Rational Design, Synthesis and Pharmacological Evaluation of the (2R)- and (2S)-Stereoisomers of 3-(2-Carboxypyrrolidinyl)-2-methyl Acetic Acid as Ligands for the Ionotropic Glutamate Receptors

Article abstract of DOI:10.1002/cmdc.201000543

In this paper we describe the rational design, synthesis and pharmacological evaluation of two new stereoisomeric (S)-glutamate (Glu) analogues. The rational design was based on hybrid structures of the natural product kainic acid, a synthetic analogue CPAA and the high-affinity Glu analogue SYM2081. Pharmacological evaluation of the two stereoisomers revealed that one stereoisomer showed a subtype selectivity profile with low micromolar affinity for GluK1 and GluK3 and a 10- to 15-fold lower affinity for GluK2. The other stereoisomer displayed full selectivity for the KA over AMPA and NMDA receptors (GluK1-3: 0.39, 0.51 and 0.099μM, respectively).

Full text of DOI:10.1002/cmdc.201000543

MED
DOI: 10.1002/cmdc.201000543
Rational Design, Synthesis and Pharmacological Evaluation
of the (2R)- and (2S)-Stereoisomers of 3-(2-Carboxy-
pyrrolidinyl)-2-methyl Acetic Acid as Ligands for the
Ionotropic Glutamate Receptors
Julie L. Rasmussen,[a] Morten Storgaard,[a] Darryl S. Pickering,[b] and Lennart Bunch*[a]
In this paper we describe the rational design, synthesis and
pharmacological evaluation of two new stereoisomeric (S)-glu-
tamate (Glu) analogues. The rational design was based on
hybrid structures of the natural product kainic acid, a synthetic
analogue CPAA and the high-affinity Glu analogue SYM2081.
Pharmacological evaluation of the two stereoisomers revealed
that one stereoisomer showed a subtype selectivity profile
with low micromolar affinity for GluK1 and GluK3 and a 10- to
15-fold lower affinity for GluK2. The other stereoisomer dis-
played full selectivity for the KA over AMPA and NMDA recep-
tors (GluK1–3: 0.39, 0.51 and 0.099mm, respectively).
Introduction
In the mammalian central nervous system (CNS), (S)-glutamate
(Glu) functions as the major excitatory neurotransmitter. Once
released from the pre-synaptic neuron into the glutamatergic
synapse, Glu activates a number of pre- and post-synaptic glu-
tamate receptors. On the basis of the pharmacological profile
and ligand selectivity studies, the Glu receptors have been di-
vided into two main classes: the fast acting ionotropic recep-
tors (iGluRs),[1] and the G protein-coupled metabotropic recep-
tors (mGluRs), which produce a much slower signal transduc-
tion through second messenger systems.[2] Uptake of Glu is
mediated by action of the excitatory amino acid transporters
(EAAT), previously termed the Glu transporters.[3] Furthermore,
the iGluRs have been divided into three groups on the basis of
ligand selectivity studies (new subunit nomenclature):[4] the a-
amino-3-hydroxy-5-methyl-4-isoxazole-propionate (AMPA) re-
ceptors comprising subunits GluA1—4, the kainic acid (KA) re-
ceptors comprising subunits GluK1–5, and the N-methyl-d-as-
partate (NMDA) receptors comprising subunits GluN1–3. Func-
tional iGluRs are tetrameric in structure comprising four subu-
nits with the ion channel pore formed in the center of the re-
ceptor.[5] For the KA receptors, subunits GluK1–3 may form
functional homomeric[6] or heteromeric[7] receptors, whereas
subunits GluK4 and GluK5[8] only form functional receptors in
combination with subunits GluK1, GluK2 or GluK3. Although
only one glutamate receptor-based drug is on the market
(memantine for the treatment of Alzheimer’s disease),[9] the
glutamatergic neurotransmitter system is believed to hold
great promise as it is involved in important neuro-physiological
processes, such as memory and learning, motor function, and
neural plasticity[10] and development.[11] Thus, psychiatric dis-
eases or disorders such as depression,[12] anxiety,[13] addic-
tion,[14] migraine,[15] and schizophrenia[16] may be directly relat-
ed to disordered glutamatergic neurotransmission.[17] More-
over, excessive Glu signaling is neurotoxic and will result in
neuronal death.[18] On this basis, it has been suggested that
neurodegenerative diseases such as Alzheimer’s, Huntington’s,
amyotrophic lateral sclerosis (ALS), cerebral stroke[16] and epi-
lepsy[19] may be the result of a malfunctioning glutamatergic
neurotransmitter system.
The usage of subtype-selective ligands is a valuable strategy
to investigate the physiological role and function of a given
Glu receptor subtype. However, despite an extensive effort, the
discovery of such compounds has proven a major challenge as
the orthosteric ligand binding pockets within an iGluR group
display high homology. In summary for KA receptor agonists,[20]
it has been shown that introduction of a substituent in the
syn-2,4-position of Glu provides highly potent and selective
homomeric GluK1 agonists.[21–23] Likewise, potent and selective
GluK1 antagonists[24] have been discovered for the homomeric
GluK1 subtype. In general, iGluR antagonism is achieved by in-
creasing the distance between the a-amino acid moiety and
the distal carboxylate group as exemplified by the recently
published KA antagonist (2S,3R)-3-(3-carboxyphenyl)-pyrroli-
dine-2-carboxylic acid.[25] Herein, we describe the rational
design and synthesis a new class of KA receptor ligands 1a/
1b, which are hybrid structures of reported KA ligands: KA,
CPAA and SYM2081. Furthermore, the pharmacological profile
[a] J. L. Rasmussen, Dr. M. Storgaard, Prof. L. Bunch
Department of Medicinal Chemistry
Faculty of Pharmaceutical Sciences, University of Copenhagen
Universitetsparken 2, 2100 Copenhagen (Denmark)
Fax: (+45)353-36041
E-mail: lebu@farma.ku.dk
[b] Prof. D. S. Pickering
Department of Pharmacology and Pharmacotherapy
Faculty of Pharmaceutical Sciences, University of Copenhagen
Universitetsparken 2, 2100 Copenhagen (Denmark)
498
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 498 – 504

Ionotropic Glutamate Receptor Ligands
of two C2 epimers 1a/1b are investigated at native iGluR re-
ceptors and homomeric KA receptor subtypes GluK1–3.
