives which
engaged our synthetic e€orts. Compound 3a (R1=CH3,
R2=H, R3=CO2CH3, X=H) was the most potent in
the series, displaying an intraperitoneal (ip) ED50 in
mice of 64.7 mg/kg and a TD50 >500 mg/kg and an
oral (po) ED50 in rats of 26.8 mg/kg and no toxicity
noted at dosages up to 500 mg/kg. In addition, in the
corneal kindling model 3a provided an ED50 of 78.3 mg/
kg, compared to phenytoin, 48.3 mg/kg, under the same
conditions.5 A brief structure activity relationship was
attempted. It was found in the same carbomethoxy ser-
ies, para substitution led to less active or inactive com-
pounds.5 However, the p-¯uoro compound (R1=CH3,
R2=H, R3=CO2CH3, X=F) was also active, providing
an ip ED50 of 159 mg/kg in mice, and a po ED50 of 49.3
mg/kg and a TD50 >230 mg/kg in rats.5 We postulated
that the activity of this analogue veri®ed Topliss's ®nd-
ings that p-¯uoro substitution produced a minimal
change in s and p e€ects compared to the unsubstituted
compound.5 In addition, the p-chloro (R1=CH3,
R2=H, R3=CO2CH3, X=Cl; +s, +p) provided
spurious results, probably due to solubility problems,
while the p-tolyl (R1=CH3, R2=H, R3=CO2CH3,
X=CH3; s, +p) and p-carboxy (R1=CH3, R2=H,
R3=CO2CH3, X=CO2H; +s, p) analogues were
inactive. Further, in the dimedone series, the unsub-
stituted benzylamine compound (3r, R1=R2=CH3,
R3=H, X=H), was active with an ED50 of 53 mg/kg
and a TD50 of 148 mg/kg in mice,5 while the comparable
cyclohexene derivative (3p, R1=R2=R3=H, X=H)
was less active.6 Additionally, it was noted that repla-
cing one of the benzylic protons with a methyl group,
producing a chiral a-phenethyl side chain, yielded iso-
mers both of which were inactive.6 In contrast, the b-
phenethyl side chain was highly active.6 We herein
report further studies on these benzylamino analogues
and the comparable vinylic benzamides.
Table 1. Benzylamines
Compound R1 R2
R3
X
Yield MP, ꢀC Clog Pa
3ab
3bd
3c
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
H
CH3
CH3 CH3
H
CH3
CH3 CH3
H
CH3
CH3 CH3
H
CH3
CH3 CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CO2CH3
CO2C2H5
CO2C(CH3)3
CO2CH3
CO2C2H5
CO2C(CH3)3 Cl
CO2CH3
CO2C2H5
CO2C(CH3)3 CH3
CO2CH3 OCH3 77 168.5±172e 2.71
CO2C2H5 OCH3 74
CO2C(CH3)3 OCH3 38
H
H
H
Cl
Cl
90
77
55
64
69
47
48
76
58
154±155c 2.79
134±135f 3.32
152±155c 4.03
173±174g 2.79
169±172h 4.03
182±185f 4.74
160±163e 3.29
134±135e 3.82
123±126e 4.53
Chemistry
3de
3e
Cyclic enaminone esters of the benzylamine series, 3
(Table 1), were synthesized from b-hydroxy keto esters,
as previously reported.5±10 The synthetic pathway is
shown in Scheme 1. The condensation reaction of ethyl
crotonate with the respective acetoacetic ester was
modi®ed as reported by Friary and co-workers for the
synthesis of the 4-carbo-tert-butoxy-5-methylcyclo-hex-
ane-1,3-dione.17 It was noted that in the synthesis of the
methyl ester, if sodium ethoxide was employed as the
base, a transesteri®cation reaction occurred and the
ethyl ester was obtained. 5-Methyl-1,3-cyclohexane-
dione was synthesized by the decarboxylation of
4-carbo-tert-butoxy-5-methylcyclohexane-1,3-dione as
reported by Friary and co-workers.17 The subsequent
condensation of the b-keto esters with the appropriate
benzylamine was varied to maximize overall yields.
3f
3gb
3h
CH3
CH3
3i
3j
3k
3l
3m
3n
154±157e 3.24
174±175c 3.95
204±208c 2.22
184±186h 2.75
172±175c 3.46
125±127f 2.41
CO2CH3
CO2C2H5
CO2C(CH3)3 CN
CN
CN
44
38
37
25
3o
3pe
3q
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
Cl
Cl
CH3
CH3
CH3
OCH3 35
OCH3 68
OCH3 82
CN
CN
CN
25 137.5±139f 2.93
3rb
3s
15
48
44
47
39
32
59
124±127f 3.45
170±172h 3.12
186±187h 3.64
159±162c 4.16
153±155c 2.91
146±148f 3.43
139±140f 3.95
159±160g 2.33
160±162h 2.85
159±160g 3.37
156±159h 1.84
H
H
3t
3u
3v
3w
3x
3y
H
H
H
H
3z
3aa
3bb
3cc
3dd
H
CH3
CH3 CH3
H
H
57
34
43
168±171i
215±217j
2.36
2.90
a
CLog P calculated on the protonated form.
Ref. 5.
Ethyl acetate.
Ref. 9.
b
c
d
e
f
Ref. 6.
Ethyl acetate-petroleum ether (bp 38±54ꢀC).
Methanol.
2-Propanol.
g
h
i
Figure 1. Sites of conguent spatial relationship (arrows) of the aroyl-
(aminoacyl)pyrroles, 1 (ref 3) and methyl 4-N-(benzylamino)-6-meth-
yl-2-oxocyclohex-3-en-1-oate, 3a (ref 3).
Ethyl acetate±ethanol (95%).
2-Propanol±acetone.
j

J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
2417
Scheme 1. Synthesis of benzylamino enaminone esters and 5-methyl-cyclohexane analogues. Conditions: (a) NaOEt or NaOMe, Á; (b) substituted
benzylamine; (c) H2SO4, Á (ref 17) (see Experimental).
Previous studies indicated that a 1:1 ratio of absolute
ethanol:ethyl acetate was employed as the solvent mix-
ture.8 The benzamides, 4 (Table 2), were synthesized by
amination of the respective b-diketones18 followed by
acylation with the corresponding benzoyl chlorides with
triethylamine as the acid scavenger. Careful monitoring
of this latter reaction was undertaken to minimize the b-
acylation (5) side product.12 However, the yields of
benzamides 4 were lower when compared to the benzyl-
amines. In view of the reported lability of the tert-
butoxy group,10 an excess of triethylamine was used in
these reactions and the isolated product was not
extracted with aqueous acid and base as indicated for
the methyl and ethyl esters, but evaporated to dryness
and chromatographed on a silica gel column. The syn-
thetic pathway is shown in Scheme 2.
extended sofa conformation without intramolecular
vinyl proton involvement. A summary of the atomic
coordinates of 4j and 4cc are provided in Tables 3 and 4.
