Chemoenzymatic synthesis of new 2,4-syn-functionalized (S)-glutamate analogues and structure-activity relationship studies at ionotropic glutamate receptors and excitatory amino acid transporters

Article abstract of DOI:10.1021/jm301433m

In the mammalian central nervous system, (S)-glutamate (Glu) is released from the presynaptic neuron where it activates a plethora of pre- and postsynaptic Glu receptors. The fast acting ionotropic Glu receptors (iGluRs) are ligand gated ion channels and are believed to be involved in a vast number of neurological functions such as memory and learning, synaptic plasticity, and motor function. The synthesis of 14 enantiopure 2,4-syn-Glu analogues 2b-p is accessed by a short and efficient chemoenzymatic approach starting from readily available cyclohexanone 3. Pharmacological characterization at the iGluRs and EAAT1-3 subtypes revealed analogue 2i as a selective GluK1 ligand with low nanomolar affinity. Two X-ray crystal structures of the key analogue 2i in the ligand-binding domain (LBD) of GluA2 and GluK3 were determined. Partial domain closure was seen in the GluA2-LBD complex with 2i comparable to that induced by kainate. In contrast, full domain closure was observed in the GluK3-LBD complex with 2i, similar to that of GluK3-LBD with glutamate bound.

Full text of DOI:10.1021/jm301433m

Article
pubs.acs.org/jmc
Chemoenzymatic Synthesis of New 2,4-syn-Functionalized
(S)‑Glutamate Analogues and Structure−Activity Relationship
Studies at Ionotropic Glutamate Receptors and Excitatory Amino
Acid Transporters
Zeinab Assaf,‡,§ Anja P. Larsen,† Raminta Venskutonyte,† Liwei Han,† Bjarke Abrahamsen,†
̇
Birgitte Nielsen,† Michael Gajhede,† Jette S. Kastrup,† Anders A. Jensen,† Darryl S. Pickering,†
Karla Frydenvang,*,† Thierry Gefflaut,*,‡,§ and Lennart Bunch*,†
†Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen, 2100
Copenhagen Ø, Denmark
‡Clermont Universite,
France
§CNRS, UMR 6296, ICCF, 63177 Aubier
́ ́
Universite Blaise Pascal, Institut de Chimie de Clermont-Ferrand, BP 10448, F-63000 Clermont-Ferrand,
̀
e, France
S
* Supporting Information
ABSTRACT: In the mammalian central nervous system, (S)-
glutamate (Glu) is released from the presynaptic neuron where
it activates a plethora of pre- and postsynaptic Glu receptors.
The fast acting ionotropic Glu receptors (iGluRs) are ligand
gated ion channels and are believed to be involved in a vast
number of neurological functions such as memory and learning, synaptic plasticity, and motor function. The synthesis of 14
enantiopure 2,4-syn-Glu analogues 2b−p is accessed by a short and efficient chemoenzymatic approach starting from readily
available cyclohexanone 3. Pharmacological characterization at the iGluRs and EAAT1−3 subtypes revealed analogue 2i as a
selective GluK1 ligand with low nanomolar affinity. Two X-ray crystal structures of the key analogue 2i in the ligand-binding
domain (LBD) of GluA2 and GluK3 were determined. Partial domain closure was seen in the GluA2-LBD complex with 2i
comparable to that induced by kainate. In contrast, full domain closure was observed in the GluK3-LBD complex with 2i, similar
to that of GluK3-LBD with glutamate bound.
task but essential, as such ligands are exceedingly valuable
compounds to be used in the study of the function of a specific
iGluR in normal and diseased states.
INTRODUCTION
■
In the mammalian central nervous system (CNS), (S)-
glutamate (Glu) functions as the major excitatory neuro-
transmitter. Once released from the presynaptic neuron into
the synapse, Glu activates a number of pre- and postsynaptic
Glu receptors. On the basis of pharmacological function and
signal transduction pathway, the Glu receptors have been
grouped into two main classes: the fast acting ionotropic Glu
receptors (iGluRs), which are ion channels,1 and the G-protein-
coupled metabotropic Glu receptors (mGluRs), which produce
a much slower signal transduction through second messenger
systems.2 Uptake of Glu from the synaptic cleft is mediated by
the excitatory amino acid transporters (EAATs), also termed
the Glu transporters.3−5 On the basis of phylogeny and ligand
affinity studies, the iGluRs are divided into three groups named
the 2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl)propanoic acid
(AMPA) receptors (subunits GluA1−4), kainic acid or kainate
(KA) receptors (subunits GluK1−5), and the N-methyl-D-
aspartate (NMDA) receptors (subunits GluN1, GluN2A−D,
and GluN3A,B). Functional iGluRs are homo- or heterotetra-
meric in subunit composition, which has been shown to
influence greatly the function of the receptor. Discovery of
subtype selective ligands for the iGluRs is a highly challenging
In this paper the chemoenzymatic synthesis of a new series of
2,4-syn substituted Glu analogues is presented together with
their pharmacological characterization at the iGluRs and the
EAAT1−3 subtypes. Furthermore, crystal structures of the
GluA2 ligand-binding domain (GluA2-LBD) construct and the
GluK3 ligand-binding domain (GluK3-LBD) construct both in
complex with 2i were obtained and new insight into the
structure−activity relationship of the AMPA and KA receptors
is disclosed.
RESULTS AND DISCUSSION
■
Chemistry. Aminotransferases (ATs) are ubiquitous
enzymes involved in proteinogenic α-amino acid metabolism
by catalyzing the reversible conversion of 2-oxocarboxylic acids
into L-α-amino acids. Most naturally occurring ATs accept Glu
and its ketoacid homologue α-ketoglutarate (KG) as preferred
substrates but are generally characterized by a relatively broad
Received: October 3, 2012
Published: February 15, 2013
© 2013 American Chemical Society
1614
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
substrate spectrum.6 Therefore, several ATs have proved to be
useful for the stereoselective synthesis of Glu analogues.
Notably, aspartate aminotransferase (AspAT) isolated from
Escherichia coli has been shown to be highly effective as a
catalyst for the stereospecific synthesis of a wide range of L-2,4-
syn-substituted Glu analogues from racemic KG 1 (Scheme 1)
Scheme 2. Synthesis of 4-Substituted KG Analogues 1a−j
Scheme 1. AspAT Catalyzed Synthesis of 4-Substituted Glu
Analogues
by kinetic resolution.7−11 Moreover, AspAT accepts cysteine
sulfinic acid (CSA) as amino donor substrate. This close
analogue of aspartate, readily prepared from L-cystine,12 is
converted into pyruvylsulfinic acid, which is highly unstable and
thus decomposes into pyruvic acid and sulfur dioxide, forcing
the needed shift in equilibrium of the transamination reaction.13
All herein reported new 2,4-syn-Glu analogues 2b−o were
prepared by enzymatic transamination of the corresponding 4-
substituted-2-oxoglutarates 1a−j (Scheme 1).
Previously, several pathways have been explored for the
synthesis of the various substituted KG analogues needed as
substrates of the transamination reaction. While a Claisen−
Johnson rearrangement proved efficient to create the carbon
skeleton of 4-alkyl analogues,11,14 alternative methodologies
were needed to prepare derivatives comprising functionalized
substituents at the 4-position.9,15 The new KG analogues 1a−j
described in the present study were prepared following an
original and highly efficient synthetic route giving access to a
variety of new functionalized derivatives.
yield from 3. Our new approach involving anhydride 5 gave 1a
with an overall yield of 84%. This result clearly demonstrates
the efficiency of this new chemical route.
Synthesis of 2-Oxoglutaric Acids. As shown in Scheme 2,
the new KG analogues 1a−j were prepared in four steps from
cyclic β-keto ester 3, itself readily prepared following a
described procedure from methyl malonate and methyl
acrylate.16 Compound 3 was reacted with isobutyl chlorofor-
mate (IBCF) and Et3N to give the corresponding enol
carbonate 4 in quantitative yield. Oxidative cleavage of 4 with
ozone afforded α-ketoester 5, also comprising a mixed carbonic
anhydride functionality. NMR analysis of the crude product
indicated that 5 was formed in high yield (>90%), and this
reactive intermediate was generally used directly in the
following step without isolation. However, a CH2Cl2 solution
of 5 could be stored at −20 °C for several weeks without
noticeable degradation. Anhydride 5 appears as a key
intermediate, as it can be reacted with numerous nucleophiles
offering an easy access to chemical diversity: The use of various
amines, hydroxylamines, and one hydrazine afforded the
corresponding amides 6a−f, hydroxamates 6g−i, and hydrazide
6j isolated after chromatography in 52−94% overall yield from
3. Finally, the lithium salts 1a−j were obtained in quantitative
yield by selective ester hydrolysis using a stoichiometric amount
of LiOH. Previously, compound 1a was prepared from direct
ozonolysis of the enamine formed by reaction of 3 with
ammonia.9 However, this approach suffered from the low yield
of the enamine ozonolysis and 1a was obtained in only 55%
As shown in Scheme 3, anhydride 5 was also reacted with
sodium azide to give the acyl azide intermediate 7, which upon
heating was converted into the isocyanate 8 via a Curtius
1
rearrangement. As judged by H NMR analysis of the crude
Scheme 3. Synthesis of the Carbamate 1k
1615
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
products, 7 and 8 were formed in nearly quantitative yield.
Isocyanate 8 was used in the next reaction without purification,
and reaction with benzyl alcohol afforded the KG 9 bearing a
benzyloxycarbonyl (Cbz) substituted amino group, in an overall
yield of 53% from 3 (five steps). It was then converted to the
dilithium salt 1k in quantitative yield. 1H NMR analyses
showed that 9 and 1k exist in solution as cyclic forms resulting
from the addition of the carbamate nitrogen to the ketone. For
cyclic dimethyl KG 9, one diastereoisomer was the major form
(>85%) in CDCl3 whereas KG 1k was mainly present in D2O
in its acyclic form.
Finally, KG 10 was obtained by dehydration of the amide
functionality of 6a in the presence of trifluoroacetic anhydride
and pyridine in 87% yield and was converted quantitatively to
the dilithium salt 1l without alteration of the nitrile
functionality (Scheme 4).
catalyzed transaminations between aspartate (40 mM) and KG
substrates 1b−l in variable concentrations allowed for the
estimation of the kinetic parameters Km and kcat (Table 1).
These results are fully consistent with the previously
observed substrate specificity of AspAT. All the newly prepared
KG analogues proved to be substrates with Km values in the
millimolar range and kcat of 2−112% relative to the natural
substrate KG. We have observed by molecular modeling of the
enzyme active site that substituents at position 4 fit in the
hydrophobic channel connecting the active site pocket to the
solvent (data not shown). This explains how bulky groups and
long chains can be accepted by the enzyme, with a preference
for hydrophobic moieties. KG analogues 1b−j bearing an
amide, hydroxamate, or hydrazide functionality are thus
relatively poor substrates compared with the 4-alkyl derivatives
described in earlier studies.11 However, the measured activities
are still valuable for synthetic purposes. The Km and kcat values
measured for 1k and 1l are noteworthy, as they fall into the
same range as the natural substrate KG. The kinetic parameters
measured for nitrile 1l were not surprising, with values
comparable to those of the 4-propyl KG analogue (Km = 0.4
mM, kcat = 53%). However, the high kcat value measured for 1k
(112%) was quite unexpected when compared to that of the
structurally close analogue 1h (4%).
Scheme 4. Synthesis of the Nitrile 1l
Preparative scale transaminations were easily carried out for
1b−l following the procedure previously described.11 Cysteine
sulfinic acid (CSA) was used as an irreversible amino donor,
and the reaction was monitored by enzymatic titration of
pyruvate formed from CSA. The reaction was stopped at
approximately 40% conversion in order to perform the kinetic
resolution of the racemic substrates. Glu analogues 2b−f, 2k,
and 2l were isolated after selective adsorption on sulfonic
Transamination Reactions. The substituted KGs 1b−l
were evaluated as substrate of E. coli AspAT on the basis of the
Michaelis−Menten model. Rate measurement of the AAT
a
Table 1. Kinetic Parameters of AspAT Catalyzed Transaminations with KG Analogues 1b−l
a
Values and standard errors were calculated from the Hanes−Woolf plot according to the least-squares method and Gauss’s error propagation law.
b
The absolute kcat value measured in our experimental conditions with KG was 39.6 1.2 s−1.
1616
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
Dowex 50 resin (H+ form) and elution with aqueous ammonia.
They were further purified by ion exchange chromatography on
cationic Dowex 1 resin (AcO− form) by elution with an AcOH
gradient. Slightly different purification protocols were used for
2g−j. Indeed, these derivatives proved to be unstable in basic
and/or acidic solutions; 2g−j were partially hydrolyzed to the
carboxylic acid derivative in aqueous ammonia, whereas 2i also
proved unstable in aqueous AcOH solution. The first
chromatographic step on Dowex 50 was thus omitted for 2g,
2h, and 2j, which were instead purified by a single
chromatographic step using Dowex 1. In the case of 2i both
chromatographic steps proved to be necessary, and elution with
neutral aqueous ammonium bicarbonate solutions from Dowex
50 (H+ form) and Dowex 1 (HCO3− form) was performed. Glu
analogues 2b−l were isolated in 35−45% yield from the
corresponding racemic KG analogues 1b−l. In all cases, only
one stereoisomer was isolated with a diastereomeric excess over
98% as judged by NMR. AspAT was previously shown to
display a very high stereoselectivity in favor of the L-2,4-syn
configuration. This stereoselectivity was unambiguously dem-
onstrated for the previously prepared 2-alkyl analogues11 as well
as for functionalized analogues including the amide or ester
functionalities.9,10 This stereoselectivity was again confirmed for
four representative compounds of this series (1c, 1i, 1h, and
1l), which gave after acidic hydrolysis the already described L-
2,4-syn-4-carboxyethyl glutamate analogue (data not shown).
This last compound was also obtained by hydrolysis of a
previously prepared amide for which the L-2,4-syn configuration
was demonstrated by NMR.9 As shown in Scheme 5, free
hydroxamate 2m and amino derivatives 2n and 2o were
prepared from 2h, 2k, and 2l, respectively.
hydroxamate functionality to the carboxylic acid and formation
of the corresponding pyroglutamate analogue. Conversely, Pd/
C catalyzed hydrogenation of 2h proved to be more efficient.