timal binding conformation is only one parameter affecting its
binding affinity. Steric clashes together with favorable interac-
tions between a ligand and a receptor play essential roles, but
also the ability of the protein to adopt a favorable conforma-
tion induced by the nature of the ligand.[28] Furthermore, the
ability of the ligand to disrupt the organized water matrix in
the receptor ligand binding domain (LBD) and exclude water
molecules on domain closure adds positively to the overall
binding (negative DG) by increasing the entropic term (DS).[29]
Using static modeling techniques, the dynamics of the protein
binding pocket, as well as the DS term of the water disruption
process, cannot be calculated, rendering essential information
outside of reach when attempting to estimate the binding af-
finity of a ligand in silico.[30]
Results and Discussion
For the Glu receptors, it has been shown from X-ray crystal
structures that Glu binds to and activates the iGluRs in a
folded conformation, whereas an extended form is observed
as the binding mode when activating the class of mGluRs
(Figure 1).[26] Furthermore, this observation has been supported
by the pharmacological evaluation of a number of conforma-
tionally restricted analogues of Glu that either mimic the
folded or the extended conformation.[26]
KA, CPAA and SYM2081 (Table 1) were submitted to an in
silico stochastic conformational search to determine their low-
energy conformations. The calculated conformations all resem-
bled the folded Glu conformation characterized as illustrated
in Figure 1. Furthermore, the calculated low-energy conforma-
tions of KA and SYM2081 were similar to conformations ob-
served when crystallized with the LBD (KA: GluA2, PDB:
1FTK,[31] and GluK2, PDB: 1TT1;[32] SYM2081: GluK2, PDB:
1SD3[32]). A three-point superimposition of calculated low-
energy conformations of KA, CPAA and SYM2081 (Figure 2a)
clearly shows that their conformations are highly similar and
perfectly mimic the folded Glu conformation. However, on
comparison with binding affinity data (Table 1) the three com-
pounds are indeed distinct. Whereas KA displays 7 nm binding
affinity for native KA receptors (subunits GluK4/5) deletion of
the iso-propenyl group (CPAA) results in a 200-fold loss of
binding affinity (1.6 mm). For KA, it has been suggested that a
p–p interaction of the iso-propylene group with residue
Tyr705 (GluK1 numbering) plays an important role for the
high-affinity binding at the KA receptors.[33] As this structural
feature is not present in CPAA and together with the fact that
Figure 1. Chemical structure of (S)-glutamic acid (Glu) and illustration of the
folded Glu-conformation and the extended Glu conformation.
However, it is noteworthy that an optimal low-energy con-
formation does not warrant high binding affinity as exempli-
fied by the highly conformationally restricted Glu analogue
DCAN (Table 1).[27] Even though this compound perfectly
mimics the folded Glu conformation, it is without affinity for
any of the iGluRs. In fact, the ability of a ligand to adopt an op-
Table 1. Chemical structures and binding affinities of selected Glu ligands at native AMPA, KA and NMDA receptors (rat brain synaptosomes) and at cloned homo-
meric rat GluK1–3.
Compd
Native receptors
KA
Cloned homomeric receptors
AMPA
NMDA
GluA2
GluK1
GluK2
GluK3
IC50 [mm]
IC50 [mm]
Ki [mm]
Ki [mm]
Ki [mm]
Ki [mm]
Ki [mm]
SYM2081[a]
KA[b]
26.6
4
1100
106
>300[i]
>100[f]
9.0
0.032
0.007
6.0
5.9
>100[c]
350
0.5
>300[i]
4.10ꢀ0.27
3.69ꢀ0.98
0.000663ꢀ0.000035
0.0759ꢀ0.0122
8.18ꢀ0.29
0.017ꢀ0.003
0.0127ꢀ0.0014
12.3ꢀ1.3
–
0.0057ꢀ0.0011
0.0328ꢀ0.0049
0.376ꢀ0.060
–
DHK[c]
>1000
CPAA[d]
DCAN[e]
BOAD
1.6
–
>1000
>100
–
>160[i]
14[f]
>100
>100
>100
2.9[f]
–
0.667ꢀ0.100
2.086ꢀ0.072
0.625ꢀ0.089
DOMO[g]
0.015
1.99ꢀ0.23
0.0011ꢀ0.0002
0.00604ꢀ0.00104
0.00384ꢀ0.00032
[a] References [21,34]. [b] References [21,22,35]. [c] References [21,36]. [d] Reference [37]. [e] Reference [27]. [f] Reference [38]. [g] References [21,39]. Where – is
shown, no data is available.
ChemMedChem 2011, 6, 498 – 504
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
499

L. Bunch et al.
MED
Figure 2. Three-point superimposition of calculated low-energy conforma-
tions (fitting atoms: ammonium and two carboxylate groups). a) KA (type
code), CPAA (purple) and SYM2081 (green); b) KA (type code), 1a (yellow)
and 1b (brown).
the binding affinity of SYM2081 at native KA receptors is as
low as 32 nm (only fivefold lower than for KA), it is within
reason to assume that such p–p interaction is not obligatory
for high-affinity binding.
