Anticonvulsant properties of various acetylhydrazones, oxamoylhydrazones and semicarbazones derived from aromatic and unsaturated carbonyl compounds

Article abstract of DOI:10.1016/S0223-5234(00)00123-9

Various acetylhydrazones, oxamoylhydrazones and semicarbazones were prepared as candidate anticonvulsants with a view to examining the viability of a putative binding site hypothesis. Atomic charge calculations were undertaken to determine the hydrogen bo

Full text of DOI:10.1016/S0223-5234(00)00123-9

Eur. J. Med. Chem. 35 (2000) 241−248
© 2000 Éditions scientifiques et médicales Elsevier SAS. All rights reserved
S0223523400001239/FLA
241
Original article
Anticonvulsant properties of various acetylhydrazones, oxamoylhydrazones and
semicarbazones derived from aromatic and unsaturated carbonyl compounds
Jonathan R. Dimmocka*, Sarvesh C. Vashishthaa, James P. Stablesb
aCollege of Pharmacy and Nutrition, University of Saskatchewan, Saskatoon, Saskatchewan S7N 5C9, Canada
bNational Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda MD 20892-9020, USA
Received 26 April 1999; revised 28 July 1999; accepted 5 August 1999
Abstract – Various acetylhydrazones, oxamoylhydrazones and semicarbazones were prepared as candidate anticonvulsants with a view to
examining the viability of a putative binding site hypothesis. Atomic charge calculations were undertaken to determine the hydrogen bonding
capacities of various molecules. The biological results obtained revealed that in general the acetylhydrazones and semicarbazones afforded
good protection against convulsions while the oxamoylhydrazones were significantly less active. These data suggest that terminal
electron-donating groups enhanced the hydrogen bonding capabilities and anticonvulsant properties of these molecules. © 2000 Éditions
scientifiques et médicales Elsevier SAS
semicarbazones / acetylhydrazones / oxamoylhydrazones / anticonvulsant / atomic charges
1. Introduction
Initially, the nature of the hydrogen bonding area was
investigated. In the absence of an identified binding site
the assumption was made that it was a peptide chain
whereby the repeating amidic functions allow hydrogen
bond formation (figure 2). If this supposition is correct,
then changing the nature of the group X should affect the
capacity of the compounds to form hydrogen bonds at the
binding site, which will be reflected by alterations in
anticonvulsant activity. The maximum electronic effect
caused by varying X would be predicted to be on atoms
6 and 7, while smaller changes in the electron densities
would be predicted to be on atoms 1–5 due to tautomer-
ism of the 5 hydrogen atom and the subsequent rearrange-
ment of single and double bonds [7]. Thus two series of
compounds were proposed whereby substituent X was
changed from an amino group in 1a–f to a methyl
function in the acetylhydrazones 2a–f and an aminocar-
bonyl moiety leading to the oxamoylhydrazones 3a–f.
The terminal amino, methyl and aminocarbonyl groups
found in series 1, 2 and 3, respectively, would be
expected to exercise different electronic effects on atoms
1–7, which in turn should affect hydrogen bonding
capabilities with the putative binding site and hence
anticonvulsant activity.
Today there is a need for new antiepileptic drugs since
approximately one quarter of epileptic patients find that
drug therapy inadequately controls their convulsions [1].
In addition, many currently used antiepileptic drugs cause
significant side effects which may limit their maximal
usefulness [2]. In the course of investigations aimed at
developing structurally novel anticonvulsants, a number
of aryl semicarbazones, including series 1, were found to
display significant activity [3, 4]. These compounds were
believed to interact at two locations on a putative binding
site designated a hydrogen bonding area and an aryl
binding site [5]. However, since the aryl group can be
replaced by other hydrophobic moieties with retention of
anticonvulsant activity [6], the portion of the binding site
with which the aryl group of series 1 and related
compounds interacts will be referred to as a hydrophobic
bonding area rather than an aryl binding site. The
principal objective of the present investigation was the
preparation of a number of analogues of 1 with a view to
further evaluating the binding site hypothesis (figure 1).
*Correspondence and reprints: dimmock@skyway.usask.ca

242
Figure 1. Structures of 1–5.

243
additional hydrophobic binding site, the synthesis and
bioevaluation of 2g–l were suggested whose bioactivity
could be compared to 2a–f. In addition 3g was also
considered for comparison with 3a.
