Synthesis and pharmacological characterization of N3-substituted willardiine derivatives: Role of the substituent at the 5-position of the uracil ring in the development of highly potent and selective GLUK5 kainate receptor antagonis

Article abstract of DOI:10.1021/jm061041u

Some N³-substituted analogues of willardiine such as 11 and 13 are selective kainate receptor antagonists. In an attempt to improve the potency and selectivity for kainate receptors, a range of analogues of 11 and 13 were synthesized with 5-sub

Full text of DOI:10.1021/jm061041u

1558
J. Med. Chem. 2007, 50, 1558-1570
Synthesis and Pharmacological Characterization of N3-Substituted Willardiine Derivatives:
Role of the Substituent at the 5-Position of the Uracil Ring in the Development of Highly Potent
and Selective GLUK5 Kainate Receptor Antagonists
Nigel P. Dolman,† Julia C. A. More,† Andrew Alt,§ Jody L. Knauss,§ Olli T. Pentika¨inen,† Carla R. Glasser,| David Bleakman,§
Mark L. Mayer,| Graham L. Collingridge,‡ and David E. Jane*,†
Department of Pharmacology, MRC Centre for Synaptic Plasticity, School of Medical Sciences, UniVersity Walk, UniVersity of Bristol, Bristol
BS8 1TD, United Kingdom, Department of Anatomy, MRC Centre for Synaptic Plasticity, School of Medical Sciences, UniVersity of Bristol
BS8 1TD, United Kingdom, Neuroscience Research, Eli Lilly and Company, Lilly Corporate Center, Indianapolis, Indiana 46285, and
Laboratory of Cellular and Molecular Neurophysiology, Porter Neuroscience Research Center, National Institute of Child Health and Human
DeVelopment, National Institutes of Health, Department of Health and Human SerVices, Bethesda, Maryland 20892
ReceiVed August 30, 2006
Some N3-substituted analogues of willardiine such as 11 and 13 are selective kainate receptor antagonists.
In an attempt to improve the potency and selectivity for kainate receptors, a range of analogues of 11 and
13 were synthesized with 5-substituents on the uracil ring. An X-ray crystal structure of the 5-methyl analogue
of 13 bound to GLUK5 revealed that there was allowed volume around the 4- and 5-positions of the thiophene
ring, and therefore the 4,5-dibromo and 5-phenyl (67) analogues were synthesized. Compound 67 (ACET)
demonstrated low nanomolar antagonist potency on native and recombinant GLUK5-containing kainate
receptors (KB values of 7 ( 1 and 5 ( 1 nM for antagonism of recombinant human GLUK5 and GLUK5/
GLUK2, respectively) but displayed IC50 values >100 µM for antagonism of GLUA2, GLUK6, or GLUK6/
GLUK2.
Introduction
ing kainate receptor antagonists.5,6 These antagonists displayed
useful selectivity for GLUK5 versus AMPA and GLUK6
-
Kainate receptors are a class of ligand-gated ion channel that
are expressed in the mammalian central nervous system and
are activated by the excitatory neurotransmitter (S)-glutamate.1
There are three groups of ligand-gated ion channels activated
by (S)-glutamate known as the (S)-2-amino-3-hydroxy-5-methyl-
4-isoxazolepropanoic acid (AMPA,a 1, Figure 1), (2S,3S,4S)-
3-carboxymethyl-4-isopropenyl-pyrrolidine-2-carboxylic acid
(kainate, 2), and N-methyl-D-aspartic acid (NMDA) receptor
subtypes.2 Kainate receptors are tetrameric assemblies of
GLUK5-7, GLUK1, and GLUK2 subunits (IUPHAR nomenclature
of the receptors that are also known as GluR5-7, KA1, and
KA2).3 AMPA receptors are tetrameric assemblies of a com-
bination of GLUA1-4 subunits (IUPHAR nomenclature of the
receptors that are also known as GluR1-4 or GluRA-D).3 The
first antagonists with significant activity at kainate receptors
were from the quinoxalinedione class of compounds, and these
included 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 3) and
the potent AMPA receptor antagonist 2,3-dihydroxy-6-nitro-7-
sulfamoyl-benzo(f)quinoxaline (NBQX, 4).4 None of these
compounds were useful as pharmacological tools for kainate
receptors as they also antagonized AMPA receptors. More
recently, a series of decahydroisoquinolines, culminating in
compound 5,5f have been reported as selective GLUK5-contain-
containing receptors and were therefore used to provide evidence
to support the role of GLUK5 in synaptic plasticity in the mossy
fiber to CA3 region of the hippocampus and in a number of
central nervous system disorders such as chronic pain, epilepsy,
ischaemia, and migraine.5,6 Interestingly, an analogue of the
natural product dysiherbaine, 6, has been reported to be a
selective GLUK5 receptor antagonist.7 However, unlike the
decahydroisoquinoline-based antagonists, 6 evoked brief seizures
followed by unconsciousness when injected into mice.8 Aside
from orthosteric antagonists, recently the first selective non-
competitive antagonist for GLUK5, 7, has been reported.9
Substitution at the 5-position of the uracil ring of the natural
product willardiine (8) led to compounds (9 and 10) with
selective agonist activity at either AMPA or kainate receptors.10
We have produced selective GLUK5-containing kainate receptor
antagonists by adding either a 2-carboxybenzyl (11) or a
2-carboxthiophene-3-ylmethyl (13) substituent to the N3-position
of the uracil ring of willardiine.11 The racemic mixture and the
pure S enantiomer 11 have been used to show that the kainate
receptor involved in the induction of long-term potentiation
(LTP) in the mossy fiber to CA3 pathway in the hippocampus
contains the GLUK5 subunit.11,12 In addition, 11 has been used
to show that GLUK5-containing kainate receptors are involved
in short-term recognition memory.13 We have reported previ-
ously that 12, the 5-iodo-substituted analogue of 11, had
enhanced GLUK5 antagonist potency and selectivity, but re-
placement of the thiophene ring of 13 with a furan ring to give
14 led to a reduction in GLUK5 receptor antagonist potency.11
While 11-14 (Figure 1) were moderately potent kainate
receptor antagonists, there is a need for more potent and selective
antagonists for in vitro and in vivo studies of GLUK5 kainate
receptor function. Herein, we report the synthesis and pharma-
cological characterization of a new series of analogues of 11
and 14 in which the 5-subsituent on the uracil ring was varied
* To whom correspondence should be addressed. Phone: +44 (0)117
9546451. Fax: +44 (0)117 9250168. E-mail: david.jane@bristol.ac.uk.
† Department of Pharmacology, University of Bristol.
‡ Department of Anatomy, University of Bristol.