The question as to whether introduction of a methyl group
in the 2-position of CPAA (corresponding to the 4-position of
SYM2081) would restore the high binding affinity at the KA re-
ceptors is intriguing and also of general interest as the new an-
alogues 1a,b would represent a new class of KA receptor li-
gands (Scheme 1). Firstly, the low-energy conformation of the
Scheme 2. Synthetic pathway towards the two stereoisomers of 2-Me-CPAA,
1a and 1b. Reagents and conditions: a) TBSCl, Et3N, CH2Cl2; b) BOC2O, Et3N,
DMAP, CH2Cl2 (87% two steps); c) LHMDS, THF, ꢁ788C, then PhSeCl; d) H2O2,
EtOAc, 08C!RT (65% two steps); e) 2-bromopropene, tBuLi, CuCN, TMSCl,
THF, ꢁ508C, then 4 (82%); f) BH3·THF, reflux, then H2O, NaOH, H2O2, 08C!
RT (42%); g) TBAF, THF, RT (91%); h) Separation of the two stereoisomers 7a
and 7b on HPLC (52%); i) RuCl3, NaIO4, H2O, MeCN, EtOAc (59%); j) HCl/diox-
ane, RT (60%).
strategy is the lack of control of the stereochemistry at the
vital C2-carbon. Thus, a successful separation of the two ste-
reoisomers is required, preferably as the last step of the syn-
thesis.
The projected synthesis of 1a,b (Scheme 2) commenced
with suitable protection of commercially available 2 to give 3
followed by introduction of the double bond to give 4, accord-
ing to literature procedures.[43] Copper-catalyzed conjugated
addition of the iso-propenyl group to 4 gave the expected ad-
dition product 5 as a single enantiomer (2,3-trans) in high yield
(82%).[38,42] The following one-pot borane-mediated reduction
of the lactam carbonyl group and hydroboration of the alkene
to give alcohols 6a/6b proceeded in 42% yield with a 3:2
ratio of the two C2 stereoisomers. Removal of the O-TBS group
was done smoothly with tetrabutylammonium fluoride (TBAF)
under standard conditions to give diols 7a/7b. Oxidation of
the two alcohols to their corresponding carboxylic acids 8a/
8b was achieved by a catalytic amount of RuIII with periodate
as a co-oxidant, according to the modified Sharpless proce-
dure.[44] Finally, deprotection of the amino group was done
with hydrochloric acid in dioxane to give 1a/1b in 60% yield.
Separation of the two C2 stereoisomers was monitored by
HPLC for each step. While work efficiency would call for sepa-
ration of stereoisomers as the final step, this was unfortunately
not possible. It turned out that only the diastereomeric diols
7a/7b displayed retention times of satisfactory separation.
Thus, the synthesis was repeated with a separation step prior
to oxidation of the diols to their corresponding carboxylic
acids, as described in the Experimental Section and Scheme 2.
The two diastereoisomers of the final products 1a/1b were ini-
Scheme 1. Chemical structures of CPAA and the rationally designed ana-
logues (2R)-methyl-CPAA (1a) and (2S)-methyl-CPAA (1b).
two stereoisomers 1a,b was determined by a stochastic con-
formational search. A subsequent three-point superimposition
(Figure 2b) revealed that the (2R)-stereoisomer 1a was optimal
at representing the Glu folded conformations of KA and
SYM2081, whereas the (2S)-stereosiomer 1b posed a flipping
of the envelope conformation of the pyrrolidine ring due to
the disfavored interaction of the 2-methyl group with the C4
ring carbon.
A structural analysis of 1a and 1b suggests that the two
target molecules may be looked at as highly functionalized
proline derivatives[40] or conformationally restricted Glu ana-
logues. Our retrosynthetic analysis advocates for (S)-pyrogluta-
minol (2) (Scheme 2) as the chiral starting synthon, which is
readily obtained from (S)-pyroglutamate.[41] This strategy will
secure the stereochemical configuration of the a-amino acid
functionality and furthermore, as the key step, the 2,3-trans-
configuration at C3 ring carbon may be fully controlled by an
addition of a suitable cuprate to the protected enone 4.[42] It
seemed attractive to generate the cuprate from 2-bromopro-
pene as the alkene would serve as a masked alcohol to be in-
troduced in a one-pot procedure together with the planned
reduction of the lactam by borane. The disadvantage of this
500
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 498 – 504

Ionotropic Glutamate Receptor Ligands
tially designated S1 and S2 awaiting C2 stereochemical assign-
ment. To unquestionably determine the stereochemical config-
uration at C2, an X-ray structure study was planned for. How-
ever, despite extensive efforts varying the parameters (solvent
and temperature), we were not successful and decided to pro-
ceed with the pharmacological characterization of S1 and S2.
Pharmacological characterization of S1 and S2 commenced
with binding affinity studies at native iGluRs (Table 2). For both
stereoisomers, the affinity for AMPA receptors was without sig-
nificance (>100 mm), whereas the binding affinity for KA recep-
tors was found to be in the low micromolar range (2.4 and
1.8 mm, respectively). The main difference was observed for
NMDA receptors as S1 displayed a 22 mm binding affinity,
development into a selective NMDA ligand and an investiga-
tion of its subtype selectivity profile at this group of Glu recep-
tors. Such plans are the topic of future work in our laborato-
ries.
Experimental Section
Chemistry
All reagents were obtained from commercial suppliers and used
without further purification. Solvents were dried by storing over
4 ꢁ molecular sieves, except for THF which was distilled over
sodium/benzophenone. Water- or air-sensitive reactions were con-
ducted in flame-dried glassware under nitrogen using a syringe/
septum cap technique. Purification
by flash chromatography was
done with silica gel size 40–63 mm
Table 2. Pharmacological evaluation of stereoisomers S1 and S2 at native AMPA, KA and NMDA receptors and
(Merck, silica gel 60) or 35–70 mm
(Fisher, silica 60 A). For TLC, Merck
TLC silica gel F254 plates were used
with appropriate spray reagents:
KMnO4 (0.5% in water), bromocre-
sol green (0.1% in EtOH) or ninhy-
cloned homomeric subtypes GluK1-3.
