Design, synthesis, and pharmacological characterization of novel, potent NMDA receptor antagonists

Article abstract of DOI:10.1021/jm049409f

The two diastereomeric pairs of acidic amino acids 5-(2-amino-2- carboxyethyl)-4,5-dihydroisoxazole-3-carboxylic acid (8A/8B) and 4-(2-amino-2-carboxyethyl)-5,5-dimethyl-4,5-dihydroisoxazole-3-carboxylic acid (10A/10B) were prepared via a strategy based on a 1,3-dipolar cycloaddition. The four amino acids were tested at ionotropic and metabotropic glutamate receptors. None of the compounds was active, neither as agonists nor as antagonists, at 1 mM on metabotropic receptors (mGluR1, -2, -4, and -5 expressed in CHO cell lines). Conversely, the pair of stereoisomers 8A/SB showed a remarkable affinity, antagonist potency, and selectivity for NMDA receptors, when tested on ionotropic glutamate receptors. The affinity of 8A proved to be 5 times higher than that of diastereomer 8B (K_i values 0.21 and 0.96 ìè, respectively). Furthermore, compounds 8A and 8B exhibited a noteworthy anticonvulsant activity in in vivo tests on DBA/2 mice. Derivative 10A was inactive at all ionotropic glutamate receptors, whereas its stereoisomer 10B displayed a seizable binding to both NMDA and AMPA receptors.

Full text of DOI:10.1021/jm049409f

6740
J. Med. Chem. 2004, 47, 6740-6748
Design, Synthesis, and Pharmacological Characterization of Novel, Potent
NMDA Receptor Antagonists
Paola Conti,† Marco De Amici,† Giovanni Grazioso,† Gabriella Roda,† Federico F. Barberis Negra,†
Birgitte Nielsen,‡ Tine B. Stensbøl,‡ Ulf Madsen,‡ Hans Bra¨uner-Osborne,‡ Karla Frydenvang,‡
Giovambattista De Sarro,§ Lucio Toma,| and Carlo De Micheli*,†
Istituto di Chimica Farmaceutica e Tossicologica, Universita` degli Studi di Milano, Viale Abruzzi 42, 20131 Milano, Italy;
Department of Medicinal Chemistry, The Danish University of Pharmaceutical Sciences, Universitetsparken,
DK-2100 Copenhagen, Denmark; Dipartimento di Medicina Sperimentale e Clinica, Universita` di Catanzaro, Italy; and
Dipartimento di Chimica Organica, Universita` degli Studi di Pavia, viale Taramelli, 12, 27100 Pavia, Italy
Received July 24, 2004
The two diastereomeric pairs of acidic amino acids 5-(2-amino-2-carboxyethyl)-4,5-dihydrois-
oxazole-3-carboxylic acid (8A/8B) and 4-(2-amino-2-carboxyethyl)-5,5-dimethyl-4,5-dihydrois-
oxazole-3-carboxylic acid (10A/10B) were prepared via a strategy based on a 1,3-dipolar
cycloaddition. The four amino acids were tested at ionotropic and metabotropic glutamate
receptors. None of the compounds was active, neither as agonists nor as antagonists, at 1 mM
on metabotropic receptors (mGluR1, -2, -4, and -5 expressed in CHO cell lines). Conversely,
the pair of stereoisomers 8A/8B showed a remarkable affinity, antagonist potency, and
selectivity for NMDA receptors, when tested on ionotropic glutamate receptors. The affinity of
8A proved to be 5 times higher than that of diastereomer 8B (Ki values 0.21 and 0.96 µM,
respectively). Furthermore, compounds 8A and 8B exhibited a noteworthy anticonvulsant
activity in in vivo tests on DBA/2 mice. Derivative 10A was inactive at all ionotropic glutamate
receptors, whereas its stereoisomer 10B displayed a seizable binding to both NMDA and AMPA
receptors.
Introduction
that these receptors are tetramers composed of at least
two NR1 subunits in combination with one or more NR2
subunits3 and, less commonly, an NR3 subunit.4,5
Glutamate (Glu) is the main excitatory neurotrans-
mitter in the mammalian central nervous system (CNS)
and mediates neurotransmission across most excitatory
synapses. Three classes of Glu-gated ion channels,
2-amino-3-(3-hydroxy-5-methyl-4-isoxazolyl)propionic acid
(AMPA), kainate (KA), and N-methyl-D-aspartic acid
(NMDA) receptors, transduce the postsynaptic signal,
i.e., Na+ and Ca2+ currents.1 Activation of the NMDA
receptors is rather complex, since it requires the binding
of both Glu, which plays the neurotransmitter role, and
glycine, which acts as a modulator.1 The correct func-
tioning of NMDA receptors, which are abundant and
ubiquitously distributed throughout the brain, is needed
to perform physiological effects of utmost importance,
i.e., phosphorylation of cAMP response element binding
protein, multiple gene activation, and long-term syn-
aptic plasticity.2 On the other hand, an overstimulation
of the NMDA receptors produces an acute neurodegen-
eration, termed excitotoxicity, which destroys the neu-
rons bearing them.2
From a therapeutic point of view, any neuronal loss
caused by Glu-induced excitotoxicity could be treated
by blocking NMDA receptors. The NMDA receptor
blocking approach could be applied to treat cerebral
ischemia, which occurs after stroke or brain trauma, as
well as neurodegenerative diseases, e.g. Parkinson’s and
Huntington’s diseases. It could also be used to treat
neurological disorders, such as epilepsy and neuropathic
pain, which are due to an overactivity of excitatory
pathways. Nevertheless, clinical trials with NMDA
receptor antagonists have so far failed due to the
occurrence of a number of adverse CNS effects, includ-
ing hallucinations, a centrally mediated increase in
blood pressure, and, at high doses, catatonia and
anesthesia.6-8 Current interest largely centers on the
development of potent and NR2B subtype-selective
antagonists,9,10 e.g. ifenprodil (1) (Figure 1), which have
demonstrated therapeutic potential for a number of
indications. In preclinical tests such compounds appear
to lack the side effects commonly associated with
nonselective NMDA antagonists.
In a recent paper11 we reported the synthesis and
pharmacological characterization of a new acidic bicyclic
amino acid (3aS*,5R*,6aS*)-5-amino-4,5,6,6a-tetrahy-
dro-3aH-cyclopenta[d]isoxazole-3,5-dicarboxylic acid (2)
(Figure 1). It displayed quite potent antagonism at the
NMDA receptors (Ki 0.37 µM) and, concominantly,
behaved as an antagonist at mGluR1 and mGluR5
(group I) and as an agonist at mGluR2. This pharma-
cological profile could suggest it as a promising lead of
The role played by NMDA receptors in excitotoxicity
has driven the search for antagonists as neuroprotective
agents. Their role in synaptic plasticity, on the other
hand, has inspired research into receptor potentiators
to treat cognitive dysfunctions.2
In the past decade, the molecular biology of the
NMDA receptor family was defined, and now we know
* To whom correspondence should be addressed. Phone: +39 02
50317538. Fax: +39 02 50317574. E-mail: carlo.demicheli@unimi.it.
† Universita` degli Studi di Milano.
‡ The Danish University of Pharmaceutical Sciences.
§ Universita` di Catanzaro.
| Universita` degli Studi di Pavia.
