N-, α-, and β-substituted 3-aminopropionic acids: Design, syntheses and antiseizure activities

Article abstract of DOI:10.1016/S0968-0896(02)00330-9

A treatment for epilepsy is proposed based on analogues of 3-aminopropionic acid (β-alanine), a putative neurotransmitter in the central nervous system (CNS). A model three point pharmacophore was proposed based on modelling data obtained from the study of antagonists for both the glial γ-aminobutyric acid (GABA)-uptake site and the glycine co-agonist site of N-methyl-D-aspartate (NMDA) receptor. Three series of 3-aminopropionic acids containing, N-, α-, and β-substituents, were designed and synthesized to probe the position and the size of a lipophilic binding pocket within the proposed pharmacophore. These analogues were tested in vivo for both their antiseizure activities and their neurologic toxicities. Among the fourteen novel 3-aminopropionic acids synthesized, eight were found to have promising antiseizure activity. This study shows that substitution on the N-terminus confers the greatest antiseizure activity, particularly against pilocarpine-induced seizures.

Full text of DOI:10.1016/S0968-0896(02)00330-9

Bioorganic & Medicinal Chemistry 11 (2003) 113–121
N-, ꢀ-, and ꢁ-Substituted 3-Aminopropionic Acids:
Design, Syntheses and Antiseizure Activities
C.Y.K. Tan,a D. Wainmana and D.F. Weavera,b,
*
aDepartment of Chemistry, Queen’s University, Kingston, Ontario, Canada K7L 3N6
bDepartments of Neurology and Chemistry, Dalhousie University, Halifax, Nova Scotia, Canada, B3H 4J3
Received 24 September 2001; accepted 28 June 2002
Abstract—A treatment for epilepsy is proposed based on analogues of 3-aminopropionic acid (b-alanine), a putative neuro-
transmitter in the central nervous system (CNS). A model three point pharmacophore was proposed based on modelling data
obtained from the study of antagonists for both the glial ꢀ-aminobutyric acid (GABA)-uptake site and the glycine co-agonist site of
N-methyl-d-aspartate (NMDA) receptor. Three series of 3-aminopropionic acids containing, N-, a-, and b-substituents, were
designed and synthesized to probe the position and the size of a lipophilic binding pocket within the proposed pharmacophore.
These analogues were tested in vivo for both their antiseizure activities and their neurologic toxicities. Among the fourteen novel
3-aminopropionic acids synthesized, eight were found to have promising antiseizure activity. This study shows that substitution on
the N-terminus confers the greatest antiseizure activity, particularly against pilocarpine-induced seizures.
# 2002 Elsevier Science Ltd. All rights reserved.
Introduction
More recent approaches to the design of antiseizure
drugs have involved (i) antagonizing excitatory neuro-
transmitters, (ii) mimicking inhibitory neuro-
transmitters, or (iii) influencing postsynaptic receptors.
Our rational design concept in this study was to simul-
taneously target both (i) and (ii) by designing com-
pounds that could potentially decrease excitation while
concomitantly increasing inhibition. This unique
approach would theoretically be of greater utility in
controlling epilepsy.
Epilepsy is a medical disorder characterized by sponta-
neous, recurrent seizures arising from excessive elec-
trical activity in a portion of the brain. Epilepsy affects
1.5–2.0% of the population in developing countries, and
0.8–1.3% in developed countries; four million people
are affected by this condition in North America and
Europe alone. Moreover, since epilepsy typically afflicts
people during their youth, its socioeconomic impact is
disproportionately high. Currently available drugs such
as phenytoin, carbamazepine, valproic acid, lamo-
trigine, and topiramate (Fig. 1) provide symptomatic
seizure suppression in only 60–70% of those receiving
treatment.1,2 These drugs are also associated with
undesirable side-effects, ranging from cosmetic (gingival
hyperplasia) to life threatening (hepatotoxicity, mega-
loblastic anemia).3ꢀ7 Consequently, a need exists for the
development of new antiseizure drugs with improved
efficacy and tolerability.
After extensive background design research, b-alanine
was selected as the prototype lead compound around
which to design new molecules. Various studies have
suggested that b-alanine may act as an inhibitory neu-
rotransmitter: b-alanine occurs naturally in the central
nervous system (CNS), is released by electrical stimula-
tion, can inhibit neuronal excitability, and has binding
sites.8ꢀ13 The existence of specific b-alanine receptors is
controversial: some investigators propose the existence
of specific receptors.14 Others, however, have clearly
shown that b-alanine binds to both glycine and ꢀ-ami-
nobutyric acid (GABA) binding sites.15 In addition,
numerous studies have verified the dual pharmacologi-
cal action of b-alanine on both glutamatergic and
Traditionally, drug design in epilepsy has targeted the
neuronal voltage gated Na+ channel protein, resulting
in the majority of currently available antiseizure drugs.
GABAergic
processes.16,17
b-Alanine
increases
GABAergic inhibition by blocking the glial GABA
uptake site,18,19 and decreases glutamatergic excitation
*Corresponding author at second address. Tel.: +1-902-494-7183; fax:
+1-902-494-1310; e-mail: weaver@chem3.chem.dal.ca
0968-0896/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(02)00330-9

114
C. Y. K. Tan et al. / Bioorg. Med. Chem. 11 (2003) 113–121
Figure 1. Structures of some current clinically available antiseizure drugs.
by binding to the glycine co-agonist site on the
N-methyl-d-aspartate (NMDA) receptor.20,21 In addi-
tion, b-amino acid active transport shuttles, capable of
transporting b-alanine and related analogues across the
blood–brain barrier, have also been identified.22ꢀ24 It is
this unique dual property of b-alanine to simultaneously
decrease excitation and increase inhibition, as well as its
ability to cross the blood–brain barrier via an active
transport shuttle, which supports our proposal of
b-amino acid analogues as potential new antiseizure
agents.