High Field NMR
To further establish a structural di€erence between
the benzylamines and benzamides, a high ®eld NMR
study was performed. We had previously shown6,7 that
the vinyl proton was deshielded due to its hydrogen-
bonding to the aromatic ring in the aniline series. In
the benzamide series, we propose that the carbonyl
oxygen forms a hydrogen bond with this proton. As
was shown from X-ray analysis of benzylamine 3a,
however, hydrogen bonding of the vinyl proton, unlike
that shown with the aniline series, does not occur.
Table 5 displays a comparison of NMR data of the
vinyl proton assignments in three individual sets of
analogues. As noted, the benzamides were all deshielded
and appeared down®eld compared to the comparable
benzylamines, thus further verifying these structural
di€erences.
X-ray Crystallography
In view of our previous reports on the intramolecular
hydrogen bonding of the vinyl hydrogen and the aromatic
ring,6,7 an X-ray di€raction study of two representative
benzamides was performed. As noted in Fig. 2 of 4j and
Fig. 3 of 4cc, strong hydrogen bonding occurred
between the vinyl proton and the carbonyl oxygen of
the amide group in each structure, providing a pseudo
three-ring structure. This pseudo three-ring con®gura-
tion was not observed with the comparable benzyl-
amines. Codding et al.,19 in the X-ray analysis of 3a
indicated intermolecular N-H...O bonding, providing an
Molecular Modeling
Molecular modeling studies were performed on the
active Carson's pyrrole, 1,3 the most active 3-amino-
pyrrole, 24 and the three active benzylamines (3a, 3q and

2418
J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
Table 2. Benzamides
Drug Development (ADD) Program, Epilepsy Branch,
Neurological Disorders Program, National Institute of
Neurological Disorders and Stroke (NINDS). These
testing procedures have been described.22±24 Phase I
evaluation of the reported benzylamines and benza-
mides involved three tests: maximal electroshock seizure
(MES), subcutaneous pentylenetetrazol (ScMet), and
neurologic toxicity (Tox) in mice. Phase I data for the
benzylamines and the comparable benzamides are
shown in Table 6. As previously reported,5,6,8 to di€er-
entiate the results between di€erent rodent species, the
most active class 1 analogues, were subsequently eval-
uated for oral (po) activity (Phase VIA) in the rat. Data
are shown in Table 6. In several instances, the threshold
tonic extension (TTE) test25 was performed. This clini-
cally nonselective, electroconvulsive seizure model has
been described previously.25 The test is similar to the
MES test but uses a lower level of electrical current. The
lower current makes the TTE test more sensitive, but
less discriminate than the MES screen. If a compound
was found to possess signi®cant activity in the TTE test
while still remaining inactive in the MES rescreen, it
becomes a candidate for more advanced testing.
Compound R1 R2
R3
X
Yield
MP, ꢀC
Clog P
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
4m
4n
4o
4p
4q
4r
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CO2CH3
CO2C2H5
CO2C(CH3)3
CO2CH3
CO2C2H5
CO2C(CH3)3 Cl
CO2CH3
CO2C2H5
CO2C(CH3)3 CH3
CO2CH3 OCH3 21
CO2C2H5 OCH3 43
CO2C(CH3)3 OCH3 46
CO2CH3
CO2C2H5
H
H
H
Cl
Cl
32
32
44
38
36
14
35
27
11
178±179a
161±163a
193±197b
200±202c
160±162a
166±168a
157±159a
125±127a
136±139a
166±168a
126±128a
142±144a
233±237d
187±189e
212±213f
172±174f
1.55
2.08
2.78
2.46
2.99
3.70
2.05
2.57
3.28
1.75
2.28
2.98
1.45
1.98
2.68
1.16
1.68
2.20
2.08
CH3
CH3
CN
CN
33
49
55
12
CO2C(CH3)3 CN
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
Cl
Cl
CH3
CH3
CH3
OCH3 32
OCH3 20
OCH3 37
CN
CN
CN
CH3
CH3 CH3
25 137.5±139a
13
28
22
18
20
28
16
146±148g
196±199g
170±171.5a 2.60
4s
4t
H
CH3
H
H
Results and Discussion
4u
4v
4w
4x
4y
4z
4aa
4bb
4cc
4dd
CH3 CH3
196±199g
197±200a
174±176a
183±186f
171±174f
166±167a
153±156g
207±210f
219±221a
238±241f
3.12
1.66
2.18
2.70
1.36
1.88
2.40
1.31
1.83
2.35
Consistent with the previous studies, our synthetic
approach utilized the Free-Wilson analysis26 success-
fully employed by Craig.27,28 The para substitution pat-
tern for this study involved: chloro (+p, +s), cyano
( p, +s), methoxy ( p, s), and methyl (+p, s)
groups which were compared to the unsubstituted ana-
logue (0 p, 0 s). As noted in Tables 1 and 2, Clog P29
measurements of the protonated benzylamines,which
predominates at physiological pH, still provided more
lipophilic compounds than the comparable benzamides.
The Clog P range for the benzylamines was 1.84 for 3bb
to 4.74 for 3f while the benzamides varied from 1.16 for
4p to 3.70 for 4f. In either series, the highly lipophilic
compounds were inactive. Class 4 benzylamines in
Table 6 which included 3s (Clog P 3.12) and 3t (3.64)
were found to be soluble with diculty and presented
problems establishing a uniform dosage, thus account-
ing for erratic test results. These current results with the
p-chloro benzylamine derivatives 3s and 3t, rearmed
our initial observation with this substituent.5 Further
data analysis on 3d±f and 3s±u showed only marginal,
class 2, activity (3d, 3u), while each comparable p-chloro
aniline moiety was highly active.5±7 Class 4 benzamides
were more numerous and included 4f, 4i (Clog P 3.70),
4m (1.45), 4q (1.68), 4aa (2.40) and 4cc (1.83). Para-
doxically, 4aa was the most active benzamide in the
series in the rat, displaying an oral MES ED50 of 183.51
mg/kg and a TD50 of >400 mg/kg, providing a protec-
tive index (TD50/ED50) of >2.18. Analysis of the esters
indicated that of the ten tert-butyl esters synthesized,
only one, 3i, possessed anticonvulsant activity. In addi-
tion to the lipophilicity of the tert-butyl group, a steric
(Es)30 factor for this group must also be taken into
consideration. Es values for the esters shows a sig-
ni®cant di€erence: 1.54 for tert-butyl; 0.07 for ethyl;
and 0.00 for methyl. The role of the ester functionality
H
CH3
H
H
CH3 CH3
H
CH3
H
H
CH3 CH3
H
CH3
H
H
6
15
15
CH3 CH3
a
95% Ethanol±ether.
Ethanol (abs.)-ligroine (bp 60±90ꢀC).
2-Propanol.