10.1021/jm049409f CCC: $27.50 © 2004 American Chemical Society
Published on Web 12/02/2004

Novel, Potent NMDA Receptor Antagonists
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 27 6741
Scheme 1
Scheme 2
alkenes is highly regioselective to yield 3,5-disubstituted
∆2-isoxazolines,14 we initially planned the synthesis of
derivatives 8. Subsequently, due to the difficulties
related to the preparation of the two diastereomers 9,
we designed their analogues 10, since the 1,3-dipolar
cycloaddition of a nitrile oxide to a trisubstituted alkene
produces 3,4,5,5-tetrasubstituted ∆2-isoxazolines, ex-
clusively.14
Figure 1. Structure of model and tested compounds.
neuroprotective agents. In in vivo tests on DBA/2 mice,
compound 2 showed a noteworthy anticonvulsant activ-
ity.11
At first sight the structure of derivative 2 can be
considered a locked conformation of 2-aminoadipic acid
(3A) (Figure 1), which is a very weak NMDA receptor
antagonist (Ki 13 µM).12 Subsequently, it was discovered
that a substantial enhancement in the NMDA potency
could be achieved by exchanging the distal carboxylate
group of 3A with the phosphonate moiety to yield
2-amino-5-phophonopentanoic acid (AP5, 3B) (Ki 0.29
µM).13 Since AP5 is devoid of any significant activity
following systemic administration in animals, modifica-
tions of its structure, e.g., a reduction in the conforma-
tional mobility, were thoroughly investigated. Interest-
ing results were achieved by incorporating the required
amino acid substructure into a cyclic framework.
Whereas derivative 4 was almost inactive (Ki >10 µM),13
its phosphono analogue 5, tagged as CGS 19755, turned
out to be one of the most potent NMDA antagonist (Ki
0.095 µM)13 so far discovered. CGS19755 is nowadays
commonly used as a radioligand in NMDA assays
(Figure 1). It is worth pointing out that the structure
of 2 can also be related to that of derivative 6 or its
phosphono analogue 7 (Figure 1), which are inactive or
marginally active.
Chemistry
The key step in the synthesis of the target compounds
8A and 8B is represented by the 1,3-dipolar cycloaddi-
tion of ethoxycarbonylformonitrile oxide, generated in
situ by treatment of ethyl 2-chloro-2-(hydroxyimino)-
acetate15 with a base, to the suitably protected racemic
allylglycine 11 (Scheme 1). The pericyclic reaction
produced a racemic mixture of the two 3,5-disubstituted
diastereoisomers 12 and 13 in a 1:1.6 ratio. In analogy,
the synthesis of target compounds 10A and 10B was
carried out by exploiting the cycloaddition reaction of
racemic dipolarophile 14 with the same nitrile oxide,
which gave a racemic mixture of cycloadducts 15 and
16 in a 1:1 ratio (Scheme 2). Since 14 is a sluggish
dipolarophile, the reaction was carried out in a micro-
wave oven.
Dipolarophile (()-1116 was prepared from commer-
cially available (()-allylglycine by standard transforma-
tions (see the Experimental Section), whereas alkene
14 was synthesized from N-(diphenylmethylene)glycine
ethyl ester following a procedure reported in the litera-
ture.17 The pairs of racemic stereoisomers 12/13 and 15/
16 were separated by silica gel column chromatography
and then transformed into final amino acids 8A, 8B,
10A, and 10B by standard methodologies.
To simplify the structure of 2 and to establish
similarities with that of derivatives 4/5 or their homo-
logues 6/7 (Figure 1), we designed regioisomers 9 and
8, respectively. Since it is well-known that the 1,3-
dipolar cycloaddition of nitrile oxides to monosubstituted
The relative configuration of the stereogenic centers
of derivatives 8A and 8B (Figure 1) was assigned

6742 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 27
Conti et al.
Figure 2. The most populated conformation for compounds
8A and 8B.
Table 1. Experimental 1H NMR Vicinal Coupling Constants
for Compounds 8A and 8B in Comparison with Their
Calculated Values
Figure 3. Perspective view of the molecular structure with
numbering for compound 13. Displacement ellipsoids are
drawn at 50% of probability level while the size of the hydrogen
atom is arbitrary.
JR,âa
JR,âb
Jâa,5
Jâb,5
J5,4a
J5,4b
8A (exp)
8A (calcd)
8B (exp)
8B (calcd)
8.1
8.3
6.7
9.0
5.3
4.7
4.4
3.4
10.2
9.0
3.4
4.4
9.3
9.4
6.6
4.7
7.3
5.0
10.5
9.6
10.7
9.4
3.9
Scheme 3
4.3
1
through an analysis of their H NMR spectra in water
combined with a theoretical investigation of their con-
formational properties carried out at the B3LYP/6-31G*
level.18,19 The possible rotamers, deriving from a rotation
around the CR-Câ and Câ-C5 single bonds of both 8A
and 8B, were taken into account and generated nine
geometries that were then fully optimized. The relative
energies of the various conformations of 8A and 8B are
strongly influenced by the solvent; as a matter of fact,
conformations presenting a hydrogen bond between the
amino group and the oxygen atom of the isoxazoline ring
show an unrealistic stabilization when calculated in the
gas phase. Hence, all energies were recalculated in a
polarizable conductor-like continuum solvation model
(C-PCM)20 to obtain values conceivable for water solu-
tions. For both diastereoisomers, the preferred confor-
mation was identified as 8A-1 and 8B-1, respectively
(Figure 2), though several other conformers contribute
with a nonnegligible percentage to the overall popula-
tions. For each conformer the 1H NMR vicinal coupling
constants were calculated by the Haasnoot et al. equa-
tion21 and were weight averaged on the basis of the
population percentages. The obtained values are re-
ported in Table 1 together with the experimental
coupling constants. As shown in Table 1, coupling
constants Jâa,5 and Jâb,5 are highly diagnostic, since their
values are quite different in the two diastereomers. The
close agreement between the experimental coupling
constants and the calculated values ensures that 5R*,RS*
is the relative configuration of 8A. Similarly, the agree-
ment between the experimental values of 8B and the
calculated ones allows the assignment of the 5R*,RR*
configuration to 8B. The relative stereochemistry of 8B
was then confirmed by an X-ray analysis carried out
on cycloadduct 13 (Figure 3).22
to overcome this difficulty could be envisaged in a
restriction of the conformational flexibility of the struc-
tures. Thus, racemic cycloadducts 15 and 16, precursors
of 10A and 10B, respectively, were transformed into the
corresponding bicyclic lactams 17 and 18 (Scheme 3),
which are characterized by a much more rigid structure.
1
The H NMR spectra of derivatives 17 and 18 showed
resolved signals and, as expected, significant differences
in the coupling pattern of the protons present in the
lactam ring, i.e., HR, Hâa, Hâb, and H4 (Figure 4). As
shown in Table 2, the Jâb,R coupling constant is quite
different in the two distereoisomers and becomes diag-
nostic in the structure assignment.
The same procedure applied to derivatives 10A and
10B failed to produce conclusive evidence on the relative
configuration of their stereogenic centers. In this case,
the values of the hydrogen vicinal coupling constants
could not be evaluated, since most of the signals
appeared as unresolved multiplets (see the Experimen-
tal Section). Thus, the relative stereochemistry of the
two diastereoisomers 10A and 10B could not be as-
signed on the basis of their 1H NMR spectra. A strategy
Figure 4. The most populated conformation for bicyclic
derivatives 17 and 18.

Novel, Potent NMDA Receptor Antagonists
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 27 6743
Table 2. Experimental 1H NMR Vicinal Coupling Constants
for Compounds 17 and 18 in Comparison with Their Calculated
Values
NMDA receptors (Ki 0.21 µM) that is only slightly lower
than that displayed by the phosphono amino acid CGS
19755 (5) (Ki 0.095 µM), one of the most potent NMDA
antagonists,13 and significantly higher than that of
derivative 7 (Ki 6.6 µM), the phosphono derivative with
the same bond connection (Figure 1). The above-
reported pharmacological data were confirmed in the
rat cortical slice model,27 where 8A antagonized the
responses induced by 10 µM NMDA (IC50 14.2 µM) and
left the responses induced by either AMPA (5 µM) or
KA (5 µM) unaffected. Generally, the activities observed
in the rat cortical slice model are significantly lower
than the affinities observed in binding assays.27 Taking
into account the excellent antagonist activity of 8A at
NMDA receptors, we decided to test compounds 8A and
8B as anticonvulsant agents. Their ability to block
seizures induced by audiogenic stimuli was evaluated
in vivo in DBA/2 mice after ip administration; their
potency, expressed as ED50 value (µM/Kg), is reported
in Table 4 and compared with that of compound 2,11
CPPene, a highly potent NMDA antagonist, and
phenytoin, a typical anticonvulsant agent. The anti-
convulsant potency of test compounds 8A and 8B is only
5-10 times lower than that of CPPene, which is
characterized by a very high NMDA receptor affinity
(Ki 0.04 µM).13 Compound 8B at all tested doses did not
show either a reduction in body temperature or behavior
changes, whereas 8A at the highest dose (100 mmol/
kg, ip) induced sedation, reduction of locomotor activity,
and a slight ataxia in three out of 10 mice. It is worth
noting that acidic amino acids 8A and 8B are active
after ip administration, thus indicating their ability to
pass the blood-brain barrier.