(ii) a lipophilic binding region (preferably aromatic),
(iii) a carboxylic acid functional group and (iv) an elec-
tron rich functionality (double bond or an oxygen)
located between the amine and the lipophilic region.
Leeson,26 in studying antagonists of the glycine co-ago-
nist site on the NMDA receptor, suggested that a phar-
macophore (Fig. 2) for this receptor has the following
parameters: (i) a secondary amine functional group, (ii)
a carboxylic acid functional group, (iii) two small lipo-
philic groups as well as a larger lipophilic group. Using
these pharmacophore models, a hybrid three point
pharmacophore (Fig. 2) incorporating the b-amino acid
backbone and a lipophilic moiety was proposed to
interact computationally with models of both the glial
GABA uptake site and the glycine subsite on the
NMDA receptor.
In this paper, we report a novel approach to the treat-
ment of epilepsy through analogues of b-alanine. A
model three-point b-amino acid pharmacophore based
on molecular modelling data obtained from studies by
N’Goka25 and Leeson26 is hypothesized. Fourteen
3-aminopropionic acid analogues have been synthesized
and tested for both their antiseizure activities as well as
their neurologic toxicities. In vivo antiseizure animal
models used in this study include pilocarpine, maximal
electroshock (MES) and pentylenetetrazole (PTZ)
assays.
We designed and synthesized three series of analogues
having lipophilic moieties in the N-, a- and b-positions
to probe the location of the lipophilic binding pocket.
C-terminus substitution in the form of an ester group
was not considered because the ester linkage is suscep-
tible to hydrolytic cleavage by various esterases. We
examined the location and the depth of the proposed
lipophilic pocket by extending the alkyl chain linking
the lipophilic moiety to the b-amino acid backbone.
Specifically, 2-phenylethyl and 3-phenylpropyl groups
were incorporated at the N-, a- and b-positions. Once
the positions and the depth responsible for the biologi-
cal activity were determined, the size of the lipophilic
pocket was probed using bulkier 2,2-diphenylethyl and
3,3-diphenylpropyl substituents. The structures and
yields of these 3-aminopropionic acids are summarized
in Table 1.
The study was divided into two parts. First, two homo-
logues for each of N-, b-, and a-substituted 3-amino-
propionic acids were synthesized to probe the position
and the depth of the corresponding lipophilic binding
pocket for the proposed pharmacophore. Second, the
size of this lipophilic binding pocket was probed using
bulkier lipophilic side chains than used in the initial
analogues. The structure–activity relationship of these
analogues will be discussed.
The desired N-substituted 3-aminopropionic acids were
prepared by a Lewis acid catalyzed Michael addition of
methyl acrylate and the appropriately substituted amine to
give their corresponding N-substituted-3-aminopropio-
nates. Subsequent saponification followed by acidification
yielded the respective N-substituted-3-aminopropionic
acids as hydrochloride salts (Scheme 1). The correspond-
ing amide derivatives of these N-substituted-3-aminopro-
pionic acids were also synthesized to investigate whether
C-terminal substitutions would affect the biological
Results and Discussion
In designing b-alanine analogues capable of interacting
in silico with both the GABA glial uptake site and the
glycine co-agonist site on the NMDA receptor, we
exploited previously proposed computational pharma-
cophore models for these receptor sites. N’Goka25 pro-
posed that the binding site for the glial GABA-uptake
receptor (Fig. 2) has the following parameters: (i) an
amine functional group (preferably a secondary amine),

C. Y. K. Tan et al. / Bioorg. Med. Chem. 11 (2003) 113–121
115
Figure 2. Pharmacophore models for GABA-uptake and glycine NMDA receptor sites25,26 and proposed common pharmacophore.
Table 1. Structures and yields (overall yield) of 3-aminopropionic acid analogues
N-substituted
Yield (%)
86.2
b-substituted
Yield (%)
50.1
a-substituted
Yield (%)
56.7
(1)
(5)
(7)
(2)
(3)
84.5
69.6
(6)
43.2
51.2
(8)
51.4
38.9
(11)
(13)
(4)
67.1
78.2
80.3
(12)
32.8
(14)
45.7
(9)
(10)

116
C. Y. K. Tan et al. / Bioorg. Med. Chem. 11 (2003) 113–121
Scheme 1. Synthesis of N-substituted-3-aminopropionic acid hydrochloride salt.
Scheme 2. Synthesis of N-substituted-3-aminopropionoamide hydrochloride salt.
Scheme 3. Synthesis of b-substituted 3-aminopropionic acid hydrochloride salt.
activity of these analogues. These analogues were pro-
duced by refluxing the hydrochloride salts with the cor-
responding amine in 2 N HCl (Scheme 2).
substitution reactions using an appropriately functiona-
lized alkyl bromide followed by N- (bromo-
methyl)phthalimide gave disubstituted diethyl
a
malonate. Hydrolysis of this intermediate afforded the
corresponding a-substituted 3-aminopropionic acid as
the hydrochloride salt (Scheme 4).
The b-substituted 3-aminopropionic acids were prepared
as described by Katritzky.27 The condensation of benzo-
triazole with benzyl carbamate and an appropriately func-
tionalized aldehyde gave the 1-benyloxycarbonylamino-1-
(benzotriazolyl)alkane intermediate. Subsequent reaction
with ethyl bromoacetate under Reformatsky type condi-
tions afforded the N-protected 3-aminoalkanoic ester. This
intermediate was then hydrolyzed with concentrated
hydrochloric acid to give the desired b-substituted 3-ami-
nopropionic acid hydrochloride salt (Scheme 3).