§ Eli Lilly and Co.
| NIH.
a Abbreviations: AMPA, (S)-2-amino-3-hydroxy-5-methyl-4-isoxazolepro-
panoic acid; kainate, (2S,3S,4S)-3-carboxymethyl-4-isopropenyl-pyrrolidine-
2-carboxylic acid; NMDA, N-methyl-D-aspartic acid; CNQX, 6-cyano-7-
nitroquinoxaline-2,3-dione; NBQX, 2,3-dihydroxy-6-nitro-7-sulfamoyl-
benzo(f)quinoxaline; DHPG, (S)-3,5-dihydroxyphenylglycine; fDR-VRP,
fast component of the dorsal root-evoked ventral root potential; ATPO, (S)-
2-amino-3-[5-tert-butyl-3-(phosphonomethoxy)-4-isoxazolyl]propionic acid.
10.1021/jm061041u CCC: $37.00 © 2007 American Chemical Society
Published on Web 03/10/2007

N3-Substituted Willardiine DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1559
Figure 1. Structures of AMPA and kainate receptor agonists and antagonists.
in an attempt to increase the potency and selectivity for GLUK5-
containing kainate receptors. In addition, the effect of adding
substituents to the thiophene ring was investigated. This novel
series of compounds was pharmacologically characterized on
both cloned and native AMPA and kainate receptors. We also
undertook a molecular modeling study to attempt to explain
the pharmacological data obtained for the compounds synthe-
sized in this study.
HCl (aq), gave amino acids 34 and 35 with enantiomeric
excesses >99% as determined by chiral HPLC. Compound 1111a
was treated with bromine in aqueous acetic acid to obtain the
5-bromo derivative (36, Scheme 1), which was purified using
Dowex 50WX8 ion-exchange resin (H+ form) and crystallization
from water, thereby avoiding the use of strongly basic Dowex
X18 resin.
The N1-protected uracils 22, 23, and 24 were required for
the synthesis of a range of 5-substituted analogues of the sub-
micromolar potent GLUK5 antagonist 13. The 5-trifluoromethyl
derivative (24) was synthesized from the corresponding uracil
(17) by the method used for the preparation of 22. In the case
of the 5-bromo derivative 23,14 the protected uracil (19) was
treated with an excess of 2-acetoxytetrahydrofuran (21), and
the resulting N1- and N1,N3-di-protected products were then
hydrolyzed with a mixture of ethanol/acetic acid11b,16 to give
almost exclusively the N1-protected uracil derivative (23) (see
Scheme 1).
The thiophene ester (39)11b,19 required for the synthesis of
the N3-substituted derivatives 52-54 was synthesized by
esterification of the carboxylic acid (37) followed by treatment
of the ester (38) with NBS under free radical conditions (Scheme
2). The initial attempt at the synthesis of the N3-substituted uracil
(54) involved the treatment of the sodium salt of 24 with
thiophene ester (39). However, attempted conversion of the
methyl ester to the corresponding carboxylic acid via base
hydrolysis of the N3-substituted intermediate led to hydrolysis
of the 5-trifluoromethyl group to a carboxylic acid. To
circumvent this problem, the commercially available methyl
ester (38) was treated with an excess of potassium t-butoxide
in diethyl ether to afford the t-butyl ester (40). Treatment of 40
with NBS under free radical conditions gave the bromomethyl
derivative (41), and this ester was used for the preparation of
the N3-substituted derivatives 52-54 (Scheme 2). Reaction of
the sodium salt of 24 with the t-butyl ester (40) gave the N3-
substituted intermediate (51) (Scheme 3). Concomitant depro-
tection of the t-butyl ester and the N1-protecting group of 51
was achieved by treatment with TFA to afford the N3-substituted
Results
Chemistry. A series of 5-substituted analogues of the
previously reported GLUK5 selective antagonist 11 were syn-
thesized, as we had previously observed that 12, the 5-iodo-
substituted analogue of 11 (Figure 1), had increased GLUK5
antagonist potency and selectivity.11a
The required 5-substituted N1-protected uracil 2214 was
synthesized from thymine (15). Heating a suspension of thymine
(15) in the presence of hexamethyldisilazane, TMSCl, and
ammonium sulfate catalyst until a clear solution resulted
produced the silylated uracil (18) (Scheme 1).15 The highly
moisture-sensitive silylated uracil (18) was then treated with a
large excess of 2-acetoxytetrahydrofuran (21), prepared by a
previously reported route.11b Treatment of the resulting mixture
of N1- and N1,N3-di-substituted products with ethanol/acetic
acid16 resulted in total hydrolysis back to the free uracil (15).
Therefore, the quantity of 2-acetoxytetrahydrofuran (21) was
reduced to only a slight excess to minimize the formation of
the N1,N3-di-substituted product. Alkylation of 22 with the
substituted benzyl bromide 2517 gave the N3-substituted uracil
27. Similarly, alkylation of commercially available ftorufur (26)
with the substituted benzyl bromide (25) gave the corresponding
5-fluoro-substituted derivative 28. Subsequent basic hydrolysis
of the methyl esters (27 and 28) afforded the free carboxylic
acids (29 and 30), which give the N3-substituted uracil deriva-
tives (31 and 32) following treatment with TFA to remove the
N1 protecting group. Treatment of the sodium salts of 31 and
32 with (S)-N-Boc-serine-â-lactone (33)18 in DMF, followed
by hydrolysis of the N-Boc-protected intermediates with 2 M

1560 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Dolman et al.
Scheme 1a
a (a) (i) HMDS, TMSCl, (NH4)2SO4, reflux; (b) (i) NaH, DMF, (ii) 25; (c) LiOH, dioxane/water (1:1), room temperature; (d) TFA, room temperature; (e)
(i) 2 equiv of NaH, DMF, room temperature, (ii) 33, (iii) 2 M HCl, 50 °C; (f) Br2/AcOH (aq).
Scheme 2a
a (a) MeOH, H2SO4, reflux; (b) NBS, benzoyl peroxide, CCl4; (c) KOtBu, Et2O; (d) Br2, AcOH, room temperature.
uracil (54). Similarly, the N3-substituted uracils 52 and 53 were
synthesized by alkylation of the corresponding N1-protected
uracils 22 and 23 with t-butyl ester (40), followed by treatment
with TFA (Scheme 3). The amino acids 55-57 were synthesized
by treatment of the sodium salts of the corresponding uracils
(52-54) with (S)-N-Boc-serine-â-lactone (33) in DMF, followed

N3-Substituted Willardiine DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1561
Scheme 3a
a (a) (i) NaH, DMF, room temperature, (ii) 41 (Scheme 2), DMF, room temperature; (b) TFA, room temperature; (c) (i) 2 equiv of NaH, DMF, room
temperature, (ii) 33 (Scheme 1), (iii) 2 M HCl, 50 °C.