Compd
Native receptors[a]
KA
Cloned homomeric receptors[b]
AMPA
NMDA
GluK1
GluK2
GluK3
IC50 [mm]
Ki [mm]
Stereoisomer S1
Stereoisomer S2
>100 2.4 [5.63ꢀ0.06]
22 [4.65ꢀ0.12] 0.23ꢀ0.035 3.83ꢀ0.54
0.33ꢀ0.043
1
drin (0.06% in EtOH). H NMR and
>100 1.8 [5.75ꢀ0.02] >100
0.39ꢀ0.051 0.51ꢀ0.020 0.099ꢀ0.008
13C NMR spectra were obtained on
a Varian Mercury Plus (300 MHz)
[a] All values were calculated from full concentration–response curves. For AMPA and KA: IC50 values with
mean pIC50 ꢀSEM in brackets; For NMDA: Ki values with mean pKi ꢀSEM in brackets. [b] Ki values reported are
the mean ꢀSEM from at least three experiments, conducted in triplicate at 12–16 drug concentrations.
and a Varian Gemini 2000 instru-
ment (75 MHz), respectively. HPLC
was done using Agilent Prep HPLC
systems with Agilent 1100 series
whereas the binding affinity of S2 was negligible (>100 mm).
At the cloned homomeric KA receptors subtype GluK1–3
(Table 2), stereoisomer S1 showed affinity for GluK1,3 in the
low micromolar range (0.23 and 0.33 mm, respectively) with a
10- to 15-fold lower affinity for GluK2 (3.83 mm). This trend in
binding affinity profile at GluK1–3 is often observed for KA li-
gands as exemplified by the Glu analogue BOAD (Table 1) and
others.[20] However, for the S2 stereoisomer the binding profile
is different. For this analogue the binding affinity at GluK1–3 is
within the same low micromolar range (0.39, 0.51 and
0.099 mm, respectively) with four- to fivefold in favor of GluK3.
pump, Agilent 1200 series diode array, multiple wavelength detec-
tor (G1365B), and Agilent PrepHT high performance preparative
cartridge column (Zorbax, 300 SB-C18 Prep HT, 21.2ꢂ250 mm,
7 mm). MS spectra were recorded using LC-MS performed using an
Agilent 1200 series solvent delivery system equipped with an auto-
injector coupled to an Agilent 6400 series triple quadrupole mass
spectrometer equipped with an electrospray ionization (ESI)
source. Gradients of 10% aqueous acetonitrile + 0.05% formic
acid (solvent A) and 90% aqueous acetonitrile + 0.046% formic
acid (solvent B) were employed. Optical rotation was measured
using a Perkin–Elmer 241 spectrometer, with a sodium lamp at
589 nm. Melting points were measured using an automated melt-
ing point apparatus, MPA100 OptiMelt (SRS) and are stated uncor-
rected.
Conclusions
(2S,3R)-N-tert-Butoxycarbonyl-2-(tert-butyldimethylsilyloxymeth-
yl)-3-(iso-propenyl)-pyrrolidine-5 one (5): A solution of 2-bromo-
propene (0.67 mL, 7.60 mmol, 2.3 equiv) in dry THF (20 mL) at
ꢁ788C was treated dropwise with tBuLi (1.7m in pentane; 9.0 mL,
15.30 mmol, 4.7 equiv) and stirred for 15 min. Subsequently, a
slurry of CuCN (342 mg, 3.80 mmol, 1.2 equiv) in dry THF (6 mL)
was added via syringe, and the reaction mixture was allowed to
reach ꢁ508C. After stirring at this temperature for 5 min, a clear,
yellow solution was obtained and then recooled to ꢁ788C. Enone
4 (1.06 g, 3.25 mmol, 1.0 equiv) dissolved in dry THF (5 mL) was
added dropwise followed by TMSCl (1.0 mL, 7.60 mmol, 2.3 equiv).
The reaction mixture was allowed to reach ꢁ508C and stirred for
1 h. The solution was quenched with saturated aq NH4Cl (50 mL)
and extracted with EtOAc (3ꢂ100 mL). The combined organic
phases were washed with brine (150 mL), dried (MgSO4), filtered
and evaporated to dryness. The crude product was purified by
flash chromatography (EtOAc/heptane, 1:8; Rf =0.24) to give 5 as a
white solid (980 mg, 82%): mp: 35.1–35.98C; [a]2d3 =ꢁ50.60;
1H NMR (CDCl3): d=4.78 (ddd, 2H, J=5, 3, 1 Hz), 3.93 (m, 2H), 3.73
Two new conformationally restricted Glu analogues 1a/1b
were designed as hybrid structures of KA, CPAA and SYM2081.
The two C2 stereoisomers were synthesized and subsequently
underwent pharmacological evaluation as potential ligands at
the iGluRs. In summary, the pharmacological profile of the S1
stereoisomer at the iGluRs follows a trend often seen for iGluR
ligands. That is a preference for native KA and NMDA over
AMPA and a preference for cloned homomeric subtypes
GluK1,3 over GluK2. On the other hand, the S2 stereoisomer is
fully selective for the KA receptors with negligible affinity for
AMPA and NMDA receptors and similar affinities at the GluK1–
3 subtypes. Given the pharmacological profile of the S2 stereo-
isomer, this analogue of the two seems most promising as a
lead structure for the discovery of new selective KA ligands by
investigation of the chemical nature of the 2-substituent. How-
ever, the S1 stereoisomer may hold potential for the further
ChemMedChem 2011, 6, 498 – 504
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
501

L. Bunch et al.
MED
(dd, 1H, J=9, 1 Hz), 2.93 (dd, 1H, J=17.1, 9.5 Hz), 2.84 (br d, 1H,
J=9 Hz), 2.33 (dd, 1H, J=17, 1 Hz), 1.79 (s, 3H), 1.54 (s, 9H), 0.89
(s, 9H), 0.07 (s, 3H), 0.06 ppm (s, 3H); 13CNMR (CDCl3): d=173.8,
149.8, 145.7, 110.8, 82.8, 63.8, 63.5, 40.0, 37.6, 28.1, 25.8, 20.2, 18.2,
ꢁ5.5, ꢁ5.5 ppm.