J4,âa
J4,âb
Jâa,R
Jâb,R
17 (exp)
17 (calcd)
18 (exp)
18 (calcd)
4.4
3.6
5.1
3.7
12.8
12.4
13.5
12.3
3.7
3.6
1.1
1.9
12.8
11.8
5.1
4.8
The conformational analysis of 17 and 18 located only
two minima for each compound, confirming that the
bicyclic skeleton is quite rigid. The only degree of
conformational freedom resides in the ethoxycarbonyl
moiety, which, however, has only a slight influence on
the geometry of the bicyclic skeletons. The preferred
conformations 17A and 18A, depicted in Figure 4,
account for 86% and 81%, respectively, of their overall
populations. The values of the coupling constants,
calculated according to the equation of Haasnoot et al.,21
are reported in Table 2 together with the corresponding
experimental ones. The close agreement between experi-
mental and theoretical data allows the attribution of
the 4R*,RR* configuration to the stereogenic centers of
17 and, consequently, to the monocyclic compounds 15
and 10A. Thus, the configuration 4R*,RS* is assigned
to the diastereoisomers 18, 16, and 10B.
Results and Discussion
The two pairs of acidic amino acids 8A/8B and 10A/
10B were assayed in vitro by means of receptor binding
techniques, second messenger assays, and electrophysi-
ological studies. The mGluR activity of the new com-
pounds was evaluated at rat mGluR1 and mGluR5
(group I), mGluR2 (representative of group II), and
mGluR4 (representative of group III), all expressed in
CHO cells.23 The compounds were tested at 1 mM for
both agonist and antagonistic activity and showed no
activity at any of the mGluRs.
The affinity for NMDA, AMPA, and KA receptors was
determined by using the radioligands [3H]CGP 39653,
[3H]AMPA, and [3H]KA, respectively.24-26 As reported
in Table 3, the pairs of racemic stereoisomers 8A/8B
and 10A/10B showed no significant affinity for AMPA
and KA receptors (IC50 > 100 µM), except for 10B, which
possesses a weak affinity for AMPA receptors (IC50 12.0
µM). In contrast to this, 8A and 8B showed a remark-
able affinity for NMDA receptors. Worth noting is that
compound 8A displays a highly selective affinity for the
An analysis of the biological data gathered in Table
3 puts in evidence that on passing from bicyclic deriva-
tive 2 to the monocyclic analogues, either 8A/8B or 10A/
10B, we lose completely the activity at all the mGluR
subtypes, whereas in some derivatives, i.e., 8A/8B, the
affinity at NMDA receptors is maintained or even
increased. If we take into account the bond connection
among the pharmacophoric groups of compounds 8A/
8B, we can find an analogy with derivative 6 or with
the phosphono analogue 7 (Figure 1), which are inactive
or marginally active.
Analysis of the structure-activity relationship can be
carried out by considering the model for the antagonist-
preferring state of the NMDA receptors reported by
Table 3. Rat Cortical Membrane Receptor Binding and in Vitro Electrophysiological Data
IC50 (µM)
[3H]CGP 39653
compda
Ki (µM)
[3H]AMPA
[3H]KA
electrophysiology
8A
0.21 [0.18; 0.23]b
0.96 [0.88; 1.1]b
>100
>100
>100
>100
>100
>100
>100
14.2 [12.9; 15.7]b,c
30.5 [24.9; 37.4]b,c
ndi
8B
>100
>100
10A
10B
79 [73; 86]b
0.37 [0.34; 0.40]b
13d,e
12.0 [11.7; 12.4]a
>100
nd
(()-2
(-)-3A
(()-4
(()-5
(()-6
(()-7
(-)-CPPene
13 [12; 15]b,c
>10f,g
0.095 ( 0.028f,h
>10f,h
6.6 ( 1.3f,h
0.04 ( 0.01f,h
a Derivatives 8A/8B and 10A/10B have been assayed as racemates. b Values are expressed as the antilog to the log mean of at least
three individual experiments. The number in brackets [min; max] indicates ( SEM according to a logaritmic distribution. c Antagonism
of NMDA (10 µM) induced responses. d Data from the literature.12 e the radioligand used is [3H]-D-APV. f Data from the literature.13
g The radioligand used is [3H]CGS19755. h The radioligand used is [3H]CPP. i nd, not determined.

6744 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 27
Conti et al.
Table 4. ED50 Values of Test Compounds against Audiogenic
As far as compounds 8A and 8B are concerned, the
higher mobility of their unconstrained side chain gener-
ated several low-energy conformations; among them,
five conformations that gave significant contributions
to the overall population are reported in Table 5.
Examination of the distances among the ionizable
groups of each conformer (Table 5) shows that none of
them matches the optimal values found for compound
2. The distances appear somewhat longer than the
values of Hutchison’s model for derivatives structurally
related to 5. Since this model accounts for the biological
activity of a series of derivatives related to 5 as well as
a number of homologues related to 2-amino-7-phos-
phonoheptanoic acid, a distance of 7.9 Å between the R
and ω carboxylate (phosphonate) groups appears to be
optimal for this set of compounds. The values found for
8A and 8B are between the values proposed for the two
models, and as a consequence, the range 5.5-7.9 Å may
be considered compatible with a productive interaction
of the ligand pharmacophoric groups with the comple-
mentary binding sites of the NMDA receptors.
Seizures Induced in DBA/2 Micea
ED50 values
(95% confidence limits)
compd
clonus
tonus
8A
17.9 (11.6-27.6)
29.9 (21.4-41.8)
28 (19-41)
13.9 (8.5-22.9)
19.7 (12.5-31.3)
22 (14-35)
1.3 (0.7-2.6)
7.3 (5.8-9.1)
8B
(()-2
(()-CPPene
phenytoin
3.5 (2.1-5.6)
9.1 (5.6-14.9)
a All data are expressed as µmol/kg for ip administration.
Similar computational studies carried out on com-
pounds 10A and 10B also showed the presence of
several minimum energy conformations (Table 5). How-
ever, most of them are characterized by too short
distances among the ionic groups. The tendency to
assume folded conformations could be the reason for the
inactivity of both isomers at the NMDA receptors.
In summary, we herein report the synthesis of four
new acidic amino acids, and two of them (8A and 8B)
behaved in vitro as potent antagonists at NMDA recep-
tors and in vivo, after ip administration in DBA/2 mice,
as effective anticonvulsants. Their biological activity
compares very well with that of phosphono derivative
5 and can be accounted for on the basis of the distances
among the three pharmacophoric groups according to
the model proposed by Hutchison for NMDA antago-
nists.28 Since the bioisosteric replacement of the ω-car-
boxylate group with the ω-phosphonate moiety strongly
increases the potency of an NMDA antagonist,13 the
synthesis of the phosphono analogues of 8A/8B will be
undertaken and the pharmacological results will further
support the analogy with derivative 5.