In order to verify that changes in the nature of
substituent X would alter the polarity of the group
interacting at the hydrogen bonding area, the decision
was made to determine the atomic charges of the atoms
1–7 noted in figure 2.
The second phase of the study was directed to exam-
ining some of the structural requirements at the hydro-
phobic bonding area. In this regard, the preparation of
series 4 was suggested. Replacement of the phenyl ring of
1a by a hydrophobic n-octyl group leading to 4a would
retain hydrophobic bonding while creating spatial prop-
erty differences between the phenyl and n-octyl groups.
Since the substitution of the phenyl ring of 1a by a
â-phenylvinyl group producing 4b led to increased anti-
convulsant activity [3], a comparison of the activity of 4b
with 4c was considered in order to provide further
information of the properties of the groups that could be
accommodated at the hydrophobic bonding area. The
remaining analogues 4d–f were designed in order to
discern the influence on anticonvulsant activity of the
length and increasing unsaturation of the n-alkyl chains.
Compounds 4b and 4e had promising anticonvulsant
activities vide infra. In order to explore further the
generality of the effect on anticonvulsant properties of
replacing the amino group of the semicarbazones by
methyl and aminocarbonyl functions, the preparation and
bioevaluation of 5a–d was suggested.
Figure 2. Proposed interactions of compounds 1–5 at a binding
site.
A previous study revealed that replacement of the
methine proton of 1a by a methyl group led to 1g with an
approximately 2-fold increase in anticonvulsant activity
when the compounds were administered by the intraperi-
toneal route to mice [3]. In order to explore whether this
observation was indicative of a general trend, which may
therefore afford some insight into the existence of an
In summary therefore, the principal aim of the present
investigation was the synthesis of the compounds in series
2–5 for anticonvulsant evaluation with a view to obtain-
Table I. Atomic charges on different atoms of various compounds in series 1–5.
Compound
X
Atomsa
4
1
2
3
5
6
7
1a
2a
3a
1g
2g
3g
4b
5a
5c
4e
5b
5d
NH2
CH3
H2NCO
NH2
CH3
H2NCO
H2N
CH3
H2NCO
H2N
CH3
H2NCO
–0.028
–0.028
–0.026
–
0.051
0.052
0.059
0.077
0.077
0.082
0.057
0.057
0.070
0.058
0.062
0.072
–0.045
–0.050
–0.056
–0.066
–0.071
–0.077
–0.047
–0.053
–0.083
–0.050
–0.061
–0.085
–0.146
–0.132
–0.119
–0.149
–0.134
–0.124
–0.146
–0.132
–0.128
–0.148
–0.134
–0.130
0.095
0.097
0.105
0.094
0.095
0.113
0.094
0.096
0.105
0.093
0.093
0.104
0.448
0.351
0.291
0.448
0.351
0.291
0.448
0.351
0.287
0.448
0.351
0.287
–0.378
–0.317
–0.303
–0.379
–0.319
–0.297
–0.378
–0.317
–0.275
–0.379
–0.320
–0.277
–
–
–0.019
–0.020
–0.014
–0.019
–0.021
–0.013
aThe atoms corresponding to the numbers are shown in figure 2.

244
Table II. Evaluation of compounds 1–5 in the mouse intraperito-
Table III. Comparison of the protection scores with various
groups of compounds in the mouse intraperitoneal MES screen.
neal MES, scPTZ and NT screensa.
Compound
MES screen scPTZ screen
NT screen
Compounds
Protection scorea
4 h
0.5 h
4 h
0.5 h
4 h
0.5 h
4 h
0.5 h
total
1a
1b
1c
1d
1e
1f
1g
2a
2b
2c
2d
2e
2f
2g
2h
2i
2j
2k
2l
3a
3b
3c
3d
3e
3f
3g
4a
4b
4c
4d
4e
4f
100
100
–
100
–
100
100
100
30
100
300
300
100
300
100
–
–
–
300
300
–
–
–
–
300
300
300
100
100
100
100
100
100
100
–
–
100
300
–
300
300
–
300
–
300
100
300
100
300
300
300
100
–
–
–
–
–
300
–
–
–
–
–
–
–
–
100
100
100
–
300
–
300
–
–
–
–
–
–
–
–
–
–
300
–
300
–
300
–
–
–
–
–
–
–
–
–
–
–
–
–
–
300
300
–
–
–
–
–
–
–
300
–
–
–
–
–
300
–
–
–
1a–f
2a–f
3a–f
1a, 2a, 3a
1g, 2g, 3g
2a–f
12
21
2
7
5
5
8
3
2
1
8
0
17
29
5
9
6
–
300
300
300
100
100
–
300
–
–
–
–
–
–
–
300
–
300
–
300
–
–
–
–
21
5
29
5
100
300
100
–
300
300
300
300
300
–
2g–l
300
300
300
300
300
300
300
300
100
300
300
100
–
–
–
–
–
–
–
–
–
aThe figures in the table are the total scores for each cluster of
compounds (see Results and discussion section regarding the
calculation of the PS values).