Debenzylation leading to 2m was the main observed reaction.
Although a small quantity (<5%) of primary amide was formed
as a result of N−O cleavage, both compounds could easily be
separated by ion exchange chromatography on Dowex 1 and
2m was isolated in 85% yield.
Hydrogenation of 2k in the presence of Pd/C also gave 2n in
quantitative yield. Attempts to prepare the amino homologue
2o by direct hydrogenation of the nitrile 2l gave mixtures of 2o
and the azepane 2p, which was isolated in yield up to 56%.
Moreover, the reaction required several days for completion
and partial epimerization was observed at position 2 (up to
10%). To circumvent these problems, the amino group of 2l
was protected with a BOC group to give 11 in 95% yield.
Nitrile reduction was then performed in the presence of Pd
hydroxide and was complete within 90 min. Cleavage of the
BOC group of 12 with 1 M HCl gave 2o which was finally
isolated after ion exchange chromatography with an overall
yield of 80% from 2l. No epimerization at the C2 position was
observed under these conditions.
Pharmacology. The newly synthesized L-2,4-syn-4-sub-
stituted Glu analogues 2b−p were evaluated as ligands for the
iGluRs. In binding studies at native AMPA, KA, and NMDA
receptors (Table 2) the Glu analogues showed mid- to high-
range micromolar affinities (IC50 of 6.5 to >100 μM) and low
receptor group selectivity except for cyano analogue 2l, which
displayed mid-nanomolar affinity (IC50 = 73 nM) at native KA
receptors corresponding to a >90-fold selectivity over native
AMPA and NMDA receptors.
First attempts of acid catalyzed removal of the tert-butyl
group of 2g were unsuccessful resulting in hydrolysis of the
Subsequently, binding affinities of 2b−p were determined at
recombinant homomeric rat KA subtypes GluK1−3 and AMPA
subtype GluA2 expressed in Sf 9 cell membranes (Table 2). The
affinity profile of all tested analogues 2b−p showed the same
trend across the three subtypes GluK1−3, displaying a relatively
highest affinity for GluK1, lowest for GluK2, and intermediate
affinity for GluK3 (Figure 1). This trend is often observed for
orthosteric GluK1−3 agonists exemplified by a large number of
L-2,4-syn Glu analogues10,11,17 as well as conformationally
restricted18−20 Glu analogues.21 Noteworthy, of the new series
reported here, N-methyl hydroxamate 2i displays low nano-
molar affinity for GluK1 (Ki = 3.4 nM) with a 200- and 35-fold
selectivity over GluK2 and GluK3, respectively (Table 2).
Subsequently, representative compounds of the Glu
analogues 2b−p were characterized pharmacologically at the
human EAAT1,2,3 subtypes stably expressed in HEK293 cells
in a [3H]-D-aspartate uptake assay (Table 3).24 While the
majority of the analogues were without inhibitory activity at any
of the three EAAT1−3 subtypes, analogue 2l displayed activity
at EAAT2, exhibiting an IC50 of 24 μM at this subtype and 9-
and >40-fold lower inhibitory potencies at EAAT1 and EAAT3,
respectively (Table 3). A comparison with previously reported
EAAT inhibitory profiles of a large number of L-2,4-syn-Glu
analogues suggests that the observed preference for EAAT2 is a
trend for this class of compounds, although the structural basis
for this remains unclear.9,11 Since analogue 2l shows nanomolar
affinity for the KA receptors and micromolar affinity for the
AMPA receptors (Table 2), it is not suitable as a
pharmacological tool for the investigation of EAAT2 function.5
However, in the same series we have previously shown that
extending the side chain results in a selective EAAT2 inhibitor
with low affinity for all of the iGluRs.9
Scheme 5. Synthesis of Hydroxamate 2m and Amino
Derivatives 2n and 2o
1617
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
Table 2. Pharmacological Evaluation of 2b−p at Native and Recombinant Homomeric Rat iGluR
a
Data are mean values of three to four individual experiments performed in triplicate. For AMPA and KA: pIC50 values with SEM in brackets. For
b
NMDA: pKi values with SEM in brackets. Mean
concentrations. Data taken from ref 22. Data taken from ref 23. Data taken from ref 10.
SEM from at least three separate experiments, conducted in triplicate at 12−16 ligand
c
d
e
X-ray Crystal Structure Study. To explore in detail the
structural basis for the high affinity binding of 2i to GluK1 and
the subtype selectivity determinants, an X-ray crystallographic
study was undertaken. However, despite several attempts to
crystallize 2i in GluK1-LBD, crystals of suitable quality were
not obtained. In contrast, crystals of 2i in the GluK3-LBD and
GluA2-LBD were achieved and their structures solved.
group P21212) with one molecule in the asymmetric unit of the
crystal (Figure 2A). Generally, the electron density is very well-
defined (for further information on data collection and
refinement statistics, see Supporting Information). In the
binding site, electron density outside the Glu parental skeleton
was clearly identified and thus 2i could be modeled herein
(Figure 2B). The ligand 2i induces only partial domain closure
(13.8°) of the GluA2-LBD (Figure 2A), which has also been
observed for the X-ray structure of the partial agonist KA in
Structure of GluA2-LBD with 2i. The structure of GluA2-
LBD in complex with 2i was solved at 1.24 Å resolution (space
1618
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
Table 3. Pharmacological Evaluation of 2b−c, 2e, 2f, 2h, 2i,
and 2k−p at EAAT1−3
Figure 1. The log(Ki) (in nM) of Glu analogues 2a−p at GluK1−3
subtypes.
GluA2-LBD (12.2°, PDB code 1FW0).25 In this respect the
binding affinity of 2i at the GluA2-LBD construct was
determined to be in the micromolar range (Ki = 126
17
μM, nH = 0.91 0.10 (n = 3)), as also observed for KA at
GluA2-LBD (Ki = 6.3 1.5 μM, nH = 1.14 0.06 (n = 3)) The
side chain comprising the hydroxamic acid functionality was
built to point toward residues Tyr471 and Pro499. However,
alternative conformations of 2i with the hydroxamic acid
moiety pointing toward residues Glu423 and Ser673 might also
be present, but because of weak electron density, alternative
conformations could not be modeled satisfactorily (see
Supporting Information).
A careful examination of the crystal structure reveals that the
amino acid part of 2i forms almost identical contacts with
surrounding residues (Pro499, Thr501, Arg506, Ser675,
Thr676, and Glu726; full-length numbering including signal
peptide) and water molecules (Figure 2B), as are observed in
the structure of GluA2-LBD in complex with Glu (PDB code
1FTJ).25 However, one contact within 3.5 Å is not seen for 2i;
the γ-carboxylate group of 2i forms hydrogen bonds to two
water molecules, whereas Glu establishes contact with three
water molecules. This difference can be explained by the partial
domain closure induced by 2i, which allows the water molecule
to move further away from the Glu-like part of 2i in order to
maintain contacts to protein backbone atoms. The distal
carboxylate group forms strong hydrogen bonding interaction
with the NH and the side chain OH of Thr676 and two water
molecules of which one is bridged to the hydroxamic acid
carbonyl group. The hydroxamic acid moiety of 2i is
accommodated in the binding site near residues Glu423,
Tyr426, Tyr471, Pro499, Thr707, Thr728, Met729, and
Tyr753. The carbonyl oxygen accepts two hydrogen bonds
from nearby water molecules, whereas the N-hydroxyl group is
in close contact with the hydroxyl group of Tyr471.
a
Data taken from ref 10.
GluA2-LBD compared to structures with Glu and KA (Figure
2C).25
Structure of GluK3-LBD with 2i. The structure of GluK3-
LBD in complex with 2i was solved at 2.65 Å resolution (Figure
2A). The complex crystallized in space group P4122 and
contains one molecule in the asymmetric unit of the crystal (for
further information on data collection and refinement statistics,
see Supporting Information). The binding affinity of 2i at the
GluK3-LBD construct was determined to be in the high
nanomolar range (Ki = 900 230 nM, nH = 0.93 0.08 (n =
An analysis of the side chain conformation in residue Met729
in GluA2-LBD disclosed it to adapt distinct conformations
depending on the bound agonist.26 This is also observed in the
GluA2-LBD:2i complex, where the hydroxamic acid moiety of
2i causes Met729 to adopt a different conformation in the
1619
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
Figure 2. Structures of GluA2-LBD and GluK3-LBD, both in complex with 2i. (A) The structures of GluA2-LBD and GluK3-LBD are shown in
cartoon representation (superimposed on lobe D1 residues) to illustrate the different degree of domain closure induced by 2i in GluA2-LBD and
GluK3-LBD. GluA2-LBD is shown in dark pink and GluK3-LBD in dark green. 2i in both structures is dark gray. (B) Zoom on the ligand-binding
site of GluA2-LBD with 2i and 2Fo − Fc omit map at 1σ carved around the ligand at 1.4 Å radius. Potential hydrogen bonds within 3.2 Å to water
molecules (red spheres) and surrounding residues are shown as black, dashed lines. (C) Overlay of GluA2-LBD in complex with 2i, glutamate (cyan,
PDB code 1FTJ),24 and KA (light gray, PDB code 1FW0).24 The differing amino acids Leu671 and Met729 of GluA2 are shown as sticks. (D) The
ligand-binding site of GluK3-LBD with 2i and 2Fo − Fc omit map at 1σ carved around the ligand at 1.6 Å radius. Potential hydrogen bonds within 3.2
Å to water molecules and surrounding residues are shown. (E) Structures of GluK3-LBD in complex with 2i and glutamate (in cyan, PDB code
3S9E),26 with focus on Asn722 shown in stick representation. Structures of GluK1-LBD and GluK2-LBD both in complex with neodysiherbaine are
shown in yellow and light gray, respectively (PDB codes 2ZNU and 3QXM).27 The residues corresponding to those of Asn722 are shown in sticks.
4)), which is approximately 10-fold lower than the binding
affinity of 2i at full-length GluK3 (123 nM). This trend has
previously been observed for the endogenous agonist Glu (7.4
μM versus 0.50 μM)27 and the conformationally restricted Glu
analogue (2S,1′R,2′S)-2-(2′-carboxycyclobutyl)glycine (CBG-
IV) (4.2 μM versus 0.33 μM).28 Similarly, a comparison of
binding affinities of agonists and antagonists at GluA2-LBD and
full-length GluA2 tabulated in Pøhlsgaard et al.25 shows that the
GluA2-LBD/full-length ratio varies between 0.3 and 5.2 for
agonists and 0.8−28.3 for antagonists. The origin for these
differences is yet to be fully understood.
Similar to Glu, the ligand 2i induces full domain closure of
the GluK3-LBD (25.2° and 24.3° for Glu and 2i, respectively)
(Figure 2A). In comparison, 2i and KA induce partial domain
1620
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
closure when crystallized in the GluA2-LBD. This difference
may be due to an otherwise buildup of a steric clash between
the terminal side chain carbon of Leu671 and the 3′-carbon
atom of 2i or the isopropenyl group of KA in GluA2-LBD
(Figure 2C). The corresponding residue in GluK3 is the less
bulky Val687, which strengthens this hypothesis, as full domain
closure of the GluK3-LBD with 2i or KA bound is indeed
observed.29 In further support of this hypothesis, it has
previously been shown that mutation of Leu671 to the less
bulky amino acid Thr in GluA230 and Val in GluA431 leads to
receptors with improved potency and efficacy of KA.
Excess electron density outside the Glu scaffold could clearly
be identified in the binding site of GluK3, corresponding to 2i
(Figure 2D). Because of the resolution (2.65 Å) of the GluK3-
LBD:2i structure, only two water molecules could be located in
the binding site. Similar interactions are observed between the
amino acid part of 2i and the nearby residues (Pro518, Thr520,
Arg525, Ala691, Thr692, and Glu739; full-length numbering
including signal peptide), as was observed in the GluK3-
LBD:Glu structure (PDB code 3S9E).27 The hydroxamic acid
moiety is accommodated in the binding pocket by displacement
of two water molecules (W1 and W2 in GluK3-LBD with Glu).
Similar to the structure of the GluA2-LBD:2i complex, the
hydroxyl group is in the vicinity of Tyr491 (2.6 Å). The
carbonyl oxygen only establishes contact with one visible water
molecule within 3.5 Å (Figure 2D).
EXPERIMENTAL SECTION
Chemistry. Melting points were determined on a Reichert hot-
■
stage apparatus and are uncorrected. IR spectra were recorded on a
Perkin-Elmer 801 spectrophotometer. H and 13C NMR spectra were
1
recorded on a Bruker Avance 400 MHz spectrometer. Chemical shifts
are reported in ppm (δ) relative to TMS as internal standard. HRMS
were recorded on a q-tof Micromass spectrometer. Optical rotations
were determined with a JASCO DIP 370 polarimeter and are reported
at the sodium D line (589 nm). Elemental analyses were performed at
the Service Central d’Analyze du CNRS, Solaize, France. Silicagel 60
(Merck, 40−63 μm) and precoated F254 plates were used for column
and TLC chromatography. HPLC analyses were performed using a
C18 column (Uptisphere ODB, 5 μm, 250 mm × 4.6 mm) in the
following conditions: flow rate 0.5 (2o, 2e) or 0.8 (2i) mL·min−1; UV
detection (210−214 nm); 30 °C; eluent consisting of H2O/CH3CN
95:5 (2i, 2o) or 20:80 (2e). All solvents were purified by distillation
following usual procedures. Cysteine sulfinic acid was prepared from
cystine following a described procedure.12 Bovine heart malic
dehydrogenase and rabbit muscle lactic dehydrogenase were purchased
from Sigma. E. coli AspAT was produced and purified following
described procedures from overexpressing E. coli strains JM103
transformed with pUC119-aspC (AspAT).33,34 Enzyme kinetic
measurements were performed at 25 °C in 0.1 M potassium
phosphate buffer, pH 7.6, Asp (40 mM), NADH (0.2 mM), keto
acid substrate (0.1−10 mM), AspAT (0.05 UI), and malic
dehydrogenase (2 UI) in a total volume of 1 mL. Rates were
calculated from the OD linear decay at 340 nm using εNADH = 6220
cm−1 M−1. All compounds undergoing pharmacological character-
ization were of ≥95% purity determined by elemental analyses or
HPLC.