(4 mL), and RuCl3·H2O (2.6 mg, 0.011 mmol, 0.02 equiv) was added.
The aqueous phase was transferred to the organic mixture and
water (2.2 mL) was added. The mixture was stirred at RT for 2.5 h.
The mixture was filtered and the filter cake washed with EtOAc.
The aqueous phase was extracted with EtOAc (10 mL) and the
combined organic layers were washed with brine (15 mL), dried
(MgSO4), filtered and evaporated to dryness. The product was puri-
fied by flash chromatography (CH2Cl2/MeOH/AcOH, 100:5:2; Rf =
0.25) to give 8a as a white foam (96 mg, 59%): [a]2d3 =ꢁ516.6;
1H NMR (CDCl3): d=4.15 (dd, 1H, J=6, 5 Hz), 3.60 (br m, 1H), 3.45
(br m, 1H), 2.70 (br m, 2H), 2.10 (m, 1H), 1.80 (m, 1H), 1.45 (d, 9H,
J=14 Hz), 1.30 ppm (d, 3H, J=7 Hz); 13C NMR (CDCl3): (rotamers in
parentheses) d=180.2 (180.0), 178.6, 155.3 (153.8), 81.2 (81.0), 61.6,
47.0 (46.0), 45.6 (45.5), 41.8 (41.6), 28.8 (28.3), 28.6 (28.5), 15.5 ppm
(14.9).
(S/R)-2-((2S,3R)-N-(tert-Butoxycarbonyl)-2-(tert-butyldimethylsilyl-
oxymethyl)pyrrolidin-3-yl)propan-1-ol (6a/6b): A solution of 5
(1.55 g, 4.2 mmol, 1.0 equiv) in dry THF (50 mL) was treated at RT
with BH3·THF complex (1m in THF; 25 mL, 25 mmol, 6.0 equiv), and
the reaction mixture was refluxed (708C) for 4 h. After cooling to
RT, an additional amount of THF (35 mL) was added and the reac-
tion mixture was cooled to 08C. Water (3 mL), NaOH (2m; 44 mL,
88 mmol, 21 equiv) and 30% w/w H2O2 in water (12 mL, 95 mmol,
23 equiv) were added carefully in this sequential fashion. The ice
bath was removed and the reaction mixture was stirred for 1 h at
RT, after which time it was quenched with saturated aq NaHCO3
(100 mL). The aqueous phase was extracted with EtOAc (3ꢂ
150 mL) and the combined organic layers were washed with brine
(200 mL), dried (MgSO4), filtered and evaporated to dryness. The
crude product was purified by flash chromatography (Et2O/hep-
tane, 2:1; Rf =0.34 and 0.25) to give a mixture of the two stereoiso-
mers 6a and 6b (3:2) as a colorless oil (670 mg, 42%): 1H NMR
(CDCl3), (two stereoisomers) d=3.80–3.40 (m, 6H, both isomers),
3.30–3.14 (m, 1H, both isomers), 2.43 (br s, 1H, minor isomer), 2.23
(m, 1H, major isomer), 2.07 (m, 1H, both isomers), 1.95 (m, 1H,
major isomer), 1.78 (m, 1H, minor isomer), 1.66 (m, 1H, minor
isomer), 1.47 (br s, 9H, both isomers), 0.98 (d, 3H, J=7 Hz, major
isomer), 0.92 (d, 3H, minor isomer), 0.91 (br s, 18H, both isomers),
0.09 ppm (br s, 12H, both isomers); 13C NMR (CDCl3) (two diaste-
reomers; rotamers in parentheses) d=154.0, 79.2, 78.8, 66.3, 65.9,
64.2, 63.4 (63.1), 61.6, 60.4 (60.3), 46.6, (46.5), 45.9 (45.8), 43.2
(42.4), 41.8 (41.4), 38.8 (38.6), 38.5 (38.3), 28.5, 28.3, 28.2, 27.0, 26.3,
25.9, 18.2, 14.3 (14.1), ꢁ5.4 ppm.
Stereoisomer 7b (100 mg, 0.35 mmol, 1.0 equiv) underwent the
same oxidative reaction conditions as described above to give 8b
1
as a white foam (60 mg, 58%): [a]2d3 =ꢁ739.0; H NMR (CDCl3): (ro-
tamers in parentheses) d=4.0 (4.1) (dd, 1H, J=8, 5 Hz), 3.65 (2.28)
(br m, 1H), 3.40 (3.55) (br m, 1H), 2.67 (b m, 1H), 2.53 (br m, 1H),
2.08 (2.0) (m, 1H), 1.64 (m, 1H), 1.45 (d, 9H, J=13 Hz), 1.25 ppm
(d, 3H, J=8 Hz); 13C NMR (CDCl3): (rotamers in parentheses) d=
180.2 (180.1), 178.0 (177.0), 155.1 (153.8), 81.2 (21.0), 63.6 (63.1),
46.8 (46.1), 45.8 (45.3), 42.4 (42.2), 28.6 (28.4), 28.2 (28.0), 15.4 ppm
(15.1).