oxamoylhydrazones 3a–g were synthesized from the
appropriate aryl aldehydes or ketones and oxamic acid
hydrazide. The semicarbazones 4a–f were prepared from
semicarbazide and various saturated and unsaturated
aliphatic aldehydes. Condensation of cinnamaldehyde
and 1-nonenal with both acetic and oxamic acid hy-
drazides led to 5a and 5b, and 5c and 5d, respectively.
The charge densities of a number of atoms in selected
compounds were calculated (table I).
–
300
300
–
–
–
–
300
–
300
300
300
300
100
–
300
300
300
300
–
–
–
3. Pharmacology
300
300
–
300
–
300
300
–
300
30
100
–
All of the compounds were injected intraperitoneally
into mice and evaluated in the maximal electroshock
(MES), subcutaneous pentylenetetrazole (scPTZ) and
neurotoxicity (NT) screens using doses of 30, 100 and
300 mg/kg at two different time intervals. These data are
presented in table II. Comparisons of the activities of
various groups of compounds are summarized in ta-
ble III. Approximately half of the compounds were also
evaluated orally in rats for activity in the MES test at
several time points (table IV).
300
300
–
–
5a
5b
5c
5d
300
300
–
–
30
30
–
–
Phenytoin
Carbamazepine
Valproic acid
–
100
300
–
300
–
100
100
–
100
300
–
aDoses of 30, 100 and 300 mg/kg of the compound were adminis-
tered and the protection and neurotoxicity measured after 0.5 and
4 h. The figures indicate the minimal dose required to cause
protection or neurotoxicity in 50% or more of the animals.
– indicates the absence of anticonvulsant activity or neurotoxicity.
4. Results and discussion
In order to study the contribution of the electronic
effects of the amino, methyl and aminomethyl groups at
the putative binding site, the atomic charges on the seven
atoms indicated in figure 2 were computed for four
groups of compounds, each group consisting of three
analogues. These data are presented in table I. The
bonding that can occur between the ligands and a peptide
chain is considered to be hydrogen bonding. In the case of
the N3 HCR bond, the proton of the peptide chain is
considered to be acidic due to the electron-withdrawing
ing a greater understanding of the nature of the binding
site with which these compounds are believed to interact.
2. Chemistry
Reaction of acetic acid hydrazide with various alde-
hydes or ketones led to the formation of 2a–l, while the

245
Table IV. Evaluation of various compounds for activity in the rat
magnitude of the positive charges on the 5 hydrogen
atoms reflects the electronic properties of group X, e.g.,
the greatest charges are on the 5 hydrogen atoms found in
the oxamoylhydrazones in which X is the most electron-
attracting group. The two remaining atoms where hydro-
gen bonding could occur, namely at positions 1 and 3, had
small atomic charges. It is likely therefore that they
would make only minor contributions to the binding of
these molecules to cellular macromolecules. In summary
therefore, the data provided by the charge density calcu-
lations support the hypothesis outlined previously that
variation in the nature of group X will markedly affect the
capacity for hydrogen bonding at a binding site.
Candidate anticonvulsants are often evaluated initially
in the MES and scPTZ screens [8]. Compounds affording
protection in the MES test may prove to be useful in
treating generalized tonic–clonic and complex partial
seizures, while activity in the scPTZ screen is claimed to
denote agents of value in treating absence seizures.
Neurotoxicity in mice may be measured by the rotorod
procedure [9], while minimal motor impairment in rats is
detected by overt evidence of ataxia and abnormal gait
and stance. These procedures were undertaken in the
present investigation.
The results of intraperitoneal screening in mice are
summarized in table II. Two general trends may be
discerned. First, the data for the MES and scPTZ tests
revealed that 74 and 92%, respectively, of the compounds
2–5 had greater activity at the end of 0.5 h than after 4 h.