Scheme 4a
a (a) (i) NaH, DMF, (ii) 43, 46, or 48 (Scheme 3), 60 °C; (b) LiOH, dioxane/water (1:1); (c) TFA, room temperature; (d) (i) 2 equiv of NaH, DMF, room
temperature, (ii) 33 (Scheme 1), (iii) 2 M HCl, 50 °C.
by removal of the Boc protecting group with 2 M HCl (aq) and
purification by ion-exchange chromatography and crystallization
from water (Scheme 3).
The amino acid (67) was obtained from 64 by the same method
employed for the synthesis of 69 (Scheme 4). The known
dibromothiophene (42)23 was obtained by treatment of the ester
(38) with excess bromine in acetic acid (Scheme 2). Bromination
of 42 with NBS under free radical conditions gave the
bromomethyl derivative (43), which was used to alkylate the
sodium salt of 22. The resultant N3-substituted intermediate (59)
was converted to the amino acid (68) using methods similar to
those used for the synthesis of 67 and 69 (Scheme 4).
Pharmacology. In Vitro Electrophysiology. Native GLUK5
Kainate Receptor Antagonist Activity. Neonatal rat dorsal root
C fibers are a convenient source of GLUK5-containing kainate
receptors24 for the assessment of the antagonist activity of N3-
substituted willardiines.11 In this assay, the known AMPA/
kainate receptor antagonist 3 (Figure 1) had low micromolar
potency as an antagonist of kainate-induced depolarizations (see
Table 1 for data), which is in agreement with a previous report.25
We have shown previously that 11 and 13 are selective sub-
micromolar potent GLUK5 receptor antagonists in this assay.11
Adding a substituent to the 5-position of the uracil ring of 11
improved GLUK5 receptor antagonist potency and selectivity
versus AMPA receptors. The rank order of antagonist potency
for the 5-substituted analogues of 11 was 5-CH3 > 5-Br > 5-I
An analogue of 55 in which the thiophene ring is replaced
with a furan ring (69) was synthesized by reacting the bro-
momethylfuran (48) with the sodium salt of the N1-protected
uracil (22) to give the N3-substituted intermediate (60) (Scheme
4). The bromomethylfuran (48)20 was synthesized from com-
mercially available furan 47 (Scheme 2) by treatment with NBS
under free radical conditions. The ester (60) was hydrolyzed
with aqueous lithium hydroxide to give 63, which was treated
with TFA to give the N3-substituted uracil (66). The amino acid
(69) was obtained by reacting the sodium salt of 66 with (S)-
N-Boc-serine-â-lactone (33) and removal of the Boc group of
the resulting intermediate with TFA (Scheme 4). The phenyl-
substituted thiophene (46) was synthesized starting from 4422
(Scheme 2), which was prepared by a previously reported
method. Esterification of 44 with methanol/H2SO4 and treatment
of the resultant ester (45) with NBS gave the required bromom-
ethylthiophene (46) (Scheme 2). Alkylation of the sodium salt
of 22 with 46 afforded the N3-substituted intermediate, which
was hydrolyzed with LiOH, and the N1 protecting group was
removed with TFA to give the desired uracil (64) (Scheme 4).

1562 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Dolman et al.
Table 1. Summary of the Activity of Novel N3-Substituted Willardiine
Analogues at AMPA and Kainate Receptors in Electrophysiological
Assaysa
by AMPA receptors expressed on neonatal rat motoneurones,26
was used as an expeditious method of investigating the relative
antagonist potency at AMPA receptors (see Table 1 for data).
All of the compounds tested were greater than 13-fold less potent
as AMPA receptor antagonists than the reference substance 4
(Figure 1). For the most potent GLUK5 kainate receptor
antagonists, apparent KD values for the antagonism of depolar-
izations of neonatal rat motoneurones induced by 9 (a selective
AMPA receptor agonist) were calculated. Apparent KD values
for antagonism of depolarizations induced by
9 for
compounds 13, 55, and 67 were 71.4 ( 8.3, 83.4 ( 4.1, and
108 ( 6 µM, respectively. Figure 2B illustrates a concentration-
response curve for (S)-5-fluorowillardiine (9)-induced depolar-
ization of neonatal rat motoneurones in the absence and presence
of 55.
Antagonist Activity on NMDA and Group I Metabotropic
Glutamate Receptors Expressed on Neonatal Rat Motoneu-
rones. Two of the most potent and selective GLUK5 receptor
antagonists 55 and 67 were tested on NMDA and group I
(mGlu1 and mGlu5) metabotropic glutamate receptors
expressed on neonatal rat motoneurones. Compounds 55 (10
µM) and 67 (1 µM) had no activity on NMDA- or (S)-3,5-
dihydroxyphenylglycine (DHPG, a mGlu1 and mGlu5 receptor
agonist)-induced depolarizations of neonatal rat motoneurones
(n ) 3).
K
D (µM)
IC50 (µM)
compound
vs kainateb
vs fDR-VRPc
3
4
1.38 ( 0.12
0.214 ( 0.043
106 ( 13d
74.1 ( 4.3d
31.2 ( 5.0d
11.7 ( 1.3d
51.2 ( 8.6
18.5 ( 4.4
63.9 ( 16.2
16.2 ( 5.4
42.9 ( 8.2
55.3 ( 9.6
16.6 ( 1.1
2.8 ( 0.8
Calcium Fluorescence Assays. Recombinant Human AMPA
and Kainate Receptor Subtypes. To further characterize the
selectivity of 55 and 67, two of the most potent and selective
GLUK5 receptor antagonists identified in the native ionotropic
glutamate receptor assays, they were tested on a range of
recombinant human AMPA and kainate receptor subtypes (see
Table 2 for data). Compounds 55 and 67 were highly potent
antagonists of cells expressing GLUK5 with Kb values of 10 (
1 and 7 ( 1 nM, respectively. Similar potencies were observed
upon testing 55 and 67 on GLUK5/GLUK2 heteromers where
Kb values of 8 ( 2 and 5 ( 1 nM, respectively, were obtained.
11
12
13
14
34
35
36
55
56
57
67
68
69
0.402 ( 0.045d
0.209 ( 0.019d
0.105 ( 0.007d
1.88 ( 0.15d
0.023 ( 0.001
0.358 ( 0.013
0.102 ( 0.020
0.018 ( 0.004
0.025 ( 0.002
0.055 ( 0.010
0.012 ( 0.001
0.010 ( 0.002
0.080 ( 0.008
25.1 ( 3.7
a All values are from three independent experiments and are expressed
as the mean ( SEM. b Apparent KD values for antagonism of kainate-
induced depolarization of neonatal rat dorsal root fibers calculated using
the Gaddum-Schild equation. c Depression of the fast component of the
dorsal root evoked ventral root potential (fDR-VRP) in the neonatal rat
spinal cord preparation, a measure of antagonist activity at AMPA receptors
expressed on motoneurones. d Values taken from refs 11a,b. e KD values
for inhibition by 13, 55, and 67 of the depolarization of neonatal rat
motoneurones induced by 9 (a selective AMPA receptor agonist) were 71.4
( 8.3, 83.4 ( 4.1, and 108 ( 6 µM (n ) 3; mean ( SEM), respectively.