Compared to the structure of GluK3-LBD:Glu and GluK3-
LBD:CBG-IV28 complexes, the side chain of Asn722 adopts a
different conformation, pointing away from the hydroxamic
acid moiety of 2i in the GluK3-LBD:2i complex. A similar
alteration in the conformation of Asn722 is observed in the
structure of GluK2-LBD (Asn721 in GluK2) in complex with
neodysiherbaine (Figure 2E).32 The high binding affinity of 2i
at GluK1 (Ki = 3.4 nM) and the selectivity versus the GluK2
and GluK3 subtypes (∼200-fold and ∼35-fold, respectively) are
remarkable. On the basis of the GluK3-LBD structure with 2i
and the GluK2 structure with neodysiherbaine, it can be
rationalized that a key factor in the origin of this selectivity may
be due to the less bulky Ser residue in GluK1 compared to Asn
in GluK2 and GluK3 (Figure 2E).
Dimethyl 4-(Isobutoxycarbonyloxy)cyclohex-3-ene-1,3-di-
carboxylate 4. To a solution of 316 (3.8 g, 17.7 mmol) in anhydrous
toluene (60 mL) were added Et3N (3.7 mL, 26.6 mmol) and isobutyl
chloroformate (2.55 mL, 19.6 mmol). The mixture was stirred at room
temperature for 3 h. After filtration and concentration under reduced
pressure, 4 was isolated as a slightly yellow oil in quantitative yield (5.6
1
g). IR (neat film) 1731, 1667 cm−1; H NMR (400 MHz, CDCl3) δ
3.96 (2H, d, J = 6.7 Hz), 3.69 (3H, s), 3.68 (3H, s), 2.74 (1H, m), 2.58
(2H, m), 2.38 (2H, m), 2.08 (1H, m), 1.98 (1H, m), 1.83 (1H, m),
0.95 (6H, d, J = 6.7 Hz); 13C NMR (100 MHz, CDCl3) δ 174.5, 165.3,
154.9, 152.3, 116.3, 74.9, 52.0, 51.8, 38.3, 28.1, 27.8, 27.4, 24.4, 18.9;
HRMS (ES+) m/z 337.1263 ([M + Na]+ C15H22NaO7 requires
337.1257).
Isobutylcarbonic-7-methoxy-4-methoxycarbonyl-6,7-dioxo-
heptanoic Anhydride 5. A solution of enol carbonate 4 (1.5 g, 4.8
mmol) in anhydrous CH2Cl2 (40 mL) was treated at −70 °C with a
mixture of O2 and O3 at a rate of 10 L/h until saturation (blue
coloration of the solution). After 10 min, the excess ozone was
eliminated by oxygen bubbling. Dimethyl sulfide (0.35 mL, 4.7 mmol)
was added, and the reaction mixture was allowed to warm to room
temperature. The solution of anhydride 5 was used directly in the next
reactions. An aliquot was concentrated under reduced pressure for
analyses. IR (neat film) 1820, 1732 cm−1; 1H NMR (400 MHz,
CDCl3) δ 4.03 (2H, d, J = 6.6 Hz), 3.89 (3H, s), 3.68 (3H, s), 3.35
(1H, m), 2.97 (2H, m), 2.55 (2H, m), 2.07−1.89 (3H, m), 0.95 (6H,
d, J = 6.7 Hz); 13C NMR (100 MHz, CDCl3) δ 191.6, 173.9, 167.2,
160.7, 149.0, 75.7, 53.2, 52.3, 40.9, 38.9, 31.3, 27.7, 25.9, 18.8; HRMS
(ES+) m/z 369.1162 ([M + Na]+ C15H22NaO9 requires 369.1148).
General Procedure for the Synthesis of 6a−j. To the solution
of anhydride 5 in CH2Cl2 was added the nucleophile HNR1R2 (1.1
mol equiv). NH3 and HNMe2 were added as 0.5−2 M solutions in
THF. MeNHOH was added as a mixture of the corresponding
hydrochloride, DIEA (1.1 mol equiv), and DMSO (3 mL). H2NOBn
and H2NOtBu were both obtained as solutions in CH2Cl2 (50 mL)
from an aqueous solution of the corresponding hydrochlorides after
adjusting the pH to 10 with 1 M NaOH and extraction with CH2Cl2
(2 × 25 mL). The reaction mixture was stirred at room temperature
for 10 min to 2 h before concentration under reduced pressure. 6a−j
were purified by flash chromatography (eluent, cyclohexane−EtOAc,
5:5 to 2:8, v/v for 6a−c,e−g,j; CH2Cl2−MeOH, 98:2 to 97:3 for 6d
CONCLUSION
■
We have presented the chemoenzymatic enantioselective
synthesis of 14 new 2,4-syn substituted Glu analogues 2b−o.
A short and efficient synthetic approach was developed based
on readily available cyclic β-keto ester 3 to access racemic 4-
substituted ketoglutarate analogues. In a highly stereoselective
way, the obtained keto acids were converted to the
corresponding Glu analogues having the L-2,4-syn configu-
ration. Pharmacological characterization at the iGluRs and
EAAT1−3 disclosed new insight into the SAR of these
receptors and transporters, and notably analogue 2i was
identified as a high affinity GluK1 ligand displaying high
subtype and iGluR class selectivity. From the two X-ray
structures obtained, 2i was found to induce partial domain
closure in GluA2-LBD, whereas at GluK3-LBD full domain
closure was seen. In all, the data presented herein add
important information to the SAR of Glu ligands and will aid
future rational design of subtype selective Glu receptor and
transporter ligands.
1621
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
and 6h; EtOAc to EtOAc−MeOH, 9:1 for 6i) and isolated as colorless
oils (6a,b,d−j) or as a white solid (6c).
Dimethyl 2-(3-Benzyloxyamino-3-oxopropyl)-4-oxogluta-
rate 6h. 6h was isolated as a colorless oil from 4 (1.20 g, 3.82
mmol): yield 792 mg, 59% from 3. Two rotamers were identified on
Dimethyl 2-(3-Amino-3-oxopropyl)-4-oxoglutarate 6a. 6a
was isolated as a slightly yellow oil from 4 (3.30 g, 10.5 mmol):
yield 2.72 g, 84% from 3; IR (neat film) 3453, 3363, 3206, 1730, 1667
1
NMR spectra. IR (neat film) 3221, 1732, 1666 cm−1; H NMR (400
MHz, CDCl3) δ 8.39 and 8.02 (1H, 2 × s broad), 7.37 (5H, m), 4.89
(2H, m), 3.86 (3H, s), 3.64 (3H, s), 3.28 (1H, dd, J = 10.0 and 19.2
Hz), 2.91 (2H, m), 2.39 and 2.11 (2H, 2 × m), 1.94 (2H, m); 13C
NMR (100 MHz, CDCl3) δ 191.9, 174.4, 169.7, 160.8, 135.4, 129.3,
128.7, 78.2, 53.2, 52.2, 40.9, 39.2, 30.4, 27.0; HRMS (ES+) m/z
352.1388 ([M + H]+, C17H22NO7 requires 352.1396). Anal.
(C17H21NO7·0.5H2O) C, H, N.
1
cm−1; H NMR (400 MHz, CDCl3) δ 5.73 (2H, s, broad), 3.86 (3H,
s), 3.67 (3H, s), 3.31 (1H, dd, J = 10.1 and 20.0 Hz), 2.96 (1H, dd, J =
4.9 and 20.0 Hz), 2.97 (1H, m), 2.29 (2H, m), 1.95 (2H, m); 13C
NMR (100 MHz, CDCl3) δ 191.9, 174.3, 174.2, 160.8, 53.1, 52.2, 40.9,
39.2, 32.8, 26.9; HRMS (ES+) m/z 268.0798 ([M + Na]+,
C10H15NaNO6 requires 268.0797). Anal. (C10H15NO6) C, H, N.
Dimethyl 2-Oxo-4-(3-oxo-3-(prop-2-ynylamino)propyl)-
glutarate 6b. 6b was isolated as a colorless oil from 4 (728 mg,
2.32 mmol): yield 616 mg, 94% from 3; IR (solution in CCl4) 3461,
2268, 1736 cm−1; 1H NMR (400 MHz, CDCl3) δ 5.78 (1H, s, broad),
4.04 (2H, dd, J = 2.5 and 5.3 Hz), 3.87 (3H, s), 3.68 (3H, s), 3.32 (1H,
dd, J = 10.0 and 19.7 Hz), 2.96 (2H, m), 2.24 (1H, m), 2.23 (1H, t, J =
2.5 Hz), 1.98 (2H, m); 13C NMR (100 MHz, CDCl3) δ 192.0, 174.4,
171.4, 160.9, 79.5, 71.8, 53.3, 52.3, 41.1, 39.3, 33.5, 29.4, 27.2; HRMS
(ES+) m/z 306.0940 ([M + Na]+, C13H17NaNO6 requires 306.0954).
Anal. (C13H17NO6) C, H, N.
Dimethyl 2-(3-Hydroxy(methyl)amino-3-oxopropyl)-4-oxo-
glutarate 6i. 6i was isolated as a colorless oil from 4 (1.50 g, 4.77
mmol): yield 672 mg, 52% from 3. Two rotamers were identified on
1
NMR spectra. IR (neat film) 3266, 1732, 1625 cm−1; H NMR (400
MHz, CDCl3) δ 8.42 (1H, s broad), 3.85 (3H, s), 3.65 (3H, s), 3.32
and 3.20 (3H, 2 × s), 3.28 (1H, dd, J = 9.2 and 19.6 Hz), 2.95 (2H,
m), 2.51 and 2.38 (2H, 2 × m), 1.93 (2H, m); 13C NMR (100 MHz,
CDCl3) δ 192.07, 191.9, 174.8, 174.4, 173.1, 160.8, 53.2, 52.2, 40.9,
39.5, 39.2, 36.2, 29.4, 26.5; HRMS (ES+) m/z 276.1069 ([M + H]+,
C11H18NO7 requires 276.1083). Anal. (C11H17NO7·0.25H2O) C, H,
N.
Dimethyl 2-Oxo-4-(3-oxo-3-(2-phenylhydrazinyl)propyl)-
glutarate 6j. 6j was isolated as a colorless oil from 4 (1.50 g, 4.77
mmol): yield 932 mg, 58% from 3. Two rotamers were identified on
NMR spectra. IR (solution in CCl4) 3434, 3351, 1732, 1699, 1604
cm−1; 1H NMR (400 MHz, CDCl3) δ 8.12 and 7.37 (1H, 2 × s), 7.20
and 7.17 (2H, 2 × t, J = 8.0 Hz), 6.87 and 6.85 (1H, 2 × t, J = 7.6 Hz),
6.76 and 6.68 (2H, 2d, J = 8.4 Hz), 5.97 and 4.98 (1H, 2 × s, broad),
3.82 and 3.81 (3H, 2 × s), 3.64 and 3.56 (3H, 2 × s), 3.26 (1H, m),
2.89 (2H, m), 2.45 and 2.25 (2H, 2 × m), 2.00−1.80 (2H, m); 13C
NMR (100 MHz, CDCl3) δ 192.0, 191.9, 178.0, 174.6, 174.5, 172.3,
160.8, 148.0, 147.2, 129.5, 129.2, 121.2, 121.1, 113.5, 112.5, 53.2, 52.2,
52.1, 40.9, 40.8, 39.3, 39.2, 31.4, 28.4, 27.0, 26.3; HRMS (ES+) m/z
359.1201 ([M + Na]+, C16H20N2NaO6 requires 359.1219). Anal.
(C16H20N2O6) C, H, N.
Dimethyl 2-Oxo-4-(3-oxo-3-phenylaminopropyl)glutarate
6c. 6c was isolated as a white solid from 4 (1.00 g, 3.18 mmol):
yield 920 mg, 90% from 3; IR (neat film) 3374, 1724, 1686, 1598
1
cm−1; H NMR (400 MHz, CDCl3) δ 8.09 (1H, s, broad), 7.47 (2H,
d, J = 8.0 Hz), 7.24 (2H, t, J = 7.9 Hz), 7.03 (1H, t, J = 7.4 Hz), 3.81
(3H, s), 3.61 (3H, s), 3.28 (1H, dd, J = 10.2 and 19.9 Hz), 2.92 (2H,
m), 2.38 (2H, m), 1.97 (2H, m); 13C NMR (100 MHz, CDCl3) δ
192.0, 174.5, 170.4, 160.7, 137.9, 128.9, 124.3, 120.0, 53.1, 52.2, 40.9,
39.2, 34.5, 27.1; HRMS (ES+) m/z 322.1281 ([M + H]+, C16H20NO6
requires 322.1291). Anal. (C16H19NO6) C, H, N.
Dimethyl 2-(3-Dimethylamino-3-oxopropyl)-4-oxoglutarate
6d. 6d was isolated as a slightly yellow oil from 4 (1.50 g, 4.77 mmol):
1
yield 1.15 g, 88% from 3; IR (neat film) 1731, 1644 cm−1; H NMR
(400 MHz, CDCl3) δ 3.84 (3H, s), 3.64 (3H, s), 3.29 (1H, dd, J = 10.0
and 19.6 Hz), 2.99−2.90 (2H, m), 2.96 (3H, s), 2.90 (3H, s), 2.34
(2H, t, J = 7.2 Hz), 1.85−1.99 (2H, m); 13C NMR (100 MHz, CDCl3)
δ 191.9, 174.6, 171.7, 160.9, 53.1, 52.0, 41.0, 39.3, 37.1, 35.5, 30.4,
26.8; HRMS (ES+) m/z 274.1280 ([M + H]+, C12H20NO6 requires
274.1291). Anal. (C12H19NO6·0.25H2O) C, H, N.