All analogues synthesized were tested in vivo for both
their antiseizure activities and their neurological toxi-
cities. Seizures are divided into two major groups: par-
tial (focal) and primary generalized. Partial seizures,
accounting for approximately 60% of all reported
human cases of epilepsy, arise from a localized area of
the brain. They are subdivided into simple partial,
complex partial, and secondarily generalized seizures.
Primary generalized seizures account for 40% of all
reported cases of epilepsy and onset with no apparent
focality. Primary generalized seizures may further be
The a-substituted 3-aminopropionic acids were synthe-
sized using a modified procedure first described by
Bohme.28 Starting with diethyl malonate, sequential
Scheme 4. Synthesis of a-substituted 3-aminopropionic acid hydrochloride salt.

C. Y. K. Tan et al. / Bioorg. Med. Chem. 11 (2003) 113–121
117
Table 2. Antiseizure and neurotoxicity activities of the 3-aminopro-
pionic acid analoguesa
data: (i) there is only one lipophilic pocket located in
close proximity to both the b-carbon and the N-termi-
nus, or (ii) there is more than one lipophilic pocket with
one being in close proximity to both the b-carbon and
the N-terminus.
Compound
Pilocarpine
MES
scPTZ
Toxicity
30 min 4 h 30 min 4 h 30 min 4 h
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Active
Active
Active
—
Active
—
—
Active
—
Active
Active
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
+
+
—
—
—
—
NE
148.6
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
To investigate these two hypotheses, analogues con-
taining a benzyl substituent in the b-position (11) as well
as the 2,2-diphenylethyl substituent in the N-, a- and
b-positions, (9), (13), (12) respectively, were synthesized
and evaluated for their antiseizure activity. The results
showed that the benzyl b-substituted analogue (11) was
active against pilocarpine-induced seizures while its
bulkier b-2,2-diphenylpropyl analogue (12) was not.
This suggested that there is indeed a lipophilic pocket
close to the b-carbon, in agreement with both hypoth-
esis (i) and (ii). However, this lipophilic pocket is rela-
tively small as demonstrated by the lack of activity upon
the incorporation of a bulkier substituent. Of the two
2,2-diphenylethyl substituted analogues, (9) and (13),
only the a-substituted analogue (13) was active. In con-
trast, the longer 3,3-diphenylpropyl homologues resul-
ted in a reversal of activity such that the N-substituted
analogue (10) was active against pilocarpine-induced
seizures while the a-substituted analogue was not. These
results suggested that two separate lipophilic pockets
are responsible for the different activities for both the N-
and the a-substituted series, supporting the second
hypothesis. Of these six analogues (9, 10, 11, 12, 13 and
14), only the N-substituted analogues (9) and (10) were
found to have activity against MES and PTZ-induced
seizures at higher doses. Why only bulky lipophilic
groups in the N-positions on the 3-aminopropionic acid
backbone influence MES and PTZ activity is not readily
apparent. Overall, the results suggest that there may be
more than one lipophilic pocket, and that a small lipo-
philic pocket is located in close proximity to both the
b-carbon and the N-terminus of 3-aminopropionic acid.
In addition, substitution on the N-terminus is the most
desirable position for antiseizure activity, particularly
against pilocarpine-induced seizures.
++ NT
—
—
—
—
—
—
+
—
—
+
—
71.6
425.8
—
—
—
—
—
—
—
—
—
—
—
Active
—
Active
Active
Carbamazepineb
Valproic acidb
8.8
271.7
Abbreviations: MES, maximal electroshock-induced seizures; scPTZ;
subcutaneous pentylenetetrazole; s, seizure; NT, not tested; NE, not
effective.
aPilocarpine test was carried out at a dose of 100 mg/kg; each analogue
was tested on 4 rats; a compound is active if 3 or 4 of the rats were
fully prevented from having seizures. The antiseizure (MES and
scPTZ) and neurotoxicity activities were determined 30 min and 4 h
after the administration of compounds. The symbols +++, ++,
and + signify activity (seizures prevented in (3 or 4)/4 rats) at 30, 100,
and 300 mg/kg, respectively; —denotes no activity observed at 300 mg/
kg. Toxicity was determined by the rotorod test.
bCommercially available antiseizure drugs. Both MES and scPTZ are
reported as the median effective dose (ED50), and toxicity is reported
as the median toxic dose (TD50) in mg/kg.31
divided into absence (petit mal), myoclonic, clonic,
tonic, tonic-clonic (grand mal) or atonic seizures. To
reflect this diversity of seizure types, all 3-aminopro-
pionic acid analogues were evaluated in three separate
antiseizure assays: pilocarpine, MES (maximal electro-
shock) and scPTZ (subcutaneous pentylenetetrazole).
Each assay models a different seizure type. Pilocarpine-
induced seizures are a model for complex partial seizures;29
MES-induced seizures model tonic-clonic seizures;30 PTZ-
induced seizures model absence seizures.31 Rotorod
neurotoxicity testing for each analogue was also per-
formed. Results are summarized in Table 2.
Most of the 14 analogues displayed no neurological
toxicity as determined by the rotorod test, except for
analogues (3), (10) and (13). However, this neurotoxi-
city was comparable to the commercially available
antiseizure drugs such as carbamazepine and valproic
acid (Table 2). Analogue (3) is an amide of the corre-
sponding N-substituted analogue (2) whereas analogue
(10) and (13) are diphenyl substituted. The addition of
the amide group (analogues 3 and 4) on the C-terminus
either depleted antiseizure activity or augmented the
neurotoxicity to the N-substituted series.