Figure 5. Hutchison’s model for the antagonist-preferring
state of the NMDA receptors.28
Table 5. Conformational Studies of Compounds 2, 8A/8B, and
10A/10B
∆Erel
(kcal/mol)
dN/Cω
(Å)
dCR/Cω
(Å)
Hutchison’s model
2-1
5.55
5.58
5.58
6.21
6.70
6.83
7.34
6.71
6.83
6.18
5.83
6.70
6.65
3.72
3.93
4.70
5.16
3.42
3.75
3.86
5.33
4.85
3.30
5.54
5.63
4.19
7.28
6.83
6.94
6.19
6.10
7.10
6.84
7.33
6.62
5.68
3.16
5.35
5.05
5.26
5.18
4.82
5.06
5.04
3.36
5.18
0.00
1.71
0.00
0.19
0.28
1.02
1.51
0.00
0.81
1.00
1.28
1.97
0.00
0.19
0.33
0.66
0.69
0.00
1.01
1.77
2.01
2.12
2-2
8A-1
8A-2
8A-3
8A-4
8A-5
8B-1
8B-2
8B-3
8B-4
8B-5
10A-1
10A-2
10A-3
10A-4
10A-5
10B-1
10B-2
10B-3
10B-4
10B-5
Experimental Section
Materials and Methods. All reagents were purchased from
Aldrich. Ethyl 2-chloro-2-(hydroxyimino)acetate was prepared
according to a literature procedure.15
1H and 13C NMR spectra were recorded on a Varian Mercury
300 (300 MHz) spectrometer in CDCl3 or D2O solution at 20
°C. Chemical shifts (δ) are expressed in ppm and coupling
constants (J) in hertz. Potentiometric titrations were per-
formed with the GLpKa apparatus (Sirius Analytical Instru-
ments Ltd, Forrest Row, East Sussex, UK) equipped with a
pH electrode, a temperature probe, an overhead stirrer, and
precision dispensers for automated distribution of the diluent
(0.15 M KCl in water) and titrants (0.5 M HCl, 0.5 M KOH).
Four separate 15 mL aqueous solutions were added to weighted
samples (1-10 mg) and acidified to pH 1.8 with HCl. The
solutions were then titrated with standardized KOH to pH 8.
The titrations were conducted under nitrogen at 25.0 ( 0.1
°C. The initial estimates of the pKa values were obtained by
difference plots (Bjerrum plots).29 These values were then
refined by a weighted nonlinear least-squares procedure.
Microwave reactions were performed with a Ethos Microsynth
Milestone apparatus. TLC analyses were performed on com-
mercial silica gel 60 F254 aluminum sheets; spots were further
Hutchison et al.28 (Figure 5). According to this model,
the optimal distances amine/ω-carboxylate and R-car-
boxylate/ω-carboxylate are almost identical, i.e., 5.55
and 5.54 Å, respectively, and correspond to the values
found in the active conformation of compound 5. On this
ground, we performed a conformational analysis of
compound 2 and found that it is characterized by only
two populated conformations. For each conformer we
measured the distances among the ionizable groups; as
shown in Table 5, the values of the most populated
conformation (2-1) are in close agreement with those
of Hutchison’s model. This result might explain the high
binding affinity of compound 2 for the NMDA receptors.

Novel, Potent NMDA Receptor Antagonists
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 27 6745
evidenced by spraying with a dilute alkaline potassium per-
manganate solution or with ninhydrin. Melting points were
determined on a Bu¨chi B-540 apparatus and are uncorrected.
Microanalyses of new compounds agreed with theoretical
values (0.3%.
under vacuum, and the residue was washed with MeOH and
Et2O to yield 37% (104 mg).
Amino Acid 8A. White prisms. Rf: 0.37 (n-butanol/water/
acetic acid 4:2:1). Mp: 161-169 °C (dec). pKa: < 2.0 (R-COOH),
2.6 (ω-COOH), 9.1 (NH3+). 1H NMR (D2O): 2.02 (ddd, J ) 8.1,
10.3, 15.1 Hz, 1H), 2.14 (ddd, J ) 3.3, 5.3, 15.1 Hz, 1H), 2.83
(dd, J ) 6.6, 18.1 Hz, 1H), 3.29 (dd, J ) 10.6, 18.1 Hz, 1H),
3.86 (dd, J ) 5.3, 8.1 Hz, 1H), 4.94 (dddd, J ) 3.3, 6.6, 10.3,
10.6 Hz, 1H). 13C NMR (D2O): 35.66, 40.16, 52.25, 80.98,
155.98, 164.83, 172.55. Anal. (C7H10N2O5): C, H, N.
(5R*,2′R*)-5-(2-Amino-2-carboxyethyl)-4,5-dihydroisox-
azole-3-carboxylic Acid (8B). Step A. Derivative 13 (800
mg, 2.32 mmol) was dissolved in MeOH (10 mL) and treated
with 1 N NaOH (10 mL) at room temperature overnight. The
disappearance of the starting material was monitored by TLC
(CHCl3/MeOH 95:5 + 50 µL AcOH). The aqueous layer was
washed with CH2Cl2, made acidic with 2 N HCl, and extracted
with AcOEt. The organic phase was dried over anhydrous Na2-
SO4, and after evaporation of the solvent, a white powder was
obtained (408 mg, 1.35 mmol, yield 58%).
Step B. The crude material, obtained from the previous
transformation (408 mg, 1.35 mmol), was treated with a 30%
CH2Cl2 solution of trifluoroacetic acid (1.00 mL,13.5 mmol) at
0 °C. The solution was stirred at room temperature for 3 h
until the disappearance of the starting material (TLC: n-
butanol/water/acetic acid 4:2:1). The volatiles were removed
under vacuum, and the residue was washed with MeOH and
Et2O to yield 41% (112 mg).
Amino Acid 8B. White prisms. Rf: 0.27 (n-butanol/water/
acetic acid 4:2:1). Mp: 175-185 °C (dec). pKa: < 2.0 (R-COOH),
2.5 (ω-COOH), 9.0 (NH3+). 1H NMR (D2O): 2.08 (ddd, J ) 3.9,
6.5, 13.7 Hz, 1H), 2.15 (ddd, J ) 4.4, 9.3, 13.7 Hz, 1H), 2.84
(dd, J ) 7.3, 18.0 Hz, 1H), 3.30 (dd, J ) 10.7, 18.0 Hz, 1H),
3.92 (dd, J ) 4.4, 6.5 Hz, 1H), 4.84 (dddd, J ) 3.9, 7.3, 9.3,
10.7 Hz, 1H). 13C NMR (D2O): 34.90, 49.91, 51.82, 80.04,
155.92, 164.71, 172.45. Anal. (C7H10N2O5): C, H, N.
Methyl 2-tert-Butoxycarbonylaminopent-4-enoate 11.16
Step A. To a stirred suspension of (()-allylglycine (2 g, 13.3
mmol) in CH2Cl2 (20 mL) was added triethylamine (3 mL, 21.5
mmol) and a solution of di-tert-butyl dicarbonate (3.77 g, 17.3
mmol) in CH2Cl2 (10 mL) cooled to 0 °C. The reaction mixture
was stirred at room temperature overnight and the progress
of the reaction was monitored by TLC (n-butanol/water/acetic
acid 4:2:1). The mixture was washed with 2 N HCl (2 × 10
mL) and dried over anhydrous Na2SO4 and the solvent
evaporated. The yield was quantitative
Step B. The crude material obtained from the previous
transformation (13.3 mmol) was dissolved in acetone (20 mL)
and then treated with solid K2CO3 (3.7 g, 26.6 mmol) and
excess methyl iodide (1.7 mL, 26.6 mmol). The mixture was
refluxed for 4 h, and the progress of the reaction was monitored
by TLC (CHCl3/MeOH 95:5 + 50µL acetic acid). The solvent
was evaporated and the residue was taken up with AcOEt (25
mL) and washed with a saturated aqueous solution of NaH-
CO3; the organic layer was dried over anhydrous Na2SO4 and
the solvent removed under vacuum. The residue was chro-
matographed on silica gel (eluant: petroleum ether/AcOEt 7:3)
to give 2.80 g (overall yield: 92%) of 11 as a colorless oil. Rf:
1
0.49 (petroleum ether/AcOEt 8:2). H NMR (CDCl3): 1.42 (s,
9H), 2.50 (m, 2H), 3.72 (s, 3H), 4.36 (dd, J ) 6.6, 13.9 Hz, 1H),
5.02 (bd, J ) 7.5 Hz, 1H), 5.14 (m, 2H), 5.68 (m, 1H). Anal.
(C11H19NO4): C, H, N.
