Novel 3-carboxy-and 3-phosphonopyrazoline amino acids as potent and selective NMDA receptor antagonists: Design, synthesis, and pharmacological characterization

Article abstract of DOI:10.1002/cmdc.201000184

The design and synthesis of new N1-substituted 3-carboxyand 3-phosphonopyrazoline and pyrazole amino acids that target the glutamate binding site of NMDA receptors are described. An analysis of the stereochemical requirements for high-affinity interaction with these receptors was performed. We identified two highly potent and selective competitive NMDA receptor antagonists, (5S,αR)-1 and (5S,αR)-4, which exhibit good in vitro neuroprotective activity and in vivo anticonvulsant activity by i.p. administration, suggesting that these molecules may have potential use as therapeutic agents.

Full text of DOI:10.1002/cmdc.201000184

DOI: 10.1002/cmdc.201000184
Novel 3-Carboxy- and 3-Phosphonopyrazoline Amino
Acids as Potent and Selective NMDA Receptor
Antagonists: Design, Synthesis, and Pharmacological
Characterization
Paola Conti,*[a] Andrea Pinto,[a] Lucia Tamborini,[a] Ulf Madsen,[b] Birgitte Nielsen,[b]
Hans Brꢀuner-Osborne,[b] Kasper B. Hansen,[b] Elisa Landucci,[c] Domenico E. Pellegrini-
Giampietro,[c] Giovambattista De Sarro,[d] Eugenio Donato Di Paola,[d] and Carlo De Micheli[a]
The design and synthesis of new N1-substituted 3-carboxy-
and 3-phosphonopyrazoline and pyrazole amino acids that
target the glutamate binding site of NMDA receptors are de-
scribed. An analysis of the stereochemical requirements for
high-affinity interaction with these receptors was performed.
We identified two highly potent and selective competitive
NMDA receptor antagonists, (5S,aR)-1 and (5S,aR)-4, which ex-
hibit good in vitro neuroprotective activity and in vivo anticon-
vulsant activity by i.p. administration, suggesting that these
molecules may have potential use as therapeutic agents.
Introduction
l-Glutamic acid [(S)-Glu] is the most abundant excitatory neu-
rotransmitter in the mammalian central nervous system (CNS),
where it is involved in the modulation of many physiological
processes such as learning, memory, and synaptic plasticity.[1]
Glutamate-activated receptors belong to two families: iono-
tropic glutamate receptors (iGluRs) and metabotropic gluta-
mate receptors (mGluRs). The iGluR family comprises three het-
erogeneous classes, named for the selective agonists N-
methyl-d-aspartic acid (NMDA), 2-amino-3-(3-hydroxy-5-
methyl-4-isoxazolyl)propionic acid (AMPA), and kainic acid
(KA).[1,2] The iGluRs exist as tetramers; to date, seven NMDA re-
ceptor subunits (GluN1, GluN2A–2D and GluNR3A–B), four
AMPA receptor subunits (GluA1–4), and five KA receptor subu-
nits (GluK1–5) have been cloned and characterized. In contrast,
the mGluRs are a heterogeneous class of G-protein-coupled re-
ceptors; to date, eight mGluR subtypes have been cloned and
classified in three groups according to their sequence homolo-
gy, second-messenger coupling, and pharmacology: group I
(mGluRs 1 and 5), group II (mGluRs 2 and 3), and group III
(mGluRs 4, 6, 7, and 8).[3,4]
and w-acidic groups is a 4–6 carbon atom chain; the w-carbox-
ylate may be replaced by a phosphonate moiety to increase
receptor affinity (Figure 1).[10] We previously disclosed the struc-
tures of two new competitive NMDA receptor antagonists,
F94b and F94c (Figure 1),[11] which are characterized by the
Figure 1. Structures of model NMDA antagonists.
It is well known that over-activation of iGluRs generates an
uncontrolled influx of calcium, triggering an excitotoxic cas-
cade that leads to neurodegeneration.[5] This phenomenon is
associated with several acute and chronic neurodegenerative
diseases, including cerebral ischemia, traumatic brain injury,
spinal injury, epilepsy, amyotrophic lateral sclerosis, and Parkin-
son’s, Alzheimer’s and Huntington’s diseases.[6,7] NMDA recep-
tors are the major mediators of excitotoxicity.[8] The blockade
of such receptors can be achieved in several ways, including
the use of competitive Glu antagonists, glycineB antagonists,
polyamine site antagonists, or open-channel blockers.[9]
[a] Prof. P. Conti, Dr. A. Pinto, Dr. L. Tamborini, Prof. C. De Micheli
Dipartimento di Scienze Farmaceutiche “P. Pratesi”
Universitꢀ degli Studi di Milano
Via Mangiagalli 25, 20133 Milano (Italy)
Fax: (+39)02-50319326
E-mail: paola.conti@unimi.it
[b] Prof. U. Madsen, B. Nielsen, Prof. H. Brꢁuner-Osborne, Dr. K. B. Hansen
Department of Medicinal Chemistry, Faculty of Pharmaceutical Sciences
University of Copenhagen (Denmark)
[c] Dr. E. Landucci, Prof. D. E. Pellegrini-Giampietro
Dipartimento di Farmacologia Preclinica e Clinica
Universitꢀ degli Studi di Firenze (Italy)
[d] Prof. G. De Sarro, Prof. E. Donato Di Paola
Dipartimento di Medicina Sperimentale e Clinica
Universitꢀ “Magna Graecia” di Catanzaro (Italy)
Competitive antagonists are typically conformationally con-
strained Glu homologues in which the spacer between the a-
ChemMedChem 2010, 5, 1465 – 1475
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1465

P. Conti et al.
MED
presence of a homologated Glu chain partially locked in an iso-
xazoline ring. In both cases, the eutomers exhibited R configu-
ration at the stereogenic a-center, as observed for a majority
of NMDA antagonists [(5S,aR)-F94b Ki =0.10 mm vs. (5R,aS)-
F94b Ki =1.9 mm; (5R,aR)-F94c Ki =0.51 mm vs. (5S,aS)-F94c
Ki =2.5 mm].[12] Further pharmacological analysis emphasized
the significance of (5S,aR)-F94b and (5R,aR)-F94c as neuropro-
tective agents, as demonstrated by the results obtained using
an in vitro model of cerebral ischemia.[12] Evaluation of the se-
lectivity of these compounds toward different NMDA receptor
subtypes revealed a preference for GluN2A and GluN2B vs.
GluN2C and GluN2D subtypes.[12]
cleus [compounds (R)-3 and (S)-3, Figure 2] was used to evalu-
ate the effects of omitting the stereogenic center in position 5
and of using a different ring conformation.
Results and Discussion
Chemistry
The target amino acids (ꢀ)-1 and (ꢀ)-2 were synthesized
through 1,3-dipolar cycloaddition of dipolarophile (ꢀ)-6[11] with
N-phenyl-ethoxycarbonylformonitrilimine, generated in situ
from its stable precursor, ethyl-2-chloro-2-(phenylhydrazono)
acetate 7, in the presence of base (Scheme 1).[13] The cycloaddi-
Based on these results, we decided to pursue the design of
new NMDA receptor antagonists with improved affinity and a
higher degree of subtype selectivity. According to our previous
experience,[13] replacement of the D2-isoxazoline ring with a
D2-pyrazoline nucleus may represent an efficient strategy to
enhance selectivity for a specific molecular target. This type of
substitution potentially improves molecular diversity by func-
tionalizing the nitrogen in position 1, where substituents could
be introduced to create additional interactions with the bind-
ing site of the target protein, in an attempt to improve both
binding affinity and selectivity. Moreover, it has been reported
that NMDA antagonists with large side groups have the poten-
tial to differentiate between GluN2 subunits.[14] Taking into ac-
count the interaction between a tyrosine residue (Y730 in
GluN2A) and the isoxazoline oxygen of our most active com-
pound,[12] we envisaged that addition of an aromatic group
(i.e. a phenyl moiety) to the nitrogen at position 1 could pre-
serve and possibly strengthen such an interaction, while fur-
ther exploring the binding pocket in this region.
We initially synthesized and evaluated racemic derivatives
(ꢀ)-1 and (ꢀ)-2 (Figure 2). Subsequently, we prepared the
pure enantiomers (5S,aR)-1, (5R,aS)-1, (5R,aR)-2, and (5S,aS)-2
to examine in detail the relevance of absolute configuration
toward biological activity. Furthermore, in an attempt to en-
hance NMDA antagonist potency, we replaced the distal car-
boxylate with a phosphonate group, a bioisostere previously
used with success in this field [compounds 4–5, Figure 2].[10] Fi-
nally, substitution of the pyrazoline moiety with a pyrazole nu-
Scheme 1. Reagents and conditions: a) NaHCO3, EtOAc, 24 h, RT; b) TFA
(30%), CH2Cl2, 3 h, 08C!RT; c) Boc2O, Et3N, CH2Cl2, 12 h, 08C!RT; d) NaOH
(1n), MeOH, 3 h, RT.
tion reaction produced a mixture of stereoisomers (ꢀ)-8 and
(ꢀ)-9 in an approximate 1:1 ratio, as estimated by 1H NMR
spectroscopy. Chromatographic separation of the two stereo-
isomers was unsuccessful at this stage; therefore, the mixture
was treated with a 30% solution of trifluoroacetic acid (TFA) in
dichloromethane to remove the N-Boc protecting group. The
corresponding primary amines (ꢀ)-10 and (ꢀ)-11 were easily
separated by flash chromatography and were then reconverted
Figure 2. Structures of target amino acids.