Dimethyl 2-(3-Diethylamino-3-oxopropyl)-4-oxoglutarate
6e. 6e was isolated as a colorless oil from 4 (1.00 g, 3.18 mmol):
yield 526 mg, 54% from 3; IR (solution in CCl4) 1761, 1741, 1667
cm−1; 1H NMR (400 MHz, CDCl3) δ 3.84 (3H, s), 3.64 (3H, s), 3.28
(5H, m), 2.95 (2H, m), 2.33 (2H, t, J = 7.2 Hz), 1.95 (2H, m), 1.14
(3H, t, J = 7.1 Hz), 1.07 (3H, t, J = 7.1 Hz); 13C NMR (100 MHz,
CDCl3) δ 192.0, 174.7, 170.7, 160.9, 53.1, 52.0, 41.9, 41.0, 40.3, 39.4,
30.2, 27.0, 14.3, 13.1; HRMS (ES+) m/z 302.1589 ([M + H]+,
C14H24NO6 requires 302.1604). Anal. (C14H23NO6·0.25H2O) C, H,
N.
Dimethyl 2-(3-Azido-3-oxopropyl)-4-oxoglutarate 7. To a
solution of anhydride 5 (prepared from 1 g of 4, 3.18 mmol) in THF
(15 mL) at −10 °C was added a solution of NaN3 (0.191 g, 3.18
mmol) in H2O (1 mL). The mixture was stirred at −10 °C for 15 min
and then at room temperature for 1 h. The reaction mixture was
diluted with EtOAc (40 mL), dried over MgSO4, and concentrated
under reduced pressure to give quantitatively a 80:20 mixture of acyl
azide 7 and isocyanate 8 isolated as a colorless oil. IR (neat film) 2273,
1
2141, 1732 cm−1; H NMR (400 MHz, CDCl3) δ 3.87 (3H, s), 3.68
(3H, s), 3.31 (1H, m), 3.05−2.90 (2H, m), 2.43 (2H, m), 2.05−1.85
(2H, m); 13C NMR (100 MHz, CDCl3) δ 191.6, 179.5, 174.0, 160.7,
53.1, 52.2, 40.7, 38.9, 34.1, 26.2.
Dimethyl 2-(2-Isocyanatoethyl)-4-oxoglutarate 8. A solution
of acyl azide 7 (prepared from 1 g of 4, 3.18 mmol) in CH2Cl2 (10
mL) was refluxed for 24 h. The solvent was removed under reduced
pressure to give 8 isolated quantitatively as a yellow oil. IR (neat film)
Dimethyl 2-Oxo-4-(3-oxo-3-(piperidin-1-yl)propyl)glutarate
6f. 6f was isolated as a colorless oil from 4 (1.50 mg, 4.77 mmol):
1
yield 1.25 g, 83% from 3; IR (neat film) 1729, 1646 cm−1; H NMR
1
2273, 1732 cm−1; H NMR (400 MHz, CDCl3) δ 3.87 (3H, s), 3.69
(400 MHz, CDCl3) δ 3.86 (3H, s), 3.66 (3H, s), 3.6−3.3 (4H, m),
3.31 (1H, dd, J = 10.1 and 19.7 Hz), 2.97 (2H, m), 2.36 (2H, m),
2.04−1.86 (2H, m), 1.63 (2H, m), 1.53 (4H, m); 13C NMR (100
MHz, CDCl3) δ 192.0, 174.6, 169.8, 160.9, 53.1, 52.0, 40.9, 39.4, 30.4,
27.0, 24.5; HRMS (ES+) m/z 314.1595 ([M + H]+, C15H24NO6
requires 314.1604). Anal. (C15H23NO6) C, H, N.
(3H, s), 3.42 (2H, t, J = 6.7 Hz), 3.31 (1H, dd, J = 7.9 and 18.0 Hz),
3.05 (1H, m), 2.97 (1H, dd, J = 4.9 and 18.3 Hz), 1.98 (1H, m), 1.82
(1H, m); 13C NMR (100 MHz, CDCl3) δ 191.6, 174.1, 160.9, 122.4,
53.3, 52.4, 40.8, 40.6, 37.3, 32.8.
1-Benzyl-2,4-dimethyl 2-Hydroxypiperidine-1,2,4-tricarbox-
ylate 9. To a solution of isocyanate 7 (0.73 g, 3 mmol) in toluene (50
mL) was added benzyl alcohol (0.31 mL, 3 mmol). The mixture was
stirred at 80 °C for 72 h before concentration under reduced pressure.
Flash chromatography (eluent, cyclohexane−EtOAc, 7:3) afforded
carbamate 9 isolated as a pale yellow oil (474 mg, 53% from 3). NMR
analyses showed 9 to be mainly present as one cyclic isomer (>85%).
IR (neat film) 3450, 1731, 1634 cm−1; 1H NMR (400 MHz, CDCl3) δ
7.33 (5H, m), 5.17−5.05 (2H, m), 4.32 (1H, s, broad), 3.92 (1H, m),
3.73 (3H, s), 3.68 (3H, s), 3.30 (1H, m), 2.86 (1H, m), 2.1−1.99 (3H,
m), 1.76 (1H, m); 13C NMR (100 MHz, CDCl3) δ 174.7, 173.1,
156.4, 135.8, 128.7, 128.4, 128.3, 82.4, 68.0, 53.5, 52.1, 41.3, 37.3, 35.1,
Dimethyl 2-Oxo-4-(3-oxo-3-tert-butoxyaminopropyl)-
glutarate 6g. 6g was isolated as a colorless oil from 4 (840 mg,
2.67 mmol): yield 0.50 g, 59% from 3. Two rotamers were identified
1
on NMR spectra. IR (neat film) 3200, 1732, 1664 cm−1; H NMR
(400 MHz, CDCl3) δ 8.04 and 7.88 (1H, 2 × s broad), 3.86 (3H, s),
3.67 (3H, s), 3.31 (1H, dd, J = 10.2 and 19.6 Hz), 2.95 (2H, m), 2.47
and 2.17 (2H, 2 × m), 1.95 (2H, m), 1.25 (9H, s); 13C NMR (100
MHz, CDCl3) δ 191.9, 174.4, 170.9, 160.8, 82.0, 53.1, 52.2, 41.0, 39.3,
30.7, 29.0, 27.3, 26.3; HRMS (ES+) m/z 318.1553 ([M + H]+,
C14H24NO7 requires 318.1553). Anal. (C14H23NO7·0.25H2O) C, H,
N.
1622
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
26.7; HRMS (ES+) m/z 374.1210 ([M + Na]+, C17H21NNaO7
requires 374.1216).
(1H, dd, J = 8.4 and 18.4 Hz), 2.87 (1H, dd, J = 5.6 and 18.4 Hz), 2.65
(1H, m), 2.22 (2H, m), 1.80 (2H, m), 1.25 (9H, s); 13C NMR (100
MHz, D2O) δ 204.3, 182.4, 173.4, 169.8, 83.9, 42.5, 41.8, 30.6, 28.2,
25.4; HRMS (ES+) m/z 290.1244 ([M − 2Li + 3H]+, C12H20NO7
requires 290.1240).
Dimethyl 2-Cyanoethyl-4-oxoglutarate 10. To a solution of 6a
(2.3 g, 9.38 mmol) in anhydrous THF (25 mL) at 0 °C were added
pyridine (1.52 mL, 18.8 mmol) and trifluoroacetic anhydride (1.44
mL, 10.2 mmol). The mixture was stirred at 0 °C for 20 min and then
at room temperature for 2 h. After addition of brine (15 mL), the
reaction mixture was extracted with EtOAc (2 × 25 mL). The
combined organic phases were dried over MgSO4 and concentrated
under reduced pressure. Flash chromatography (eluent, cyclohexane−
EtOAc, 1:1) afforded nitrile 10 (1.84 g, 87%) isolated as a colorless oil.
IR (neat film) 2220, 1731 cm−1; 1H NMR (400 MHz, CDCl3) δ 3.88
(3H, s), 3.71 (3H, s), 3.33 (1H, m), 3.03 (2H, m), 2.47 (2H, t, J = 8.0
Hz), 2.05 (1H, m), 1.92 (1H, m); 13C NMR (100 MHz, CDCl3) δ
191.3, 173.3, 160.7, 118.8, 53.4, 52.6, 40.5, 38.8, 27.2, 15.4; HRMS (ES
+) m/z 250.0698 ([M + Na]+, C10H13NNaO5 requires 250.0691).
Anal. (C10H13NO5) C, H, N.
General Procedure for the Synthesis of Lithium Oxogluta-
rates 1b−l. To a solution of 6b−j, 9, or 10 (1 mmol) in MeOH (5
mL) was added dropwise a freshly prepared 0.4 M solution of LiOH
(5.25 mL, 2.1 mmol). The mixture was stirred at room temperature for
16 h. After evaporation of MeOH, the pH of the aqueous solution was
adjusted to 7.6 by addition of Dowex 50WX8 resin (H+ form). The
resin was removed by filtration before evaporation of the water under
reduced pressure. 1a−l were isolated in quantitative yields as white or
yellow solids.
Dilithium 2-(3-Benzyloxyamino-3-oxopropyl)-4-oxogluta-
rate 1h. 1h was isolated as a yellow solid from 6h (634 mg, 1.80
1
mmol): yield 594 mg, 98%; H NMR (400 MHz, D2O) δ 7.44 (5H,
m), 4.88 (2H, s), 2.98 (1H, dd, J = 8.4 and 18.4 Hz), 2.78 (1H, dd, J =
5.6 and 18.4 Hz), 2.56 (1H, m), 2.10 (2H, t, J = 8.0 Hz), 1.67 (2H,
m); 13C NMR (100 MHz, D2O) δ 204.6, 182.3, 172.1, 169.8, 134.6,
129.8, 129.2, 128.7, 78.1, 42.3, 41.7, 30.4, 28.0; HRMS (ES+) m/z
290.1244 ([M − 2Li + 3H]+, C12H20NO7 requires 290.1240).
Dilithium 2-(3-Hydroxy(methyl)amino-3-oxopropyl)-4-oxo-
glutarate 1i. 1i was isolated as a yellow solid from 6i (430 mg,
1
1.56 mmol): yield 395 mg, 98%; H NMR (400 MHz, D2O) δ 3.36
and 3.22 (3H, 2 × s), 3.04 (1H, dd, J = 9.0 and 18.4 Hz), 2.87 (1H, dd,
J = 5.6 and 18.4 Hz), 2.66 (1H, m), 2.52 and 2.40 (2H, 2 × m), 1.85−
1.60 (2H, m); 13C NMR (100 MHz, D2O) δ 204.4, 182.8, 175.1,
169.7, 42.6, 42.0, 35.9, 29.6, 27.4; HRMS (ES+) m/z 254.0863 ([M −
Li + 2H]+, C9H13LiNO7 requires 254.0852).
Dilithium 2-Oxo-4-(3-oxo-3-(2-phenylhydrazinyl)propyl)-
glutarate 1j. 1j was isolated as a yellow solid from 6j (537 mg,
1
1.59 mmol): yield 491 mg, 96%; H NMR (400 MHz, D2O) δ 7.33
(2H, t, J = 7.6 Hz), 6.99 (1H, t, J = 7.6 Hz), 6.91 (2H, d, J = 7.6 Hz),
3.11 (1H, dd, J = 8.4 and 18.4 Hz), 2.91 (1H, dd, J = 5.6 and 18.4 Hz),
2.71 (1H, m), 2.38 (2H, m), 1.86 (2H, m); 13C NMR (100 MHz,
D2O) δ 204.3, 182.5, 176.1, 169.8, 147.3, 129.5, 121.2, 113.4, 42.6,
42.0, 31.8, 28.2; HRMS (ES+) m/z 315.1170 ([M − Li + 2H]+,
C14H16LiN2O6 requires 315.1168).
Dilithium 2-(2-Benzyloxycarbonylaminoethyl)-4-oxogluta-
rate 1k. 1k was isolated as a yellow solid from 9 (472 mg, 1.34
mmol): yield 440 mg, 98%. NMR analyses showed 1k to be mainly
present in D2O as a 7:3 mixture of the linear form and one cyclic form.
Linear form: 1H NMR (400 MHz, D2O) δ 7.39 (5H, m), 5.06 (2H, s),
3.10 (2H, m), 3.02 (1H, dd, J = 8.8 and 18.8 Hz), 2.83 (1H, dd, J = 5.6
and 18.8 Hz), 2.63 (1H, m), 1.75−1.55 (2H, m); 13C NMR (100
MHz, D2O) δ 204.3, 182.9, 169.7, 158.2, 136.5, 128.7, 128.3, 127.6,
66.8, 42.0, 40.7, 38.7, 31.9. Cyclic form: 1H NMR (400 MHz, D2O) δ
7.39 (5H, m), 5.10 (2H, s), 3.86 (1H, m), 3.20 (1H, m), 2.51 (1H, m),
1.95 (3H, m), 1.67 (1H, m); 13C NMR (100 MHz, D2O) δ 183.8,
178.2, 157.4, 136.1, 128.7, 128.3, 127.7, 84.2, 67.6, 41.8, 38.6, 37.9,
27.1; HRMS (ES−) m/z 322.0927 ([M − 2Li + H]−, C15H15LiNO7
requires 322.0927).
Dilithium 2-Oxo-4-(3-oxo-3-(prop-2-ynylamino)propyl)-
glutarate 1b. 1b was isolated as a yellow solid from 6b (518 mg,
1
2.29 mmol): yield 489 mg, 99%; H NMR (400 MHz, D2O) δ 3.95
(2H, s), 3.06 (1H, dd, J = 8.4 and 18.4 Hz), 2.87 (1H, dd, J = 5.8 and
18.4 Hz), 2.63 (1H, m), 2.28 (3H, m), 1.78 (2H, m); 13C NMR (100
MHz, D2O) δ 202.3, 180.5, 173.9, 167.8, 77.3, 69.7, 40.5, 39.9, 31.5,
26.6, 26.1; HRMS (ES+) m/z 262.0919 ([M − Li + 2H]+,
C11H13LiNO6 requires 262.0903).
Dilithium 2-Oxo-4-(3-oxo-3-phenylaminopropyl)glutarate
1c. 1c was isolated as a white solid from 6c (473 mg, 1.47 mmol):
1
yield 448 mg, 99%; H NMR (400 MHz, D2O) δ 7.43 (4H, m), 7.27
(1H, m), 3.11 (1H, dd, J = 8.5 and 18.5 Hz), 2.92 (1H, dd, J = 5.6 and
18.5 Hz), 2.72 (1H, m), 2.44 (2H, m), 1.89 (2H, m); 13C NMR (100
MHz, D2O) δ 204.4, 182.7, 174.9, 169.9, 136.8, 129.2, 125.6, 122.2,
42.7, 42.1, 34.5, 28.3; HRMS (ES+) m/z 300.1065 ([M − Li + 2H]+,
C14H15LiNO6 requires 300.1059).