(S/R)-2-((2S,3R)-2-(Carboxy)pyrrolidin-3-yl)propanoic acid (1a/
1b): For stereoisomer S1: 8a (95 mg, 0.33 mmol, 1.0 equiv) was
dissolved in dioxane (5 mL) and cooled to 08C. HCl (4m in dioxane;
1 mL, 4.0 mmol, 12.1 equiv) was added dropwise, and the ice bath
was removed and the solution stirred at RT for 18 h. The solvent
was evaporated to dryness, and the compound purified by HPLC
(mobile phase: 0.1% trifluoroacetic acid (TFA) in water, tR =
5.58 min.). The compound was dissolved in 1m HCl and evaporat-
ed to dryness to give S1 as the HCl-salt (36 mg, 60%): mp: 121.3–
173.98C (decomposition); [a]2d3 =211.96; 1H NMR (D2O): d=4.29
(dd, 1H, J=7, 3 Hz), 3.50–3.30 (br m, 2H), 2.70–2.60 (br m, 2H),
2.27 (m, 1H), 1.87 (m, 1H), 1.25 ppm (dd, 3H, J=7, 4 Hz); 13C NMR
(D2O): d=178.6, 172.0, 61.7, 45.7, 45.0, 41.0, 27.7, 15.1 ppm; LC-MS
(ESI): m/z [M+H]+ calcd for C8H13NO4: 188.08, found: 188.0.
(S/R)-2-((2S,3R)-N-(tert-Butoxycarbonyl)-2-(hydroxymethyl)pyrro-
lidin-3-yl)propan-1-ol (7a/7b): A solution of 7a/7b (750 mg,
2.0 mmol, 1.0 equiv) in dry THF (20 mL) was treated with TBAF (1m
in THF; 3.4 mL, 3.4 mmol, 1.7 equiv). The reaction mixture was
stirred for 1 h at RT and then quenched with saturated aq NaHCO3
(10 mL) and water (10 mL). The aqueous layer was extracted with
EtOAc (3ꢂ50 mL), and the combined organic layers were washed
with brine (150 mL), dried (MgSO4), filtered, and evaporated to dry-
ness. The crude product was purified by flash chromatography
(EtOAc/heptane, 9:1; Rf =0.25), to give the two stereoisomers 7a
and 7b (3:2) as a colorless oil (473 mg, 91%).
For stereoisomer S2: 8b (65 mg, 0.23 mmol, 1.0 equiv) was submit-
ted to the same reaction conditions as described above for S1, to
give S2 as a white solid (25 mg, 60%, tR =6.69 min): mp: 122.4–
1
142.38C (decomposition); [a]2d3 =ꢁ0.806; H NMR (D2O): d=4.17 (d,
1H, J=6 Hz), 3.50–3.30 (br m, 2H), 2.78 (m, 2H), 2.26 (1H), 1.87 (m,
1H), 1.22 (br d, 3H, J=6 Hz); 13C NMR (D2O): d=179.2, 171.5, 61.2,
45.7, 44.7, 41.3, 27.5, 14.0 ppm; LC-MS (ESI): m/z [M+H]+ calcd for
C8H13NO4: 188.08, found: 188.0.
Separation of the two stereoisomers was carried out on HPLC with
15% CH3CN in water as a mobile phase and a flow rate of
20 mLminꢁ1 with UV detection at 200 nm. Stereoisomer 7a: tR =
1
27.4 min; yield=147 mg; de>98%; [a]2d3 =ꢁ399.9; H NMR (CDCl3):
d=3.65–3.49 (m, 6H), 3.21 (m, 1H), 1.94 (m, 1H), 1.76 (m, 1H), 1.65
(m, 1H), 1.47 (s, 10H), 1.0 ppm (d, 3H, J=7 Hz); 13C NMR (CDCl3):
d=156.5, 80.3, 67.1, 65.8, 62.0, 46.6, 43.0, 38.4, 28.6, 28.3,
15.0 ppm. Stereoisomer 7b: tR =37.9 min; yield=124 mg; de=
98%; [a]2d3 =ꢁ307.5; 1H NMR (CDCl3): d=3.74–3.46 (m, 6H), 3.15
(ddd, 1H, J=10, 7 Hz), 1.94 (m, 1H), 1.82 (m, 1H), 1.70 (m, 1H),
1.55 (m, 1H), 1.45 (s, 9H), 0.92 ppm (d, 3H, J=7 Hz); 13C NMR
(CDCl3): d=156.5, 80.4, 66.8, 66.4, 63.1, 46.6, 42.9, 37.6, 28.6, 26.7,
13.8 ppm. Total yield after separation by HPLC: 52%.
Modeling study
The modeling study was performed using the Molecular Operating
Environment (MOE) software package (v2009.10; Chemical Com-
puting Group, 2009) using the built-in mmff94x force field and the
GB/SA continuum solvent model. The compound was submitted to
a stochastic conformational search and with respect to its global
minimum (DG in kcalmolꢁ1
) returned conformations above
+7 kcalmolꢁ1 were discarded. The g-carboxylate group was pro-
tonated prior to execution of the conformational search, as this
gave a larger and thus more reliable number of output conforma-
tions. The superimposition of ligands was carried out using the
built-in function in MOE, by fitting the ammonium group and the
two carboxylate groups.
(S/R)-2-((2S,3R)-N-(tert-Butoxycarbonyl)-2-(carboxy)pyrrolidin-3-
yl)propanoic acid (8a,b): Stereoisomer 7a (147 mg, 0.57 mmol,
1.0 equiv) was dissolved in EtOAc (4.3 mL) and CH3CN (4.3 mL).
NaIO4 (995 mg, 4.65 mmol, 8.2 equiv) was dissolved in water
502
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 498 – 504

Ionotropic Glutamate Receptor Ligands
[19] G. M. Alexander, D. W. Godwin, Epilepsy Res. 2006, 71, 1–22.
[20] L. Bunch, P. Krogsgaard-Larsen, Med. Res. Rev. 2009, 29, 3–28.
[21] E. Sagot, D. S. Pickering, X. Pu, M. Umberti, T. B. Stensbøl, B. Nielsen, M.
Chapelet, J. Bolte, T. Gefflaut, L. Bunch, J. Med. Chem. 2008, 51, 4093–
4103.
[22] S. R. Baker, D. Bleakman, J. Ezquerra, B. A. Ballyk, M. Deverill, K. Ho, R. K.