1466
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1465 – 1475

NMDA Receptor Antagonists
into (ꢀ)-8 and (ꢀ)-9, respectively, using standard reaction con-
ditions.
Relative
configuration
of
stereoisomers
(ꢀ)-8 and (ꢀ)-9 was later established by X-ray analysis carried
out on cycloadduct (ꢁ)-8. Amino acids (ꢀ)-1 and (ꢀ)-2 were
ultimately obtained from (ꢀ)-8 and (ꢀ)-9 in two steps by alka-
line hydrolysis of the two ester functionalities, followed by TFA
removal of the N-Boc protecting group, as described above.
The synthetic approach described for the racemates was
also applied to the synthesis of enantiomerically pure amino
acids (5S,aR)-1, (5R,aR)-2, (5R,aS)-1, and (5S,aS)-2, starting from
enantiopure dipolarophile (R)-6[15] or (S)-6,[16] respectively
(Scheme 1). Intermediate (ꢁ)-8 was crystallized successfully
from diisopropyl ether/hexane and, thus, was able to be sub-
mitted for X-ray analysis. Due to the absence of anomalous
scattering, we were only able to assign relative spatial arrange-
ment and not absolute configuration of the two stereocenters.
Nevertheless, because the absolute configuration of the a-
amino acid carbon was retained from the starting material (S)-
6 and therefore known, the configuration 5R,aS was assigned
to compound (ꢁ)-8 (Figure 3). The absolute configurations of
(+)-8, (ꢁ)-9 and (+)-9 were similarly assigned.
Scheme 2. Reagents and conditions: a) PDC, DMF, 4 h, RT; b) NaOH (1n),
MeOH, 3 h, RT; c) TFA (30%), CH2Cl2, 3 h, 08C!RT.
phile (R)-6 or (S)-6, respectively (Scheme 3). Cycloadducts
(5S,aR)-14 and (5R,aR)-15 were separated by flash chromatog-
raphy and submitted to hydrolysis with aqueous 1n sodium
hydroxide. Subsequent treatment with bromotrimethylsilane
(TMSBr; 10 equiv) afforded the final products (5S,aR)-4 and
(5R,aR)-5 (Scheme 3). The same reaction sequence was applied
for the synthesis of (5R,aS)-4 and (5S,aS)-5, beginning from
alkene (S)-6.
Figure 3. Perspective view of molecular structure of (ꢁ)-8. Displacement el-
lipsoids are drawn at 20% probability level.
To obtain the corresponding pyrazole derivatives, a mixture
of compounds (5S,aR)-8 and (5R,aR)-9 was oxidized with pyri-
dinium dichromate (PDC) in DMF to yield pyrazole (R)-12,
which was submitted to standard deprotection procedures to
yield final amino acid (R)-3 (Scheme 2). The same approach
was used to prepare (S)-3, beginning from a mixture of two
diastereomeric pyrazolines exhibiting the S configuration at
the amino acid stereogenic center (Scheme 2).
Synthesis of the new amino acids (5S,aR)-4, (5R,aR)-5,
(5R,aS)-4, and (5S,aS)-5 was achieved by using our recently
published method for the preparation of 5-substituted 3-phos-
phono-pyrazolines.[17] The procedure involves in situ genera-
tion of dimethoxyphosphorylformonitrilimine from its stable
precursor 13 and 1,3-dipolar cycloaddition to chiral dipolaro-
Scheme 3. Reagents and conditions: a) NaHCO3, EtOAc, 24 h, RT; b) NaOH
(1n), MeOH, 3 h, RT; c) TMSBr, CH2Cl2, 12 h, RT.
ChemMedChem 2010, 5, 1465 – 1475
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1467

P. Conti et al.
MED
Pharmacology
All new derivatives were subjected to in vitro assays, evaluat-
ing receptor binding to rat cortical membranes, using the radi-
oligands [3H]CGP39653, [3H]AMPA, and [3H]KA as representa-
tives for NMDA, AMPA, and KA receptors, respectively
(Table 1).[18–20] Amino acid (ꢀ)-2 displayed a fivefold increase in
the binding affinity for NMDA receptors, relative to (ꢀ)-F94c
(Ki =0.20 mm vs. 0.96 mm, Table 1). Additionally, compound
Figure 4. Compounds A) (ꢀ)-1 and B) (ꢀ)-2 were evaluated at concentra-
tions of 0.3 and 3 mm for their ability to inhibit current responses to 10 mm
Glu in Xenopus oocytes expressing cloned rat NMDA receptors, containing
GluN1 in combination with either GluN2A, GluN2B, GluN2C, or GluN2D. Re-
sponses are normalized to the control response using 10 mm Glu alone. Data
represent the mean ꢀSEM from 3–4 oocytes for each compound toward
NMDA receptor subtypes.
Table 1. Receptor binding affinities at native rat iGluRs.[a]
Ki [mm]
[3H]CGP39653
IC50 [mm]
[3H]AMPA
Compound
[3H]KAIN
(ꢀ)-F94b[b]
(5R,aS)-F94b[b]
(5S,aR)-F94b[b]
(ꢀ)-F94c[b]
(5R,aR)-F94c[b]
(5S,aS)-F94c[b]
(ꢀ)-1
0.21 [0.18;0.23]
1.9 [1.8;2.0]
0.10 [0.09;0.12]
0.96 [0.88;1.1]
0.51 [0.49;0.53]
2.5 [2.4;2.7]
0.026 [0.024;0.028]
0.017 [0.016;0.019]
5.5 [4.8;6.3]
0.20 [0.18;0.22]
5.4 [4.7;6.2]
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
the results of this investigation show that (ꢀ)-1 and (ꢀ)-2
share with model compounds F94b and F94c a similar NMDA
receptor subtype preference for GluN2A and GluN2B over
GluN2C and GluN2D.
Comparative analysis of the binding affinities obtained for
pure enantiomers suggests that the most active antagonists
are (5S,aR)-1 (Ki: 0.017 mm, Table 1) and (5S,aR)-4 (Ki: 0.031 mm,
Table 1), which correlate stereochemically to our model com-
pound (5S,aR)-F94b (Ki: 0.10 mm, Table 1). This is in agreement
with the general rule that NMDA receptor antagonists possess
an R configuration at the a-amino acid center. In addition, it
appears evident that the presence of a stereocenter in posi-
tion 5 of the heterocyclic nucleus plays a pivotal role in deter-
mining the correct orientation of the pharmacophoric groups
to maximize interactions with the target protein. When the w-
acidic group and the side chain at C5 are in the same plane, as
observed for pyrazole derivatives (R)-3 or (S)-3, the affinity for
NMDA receptors is no longer seen (Ki: 48 and >100 mm, re-
spectively, Table 1). The bioisosteric replacement of the distal
carboxylate with a phosphonate group, surprisingly, did not
produce the expected increase in affinity; the two compounds
exhibited nearly identical activity in this regard.
(5S,aR)-1
(5R,aS)-1
(ꢀ)-2
(5R,aR)-2
(5S,aS)-2
(R)-3
0.18 [0.15;0.22]
48 [45;51]
(S)-3
>100
(5S,aR)-4
(5R,aS)-4
(5R,aR)-5
(5S,aS)-5
0.031 [0.026;0.035]
0.87 [0.76;0.99]
1.1 [1.0;1.3]
2.3 [2.2;2.4]
[a] Values are expressed as the antilog to the log mean of at least three
individual experiments; values in brackets [min;max] indicate ꢀSEM ac-
cording to a logarithmic distribution of Ki. [b] Data from reference [12].
(ꢀ)-1 showed a NMDA affinity eightfold higher than (ꢀ)-F94b
(Ki =0.026 mm vs. 0.21 mm, Table 1), therefore behaving as a
very potent NMDA receptor ligand. Compounds (ꢀ)-1 and
(ꢀ)-2 exhibited no activity at other iGlu receptors (Table 1).
Moreover, they were inactive as either agonists or antagonists
at up to 1 mm concentration toward rat mGluR1 (group I),
mGluR2 (group II), and mGluR4 (group III) receptors expressed
in CHO cells.[21]
Due to the high potency of our two antagonists, we decided
to extend the investigation of their pharmacological profiles,
and to evaluate their potential use as neuroprotectants. Select-
ed compounds (5S,aR)-1 and (5S,aR)-4 were examined as po-
tential neuroprotective agents in cultured murine cortical cells
exposed to oxygen-glucose deprivation (OGD) or NMDA at
300 mm; these assays represent well-established in vitro models
of cerebral ischemia and excitotoxicity (Figure 5). As previously
described,[23] exposure to OGD for 60 min or to 300 mm NMDA
for 10 min produced an intermediate level of neuronal
damage, as measured by the release of lactate dehydrogenase
(LDH) into the bathing medium, which is approximately 75%
(OGD) and 70% (NMDA) of that observed upon exposing the
cultures to 1 mm Glu for 24 h (maximal damage).