Dilithium 2-(3-Dimethylamino-3-oxopropyl)-4-oxoglutarate
1d. 1d was isolated as a yellow solid from 6d (824 mg, 3.02 mmol):
1
Dilithium 2-Cyanoethyl-4-oxoglutarate 1l. 1l was isolated as a
yellow solid from 10 (1.20 g, 5.28 mmol): yield 1.10 g, 99%; 1H NMR
(400 MHz, D2O) δ 2.95 (1H, dd, J = 8.4 and 18.8 Hz), 2.77 (1H, dd, J
= 5.6 and 18.8 Hz), 2.59 (1H, m), 2.37 (2H, m), 1.73 (2H, m); 13C
NMR (100 MHz, D2O) δ 203.8, 181.8, 169.5, 121.2, 42.2, 41.7, 27.5,
14.8; HRMS (ES−) m/z 198.0392 ([M − 2Li + H]+, C8H8NO5
requires 198.0402).
yield 760 mg, 98%; H NMR (400 MHz, D2O) δ 3.05 (3H, s), 3.04
(1H, dd, J = 8.4 and 18.4 Hz), 2.90 (3H, s), 2.86 (1H, dd, J = 5.6 and
18.4 Hz), 2.65 (1H, m), 2.40 (2H, m), 1.74 (2H, m); 13C NMR (100
MHz, D2O) δ 204.4, 182.6, 175.5, 169.9, 42.6, 41.7, 37.6, 35.5, 30.9,
27.8; HRMS (ES+) m/z 252.1059 ([M − Li + 2H]+, C10H15LiNO6
requires 252.1059).
Dilithium 2-(3-Diethylamino-3-oxopropyl)-4-oxoglutarate
1e. 1e was isolated as a white solid from 6e (341 mg, 1.13 mmol):
yield 307 mg, 95%; 1H NMR (400 MHz, D2O) δ 3.35 (2H, q, J = 7.2
Hz), 3.30 (2H, q, J = 7.2 Hz), 3.03 (1H, dd, J = 8.4 and 18.4 Hz), 2.84
(1H, dd, J = 5.6 and 18.4 Hz), 2.63 (1H, m), 2.37 (2H, m), 1.73 (2H,
m), 1.15 (3H, t, J = 7.2 Hz), 1.05 (3H, t, J = 7.2 Hz); 13C NMR (100
MHz, D2O) δ 204.3, 182.6, 174.7, 169.8, 43.0, 42.7, 42.1, 40.7, 30.7,
28.3, 13.2, 12.0; HRMS (ES+) m/z 274.1303 ([M − 2Li + 3H]+,
C12H20NO6 requires 274.1291).
Dilithium 2-Oxo-4-(3-oxo-3-(piperidin-1-yl)propyl)glutarate
1f. 1f was isolated as a white solid from 6f (1.16 g, 3.70 mmol): yield
1.09 g, 99%; 1H NMR (400 MHz, D2O) δ 3.49 (4H, m), 3.07 (1H, dd,
J = 8.4 and 18.4 Hz), 2.88 (1H, dd, J = 5.6 and 18.4 Hz), 2.67 (1H,
m), 2.43 (2H, m), 1.75 (2H, m), 1.70−1.50 (6H, m); 13C NMR (100
MHz, D2O) δ 204.4, 182.6, 173.5, 169.9, 47.5, 43.4, 42.8, 42.1, 31.1,
28.3, 26.1, 25.3, 23.7; HRMS (ES+) m/z 292.1388 ([M − Li + 2H]+,
C13H19LiNO6 requires 292.1372).
General Procedure for the Synthesis of L-2,4-syn-Glu
Analogues 2b−l. To a solution of racemic 1b−l (0.5−5 mmol) in
water (25−250 mL) was added cysteine sulfinic acid (0.5−5 mmol).
The pH of the solution was adjusted to 7.6 with 1 M NaOH before the
addition of E. coli AspAT (10−200 units; the quantity of enzyme was
adjusted to attain 40% conversion in 2−4 h). The mixture was stirred
slowly at room temperature, and the reaction was monitored by
titration of pyruvate. The 5 μL aliquots of the reaction mixture were
added to 995 μL of 0.1 M potassium phosphate buffer, pH 7.6,
containing NADH (0.2 mM) and lactate dehydrogenase (1 unit).
Pyruvate concentration was calculated from the ΔOD measured at 340
nm using εNADH = 6220 M−1 cm−1. When a conversion rate of 40%
was reached, the reaction mixture containing 2b−f,k,l was rapidly
passed through a column of Dowex 50WX8 resin (H+ form, 25 mL).
The column was then washed with water (100 mL) until complete
elution of CSA and then eluted with 1 M NH4OH. The ninhydrin
positive fractions were combined and concentrated to dryness under
reduced pressure. The residue was diluted in water (5 mL) before
adsorption on a column of Dowex 1X8 resin (200−400 Mesh, AcO−
Dilithium 2-Oxo-4-(3-oxo-3-tert-butoxyaminopropyl)-
glutarate 1g. 1g was isolated as a white solid from 6g (400 mg,
1
1.26 mmol): yield 367 mg, 97%; H NMR (400 MHz, D2O) δ 3.05
1623
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
form, 1.5 × 12 cm). The column was washed with water (50 mL) and
then eluted with an AcOH gradient (0.1−1 M). The ninhydrin
positive fractions were combined and concentrated under reduced
pressure. The reaction mixtures containing 2g,h,j were purified directly
on Dowex 1X8 resin as described above. In the case of 2i, elution from
Dowex 50WX8 resin was done with 1 M ammonium bicarbonate and
172.1, 134.6, 129.9, 129.2, 128.7, 78.2, 53.3, 44.4, 32.9, 30.4, 28.3;
HRMS (ES-) m/z 323.1235 ([M − H]−, C15H19N2O6 requires
323.1243). Anal. (C15H20N2O6·0.25H2O) C, H, N.
(2S,4R)-4-(3-Hydroxy(methyl)amino-3-oxopropyl)glutamic
Acid 2i. 2i was isolated from 1i (211 mg, 0.82 mmol): yield 78 mg,
39%; mp 96 °C; [α]25D +8.0 (c 0.5, H2O). NMR analyses showed 2i to
1
−
exist as a 80:20 mixture of rotamers in D2O. H NMR (400 MHz,
elution from the Dowex 1X8 resin (HCO3 form) was done with an
D2O) δ 3.71 (1H, dd, J = 3.9 and 9.3 Hz), 3.37 and 3.23 (3H, 2 × s),
2.52 (2H, t, J = 8 Hz), 2.39 (1H, m), 2.18 (1H, ddd, J = 3.9, 9.8, and
14.2 Hz), 1.91 (1H, ddd, J = 4.4, 9.3, and 14.3 Hz), 1.88−1.66 (2H,
m); 13C NMR (100 MHz, D2O) δ 182.0, 175.1, 174.3, 53.3, 44.3, 38.7,
35.9, 33.1, 33.0, 30.2, 29.5, 27.8, 27.6; HRMS (ES−) m/z 247.0934
([M − H]−, C9H15N2O6 requires 247.0930). Anal. Calcd for
C9H16N2O6·0.75NH3·H2O: C, 38.74, H, 7.31 N, 13.80. Found: C,
38.81, H, 6.87, N, 14.17. 99.1% purity by HPLC analysis.
ammonium bicarbonate gradient (0.1−0.5 M). All the residues
obtained after chromatography were dissolved in water (5 mL), and
the solutions were lyophilized to afford 2b−l as white solids.
(2S,4R)-4-(3-Oxo-3-(prop-2-ynylamino)propyl)glutamic Acid
2b. 2b was isolated from 1b (254 mg, 0.95 mmol): yield 95 mg, 39%;
1
mp 110 °C; [α]25 +1.0 (c 1.5, H2O); H NMR (400 MHz, D2O) δ
D
3.95 (2H, d, J = 2.5 Hz), 3.79 (1H, dd, J = 5.4 and 8.3 Hz), 2.62 (1H,
m), 2.61 (1H, t, J = 2.4 Hz), 2.40−2.25 (3H, m), 2.00−1.86 (3H, m);
13C NMR (100 MHz, D2O) δ 178.6, 175.3, 173.6, 79.6, 71.7, 53.0,
41.5, 32.8, 32.3, 28.7, 27.8; HRMS (ES+) m/z 257.1149 ([M + H]+,
C11H17N2O5 requires 257.1137). Anal. (C11H16N2O5·0.5H2O) C, H,
N.
(2S,4R)-4-(3-Oxo-3-(2-phenylhydrazinyl)propyl)glutamic
Acid 2j. 2j was isolated from 1j (141 mg, 0.44 mmol): yield 53 mg,
39%; mp 136 °C; [α]25D +8.0 (c 0.6, MeOH). NMR analyses showed
1
2j to exist as 93:7 mixture of rotamers in D2O. H NMR (400 MHz,
D2O) δ 7.31 (2H, m), 6.97 (1H, m), 6.89 (2H, m), 3.86 and 3.77 (1H,
2 × dd, J = 6.0 and 8.0 Hz), 2.68 (1H, m), 2.53 and 2.42 (2H, 2 × m),
2.33 and 2.25 (1H, 2 × ddd, J = 5.8, 9.6, and 14.9 Hz), 2.03−1.83 (3H,
m); 13C NMR (100 MHz, D2O) δ 178.1, 175.4, 173.0, 147.2, 129.6,
129.5, 121.2, 120.9, 113.4, 112.6, 52.6, 41.2, 32.2, 30.9, 27.7; HRMS
(ES+) m/z 310.1408 ([M + H]+, C14H20N3O5 requires 310.1403).
Anal. (C14H19N3O5·1.5H2O) C, H, N.
(2S,4R)-4-(2-Benzyloxycarbonylaminoethyl)glutamic Acid
2k. 2k was isolated from 1k (410 mg, 1.23 mmol): yield 159 mg,
40%; mp 97 °C; [α]25D +8.1 (c 0.5, H2O); 1H NMR (400 MHz, D2O)
δ 7.43 (5H, m), 5.10 (2H, s), 3.74 (1H, dd, J = 5.6 and 7.7 Hz), 3.19
(2H, m), 2.62 (1H, m), 2.25 (1H, ddd, J = 5.2, 9.6, and 14.8 Hz),
1.99−1.71 (3H, m); 13C NMR (100 MHz, D2O) δ 179.0, 173.6, 158.3,
136.5, 128.8, 128.3, 127.6, 66.9, 53.0, 39.7, 38.1, 32.3, 31.7; HRMS
(ES−) m/z 323.1259 ([M − H]−, C15H19N2O6 requires 323.1243).
Anal. (C15H20N2O6·0.25H2O) C, H, N.
(2S,4R)-4-(2-Cyanoethyl)glutamic Acid 2l. 2l was isolated from
1l (960 mg, 4.55 mmol): yield 364 mg, 40%; mp 110 °C; [α]25D +15.8
(c 0.7, H2O); 1H NMR (400 MHz, D2O) δ 3.79 (1H, dd, J = 5.6 and
7.6 Hz), 2.68 (1H, m), 2.56 (2H, m), 2.28 (1H, ddd, J = 5.6, 9.6, and
15.0 Hz), 2.06−1.88 (3H, m); 13C NMR (100 MHz, D2O) δ 177.8,
173.4, 120.8, 52.8, 41.2, 32.1, 27.3, 14.4; HRMS (ES−) m/z 199.0726
([M − H]−, C8H11N2O4 requires 199.0719). Anal. (C8H12N2O4) C,
H, N.
(2S,4R)-4-(3-Hydroxyamino-3-oxopropyl)glutamic Acid 2m.
To a solution of 2h (60 mg, 0.185 mmol) in H2O (10 mL) was added
10% Pd/C (40 mg). The mixture was stirred for 2 h under hydrogen
atmosphere, using a balloon. Pd/C was removed by filtration through
a membrane (0.2 μm), and the filtrate was concentrated under reduced
pressure. Ion exchange chromatography on Dowex 1X8, performed as
(2S,4R)-4-(3-Oxo-3phenylaminopropyl)glutamic Acid 2c. 2c
was isolated from 1c (215 mg, 0.70 mmol): yield 85 mg, 41%; mp 172
1
°C; [α]25 −44.6 (c 0.8, MeOH); H NMR (400 MHz, D2O) δ 7.38
D
(4H, m), 7.20 (1H, m), 3.70 (1H, dd, J = 3.4 and 9.0 Hz), 2.45−2.3
(3H, m), 2.18 (1H, m), 1.95−1.75 (3H, m); 13C NMR (100 MHz,
D2O) δ 182.3, 174.7, 174.4, 136.7, 129.1, 125.6, 122.1, 53.3, 44.6, 34.3,
33.1, 28.5; HRMS (ES+) m/z 295.1286 ([M + H]+, C14H19N2O5
requires 295.1294). Anal. (C14H18N2O5) C, H, N.
(2S,4R)-4-(3-Dimethylamino-3-oxopropyl)glutamic Acid 2d.
2d was isolated from 1d (211 mg, 0.82 mmol): yield 85 mg, 42%; mp
1
145 °C; [α]25 −7.0 (c 0.7, MeOH); H NMR (400 MHz, D2O) δ
D
3.81 (1H, dd, J = 5.2 and 8.0 Hz), 3.07 (3H, s), 2.93 (3H, s), 2.65 (1H,
m), 2.49 (2H, t, J = 7.6 Hz), 2.29 (1H, ddd, J = 5.6, 9.6, and 14.8 Hz),
2.00−1.85 (3H, m); 13C NMR (100 MHz, D2O) δ 178.7, 174.9, 173.6,
53.0, 41.6, 37.5, 35.5, 32.4, 30.2, 27.5; HRMS (ES−) m/z 245.1143
([M − H]−, C10H17N2O5 requires 245.1137). Anal. (C10H18N2O5·
H2O) C, H, N.