Ethyl (5R*,2′S*)-5-(2-tert-Butoxycarbonylamino-2-meth-
oxycarbonylethyl)-4,5-dihydroisoxazole-3-carboxylate (12)
and Ethyl (5R*,2′R*)-5-(2-tert-Butoxycarbonylamino-2-
methoxycarbonylethyl)-4,5-dihydroisoxazole-3-carboxy-
late (13). To a solution of 11 (2.8 g, 12.30 mmol) in AcOEt (40
mL) was added ethyl 2-chloro-2-(hydroxyimino)acetate15 (2.79
g, 18.45 mmol) and NaHCO3 (5 g). The mixture was vigorously
stirred for 3 days at room temperature; the progress of the
reaction was monitored by TLC (petroleum ether/AcOEt 7:3).
Water was added to the reaction mixture and the organic layer
was separated and dried over anhydrous Na2SO4. The crude
material, obtained after evaporation of the solvent, was
chromatographed on silica gel (petroleum ether/AcOEt 4:1) to
give 1.58 g of 12 and 2.52 g of 13. The overall yield was 97%.
Compound 12. Colorless oil. Rf: 0.23 (petroleum ether/
Ethyl
2-tert-Butoxycarbonylamino-5-methylhex-4-
enoate (14). Step A. Under a nitrogen atmosphere at -78
°C, DMPU (3.25 mL, 26.84 mmol) and N-(diphenylmethylene)-
glycine ethyl ester (7.17 g, 26.84 mmol) were added to a
solution of 2.5 M BuLi in hexane (10.74 mL, 26.84 mmol) and
diisopropylamine (3.76 mL, 26.84 mmol) in THF (10 mL). The
mixture was stirred for 10 min and then 1-bromo-3-methyl-
2-butene (3.09 mL, 26.84 mmol) was added. The progress of
the reaction was monitored by TLC (petroleum ether/AcOEt
9:1). Stirring was continued for 2 h more at -78 °C and then
a saturated solution of ammonium chloride was added and the
mixture was allowed to reach room temperature. THF was
evaporated and the aqueous layer extracted with Et2O. The
organic phase was dried over anhydrous Na2SO4 and the
solvent removed under vacuum. Hydrolysis of the alkylated
imine with 1 N aqueous HCl/THF afforded the crude amine
(3.67 g) in 80% yield.
1
AcOEt 7:3). H NMR (CDCl3): 1.36 (t, J ) 7.2 Hz, 3H), 1.43
(s, 9H), 2.14 (m, 2H), 2.91 (dd, J ) 10.5, 17.8 Hz, 1H), 3.37
(dd, J ) 7.2, 17.8 Hz, 1H), 3.78 (s, 3H), 4.32 (q, J ) 7.2 Hz,
2H), 4.34 (m, 1H), 4.92 (m, 1H), 5.34 (bd, J ) 6.7 Hz, 1H).
Anal. (C15H24N2O7): C, H, N.
Compound 13. Colorless needles from diisopropyl ether.
1
Mp: 94.5 °C. Rf: 0.21 (petroleum ether/AcOEt 7:3). H NMR
(CDCl3): 1.36 (t, J ) 7.0 Hz, 3H), 1.43 (s, 9H), 2.00 (m, 1H),
2.22 (ddd, J ) 7.0, 8.0, 12.3 Hz, 1H), 2.88 (dd, J ) 7.0, 16.1
Hz, 1H), 3.33 (dd, J ) 9.1, 16.1 Hz, 1H), 3.75 (s, 3H), 4.34 (q,
J ) 7.0 Hz, 2H), 4.46 (m, 1H), 4.91 (m, 1H), 5.3 (bd, J ) 7.5
Hz, 1H). Anal. (C15H24N2O7): C, H, N.
Step B. At 0 °C a solution of di-tert-butyl dicarbonate (5.71
g, 26.16 mmol) in CH2Cl2 (10 mL) was added to a solution of
the crude amine (3.67 g, 21.47 mmol) and triethylamine (3.95
mL, 28.36 mmol) in CH2Cl2 (20 mL). The mixture was stirred
at room temperature overnight and the progress of the reaction
was monitored by TLC (AcOEt). The mixture was washed with
2 N HCl, the organic phase was dried over anhydrous Na2-
SO4, and the crude material, obtained after the evaporation
of the solvent, was chromatographed on silica gel (petroleum
ether/AcOEt 95:5) to give 5.24 g (yield 90%) of 14 as a colorless
(5R*,2′S*)-5-(2-Amino-2-carboxyethyl)-4,5-dihydroisox-
azole-3-carboxylic Acid (8A). Step A. Derivative 12 (800
mg, 2.32 mmol) was dissolved in MeOH (10 mL) and treated
with 1 N NaOH (10 mL) at room temperature overnight. The
disappearance of the starting material was monitored by TLC
(CHCl3/MeOH 95:5 + 50 µL acetic acid). The aqueous layer
was washed with CH2Cl2, made acidic with 2 N HCl, and
extracted with EtOAc. The organic phase was dried over
anhydrous Na2SO4 and after evaporation of the solvent a white
powder was obtained (420 mg, 1.39 mmol, yield 60%).
Step B. The crude material, obtained from the previous
transformation (420 mg, 1.39 mmol), was treated with a 30%
CH2Cl2 solution of trifluoroacetic acid (1.07 mL, 13.9 mmol)
at 0 °C. The solution was stirred at room temperature for 3 h
until the disappearance of the starting material (TLC: n-
butanol/water/acetic acid 4:2:1). The volatiles were removed
1
oil. Rf: 0.52 (petroleum ether/AcOEt 9:1). H NMR (CDCl3):
1.22 (t, J ) 7.3 Hz, 3H), 1.42 (s, 9H), 1.60 (s, 3H), 1.64 (s, 3H),
2.50 (m, 2H), 4.08 (q, J ) 7.3 Hz, 2H), 4.30 (m, 1H), 5.02 (m,
1H), 5.02 (m, 1H). Anal. (C14H25NO4): C, H, N.
Ethyl
(4R*,2′R*)-4-(2-tert-Butoxycarbonylamino-2-
ethoxycarbonylethyl)-5,5-dimethyl-4,5-dihydroisoxazole-
3-carboxylate (15) and Ethyl (4R*,2′S*)-4-(2-tert-Butoxy-
carbonylamino-2-ethoxycarbonylethyl)-5,5-dimethyl-4,5-
dihydroisoxazole-3-carboxylate (16). To a solution of 14
(5.0 g, 18.45 mmol) in AcOEt (40 mL) was added ethyl 2-chloro-
2-(hydroxyimino)acetate (3.35 g, 22.14 mmol) and NaHCO3 (5

6746 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 27
Conti et al.
g). The mixture was irradiated in a microwave oven at 50 W
for 20 min in a sealed vessel. The same amount of ethyl
2-chloro-2-(hydroxyimino)acetate was then added and the
reaction was heated again for 20 min. The progress of the
reaction was monitored by TLC (petroleum ether/AcOEt 9:1).
Water was added to the reaction mixture and the organic layer
was separated and dried over anhydrous Na2SO4. The crude
material, obtained after evaporation of the solvent, was
chromatographed on silica gel (petroleum ether/AcOEt 9:1) to
give 1.85 g of 15 and 1.80 g of 16. The overall yield was 52%.
Ethyl (3aR*,5R*)-3,3-Dimethyl-7-oxo-3,3a,4,5,6,7-hexahy-
droisoxazolo[3,4-c]pyridine-5-carboxylate (17) and Ethyl
(3aR*,5S*)-3,3-Dimethyl-7-oxo-3,3a,4,5,6,7-hexahydroisox-
azolo[3,4-c]pyridine-5-carboxylate (18). Derivative 15 (100
mg, 0.258 mmol) was treated with a 30% CH2Cl2 solution of
trifluoroacetic acid (168 µL, 2.58 mmol) at 0 °C. The solution
was stirred at room temperature for 3 h until the disappear-
ance of the starting material (TLC: AcOEt). The volatiles were
removed under vacuum, and the residue was dissolved in a
saturated solution of NaHCO3. The aqueous layer was ex-
tracted with AcOEt, the organic phase was dried over anhy-
drous Na2SO4, and the solvent evaporated to give lactam 17
in 90% yield (56 mg, 0.232 mmol).