Kamboj, I. Collado, C. Dominguez, A. Escribano, A. I. Mateo, C. Pedregal,
A. Rubio, Bioorg. Med. Chem. Lett. 2000, 10, 1807–1810.
[23] L. Bunch, T. H. Johansen, H. Brꢃuner-Osborne, T. B. Stensbol, T. N. Johan-
sen, P. Krogsgaard-Larsen, U. Madsen, Bioorg. Med. Chem. 2001, 9, 875–
879; L. Bunch, T. Gefflaut, S. Alaux, E. Sagot, B. Nielsen, D. S. Pickering,
Eur. J. Pharmacol. 2009, 609, 1–4.
[24] A. M. Larsen, L. Bunch, ACS Chem. Neurosci. 2011, DOI: 10.1021/
cn1001039.
[25] A. M. Larsen, R. Venskutonytꢄ, B. Nielsen, D. S. Pickering, L. Bunch, ACS
Chem. Neurosci. 2011, DOI: 10.1021/cn100093f.
[26] S. B. Vogensen, J. R. Greenwood, L. Bunch, R. P. Clausen, Curr. Top. Med.
Chem. 2011, in press.
[27] L. Bunch, T. Liljefors, J. R. Greenwood, K. Frydenvang, H. Brꢃuner-Os-
borne, P. Krogsgaard-Larsen, U. Madsen, J. Org. Chem. 2003, 68, 1489–
1495.
Pharmacology
Native receptor binding assays: Affinities for native AMPA, KA and
NMDA receptors in rat cortical synaptosomes were determined
using 5 nm (RS)-[3H]AMPA (55.5 Cimmolꢁ1),[45] 5 nm [3H]KA
(58.0 Cimmolꢁ1 [46]
)
and 2 nm [3H]CGP 39653 (Kd =6 nm,
50.0 Cimmolꢁ1),[47] respectively, with minor modifications as previ-
ously described.[48]
Recombinant receptor binding assays: Sf9 cells were cultured and
infected with recombinant baculovirus of cloned rat GluA2(R)o or
GluK1–3 and membranes prepared and used for binding as previ-
ously detailed.[21,49] The binding affinity of 1 was determined from
competition
experiments
with
2–5 nm
(R,S)-[3H]AMPA
(42.1 Cimmolꢁ1; PerkinElmer, Waltham, MA, USA) at GluA2(R)o or 1–
5 nm [3H]SYM2081 (40 Cimmolꢁ1; ARC, St. Louis, MO, USA) at
GluK1(Q)1b, GluK2(V,C,R)a and GluK3a. Italic letters in parentheses in-
dicate the RNA-edited isoforms of the subunits used. The IC50, Hill
coefficient and Ki values for 1 were evaluated as previously de-
scribed.[50]
[28] A. M. Davis, S. A. St-Gallay, G. J. Kleywegt, Drug Discovery Today 2008,
13, 831–841; C. K. Tygesen, J. S. Rasmussen, S. V. Jones, A. Hansen, K.
Hansen, P. H. Andersen, Proc. Natl. Acad. Sci. USA 1994, 91, 13018–
13022.
Acknowledgements
[29] A. M. Davis, S. J. Teague, G. J. Kleywegt, Angew. Chem. 2003, 115, 2822–
2841; Angew. Chem. Int. Ed. 2003, 42, 2718–2736.
[30] D. Tarasov, D. Tovbin, J Mol Model 2009, 15, 329–341; B. Korczak, S. L.
Nutt, E. J. Fletcher, K. H. Hoo, C. E. Elliott, V. Rampersad, E. A. McWhinnie,
R. K. Kamboj, Recept. Channels 1995, 3, 41–49.
We would like to thank The Lundbeck Foundation, The Carlsberg
Foundation and GluTarget for financial support.
Keywords: glutamate receptors · kainate · neurodegenerative
diseases · neurotoxicity · neurotransmitters
[31] N. Armstrong, E. Gouaux, Neuron 2000, 28, 165–181.
[32] M. L. Mayer, Neuron 2005, 45, 539–552.
[33] K. Hashimoto, T. Matsumoto, K. Nakamura, S. Ohwada, T. Ohuchi, M. Ho-
rikawa, K. Konno, H. Shirahama, Bioorg. Med. Chem. 2002, 10, 1373–
1379.
[34] L. M. Zhou, Z. Q. Gu, A. M. Costa, K. A. Yamada, P. E. Mansson, T. Giorda-
no, P. Skolnick, K. A. Jones, J. Pharmacol. Exp. Ther. 1997, 280, 422–427;
Z. Q. Gu, D. P. Hesson, J. C. Pelletier, M. L. Maccecchini, L. M. Zhou, P.
Skolnick, J. Med. Chem. 1995, 38, 2518–2520; H. Brꢃuner-Osborne, B.
Nielsen, T. B. Stensbol, T. N. Johansen, N. Skjaerbaek, P. Krogsgaard-
Larsen, Eur. J. Pharmacol. 1997, 335, R1-R3.
[35] P. Conti, M. De Amici, G. De Sarro, M. Rizzo, T. B. Stensbol, H. Brauner-
Osborne, U. Madsen, L. Toma, C. De Micheli, J. Med. Chem. 1999, 42,
4099–4107.
[36] K. Shimamoto, B. LeBrun, Y. Yasuda-Kamatani, M. Sakaitani, Y. Shigeri, N.
Yumoto, T. Nakajima, Mol. Pharmacol. 1998, 53, 195–201.
[37] J. D. Sonnenberg, H. P. Koch, C. L. Willis, F. Bradbury, D. Dauenhauer, R. J.
Bridges, A. R. Chamberlin, Bioorg. Med. Chem. Lett. 1996, 6, 1607–1612.
[38] L. Bunch, B. Nielsen, A. A. Jensen, H. Brauner-Osborne, J. Med. Chem.
2006, 49, 172–178.