(2S,4R)-4-(3-Diethylamino-3-oxopropyl)glutamic Acid 2e. 2e
was isolated from 1e (131 mg, 0.46 mmol): yield 44 mg, 35%; mp 145
1
°C; [α ]25 +42.1 (c 1, 6N HCl); H NMR (400 MHz, D2O) δ 3.79
D
(1H, dd, J = 5.5 and 8.3 Hz), 3.39 (2H, q, J = 7.2 Hz), 3.34 (2H, q, J =
7.2 Hz), 2.66 (1H, m), 2.47 (2H, m), 2.30 (1H, ddd, J = 5.5, 9.6, and
14.9 Hz), 2.00−1.85 (3H, m), 1.17 (3H, t, J = 7.2 Hz), 1.09 (3H, t, J =
7.2 Hz); 13C NMR (100 MHz, D2O) δ 178.7, 173.9, 173.5, 52.9, 42.9,
41.6, 40.8, 32.4, 29.9, 27.8, 13.1, 12.0; HRMS (ES+) m/z 297.1418
([M + Na]+, C12H22N2NaO5 requires 297.1426). Anal. Calcd for
C12H22N2O5·0.5H2O: C, 50.87, H, 8.18, N, 9.88. Found: C, 50.98, H,
7.75, N, 9.92. 95.4% purity by HPLC analysis.
(2S,4R)-4-(3-Oxo-3-(piperidin-1-yl)propyl)glutamic Acid 2f.
2f was isolated from 1f (255 mg, 0.86 mmol): yield 86 mg, 35%; mp
170 °C; [α]25D +41.2 (c 0.8, 6 N HCl); 1H NMR (400 MHz, D2O) δ
3.79 (1H, dd, J = 5.4 and 8.3 Hz), 3.49 (4H, m), 2.64 (1H, m), 2.49
(2H, t, J = 7.8 Hz), 2.30 (1H, ddd, J = 5.4, 9.6, and 14.9 Hz), 2.00−
1.85 (3H, m), 1.70−1.50 (6H, m); 13C NMR (100 MHz, D2O) δ
178.6, 173.5, 172.8, 52.9, 47.3, 43.4, 41.6, 32.4, 30.2, 25.9, 25.2, 23.6;
HRMS (ES+) m/z 309.1430 ([M + Na]+, C13H22N2NaO5 requires
309.1426). Anal. (C13H22N2O5) C, H, N.
described for 2h, afforded 2m (37 mg, 85%) as a pale yellow solid. Mp
1
87 °C; [α]25 +5.4 (c 0.5, H2O); H NMR (400 MHz, D2O) δ 3.79
D
(1H, dd, J = 5.4 and 8.2 Hz), 2.61 (1H, m), 2.34−2.17 (3H, m), 2.00−
1.85 (3H, m); 13C NMR (100 MHz, D2O) δ 178.6, 173.6, 172.0, 53.0,
41.5, 32.3, 29.8, 27.7; HRMS (ES−) m/z 233.0773 ([M − H]−,
C8H13N2O6 requires 233.0774). Anal. (C8H14N2O6·0.5H2O) C, H, N.
(2S,4R)-4-(2-Aminoethyl)glutamic Acid 2n. To a solution of 2k
(87 mg, 0.27 mmol) in H2O (5 mL) was added 10% Pd/C (30 mg).
The mixture was stirred for 3 h at room temperature under hydrogen
atmosphere, using a balloon. Pd/C was removed by filtration through
a membrane (0.2 μm), and the filtrate was concentrated under reduced
pressure. The residue was dissolved in H2O (2 mL) and the solution
poured on a column of Dowex 1X8 (200−400 mesh, AcO− form, 1.5
cm × 10 cm). The column was eluted with H2O, and the ninhydrin
positive fractions were combined and concentrated under reduced
pressure. The residue was dissolved in H2O and the solution was
(2S,4R)-4-(3-Oxo-3-tert-butoxyaminopropyl)glutamic Acid
2g. 2g was isolated from 1b (300 mg, 1.00 mmol): yield 116 mg,
1
40%; mp 137 °C; [α]25 +7.6 (c 0.7, MeOH); H NMR (400 MHz,
D
D2O) δ 3.79 (1H, dd, J = 5.5 and 8.2 Hz), 2.61 (1H, m), 2.28 (3H, m),
1.92 (3H, m), 1.24 (9H, s); 13C NMR (100 MHz, D2O) δ 178.3,
173.6, 172.9, 84.3, 52.9, 41.5, 32.7, 30.3, 28.2, 25.3; HRMS (ES+) m/z
291.1546 ([M + H]+, C12H23N2O6 requires 291.1556). Anal.
(C12H22N2O6·H2O) C, H, N.
(2S,4R)-4-(3-Benzyloxyamino-3-oxopropyl)glutamic Acid
2h. 2h was isolated from 1h (246 mg, 0.73 mmol): yield 100 mg,
lyophilized to afford 2n as a slightly yellow solid (50 mg, 97%). Mp 98
°C; [α]25 −37.5 (c 0.6, H2O); H NMR (400 MHz, D2O) δ 3.74
1
1
42%; mp 137 °C; [α]25 −5.1 (c 0.8, MeOH); H NMR (400 MHz,
D
D
D2O) δ 7.48 (5H, m), 4.91 (2H, s), 3.67 (1H, dd, J = 3.8 and 9.3 Hz),
2.30 (1H, m), 2.18−2.05 (3H, m), 1.86 (1H, ddd, J = 4.0, 9.2, and 13.6
Hz), 1.82−1.65 (2H, m); 13C NMR (100 MHz, D2O) δ 182.0, 174.4,
(1H, dd, J = 4.3 and 8.0 Hz), 3.00 (2H, t, J = 7.9 Hz), 2.41 (1H, m),
2.16 (1H, ddd, J = 4.3, 9.2, and 14.7 Hz), 2.00−1.77 (3H, m); 13C
NMR (100 MHz, D2O) δ 181.5, 174.3, 53.2, 42.4, 37.8, 32.9, 29.9;
1624
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
HRMS (ES−) m/z 189.0883 ([M − H]−, C7H13N2O4 requires
189.0875). Anal. (C7H14N2O4·0.75H2O) C, H, N.
isolation of a fraction (60 mg, 40%) with a diastereomeric excess over
1
98% as judged by H NMR. 2p was isolated as a white solid. Mp 70
1
(2S,4R)-4-(2-Cyanoethyl)-N-tert-butoxycarbonylglutamic
Acid 11. To a solution of 2l (200 mg, 1 mmol) in H2O (6 mL) was
added NaHCO3 (504 mg, 6.0 mmol). The solution was cooled to 0
°C, and a solution of BOC2O (0.55 g, 2.5 mmol) in 1,4-dioxane (2.5
mL) was added. The mixture was stirred at room temperature for 72 h
°C; [α]25 −29.2 (c 0.5, H2O); H NMR (400 MHz, D2O) δ 3.82
D
(1H, d, J = 11.2 Hz), 3.30 (2H, m), 2.76 (1H, m), 2.62 (1H, m), 2.16
(1H, m), 2.01 (2H, m), 1.93−1.70 (2H, m); 13C NMR (100 MHz,
D2O) δ 179.3, 173.8, 59.5, 44.8, 43.4, 31.2, 29.2, 22.6; HRMS (ES−)
m/z 186.0763 ([M − H]−, C8H12NO4 requires 186.0766). Anal.
(C8H13NO4·0.5H2O) C, H, N.
−
and then poured on a column of Dowex 1X8 (200−400 mesh, HCO3
form, 1.5 cm × 5 cm). The column was washed with water (25 mL)
and then eluted with an ammonium bicarbonate gradient (0.1−1 M).
The ninhydrin positive fractions were combined and concentrated
under reduced pressure to afford 11 (270 mg, 90%) as a white solid.
NMR analyses showed 11 to exist as a 1:3 mixture of rotamers in D2O.
Mp 87 °C; [α]25D +1.0 (c 2, H2O); 1H NMR (400 MHz, D2O) δ 3.76
(1H, m), 2.53−2.32 (3H, m), 1.98 (1H, m), 1.80 (2H, m), 1.61 (1H,
m), 1.40 (9H, s); 13C NMR (100 MHz, D2O) δ 183.3, 180.0, 157.8,
157.6, 121.3, 56.1, 55.0, 45.2, 34.6, 34.3, 28.2, 27.7, 27.5, 14.9; HRMS
(ES−) m/z 299.1253 ([M − H]−, C13H19N2O6 requires 299.1243).
(2S,4R)-4-(3-Aminopropyl)-N-tert-butoxycarbonylglutamic
Acid 12. To a solution of 11 (270 mg, 0.9 mmol) in MeOH (10 mL)
and AcOH (10 mL) was added 20% Pd(OH)2/C (50 mg). The
mixture was stirred for 4 h at room temperature under hydrogen
atmosphere, using a balloon. Pd(OH)2/C was removed by filtration
through a membrane (0.2 μm), and the filtrate was concentrated under
reduced pressure. The residue was diluted in water (5 mL) before
adsorption on a column of Dowex 1X8 resin (200−400 mesh, AcO−
form, 1.5 cm × 6 cm). The column was washed with water (25 mL)
and then eluted with an AcOH gradient (0.1−1 M). The ninhydrin
positive fractions were combined and concentrated under reduced
pressure. The residue was dissolved in H2O (5 mL) and the solution
Binding Pharmacology. Native iGluRs. All native receptor
binding experiments were performed using rat brain synaptic
membranes of cortex and the central hemispheres from male SPRD
rats with tissue preparation as described in the literature.35
Characterization at AMPA, KA, and NMDA receptors was determined
using 5 nM [3H]AMPA,36 5 nM [3H]KA,37 and 2 nM [3H]CGP
3965338 with some modifications: On the day of the assay, frozen
membranes were quickly thawed and homogenized in 50 volumes of
ice-cold 30 mM Tris-HCl buffer, pH 7.4, containing 2.5 mM CaCl2
(for [3H]AMPA binding), 50 mM Tris-HCl buffer, pH 7.4 (for
[3H]KA binding), or 50 mM Tris-HCl buffer, pH 7.4, containing 2.5
mM CaCl2 (for [3H]CGP 39653 binding), before centrifugation
(48000g for 10 min). This washing step was repeated four times. In
the [3H]AMPA binding experiment, 100 mM KSCN was added to the
buffer during the final wash and during incubation. The final pellet was
resuspended in ice-cold buffer corresponding to approximately 0.4−
0.5 mg of protein/mL. The binding was carried out in 96-wells plates
by incubation of 200 μL of membrane suspension, 25 μL of
[3H]ligand, and 25 μL of test substance in various concentrations at
0 °C for 30 min ([3H]AMPA binding) or for 60 min ([3H]KA and
[3H]CGP binding). The reaction was terminated by rapid filtration
through GF/C filters (Perkin-Elmer Life Sciences), using a 96-well
Packard FilterMate cell harvester, followed by washing with 3 × 250
μL of ice-cold binding buffer. To the dried filters was added Microscint
scintillation fluid (Perkin-Elmer Life Sciences), and the amount of
filter bound radioactivity was quantified in a Packard TopCount
microplate scintillator counter. The experiments were performed in
triplicate at least three times for each compound. Nonspecific binding
was determined using 1.0 mM Glu. The binding data were analyzed by
a nonlinear regression curve-fitting procedure using GraphPad Prism,
version 5.04 (GrapPad Software, CA, U.S.).
Recombinant iGluRs. Radioligand binding assays were performed
as previously described at full length recombinant rat iGluRs:
GluA2(R)o, GluK1(Q)1b, GluK2(VCR)a, GluK3a, expressed in Sf9
cell membranes.39 (RS)-[3H]AMPA (45.8 Ci/mmol, PerkinElmer,
Waltham, MA) was used as the radioligand for GluA2(R)o (Kd = 16.8
nM). [3H]-(2S,4R)-4-Methylglutamic acid (40 Ci/mmol, ARC, St.
Louis, MO) was used as the radiolabel for GluK1−3: GluK1(Q)1b, Kd
= 0.66 nM; GluK2(VCR)a, Kd = 17 nM; GluK3a, Kd = 5.7 nM.
Soluble iGluR LBDs. Binding of 2i was carried out at the LBD
constructs as previously described.40,41 GluK3 LBD protein (50 ng)
was incubated with 5−10 nM [3H]-(2S,4R)-4-methylglutamic acid for
1 h at 4 °C in assay buffer (50 mM Tris-HCl, 10% (v/v) glycerol, pH
7.1 at 4 °C); GluK3 LBD Kd = 63 nM.26 Nonspecific binding was
determined in the presence of 1 mM (S)-glutamate. Competition
experiments were carried out in the presence of 16 concentrations of
2i (10 nM to 0.25 mM) in triplicate. Samples were filtered onto
Whatman (Dassel, Germany) 0.2 μm ME 24 mixed cellulose ester
filters and were washed twice with 1.5 mL of ice-cold assay buffer.
Filters were dried at 70 °C for 1 h, and 1.5 mL of Filter-Count
(PerkinElmer) scintillation fluid added. After dissolution of the filters
radioactivity was determined by liquid scintillation counting (TriCarb
2900, PerkinElmer).
was lyophilized to afford 12 as a white solid (256 mg, 93%). Mp 127
1
°C; [α]25 +0.8 (c 1, H2O); H NMR (400 MHz, D2O) δ 3.95 (1H,
D
dd, J = 4.2 and 9.4 Hz), 3.00 (2H, t, J = 6.2 Hz), 2.48 (1H, m), 2.14
(1H, m), 1.77 (1H, m), 1.72−1.56 (4H, m), 1.43 (9H, s); 13C NMR
(100 MHz, D2O) δ 180.5, 178.2, 157.4, 157.6, 81.1, 53.8, 42.8, 39.2,
33.9, 28.9, 27.7, 24.7; HRMS (ES−) m/z 303.1547 ([M − H]−,
C13H23N2O6 requires 303.1556).