Cycloadduct 15. Yellow oil. Rf: 0.43 (petroleum ether/
1
AcOEt 9:1). H NMR (CDCl3): 1.27 (t, J ) 7.3 Hz, 3H), 1.37
(t, J ) 7.3 Hz, 3H), 1.38 (s, 3H), 1.45 (s, 9H), 1.57 (s, 3H), 1.84
(m, 1H), 2.09 (m, 1H), 3.15 (dd, J ) 2.2, 10.6 Hz, 1H), 4.18 (q,
J ) 7.3 Hz, 2H), 4.32 (q, J ) 7.3 Hz, 2H), 4.33 (m, 1H), 5.15
(bd, J ) 9.2 Hz, 1H). Anal. (C18H30N2O7): C, H, N.
The same procedure applied to derivative 16 gave lactam
18 in comparable yield.
Lactam 17. Colorless oil. Rf: 0.50 (AcOEt). 1H NMR
(CDCl3): 1.22 (s, 3H), 1.30 (t, J ) 7.3 Hz, 3H), 1.57 (s, 3H),
1.83 (m, 1H), 2.45 (ddd, J ) 12.8, 4.4, 3.7 Hz, 1H), 3.22 (dd, J
) 12.8, 4.4 Hz, 1H), 4.23 (dd, J ) 12.8, 3.7 Hz, 1H), 4.28 (q, J
) 6.9 Hz, 2H), 6.54 (bs, 1H). Anal. (C11H16N2O4): C, H, N.
Lactam 18: Colorless oil. Rf: 0.43 (AcOEt). 1H NMR
(CDCl3): 1.22 (s, 3H), 1.30 (t, J ) 7.3 Hz, 3H), 1.57 (s, 3H),
2.08 (ddd J ) 13.5, 13.5, 5.1 Hz, 1H), 2.38 (ddd, J ) 13.5, 5.1,
1.1 Hz, 1H), 3.12 (dd, J ) 13.5, 5.1 Hz, 1H), 4.24 (q, J ) 7.0
Hz, 2H), 4.23 (m, 1H), 7.40 (bs, 1H). Anal. (C11H16N2O4): C,
H, N.
Cycloadduct 16. Yellow oil. Rf: 0.33 (petroleum ether/
1
AcOEt 9:1); H NMR (CDCl3): 1.30 (t, J ) 7.3 Hz, 3H), 1.37
(t, J ) 7.3 Hz, 3H), 1.39 (s, 3H), 1.44 (s, 9H), 1.47 (s, 3H), 2.05
(m, 2H), 3.22 (m, 1H), 4.23 (q, J ) 7.3 Hz, 2H), 4.34 (q, J )
7.3 Hz, 2H), 4.35 (m, 1H), 5.22 (bd, J ) 7.7 Hz, 1H). Anal.
(C18H30N2O7): C, H, N.
(4R*,2′R*)-4-(2-Amino-2-carboxyethyl)-5,5-dimethyl-
4,5-dihydroisoxazole-3-carboxylic Acid (10A). Step A.
Derivative 15 (500 mg, 1.29 mmol) was dissolved in EtOH (10
mL) and treated with 1 N NaOH (4 mL) at room temperature
overnight. The disappearance of the starting material was
monitored by TLC (CHCl3/MeOH 8:2 + 100 µL acetic acid).
The aqueous layer was washed with CH2Cl2, made acidic with
2 N HCl, and extracted with AcOEt. The organic phase was
dried over anhydrous Na2SO4 and, after evaporation of the
solvent, a white powder was obtained (383 mg, 1.16 mmol,
yield 90%).
Step B. The crude material, obtained from the previous
transformation (383 mg, 1.16 mmol), was treated with a 30%
CH2Cl2 solution of trifluoroacetic acid (894 µL, 11.6 mmol) at
0 °C. The solution was stirred at room temperature for 3 h
until the disappearance of the starting material (TLC: n-
butanol/water/acetic acid 4:2:1). The volatiles were removed
under vacuum, and the residue was crystallized from EtOH
and Et2O to yield 38% (101 mg).
X-ray Analysis of Ethyl (5R*,2′R*)-5-(2-tert-Butoxycar-
bonylamino-2-methoxycarbonylethyl)-4,5-dihydroisox-
azole-3-carboxylate (13). Crystal Data. Single crystals
suitable for X-ray diffraction studies were grown from di-n-
propyl ether. C15H24N2O7, Mr 344.4, orthorhombic, space group
Pbca (No 61), a ) 11.721(2) Å, b ) 9.682(2) Å, c ) 31.642(5)
Å, V ) 3590.8(11) Å3, Z ) 8, Dc ) 1.274 Mg/m3, F(000) ) 1472,
µ(Mo KR) ) 0.101 mm-1, crystal size: 0.06 × 0.26 × 0.3 mm.
Data Collection and Reduction. A single crystal was
mounted and immersed in a stream of nitrogen gas [T ) 122-
(1) K]. Data were collected, using the graphite-monochromated
Mo KR radiation source (λ ) 0.71070 Å) on a KappaCCD
diffractometer. Data collection and cell refinement were
performed using COLLECT30 and DIRAX.31 Data reduction
was performed using EvalCCD.32 Correction for absorption was
performed using SADABS.33
Amino Acid 10A. White prisms. Rf: 0.31 (n-butanol/water/
acetic acid 4/2/1). Mp: 160-170 °C (dec). pKa: < 2.0 (R-COOH),
Structure Solution and Refinement. Positions of all non-
hydrogen atoms were found by direct methods (SHELXS97).34,35
Full-matrix least squares refinements (SHELXL97),36 were
performed on F2, minimizing Σw(Fo2 - kFc2)2, with anisotropic
displacement parameters of the non-hydrogen atoms. The
position of most of the hydrogen atoms were located in
subsequent difference electron density maps, and except for
the hydrogen atoms of methyl groups, all hydrogen atoms were
1
2.5 (ω-COOH), 8.8 (NH3+). H NMR (D2O): 1.29 (s, 3H), 1.33
(s, 3H), 1.93 (m, 1H), 2.20 (m, 1H), 3.27 (dd, J ) 6.4, 6.4 Hz,
1H), 3.98 (dd, J ) 6.9, 6.9 Hz, 1H). 13C NMR (D2O): 20.21,
26.72, 28.62, 50.39, 52.49, 90.96, 158.79, 165.16, 172.47. Anal.
(C9H14N2O5): C, H, N.
(4R*,2′S*)-4-(2-Amino-2-carboxyethyl)-5,5-dimethyl-
4,5-dihydroisoxazole-3-carboxylic Acid (10B). Step A.
Derivative 16 (500 mg, 1.29 mmol) was dissolved in EtOH (10
mL) and treated with 1 N NaOH (4 mL) at room temperature
overnight. The disappearance of the starting material was
monitored by TLC (CHCl3/MeOH 8:2 + 100 µL acetic acid).
The aqueous layer was washed with CH2Cl2, made acidic with
2 N HCl, and extracted with AcOEt. The organic phase was
dried over anhydrous Na2SO4 and, after evaporation of the
solvent, a white powder was obtained (362 mg, 1.09 mmol,
yield 85%).
Step B. The crude material obtained from the previous
transformation (362 mg, 1.09 mmol) was treated with a 30%
CH2Cl2 solution of trifluoroacetic acid (839 µL, 10.9 mmol) at
0 °C. The solution was stirred at room temperature for 3 h
until the disappearance of the starting material (TLC: n-
butanol/water/acetic acid 4:2:1). The volatiles were removed
under vacuum, and the residue was crystallized from EtOH
and Et2O to yield 36% (90 mg).
refined with fixed isotropic displacement parameters (Uiso
)
1.2Ueq for NH, CH, and CH2). The positions of the hydrogen
atoms of methyl groups were included in idealized positions
and refined with a riding model with fixed isotropic displace-
ment parameter (Uiso ) 1.5Ueq) of the parent atom. Extinction
correction was applied;36 extinction coefficient ) 0.0019(2).