In probing the location and the depth of the lipophilic
pocket, analogues containing either 2-phenylethyl or
3-phenylpropyl substituents in one of the N-, a- or
b-positions were tested for their antiseizure activity.
Both the N-substituted 2-phenylethyl- (1) and the
3-phenylpropyl-3-aminopropionic acids (2) were found
to have antiseizure activity against pilocarpine-induced
seizures. Of the b-substituted analogues, only the
2-phenylethyl analogue (5) was active against pilo-
carpine-induced seizures, whereas the 3-phenylpropyl
analogue (6) was not. In contrast, only the a-substituted
3-phenylpropyl analogue (8) showed activity against
pilocarpine-induced seizures. All six analogues were
inactive in the MES and PTZ models. These analogues
exhibit selectivity towards pilocarpine-induced seizures
and therefore may be of interest in the treatment of
complex partial seizures. Two possible structural
hypotheses were proposed to explain these biological
In conclusion, we have proposed a unique approach to
the treatment of epilepsy based on b-alanine, a putative
neurotransmitter in the CNS. Interestingly, these com-
pounds had activity against pilocarpine induced seizures
(a model of complex partial seizures) but less activity
against MES or PTZ induced seizures. This observation
is noteworthy given the need for drugs specific for com-
plex partial seizures and may reflect the high density of
GABA and NMDA receptor populations in the limbic

118
C. Y. K. Tan et al. / Bioorg. Med. Chem. 11 (2003) 113–121
regions of the brain from which complex partial seizures
preferentially originate. As a backbone upon which to
design anticonvulsants, b-alanine is active at both
GABA and NMDA receptor populations. Future bind-
ing studies will evaluate the structure–activity relation-
ships of the binding of these b-amino acid analogues to
GABA-A and NMDA-gly receptor sites.
Hydrochloride salt of N-(2,2-diphenylethyl)-3-aminopro-
pionic acid (9). White crystals (9) were obtained, mp:
197–198 ꢂC; IR: 3421, 1718, 1600, 1562 cmꢀ1; 1H NMR:
d 7.29–7.24 (m, 8H), 7.21–7.12 (m, 2H), 4.34 (t, 1H,
J=8.03 Hz), 3.69 (d, 2H, J=8.03 Hz), 3.24 (t, 2H,
J=6.79 Hz), 2.64 (t, 2H, J=6.92 Hz); 13C NMR:
172.71, 140.27, 129.15, 127.84, 127.61, 51.66, 48.46,
44.18, 29.52; Anal. calcd for C17H20NO2Cl: C, 66.77; H,
6.59; N, 4.58. Found: C, 66.40; H, 6.60; N, 4.59.
Experimental
Chemistry
Hydrochloride salt of N-(3,3-diphenylpropyl)-3-amino-
propionic acid (10). 10 as white crystals was obtained,
mp 179–180 ꢂC; IR: 3447, 1742, 1589 cmꢀ1; 1H NMR: d
7.44–7.42 (m, 8H), 7.14–7.00 (m, 2H), 3.92 (t, 1H,
J=7.95 Hz), 3.05 (t, 2H, J=6.45 Hz), 2.83–2.78 (m,
2H), 2.60 (t, 2H, J=6.45 Hz), 2.38–2.30 (m, 2H); 13C
NMR: 172.92, 143.77, 128.88, 127.75, 126.85, 48.75,
47.24, 43.46, 31.62, 30.25; Anal. calcd for
C18H22NO2Cl: C, 67.60; H, 6.93; N, 4.38. Found: C,
67.54; H, 7.08; N, 4.31.
Proton (1H) and carbon (13C) nuclear magnetic reso-
nance (NMR) spectra were recorded on a Bruker
Avance 300 (300 MHz) spectrometer with CD3OD as
the solvent unless otherwise stated. Infrared (IR) spec-
tra were recorded on a Bomem MB-120 spectro-
photometer using KBr disks. Melting points (mp) were
determined using a Mel-Temp II capillary apparatus
and are uncorrected. Elemental analyses were per-
formed by G-C-L Laboratories (Guelph, Canada). Sol-
vents were purified using standard methods.
General procedure for N-substituted-3-aminopropiona-
mide hydrochloride salt. The corresponding amine (1.0
equiv) was added to the N-substituted-3-aminopro-
pionic acid hydrochloride salt in 150 mL of 2 N HCl
aqueous. The reaction mixture was allowed to reflux
overnight and was evaporated to dryness under reduced
pressure to give a white solid. Subsequent recrystalliza-
tion from EtOAc then afforded the final product.
General procedure for N-substituted-3-aminopropionic
acid hydrochloride salt. Under N2, a solution of methyl
acrylate (1.0 equiv) in 50 mL dry CH2Cl2 was added via
a dropping funnel to a stirred mixture of amine (1.0
equiv) and MgBr2.OEt2 (0.3 equiv) in 100 mL dry
CH2Cl2. The reaction mixture was allowed to stir at
room temperature until completion, as determined by
TLC. The reaction mixture was then quenched with
H2O. The organic layer was separated and concentrated
under reduced pressure to give a colourless oil. It was
dissolved in minimal amount of MeOH and saponi-
fied with aqueous NaOH (1.0 equiv). The reaction
mixture was washed 3 times with Et2O and acidified
with 1.0 N aqueous HCl (pH=1). The reaction mix-
ture was evaporated to dryness under reduced pres-
sure to give a white solid (containing the product and
NaCl). NaCl was removed by repeatedly washing and
drying with absolute EtOH (3ꢁ50 mL). The final
product was recrystallised with EtOH/EtOAc to give
white crystals.