To confirm that (ꢀ)-1 and (ꢀ)-2 behave as antagonists, and
to evaluate their subtype selectivity, the new ligands were
tested on Xenopus oocytes expressing GluN1, in combination
with either GluN2A, GluN2B, GluN2C, or GluN2D subunits.[12]
Potential antagonists were evaluated at concentrations of 0.3
and 3 mm for inhibition of current responses to 10 mm Glu
(Figure 4). At 3 mm, both compounds antagonized responses
from GluN2A, and GluN2B and exhibited no or little antagonis-
tic effects toward GluN2C and GluN2D. Consistent with the
binding data, derivative (ꢀ)-1 demonstrated greater inhibition
of the Glu response than (ꢀ)-2 at 3 mm. The two antagonists
have a preference for GluN2A and GluN2B, as the potency of
Glu at the four receptor subtypes is similar (Glu EC50 values of
GluN1/GluN2A, GluN1/GluN2B, GluN1/GluN2C, and GluN1/
GluN2D are 2.9, 1.8, 1.0, and 0.5 mm, respectively).[22] Therefore,
Upon addition to the incubation medium during OGD or
NMDA exposure, and following the subsequent 24 h recovery
period, compounds (5S,aR)-1 and (5S,aR)-4 attenuated OGD
and NMDA-induced neuronal death in a dose-dependent
manner, with the maximal effect observed at 1 mm. As shown
1468
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1465 – 1475

NMDA Receptor Antagonists
nist CPPene (Figure 1), with higher potency than those report-
ed for AP7 or certain well-known anticonvulsant drugs (Table 2
and Figure 6). Following i.p. administration, (5S,aR)-1 exhibited
maximal protection at 15 min, while (5S,aR)-4 exerted maximal
protection between 30 and 90 min (Figure 7). Again, it was ob-
Table 2. Anticonvulsant activity of (5S,aR)-1, (5S,aR)-4, CPPene, AP7, fel-
bamate, valproate, and phenytoin against audiogenic seizures in DBA/2
mice.[a]
Compound
ED50 [mmolkgꢁ1
]
Clonus
Tonus
(5S,aR)-1
(5S,aR)-4
CPPene
2.7 (1.2–5.8)
2.8 (1.5–5.5)
1.8 (1.2–2.3)
0.5 (0.2–0.9)
0.3 (0.2–0.7)
0.8 (1.4–1.4)
AP7
30.5 (22.3–41.8)
23.5 (15.3–36.1)
97.0 (51.0–185.0)
187.0 (127.0–273.0)
7.3 (5.8–9.1)
felbamate
valproate
phenytoin
204.0 (149.0–282.0)
258.0 (196.0–341.0)
9.1 (5.6–14.9)
[a] All data were calculated according the method of Litchfield and Wil-
coxon[32] and represent the 95% confidence limits of ED50 values. A mini-
mum of 32 animals (i.p. administration) were used to calculate each ED50
value.
Figure 5. Effects of (5S,aR)-1 and (5S,aR)-4 on mixed mouse cortical cultures
exposed to A) OGD or B) 300 mm NMDA. Under basal conditions, drugs were
added to the medium, and neuronal death was assessed after 24 h by meas-
uring LDH release in the medium. OGD was applied for 60 min and 300 mm
NMDA for 10 min in cortical cultures. For these experiments, drugs were
added 15 min before OGD or NMDA and were left in the medium during
OGD and NMDA treatment and the subsequent 24 h recovery period. Data
are expressed as a percentage of OGD and NMDA-induced neuronal
damage. Values are the mean ꢀSEM of at least six experiments performed
in triplicate (*p<0.05, **p<0.01 versus the corresponding control). Statisti-
cal analyses was performed using analysis of variance and Tukey’s w test for
multiple comparisons.
in Figure 5, amino acid (5S,aR)-4 displayed a greater neuropro-
tective effect than (5S,aR)-1 in both assays. Derivative (5S,aR)-4
attenuates NMDA and OGD injury by approximately 78ꢀ3.5%
and 85ꢀ5%, respectively, while (5S,aR)-1 does so by 70ꢀ3%
and 64ꢀ5.5%, respectively. Thus, in contrast to the results ob-
tained in binding experiments, the bioisosteric replacement of
the w-carboxylate with a phosphonate moiety yielded a sizable
improvement in biological activity in these assays.
Figure 6. Dose–response curves for the anticonvulsant effects of (5S,aR)-1
and (5S,aR)-4 following i.p. administration. Abscissae correspond to dosage,
and ordinates show: A) percent of clonic seizures and B) percent of tonic
seizures. Mice (in groups of 10 per dose) were pretreated with increasing
The same compounds were also tested in vivo in DBA/2
mice to evaluate their ability to protect the animals against au-
diogenic seizures. Both compounds were able to produce
dose-dependent protection against the tonic and clonic phase
of these seizures. The ED50 values for (5S,aR)-1 and (5S,aR)-4
are similar to those of the well-known NMDA receptor antago-
~
&
doses of (5S,aR)-1 ( ) or (5S,aR)-4 ( ), and auditory stimulation was applied
for 30 min following drug administration (**p<0.01 versus the correspond-
ing control). Statistical analysis was performed using Fisher’s exact probabili-
ty test.
ChemMedChem 2010, 5, 1465 – 1475
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1469

P. Conti et al.
MED
served that replacement of the w-carboxylate with a phospho-
nate group produces significant modification of the pharmaco-
logical profile. In fact, although we obtained nearly identical
ED50 values for anticonvulsant activity, a result that seems to
parallel the data of binding experiments, the two compounds
displayed appreciable differences in pharmacokinetic profiles
with a significant impact on overall pharmacological activity.
Indeed, (5S,aR)-4 had a slower onset of action than (5S,aR)-1, a
delay that may be attributed to a different absorption rate into
the CNS.
sible stereoisomers, as well as by abolishing the stereocenter
associated with the heterocycle. Our goal to improve the po-
tency of lead compounds F94b and F94c was successfully ac-
complished. Moreover, the introduction of a phosphonic acid
group has emphasized the possibility for modification of not
only the pharmacodynamic but also, interestingly, the pharma-
cokinetic profile of the molecule.
The ligands described are able to differentiate between the
GluN2A/B subtypes over the GluN2C/D subtypes; however,
higher selectivity toward a single receptor subtype is desired.
Work is underway to identify suitable substituents for the N1
position in order to further enhance interactions with an
amino acid residue of the active site, which is predicted to in-
crease the affinity for a specific receptor subunit, particularly
the GluN2B subtype.
Additionally, Figure 7 clearly shows that (5S,aR)-4 offers a
longer period of protection against seizures, a result that, to-
gether with the notable neuroprotective activity described
above, suggests this compound may be a promising neuropro-
tective agent. Further pharmacological investigations, as well
as evaluation of pharmacokinetic and toxicological parameters,
will be pursued and the results reported in due course.
Experimental Section
Materials and methods
1H NMR and 13C NMR spectra were recorded with a Varian Mercury
300 (300 MHz) spectrometer. Chemical shifts (d) are expressed in
ppm, and coupling constants (J) are expressed in Hz. Rotary power
determinations were carried out using a Jasco P-1010 spectropo-
larimeter, coupled with a Haake N3-B thermostat. TLC analyses
were performed on commercial silica gel 60 F254 aluminum sheets;
spots were further evidenced by spraying with a dilute alkaline po-
tassium permanganate solution or ninhydrin. Melting points were
determined on a model B 540 Bꢂchi apparatus and are uncorrect-
ed. Microanalyses (C, H, N) of new compounds were within ꢀ0.4%
of theoretical values.
Ethyl (5R*,aS*)-5-(2-tert-butoxycarbonylamino-2-methoxycarbo-
nylethyl)-4,5-dihydro-1-phenyl-1H-pyrazole-3-carboxylate [(ꢀ)-8]
and ethyl (5R*,aR*)-5-(2-tert-butoxycarbonylamino-2-methoxy-
carbonylethyl)-4,5-dihydro-1-phenyl-1H-pyrazole-3-carboxylate
[(ꢀ)-9]. Ethyl-2-chloro-2-(phenylhydrazone)acetate 7[13] (2.22 g,
9.81 mmol) and solid NaHCO3 (2.75 g, 32.7 mmol) were added to a
solution of methyl 2-tert-butoxycarbonylaminopent-4-enoate (ꢀ)-
6[11] (1.50 g, 6.54 mmol) in EtOAc (25 mL). The mixture was stirred
vigorously for 24 h; progress of the reaction was followed by TLC
[petroleum ether (PE)/EtOAc, 7:3]. H2O (5 mL) was added, and the
organic layer was separated and dried over anhydrous Na2SO4. The
crude material obtained upon evaporation of the solvent was puri-
fied by column chromatography on silica gel (PE/EtOAc, 85:15) to
give an inseparable mixture of two diastereomers: (ꢀ)-8 and (ꢀ)-9
(1.92 g, 70%). Rf =0.28 (PE/EtOAc, 4:1).