Refinement (250 parameters, 3149 unique reflections) con-
2
verged at RF ) 0.048, wRF ) 0.085 [2581 reflections with Fo
> 4σ(Fo); w-1 ) (σ2(Fo )+(0.0251P)2 + 2.4415P), where P ) (Fo
2
2
+ 2Fc2)/3; S ) 1.167]. The residual electron density varied
between -0.19 and 0.20 e Å-3. Complex scattering factors for
neutral atoms were taken from the International Tables for
Crystallography as incorporated in SHELXL97.36
Further details of crystal structure of compound 13 can be
obtained, free of charge, on application to CCDC (CCDC
241975), 12 Union Road, Cambridge CB2 lEZ UK.
Pharmacology. Receptor Binding and Electrophysi-
ology Assays at iGluRs. The membrane preparations used
in receptor-binding experiments were prepared according to
Ransom and Stec37 with slight modifications, as previously
described.38
Amino Acid 10B. White prisms. Rf: 0.30 (n-butanol/water/
acetic acid 4:2:1). Mp: 120-130 °C (dec). pKa: < 2.0 (R-COOH),
2.15 (ω-COOH), 8.95 (NH3+). 1H NMR (D2O): 1.22 (s, 3H), 1.29
(s, 3H), 2.05 (m, 1H), 2.15 (m, 1H), 3.17 (m, 1H), 3.93 (m, 1H).
13C NMR (D2O): 20.04, 26.72, 28.49, 49.13, 51.72, 91.36,
158.06, 165.06, 172.17. Anal. (C9H14N2O5): C, H, N.
Affinity for AMPA,25 KA,26 and NMDA24 receptor sites was
determined using 5 nM [3H]AMPA, 5 nM [3H]kainic acid in
the absence of CaCl2, and 2 nM [3H]CGP 39653 ([3H]-(()-(E)-

Novel, Potent NMDA Receptor Antagonists
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 27 6747
(4) Das, S.; Sasaki, Y. F.; Rothe, T.; Premkumar, L. S.; Takasu, M.;
Crandall, J. E.; Dikkes, P.; Conner, D. A.; Rayudu, P. V.; Cheung,
W.; Chen, H. S. V.; Lipton, S. A.; Nakanishi, N. Increased NMDA
current and spine density in mice lacking the NMDA receptor
subunit NR3A. Nature (London) 1998, 393, 377-381.
(5) Chatterton, J. E.; Awobuluyl, M.; Premkumar, L. S.; Takanishi,
H.; Talantova, M.; Shin, Y.; Cui, J.; Tu, S.; Sevarino, K. A.;
Nakanishi, N.; Tong, G.; Lipton, S. A.; Zhang, D. Excitatory
glycine receptors containing the NR3 family of NMDA receptor
subunits. Nature (London) 2002, 415, 793-798.
2-amino-4-propyl-5-phosphono-3-pentenoate), respectively. The
tritiated ligands were all purchased from New England
Nuclear (NEN).
The amount of bound radioactivity was determined using a
Packard TOP-COUNT microplate scintillation counter.
The rat cortical wedge preparation, in a modified version,
was used for the determination of the depolarizing effects of
the compounds under study.27 Tests for agonist activities of
8A/8B and 10A/10B were performed by their application for
90 s. Tests for antagonist activities of 8A/8B and 10A/10B
were performed by their preapplication for 90 s followed by
coapplication of the compound under study with a standard
agonist, AMPA (5 µM), NMDA (10 µM), or KA (5 µM), for
further 90 s.
(6) Kemp, J. A.; Kew, J. N. C.; Gill, R. Handbook of Experimental
Pharmacology; Jonas, P., Monyer, H., Eds.; Springer: Berlin,
1999; Vol 141, pp 495-527.
(7) Lees, K. R.; Asplund, K.; Carolei, A.; Davis, S. M.; Diener, H.
C.; Kaste, M.; Orgogozo, J. M.; Whitehead, J. Glycine antagonist
(gavestinel) in neuroprotection (GAIN International) in patients
with acute stroke: A randomised controlled trial. GAIN Inter-
national Investigators. Lancet 2000, 355, 1949-1954.
(8) Sacco, R. L.; DeRosa, J. T.; Haley, E. C.; Levin, B.; Ordronneau,
P.; Phillips, S. J.; Rundek, T.; Snipes, R. G.; Thompson, J. L. P.
Glycine antagonist in neuroprotection for patients with acute
stroke: GAIN Americas: A randomized controlled trial. J. Am.
Med. Assoc. 2001, 285, 1719-1728.
(9) Nikam, S. S.; Meltzer, L. T. NR2B selective NMDA receptor
antagonists. Curr. Pharm. Des. 2002, 8, 845-855.
(10) Gotti, B.; Duverger, D.; Bertin, J.; Carter, C.; Dupont, R.; Frost,
J.; Gaudilliere, B.; MacKenzie, E. T.; Rousseau, J.; Scatton, B.
Ifenprodil and SL 82.0715 as cerebral anti-ischemic agents. I.
Evidence for efficacy in models of focal cerebral ischemia. J.
Pharmacol. Exp. Ther. 1988, 247, 1211-1221.
(11) Conti, P.; De Amici, M.; Joppolo di Ventimiglia, S.; Stensbøl, T.
B.; Madsen, U.; Bra¨uner-Osborne, H.; Russo, E.; De Sarro, G.;
Bruno, G.; De Micheli, C. Synthesis and Anticonvulsant Activity
of Novel Bicyclic Acidic Amino Acids. J. Med. Chem. 2003, 46,
3102-3108.
(12) Olverman, H. J.; Jones, A. W.; Mewett, K. N.; Watkins, J. C.
Structure/activity relations of N-methyl-D-aspartate receptor
ligands as studied by their inhibition of [3H]D-2-amino-5-
phosphonopentanoic acid binding in rat binding membranes.
Neuroscience 1988, 26, 17-31.
(13) Johnson, G.; Ornstein, P. L. Competitive NMDA Antagonistss
A Comprehensive Analysis of Molecular Biological, Structure
Activity and Molecular Modeling Relationships. Curr. Pharm.
Des. 1996, 2, 331-356.
(14) Gru¨nanger, P.; Vita-Finzi, P. In The Chemistry of Heterocyclic
Compounds; Taylor, E. C., Ed., J. Wiley & Sons: New York,
1991; Vol. 49.
Data Analysis. Binding data were analyzed by the non-
linear curve fitting program GRAFIT 3.0.39 Data were fitted
to the following equation: B ) 100 - (100 × [inhibition]n)/
(IC50 + [inhibitor]n), where B is the binding, expressed as a
n
percentage of total specific binding, and n is the Hill coefficient.
Metabotropic Testing. The mGluR subtypes mGluR1R,
mGluR2, mGluR4a and mGluR5a were expressed in Chinese
hamster ovary (CHO) cell lines. Cell were maintained and
tested as previously described.23
In Vivo Pharmacology. Male DBA/2 mice (6-12 g; 22-
26 days old) were used. The animals were housed in groups of
10 in PVC cages (260 mm × 440 mm × 120 mm) at a
temperature of 21-23 °C and a relative humidity of 57 ( 2%;
a 12 h light/dark cycle was applied (light on in the interval
7:00 a.m. to 7:00 p.m.). Food and water were available ad
libitum.
Procedure. DBA/2 mice were exposed to a loud sound
stimulus which produced a whole-body clonus.40 Drug (or
vehicle, water) was administered intraperitoneally (ip) in a
volume of 100 µL/10 g body weight. Prior to the test for sound-
induced seizures, mice were observed for drug related abnor-
mal motor behavior effects. Anticonvulsant tests were carried
out on individual mice 45 min after drug (or vehicle) admin-
istration under a Perspex dome (58 cm in diameter) fitted with
a doorbell generating a sound of 110 dB for a period of 60s or
until the onset of a clonic seizure. The sound stimulus produced
a sequential seizure response, consisting of a wild running
phase (score 1), clonic seizures (score 2), tonic extension (score
3), and occasionally respiratory arrest (score 4) as previously
reported.40 The compounds were dissolved in NaOH (0.1 mM)
and the pH adjusted to 7.3-7.7 with a phosphate buffer
solution for ip injection.