Thus, in general, these compounds are short acting
anticonvulsants. Second, protection was afforded by 79
and 41% of the compounds 2–5 in the MES and scPTZ
screens, respectively. In addition, 55% of the compounds
had greater activity in the MES test rather than the scPTZ
screen, while for the remaining cases equal activity was
demonstrated. Furthermore, as will be presented later, in
the rat oral MES screen 1b, 1g, 2f and 4b afforded
protection, but these compounds were inactive in the
scPTZ test using the same doses. Thus the majority of the
compounds displayed preferential or exclusive MES-
selectivity and hence further discussion in regard to the
binding site hypothesis and structure–activity relation-
ships will be devoted to the data provided from MES
screens only.
In order to compare the activities of various clusters of
candidate anticonvulsants in the mouse intraperitoneal
MES screen, compounds which were active at 30, 100
and 300 mg/kg were assigned scores of 10, 3 and 1,
respectively. In this way, the protection score (PS) values
for different groups of compounds were obtained and are
presented in table III. A comparison of the figures for
1a–f, 2a–f and 3a–f revealed that the order of anticon-
oral MES screena.
Compound
Time of test (h)
0.25
0.5
1
2
4
1a
1b
1c
1d
1e
1f
1g
2a
2c
2d
2f
2g
3g
4b
4c
4e
4
–
–
3
–
2
4
–
–
1
1
1
2
4
–
4
1
1
2
–
1
4
2
–
–
–
4
4
1
–
1
1
–
1
4
3
4
3
1
1
–
4
4
4
4
–
–
4
4
–
2
2
1
–
–
4
3
3
2
–
2
–
3
4
4
3
–
–
4
1
–
2
1
2
–
–
4
4
2
–
1
1
2
3
–
1
4
–
–
3
–
–
1
–
–
–
–
1
3
1
1
–
–
–
3
4f
5a
5b
5c
Phenytoin
aThe doses administered were 50 mg/kg (1a–g and 4b), 25 mg/kg
(2f) and 30 mg/kg in the remaining cases, including phenytoin.
Four rats were used for each compound and the figures in the table
indicate the number of rats protected. – indicates an absence of
anticonvulsant activity.
influence of the adjacent carbonyl group; however, it is
possible that it is a dispersion force rather than a
hydrogen bond. Nevertheless, from the standpoint of the
arguments developed in this study, the term hydrogen
bonding will be used in all four ligand–peptide interac-
tions as illustrated in figure 2. The magnitude of the
charges on the atoms which could be involved in hydro-
gen bonding is in the order of 7 > 5 > 3 > 1. In particular
the large negative charge on the 7 oxygen atom would be
predicted to exert an enormous effect on hydrogen
bonding at a binding site. In each of the 4 groups of
compounds, the negative electronic charge of the 7
oxygen atom was in the order of amino > methyl >
aminocarbonyl. The average percentage reductions in the
atomic charges on the 7 oxygen atom of the acetyl and
oxamoyl hydrazones compared to the corresponding
semicarbazones were 16 and 24, respectively.
The second portion of these molecules, where hydrogen
bonding could occur, is at position 5. In this case,
hydrogen bonding will be facilitated as the positive charge
on the 5 hydrogen atom increases. With the exception of
4e and 5b which have the same electron densities, the

246
vulsant activity was the acetylhydrazones > semicarba-
zones > oxamoylhydrazones. While the greater activity of
the acetylhydrazones than the semicarbazones was unan-
ticipated, the results indicate that the oxamoylhydra-
zones, having the smallest negative atomic charges on the
7 oxygen atom, possessed the lowest anticonvulsant
activity. It is conceivable that the terminal methyl group
in 2a–f forms van der Waals bonding with a correspond-
ing area of the binding site; such interactions would not
occur when X is either an amino or aminocarbonyl group.
A comparison of the PS values for 1a, 2a and 3a with
1g, 2g and 3g revealed that greater anticonvulsant activity
was displayed by the nor analogues 1a, 2a and 3a. In
order to evaluate the generality of this observation, the PS
figures for 2a–f and 2g–l were contrasted. The data in
table III reveal that the acetylhydrazones having a me-
thine proton were significantly more active than the homo
analogues. Thus a site of steric repulsion may be present
on the binding site hindering alignment of the molecules
and hence lowering anticonvulsant activity.