Methanol±EtOAc-acetone.
2-Propanol±EtOAc.
EtOAc.
b
c
d
e
f
g
Methylene chloride.
3r). The compounds were modeled and minimized by
Spartan programs20 using AM1 model with geometry
optimization. These conformations were transferred to
the SYBYL21 program where they were superimposed
via the MULTIFIT program. Each structure was con-
strained to superimpose the phenyl centroid, the pyrrole
nitrogen (in 1 and 2, or the benzylamine nitrogen in 3a,
3q and 3r), and the carbonyl oxygen. The RMS devia-
tions from the templates of the two pyrroles were 0.77 A
for 1 and 0.78 A for 2, while the mean RMS deviation
of the ®ve molecules was the same in both cases: 0.595
A. The results of the MULTIFIT program are shown in
Figure 4.
Pharmacology
Pharmacological testing of the compounds listed in
Tables 1 and 2 have been provided by the Antiepileptic

J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
2419
Scheme 2. Synthesis of vinylic benzamido compounds. Conditions: (a) H2SO4, ~ (ref. 17); (b) NH3, ~; (c) substituted benzoyl chloride NEt3 (see
Experimental).
amines, was due to the presence of the methylene bridge
which blocks this contribution, thus, comparison to the
active p-substituted aniline derivatives was unsound.5
ED50 analysis for 3p and 3r are currently underway and
will be reported shortly.
In the benzamide series, none of the unsubstituted esters
was active, while dimedone derivative 4r was the most
active analogue in this study. Of interest in this series
was the signi®cant anti-pentylenetetrazol activity shown
Figure 2. X-ray crystal structure of methyl 4-N-[(p-methoxy)-benz-
amido]-6-methyl-2-oxocyclohex-3-en-1-oate (4j).
by a number of analogues. Para-tolyl derivatives 4g and
4h, p-methoxy 4j and 4z, p-chloro 4s, and unsubstituted
4p all possessed signi®cant anti-scMet activity which
was not observed in the benzylamine series. The inter-
pretation of this signi®cant change in protection may be
explained by analysis of the three dimensional behavior
of these series. It was shown in X-ray crystal studies,
high ®eld NMR analysis and molecular modeling
Figure 3. X-ray crystal structure of 3-N-{(p-cyano)benzamido]-5-
methylcyclohex-2-en-1-one (4cc).
examination, that these two series are distinct. Thus,
binding of an active benzylamine to a putative receptor
would be di€erent than that of the active benzamide.
in contributing to anticonvulsant activity was most
dramatically shown in the benzylamine series with pro-
totype 3a (Clog P 2.79). Changing the ester function-
ality to the ethyl ester 3b (Clog P 3.32) or the tert butyl
ester 3c (Clog P 4.03) completely abolished activity.
Further, in this series, only the unsubstituted com-
pounds 3a, 3p 3q and 3r possessed signi®cant anti MES
activity rearming our initial conclusion5 that a steric
factor was most probably operating on the aromatic
group as well. The lack of the -I e€ect in the benzyl-
The fact that only the unsubstituted benzylamines are
MES-active, while only 4r of the unsubstituted benz-
amides protects the animals from electroshock seizures
further veri®es this hypothesis. Further, the lack of
scMet activity in the benzylamines parallels that of the
aniline derivatives,5,6,8 and further di€erentiates them
from the benzamides.
To further justify the pseudo three ring hypothesis, a
search of the Cambridge Structural Database31 for

2420
J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
Table 3. Atomic coordinates [Â104] and equivalent isotropic dis-
placement parameters [A2Â103] for methyl 4-N-[(p-methoxy)benz-
amido]-6-methyl-2-oxocyclohex-3-en-1-oate, 4j. U (eq) is de®ned as
one third of the orthogonalized Uij tensor
Table 4. Atomic coordinates [Â104] and equivalent isotropic dis-
placement parameters [A2Â103] for 3-N-[(p-cyano)benzamido]-5-meth-
ylcyclohex-2-en-1-one, 4c. U (eq) is de®ned as one third of the ortho-
gonalized Uij tensor
x
y
z
U (eq)
x
y
z
U (eq)
O (1)
O (2)
N (1)
N (111)
C (1)
C (2)
C (3)
C (4)
C (41)
C (42)
C (5)
C (6)
C (7)
332 (1)
192 (1)
829 (1)
1149 (3)
204 (2)
181 (2)
780 (2)
1366 (2)
1345 (2)
1961 (2)
1372 (2)
758 (2)
363 (2)
574 (2)
232 (2)
396 (2)
886 (2)
1034 (2)
1225 (2)
1073 (2)
8647 (2)
8829 (2)
9009 (2)
9835 (4)
8817 (3)
8756 (3)
8849 (3)
8516 (3)
7455 (3)
8793 (3)
9006 (3)
8928 (2)
8989 (2)
9152 (3)
8724 (3)
8884 (3)
9486 (3)
9678 (4)
9924 (3)
9753 (3)
5409 (2)
2950 (2)
3311 (2)
567 (2)
82 (1)
71 (1)
54 (1)
118 (2)
57 (1)
56 (1)
59 (1)
53 (1)
76 (1)
73 (1)
60 (1)
49 (1)
52 (1)
54 (1)
63 (1)
70 (1)
66 (1)
84 (1)
68 (1)
59 (1)
O (13)
O (17)
O (111)
O (141)
O (142)
O (23)
O (27)
O (211)
O (241)
O (242)
N (1)
3471 (5)
2543 (5)
235 (4)
3181 (6)
5696 (6)
2290 (4)
1428 (5)
6740 (4)
3213 (5)
992 (6)
2596 (5)
1876 (5)
3082 (5)
3122 (6)
3497 (7)
4053 (15)
4193 (10)
6054 (13)
3588 (34)
4478 (25)
4580 (16)
3545 (9)
2306 (6)
1675 (5)
971 (6)
120 (10)
693 (6)
850 (6)
1339 (6)
780 (6)
166 (6)
1331 (6)
1453 (7)
1847 (8)
3528 (13)
100 (7)
2620 (2)
1142 (2)
90 (2)
1518 (2)
533 (2)
86 (1)
88 (1)
77 (1)
122 (2)
90 (1)
68 (1)
80 (1)
75 (1)
98 (1)
109 (2)
60 (1)
59 (1)
51 (1)
59 (1)
71 (2)
156 (5)
71 (2)
122 (3)
57 (5)
65 (6)
105 (3)
62 (2)
61 (1)
52 (1)