Hydrochloride salt of N0-(20-phenylethyl)-N-(2-phenyl-
ethyl)-3-aminopropionoamide (3). White crystals of 3
were obtained, mp: 242–243 ꢂC; IR: 3314, 1649,
1454 cmꢀ1; 1H NMR: d 7.31–7.06 (m, 10H), 3.35 (t, 4H,
J=6.48 Hz), 3.18–3.14 (m, 4H), 2.96–2.89 (m, 2H), 2.71
(t, 2H, J=7.10 Hz), 2.53 (t, 2H, J=6.50 Hz); 13C NMR:
170.88, 139.20, 136.65, 129.04, 128.78, 128.53, 127.35,
126.43, 44.11, 40.97, 37.98, 35.42, 33.72, 30.67; Anal.
calcd for C19H25N2OCl: C, 68.56; H, 7.57; N, 8.42.
Found: C, 68.90; H, 7.87; N, 8.54.
Hydrochloride salt of N0-(30-phenylpropyl)-N-(3-phenyl-
propyl)-3-aminopropionoamide (4). 4 as white crystals
was obtained, mp 173–174 ꢂC; IR: 3318, 1642,
1
1578 cmꢀ1; H NMR: d 7.26–7.07 (m, 10H), 3.20–3.14
Hydrochloride salt of N-(2-phenylethyl)-3-aminopropio-
nic acid (1). White crystals of 1 were obtained, mp:
193–194 ꢂC; IR: 3462, 1741, 1604, 1577 cmꢀ1; 1H NMR:
d 7.21–7.17 (m, 5H), 3.25–3.16 (m, 4H), 2.94–2.89 (m,
2H), 2.72–2.67 (m, 2H); 13C NMR: 171.30, 136.71,
129.00, 128.87, 127.30, 49.30, 43.46, 32.22, 30.22; Anal.
calcd for C11H16NO2Cl: C, 57.52; H, 7.02; N, 6.10.
Found: C, 57.12; H, 7.40; N, 5.72.
(m, 4H), 2.99–2.95 (m, 2H), 2.67 (t, 2H, J=5.75 Hz),
2.61–2.56 (m, 4H), 2.01–1.93 (m, 2H), 1.80–1.73 (m,
2H); 13C NMR: 170.75, 141.83, 140.52, 128.58, 128.34,
126.34, 125.86, 47.32, 43.94, 39.04, 33.10, 32.41, 31.00,
30.74, 27.98; Anal. calcd for C21H29N2OCl: C, 69.88; H,
8.10; N, 7.76. Found: C, 69.77; H, 8.09; N, 7.73.
General procedure for ꢁ-substituted-3-aminopropionic
acid hydrochloride salt. (1) Benzotriazole (1.0 equiv),
benzyl carbamate (1.0 equiv), an aldehyde (1.0 equiv),
and a catalytic amount of p-toluenesulfonic acid mono-
hydrate (100 mg) were refluxed in toluene (150 mL)
using a Dean-Stark trap until the theoretical amount of
water was produced. The reaction mixture was con-
centrated under reduced pressure to produce 1-benzy-
loxycarbonylamino-1-(1-benzotriazolyl)alkane as an
oily residue which was used directly in the next step. (2)
Under N2, chloromethylsilane (0.3 equiv) was added
Hydrochloride salt of N-(3-phenylpropyl)-3-aminopropio-
nic acid (2). 2 as white crystals was obtained, mp: 141–
142 ꢂC; IR: 2954, 1729, 1604, 1573 cmꢀ1; H NMR: d
1
7.12–6.96 (m, 5H), 3.06 (t, 2H, J=6.72 Hz), 2.85 (t, 2H,
J=7.92 Hz), 2.66–2.56 (m, 4H), 1.89–1.78 (m, 2H); 13C
NMR: 172.53, 140.68, 128.67, 128.46, 126.42, 47.71,
43.48, 32.53, 30.13, 27.93. Anal. calcd for
C12H18NO2Cl: C, 59.14; H, 7.44; N, 5.75. Found: C,
59.30; H, 7.84; N, 5.49.

C. Y. K. Tan et al. / Bioorg. Med. Chem. 11 (2003) 113–121
119
into a solution of dried THF (100 mL) with Zn powder
(1.5 equiv). The reaction mixture was stirred at room
room temperature until all the Na was dissolved. Die-
thylmalonate (1.1 equiv) was added and the reaction
mixture was heated to 40 ꢂC. An alkyl bromide (1.0
equiv) was then added via a dropping funnel. The reac-
tion mixture was then refluxed until completion, as
determined by TLC. The NaBr was filtered off and the
filtrate was concentrated under reduced pressure to give
a pale yellow oil. Excess diethylmalonate was distilled
off by high vacuum distillation to yield the mono-
substituted-diethylmalonate. (2) Under N2, Na (1.0
equiv) was added to dry THF (150 mL) containing the
monosubstituted-diethylmalonate and the reaction mix-
ture was warmed to 40 ꢂC and stirred until the Na was
completely dissolved. N-(bromomethyl)phthalimide (1.0
equiv) dissolved in dry THF (50 mL) was added via a
dropping funnel. Once the N-(bromomethyl)phthali-
mide was added, the reaction mixture was refluxed until
completion, as determined by TLC. NaBr was filtered
off and the filtrate was concentrated under reduced
pressure to give a yellow oil. This yellowish oil was fur-
ther dissolved in 100 mL of EtOAc and washed with
H2O (3ꢁ50 mL) before the EtOAc was dried over
MgSO4 and concentrated under reduced pressure to
give the disubstituted diethylmalonate as a pale yellow
solid. It was then recrystallized from EtOH/H2O to
yield a white solid. (3) The disubstituted-diethylmalo-
nate was hydrolysed to yield the a-substituted-3-amino-
propionic acid hydrochloride salt either by mineral acid
or by hydrazine hydrate followed by acids.