Replacement of the phenyl ring of 1a (PS = 3) by an
n-octyl group led to 4a (PS = 1) with reduced activity.
However the preparation of analogues in which an
olefinic double bond was attached to the carbimino
(C=N) group led to increases (4b, 4c and 4e; PS = 4) or
retention (4d and 4f; PS = 3) of activity. Other structural
features incorporated into the R group of series 4 namely
an aryl ring (4b), alkyl chains of varying lengths (4c–e)
and an additional olefinic double bond (4f) seemed to be
relatively unimportant contributors to bioactivity.
The biodata generated in the mouse intraperitoneal
MES screen for series 5 confirmed two generalizations
noted earlier. First, as observed in the case of 4b and 4e,
the hydrophobic bonding area can accommodate both the
â-phenylvinyl and 1-octenyl groups present in 5a and 5b;
both compounds have a PS value of 4. Second, while the
acetylhydrazones 5a and 5b have the same PS value of 4
as the corresponding semicarbazones 4b and 4e, the
oxamoylhydrazones 5c (PS = 0) and 5d (PS = 1) are
much less potent. These data support the conclusion
drawn earlier that in order for significant anticonvulsant
activity to be displayed, group X in figure 2 may be an
amino or methyl but not an aminocarbonyl function.
In order to confirm the utility of these alkylidene
semicarbazones, quantification of a representative com-
pound 4c was undertaken. The ED50 figure in the MES
screen was 21.3 mg/kg and a protection index (PI: i.e.,
TD50 in the rotorod screen/ED50 in the anticonvulsant
test) was 12.4. These figures compare favourably with 1a
for which the ED50 and PI values were 69.7 and 2.93,
respectively [3]. In addition, while 4c has approximately
one-third of the potency of phenytoin (ED50 in the MES
screen = 6.32; PI = 6.52), the PI value of 4c is virtually
double that of the established drug. Thus 4c and related
compounds serve as useful molecules for further devel-
opment.
In the case of those series in which there was a variation
in the aryl substituent patterns, the groups in 1a–e, 2a–e
and 2g–k, and 3a–e were chosen for a Topliss analysis
[10]. The 4-bromo substituent in 2f and 2l, and 3f was
used owing to the promising anticonvulsant activity of
1f [11]. In the light of the results obtained, a Topliss
analysis did not reveal any parameter dependencies nor
did the 4-bromo derivatives 2f and 2l, and 3f demonstrate
markedly superior activity than their analogues.
A number of compounds were examined for activity in
the rat oral MES screen and the results are summarized in
table IV. The initial screening using 1a–g and 4b em-
ployed doses of 50 mg/kg. However, since protection in
all animals was often obtained, the doses of the analogues
were generally reduced to 30 mg/kg in order to detect the
most active compounds. The ED50 figures in the rat oral
MES screen for 1a, 1c, 1f and 1g were 20–23 mg/kg [3,
4] and the related acetylhydrazones possessed compa-
rable (2c and 2f) or reduced (2a and 2g) activity when
30 mg/kg of the compound was administered. In general
therefore, the semicarbazones afforded greater protection
than the corresponding acetylhydrazones in the rat oral
MES screen. This trend was also observed by noting that
the semicarbazones 4b and 4e were more active than the
analogous acetylhydrazones 5a and 5b. These observa-
tions support the hypothesis outlined previously that the
greater electron-donating capabilities of the amino group
compared to the acetyl function leading to greater hydro-
gen bond formation favours bioactivity. The promising
results obtained in the mouse intraperitoneal MES screen
for the semicarbazones which contained olefinic linkages
were also noted when administered orally to rats. Thus
protection in the MES screen was 100% (4b, 4c and 4e)
or 75% (4f) affording further evidence towards the
profitability of developing this series of compounds. At
the doses utilized in the MES screen, none of the
compounds displayed neurotoxicity and 1b, 1g, 2f and 4b
gave no protection in the scPTZ test.
5. Conclusions
The synthesis of a number of acetylhydrazones, oxam-
oylhydrazones and semicarbazones of aryl aldehydes and
related compounds as candidate anticonvulsants is re-
ported in this study. Measurements of the atomic charges
on certain atoms of representative compounds revealed
changes in electronic properties caused by varying the
terminal group from amino to methyl and aminocarbonyl

247
1
functions. Evaluation in the mouse intraperitoneal MES
and scPTZ screens showed that the compounds were
rapidly acting and afforded greater protection in the MES
screen. The biodata generated suggested that a site of
steric repulsion close to atom 1 (figure 2) is possible.