Neither of these compounds had significant activity at GLUA2
,
GLUK6, or GLUK6/GLUK2 receptors (IC50 values >100 µM).
Figure 2C illustrates concentration response curves for antago-
nism of glutamate-induced Ca2+ fluorescence by 55 in cell lines
expressing human GLUK5, GLUK5/GLUK2, or GLUK6
.
Modeling Study. To understand the pharmacological data
obtained for the new compounds, we undertook a molecular
modeling study in which the recently reported X-ray crystal
structures of 11 and 55 bound to the ligand binding core of
21
GLUK5 were used as a template. To investigate the reasons
> 5-F ≈ 5-H. A series of 5-substituted analogues of 13 was
synthesized, and again the 5-methyl-substituted analogue (55)
gave the most potent GLUK5 receptor antagonist (see Figure
2A for a concentration-response curve for kainate-induced
depolarization of dorsal root C fibers in the absence and presence
of 55) with the rank order of potency being 5-CH3 > 5-Br >
5-CF3. We have previously reported that 14, the analogue of
13 with a furan rather than a thiophene ring, was 18-fold less
potent as a GLUK5 receptor antagonist.11b However, the 5-methyl
analogue 69 was >20-fold more potent than the parent
compound 14, but 69 was still 4-fold less potent than the
corresponding thiophene ring-containing compound 55. Adding
substituents to the thiophene ring of 55 improved potency as
both 67 and 68 proved to be highly potent antagonists of native
GLUK5 receptors.
for the low affinity of 67 for AMPA receptors, attempts were
made to dock it into the ligand binding core of GLUA2
.
Protein Structures. The crystal structure of the ligand
binding domain of GLUA2 in complex with bound antagonist
ligand (S)-2-amino-3-[5-tert-butyl-3-(phosphonomethoxy)-4-
isoxazolyl]propionic acid (ATPO) (PDB code: 1N0T)27 and the
GLUK5 ligand binding core in complex with the antagonist
ligands 55 (PDB code: 2F34)21 and 11 (PDB code: 2F35)21
were used in this study. Hydrogen atoms for all protein structures
and models were added using the program Reduce.28 The
flexibility of amino acid side chains was modeled by using the
amino acid side-chain rotamer library29 incorporated within the
BODIL30 modeling environment.
Ligand Structures. Ligand structures were optimized quan-
tum mechanically (QM) with Gaussian 0331 at the B3LYP/6-
31+G* level using a continuum solvent model (water, using
the PCM model of Gaussian 03).
Native AMPA Receptor Antagonist Activity. The ability
of compounds to reduce the fast component of the dorsal root-
evoked ventral root potential (fDR-VRP), which is mediated

N3-Substituted Willardiine DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1563
Figure 2. Activity of 55 on native AMPA and kainate receptors expressed on neonatal rat spinal cord and cloned kainate receptor subunits. (A)
Concentration-response curves for kainate on the dorsal root in the absence and presence of 500 nM 55. The apparent KD value was 18 ( 4 nM
(n ) 3; mean ( SEM). (B) Concentration-response curves for the AMPA receptor agonist 9 in the absence and presence of 200 µM 55, obtained
by recording the ventral root response of the spinal cord. The apparent KD value was 83.4 ( 4.1 µM (n ) 3; mean ( SEM). (C) Inhibition of
glutamate-induced calcium influx by 55 in cells expressing human GLUK5 (O), GLUK5/GLUK2 (∆), or GLUK6 (0) receptors.
Table 2. Summary of the Activity on Recombinant Human AMPA and Kainate Receptors of N3-Substituted Willardiine Analogues as Antagonists of
Glutamate-Stimulated Ca2+ Influx
compound
GLUA2
GLUK5
GLUK5/GLUK2
GLUK5/GLUK6
GLUK6
GLUK6/GLUK2
4a
2.5 ( 0.3
>300
37 ( 9
50 ( 10
1.0 ( 0.4
0.18 ( 0.02
0.008 ( 0.002
0.005 ( 0.001
28 ( 8
0.8 ( 0.1
0.12 ( 0.01
29 ( 14
>300
>100
>100
>100
135 ( 91
>300
11b
13b
55b
67b
0.6 ( 0.1
>100
0.12 ( 0.03
0.010 ( 0.001
0.007 ( 0.001
>100
>100
>100
>100c
>100
a Data for compound 4 represent IC50 values (µM, n ) 3, mean ( SEM) in all assays. b Data represent Kb values (µM, n ) 3, mean ( SEM) for GLUK5
GLUK5/GLUK2, and GLUK5/GLUK6 and IC50 values (µM, n ) 3, mean ( SEM) for GLUA2, GLUK6, and GLUK6/GLUK2
30 µM, 27 ( 7% inhibition at 100 µM.
,
.
c No significant inhibition up to
Ligand Docking. QM optimized ligands were docked flexibly
into the GLUA2 and GLUK5 receptors with Gold 2.2.32,33 The
search area was limited to a 15 Å radius sphere centered at the
binding site.
Thus, neither the hydrogen atom in 11 nor the fluoro group in
35 can form favorable interactions with the surrounding groups.
When the hydrogen atom is replaced by an iodo (Figure 3B) or
bromo group, there is a steric clash between the ligand and the
receptor. Therefore, the receptor must adjust itself in this region
to be able to accommodate those groups, which in turn affects
the intramolecular energy of the receptor (and the ligand binding
affinity value). The 5-methyl group of 34, however, can be
perfectly accommodated by the receptor (Figure 3C). Note that
the 5-iodo, 5-bromo, or 5-methyl group can replace both the
wat148 and the unorganized water molecule. The rank order of
potency of 5-substituted analogues of 13 on native GLUK5 was
5-CH3 > 5-Br > 5-CF3. This rank order can be explained in a
way similar to that for 11 (see above and Figure 3). The
additional substituent in this series, 5-CF3 (compound 57), is
fairly negatively charged due to the electronegative fluorine
atoms, and thus the environment surrounding the 5-substituent,
which consists of hydrophobic residues (charge ≈ 0) and
acceptor atoms (negatively charged), is not optimal for favorable
binding.