Figure 7. Anticonvulsant effects following i.p. administration of 10 mmolkgꢁ1
~
&
of (5S,aR)-1 ( ) and (5S,aR)-4 ( ), plotted against audiogenic seizures in
^
DBA/2 mice; : vehicle. Abscissa corresponds to time following i.p. adminis-
tration of drug. Groups of 10 mice were used for determination of each
point, which represents the mean seizure score ꢀSEM against audiogenic
seizures in DBA/2 mice (**p<0.01 versus the corresponding control). Statisti-
cal analysis was performed using Fisher’s exact probability test.
Conclusions
In the present work we have described the development of
new, highly potent, NMDA receptor antagonists with promis-
ing pharmacological activity as neuroprotective agents. These
agents may potentially be used to counteract damages in-
duced by cerebral ischemia, and are characterized by good
in vivo anticonvulsant activity. The rationale for their design
was based on the structural modification of previously de-
scribed NMDA antagonists that we decided to further modify
in order to strengthen interactions with the receptor binding
site. To this end, we replaced the original D2-isoxazoline ring
with an N1-substituted pyrazoline and, subsequently, replaced
the distal carboxylate with a bioisosteric phosphonate group.
Finally, we investigated the relevance of stereochemistry
toward biological activity by preparing and evaluating all pos-
Ethyl
(5R*,aS*)-5-(2-amino-2-methoxycarbonylethyl)-4,5-dihy-
dro-1-phenyl-1H-pyrazole-3-carboxylate [(ꢀ)-10] and ethyl
(5R*,aR*)-5-(2-amino-2-methoxycarbonylethyl)-4,5-dihydro-1-
phenyl-1H-pyrazole-3-carboxylate [(ꢀ)-11]. The mixture of com-
pounds (ꢀ)-8 and (ꢀ)-9 (1.92 g, 4.57 mmol) was treated with a
30% CH2Cl2 solution of TFA (11.6 mL) at 08C. The solution stirred at
room temperature for 3 h, and the reaction was followed by TLC
(PE/EtOAc, 7:3). A saturated solution of NaHCO3 (50 mL) and CH2Cl2
(35 mL) were then added, and the organic layer was separated,
dried over anhydrous Na2SO4, and evaporated under reduced pres-
sure. The crude material was purified on silica gel (PE/EtOAc, 3:7)
to give (ꢀ)-10 (0.49 g, 34%) and (ꢀ)-11 (0.53 g, 36%) as yellow
oils. The amines were used directly in the subsequent synthetic
step without further purification. (ꢀ)-10: Rf =0.64 (EtOAc); (ꢀ)-11
Rf =0.50 (EtOAc).
1470
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1465 – 1475

NMDA Receptor Antagonists
Ethyl (5R*,aS*)-5-(2-tert-butoxycarbonylamino-2-methoxycarbo-
nylethyl)-4,5-dihydro-1-phenyl-1H-pyrazole-3-carboxylate [(ꢀ)-8]
from [(ꢀ)-10]. Et3N (0.32 mL, 2.25 mmol) and a solution of di-tert-
butyl dicarbonate (425 mg, 1.95 mmol) in CH2Cl2 (3 mL) were
added to a stirred solution of (ꢀ)-10 (0.49 g, 1.50 mmol) in CH2Cl2
(10 mL) at 08C. The reaction mixture stirred at room temperature
overnight and was monitored by TLC (EtOAc). The solvent was
evaporated under reduced pressure, and the crude material was
purified by column chromatography on silica gel (PE/EtOAc, 85:15)
to give (ꢀ)-8 (0.50 g, 80%) as a pale-yellow oil; Rf =0.28 (PE/EtOAc,
4:1); 1H NMR ([D6]DMSO, 1008C): d=1.28 (t, J=7.0, 3H); 1.42 (s,
9H); 1.83 (ddd, J=4.1, 10.2, 14.0, 1H); 2.20 (ddd, J=2.6, 10.7, 14.0,
1H); 2.96 (dd, J=5.6, 17.7, 1H); 3.23 (dd, J=11.7, 17.7, 1H); 3.60 (s,
3H); 4.10 (ddd, J=4.1, 8.0, 10.7, 1H); 4.24 (q, J=7.0, 2H); 4.60
(dddd, J=2.6, 5.6, 10.2, 11.7, 1H); 6.92 (t, J=7.2, 1H); 7.12 (d, J=
8.0, 1H); 7.18 (d, J=9.0, 2H); 7.27 (dd, J=7.2, 9.0, 2H); 13C NMR
([D6]DMSO): d=14.93, 28.82, 31.80, 36.38, 50.55, 52.69, 58.57, 61.11,
79.25, 114.99, 121.52, 129.92, 139.75, 142.41, 156.34, 162.76,
173.10; Anal. calcd for C21H29N3O6: C 60.13, H 6.97, N 10.02, found:
C 60.18, H 7.04, N 9.90.
for C21H29N3O6: C 60.13, H 6.97, N 10.02, found: C 60.22, H 7.13, N
9.77.
(5R*,aR*)-5-(2-Amino-2-carboxyethyl)-4,5-dihydro-1-phenyl-1H-
pyrazole-3-carboxylic acid [(ꢀ)-2]. A) Compound (ꢀ)-9 (0.49 g,
1.16 mmol) was dissolved in MeOH (10 mL) and treated with 1n
aqueous NaOH (3.5 mL). Disappearance of the starting material
was monitored by TLC (PE/EtOAc, 7:3). Upon evaporation of the
MeOH, the aqueous layer was washed with Et2O, acidified with 2n
aqueous HCl, and extracted with EtOAc. The organic layer was
dried over anhydrous Na2SO4 and, upon evaporation of the solvent,
the diacid derivative (0.37 g, 86%) was obtained as a yellow solid.
B) The diacid (0.37 g, 1.00 mmol) was treated with a 30% CH2Cl2
solution of TFA (0.77 mL, 10.00 mmol) at 08C. The solution stirred
at room temperature for 3 h, and the reaction was followed by TLC
(CHCl3/MeOH, 9:1 + 1% AcOH). Volatiles were removed under re-
duced pressure, and the residue was crystallized from EtOH/Et2O
to give (ꢀ)-2 (0.16 g, 70%) as a yellow powder; mp: >1888C
(dec.); Rf =0.44 (BuOH/H2O/AcOH, 4:2:1); 1H NMR (D2O + 1 drop
CF3COOD): d=1.66 (ddd, J=3.8, 10.2, 14.0, 1H); 2.13 (ddd, J=3.0,
11.0, 14.0, 1H); 2.76 (dd, J=5.1, 17.6, 1H); 3.09 (dd, J=12.1, 17.6,
1H); 3.80 (dd, J=3.8, 11.0, 1H); 4.60 (m, 1H); 6.80 (t, J=7.7, 1H);
6.93 (d, J=7.7, 2H); 7.18 (t, J=7.7, 2H); 13C NMR (D2O + 1 drop
CF3COOD): d=31.55, 35.52, 49.02, 58.72, 115.14, 122.41, 129.72,
139.56, 141.23, 165.65, 171.17; Anal. calcd for C13H15N3O4: C 56.31,
H 5.45, N 15.15, found: C 56.19, H 5.63, N 14.95.
(5R*,aS*)-5-(2-Amino-2-carboxyethyl)-4,5-dihydro-1-phenyl-1H-
pyrazole-3-carboxylic acid [(ꢀ)-1]. A) Compound (ꢀ)-8 (0.50 g,
1.20 mmol) was dissolved in MeOH (5 mL) and treated with 1n
aqueous NaOH (3.6 mL). Disappearance of the starting material
was monitored by TLC (PE/EtOAc, 7:3). Upon evaporation of the
MeOH, the aqueous layer was washed with Et2O, acidified with 2n
aqueous HCl, and extracted with EtOAc. The organic layer was
dried over anhydrous Na2SO4 and, upon evaporation of the solvent,
the diacid derivative (0.44 g, 98%) was obtained as a yellow solid.
B) The diacid derivative (0.44 g, 1.18 mmol) was treated with a
30% CH2Cl2 solution of TFA (3 mL) at 08C. The solution stirred at
room temperature for 3 h, and the reaction was followed by TLC
(CHCl3/MeOH, 9:1 + 1% AcOH). Volatiles were removed under re-
duced pressure and the residue was crystallized from EtOH/Et2O to
give (ꢀ)-1 (0.17 g, 53%) as a yellow powder; mp: >1588C (dec.);
Rf =0.42 (BuOH/H2O/AcOH, 4:2:1); 1H NMR (D2O +1 drop
CF3COOD): d=1.96 (ddd, J=6.3, 9.3, 15.0, 1H); 2.04 (ddd, J=3.0,
7.7, 15.0, 1H); 2.77 (dd, J=5.5, 17.9, 1H); 3.09 (dd, J=12.1, 17.9,
1H); 3.92 (dd, J=6.3, 7.7, 1H); 4.65 (m, 1H); 6.85 (t, J=7.7, 1H);
6.96 (d, J=7.7, 2H); 7.18 (t, J=7.7, 2H); 13C NMR (D2O + 1 drop
CF3COOD): d=32.14, 35.80, 49.84, 58.48, 115.61, 122.66, 129.76,
139.71, 141.37, 165.65, 171.14; Anal. calcd for C13H15N3O4: C 56.31,
H 5.45, N 15.15, found: C 56.26, H 5.55, N 14.96.