(15) Kozikowsky, A. P.; Adamczyk, M. Methods for the stereoselective
cis-cyanohydroxylation of olefins. J. Org. Chem. 1983, 48, 366-
372.
(16) Biagini, S. G. C.; Gibsone´Thomas, S. E.; Keen, S. P. Cross-
metathesis of unsaturated R-amino acid derivatives. J. Chem.
Soc., Perkin Trans. 1 1998, 2485-2500.
(17) Bachi, M. D.; Balanov, A.; Bar-Ner, N. Thiol-mediated free
radical cyclization of alkenyl and alkynyl isocyanides. J. Org.
Chem. 1994, 59, 7752-7758.
Acknowledgment. Professor Shigetada Nakanishi
is gratefully acknowledged for the kind gift of the
mGluR expressing CHO cell lines. This work was
financially supported by MIUR (COFIN 2003) Rome and
Universita` degli Studi di Milano (FIRST). H.B.O. was
supported by the Danish Medical Research Council the
Augustinus Foundation and Skibsreder Per Henriksen
Foundation. The technical assistance of Mr. F. Hansen
(University of Copenhagen) in collecting X-ray data is
gratefully acknowledged.
(18) (a) Hohenberg, P.; Kohn, W. Inhomogeneous Electron Gas. Phys.
Rev. 1964, 136, B864. (b) Kohn, W.; Sham, L. J. Self-Consistent
Equations Including Exchange and Correlation Effects. Phys.
Rev. 1965, 140, A1133.
(19) Lee, C.; Yang, W.; Parr, R. G. Development of the Colle-Salvetti
correlation-energy formula into a functional of the electron
density. Phys. Rev. B 1988, 37, 785-789. Becke, A. D. Density-
functional thermochemistry. III. The role of exact exchange. J.
Chem. Phys. 1993, 98, 5648-5652.
(20) Barone, V.; Cossi, M. J. Quantum Calculation of Molecular
Energies and Energy Gradients in Solution by a Conductor
Solvent Model. Phys. Chem. A 1998, 102, 1995-2001.
(21) Haasnoot, C. A. G.; de Leeuw, F. A. A. M.; Altona, C. The relation
between proton-proton NMR coupling constants and substitu-
ent electronegativities. I. An empirical generalization of the
Karplus equation. Tetrahedron 1980, 36, 2783.
(22) ORTEPII, Johnson, C. K. ORTEPII. A FORTRAN thermal-
ellipsoid plot program for crystal structure illustrations. Report
ORNL-5138; Oak Ridge National Laboratory: Oak Ridge, TN,
1976.
Supporting Information Available: Tables containing
the X-ray crystallographic crystal data of derivative 13. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(23) Bjerrum, E. J.; Kristensen, A. S.; Pickering, D. S.; Greenwood,
J. R.; Nielsen, B.; Liljefors, T.; Schousboe, A.; Bra¨uner-Osborne,
H.; Madsen, U. Design, synthesis, and pharmacology of a highly
subtype-selective GluR1/2 agonist, (RS)-2-amino-3-(4-chloro-3-
hydroxy-5-isoxazolyl)propionic acid (Cl-HIBO), J. Med. Chem.
2003, 46, 2246-2249.
(24) Sills, M. A.; Fagg, G.; Pozza, M.; Angst, C.; Brundish, D. E.; Hurt,
S. D.; Wilusz, E. J.; Williams, M. [3H]CGP 39653: A new
N-methyl-D-aspartate antagonist radioligand with low nanomo-
lar affinity in rat brain. Eur. J. Pharmacol. 1991, 192, 19-24.
References
(1) Dingledine, R.; Borges, K.; Bowie, D.; Traynelis, S. F. The
glutamate receptor ion channels. Pharmacol. Rev. 1999, 51,
7-61.
(2) Bra¨uner-Osborne, H.; Egebjerg, J.; Nielsen, E. Ø.; Madsen, U.;
Krogsgaard-Larsen, P. Ligands for glutamate receptors: Design
and therapeutic prospects. J. Med. Chem. 2000, 43, 2609-2645.
(3) McBain, C. J.; Mayer, M. L. N-methyl-D-aspartic acid receptor
structure and function. Physiol. Rev. 1994, 74, 723-760.

6748 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 27
Conti et al.
(25) Honore´, T.; Nielsen, M. Complex structure of quisqualate-
sensitive glutamate receptors in rat cortex. Neurosc. Lett. 1985,
54, 27-32.
(33) Sheldrick, G. M. SADABS. Program for absorption correction,
Bruker AXS analytical X-ray systems, Madison, WI, 1996.
(34) Sheldrick, G. M. Phase annealing in SHELX-90: Direct methods
for larger structures. Acta Crystallogr. 1990, A46, 467-473.
(35) Sheldrick, G. M. SHELXS97. Program for the solution of crystal
structures, University of Go¨ttingen, Germany, 1997.
(36) Sheldrick, G. M. SHELXL97. Program for crystal structure
refinement, University of Go¨ttingen, Germany, 1997.
(37) Ransom, R. W.; Stec, N. L. Cooperative modulation of [3H]MK-
801 to the N-methyl-D-aspartate receptor ion channel complex
by L-glutamate, glycine and polyamines. J. Neurochem. 1988,
51, 830-836.
(38) Stensbøl, T. B.; Johansen, T. N.; Egebjerg, J.; Ebert, B.; Madsen,
U.; Krogsgaard-Larsen, P. Resolution, absolute stereochemistry
and molecular pharmacology of the enantiomers of ATPA. Eur.
J. Pharmacol. 1999, 380, 153-162.
(39) Leatherbarrow, R. J. GraFit version 3.0; Erithacus Software:
Staines, UK, 1992.
(26) Braitman, D. J.; Coyle, J. T. Inhibition of [3H]kainic acid receptor
binding by divalent cations correlates with ion affinity for the
calcium channel. Neuropharmacology 1987, 26, 1247-1251.
(27) Harrison, N. L.; Simmonds, M. A. Quantitative studies on some
agonists of N-methyl-D-aspartate in slices of rat cerebral cortex.
Br. J. Pharmacol. 1985, 84, 381-391.
(28) Hutchison, A. J.; Williams, M.; Angst, C.; de Jesus, R.; Blan-
chard, L.; Jackson, R. H.; Wilusz, E. J.; Murphy, D. E.; Bernard,
P. S.; Schneider, J.; Campbell, T.; Guida, W. Sills, M. A.
4-(Phosphonoalkyl)- and 4-(Phosphonoalkenyl)-2-piperidinecar-
boxylic Acids: Synthesis, Activity at N-Methyl-D-aspartic Acid
Receptors, and Anticonvulsant Activity. J. Med. Chem. 1989,
32, 2171-2178.
(29) Avdeef, A. pH-metric log P. II: Refinement of partition coef-
ficients and ionization constants of multiprotic substances. J.
Pharm. Sci. 1993, 82, 183-190.
(30) COLLECT. Nonius B. V., Delft, The Netherlands 1999.
(31) Duisenberg, A. J. M. Indexing in single-crystal diffractometry
with an obstinate list of reflections. J. Appl. Crystallogr. 1992,
25, 92-96.
(32) Duisenberg, A. J. M.; Eval C. C. D. PhD Thesis, University of
Utrecht, The Netherlands 1998.
(40) De Sarro, G.; Ongini, E.; Bertorelli, R.; Aguglia, U.; De Sarro,
A. Excitatory amino acid neurotransmission through both
NMDA and non-NMDA receptors is involved in the anticonvul-
sant activity of felbamate in the DBA/2 mice. Eur. J. Pharmacol.
1994, 262, 11-19.
JM049409F