Confirmation that the hydrophobic bonding area could
accommodate groups other than aryl rings was noted and
in particular 1-nonenal semicarbazone 4c is clearly a
useful lead compound. Most of the semicarbazones in
series 1 and 4 afforded complete protection in the rat oral
MES screen and were bereft of NT at the doses em-
ployed. In general the results obtained revealed that the
acetylhydrazones and semicarbazones afforded greater
protection in the MES screens than the corresponding
oxamoylhydrazones which may have been due to the
higher negative charges on the 7 oxygen atom with the
more active compounds.
5b: 69–71°, 74%. The H-NMR spectrum of a represen-
tative compound 2d was as follows: δ (CDCl3): 2.36 (s,
3H, CH3), 2.37 (s, 3H, CH3), 7.18 (d, 2H, 3,5 aryl H),
7.55 (d, 2H, 2,6 aryl H), 7.80 (s, 1H, CH=N), 10.11 (s,
1H, NH) ppm.
6.1.2. Synthesis of compounds 3a–g, and 5c and 5d
A mixture of the appropriate aldehyde and ketone
(0.01 mol) and oxamic acid hydrazide (0.01 mol) in
ethanol (70% v/v, 100 mL) was heated under reflux for
3 h. On cooling, the precipitate was collected, washed
thoroughly with ethanol (70% v/v) and dried to yield the
pure oxamoylhydrazones. The m.p.’s and yields are: 3a:
286–288° (lit. [17] 215–216°), 92%; 3b: 312–314°, 93%;
3c: 286–290°, 91%; 3d: 294–296°; 85%; 3e: 290–292°,
78%; 3f: > 300° (lit. [17] > 300°), 96%; 3g: 211–213°,
77%; 5c: 304–306°, 92% and 5d: 245° (dec), 90%. The
1H-NMR spectrum of a representative compound 3d was
as follows: δ (DMSO-d6): 2.33 (s, 3H, CH3), 7.26 (d, 2H,
3,5 aryl H), 7.57 (d, 2H, 2,6 aryl H), 7.93 (s, 1H, CH=N),
8.27 (s, 1H, CONH2), 8.51 (s, 1H, CONH2), 12.02 (s, 1H,
CONHN) ppm.
6. Experimental protocols
6.1. Chemistry
Melting points are uncorrected and are quoted in
degrees Centigrade. Elemental analyses (C, H, N) were
undertaken on 2a–e and 2g–l, 3a–g, 4a and 4c–f, and
5a–d by Mr K. Thoms, Department of Chemistry, Uni-
versity of Saskatchewan and were within 0.4% of the
calculated values. All compounds were shown to be
homogenous by TLC using silica gel 60 F254 precoated
plastic sheets and a solvent system of chloroform:metha-
nol (7:3). 1H-NMR spectroscopy utilized a Bruker AMX
500 FT (500 MHz) instrument.
6.1.3. Synthesis of compounds 4a–f
Compound 4b has been prepared previously [3]. The
semicarbazones 4a and 4c–f were synthesized as follows.