Synthesis of (5S,aR)-1, (5R,aS)-1, (5R,aR)-2, and (5S,aS)-2. The
synthetic procedure described for racemates (ꢀ)-1 and (ꢀ)-2 was
also used to obtain the pure enantiomers (5S,aR)-1, (5R,aR)-2,
(5R,aS)-1 and (5S,aS)-2, starting from alkenes (R)-6[15] and (S)-6,[16]
respectively.
(5S,aR)-8 was crystallized from hexane/diisopropyl ether as white
prisms; mp: 63–658C; [a]2D0 =+313.1 (c=0.91; CHCl3); Anal. calcd
for C21H29N3O6: C 60.13, H 6.97, N 10.02, found: C 60.08, H 6.85, N
10.11.
(5R,aS)-8 was crystallized from hexane/diisopropyl ether as white
prisms; mp: 63–658C; [a]2D0 =ꢁ315.0 (c=1.00; CHCl3); Anal. calcd
for C21H29N3O6: C 60.13, H 6.97, N 10.02, found: C 60.23, H 6.94, N
10.00.
(5R,aR)-9 was obtained as a pale-yellow oil; [a]2D0 =ꢁ398.1 (c=
0.93; CHCl3); Anal. calcd for C21H29N3O6: C 60.13, H 6.97, N 10.02,
found: C 60.30, H 6.79, N 10.15.
(5S,aS)-9 was obtained as a pale-yellow oil; [a]2D0 =+395.1 (c=
0.95; CHCl3); Anal. calcd for C21H29N3O6: C 60.13, H 6.97, N 10.02,
found: C 60.21, H 6.83, N 9.89.
Ethyl (5R*,aR*)-5-(2-tert-butoxycarbonylamino-2-methoxycarbo-
nylethyl)-4,5-dihydro-1-phenyl-1H-pyrazole-3-carboxylate (ꢀ)-9
from (ꢀ)-11. Et3N (0.37 mL, 2.49 mmol) and a solution of di-tert-
butyl dicarbonate (0.47 g, 2.16 mmol) in CH2Cl2 (5 mL) were added
to a stirred solution of (ꢀ)-11 (0.53 g, 1.66 mmol) in CH2Cl2 (15 mL)
at 08C. The reaction mixture stirred at room temperature overnight
and was followed by TLC (EtOAc). The solvent was evaporated
under reduced pressure, and the crude material was purified by
column chromatography on silica gel (PE/EtOAc, 85:15) to give
(ꢀ)-9 (0.49 g, 70%) as a pale-yellow oil. Rf =0.28 (PE/EtOAc, 4:1);
1H NMR ([D6]DMSO, 1008C): d=1.28 (t, J=7.0, 3H); 1.40 (s, 9H);
1.74 (ddd, J=7.0, 9.7, 14.0, 1H); 2.24 (ddd, J=3.0, 6.7, 14.0, 1H);
2.97 (dd, J=5.5,17.9, 1H); 3.23 (dd, J=11.7, 17.9, 1H); 3.67 (s, 3H);
4.13 (ddd, J=6.7, 7.0, 7.0, 1H); 4.24 (q, J=7.0, 2H); 4.64 (dddd, J=
3.0, 5.5, 9.7, 11.7, 1H); 6.92 (d, J=7.3, 1H); 6.92 (d, J=7.0, 1H) 7.14
(d, J=8.8, 2H); 7.30 (dd, J=7.3, 8.8, 2H); 13C NMR ([D6]DMSO): d=
14.91, 28.83, 33.57, 37.57, 51.56, 52.76, 59.24, 61.13, 79.23, 114.89,
121.60, 130.03, 140.19, 142.55, 155.97, 162.75, 173.08; Anal. calcd
(5S,aR)-1 was crystallized from EtOH/Et2O as yellow prisms; mp:
>1928C (dec.); [a]2D0 =+334.7 (c=0.10; H2O); Anal. calcd for
C13H15N3O4: C 56.31, H 5.45, N 15.15, found: C 56.51, H 5.55, N
14.90.
(5R,aS)-1 was crystallized from EtOH/Et2O as yellow prisms; mp:
>1928C (dec.); [a]D20
C13H15N3O4: C 56.31, H 5.45, N 15.15, found: C 56.48, H 5.50, N
15.23.
=
ꢁ336.2 (c=0.11; H2O); Anal. calcd for
(5R,aR)-2 was crystallized from EtOH/Et2O as yellow prisms; mp:
>2188C (dec.); [a]2D0 =ꢁ394.0 (c=0.10; H2O); Anal. calcd for
C13H15N3O4: C 56.31, H 5.45, N 15.15, found: C 56.57, H 5.23, N
15.31.
(5S,aS)-2 was crystallized from EtOH/Et2O as yellow prisms; mp:
>2188C (dec.); [a]2D0 =+396.2 (c=0.10; H2O); Anal. calcd for
ChemMedChem 2010, 5, 1465 – 1475
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1471

P. Conti et al.
MED
C13H15N3O4: C 56.31, H 5.45, N 15.15, found: C 56.49, H 5.58, N
14.92.
methyl]-phosphonate 13[17] (2.61 g, 8.50 mmol) and solid NaHCO3
(2.40 g, 28.57 mmol) were added to a solution of (R)-6 (1.30 g,
5.70 mmol) in EtOAc (25 mL). The reaction mixture stirred vigorous-
ly for 24 h and was monitored by TLC (PE/EtOAc, 3:7). H2O (2 mL)
was added, and the organic layer was isolated and dried over an-
hydrous Na2SO4. Upon evaporation of the solvent, the crude mate-
rial was purified by flash chromatography on silica gel (PE/EtOAc,
2:3) to give pure (5R,aR)-15 (690 mg, 27%) and (5S,aR)-14 (1.02 g,
39%).
Ethyl (R)-5-(2-tert-butoxycarbonylamino-2-methoxycarbonyleth-
yl)-1-phenyl-1H-pyrazole-3-carboxylate [(R)-12]. PDC (6.71 g,
17.85 mmol) was added to a mixture of (5S,aR)-8 and (5R,aR)-9
(0.50 g, 1.19 mmol) in DMF (7 mL). The solution stirred at room
temperature for 4 h and the reaction was followed by TLC (PE/
EtOAc, 4:1). H2O (10 mL) was added, and the solution was extract-
ed with EtOAc (3ꢃ30 mL). The combined organic layers were
washed with brine, dried over anhydrous Na2SO4, and evaporated
under reduced pressure. The crude material was purified on silica
gel (PE/EtOAc, 4:1) to give (R)-12 (0.39 g, 80%), which was crystal-
lized from diisopropyl ether as white prisms; mp: >110 8C (dec.);
Rf =0.28 (PE/EtOAc, 4:1); [a]2D0 =ꢁ35.2 (c=0.22; CHCl3); 1H NMR
([D6]DMSO, 1008C): d=1.30 (t, J=7.0, 3H); 1.34 (s, 9H); 3.03 (dd,
J=9.0, 15.7, 1H); 3.13 (dd, J=5.4, 15.7, 1H); 3.58 (s, 3H); 4.32 (m,
1H); 4.35 (q, J=7.0, 2H); 6.75 (s, 1H); 6.92 (d, J=7.0, 1H); 7.42–
7.60 (m, 5H); 13C NMR ([D6]DMSO): d=14.90, 28.14, 28.73, 52.58,
52.80, 60.97, 79.50, 109.29, 126.41, 129.74, 130.10, 139.28, 142.04,
143.57, 155.98, 162.33, 172.23; Anal. calcd for C21H27N3O6: C 60.42,
H 6.52, N 10.07, found: C 60.69, H 6.63, N 10.18.
(5S,aR)-14 was obtained as a pale-yellow oil. Rf =0.26 (PE/EtOAc,
3:7); [a]2D0 =+149.5 (c=0.85; CHCl3); 1H NMR ([D6]DMSO, 1008C):
d=1.40 (s, 9H); 1.75 (ddd, J=4.4, 10.4, 14.8, 1H); 2.20 (dddd, J=
1.1, 2.9, 10.7, 14.8, 1H); 2.91 (ddd, J=1.4, 5.8, 17.6, 1H); 3.23 (ddd,
J=1,6, 11.3, 17.6, 1H); 3.61 (s, 3H); 3.74 (d, J=11.0, 1H); 3.75 (d,
J=11.0, 1H); 4.10 (ddd, J=4.4, 8.2, 10.7, 1H); 4.49 (dddd, J=2.9,
5.8, 10.4, 11.3, 1H); 6.90 (m, 1H); 6.92 (m, 1H); 7.16 (d, 2H); 7.27
(d, 2H); 13C NMR ([D6]DMSO): d=28.82, 31.69, 39.79 (d, JC–P
=
19 Hz), 50.72, 52.69, 53.72, 53.80, 57.39, 79.28, 114.81, 121.24,
129.87, 139.24 (d, JC–P =231 Hz), 142.80, 156.38, 173.03; 31P NMR
(CDCl3): d=12.89; Anal. calcd for C20H30N3O7P: C 52.74, H 6.64, N
9.23, found: C 52.90, H 6.72, N 9.17.