Figures 3 and 4 were prepared by using Bodil30 and
Raster3D.34
Effect of Substituent at the 5-Position of the Uracil Ring
on Binding Affinity for GLUK5. The position of the willardiine
ring is driven by the interactions with the S2-domain of GLUK5,
where the key interactions of the carboxylic acid group attached
to the phenyl ring of 11 are with Ser674 and Thr675. The
interaction with Ser674 is via a water molecule, and the
carboxylate group attached to the phenyl group of 11 also
interacts with the main-chain amino and side-chain hydroxyl
group of Thr675, as seen in the crystal structure of 11 bound to
21
the ligand binding core of GLUK5
.
Thus, the binding of
5-substituted analogues of 11 can be studied by using the crystal
structure of GLUK5 with bound 11 and by simply replacing the
hydrogen atom at the 5-position to an iodo (12), bromo (36),
methyl (34), or fluoro (35) group. As seen in the crystal
structure, the hydrogen atom at the 5-position of the uracil ring
of 11 is located next to the oxygen atom of water molecule
148 (numbering from the X-ray structure), and there is an empty
space in between the ligand, wat148, wat201, Glu426, Tyr429,
Pro501, and Tyr749 (black spot in Figure 3A). This space is
large enough to accommodate a water molecule, which,
however, cannot hydrogen bond with any neighboring groups
and thus is unorganized and not seen in the crystal structure.
Binding Mode of 67 in Ligand Binding Core of GLUK5
.
As 67 is a derivative of 55 containing an additional phenyl group
at the 5-position of the thiophene ring, it is expected that it
would bind in a very similar way to 55. The better binding
affinity of 67 is due to additional hydrophobic interactions with
receptor. The QM optimized 67 was docked into the crystal
structure of GLUK5 with bound antagonist ligand 55 (55 was
removed prior to docking). The docking revealed that the phenyl

1564 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Dolman et al.
hydrogen atoms (0.66 Å) and the QM calculated energy (QMopt
- Qmdocked ) -2.1 kcal/mol).
Subtype Selectivity of 67. It was not possible to dock 67
into the GLUA2 binding core conformation adopted when ATPO
binds. The investigation of the binding mode of 67 in GLUA2
required a receptor model of GLUA2 in which the conformation
is similar to that of GLUK5 with 55 bound. It is likely that
GLUA2 can adjust itself into a conformation similar to that
observed in the GLUK5 crystal structure.35 The crystal structure
of GLUK5 with bound 5521 was used as a template for the
relative positions of the S1-S2 domain in the GLUA2 model,
while the main-chain and amino acid conformations were taken
from the crystal structure of GLUA2 with bound antagonist
ligand ATPO.27 When the binding mode of 67 in the GLUA2
model is studied, it is obvious that the ligand cannot bind in a
way similar to that in GLUK5, because if the docked ligand
interacts with the main-chain amino and side-chain hydroxyl
groups of Thr655, there is a steric clash between the ligand
and the receptor, either with Leu650 or with Gly653. In
particular, the bulkier Leu650 in GLUA2 does not allow the
thiophene ring and 5-phenyl group to be located in a position
similar to that observed in GLUK5 where these groups interact
with Val670. Therefore, the thiophene-phenyl ring system is
forced to bend toward the S1-domain of GLUA2; however,
Gly653 still overlaps with the thiophene group, and thus, if this
overlap is removed, the ligand would also lose partially its
interactions with Thr655, which are essential for the high
binding affinity in GLUK5. Thus, it is clear that the binding of
67 in GLUA2 would be of much lower affinity than when this
compound binds to GLUK5
.
Figure 3. The bulkiness of substituents at the 5-position of the
uracil ring of 11 and its derivatives is crucial for ligand binding
affinity (figure shown in stereo). While 11 (A) leaves an empty
hydrophobic cavity (black spot) that might be filled by a water molecule,
the 5-iodo analogue, 12 (B), is too bulky. However, the 5-methyl group
of 34 (C) is a perfect size. The surrounding residues (blue surface),
water molecule (red surface), and the ligand (11, green surface; 12,
yellow surface; 34, orange surface) are shown in a ball-and-stick
representation.
Discussion
When tested on kainate receptors on rat dorsal root C fibers
and AMPA receptors on neonatal rat motoneurones, the N3-(2-
carboxybenzyl) derivative of willardiine, 11, was found to be a
potent and selective kainate receptor antagonist.11a,12 Replace-
ment of the benzene ring of 11 with a thiophene ring to give
13 increased potency and selectivity for kainate receptors.11b
We have previously determined that the selectivity of willardiine
derivatives as AMPA or kainate receptor agonists depended on
the 5-substituent on the uracil ring. Indeed, 5-fluorowillardiine
(9) was a potent and selective AMPA receptor agonist, whereas
5-iodowillardiine (10) was a potent and selective GLUK5 kainate
receptor agonist.10a-d We therefore synthesized a range of
5-substituted analogues of 11 and 13 to investigate whether
5-substitution of the uracil ring has a marked effect on antagonist
potency and selectivity. Inspection of the potencies of a series
of 5-substiuted analogues of 11 on native GLUK5-containing
kainate receptors expressed on dorsal root C fibers (see Table
1) showed that the rank order of potency was 5-CH3 > 5-Br >
5-I > 5-F > 5-H, suggesting that a relatively small hydrophobic
substituent was required for optimal activity. This rank order
of potency is different from that observed in the agonist series
of willardiine derivatives where the rank order of potency on
GLUK5-containing kainate receptors expressed on dorsal root
ganglion cells was 5-I > 5-Br > 5-Cl > 5-F > 5-H >
5-CH3,10a-d suggesting that the binding pocket with which the
5-substituent interacts in the closed agonist bound form of the
GLUK5 receptor has different structural requirements from that
of the open antagonist bound form. In addition, in the agonist
series, the 5-substituent increases the acidity of the uracil ring,
and as the ionized form binds with higher affinity, an electron-
withdrawing substituent at the 5-position of the uracil ring is
favorable. However, in the antagonist series, the uracil ring
cannot ionize and so electronic factors are less important.
Figure 4. (A) Interactions of 67 with the S1S2 domain of GLUK5
.
The R-carboxyl group forms a salt bridge with R508 and a hydrogen
bond with T503, while the R-amino group forms a hydrogen bond with
P501, T503, and E723 (the latter via a water molecule). The carboxylic
acid group attached to the thiophene ring forms hydrogen-bond
interactions with S674 and T675. The 5-methyl group of 67 projects
into a hydrophobic cavity formed by Y429, P501, and the CH2 groups
of E426. The thiophene ring of 67 forms hydrophobic interactions with
V670 and M722. (B) The higher binding affinity of 67 as compared to
55 is due to additional hydrophobic interactions with residues V670
and N705.
group can form two favorable hydrophobic interactions with
Cγ1H3 of Val670 and CâH2 of Asn705 (Figure 4), and this
shields them from solvent. The conformation of docked 67 is
practically the same as the QM optimized structure, when
measuring both the root-mean-square deviation of all non-

N3-Substituted Willardiine DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1565
All of the 5-substituted analogues of 11 were much more
potent on GLUK5-containing kainate receptors than on native
AMPA receptors; nonetheless, it is of interest to compare the
rank order of potency in the antagonist series to that of
5-substituted analogues of willardiine in the agonist series.