A solution of the appropriate aldehyde (0.01 mol) in
ethanol (95% v/v, 25 mL) was added to a solution of
semicarbazide hydrochloride (0.01 mol) and sodium ac-
etate (0.013 mol) in water (10 mL). The mixture was
stirred at room temperature for 1 h, cooled (4 °C) and the
precipitates were collected and recrystallized from etha-
nol (80% v/v, 4a) or ethanol (4c–f) to give the semicar-
bazones. The m.p.’s and yields are: 4a: 93–95° (lit. [18]
100°), 92%; 4c: 160–161° (lit. [19] 161.5–162.5°), 92%;
4d: 166–168° (lit. [20] 163°), 72%; 4e: 162–164°
(lit. [19] 161–162°), 96% and 4f: 193–195° (lit. [21]
6.1.1. Synthesis of compounds 2a–l, and 5a and 5b
The acetylhydrazone 2f was prepared by the method
described previously [12]. The remaining analogues were
synthesized as follows. A mixture of the appropriate
aldehyde or ketone (0.01 mol) and acetic acid hydrazide
(0.01 mol) in ethanol (25 mL) was stirred at room tem-
perature for 3 h (2a–e, and 5a and 5b), 24 h (2i and 2l),
48 h (2g, 2h and 2j) or 72 h (2k). The mixture was
concentrated in vacuo and the precipitated compound was
collected by filtration and recrystallized from ethanol
(2a–c, 2l and 5a), ethyl acetate (2d, 2e and 2i), ethanol
(70% v/v, 2g, 2h, 2j and 2k) or hexane (5b) to yield the
acetylhydrazones. The m.p.’s and yields are: 2a:
133–135° (lit. [13] 136°), 42%; 2b: 136–138°, 84%; 2c:
198–200°, 76%; 2d: 128–130° (lit. [14] 132°), 27%; 2e:
129–131° (lit. [15] 139°), 62%; 2g: 120–122° (lit. [16]
131–132°), 71%; 2h: 163–165°, 83%; 2i: 153–155°,
75%; 2j: 160–162°, 69%; 2k: 150–152°, 92%; 2l:
176–178°, 75%; 5a: 164–166° (lit. [15] 171°), 81% and
1
211–213°), 93%. The H-NMR spectrum of a represen-
tative compound 4c was as follows: δ (DMSO-d6): 0.84
(t, 3H, CH3), 1.24 [def s, 10H, (CH2)5], 1.26 [def t, 2H,
CH3 (CH2)5CH2], 2.12 (q, 2H, CH2CH=CH), 5.95–6.07
(m, 2H, CH=CH), 6.18 (s, 2H, NH2), 7.47 (d, 1H,
CH=N), 9.92 (s, 1H, NHCONH2) ppm.
6.1.4. Charge density calculations
The structures of the compounds 1a and 1g, 2a and 2g,
3a and 3g, 4b and 4e and 5a–d were built and minimized
using the HyperChem molecular modelling pro-
gramme [22]. Initially the molecules were optimized
using the MM+ molecular mechanics module with the
Polak-Ribiere algorithm (conjugant gradient). Molecules
were deemed to be minimized when there was a mini-
mum energy difference between cycles of less than

248
0.001 kcal/mol. Non-bonded electrostatic interactions
were calculated using bond dipole interactions. The
minimum energy conformers from the MM+ calculations
were obtained and the CNDO semiempirical force field
was used to calculate their wave functions from which the
charges listed in table I were generated.
and the University of Saskatchewan who awarded a
University Graduate Scholarship to S.C. Vashishtha. In
addition, gratitude is expressed to the Antiepileptic Drug
Development Program, NIH, USA, contracted through
the laboratories of Drs H. Wolf, S. White and M. Franklin
at the University of Utah, Salt Lake City, USA, for
generation of the biodata described herein. The advice of
Dr Z. Zimpel, Department of Chemistry, University of
Saskatchewan, regarding the atomic charge calculations
is gratefully acknowledged. Mrs S. Thiessen, Mrs Z.
Dziadyk, Mrs D. Johnson and Mrs C. Jamont are thanked
for typing various drafts of the manuscript.
6.2. Anticonvulsant evaluations
The screening of the compounds for anticonvulsant
activity and neurotoxicity was undertaken by the National
Institute of Neurological Disorders and Stroke, Bethesda,
USA according to their protocols [8].
Side effects were noted in the mouse intraperitoneal
NT screen in the case of the following compounds (dose
of compound in mg/kg and time of observation in h in
parentheses): anaesthesia: 1g (300, 0.5), loss of righting
reflex: 2b (300, 0.5), 4d (300, 0.5), death: 2g (300, 0.5
and 4), 2j (300, 4), 4a (300, 0.5), 4d (300, 0.5 and 4),
unable to grasp rotorod: 2a (300, 0.5), 2b (300, 4), 2c
(300, 4), 2d (300, 0.5 and 4), 2f (300, 0.5 and 4), 2h (300,
4), 2i (300, 4), 2k (300, 4), 2l (300, 4), 4a (300, 0.5), 5a
(300, 4), 5b (300, 0.5). The neurological disorders noted
in the scPTZ test were as follows: tonic extension: 2a
(100, 4), 2h (30, 0.5), 2j (30 and 300, 0.5), 3f (100, 0.5),
5b (30, 0.5), continuous seizure activity: 2a (300, 4), 3f
(100, 0.5), 5a (30, 100 and 300, 0.5), 5b (300, 4), death
following tonic extension: 2i (300, 4), death following
clonic seizure: 2l (100 and 300, 4). In the rat oral scPTZ
screen, death following continuous seizure activity was
noted 0.25, 0.5 and 2 h after administration of 25 mg/kg
of 2f. In the case of phenytoin, NT was determined using
a dose of 1 000 mg/kg at the following times (number of
rats displaying neurotoxicity out of eight) ie. 0.25 (0), 0.5
(0), 1 (0), 2 (0), 4 (0), 6 (2), 8 (2) and 24 (1) h.