(R)-5-(2-Amino-2-carboxyethyl)-1-phenyl-1H-pyrazole-3-carboxyl-
ic acid [(R)-3]. A) Compound (R)-12 (0.39 g, 0.95 mmol) was dis-
solved in MeOH (10 mL) and treated with 1n aqueous NaOH
(2.85 mL). Disappearance of the starting material was monitored
by TLC (PE/EtOAc, 7:3). Upon evaporation of MeOH, the aqueous
layer was washed with Et2O, acidified with 2n aqueous HCl, and
extracted with EtOAc. The organic layer was dried over anhydrous
Na2SO4 and, upon evaporation of the solvent, the diacid derivative
(0.32 g, yield 90%) was obtained as a white solid. B) The diacid
(0.32 g, 0.85 mmol) was treated with a 30% CH2Cl2 solution of TFA
(2.2 mL) at 08C. The solution stirred at room temperature for 3 h,
and the reaction was followed by TLC (CHCl3/MeOH, 9:1 + 1%
AcOH). Volatiles were removed under reduced pressure, and the
residue was crystallized from EtOH/Et2O to give (R)-3 (0.14 g, 73%)
as white prisms; mp: >2308C (dec.); Rf =0.20 (BuOH/H2O/AcOH
(5R,aR)-15 was obtained as a pale-yellow oil. Rf =0.28 (PE/EtOAc,
3:7); [a]2D0 =ꢁ195.0 (c=0.98; CHCl3); 1H NMR ([D6]DMSO, 1008C):
d=1.40 (s, 9H); 1.68 (ddd, J=7.0, 10.0, 14.0, 1H); 2.23 (dddd, J=
1.1, 3.0, 7.0, 14.0, 1H); 2.94 (ddd, J=1.4, 5.5, 17.6, 1H); 3.23 (ddd,
J=1,6, 11.3, 17.6, 1H); 3.67 (s, 3H); 3.74 (d, J=11.3, 1H); 3.75 (d,
J=11.3, 1H); 4.12 (ddd, J=7.0, 1H); 4.52 (dddd, J=3.0, 5.5, 10.0,
11.3, 1H); 6.90 (t, 1H); 6.92 (m, 1H); 7.10 (d, 2H), 7.28 (d, 2H);
13C NMR ([D6]DMSO): d 28.82, 33.56, 39.57 (d, JC–P =19 Hz) 51.62,
52.76, 53.72, 53.79, 57.92, 79.25, 114.68, 121.31, 130.00, 140.02 (d,
J
C–P =231 Hz), 142.96, 156.01, 173.07; 31P NMR (CDCl3): d=12.86;
Anal. calcd for C20H30N3O7P: C 52.74, H 6.64, N 9.23, found: C 52.86,
H 6.75, N 9.08.
(5S,aR)-5-(2-Amino-2-carboxyethyl)-4,5-dihydro-1-phenyl-1H-pyr-
azole-3-phosphonic acid [(5S,aR)-4]. A) Compound (5S,aR)-14
(1.02 g, 2.24 mmol) was dissolved in MeOH (5 mL) and treated with
1n aqueous NaOH (5 mL). Disappearance of the starting material
was monitored by TLC (PE/EtOAc, 7:3). Upon evaporation of the
MeOH, the aqueous layer was washed with Et2O, acidified with 2n
aqueous HCl, and extracted with EtOAc. The organic layer was
dried over anhydrous Na2SO4, and, upon evaporation of the sol-
vent, the carboxylic acid intermediate was obtained as a colorless
oil. B) Bromotrimethylsilane (TMSBr; 3 mL, 22.5 mmol) was added
to a solution of the carboxylic acid derivative in CH2Cl2 (10 mL).
The reaction stirred at room temperature overnight. The solvent
was evaporated under reduced pressure, and MeOH (5 mL) was
added. The reaction mixture stirred for 1 h, followed by evapora-
tion of the solvent to give pure (5S,aR)-4 (480 mg, 67%) as a white
hygroscopic powder; mp: >2108C (dec.); Rf =0.34 (BuOH/H2O/
4:2:1); [a]2D0 =+3.7 (c=0.10; H2O); 1H NMR (D2O
+ 1 drop
CF3COOD): d=3.13 (dd, J=6.8, 16.0, 1H); 3.23 (dd, J=6.8, 16.0,
1H); 3.90 (t, J=6.8, 1H); 6.76 (s, 1H); 7.22 (m, 2H); 7.35 (m, 3H);
13C NMR (D2O + 1 drop CF3COOD): d=26.21, 51.53, 110.02, 126.03,
129.90, 130.23, 137.35, 139.36, 143.50, 164.88, 170.12; Anal. calcd
for C13H13N3O4: C 56.72, H 4.76, N 15.27, found: C 56.52, H 4.64, N
15.35.
(S)-5-(2-Amino-2-carboxyethyl)-1-phenyl-1H-pyrazole-3-carboxyl-
ic acid [(S)-3]. The same procedure described for the synthesis of
(R)-3 was used to prepare (S)-3, starting from a mixture of the two
diastereomeric pyrazolines (5R,aS)-8 and (5S,aS)-9.
(S)-12 was crystallized from diisopropyl ether as white prisms; mp:
>110 8C (dec.); [a]2D0 =+34.5 (c=0.25; CHCl3); Anal. calcd for
C21H27N3O6: C 60.42, H 6.52, N 10.07, found: C 60.52, H 6.44, N
10.00.
1
AcOH 4:2:1); [a]2D0 =+116.5 (c=0.13; H2O); H NMR (D2O): d=1.90–
2.10 (m, 2H); 2.80 (dd, J=4.7, 17.9, 1H); 3.16 (dd, J=10.7, 17.9,
1H); 3.92 (dd, J=6.0, 7.7, 1H); 4.42 (dddd, J=4.7, 5.0, 10.0, 10.7,
1H); 6.84 (t, J=7.7, 1H); 7.01 (d, J=7.7, 2H); 7.20 (t, J=7.7, 2H);
(S)-3 was crystallized from EtOH/Et2O as white prisms; mp:
>2308C (dec.); [a]2D0 =ꢁ4.4 (c=0.10; H2O); Anal. calcd for
C13H13N3O4: C 56.72, H 4.76, N 15.27, found: C 56.89, H 4.71, N
15.44.
13C NMR (D2O): d=31.92, 39.85 (d,
JC–P =20 Hz), 50.16, 57.28,
115.65, 121.68, 129.58, 143.26, 148.12 (d, JC–P =215 Hz), 171.42;
31P NMR (D2O): d=4.80; Anal. calcd for C12H16N3O5P: C 46.01, H
5.15, N 13.41, found: C 45.95, H 5.23, N 13.26.
Dimethyl (5S,aR)-5-(2-tert-butoxycarbonylamino-2-methoxycar-
bonylethyl)-4,5-dihydro-1-phenyl-1H-pyrazole-3-phosphonate
[(5S,aR)-14] and dimethyl (5R,aR)-5-(2-tert-butoxycarbonylami-
no-2-methoxycarbonylethyl)-4,5-dihydro-1-phenyl-1H-pyrazole-
3-phosphonate [(5R,aR)-15]. Dimethyl [bromo(phenylhydrazono)-
(5R,aR)-5-(2-Amino-2-carboxyethyl)-4,5-dihydro-1-phenyl-1H-
pyrazole-3-phosphonic acid [(5R,aR)-5]. A) Compound (5R,aR)-15
(690 mg, 1.52 mmol) was dissolved in MeOH (5 mL) and treated
with 1n aqueous NaOH (5 mL). Disappearance of the starting ma-
1472
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1465 – 1475

NMDA Receptor Antagonists
terial was monitored by TLC (PE/EtOAc, 7:3). Upon evaporation of
the MeOH, the aqueous layer was washed with Et2O, acidified with
2n aqueous HCl, and extracted with EtOAc. The organic layer was
dried over anhydrous Na2SO4 and, upon evaporation of the solvent,
the carboxylic acid intermediate was obtained as a colorless oil.
B) TMSBr (2 mL, 15.0 mmol) was added to a solution of the carbox-
ylic acid derivative in CH2Cl2 (10 mL). The reaction stirred at room
temperature overnight. The solvent was evaporated under reduced
pressure, and MeOH (5 mL) was added. The reaction mixture
stirred for 1 h, followed by evaporation of the solvent to give pure
(5R,aR)-5 (320 mg, 66%) as a white hygroscopic powder; mp:
>2158C (dec.); Rf =0.37 (BuOH/H2O/AcOH 4:2:1); [a]2D0 =ꢁ219.2
of the diffraction frames was performed using the SAINT-NT pro-
gram (SAINT v. 7.23A User’s Manual (2003), Bruker AXS Inc., Madi-
son, WI, USA), while the SORTAV program was employed to analyze
and merge the raw data.[24,25] Due to the low absorption coeffi-
cient, an absorption correction was deemed unnecessary. Notably,
due to the small dimensions and poor scattering ability of the
sample, the majority (~80%) of the final, merged, 5310 independ-
ent data was weak, with I<2s(I).