Inspection of the data for 5-substituted derivatives of 11 as
antagonists of AMPA receptors expressed on motoneurones
reveals the following rank order of potency: 5-F > 5-CH3 >
5-Br > 5-I > 5-H, suggesting that a small substituent is preferred
at the 5-position of the uracil ring. This is similar to that
observed with the agonist series of willardiines where the rank
order of potency was 5-F > 5-Br > 5-I > 5-H > 5-CH3.10a-d
The low agonist potency of the 5-methyl-substituted analogue
of willardiine on AMPA receptors is likely due to the reduced
ionization of the uracil ring.
in the binding pocket. The difference in rank order of potency
for 5-substituted willardiine derivatives in the agonist and
antagonist series may be due to the large degree of opening of
the antagonist bound GLUK5 ligand binding core. We have
reported that the GLUK5 ligand binding core with either 11 or
55 bound is much more open than previously described
antagonist crystal structures from the AMPA or NMDA receptor
families.21 This large degree of opening in the antagonist bound
structure of GLUK5 is likely to have altered the geometry of
the pocket in which the 5-substituent is incorporated.
It was apparent from the X-ray crystal structure of 55 bound
to GLUK5 that there was a considerable amount of space around
the 4- and 5-positions of the thiophene ring. To probe the
volume of this free space in the receptor, the willardiine
derivatives 67 and 68 were synthesized. It was also postulated
that the addition of these groups to the thiophene ring would
increase the overall lipophilicity of the molecule, making them
more suitable for systemic administration. Both 67 and 68
proved to be potent GLUK5 receptor antagonists with 67 being
more selective for GLUK5 versus AMPA receptors than 68 or
the parent compound 55. The ∼6-fold increase in AMPA
receptor antagonist potency for 68 as compared to the parent
compound 55 is likely due to additional interactions between
one or both of the bromo groups and the S1-domain of the
AMPA receptor. Docking of 67 into the GLUK5/55 X-ray crystal
structure suggested that the small increase in potency as
compared to 55 is due to an additional hydrophobic interaction
of the phenyl ring of 67 with Val670 and the CH2 group of
Asn705 (see modeling section). The small improvement in
selectivity of 67 as compared to 55 for kainate versus AMPA
receptors is likely due to a steric interaction between the
5-phenyl substituent and the leucine residue present in all AMPA
Having established that 5-methyl- or 5-bromo-substitution was
optimal for GLUK5 antagonist activity in the series of com-
pounds based on 11, we synthesized the 5-bromo and 5-methyl
analogues of 13. We also synthesized the 5-trifluoromethyl
analogue 57, as in the agonist series 5-CF3 substitution of
willardiine led to a potent and selective GLUK5 receptor agonist
and in addition the 5-CF3 group is more hydrophobic than a
5-methyl group. The rank order of antagonist potency on native
kainate receptors for this series of compounds was similar to
that observed with the compounds based on 11, that is, 5-CH3
> 5-Br > 5-CF3 > 5-H. For native AMPA receptors, the rank
order of antagonist potency for 5-substituted analogues of 13
was 5-CH3 > 5-H > 5-Br > 5-CF3. As in the case of
5-substituted analogues of 11, a 5-methyl substituent was
preferred to 5-Br; however, in the series of compounds based
on 13, 5-Br is less well accommodated than 5-H, although the
difference is marginal.
receptor subunits that correspond to Val670 in GLUK5
.
Replacement of the thiophene ring of compounds 13 and 55
with a furan ring reduced GLUK5 receptor antagonist potency,
but 14 was ∼3-fold more potent than the corresponding
thiophene derivative 13 as an AMPA receptor antagonist.
However, 55 was slightly more potent than the corresponding
furan derivative 69 as an AMPA receptor antagonist. It would
appear that the oxygen atom in the furan ring can form additional
interactions with the ligand binding core of AMPA receptors
presumably via hydrogen bonding. The thiophene ring of 13
and 55 is preferred to a furan ring for interaction with GLUK5
receptors due to the ability to form better hydrophobic contacts
with Val670 and Met722 in the GLUK5 ligand binding core (see
modeling section). The greater potency of the 5-methyl deriva-
tive 69 as compared to the parent compound 14 may be due to
the favorable 5-methyl group interaction with a hydrophobic
pocket in GLUK5, which offsets the reduced hydrophobic
interaction with Val670 and Met722.
Characterization of compounds 11, 13, 55, and 67 on cloned
kainate receptors expressed in HEK293 cells suggests that these
compounds have high affinity for GLUK5 regardless of whether
the tetrameric receptor complex is made up of GLUK5 subunits
alone or is a combination of GLUK5 with GLUK2 or GLUK6. A
similar result was obtained with the decahydroisoquinoline series
of GLUK5 selective competitive antagonists.5 However, the
noncompetitive GLUK5 receptor antagonist 7 (Figure 1) will only
antagonize tetramers comprised of GLUK5; heteromeric GLUK5
receptors are not blocked.9a Thus, antagonists such as 67 can
be used in combination with 7 to help determine whether
homomeric or heteromeric GLUK5 kainate receptors are ex-
pressed in native tissue.
The high degree of selectivity of 67 for GLUK5-containing
kainate receptors was demonstrated by the weak activity of 67
on native AMPA, NMDA, and group I mGlu receptors
expressed on neonatal rat motoneurones and on cloned human
AMPA and GLUK6 kainate receptors. Thus, 67 can be used to
investigate the role of GLUK5 kainate receptors in CNS function
such as synaptic plasticity and in models of neuropathic pain,
epilepsy, anxiety, and migraine.
We have recently reported the X-ray crystal structures of 11
21
and 55 bound to the ligand binding core of GLUK5
.
This
allowed us to obtain a greater insight into the structural
requirements for the optimal binding of N3-substituted willar-
diine derivatives to GLUK5. In addition, modeling studies
allowed us to rationalize the rank order of antagonist potencies
for 5-substituted analogues of 11 and 13. Inspection of the
crystal structure revealed that the 5-methyl substituent of 55
displaced a water molecule in the GLUK5 receptor binding
pocket and was able to fit into a cavity lined with hydrophobic
substituents (Tyr429, Pro501, Tyr749, and CH2 groups of
Glu426), thus enhancing the binding affinity of the ligand.