65 (1)
96 (2)
57 (1)
64 (1)
61 (1)
54 (1)
60 (1)
57 (1)
64 (1)
72 (2)
134 (3)
72 (2)
99 (2)
62 (1)
50 (1)
50 (1)
65 (1)
68 (1)
93 (2)
57 (1)
60 (1)
59 (1)
3682 (2)
3824 (3)
3313 (3)
1679 (2)
615 (2)
2771 (2)
3997 (3)
3803 (3)
25 (3)
1186 (4)
1134 (3)
7260 (2)
6597 (2)
4279 (2)
5662 (3)
6237 (4)
562 (3)
4980 (3)
626 (3)
1574 (3)
1754 (4)
773 (6)
1096 (4)
1410 (7)
93 (12)
188 (10)
1133 (6)
336 (4)
284 (3)
103 (3)
878 (4)
1034 (5)
79 (3)
950 (4)
4354 (2)
5104 (2)
5506 (2)
5123 (2)
5026 (2)
5521 (2)
4414 (2)
4026 (2)
2816 (2)
2086 (2)
1560 (2)
870 (2)
N (2)
576 (2)
789 (3)
745 (3)
C (11)
C (12)
C (13)
C (14)
C (141)
C (142)
C (15)
C (15A)
C (151)
C (16)
C (17)
C (18)
C (110)
C (11A)
C (111)
C (112)
C (113)
C (21)
C (22)
C (23)
C (24)
C (241)
C (242)
C (25)
C (251)
C (26)
C (27)
C (28)
C (29)
C (210)
C (21A)
C (211)
C (212)
C (213)
C (8)
C (9)
1537 (4)
2466 (5)
3268 (4)
4058 (5)
2576 (6)
2418 (15)
3368 (5)
1686 (3)
131 (3)
1073 (3)
2066 (3)
4515 (4)
2843 (3)
2723 (3)
1854 (3)
1235 (3)
1138 (3)
1823 (3)
2756 (3)
3553 (3)
4865 (5)
2985 (3)
3850 (4)
2123 (3)
887 (3)
C (10)
C (11)
C (111)
C (12)
C (13)
711 (2)
1 (2)
1234 (2)
1921 (2)
Table 5. Vinyl proton assignments
850 (3)
5212 (3)
6191 (4)
6382 (4)
5469 (3)
5843 (4)
5916 (8)
4550 (4)
3577 (4)
4273 (3)
5231 (3)
5231 (3)
5944 (4)
5602 (4)
3188 (4)
4534 (3)
3823 (4)
4169 (3)
Compound
R
X
d, ppm Compound
R
X
d, ppm
3aa
3bd
3f
CH3
C2H5
H
H
4.98b
5.22b
4a
4b
4f
CH3
C2H5
C(CH3)3 Cl 6.76c
H
H
6.82c
6.82c
C(CH3)3 Cl 5.17b
112 (10)
646 (7)
a
Ref 5.
b
3457 (6)
3457 (6)
3904 (7)
5015 (6)
7398 (9)
5698 (6)
5307 (6)
4209 (6)
Solvent: CDCl3.
Solvent: (CD3)2SO.
Ref 6.
c
887 (3)
d
1741 (3)
2334 (4)
2612 (5)
2113 (3)
1261 (3)
662 (3)
analogue 6, in which the ortho amino group was intra-
molecularly hydrogen bonded to the carboxamide oxygen.
potential intermolecular or intramolecular hydrogen
bonding interactions between vinyl protons as donors
and oxygen atoms (selecting a screening H...O distance
of between 1.8 and 2.5 A) as acceptors gave a total of
11,241 hits. While some of these could be discarded as
not being hydrogen bonding in character based on their
small C-H...O angle, nevertheless the mean H....O dis-
tance was 2.448 A. The distribution of C-H....O angles
was bimodal with a minor cluster centered between 95
and 100ꢀ and a main cluster centered between 155 and
160ꢀ. These results clearly show that a hydrogen-bond-
ing interaction between a vinylic proton and oxygen is a
relatively common occurrence in the solid state. The
pseudo ring hypothesis is not novel. Rososowsky32
postulated a ``pseudo-ring'' system in explaining the
activity of a 2-amino-3-carbamoylpyrazino folic acid
Molecular modeling analysis of the active benzylamines
and the active aroylpyrroles (Fig. 4) further substantiate
the conclusion that these molecules are highly congruent
in their three-dimensional conformation.
Conclusion
The anticonvulsant analysis of the benzylamines and
benzamides reveal di€erent activity pro®les, due most
probably to their three dimensional conformations.
These conformational di€erences may in¯uence the
potential binding interactions in protecting against

J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
2421
solvents using tetramethylsilane as an internal reference.
TLC analysis employed ethyl acetate:cyclohexane (3:1)
elution solvent mixture and 5-Â10 cm and 5-Â20 cm
¯uorescent plates (Whatman silica gel 60A). Elemental
analyses (C, H, N, and halogen) were performed by
Schwarzkopf Microanalytical Laboratory, Woodside,
NY 11377. X-ray crystal analysis was performed on
a Nicolet P3 di€ractometer. 4-Carbo-tert-butoxy-5-meth-
ylcyclohexane-1,3-dione,8,17 4-carbomethoxy-5-methyl-
cyclohexane-1,3-dione,5 4-carbethoxy-5-methylcyclohex-
ane-1,3-dione5 and 5-methyl-1,3-cyclohexanedione17 were
prepared by literature methods. 1,4-Cyclohexanedione
and 5,5-dimethyl-1,3-cyclohexanedione (dimedone)
were obtained from Aldrich Chemical Company and
used without further puri®cation.
Typical experiments illustrating the general procedures
for the preparation of the benzylamines and benzamides
are described below.
General procedure for the preparation of benzylamines
tert-Butyl 4-N-(benzylamino)-6-methyl-2-oxocyclohex-3-
en-1-oate (3c). Into a single-neck 100 mL round bottom
¯ask ®tted with a magnetic stirring bar containing 25
mL of absolute ethanol and 30 mL of ethyl acetate was
added 1.80 g (8 mmol) of 4-carbo-tert-butoxy-5-
methylcyclohexane-1,3-dione, and benzylamine (1.04 g,
9.7 mmol) and the mixture stirred and re¯uxed for 4 h.
Removal of the ca. one-half of the solvent mixture and
stirring overnight at room temperature produced the
crude product which was recrystallized twice from ethyl
acetate to provide 3c in 55% yield, mp 152±155ꢀC as
white crystalline needles: 1H NMR d (CDCl3): 1.08 (3H,
d, J=6.60 Hz, CH3); 1.49 (9H, s, 3ÂCH3); 1.45±3.21
(3H, m, C1, C5 of cyclohexene ring); 3.01 (1H, d,
J=11.00 Hz, C6 of cyclohexene ring); 5.05 (1H, br s,
NH split by CH2); 5.23 (1H, s, ˆCH); 7.25±7.40 (5H, m,
aromatic ring). Anal. calcd for C19H25NO3: C, 72.35; H,
7.99; N, 4.44. Found C, 72.54; H, 8.05; N, 4.68.