Structure solution and refinement
1
(c=0.11; H2O); H NMR (D2O): d=1.70 (ddd, J=4.0, 10.7, 14,5, 1H);
The structure was solved using direct methods and was subse-
quently refined against the final set of 5310 independent reflec-
tions. Due to the lack of anomalous scattering in the title com-
pound, the absolute configuration of the C7 stereogenic center (R)
was inferred, taking into consideration the fact that the stereo-
chemistry of the C14 carbon (S) was precisely known and the title
compound crystallized in a non-centrosymmetric space group. The
program SHELX97[26] was used throughout as a component of the
WinGX program package.[27] Positions of the hydrogen atoms were
determined by applying suitable riding motion constraints. The ter-
minal methoxy moiety was determined to be disordered between
two accessible conformations, with estimated occupation probabil-
ities as high as 70 and 30%, respectively. The thermal motion of C
and O atoms belonging to disordered chains was considered iso-
tropic. A list of relevant final least-squares parameters follows
(more information can be found in the deposited Crystallographic
Information File): R(F)=0.0941 (for 1093 independent reflections;
R(F)=0.3470 (all independent data); wR(F2)=0.2970 (all independ-
ent data); goodness-of-fit: 0.831; residual Fourier density in the
unit cell, D1max/min =+0.22 (4)/ꢁ0.16 (4) eꢄꢁ3. CCDC 781612 [(ꢁ)-
(5R,aS)-8] contains the supplementary crystallographic data for this
paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif.
2.16 (ddd, J=2.2, 11.0, 14.5, 1H); 2.83 (ddd, J=1.4, 4.7, 17.6, 1H);
3.15 (ddd, J=1.4, 11.0, 17.6, 1H); 3.91 (dd, J=4.0, 11.0, 1H); 4.52
(dddd, J=2.2, 4.7, 10.7, 11.0, 1H); 6.83 (t, J=7.4, 1H); 6.98 (d, J=
7.7, 2H); 7.18 (dd, J=7.4, 7.7, 2H); 13C NMR (D2O): d=31.26, 39.57
(d, JC–P =19 Hz), 49.57, 57.55, 115.22, 121.45, 129.77, 143.35, 148.62
(d, JC–P =214 Hz), 171.60; 31P NMR (D2O): d=5.16; Anal. calcd for
C12H16N3O5P: C 46.01, H 5.15, N 13.41, found: C 46.29, H 4.97, N
13.36.
Synthesis of (5R,aS)-4 and (5S,aS)-5. The synthetic procedures de-
scribed for (5S,aR)-4 and (5R,aR)-5 were also used for synthesis of
enantiomers (5R,aS)-4 and (5S,aS)-5, starting from alkene (S)-6.
(5R,aS)-14 was obtained as a pale-yellow oil; [a]2D0 =ꢁ149.0 (c=
0.90; CHCl3); Anal. calcd for C20H30N3O7P: C 52.74, H 6.64, N 9.23,
found: C 52.61, H 6.84, N 9.19.
(5R,aS)-4 was obtained as a white hygroscopic powder; mp:
>2108C (dec.); [a]2D0 =ꢁ115.3 (c=0.11; H2O); Anal. calcd for
C12H16N3O5P: C 46.01, H 5.15, N 13.41, found: C 46.31, H 5.33, N
13.27.
(5S,aS)-15 was obtained as a pale-yellow oil; [a]2D0 =+194.5 (c=
0.95; CHCl3); Anal. calcd for C20H30N3O7P: C 52.74, H 6.64, N 9.23,
found: C 52.90, H 6.48, N 9.11.
(5S,aS)-5 was obtained as a white hygroscopic powder; mp:
>2158C (dec.); [a]2D0 =+220.5 (c=0.10; H2O); Anal. calcd for
C12H16N3O5P: C 46.01, H 5.15, N 13.41, found: C 46.18, H 5.24, N
13.53.
Pharmacology
iGlu receptor binding assays
All binding assays were performed using rat brain synaptic mem-
branes from male Sprague–Dawley rats, and tissue preparations
were prepared as previously described.[28] Affinities for NMDA,[18]
AMPA,[19] and KA[20] receptors were determined using 2 nm
X-ray analysis of ethyl (5R,aS)-5-(2-tert-butoxycarbonylamino-
2-methoxycarbonylethyl)-4,5-dihydro-1-phenyl-1H-pyrazole-3-
carboxylate [(ꢁ)-(5R,aS)-8]
Crystals suitable for X-ray diffraction studies were grown by slow
evaporation at 48C from a 1:1 mixture of diisopropyl ether and
hexane. A very thin crystal (0.200ꢃ0.075ꢃ0.025 mm3) was chosen
for an experiment directed toward solving the structure.
[3H]CGP39653
(50 Cimmolꢁ1
,
Kd =6 nm),
5 nm
[3H]AMPA
(45.5 Cimmolꢁ1), and 5 nm [3H]KA (58 Cimmolꢁ1), respectively, with
minor modifications as previously described.[29] Nonspecific binding
was determined using 1 mm (S)-Glu.
Crystal data
Electrophysiology of NMDA receptors in Xenopus oocytes
Title compound: C21H29N3O6, Mr =419.5 Da; orthorhombic; space
group: P212121 (No 19); unit cell (ꢄ, ꢄ3): a=5.4027 (11), b=19.2240
(38), c=22.3810 (45), V=2325 (1); Z: 4; Dc: 1.199 gcmꢁ3; F(000):
896; m: 0.09 mmꢁ1
For expression of NMDA receptors, cRNA for rat GluN1—1A,
GluN2A, GluN2B, GluN2C, and GluN2D were transcribed in vitro
and injected into Xenopus oocytes as previously described.[12] Re-
cordings were performed 2–4 days after injection, using whole-cell
two-electrode voltage clamp from ꢁ40 mV to ꢁ80 mV in extracel-
lular solution containing 115mM NaCl, 2.5mM KCl, 1.9mM BaCl2,
and 10mM HEPES (pH 7.6). During recording, 10 mm glycine was in-
cluded in all agonist and/or antagonist applications. Electrophysio-
logical data were acquired and processed as previously de-
scribed.[12]
.
Data collection and reduction
A total of 34927 reflections within an entire sphere (below; sinq/
lmax =0.65 ꢄꢁ1) were collected using a graphite-monochromated
Mo Ka radiation source (l=0.71073 ꢄ) on a Bruker AXS Smart Apex
diffractometer, equipped with an Apex II CCD detector. Integration
ChemMedChem 2010, 5, 1465 – 1475
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1473

P. Conti et al.
MED
determined for control cultures not exposed to OGD and was sub-
tracted from all experimental values.
Determination of mGluR activity
The mGluR subtypes mGluR1A, mGluR2 and mGluR4A were ex-
pressed in Chinese hamster ovary (CHO) cell lines. Cells were main-
tained and tested as previously described by measurement of in-
tracellular calcium (mGluR1) or cyclic AMP (mGluR2, 4) levels.[21]
Compounds (ꢀ)-1 and (ꢀ)-2 were evaluated at concentrations up
to 1 mm.
In vivo pharmacology
DBA/2 mice (8–12 g, 22–25 days old) were purchased from Charles
River (Calco, Como, Italy). Procedures involving animals and their
care were conducted in conformity with international and national
laws and policies (European Communities Council Directive of No-
vember 24, 1986, 86/609EEC). Groups of 10 mice of either gender
were exposed to auditory stimulation for 15, 30, 60, 90 and
120 min following administration of vehicle or each dose of drugs.
Compounds were administered i.p. (0.1 mL per 10 g body weight)
as a freshly prepared solution in 50% DMSO and 50% sterile 0.9%
NaCl solution. This latter protocol has been validated several times
for systemic and focal intracerebral microinjections for in vitro elec-
trophysiological studies.[30] Individual mice were placed under a
hemispheric perspex dome (1: 58 cm); 60 s was allowed for habit-
uation and assessment of locomotor activity. Auditory stimulation
(12–16 kHz, 109 dB) was applied for 60 s, or until tonic extension
occurred and induced a sequential seizure response in control
DBA/2 mice, consisting of an early wild running phase, followed by
generalized myoclonus and tonic flexion and extension, sometimes
followed by respiratory arrest.[31] Behavioral changes were observed
and recorded during the period between drug administration and
auditory testing. The ED50 values for each phase of the audiogenic
seizure were determined for each dose of compound administered,
and dose–response curves were fitted using a computer method
of Litchfield and Wilcoxon.[32]
Oxygen-glucose deprivation (OGD) and NMDA neurotoxicity
in cortical cell cultures
Primary cultures of mixed cortical cells containing both neuronal
and glial elements were prepared as previously described.[23] Briefly,
cerebral cortices were dissected from fetal mice at 14–15 days of
gestation, minced using medium stock [MS, composed of Eagle’s
minimal essential medium (MEM, with Earle’s salts, glutamine- and
NaHCO3-free) supplemented with 38 mm NaHCO3, 22 mm glucose,
100 UmLꢁ1 penicillin, and 100 mgmLꢁ1 streptomycin], and incubat-
ed for 10 min at 378C in MS with 0.25% trypsin and 0.05% DNAse.