temperature
for
15 min
then
1-benzylox-
ycarbonylamino-1-(1-benzotriazolyl)alkane in dry THF
(30 mL) was slowly added via a dropping funnel. The
reaction mixture was heated briefly to reflux and
allowed to cool before ethyl bromoacetate (1.0 equiv)
was added. The suspension was then heated to reflux
and the nitrogen inlet was removed. The grey powder
soon turned into thin white suspension, and the reaction
was then allowed to reflux until completion, as deter-
mined by TLC. The reaction mixture was poured into
concentrated NH4OH (40 mL) containing ice (20 g), and
stirred until all solids dissolved. The organic layer was
extracted with Et2O (3ꢁ70 mL), washed with H2O
(3ꢁ70 mL) and dried over MgSO4 before the solvent
was removed under reduced pressure to give 3-dialkyla-
minoalkanoic ester adduct. (3) The ester adduct was
refluxed in 1:1:1 (HCl/H2O/HCO2H) until the hydro-
lysis was completed, as determined by TLC. The reac-
tion mixture was then evaporated to dryness under
reduced pressure and was recrystallized from EtOH/
EtOAc to give a b-substituted-3-aminopropionic acid
hydrochloride salt as white solids.
Hydrochloride salt of 3-amino-3-(2-phenylethyl)propionic
acid (5). White crystals of 5 were obtained, mp 183–
184 ꢂC; IR: 3010, 1598 cmꢀ1; H NMR: d 7.06–6.89 (m,
1
5H), 3.41–3.20 (m, 1H), 2.62–2.42 (m, 4H), 1.83–1.69
(m, 2H); 13C NMR: 173.93, 141.90, 129.91, 129.65,
127.65, 50.15, 37.04, 35.80, 32.62; Anal. calcd for
C11H16NO2Cl: C, 57.52; H, 7.02; N, 6.10. Found: C,
57.33; H, 7.31; N, 6.02.
With mineral acid, the disubstituted-diethylmalonate
was refluxed in a mixture of 1:1:3 (H2O/HCO2H/con-
centrated HCl) until the hydrolysis was complete, as
determined by TLC. The reaction mixture was then
evaporated to dryness under reduced pressure to give
white solids (a-substituted-3-aminopropionic acid
hydrochloride salt and phthalic acid). The a-sub-
stituted-3-aminopropionic acid hydrochloride salt was
isolated by strong acid ion exchange column (Amberlite
IR-120 plus), where it was eluted by 4 N NH4OH, eva-
porated to dryness and acidified with 1 N HCl. a-Sub-
stituted-3-aminopropionic acid hydrochloride salt was
than recrystallized from EtOH/EtOAc.
Hydrochloride salt of 3-amino-3-(3-phenylpropyl)propio-
nic acid (6). Compound 6 as white crystals was
;
obtained, mp 230–231 ꢂC; IR: 3004, 1615 cmꢀ1 1H
NMR: d 7.06–6.89 (m, 5H), 3.41–3.20 (m, 1H), 2.62–
2.42 (m, 4H), 1.83–1.69 (m, 2H); 13C NMR: 173.93,
141.90, 129.91, 129.65, 127.65, 50.15, 37.04, 35.80,
32.62. Anal. Calcd. for C12H18NO2Cl: C, 59.14; H, 7.44;
N, 5.75. Found: C, 59.31; H, 7.49; N, 5.65.
Hydrochloride salt of 3-amino-3-benzylpropionic acid
(11). White crystals (11) were obtained, mp 172–174 ꢂC;
With hydrazine hydrate, the disubstituted-diethylmalo-
nate was dissolved in warm EtOH. Hydrazine hydrate
(1.0 equiv) was added into the mixture and stirred for
15 min before 15 mL of 1 N HCl was added. The reac-
tion mixture was warmed to 40ꢂC before it was evapo-
rated to dryness to yield a white solid. The white solid
was then taken up with MeOH and the undissolved
solid was filtered off. The filtrate was evaporated to
dryness to give a white solid. The white solid was
refluxed in a mixture of 1:1:3 (H2O/HCO2H/con-
centrated HCl) until the hydrolysis was completed, as
determined by TLC. The reaction mixture was evapo-
rated to dryness and the white solid was recrystallized
with EtOH/EtOAc to yield the a-substituted-3-amino-
propionic acid as a hydrochloride salt.
1
IR: 3007, 1610 cmꢀ1; H NMR: d 7.24–7.15 (m, 5H),
3.70–3.62 (m, 1H), 2.99–2.80 (m, 2H), 2.53–2.49 (m, 2H);
13C NMR: 173.50, 136.70, 130.48, 130.10, 128.59, 50.95,
39.35, 36.07; Anal. calcd for C10H14NO2Cl: C, 55.69; H,
6.54; N, 6.49. Found: C, 55.81; H, 6.69; N, 6.52.