References
[1] Sanders J.W., Epilepsia 34 (1993) 1007–1016.
[2] Edafiogho I.O., Scott K.R., in: Wolf M.E. (Ed.), Burger’s Medicinal
Chemistry and Drug Discovery, John Wiley and Sons Inc., NY,
1996, p. 175.
[3] Dimmock J.R., Sidhu K.K., Thayer R.S., Mack P., Duffy M.S., Reid
R.S. et al., J. Med. Chem. 36 (1993) 2243–2252.
[4] Dimmock J.R., Sidhu K.K., Tumber S.D., Basran S.K., Chen M.,
Quail J.W. et al., Eur. J. Med. Chem. 30 (1995) 287–301.
[5] Dimmock J.R., Pandeya S.N., Quail J.W., Pugazhenthi U., Allen
T.M., Kao G.Y., Balzarini J., De Clercq E., Eur. J. Med. Chem. 30
(1995) 303–314.
[6] Dimmock J.R., Vashishtha S.C., Stables J.P., Pharmazie 50 (1995)
823–824.
[7] Scovill J.P., Klayman D.L., Franchino C.F., J. Med. Chem. 25 (1982)
1261–1264.
[8] Stables J.P., Kupferberg H.J., in: Avanzini G., Tanganelli P., Avoli
M. (Eds.), Molecular and Cellular Targets for Antiepileptic Drugs,
John Libbey & Co. Ltd., London, 1997, pp. 191–198.
[9] Dunham M.S., Miya T.A., J. Am. Pharm. Ass. Sci. Ed. 46 (1957)
208–209.
[10] Topliss J.G., J. Med. Chem. 20 (1977) 463–469.
[11] Dimmock J.R., Baker G.B., Epilepsia 35 (1994) 648–655.
Quantitation of 4c in the mouse i.p. MES, scPTZ and
NT screens was undertaken at the time of peak effect,
namely 1 h. The ED50 and TD50 figures (95% confidence
intervals) in the MES, scPTZ and NT screens for 4c were
21.26 (14.44–30.61), 53.88 (34.51–74.05) and 263.85
(140.96–628.86) mg/kg, respectively. The related figures
for phenytoin were 6.32 (5.44–7.23), > 50 and 41.2
(36.9–46.1) mg/kg, respectively, determined 1, 1 and
0.5 h, respectively, after administration of the drug.
[12] Dimmock J.R., Puthucode R.N., Lo M.S., Quail J.W., Yang J.,
Stables J.P., Pharmazie 51 (1996) 83–88.
[13] Horner L., Fernekess H., Chem. Ber. 94 (1961) 712–724.
[14] Gillis B.T., Schimmel K.F., J. Org. Chem. 27 (1962) 413–417.
[15] Grammatakis P., Bull. Soc. Chim. France (1950) 690–698.
[16] Turner R.A., J. Am. Chem. Soc. 69 (1947) 875–877.
[17] Moshchitskee S.O., Sologub L.S., Pavlenko A.F., Akkerman V.P., Zh.
Org. Chem. 2 (1966) 2164–2167; Chem. Abstr. 66 (1967) 75960j.
[18] Buckingham J., Dictionary of Organic Compounds, volume 5, 5th
edition, Chapman and Hall, NY, 1982, p. 4849.
[19] Buckingham J., Dictionary of Organic Compounds, volume 5, 5th
edition, Chapman and Hall, NY, 1982, p. 6449.
Acknowledgements
[20] Romburgh P.V., Rec. Trav. Chim. 57 (1938), 494–499.
[21] Braude E.A., Evans E.A., J. Chem. Soc. (1955), 334–337.
The authors thank Apotex Inc. of Toronto, Ontario,
Canada for providing financial assistance for this study
[22] HyperChem Release 5.0 for Windows (1996), Hypercube, Inc.,
Waterloo, Ontario, Canada.