[39] S. Kwak, H. Aizawa, M. Ishida, H. Shinozaki, Neurosci. Lett. 1992, 139,
114–117; A. L. Smith, R. A. J. McIlhinney, Br. J. Pharmacol. 1992, 105,
83–86.
[1] P. E. Chen, D. J. A. Wyllie, Br. J. Pharmacol. 2006, 147, 839–853.
[2] F. Ferraguti, R. Shigemoto, Cell Tissue Res. 2006, 326, 483–504.
[3] G. Campiani, C. Fattorusso, M. De Angelis, B. Catalanotti, S. Butini, R.
Fattorusso, I. Fiorini, V. Nacci, E. Novellino, Curr. Pharm. Des. 2003, 9,
599–625; R. P. Seal, S. G. Amara, Annu. Rev. Pharmacool. Toxicol. 1999,
39, 431–456; L. Bunch, M. N. Erichsen, A. A. Jensen, Expert Opin. Ther.
Targets 2009, 13, 719–731.
[4] G. L. Collingridge, R. W. Olsen, J. Peters, M. Spedding, Neuropharmacolo-
gy 2009, 56, 2–5.
[5] A. A. Kaczor, D. Matosiuk, Curr. Med. Chem. 2010, 17, 2608–2635.
[6] B. Sommer, N. Burnashev, T. A. Verdoorn, K. Keinanen, B. Sakmann, P. H.
Seeburg, EMBO J. 1992, 11, 1651–1656.
[7] C. H. Cui, M. L. Mayer, J. Neurosci. 1999, 19, 8281–8291.
[8] J. Lerma, FEBS Lett. 1998, 430, 100–104.
[9] K. McKeage, CNS Drugs 2009, 23, 881–897.
[10] J. R. Mellor, Biochem. Soc. Trans. 2006, 34, 949–951.
[11] R. Chittajallu, S. P. Braithwaite, V. R. J. Clarke, J. M. Henley, Trends Phar-
macol. Sci. 1999, 20, 26–35; D. Bleakman, M. R. Gates, A. M. Ogden, M.
Mackowiak, Curr. Pharm. Des. 2002, 8, 873–885; D. Bleakman, D. Lodge,
Neuropharmacology 1998, 37, 1187–1204; P. Pinheiro, C. Mulle, Cell
Tissue Res. 2006, 326, 457–482; J. Lerma, A. V. Paternain, A. Rodriguez-
Moreno, J. C. Lopez-Garcia, Physiol. Rev. 2001, 81, 971–998; J. Lerma,
Nat. Rev. Neurosci. 2003, 4, 481.
[40] U. Kazmaier, C. Schmidt, Synlett 2009, 1136–1140.
[41] S. K. Panday, J. Prasad, D. K. Dikshit, Tetrahedron: Asymmetry 2009, 20,
1581–1632.
[42] L. Bunch, P. Krogsgaard-Larsen, U. Madsen, Synthesis 2002, 31–33.
[43] C. M. Acevedo, E. F. Kogut, M. A. Lipton, Tetrahedron 2001, 57, 6353–
6359.
[44] P. H. J. Carlsen, T. Katsuki, V. S. Martin, K. B. Sharpless, J. Org. Chem.
1981, 46, 3936–3938.
[45] T. Honorꢄ, M. Nielsen, Neurosci. Lett. 1985, 54, 27–32.
[46] D. J. Braitman, J. T. Coyle, Neuropharmacology 1987, 26, 1247–1251.
[47] 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.
[48] M. B. Hermit, J. R. Greenwood, B. Nielsen, L. Bunch, C. G. Jorgensen, H. T.
Vestergaard, T. B. Stensbol, C. Sanchez, P. Krogsgaard-Larsen, U. Madsen,
H. Brꢃuner-Osborne, Eur. J. Pharmacol. 2004, 486, 241–250; R. P. Clau-
sen, K. B. Hansen, P. Cali, B. Nielsen, J. R. Greenwood, M. Begtrup, J.
Egebjerg, H. Brauner-Osborne, Eur. J. Pharmacol. 2004, 499, 35–44.
[12] S. Chaki, T. Okubo, Y. Sekiguchi, Recent Pat. Anti-Cancer Drug Discovery
2006, 1, 1–27.
[13] D. Michelson, L. R. Levine, M. A. Dellva, P. Mesters, D. D. Schoepp, E. Du-
nayevich, G. D. Tollefson, Neuropharmacology 2005, 49, 257–257.
[14] P. W. Kalivas, Dialogues Clin. Neurosci. 2007, 9, 389–397; P. M. t. Lea, A. I.
Faden, CNS Drug Rev. 2006, 12, 149–166.
[15] M. Vikelis, D. D. Mitsikostas, CNS & neurological disorders drug targets
2007, 6, 251–257.
[16] K. W. Muir, Curr. Opin. Pharmacol. 2006, 6, 53–60.
[17] A. C. Foster, J. A. Kemp, Curr. Opin. Pharmacol. 2006, 6, 7–17; P. M.
Beart, R. D. O’Shea, Br. J. Pharmacol. 2007, 150, 5–17; G. J. Lees, Drugs
2000, 59, 33–78.
[18] A. Almeida, S. J. Heales, J. P. Bolanos, J. M. Medina, Brain Res. 1998, 790,
209–216.
ChemMedChem 2011, 6, 498 – 504
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
503

L. Bunch et al.
MED
[49] S. B. Vogensen, R. P. Clausen, J. R. Greenwood, T. N. Johansen, D. S. Pick-
ering, B. Nielsen, B. Ebert, P. Krogsgaard-Larsen, J. Med. Chem. 2005, 48,
3438–3442.
[50] B. S. Nielsen, T. G. Banke, A. Schousboe, D. S. Pickering, Eur. J. Pharmacol.
1998, 360, 227–238.
Received: December 16, 2010
Published online on January 25, 2011
504
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 498 – 504