Hydrochloride salt of 3-amino-3-(2,2-diphenylethyl)pro-
pionic acid (12). White crystals of 12 were obtained, mp
207–209 ꢂC; IR: 3002, 1594 cmꢀ1; 1H NMR: d 7.27–7.00
(m, 10H), 4.04 (t, 1H, J=8.10 Hz), 3.23–3.14 (m, 1H),
2.71–2.49 (m, 2H), 2.44–2.29 (m, 2H); 13C NMR:
173.58, 144.46, 129.81, 128.81, 127.80, 48.35, 48.20,
39.03, 36.60; Anal. calcd for C17H20NO2Cl: C, 66.77; H,
6.59; N, 4.58. Found: C, 66.47; H, 6.72; N, 4.65.
1
General procedure for ꢀ-substituted-3-aminopropionic
acid hydrochloride salt. (1) Under N2, Na (1.0 equiv) in
dry absolute EtOH (100 mL) was allowed to stir at
2-Phenylethyl-diethylmalonate (I). H NMR (CDCl3): d
7.32–7.20 (m, 5H), 4.23 (q, 4H, J=7.20 Hz), 3.36 (t, 1H,
J=7.50 Hz), 2.68 (t, 2H, J=7.50 Hz), 2.28–2.20 (m,

120
C. Y. K. Tan et al. / Bioorg. Med. Chem. 11 (2003) 113–121
2H), 1.30 (t, 6H, J=7.20 Hz); 13C NMR: 170.39, 133.21,
128.54, 128.46, 126.19, 61.37, 51.29, 33.32, 30.34, 14.09.
150 mA in rat) for 0.2 s via corneal electrodes primed with
an electrode solution containing an anesthetic agent. The
protection was defined as the abolition of hind-leg tonic
maximal extension component of the seizure.
3-Phenylpropyl-phthalimidomethyl-diethylmalonate (II).
II as white crystals was obtained, mp 74–76 ꢂC; 1H
NMR (CDCl3): d 7.88–7.85 (m, 2H), 7.75–7.72 (m, 2H),
7.31–7.19 (m, 5H), 4.32–4.24 (q, 4H, J=7.20 Hz), 2.79–
2.73 (m, 2H), 2.14–2.08 (m, 2H), 1.33 (t, 6H,
J=7.20 Hz); 13C NMR: 171.90, 170.23, 143.32, 136.13,
133.80, 130.57, 130.32, 127.95, 125.15, 63.77, 59.38,
41.36, 34.79, 32.19, 15.98.
The scPTZ seizure threshold assay was carried out by
an ip administration of PTZ (85 mg/kg in mice and
70 mg/kg in rats). Animals were observed over a 30 min
period and the absence of clonic spasms in the observed
time period were defined as protect.
Minimal neurotoxicity was measured by the rotorod
test. Mice were placed on a 1-in diameter knurled plastic
rod rotating at 6 rpm after the administration of the
drug, and their ability to maintain their balance was
tested. Neurological deficit was indicated by the inabil-
ity of the animal to maintain its equilibrium for 1 min
on the rotating rod in each of three trials.
Hydrochloride salt of 3-amino-2-(2-phenylethyl)propionic
acid (7). White crystals of 7 were obtained, mp 217–
218 ꢂC; IR: 3013, 1608 cmꢀ1; H NMR: d 7.17–7.01 (m,
1
5H), 3.11–2.91 (m, 2H), 2.65–2.54 (m, 3H), 1.94–1.79
(m, 2H); 13C NMR: 175.98, 142.23, 129.56, 129.50,
127.25, 43.38, 41.39, 33.81, 32.78; Anal. calcd for
C11H16NO2Cl: C, 57.52; H, 7.02; N, 6.10. Found: C,
57.44; H, 7.40; N, 6.07.
The pilocarpine assay was performed in the Division of
Neurology, Queen’s University. All compounds were
evaluated in male Sprague–Dawley rats. The tested
compound (dissolved in 10% dimethylsulfoxide saline
solution) was injected by ip (100 mg/kg). The rat was
then observed for 20 min before pilocarpine (350 mg/kg)
was administrated. The rat was further observed for
another 30 min and the absence of clonic spasms were
defined as protection.
Hydrochloride salt of 3-amino-2-(3-phenylpropyl)propio-
nic acid (8). 8 as white crystals was obtained, mp 166–
168 ꢂC; IR: 3009, 1605 cmꢀ1; H NMR: d 7.15–6.99 (m,
1
5H), 3.04–2.86 (m, 2H), 2.64–2.49 (m, 43H), 1.61–1.52
(m, 4H); 13C NMR: 176.12, 142.99, 129.44, 129.40,
126.94, 43.72, 41.35, 36.47, 30.46, 29.68; Anal. calcd for
C12H18NO2Cl: C, 59.14; H, 7.44; N, 5.75. Found: C,
58.98; H, 7.59; N, 5.69.
Hydrochloride salt of 3-amino-2-(2,2-diphenylethyl)pro-
pionic acid (13). White crystals of 13 were obtained, mp:
Acknowledgements
199–200 ꢂC; IR: 3011, 1603 cmꢀ1; H NMR: d 7.23–7.17
C.Y.K.T. acknowledges a Natural Sciences and Engi-
neering Research Council (NSERC) of Canada Scho-
larship. D.F.W. acknowledges operating grants from
the Medical Research Council of Canada, NSERC and
Neurochem Inc. We would also like to thank the ADD
of NIH for carrying out MES, scPTZ and rotorod neu-
rotoxicity testing.
1
(m, 8H), 7.11–7.07 (m, 2H), 4.04 (t, 1H, J=7.20 Hz),
3.05–2.98 (m, 2H), 2.48–2.40 (m, 2H), 2.15–2.14 (m, 1H);
13C NMR: 174.97, 144.21, 128.73, 128.06, 126.65, 48.86,
41.31, 40.62, 35.72; Anal. calcd for C17H20NO2Cl: C,
66.77; H, 6.59; N, 4.58. Found: C, 66.94; H, 6.99; N, 4.52.