Figure 4. Multi®t modeling of the ®ve active anticonvulsants 1, 2, 3a,
3q and 3r with the three atoms/centroid encircled. Colors: chlorine=-
green; oxygen=red; nitrogen=blue.
tert-Butyl 4-N-[(p-methyl)benzylamino]-6-methyl-2-oxo-
cyclohex-3-en-1-oate (3i;). Employing a similar proce-
dure, 4-carbo-tert-butoxy-5-methylcyclohexane-1,3-dione,
and 4-methylbenzylamine produced an oil after stirring
overnight and volume reduction. To the oil was added
ca. 10 mL of anhydrous ether, while stirring and after 30
min, a crude solid was obtained, which after recrystalli-
zation twice from ethyl acetate gave 3i in 58% yield, mp
electroshock and/or pentylenetetrazol induced seizures.
Further research in this area is continuing and will be
reported shortly.
123±126ꢀC as a white crystalline powder: H NMR d
1
((CD3)2SO): 0.89 (3H, d, J=5.88 Hz, CH3); 1.97±3.25
(4H, m, cyclohexene ring); 3.47 (3H, s, CH3); 3.70 (3H,
s, OCH3); 4.11 (2H, s, CH2 of NHCH2); 4.77 (1H, s,
ˆCH); 6.80±7.30 (4H, m, aromatic ring); 7.64 (1H, s,
NH). Anal. calcd for C20H27NO3: C, 72.92; H, 8.26; N,
4.25. Found C, 73.06; H, 8.06; N, 4.45.
Experimental
Chemistry
Melting points were determined on a Thomas-Hoover
capillary melting point apparatus and were uncorrected.
Observed boiling points were also uncorrected. Infrared
spectra were recorded with samples in KBr, as diluted
chloroform solutions in matched sodium chloride cells,
or neat on a Perkin±Elmer 1660 series FTIR spectro-
Ethyl 4-N-[(p-methoxy)benzylamino]-6-methyl-2-oxocyclo-
hex-3-en-1-oate (3k). Employing a similar procedure,
4-carbethoxy-5-methylcyclohexane-1,3-dione, and 4-
methylbenzylamine produced crude solid after stirring
overnight, which after recrystallization twice from ethyl
acetate gave 3k in 74% yield, mp 154±157ꢀC as a white
1
photometer. H NMR spectra were recorded on a Gen-
eral Electric QE 300-MHz spectrometer in deuterated

2422
J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
Table 6. Anticonvulsant evaluation
Benzylamines
Benzamides
Compound
Anticonvulsant resultsa
Compound
Anticonvulsant resultsa
3a
Phase I (mice): Class 1 (MES)
Mice: ED50 64.7 mg/kg; TD50 >500 mg/kg;
Rats: ED50 26.8 mg/kg TD50 >500 mg/kg
Phase I (mice): Class 3
4a
Phase I (mice): Class 3
3b
3c
3d
3e
3f
4b
4c
4d
4e
4f
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 4
Phase I (mice): Class 3
Phase I (mice): Class 2 (MES)
Phase I (mice): Class 3
Phase I (mice): Class 3
3g
3h
3i
3j
3k
3l
3m
3n
3o
3p
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 1 (MES)
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 1 (MES)
Rat po identi®cation: Active at 50 mg/kg (30 min±4 h)
No toxicity at 50 mg/kg at all times
Phase I (mice): Class 1 (MES)
Rats: ED50 55.36 mg/kg TD50 >500 mg/kg
Phase I (mice): Class 1 (MES)
Rat po identi®cation: Active at 30 mg/kg (30 min±4 h)
No toxicity at 50 mg/kg at all times 1/1 died at 4 h
Phase I (mice): Class 4
4g
4h
4i
4j
4k
4l
4m
4n
4o
4p
Phase I (mice): Class 1 (ScMet)
Phase I (mice): Class 1 (ScMet)
Phase I (mice): Class 4
Phase I (mice): Class 1 (MES/ScMet)
Phase I (mice): Class 2 (MES)
Phase I (mice): Class 3
Phase I (mice): Class 4
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 2 (ScMet) TTEb
Active at 100 mg/kg MES
Con®rmation: 0/4 at all times at 100 mg/kg
Phase I (mice): Class 4
3q
3r
4q
4r
Phase I (mice): Class 1 (MES)
Rat po identi®cation: Active at 30 mg/kg (30 min±4 h)
No toxicity at 30 mg/kg at all times
Phase I (mice): Class 1 (MES/ScMet)
Rat po identi®cation: No protection at 30 mg/kg at all
times
3s
4s
3t
3u
3v
3w
3x
3y
3z
Phase I (mice): Class 4
4t
4u
4v
4w
4x
4y
4z
Phase I (mice): Class 3
Phase I (mice): Class 1 (MES)
Phase I (mice): Class 3 TTE: Inactive at 100 mg/kg
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 1 (MES)
Phase I (mice): Class 1 (ScMet)
Phase I (mice): Class 4 (MES)
Phase I (mice): Class 2 (MES)
Phase I (mice): Class 2 (MES)
Phase I (mice): Class 3
Phase I (mice): Class 3
Phase I (mice): Class 2 (MES)
Phase I (mice): Class 3
Phase I (mice): Class 3
3aa
4aa
Rat: ED50 (MES) 183.51 mg/kg TD50 >400 mg/kg
Phase I (mice): Class 1 (MES/ScMet)
Phase I (mice): Class 4 (MES/ScMet)
3bb
3cc
Phase I (mice): Class 2 (MES)
Phase I (mice): Class 1 (MES) 8/8
Toxic at 100 mg/kg 4/4 Toxic at 300 mg/kg
Phase I (mice): Class 3
4bb
4cc
3dd
4dd
Phase I (mice): Class 1 (MES)
a
Phase I in mice activity-class 1=activity at 100 mg/kg or <; class 2=activity between 100 and 300 mg/kg; class 3=no activity at 300 mg/kg; class
4=activity was inconsistent.
TTE=Threshold tonic extension test (see Experimental).
b
crystalline powder: 1H NMR d ((CD3)2SO): 0.89 (3H, d,
J=5.86 Hz, CH3); 1.13 (3H, t, J=7.33 Hz CH3 of
CH2CH3); 1.82±3.56 (4H, m, cyclohexene ring); 3.69
(3H, s, OCH3 ); 4.04 (2H, q, J=7.33 Hz, CH2 of
CH2CH3); 4.11 (2H, d, J=7.26 Hz, CH2); 4.77 (1H, s,
ˆCH); 6.80±7.28 (4H, m, aromatic ring); 7.64 (1H, s,
NH). Anal. calcd for C18H23NO4: C, 68.12; H, 7.30; N,
4.41. Found C, 68.22; H, 7.34; N, 4.68.
CH2); 5.25 (1H, s, ˆCH); 6.93±7.67 (4H, m, aromatic
ring); 9.08 (1H, s, NH). Anal. calcd for C17H18N2O3: C,
68.44; H, 6.08; N, 9.39. Found C, 68.49; H, 6.04; N,
9.44.