(2S,4R)-4-(3-Aminopropyl)glutamic Acid 2o. A solution of 12
(125 mg, 0.41 mmol) in 1 M HCl (5 mL) was stirred at room
temperature for 3 h before concentration under reduced pressure. The
residue was dissolved in H2O (10 mL) and the pH adjusted to 12 with
0.2 M NaOH. The solution was poured on a column of Dowex
50WX8 resin (H+ form, 1.5 cm × 10 cm). The column was washed
with H2O (50 mL) and then eluted with 1 M NH3. The ninhydrin-
positive fractions were combined and concentrated under reduced
pressure. The residue was dissolved in H2O (2 mL) and the solution
poured on a column of Dowex 1X8 (200−400 mesh, AcO− form, 1.5
cm × 10 cm). The column was eluted with H2O, and the ninhydrin-
positive fractions were combined and concentrated under reduced
pressure. The residue was dissolved in H2O (4 mL) and the solution
was lyophilized to afford 2o (73 mg, 80%) as a white solid. Mp 140
°C; [α]25D −9.0 (c 0.6, H2O); 1H NMR (400 MHz, D2O) δ 3.72 (1H,
dd, J = 4.3 and 8.4 Hz), 3.00 (2H, t, J = 7.1 Hz), 2.38 (1H, m), 2.14
(1H, ddd, J = 4.3, 9.0, and 14.1 Hz), 1.92 (1H, ddd, J = 4.9, 8.4, and
14.0 Hz), 1.72−1.50 (4H, m); 13C NMR (100 MHz, D2O) δ 182.7,
174.4, 53.3, 44.4, 39.2, 33.0, 29.0, 24.8; HRMS (ES−) m/z 203.1045
([M − H]−, C8H15N2O4 requires 203.1032). Anal. Calcd for
C8H16N2O4·1.5H2O: C, 41.55, H, 8.28, N, 12.11. Found: C, 41.62,
H, 7.65, N, 12.03. 96.9% purity by HPLC analysis.
(2S,4R)-Azepane-2,4-dicarboxylic Acid 2p. To a solution of 2l
(160 mg, 0.8 mmol) in H2O (20 mL) was added 10% Pd/C (100 mg).
The mixture was stirred for 72 h at room temperature under hydrogen
atmosphere, using a balloon. Pd/C was removed by filtration through
a membrane (0.2 μm), and the filtrate was concentrated under reduced
pressure. The residue was dissolved in H2O (2 mL) and the solution
poured on a column of Dowex 1X8 (200−400 mesh, AcO− form, 1.5
cm × 15 cm). The column was first eluted with H2O to afford 2o (65
mg, 39%) as a 85:15 mixture of diastereomers. Further elution with an
AcOH gradient afforded 2p (89 mg, 59%) as a 94:6 mixture of
diastereomers. A second chromatography on Dowex 1X8 allowed the
GluA2 LBD protein (50 ng) binding was similarly measured in
assay buffer (50 mM Tris-HCl, 100 mM KSCN, 2.5 mM CaCl2, 10%
(v/v) glycerol, pH 7.2 at 4 °C) using 2−5 nM (RS)-[5-methyl-3H]-
AMPA at 11 concentrations of 2i (1 μM to 1 mM) in triplicate. The
radioligand Kd value (12.8 nM) at GluA2 LBD was previously
determined.42
Competition data were analyzed using Grafit, version 3.00
(Erithacus Software Ltd., Horley, U.K.), and fit as previously
described43 to determine 2i Ki.
1625
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
EAAT Pharmacology. Cell Culture. The construction and
maintenance of stable HEK293 cell lines expressing the human
EAAT subtypes EAAT1, EAAT2, and EAAT3 have been described
previously.24 The cell lines were grown in culture medium [Dulbecco’s
modified Eagle medium (DMEM) supplemented with penicillin (100
U/mL), streptomycin (100 μg/mL), and 5% dialyzed fetal bovine
serum] supplemented with 1 mg/mL G-418 at 37 °C in a humidified
5% CO2 incubator.
optimization (MMFF94s [MacroModel, version 9.9; Schrodinger, LLC:
̈
New York, NY, 2011]). The two structures were refined using
PHENIX, and between refinements the model was inspected and
corrected in COOT. The GluA2-LBD complex was refined using
anisotropic B-factors and riding hydrogen atoms, whereas the GluK3-
LBD structure was refined using TLS and grouped B-factors. 2i, water,
and glycerol molecules, as well as chloride, lithium, and potassium
ions, were manually modeled into the electron densities. For statistics
on data collection and refinements, see Table S1 in Supporting
Information.
Structure Analysis. Domain closure of GluA2-LBD and GluK3-
LBD in complex with 2i, both compared to the apo structure of GluA2
(PDB entry 1FTO, molA),25 was calculated using the DynDom
server.52 Ligand−protein interactions were identified and distances
measured in COOT. Figures were prepared using PyMOL [The
[3H]-D-Asp Uptake Assay. The characterization of the compounds
was performed in a [3H]-D-Asp uptake assay essentially as previously
described.24 Briefly, cells were split into white 96-well plates
(PerkinElmer, Boston, MA). At 16−24 h later the culture medium
was aspirated, and cells were washed twice with 75 μL of assay buffer
(Dulbecco’s phosphate buffered solution supplemented with 20 mM
HEPES, 2 mM CaCl2, and 1 mM MgCl2, pH 7.4). Then 50 μL of assay
buffer supplemented with 30 nM [3H]-D-Asp and the various
compounds were added to the various wells, and the plate was
incubated at 37 °C for 6 min. The assay mixtures were quickly
removed from the wells, which were then washed with 3 × 75 μL of
ice-cold assay buffer, and 150 μL of Microscint scintillation fluid
(PerkinElmer, Boston, MA) was added to each well. The plate was
shaken for at least 1 h and counted in a TopCounter (PerkinElmer,
Boston, MA). The experiments were performed in duplicate 3−4
times for each compound. Concentration−inhibition curves were
generated by nonweighted least-squares fits using the program
GraphPad Prism (GraphPad Software, CA, U.S.).
X-ray Structure Determinations. Crystallization of GluA2-LBD
in Complex with 2i. The rat GluA2-LBD (GluR2-S1S2J)25 has been
expressed and purified essentially as previously described44 except that
trypsin digestion was used to remove the His-tag (buffer: 20 mM
NaOAc, 1 mM Glu, 10 mM EDTA, 5 mM (RS)-methionine, 10 mM
NaCl, pH 5.5) followed by ion-exchange chromatography (HiTrap
SP).28 GluA2-LBD in complex with 2i was crystallized using the
hanging drop vapor diffusion method at 6 °C. The drop contained 1
μL of complex solution (4.9 mg/mL GluA2-LBD and 8.4 mM 2i in 10
mM HEPES, pH 7.0, 20 mM NaCl, and 1 mM EDTA) and 1 μL of
reservoir solution of 15.2% PEG4000, 0.1 M Li2SO4, 0.1 M
phosphate−citrate, pH 4.5. Reservoir volume was 0.5 mL. The
crystals appeared within 1 week.
Crystallization of GluK3-LBD in Complex with 2i. The rat GluK3-
LBD (GluK3-S1S2) was prepared and purified as previously
reported.27 Crystallization of GluK3-LBD with Glu was carried out
by the hanging drop vapor diffusion method at 6 °C. The protein−Glu
solution contained 6.2 mg/mL GluK3-LBD in 10 mM HEPES, 20 mM
NaCl, and 1 mM Glu, pH 7.0. Crystals were grown in drops of 1 μL of
protein solution and 1 μL of reservoir solution of 1.8 M sodium/
potassium phosphate, pH 8.2. The reservoir volume was 0.5 mL. After
growth, the drops containing crystals were added 1 μL of 20 mM 2i in
10 mM HEPES, 20 mM NaCl, and 1 mM EDTA, pH 7.0, and were
allowed to equilibrate for 3 weeks at 6 °C.
Data Collection and Processing. Crystals of both complexes were
flash cooled in liquid nitrogen after soaking them into cryo buffer
consisting of reservoir solution plus 20% glycerol. X-ray data were
collected at the I911-3 beamline (MAX-Lab, Lund, Sweden) at a
wavelength of 1.0000 Å. Full data sets were collected to 1.24 Å
resolution for the GluA2 complex and to 2.65 Å resolution for the
GluK3 complex. Data processing was performed using iMosflm45 and
the CCP4 suite of programs.46 Structure determination was carried out
using the program PHASER47 implemented in CCP4. The GluA2-
LBD with (S)-ACPA (PDB entry 1M5E, molA)42 was used as a search
model for the GluA2 complex, and the structure of GluK3-LBD in
complex with Glu (PDB entry 3S9E, molA)27 was used as search
model for the GluK3 complex, including protein atoms only. Further
model building was performed by the program ArpWarp48 in CCP4
(GluA2 complex) and Autobuild in PHENIX49 (GluK3 complex). A
few residues that could not be built in automatically were added
manually using the program COOT.50 The ligand coordinates were
PyMOL Molecular Graphics System; Schrodinger, LLC: New York, NY,
̈
2006].
ASSOCIATED CONTENT
■
S
* Supporting Information
Combustion analysis data for 2b−p, 6a−j, and 10; H and 13C
1
NMR spectra of compounds 2b−p; HPLC analysis results of
2e, 2i, and 2o; statistics on X-ray data collection and
refinements as well as Figure S1 with zoom on the ligand-
binding site of GluA2-LBD with 2i illustrating the alternative
conformation. This material is available free of charge via the
Internet at http://pubs.acs.org.
Accession Codes
The coordinates and structure factors have been deposited in
the Protein Data Bank: GluA2-LBD with 2i (PDB entry 4IGT)
and GluK3-LBD with 2i (PDB entry 4IGR).
AUTHOR INFORMATION
Corresponding Author
■
*For K.F. (structural biology work): phone, +45 35336207; e-
mail, karla.frydenvang@sund.ku.dk. For T.G. (synthetic work):
phone, +33 473407866; fax, +33 473407717; e-mail, thierry.
gefflaut@univ-bpclermont.fr. For L.B.: phone, +45 35336244;
e-mail, lebu@sund.ku.dk.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the French National Center for Scientific Research
(CNRS), The Carlsberg Foundation, the Novo Nordisk
Foundation, GluTarget, Danscatt, and The Danish Medical
Research Council. We are grateful to Prof. Kagamiyama’s group
(Osaka Medical College, Japan) for providing us with the AAT
overexpressing E. coli strain. Dr. Susan Amara is thanked for
providing us with the EAAT cDNAs. Heidi Peterson is thanked
for help with the GluA2 protein. MAX-lab, Lund, Sweden, is
thanked for providing beam time.
ABBREVIATIONS
■
KG, α-ketoglutarate; AMPA, 2-amino-3-(3-hydroxy-5-methyl-
isoxazol-4-yl)propanoic acid; AspAT, aspartate aminotransfer-
ase; CSA, cysteine sulfinic acid; EAAT, excitatory amino acid
transporter; iGluR, ionotropic Glu receptor; KA, kainic acid;
LBD, ligand binding domain; mGluR, metabotropic Glu
receptor; NMDA, N-methyl-D-aspartate
REFERENCES
created in Maestro [Maestro, version 9.2; Schrodinger, LLC: New
̈
■
York, NY, 2011] and fitted into the electron density. Topology and
(1) Traynelis, S. F.; Wollmuth, L. P.; McBain, C. J.; Menniti, F. S.;
Vance, K. M.; Ogden, K. K.; Hansen, K. B.; Yuan, H.; Myers, S. J.;
parameter files for 2i were obtained using eLBOW51 after geometry
1626
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
Dingledine, R. Glutamate receptor ion channels: structure, regulation,
and function. Pharmacol. Rev. 2010, 62, 405−496.
(19) Faure, S.; Jensen, A. A.; Maurat, V.; Gu, X.; Sagot, E.; Aitken, D.
J.; Bolte, J.; Gefflaut, T.; Bunch, L. Stereoselective chemoenzymatic
synthesis of the four stereoisomers of L-2-(2-carboxycyclobutyl)glycine
and pharmacological characterization at human excitatory amino acid
transporter subtypes 1, 2, and 3. J. Med. Chem. 2006, 49, 6532−6538.
(20) Conti, P.; De Amici, M.; De Sarro, G.; Rizzo, M.; Stensbol, T.
B.; Brauner-Osborne, H.; Madsen, U.; Toma, L.; De Micheli, C.
Synthesis and enantiopharmacology of new AMPA-kainate receptor
agonists. J. Med. Chem. 1999, 42, 4099−4107.
(21) Bunch, L.; Krogsgaard-Larsen, P. Subtype selective kainic acid
receptor agonists: discovery and approaches to rational design. Med.
Res. Rev. 2009, 29, 3−28.
(22) Johansen, T. N.; Janin, Y. L.; Nielsen, B.; Frydenvang, K.;
Brauner-Osborne, H.; Stensbol, T. B.; Vogensen, S. B.; Madsen, U.;
Krogsgaard-Larsen, P. 2-Amino-3-(3-hydroxy-1,2,5-thiadiazol-4-yl)-
propionic acid: resolution, absolute stereochemistry and enantiophar-
macology at glutamate receptors. Bioorg. Med. Chem. 2002, 10, 2259−
2266.
(2) Ferraguti, F.; Shigemoto, R. Metabotropic glutamate receptors.
Cell Tissue Res. 2006, 326, 483−504.
(3) Campiani, G.; Fattorusso, C.; De Angelis, M.; Catalanotti, B.;
Butini, S.; Fattorusso, R.; Fiorini, I.; Nacci, V.; Novellino, E. Neuronal
high-affinity sodium-dependent glutamate transporters (EAATs):
targets for the development of novel therapeutics against neuro-
degenerative diseases. Curr. Pharm. Des. 2003, 9, 599−625.
(4) Seal, R. P.; Amara, S. G. Excitatory amino acid transporters: a
family in flux. Annu. Rev. Pharmacool. Toxicol. 1999, 39, 431−456.
(5) Bunch, L.; Erichsen, M. N.; Jensen, A. A. Excitatory amino acid
transporters as potential drug targets. Expert Opin. Ther. Targets 2009,
13, 719−31.
(6) Hwang, B.-Y.; Cho, B.-K.; Yun, H.; Koteshwar, K.; Kim, B.-G.
Revisit of aminotransferase in the genomic era and its application to
biocatalysis. J. Mol. Catal. B: Enzym. 2005, 37, 47−55.
(7) Helaine, V.; Bolte, J. Chemoenzymatic synthesis of (4S)- and
(4R)-4-methyl-2-oxoglutaric acids, precursors of glutamic acid
analogues. Eur. J. Org. Chem. 1999, 3403−3406.