Hydrochloride salt of 3-amino-2-(3,3-diphenylpropyl)pro-
pionic acid (14). White crystals (14) were obtained, mp
193–194 ꢂC; IR: 3001, 1593 cmꢀ1; 1H NMR: d 7.17–7.03
(m, 10H), 3.84 (t, 1H, J=7.80 Hz), 3.09–3.02 (m, 1H),
2.92–2.86 (m, 1H), 2.66–2.61 (m, 1H), 2.11–1.99 (m,
2H), 1.60–1.57 (m, 2H); 13C NMR: 176.06, 145.84,
129.49, 128.80, 127.26, 52.33, 43.75, 41.39, 33.61, 29.55;
Anal. calcd for C18H22NO2Cl: C, 67.60; H, 6.93; N,
4.38. Found: C, 67.26; H, 7.05; N, 4.12.
References and Notes
1. Bazil, C. W.; Pedley, T. A. Annu. Rev. Med. 1998, 49, 135.
2. Perucca, E. Br. J. Clin. Pharmacol. 1996, 42, 531.
3. Eadie, M. J. Drugs 1984, 27, 328.
4. Leppik, I. E. Epilepsia 1994, 35, S29.
5. Brodie, J. M. Lancet 1992, 339, 1397.
6. Wagner, M. L. Am. J. Hosp. Pharm. 1994, 51, 1657.
7. Davies-Jones, G. A. B. Anticonvulsants: Side Effects of
Drugs; 11th ed.; Elsevier Science: New York, 1988.
8. Del Rio, R. M.; Munoz, L.; DeFeudis, F. Exp. Brain Res.
1977, 28, 225.
9. Riddall, D.; Leach, M.; Davidson, J. J. Neurochem. 1976,
27, 835.
10. Martin, D. L.; Shain, W. J. Biol. Chem. 1979, 254, 7076.
11. Hosli, L. Neurosciene 1980, 5, 145.
12. Sandberg, M.; Jacobson, I. J. Neurochem. 1981, 1352.
13. Toggenberg, G.; Felix, D.; Cuenod, M.; Henke, H.
J. Neurochem. 1982, 39, 176.
14. DeFeudis, F.; Del Rio, R. M. Gen. Pharmacol. 1977, 8,
177.
Pharmacological methods
Maximal electroshock (MES), subcutaneous pentylene-
tetrazole (scPTZ), and rotorod neurotoxicity testing
were carried out by the Antiepileptic Drug Develop-
ment (ADD) Program, Epilepsy Branch, National
Institutes of Health, Bethesda, MD.32 All compounds
were evaluated in either male Carworth Farms #1 mice
or male Sprague–Dawley rats. Each compound was
administrated intraperitoneally (ip) at three different
dose levels 30, 100, and 300 mg/kg.
15. Lopez, A. Adv. Exp. Med. Biol. 1982, 139, 293.
16. Horikoshi, T.; Asanuma, A.; Yanagisawa, K.; Anzai, K.;
Goto, S. Mol. Brain Res. 1988, 4, 97.
MES seizures were induced 30 min after drug treatment
by the application of a 60 Hz current (50 mA in mice and

C. Y. K. Tan et al. / Bioorg. Med. Chem. 11 (2003) 113–121
121
17. Wu, F. S.; Gibbs, T. T.; Farb, D. Eur. J. Pharmacol. 1993,
246, 239.
25. N’Goka, V.; Schlewer, G.; Linget, J.; Chambon, J.; Wer-
muth, C. J. Med. Chem. 1991, 34, 2547.
18. Breckenridge, R. J.; Nicholson, S.; Nicol, A.; Suckling, C.
Biochem. Pharmacol. 1981, 30, 3045.
19. Moronio, F.; Mulas, A.; Moneti, G.; Corraetti, R.; Pepeu,
G. Ann. Super. Sanita 1982, 18, 49.
26. Leeson, P.; Iverson, L. J. Med. Chem. 1994, 37, 4053.
27. Katritzky, A. R.; Yannakopoulou, K. Synthesis 1989, 747.
28. Bohme, H.; Broese, R.; Eiden, F. Chem. Ber. 1959, 92,
1258.
20. Ogita, K.; Suzuki, T.; Yoneda, Y. Neuropharmacology
1989, 28, 1263.
29. Eadie, M. J. Clinical Use of Antiepileptic Drugs; Springer:
Berlin, 1985.
21. Pullan, L.; Powel, R. Neurosci. Lett. 1992, 148, 199.
22. Qing-Rong, L.; Lopez-Corcuera, B.; Nelson, H.; Mndi-
yan, S. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 12145.
23. Hitzemann, R.; Loh, H. J. Neurochem. 1978, 30, 471.
24. Mayor, F.; Valdvieso, F.; Gimenez, C.; Aragon, M. Prog.
Bioorg. Chem. Mol. Biol. Proc. Int. Symp. Front. Bioorg.
Chem. Mol. Bio 1984, 287.
30. Swinyard, E. A. J. Am. Pharm. Assoc. 1949, 38, 201.
31. Porter, R. J.; Cereghino, J. J.; Gladding, G. D.; Hessie,
B. J.; Kupferbug, H. J.; Scoville, B.; White, B. G. Cleve. Clin.
Q. 1984, 51, 293.
32. Stables, J. P.; Kupferberg, H. J. The NIHAnticonvulsant
Drug Development (ADD) Program: preclinical anticonvulsant
screening project; John Libbey, 1997.