3-N-[(p-chloro)benzylamino]-5-methylcyclohex-3-enone
(3t). Employing a similar procedure, 5-methylcyclo-
hexane-1,3-dione, and 4-chlorobenzylamine produced
crude solid after stirring overnight, which after recrys-
tallization twice from 2-propanol gave 3t in 44% yield,
mp 186±187ꢀC as ®ne yellow needles: 1H NMR d
((CD3)2SO): 1.03 (3H, d, J=5.86 Hz, CH3 or C6); 2.88
(1H, d, J=9.52 Hz, CH at C6); 3.30 (2H, s, CH2); 4.31
(2H, t, J=3.67 Hz, CH2 of NHCH2); 4.88 (1H, s,
ˆCH); 6.58 (1H, br s, NH); 7.37 (5H, m, aromatic ring).
Anal. calcd for C18H20ClNO4: C, 67.33; H, 6.46; Cl,
13.44; N, 5.31. Found C, 68.53; H, 6.94; Cl, 13.49; N,
6.55.
Methyl 4-N-[(p-cyano)benzylamino]-6-methyl-2-oxocyclo-
hex-3-en-1-oate (3m). Employing a similar procedure,
4-carbomethoxy-5-methylcyclohexane-1,3-dione, and 4-
aminobenzyl cyanide produced crude solid after stirring
overnight, which after recrystallization twice from ethyl
acetate provided 3m in 44% yield, mp 204±208ꢀC as a
orange crystalline powder: 1H NMR d ((CD3)2SO): 0.95
(3H, d, J=7.33 Hz, CH3 on C6); 1.97±3.24 (4H, m,
cyclohexene ring); 3.61 (3H, s, CO2CH3); 3.99 (2H, s,

J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
2423
General procedure for the preparation of benzamides
residue which was recrystallized from 2-propanol.
Compound 4d was produced in 38% yield and occurred
as light yellow crystals, mp 200±202ꢀC: 1H NMR: d
((CD3)2SO): 1.05 (3H, d, J=7.33 Hz, C6 CH3); 2.13±
3.72 (4, m, cyclohexene ring); 3.47 (3H, s, OCH3); 6.75
(1H, s, ˆCH); 7.76±8.27 (4H, m, aromatic ring); 10.11
(1H, s, NH). Anal. calcd for C16H16ClNO4: C, 59.73; H,
5.01; Cl, 11.02; N, 4.35. Found C, 59.99; H, 4.94; Cl,
10.96; N, 4.38.
tert-Butyl 4-N-(benzamido)-6-methyl-2-oxocyclohex-3-
en-1-oate (4c). Into a 1L three neck ¯ask equipped with
a magnetic stirrer, gas bubbler, Dean-Stark trap and
condenser, was introduced 4-carbo-tert-butoxy-5-
methylcyclohexane-1,3-dione (55.5 g, 0.25 mol) and 500
mL of benzene. After the mixture was heated to re¯ux,
dry ammonia was introduced. Within 5 min, a thick
amorphous precipitate of the crude 6-carbo-tert-butoxy-
3-amino-5-methylcyclohex-2-enone formed. The mix-
ture was re¯uxed an additional 1 h. On cooling, the
slurry was ®ltered, washed with anhydrous ether and air
dried. This crude product could be used as such for the
subsequent reaction.
Ethyl 4-N-[(p-methoxy)benzamido]-6-methyl-2-oxocyclo-
hex-3-en-1-oate (4k). The amination reaction was mod-
i®ed to include 4-carbethoxy-5-methyl-cyclohexane-1,3-
dione and 0.025 mol of p-toluenesulfonic acid as cata-
lyst. 6-Carbethoxy-3-amino-5-methyl-2-cyclohexenone
did not precipitate immediately, but did so on tritura-
tion with anhydrous ether and hexane. The acylation
reaction and workup proceeded as previously indicated
for 4d to yield 4k, 43% as yellow needles, from ethanol-
ether, mp 126±128ꢀC: 1H NMR: d (CDCl3): 1.00 (3H, d,
J=6.59 Hz, CH3); 1.20 (3H, t, J=6.96 Hz, CH3 or
CH2CH3); 2.28±2.85 (3H, m, cyclohexene ring); 3.19
(1H, d, J=11.72 Hz, C1 trans H); 3.84 (3H, s, OCH3);
4.13 (2H, q, J=6.96 Hz, CH2 of CH2CH3); 6.78 (1H, s,
ˆCH); 6.99±7.97 (4H, dd, J=8.79 Hz, aromatic ring);
9.99 (1H, s, NH). Anal. calcd for C18H21NO5: C, 65.24;
H, 6.39; N, 4.23. Found C, 65.29; H, 6.44; N, 4.37.
Into a 500 mL three neck ¯ask ®tted with a condenser,
pressure-equalizing dropping funnel, a magnetic stirrer
and a gas inlet tube, was added 6-carbo-tert-butoxy-3-
amino-5-methyl-cyclohexanone, (11.25 g, 50 mmol) and
triethylamine (10.11 g, 100 mmol) to 300 mL of dry
acetone under a N2 atmosphere. Upon re¯ux, benzoyl
chloride (7.14 g, 51 mmol), dissolved in 50 mL of dry
methylene chloride, was added. The reaction mixture
was re¯uxed for 2.5 h. Once the reaction had cooled, the
mixture was ®ltered to remove triethylamine hydro-
chloride. The solvent was evaporated in vacuo. The
crude product was dissolved in 50 mL of methylene
chloride and washed twice with an equivalent volume of
saturated KHCO3 and water and dried (Na2SO4). The
organic layer was evaporated in vacuo at a temperature
not exceeding 50ꢀC, taken up with a minimum amount
of methylene chloride and chromatographed on a silica
gel column (10 g of silica gel per 1 g of product). The
residue was recrystallized from absolute ethanol:ligroine
(bp 60±90ꢀC), to provide 4c in a 44% yield which
occurred as bright yellow plates, mp 193±197ꢀC: 1H
NMR d ((CD3)2SO): 1.06 (3H, d, J=6.59 Hz, C6 CH3);
1.18 (9H, s, 3ÂCH3); 2.18±3.79 (4H, m, cyclohexene
ring); 6.70 (1H, s, ˆCH); 7.44±8.28 (5H, m, aromatic
ring); 9.99 (1H, s, NH). Anal. calcd for C19H23NO4: C,
69.28; H, 7.04; N, 4.25. Found C, 69.38; H, 7.09; N, 6.95.