Enzymatic digestion was terminated by a second incubation
(10 min at 378C) in MS, supplemented with 10% heat-inactivated
horse serum and 10% fetal bovine serum, after which the cells
were mechanically disrupted and counted. After brief centrifuga-
tion, cells were resuspended (approximately 4ꢃ105 cellsmLꢁ1) and
plated in 15 mm multi-well vessels on a layer of confluent astro-
cytes using a plating medium of MS supplemented with 10% heat-
inactivated horse serum, 10% fetal bovine serum, and 2 mm gluta-
mine. Cultures were stored at 378C with 100% humidity and a
95% air/5% CO2 atmosphere. Following 4–5 days in vitro (DIV),
non-neuronal cell division was halted by the application of 3 mm
cytosine arabinoside for 24 h. Cultures were then shifted to a main-
tenance medium, identical to the plating medium but lacking fetal
bovine serum, which was partially replaced twice weekly. Experi-
ments were performed using mature cultures (14–15 DIV).
Acknowledgements
This work was supported by the Italian Ministry of Education
(MIUR-PRIN 2007), the GluTarget Center of Excellence, and the
Carlsberg Foundation. We gratefully acknowledge Dr. L. Lo Presti
for X-ray crystallographic analysis.
OGD was induced as previously described.[23] Briefly, the culture
medium was replaced by thorough exchange with a glucose-free
balanced salt solution (116mM NaCl, 5.4mM KCl, 0.8mM MgSO4,
1mM NaH2PO4, 26mM NaHCO3, 1.8mM CaCl2, and 10 mgmLꢁ1
phenol red), previously saturated with 95% N2/5% CO2 and heated
at 378C. Multi-well plates were then placed in an airtight incuba-
tion chamber, equipped with inlet and outlet valves, and 95% N2/
5% CO2 was blown through the chamber for 10 min to ensure
maximal removal of oxygen. The chamber was then sealed and in-
cubated at 378C for 60 min. OGD was terminated by removing the
cultures from the chamber, replacing the exposure solution with
oxygenated MS, and returning the multi-well plates to the incuba-
tor under normoxic conditions. NMDA neurotoxicity was obtained
by exposing cultures to 300 mm NMDA for 10 min (mild insult).
During this exposure, the original culture medium was replaced by
thorough exchange with a HEPES controlled salt solution (HCSS:
120mM NaCl, 5.4mM KCl, 0.8mM MgCl2, 1.8mM CaCl2, 20mM
HEPES, 15mM glucose, 10mM NaOH and 10 mgmLꢁ1 phenol red).
Upon exposure to NMDA, cultures were rinsed with HCSS, original
medium was restored, and cells were stored at 378C in 100% hu-
midity and a 95% air/5% CO2 atmosphere for 24 h, followed by as-
sessment of the extent of neuronal death. To achieve maximal neu-
ronal injury, the cultures were exposed for 24 h to 1 mm Glu in MS
at 378C, with 100% humidity and a 95% air/5% CO2 atmosphere.
Cell damage was measured by the amount of lactate dehydrogen-
ase (LDH) released from injured cells into the extracellular fluid
24 h after exposure to OGD or Glu.[23] Background LDH release was
Keywords: 1,3-dipolar cycloaddition · anticonvulsant activity ·
neuroprotection · NMDA receptor antagonists · pyrazolines
[1] H. V. Wheal, A. M. Thomson (Eds.), Excitatory Amino Acids and Synaptic
Transmission, Academic Press, London (UK), 1995.
[2] D. T. Monaghan, R. J. Wenthold (Eds.), The Ionotropic Glutamate Recep-
tors, Humana Press, Totowa, NJ (USA), 1997.
[3] P. J. Conn, J. Patel (Eds.), The Metabotropic Glutamate Receptors,
Humana Press, Totowa, NJ (USA), 1994.
[4] F. Moroni, F. Nicoletti, D. E. Pellegrini-Giampietro (Eds.), Metabotropic
Glutamate Receptors and Brain Function, Portland Press, London (UK),
1998.
[5] D. W. Choi, J. Neurobiol. 1992, 23, 1261–1276.
[6] C. G. Parsons, W. Danysz, G. Quack, Drug News Perspect. 1998, 11, 523–
569.
[7] D. Bowie, CNS Neurol. Disord. Drug Targets 2008, 7, 129–143.
[8] E. A. Waxman, D. R. Lynch, Neuroscientist 2005, 11, 37–49.
[9] P. Paoletti, J. Neyton, Curr. Opinion Pharmacol. 2007, 7, 39–47.
[10] G. Johnson, P. L. Ornstein, Curr. Pharm. Des. 1996, 2, 331–356.
[11] P. Conti, M. De Amici, G. Grazioso, G. Roda, F. Barberis Negra, B. Nielsen,
T. B. Stensbøl, U. Madsen, H. Brꢀuner-Osborne, K. Frydenvang, G. De Sar-
ro, L. Toma, C. De Micheli, J. Med. Chem. 2004, 47, 6740–6748.
[12] P. Conti, M. De Amici, G. Grazioso, G. Roda, A. Pinto, K. B. Hansen, B.
Nielsen, U. Madsen, H. Brꢀuner-Osborne, J. Egebjerg, V. Vestri, D. E. Pel-
legrini-Giampietro, P. Sibille, F. C. Acher, C. De Micheli, J. Med. Chem.
2005, 48, 6315–6325.
1474
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1465 – 1475

NMDA Receptor Antagonists
[13] P. Conti, S. Joppolo di Ventimiglia, A. Pinto, L. Tamborini, F. S. Menniti,
J. T. Lazzaro, C. De Micheli, Chem. Biodiversity 2008, 5, 657–663.
[14] L. Kinarsky, B. Feng, D. A. Skifter, R. M. Morley, S. Sherman, D. E. Jane,
D. T. Monaghan, J. Pharmacol. Exp. Ther. 2005, 313, 1066–1074.
[15] G. F. Busscher, F. P. J. T. Rutjes, F. L. Van Delft, Tetrahedron Lett. 2004, 45,
3629–3632.
[24] R. H. Blessing, Crystallogr. Rev. 1987, 1, 3–58.
[25] R. H. Blessing, D. A. Langs, J. Appl. Crystallogr. 1987, 20, 427–428.
[26] G. M. Sheldrick, Acta Crystallogr. 2008, A64, 112–122.
[27] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837–838.
[28] R. W. Ransom, N. L. Stec, J. Neurochem. 1988, 51, 830–836.
[29] R. P. Clausen, K. B. Hansen, P. Cali, B. Nielsen, J. R. Greenwood, M. Beg-
trup, J. Egebjerg, H. Brꢀuner-Osborne, Eur. J. Pharmacol. 2004, 499, 35–
44.
[16] S. Collet, P. Bauchat, R. Danion-Bougot, D. Danion, Tetrahedron: Asym-
metry 1998, 9, 2121–2131.
[17] P. Conti, A. Pinto, L. Tamborini, V. Rizzo, C. De Micheli, Tetrahedron 2007,
63, 5554–5560.
[30] a) E. Russo, A. Costanti, G. Ferreri, R. Citraro, G. De Sarro, Neuropharma-
cology 2004, 46, 865–878; b) E. D. Di Paola, P. Gareri, A. Davoli, S. Grat-
teri, F. Scicchitano, C. Naccari, G. De Sarro, Epilepsy Res. 2007, 75, 112–
121; c) R. Gitto, R. Caruso, B. Pagano, L. De Luca, R. Citraro, E. Russo, G.
De Sarro, A. Chimirri, J. Med. Chem. 2006, 49, 5618–5622.
[31] G. De Sarro, M. J. Croucher, B. S. Meldrum, Neuropharmacology 1984, 23,
526–530.
[18] M. A. Sills, G. Fagg, M. Pozza, C. Angst, D. E. Brundish, S. D. Hurt, E. J.
Wilusz, M. Williams, Eur. J. Pharmacol. 1991, 192, 19–24.
[19] T. Honorꢅ, M. Nielsen, Neurosci. Lett. 1985, 54, 27–32.
[20] D. J. Braitman, J. T. Coyle, Neuropharmacology 1987, 26, 1247–1251.
[21] E. J. Bjerrum, A. S. Kristensen, D. S. Pickering, J. R. Greenwood, B. Niel-
sen, T. Liljefors, A. Schousboe, H. Brꢀuner-Osborne, U. Madsen, J. Med.
Chem. 2003, 46, 2246–2249.
[32] J. Litchfield, F. Wilcoxon, J. Pharmacol. Exp. Ther. 1949, 96, 99–113.
[22] K. B. Hansen, H. Brꢀuner-Osborne, J. Egebjerg, Comb. Chem. High
Throughput Screening 2008, 11, 304–315.
Received: April 30, 2010
Revised: June 23, 2010
[23] D. E. Pellegrini-Giampietro, A. Cozzi, F. Peruginelli, P. Leonardi, E. Meli, R.
Pellicciari, F. Moroni, Eur. J. Neurosci. 1999, 11, 3637–3647.
Published online on July 27, 2010
ChemMedChem 2010, 5, 1465 – 1475
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1475