Modeling studies suggested that the 5-methyl substituent is the
optimal size to fit into this pocket as larger substituents can
only be accommodated via structural rearrangement of residues
Conclusion
We have previously reported that some N3-substituted wil-
lardiine derivatives are selective GLUK5-containing kainate
receptor antagonists.2,11a,b,12 This was particularly so for com-
pounds 11 and 13, with the latter being the most potent and
selective. We have now conducted structure-activity relation-
ship studies centered around these two compounds and have
found that by 5-methyl substitution of the uracil ring we can
improve both potency and selectivity for kainate receptors. In

1566 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Dolman et al.
s, 1H, NH); 13C NMR (DMSO-d6, 67.94 MHz) δ 12.1, 24.0, 31.3,
69.0, 85.5, 109.1, 136.0, 150.4, 163.9; MS (electrospray) 195 [M
- H]-.
Compounds 23 and 24 were synthesized using a procedure
similar to that described for 22; see the Supporting Information
for more details.
addition, through analysis of the X-ray crystal structures of 11
and 55 docked into the ligand binding core of GLUK5, we were
able to design even more potent compounds by adding substit-
uents to the thiophene ring, thereby producing compounds 67
(ACET) and 68, with the former being one of the most potent
and selective GLUK5-containing kainate receptor antagonists
identified to date. Compound 67 will be a useful pharmacologi-
cal tool to study the function of GLUK5-containing kainate
receptors in the CNS. Further work is in progress to investigate
the usefulness of 67 for in vivo studies of CNS disorders.
Methyl 3-Bromomethyl-5-phenylthiophene-2-carboxylate (46).
Benzoyl peroxide (0.04 g) was added to a solution of methyl
3-methyl-5-phenylthiophene-2-carboxylate (45) (11.6 g, 50.0 mmol)
in carbon tetrachloride (500 mL). A 100 W lamp was used to
irradiate the mixture, which subsequently heated to reflux, and NBS
(8.01 g, 45.0 mmol) was added to the mixture in four equal portions
over 4 h. The mixture was allowed to react for a further 2 h, cooled
to room temperature, and filtered. The filtrate was concentrated
under reduced pressure to a quarter of its original volume, and then
extracted with water (3 × 100 mL), and once with saturated brine
solution (50 mL). The organic layer was dried (MgSO4) and
concentrated under reduced pressure to give 46 (14.76 g, 70-80%
Experimental Section
Chemistry. General Procedures. Proton NMR spectra were
measured on a Jeol NMR spectrometer at either 300.40 or 270.18
MHz. Carbon NMR spectra were run on a 300.40 MHz Jeol NMR
spectrometer at 75.45 or a 270.17 MHz Jeol spectrometer at 67.94
MHz unless otherwise stated. 3-(Trimethylsilyl)propionic-2,2,3,3-
d4 acid sodium salt in D2O, or tetramethylsilane in CDCl3, DMSO-
d6, or TFA-d were used as internal standards. Elemental analyses
were performed by Medac Ltd., Englefield Green, UK. Melting
points were determined in capillary tubes on Electrothermal IA9100
electronic melting point equipment. Determinations of optical
rotation were carried out in 6 M HCl at room temperature with a
wavelength of 589 nm at the University of Warwick. Thin layer
chromatography was performed on Merck silica gel 60 F254 plastic
sheets. Silica gel for flash chromatography was silica gel 60 (220-
440 mesh) from Fluka chemicals, U.K. The eluents for thin layer
chromatography of amino acids included [2 (pyridine:acetic acid:
water (3:8:11)):3 (n-butanol)] and [propan-2-ol:35% aqueous am-
monia (70:30)]. Amino acids were detected by spraying plates with
a 2% solution of ninhydrin in 70% ethanol. Purity of amino acids
was also determined by high voltage paper electrophoresis, using
pH 4 buffer and an applied voltage of 4 kV. Ion-exchange resin
chromatography was carried out using Dowex 50WX8-400 acid
form resin and the acetate or hydroxide form of Dowex X18-400
obtained from Aldrich Chemical Co., UK, or Biorad AG1X8 from
Bio-Rad, UK. Chiral HPLC was carried out using a RSTech,
Chirosil RCA+, 150 × 4.6 column, flow rate 1.0 mL/min, detection
220 nM with a mobile phase consisting of methanol/5 mM
perchloric acid (85:15). All solids were dried over P2O5 in vacuo
for 3 days prior to reaction. With the exception of compounds 22
and 36, all of the amino acids synthesized were washed with
anhydrous ethanol followed by diethyl ether to remove surface
water. The petroleum ether utilized in each case had a boiling point
range of 40-60 °C. All anhydrous reactions were conducted under
argon. All reagents and dry solvents were obtained from the Aldrich
Chemical Co., UK.
5-Methyl-1-(tetrahydrofuran-2-yl)pyrimidine-2,4-dione (22).
To a suspension of thymine (15) (10.0 g, 79.2 mmol) in hexam-
ethyldisilazane (72.6 mL, 317 mmol) were added chlorotrimeth-
ylsilane (10.3 mL, 79.2 mmol) followed by ammonium sulfate (0.10
g, 0.7 mmol) under a dry argon atmosphere with vigorous exclusion
of moisture. The mixture was heated under reflux conditions for
18 h and then concentrated under reduced pressure (1 mmHg, 60
°C) to leave a colorless oil, which was dissolved in dry acetonitrile
(50 mL) under a dry argon atmosphere. To this was added
2-acetoxytetrahydrofuran (21) (12.36 g, 95.04 mmol), and the
mixture was stirred at room temperature for 18 h. The mixture
was concentrated under reduced pressure, and the residue was
purified using silica gel chromatography, eluting with ethyl
acetate/petroleum ether on a gradient elution system [(3:2)-(4:1)]
to give 22 (8.37 g, 54%) as a white solid, which was further
purified by crystallization from methanol to give clear prisms.
mp 181-182 °C; lit. mp 182-183 °C;14 1H NMR (DMSO-d6,
270.17 MHz) δ 1.80 (d, J ) 1.0 Hz, HCCCH3), 1.89-2.04 (m,
3H, CH2CH2 + CH2CH2), 2.16-2.28 (m, 1H, CH2CH2), 3.76-
3.84 (m, 1H, CHOCH2), 4.13-4.21 (m, 1H, CHOCH2), 5.93-5.97
(m, 1H, CHOCH2), 7.41 (d, J ) 1.0 Hz, 1H, HCCCH3), 11.26 (br
1
pure by H NMR) as an orange colored oil, and this was used
without further purification in the next step. 1H NMR (CDCl3,
399.78 MHz) δ 3.91 (s, 3H, CH3Ar), 4.91 (s, 2H, CH2Br), 7.37 (s,
1H, Ar), 7.34-7.43 (m, 3H, Ph), 7.60-7.63 (m, 2H, Ph); MS
(electrospray) 334 [M + Na]+.