3-N-[(p-methyl)benzamido]-5-methylcyclohex-3-enone
(4v). Amination of 1,3-cyclohexanedione as in 4d pro-
vided 3-aminocyclohex-2-enone as a brown, low melting
solid. The acylation reaction and workup proceeded as
previously indicated for 4d to yield 4v, 20%, as yellow
needles, from ethanol-ether, mp 197±200ꢀC: 1H NMR: d
(CDCl3): 1.99±2.16 (2H, m, CH2); 2.26 (2H, t, CH2);
2.48 (3H, s, CH3); 2.61±2.86 (2H, t, CH2); 6.60 (1H, s,
ˆCH); 7.23±7.82 (4H, dd, aromatic ring); 7.90 (1H, S,
NH). Anal. calcd for C14H15NO2: C, 73.34; H, 6.59; N,
6.11. Found C, 73.47; H, 6.49; N, 6.38.
3-N-[(p-cyano)benzamido]-5,5-dimethylcyclohex-3-enone
(4dd). Amination of dimedone under similar conditions
as in 4d provided 3-amino-5,5-dimethyl-cyclohex-2-
enone which did not precipitate immediately, but did so
on trituration with anhydrous ether. The acylation
reaction and work up proceeded as previously indicated
for 4d to yield 4dd, 15% as a yellow crystalline powder,
from ethyl acetate, mp 238±241ꢀC: 1H NMR: d
((CD3)2SO): 1.00 (6H, s, 2ÂCH3); 2.20 (2H, s, CH2);
2.50 (2H, s, CH2); 6.80 (1H, s, ˆCH); 8.00 (4H, m,
aromatic ring); 10.10 (1H, s, NH). Anal. calcd for
C16H16N2O2: C, 71.62; H, 6.01; N, 10.44. Found C,
71.67; H, 6.34; N, 10.78.
Methyl 4-N-[(p-chloro)benzamido]-6-methyl-2-oxocyclo-
hex-3-en-1-oate (4d). The amination reaction was
modi®ed using the above molar quantity of 4-carbo-
methoxy-5-methylcyclohexane-1,3-dione and 0.025 mol
of p-toluenesulfonic acid as catalyst. 6-Carbomethoxy-
3-amino-5-methylcyclohex-2-enone did not precipitate
immediately. The benzene was decanted and the red-
dish-brown residue triturated with anhydrous ether
until cloudy and the mixture refrigerated. After 14 days,
a crystalline product was obtained. Seeding with this
crystalline product readily converted subsequent runs
into solids on overnight refrigeration. The acylation
reaction proceeded as previously indicated to yield a
residue which was washed with fresh methylene chloride
and the oil that remained, after evaporation, was taken
up with 100 mL of methylene chloride, washed succes-
sively with an equivalent volume of saturated KHCO3
(twice), 1 N HCl, and water. The organic layer was
dried (Na2SO4) and evaporated under reduced pressure
at a temperature not exceeding 50ꢀC to yield a solid
X-ray crystal analysis
Methyl 4-N-[(p-methoxy)benzamido]-6-methyl-2-oxocy-
clohex-3-en-1-oate, 4j, and 3-N-[(p-cyano)benzamido]-5-
methylcyclohex-2-en-1-one, 4cc, were recrystallized
from an ethanol:water mixture. All experimental details
related to the structural analysis are provided in Figs. 2
and 3 and Tables 3 and 4 and supplemental material.
The structure was solved by direct methods of the

2424
J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
ShelXTLPC program and re®ned by the ShelXTL
program.33
Twenty mice were pretreated with 100 mg/kg of the test
compound. At several time intervals (15 min, 30 min, 1,
2 and 4 h) post treatment with the test compound, four
mice at each time point were challenged with 12.5 mA
of electrical current for 0.2 s via corneal electrodes. This
stimulation produced a TTE seizure in the animals. For
each time interval, results were expressed as a ratio of
the number of animals protected over the number of
animals tested.
Molecular modeling
Initial computations on the benzamides and benzyla-
mines were performed on a Silicon Graphics Personal
Iris 4D/35 workstation running Molecular Simulations
Quanta and CHARMM molecular mechanics software.34
Each structure in Tables 1 and 2 was initially minimized
employing steepest descents (100 minimization steps)
and subsequently by adopted-basis Newton Raphson
(ABNR, 1000 steps). These individual minima were
saved for a conformational search, using torsion angles
and restrictions previously reported.6 The active analo-
gues in each series were superimposed and their mole-
cular volumes calculated and compared to the inactive
analogues. There was no di€erence between the active
Supplemental material
Additional X-ray crystal data are provided for 4j (4
tables) and 4cc (4 tables). Unit cells of 4j and 4cc and
are available from the authors.
Acknowledgements
and inactive analogues (union volume(inactive benzamides)
=
We thank the Minority Biomedical Research Support
Program (GM08244-06), the Mordecai Wyatt Johnson
Research Grant and the UÃStar Grant Program for
support of this investigation; and the Graduate School
of Arts and Sciences for support of the high resolution
NMR spectrometer. We extend sincere thanks to Dr.
Harold R. Almond, R.W. Johnson Pharmaceutical
Research Institute for sharing his molecular modeling
laboratory for the MULTIFIT analysis. We also wish
to thank Dr. Gary O. Rankin, Marshall University,
School of Medicine, Department of Pharmacology,
Huntington, WV, Dr. C. Randall Clark, Auburn Uni-
versity, Department of Pharmaceutical Sciences,
Auburn, AL, and Dr. John R. Carson, R.W. Johnson
Pharmaceutical Research Institute for helpful discus-
sions and to Dr. Penelope W. Codding, University of
Victoria for a helpful critique.
371.87 A v. union volume(active benzamides)=385.14 A).
Structures 1, 2, 3a, 3q and 3r were modeled with the
Spartan programs20 using the AM1 model with geo-
metry optimization. These individual minima were
transferred to the SYBYL21 program where they were
superimposed via the MULTIFIT program.21 The
structures were constrained to superimpose the phenyl
centroid, the pyrrole nitrogen (in 1 and 2, or the benzy-
lamine nitrogen in 3a, 3q and 3r), and the carbonyl
oxygen. The analysis did not employ a reference struc-
ture (i.e. phenytoin) to which the other molecules were
®t. Quite the opposite, all molecules were allowed to ®t
each other simultaneously, resulting in a ``consensus
conformation'' (multi®t cluster).35 The results of the
MULTIFIT program are shown in Fig. 4.
Pharmacology
Initial evaluations for anticonvulsant activity were per-
formed by the Antiepileptic Drug Development (ADD)
Program, Epilepsy Branch, Neurological Disorders
Program, National Institute of Neurological Disorders
and Stroke and included phases I, II, VIA and VIB test
procedures which have been described.22±24 These tests
were performed in male Carworth Farms no. 1 (CF1)
mice (Phases I and II) or male Sprague±Dawley rats
(Phase VIA and VIB). Phase I and Phase VIA of the
evaluation included three tests: maximal electroshock
(MES), subcutaneous pentylenetetrazol (ScMet), and
the rotorod test for neurological toxicity (Tox). Com-
pounds were suspended in 0.5% aqueous methylcellu-
lose and were administered at three dosage levels (30,
100 and 300 mg/kg) with anticonvulsant activity and
motor impairment noted 30 min and 4 h after adminis-
tration. Phase II and phase VIB testing quantitated the
anticonvulsant activity and motor impairment observed
for the most promising compounds in phase I. Phase II
quanti®ed data in CF1 mice by intraperitoneal (ip)
administration, while phase VIB provided oral rat data
comparable to phase II ip data in mice. Data for the
respective evaluations are shown in Table 5. The TTE
test25 performed on 4p and 4v is described as follows.