(8) Helaine, V.; Rossi, J.; Gefflaut, T.; Alaux, S.; Bolte, J. Synthesis of
4,4-disubstituted L-glutamic acids by enzymatic transamination. Adv.
Synth. Catal. 2001, 343, 692−697.
(9) Sagot, E.; Jensen, A. A.; Pickering, D. S.; Pu, X.; Umberti, M.;
Stensbøl, T. B.; Nielsen, B.; Assaf, Z.; Aboab, B.; Bolte, J.; Gefflaut, T.;
Bunch, L. Chemo-enzymatic synthesis of (2S,4R)-2-amino-4-(3-(2,2-
diphenylethylamino)-3-oxopropyl)pentanedioic acida novel selec-
tive inhibitor of human excitatory amino acid transporter subtype 2. J.
Med. Chem. 2008, 51, 4085−4092.
(10) Sagot, E.; Pickering, D. S.; Pu, X.; Umberti, M.; Stensbøl, T. B.;
Nielsen, B.; Chapelet, M.; Bolte, J.; Gefflaut, T.; Bunch, L. Chemo-
enzymatic synthesis of a series of 2,4-syn-functionalized (S)-glutamate
analogues: new insight into the structure−activity-relation of
ionotropic glutamate receptor subtypes 5, 6 and 7. J. Med. Chem.
2008, 51, 4093−4103.
(11) Alaux, S.; Kusk, M.; Sagot, E.; Bolte, J.; Jensen, A. A.; Brauner-
Osborne, H.; Gefflaut, T.; Bunch, L. Chemoenzymatic synthesis of a
series of 4-substituted glutamate analogues and pharmacological
characterization at human glutamate transporters subtypes 1−3. J.
Med. Chem. 2005, 48, 7980−7992.
(12) Emiliozzi, R.; Pichat, L. A simple method for the preparation of
cysteinesulfinic acid. Bull. Soc. Chim. Fr. 1959, 1887−1888.
(13) Jenkins, W. T.; D’Ari, L. Preparation of the phosphopyridox-
amine form of the glutamic-aspartic transaminase. Biochem. Biophys.
Res. Commun. 1966, 22, 376−82.
(14) Helaine, V.; Rossi, J.; Bolte, J. A new access to alkyl-alpha-
ketoglutaric acids, precursors of glutamic acid analogues by enzymatic
transamination. Application to the synthesis of (2S,4R)-4-propyl-
glutamic acid. Tetrahedron Lett. 1999, 40, 6577−6580.
(15) Xian, M.; Alaux, S.; Sagot, E.; Gefflaut, T. Chemoenzymatic
synthesis of glutamic acid analogues: substrate specificity and synthetic
applications of branched chain aminotransferase from Escherichia coli.
J. Org. Chem. 2007, 72, 7560−7566.
(16) Chakrabortty, P. N.; Dasgupta, R.; Dasgupta, S. K.; Ghosh, S. R.;
Ghatak, U. R. Synthetic studies towards complex diterpenoidsVI:
New synthetic routes to tetracyclic bridged-bicyclo[3.2.1]octane
intermediates by intramolecular alkylation reactions through α-
diazomethyl ketones of hydroaromatic γ,δ-unsaturated acids. Tetrahe-
dron 1972, 28, 4653−4665.
(23) Clausen, R. P.; Hansen, K. B.; Cali, P.; Nielsen, B.; Greenwood,
J. R.; Begtrup, M.; Egebjerg, J.; Brauner-Osborne, H. The respective N-
hydroxypyrazole analogues of the classical glutamate receptor ligands
ibotenic acid and (RS)-2-amino-2-(3-hydroxy-5-methyl-4-isoxazolyl)-
acetic acid. Eur. J. Pharmacol. 2004, 499, 35−44.
(24) Jensen, A. A.; Brauner-Osborne, H. Pharmacological character-
̈
ization of human excitatory amino acid transporters EAAT1, EAAT2
and EAAT3 in a fluorescence-based membrane potential assay.
Biochem. Pharmacol. 2004, 67, 2115−2127.
(25) Armstrong, N.; Gouaux, E. Mechanisms for activation and
antagonism of an AMPA-sensitive glutamate receptor: crystal
structures of the GluR2 ligand binding core. Neuron 2000, 28, 165−
181.
(26) Pøhlsgaard, J.; Frydenvang, K.; Madsen, U.; Kastrup, J. S.
Lessons from more than 80 structures of the GluA2 ligand-binding
domain in complex with agonists, antagonists and allosteric
modulators. Neuropharmacology 2011, 60, 135−150.
(27) Venskutonyte, R.; Frydenvang, K.; Gajhede, M.; Bunch, L.;
Pickering, D. S.; Kastrup, J. S. Binding site and interlobe interactions of
the ionotropic glutamate receptor GluK3 ligand binding domain
revealed by high resolution crystal structure in complex with (S)-
glutamate. J. Struct. Biol. 2011, 176, 307−314.
(28) Juknaite, L.; Venskutonyte, R.; Assaf, Z.; Faure, S.; Gefflaut, T.;
Aitken, D. J.; Nielsen, B.; Gajhede, M.; Kastrup, J. S.; Bunch, L.;
Frydenvang, K.; Pickering, D. S. Pharmacological and structural
characterization of conformationally restricted (S)-glutamate ana-
logues at ionotropic glutamate receptors. J. Struct. Biol. 2012, 180, 39−
46.
(29) Venskutonyte, R.; Frydenvang, K.; Hald, H.; Rabassa, A. C.;
Gajhede, M.; Ahring, P. K.; Kastrup, J. S. Kainate induces various
domain closures in AMPA and kainate receptors. Neurochem. Int.
2012, 61, 536−545.
(30) Armstrong, N.; Mayer, M.; Gouaux, E. Tuning activation of the
AMPA-sensitive GluR2 ion channel by genetic adjustment of agonist-
induced conformational changes. Proc. Natl. Acad. Sci. U.S.A. 2003,
100, 5736−5741.
(31) Madden, D. R.; Cheng, Q.; Thiran, S.; Rajan, S.; Rigo, F.;
Keinanen, K.; Reinelt, S.; Zimmermann, H.; Jayaraman, V. Stereo-
chemistry of glutamate receptor agonist efficacy: engineering a dual-
specificity AMPA/kainate receptor. Biochemistry 2004, 43, 15838−
15844;Erratum: Biochemistry 2006, 45, 3472.
(32) Unno, M.; Shinohara, M.; Takayama, K.; Tanaka, H.; Teruya,
K.; Doh-ura, K.; Sakai, R.; Sasaki, M.; Ikeda-Saito, M. Binding and
selectivity of the marine toxin neodysiherbaine A and its synthetic
analogues to GluK1 and GluK2 kainate receptors. J. Mol. Biol. 2011,
413, 667−683.
(17) Bunch, L.; Gefflaut, T.; Alaux, S.; Sagot, E.; Nielsen, B.;
Pickering, D. S. Pharmacological characterization of (4R)-alkyl
glutamate analogues at the ionotropic glutamate receptors. Focus on
subtypes iGlu(5-7). Eur. J. Pharmacol. 2009, 609, 1−4.
(18) Bunch, L.; Nielsen, B.; Jensen, A. A.; Brauner-Osborne, H.
Rational design and enantioselective synthesis of (1R,4S,5R,6S)-3-
azabicyclo[3.3.0]octane-4,6-dicarboxylic acida novel inhibitor at
human glutamate transporter subtypes 1, 2, and 3. J. Med. Chem.
2006, 49, 172−178.
(33) Kamitori, S.; Hirotsu, K.; Higuchi, T.; Kondo, K.; Inoue, K.;
Kuramitsu, S.; Kagamiyama, H.; Higuchi, Y.; Yasuoka, N.; Kusunoki,
M.; Matsuura, Y. Overproduction and preliminary-X-ray character-
ization of aspartate-aminotransferase from Escherichia-coli. J. Biochem.
1987, 101, 813−816.
1627
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628

Journal of Medicinal Chemistry
Article
(34) Gefflaut, T.; Assaf, Z.; Sancelme, M. Preparation of Glutamate
Analogues by Enzymatic Transamination. In Unnatural Amino Acids;
Pollegioni, L., Servi, S., Eds.; Humana Press: New York, 2012; Vol.
794, pp 55−72.
(35) Ransom, R. W.; Stec, N. L. Cooperative modulation of [H-3]
Mk-801 binding to the N-methyl-D-aspartate receptor-ion channel
complex by L-glutamate, glycine, and polyamines. J. Neurochem. 1988,
51, 830−836.
(36) Honore, T.; Nielsen, M. Complex structure of quisqualate-
sensitive glutamate receptors in rat cortex. Neurosci. Lett. 1985, 54,
27−32.
(37) Braitman, D. J.; Coyle, J. T. Inhibition of [H-3] kainic acid
receptor-binding by divalent-cations correlates with ion affinity for the
calcium-channel. Neuropharmacology 1987, 26, 1247−1251.
(38) Sills, M. A.; Fagg, G.; Pozza, M.; Angst, C.; Brundish, D. E.;
Hurt, S. D.; Wilusz, E. J.; Williams, M. [H-3] Cgp-39653a new N-
methyl-D-aspartate antagonist radioligand with low nanomolar affinity
in rat-brain. Eur. J. Pharmacol. 1991, 192, 19−24.
structure solution. Acta Crystallogr., Sect D: Biol. Crystallogr. 2010, 66,
213−221.
(50) Emsley, P.; Lohkamp, B.; Scott, W. G.; Cowtan, K. Features and
development of Coot. Acta Crystallogr., Sect D: Biol. Crystallogr. 2010,
66, 486−501.
(51) Moriarty, N. W.; Grosse-Kunstleve, R. W.; Adams, P. D.
Electronic ligand builder and optimization workbench (eLBOW): a
tool for ligand coordinate and restraint generation. Acta Crystallogr.,
Sect D: Biol. Crystallogr. 2009, 65, 1074−1080.
(52) Hayward, S.; Lee, R. A. Improvements in the analysis of domain
motions in proteins from conformational change: DynDom version
1.50. J. Mol. Graphics Modell. 2002, 21, 181−183.
(39) Clausen, R. P.; Naur, P.; Kristensen, A. S.; Greenwood, J. R.;
Strange, M.; Brauner-Osborne, H.; Jensen, A. A.; Nielsen, A. S.;
Geneser, U.; Ringgaard, L. M.; Nielsen, B.; Pickering, D. S.; Brehm, L.;
Gajhede, M.; Krogsgaard-Larsen, P.; Kastrup, J. S. The glutamate
receptor GluR5 agonist (S)-2-amino-3-(3-hydroxy-7,8-dihydro-6H-
cyclohepta[d]isoxazol-4-yl)propionic acid and the 8-methyl analogue:
synthesis, molecular pharmacology, and biostructural characterization.
J. Med. Chem. 2009, 52, 4911−4922.
(40) Nielsen, B. B.; Pickering, D. S.; Greenwood, J. R.; Brehm, L.;
Gajhede, M.; Schousboe, A.; Kastrup, J. S. Exploring the GluR2 ligand-
binding core in complex with the bicyclical AMPA analogue (S)-4-
AHCP. FEBS J. 2005, 272, 1639−1648.
(41) Frydenvang, K.; Lash, L. L.; Naur, P.; Postila, P. A.; Pickering,
D. S.; Smith, C. M.; Gajhede, M.; Sasaki, M.; Sakai, R.; Pentikainen, O.
T.; Swanson, G. T.; Kastrup, J. S. Full domain closure of the ligand-
binding core of the ionotropic glutamate receptor iGluR5 induced by
the high affinity agonist dysiherbaine and the functional antagonist 8,9-
dideoxyneodysiherbaine. J. Biol. Chem. 2009, 284, 14219−14229.
(42) Hogner, A.; Kastrup, J. S.; Jin, R.; Liljefors, T.; Mayer, M. L.;
Egebjerg, J.; Larsen, I. K.; Gouaux, E. Structural basis for AMPA
receptor activation and ligand selectivity: crystal structures of five
agonist complexes with the GluR2 ligand-binding core. J. Mol. Biol.
2002, 322, 93−109.
(43) Nielsen, B. S.; Banke, T. G.; Schousboe, A.; Pickering, D. S.
Pharmacological properties of homomeric and heteromeric GluR1(o)
and GluR3(o) receptors. Eur. J. Pharmacol. 1998, 360, 227−238.
(44) Ramanoudjame, G.; Du, M.; Mankiewicz, K. A.; Jayaraman, V.
Allosteric mechanism in AMPA receptors: a FRET-based investigation
of conformational changes. Proc. Natl. Acad. Sci. U.S.A. 2006, 103,
10473−10478.
(45) Leslie, A. G. W.; Powell, H. R. Processing Diffraction Data with
Mosflm. Evol. Methods Macromol. Crystallogr. 2007, 245, 41−51.
(46) Winn, M. D.; Ballard, C. C.; Cowtan, K. D.; Dodson, E. J.;
Emsley, P.; Evans, P. R.; Keegan, R. M.; Krissinel, E. B.; Leslie, A. G.
W.; McCoy, A.; McNicholas, S. J.; Murshudov, G. N.; Pannu, N. S.;
Potterton, E. A.; Powell, H. R.; Read, R. J.; Vagin, A.; Wilson, K. S.
Overview of the CCP4 suite and current developments. Acta
Crystallogr., Sect D: Biol. Crystallogr. 2011, 67, 235−242.
(47) McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn, M.
D.; Storoni, L. C.; Read, R. J. Phaser crystallographic software. J. Appl.
Crystallogr. 2007, 40, 658−674.
(48) Cohen, S. X.; Ben Jelloul, M.; Long, F.; Vagin, A.; Knipscheer,
P.; Lebbink, J.; Sixma, T. K.; Lamzin, V. S.; Murshudov, G. N.;
Perrakis, A. ARP/wARP and molecular replacement: the next
generation. Acta Crystallogr., Sect D: Biol. Crystallogr. 2008, 64, 49−60.
(49) Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis,
I. W.; Echols, N.; Headd, J. J.; Hung, L. W.; Kapral, G. J.; Grosse-
Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R.
J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P. H.
PHENIX: a comprehensive Python-based system for macromolecular
1628
dx.doi.org/10.1021/jm301433m | J. Med. Chem. 2013, 56, 1614−1628