References
1. Coatsworth, J. J. NINDS monograph No. 12, HEW pub-
lication no. (NIH) 73-51. U.S. Government Printing Oce,
1971.
2. Trimble, M. R. In New Anticonvulsants, Advances in the
Treatment of Epilepsy; J. Wiley and Sons: New York, 1994.
3. Carson, J. R.; Carmosin, R. J.; Pitis, P. M.; Vaught, J. L.;
Almond, H. R.; Stables, J. P.; Wolf, H. H.; Swinyard, E. A.;
White, H. S. J. Med. Chem. 1997, 40, 1578.
4. Unverferth, K.; Engel, J.; Hofgen, N.; Rostock, A.; Gun-
ther, R.; Lankau, H.-J.; Manzer, M.; Rolfs, A.; Liebscher, J.;
Muller, B.; Hofmann, H.-J. J. Med. Chem. 1998, 41, 63.
5. Eda®ogho, I. O.; Hinko, C. N.; Chang, C.; Moore, J. A.;
Mulzac, D.; Nicholson, J. M.; Scott, K. R. J. Med. Chem.
1992, 35, 2798.
6. Scott, K. R.; Eda®ogho, I. O.; Richardson, E. L.; Farrar,
V. A.; Moore, J. A.; Tietz, E. I.; Hinko, C. N.; Chang, H.;
El-Assadi, A.; Nicholson, J. M. J. Med. Chem. 1993, 36, 1947.
7. Roberts, R.R.; Eda®ogho, I.O.; Scott, K.R. US Patent 5
580 894 1996.
8. Scott, K. R.; Rankin, G. O.; Stables, J. P.; Alexander, M.
S.; Eda®ogho, I. O.; Farrar, V. A.; Kolen, K. R.; Moore, J.
A.; Sims, L. D.; Tonnu, A. D. J. Med. Chem. 1995, 38, 4033.
9. Eda®ogho, I. O.; Moore, J. A.; Alexander, M. S.; Scott, K.
R. J. Pharm. Sci. 1994, 83, 1155.

J. E. Foster et al. / Bioorg. Med. Chem. 7 (1999) 2415±2425
2425
10. Laws, M. L.; Roberts, R. R.; Nicholson, J. M.; Butcher,
R.; Stables, J. P.; Goodwin, A. M.; Smith, C. A.; Scott, K. R.
Bioorganic and Med. Chem. 1998, 6, 2289.
11. Jones, R. A.; Bean, G. P. The Chemistry of Pyrroles; Aca-
demic Press: New York; pp. 284±289.
Development Program, National Institutes of Health, DHEW
Publ (NIH) (U.S.) 1978, NIH 78-1093.
23. Porter, R. J.; Cereghino, J. J.; Gladding, G. D.; Hessie,
B. J.; Kupferberg, H. J.; Scoville, B.; White, B. G. Cleveland
Clin. Q. 1984, 51, 293.
12. Cook, A. G. Enamines: Synthesis and Reactions; 2nd ed.;
Marcel Dekker: New York, 1988; pp.16±17.
24. Krall, R. L.; Penry, J. K.; White, B. G.; Kupferberg, H. J.;
Swinyard, E. A. Epilepsia 1978, 19, 400.
13. Eda®ogho, I. O.; Scott, K. R. Anticonvulsants, In Burger's
Medicinal Chemistry and Drug Discovery; 5th ed.; Wol€, M. E.
Ed.; J. Wiley and Sons: New York, 1996; pp. 175±260.
14. Musso, D. L.; Boswell, G. E.; Mehta, N. B.; Soroko, F. E.;
Burchall, C. J. Eur. J. Med. Chem. 1996, 31, 331.
15. Moreau, S.; Coudert, P.; Rubat, C.; Vallee-Goyet, D.;
Gardette, D.; Gramain, J.-C.; Couquelet, J. Bioorganic and
Med. Chem. 1998, 6, 983.
25. Piredda, S. G.; Woodhead, J. H.; Swinyard, E. A. J.
Pharmacol. Exp. Ther. 1985, 232, 741.
26. Free, S. M., Jr.; Wilson, J. W. J. Med. Chem. 1964, 7, 395.
27. Craig, P. N. J. Med. Chem. 1971, 14, 680.
28. Craig, P. N. J. Med. Chem. 1972, 15, 144.
29. MacLogP Program; version 2.0, BioByte Corp. Claremont,
CA 91711.
30. Taft, R. W. In Steric E€ects in Organic Chemistry; Newman
M. S. Ed. 1956; J. Wiley and Sons: New York; pp. 556±675.
31. Allen, F. H.; Kennard, O. Chem. Design Automation News
1993, 8, 31.
16. Michon, V.; Herve du Penhoat, C.; Tombret, F.; Gillardin,
J. M.; Lepage, F.; Berthon, L. Eur. J. Med. Chem. 1995, 30,
147.
17. Friary, R. J.; Gilligan, J. M.; Szajewski, R. P.; Falci, K. F.;
Franck, R. W. J. Org. Chem. 1973, 38, 3487.
32. Rosowsky, A.; Forsch, R. A.; Moran, R. G. J. Med.
Chem. 1991, 34, 463.
18. Zymalkowski, F.; Rimek, H. Archiv der Pharmazie 1961,
294, 759.
19. Kubicki, M.; Codding, P. W. Acta Cryst. 1993, C49, 2045.
20. Spartan SGI Version 4.1.1, Wavefunction, 18401 Von
Karman Ave., #370, Irvine, CA 92715. # 1991±1996 Wave-
function.
21. SYBYL, Version 6.04, Tripos Associates, 1699 S. Hanley
Road, St. Louis, MO 63144.
22. Anticonvulsant Screening Project, Antiepileptic Drug
33. Sheldrick, G.M. SelXLTLPC and ShelXTL. Program for
crystal structure determination, 1996, Cambridge University,
England.
34. Quanta Release 3.2.3, Molecular Simulations Corporation,
Waltham, MA 1992.
35. Bigge, C. F.; Malone, T. C.; Boxer, P. A.; Nelson, C. B.;
Ortwine, D. F.; Schelkun, R. M.; Retz, D. M.; Lescosky, L. J.;
Borosky, S. A.; Vartanian, M. G.; Schwarz, R. D.; Campbell,
G. W.; Robichaud, L. J.; Watjen, F. J. Med. Chem. 1995, 38, 3720.