Compounds 41, 43, and 48 were synthesized using a procedure
similar to that described for 46; see the Supporting Information
for more details.
3-(2-Methoxycarbonyl-5-phenylthiophene-3-yl-methyl)-5-meth-
yl-1-(tetrahydrofuran-2-yl)pyrimidine-2,4-dione (58). A 60%
suspension of sodium hydride in mineral oil (0.67 g, 17 mmol)
was added to a solution of 5-methyl-1-(tetrahydrofuran-2-yl)-
pyrimidine-2,4-dione (22) (3.0 g, 15 mmol) in anhydrous DMF (30
mL), and the mixture was stirred at room temperature under a dry
argon atmosphere for 18 h. A solution of crude methyl 3-bromom-
ethyl-5-phenylthiophene-2-carboxylate (46) (7.0 g, 18 mmol) in dry
DMF (10 mL) was added, and the reaction mixture was stirred at
room temperature for 48 h. The mixture was concentrated under
reduced pressure (1 mmHg, 60 °C), and the resulting residue was
suspended in ethyl acetate (100 mL) and washed with water (2 ×
50 mL) and saturated brine solution (50 mL). The organic extract
was dried (MgSO4), filtered, and concentrated under reduced
pressure. The residue was purified using silica gel chromatography,
eluting with ethyl acetate/petroleum ether (1:1) to give 58 (7.2 g,
46%) as a white solid. mp 145-146 °C; 1H NMR (CDCl3, 270.17
MHz) δ 1.99 (d, J ) 1.3 Hz, 3H, HCCCH3), 1.99-2.09 (m, 3H,
CH2CH2 + CH2CH2), 2.36-2.42 (m, 1H, CH2CH2), 3.92 (s, 3H,
CO2CH3), 3.94-4.03 (m, 1H, CHOCH2), 4.19-4.27 (m, 1H,
CHOCH2), 5.56 (s, 2H, CH2Ar), 6.04-6.08 (m, 1H, CHOCH2),
6.94 (s, 1H, Ar), 7.20 (d, J ) 1.3 Hz, 1H, HCCCH3), 7.37-7.40
(m, 3H, Ph), 7.53-7.57 (m, 2H, Ph); 13C NMR (CDCl3, 67.94 MHz)
δ 13.5, 24.1, 32.9, 40.0, 52.0, 77.0, 87.5, 109.4, 123.7, 126.1, 126.2,
128.7, 128.9, 133.2, 133.5, 145.8, 148.9, 150.7, 162.7, 163.4; MS
(electrospray) 449 [M + Na]+.
Compounds 27, 28, 49, 50, 51, 59, and 60 were synthesized using
a procedure similar to that described for 58; see the Supporting
Information for more details.
3-(2-Carboxy-5-phenylthiophene-3-yl-methyl)-5-methyl-1-
(tetrahydrofuran-2-yl)pyrimidine-2,4-dione (61). A solution of
lithium hydroxide monohydrate (0.62 g, 15 mmol) in water (30
mL) was added to a solution of 3-(2-methoxycarbonyl-5-phenylth-
iophene-3-yl-methyl)-5-methyl-1-(tetrahydrofuran-2-yl)pyrimidine-
2,4-dione (58) (2.9 g, 6.7 mmol) in dioxane (30 mL), and the
mixture was stirred at room temperature for 18 h. The solution
was acidified to pH 3 with 2 M aqueous HCl and then evaporated
under reduced pressure. The residue was taken up in dichlo-
romethane/ethanol (95:5) and washed with water (50 mL) and once
with a saturated brine solution (50 mL), dried (MgSO4), and
concentrated under reduced pressure to leave 61 (2.58 g, 93%) as
a white solid. mp >300 °C; 1H NMR (CDCl3, 270.17 MHz) δ 1.91
(s, 3H, HCCCH3), 1.91-2.22 (m, 3H, CH2CH2 + CH2CH2), 2.18-
2.33 (m, 1H, CH2CH2), 3.8-3.89 (m, 1H, CHOCH2), 4.21-4.28
(m, 1H, CHOCH2), 5.29 (s, 2H, CH2Ar), 5.98-6.01 (m, 1H,

N3-Substituted Willardiine DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1567
CHOCH3), 7.09 (s, 1H, Ar), 7.35-7.46 (m, 3H, Ph), 7.56 (s, 1H,
HCCCH3), 7.65-7.69 (m, 2H, Ph), 13.2 (br s, 1H, CO2H); MS
(electrospray) 411 [M - H]-.
Compounds 29, 30, 62, and 63 were synthesized by a procedure
similar to that described for 61. The carboxylic acids 52, 53, and
54 were synthesized by hydrolysis of the t-butyl ester. See the
Supporting Information for more details.
white needles. mp 195-197 °C (dec); [R]25 ) -8.4 (c 0.105, 6
D
M HCl); ee >99% as determined by chiral HPLC; 1H NMR (TFA-
d, 270.17 MHz) δ 4.66 (dd, JBA ) 15.8 Hz, JBX ) 5.9 Hz, 1H,
CHCH2), 4.72 (dd, JAB ) 15.8 Hz, JAX ) 3.0 Hz, 1H, CHCH2),
5.01 (dd, JAX + JBX ) 8.3 Hz, 1H, CHCH2), 5.81 (s, 2H, CH2Ph),
7.15 (d, J ) 7.6 Hz, 1H, Ph), 7.47-7.53 (m, 1H, Ph), 7.60 (dt, J
) 7.6 Hz, J ) 1.3 Hz, 1H, Ph), 7.86 (d, J ) 4.62, 1H, HCCFCO),
8.24 (dd, J ) 7.6 Hz, J ) 1.3 Hz, 1H, Ph); 13C NMR (D2O + DCl
[pH 1], 67.94 MHz) δ 46.9, 52.1, 54.7, 129.1, 130.7, 132.1, 132.3,
132.8, 133.7, 136.0, 138.7, 154.5, 162.1 (d, J ) 25.4 Hz), 171.8,
174.0; MS (electrospray) 352.1 [M + H]+; 374.1 [M + Na]+.
3-(2-Carboxy-5-phenylthiophene-3-yl-methyl)-5-methylpyri-
midine-2,4-dione (64). A solution of 3-(2-carboxy-5-phenylth-
iophene-3-yl-methyl)-5-methyl-1-(tetrahydrofuran-2-yl)pyrimidine-
2,4-dione (61) (2.51 g, 6.09 mmol) in TFA (50 mL) was stirred
for 48 h at room temperature. The mixture was concentrated under
reduced pressure to leave a dark residue, which was suspended in
diethyl ether (50 mL) and filtered. The solid was washed with
diethyl ether (2 × 50 mL) and air-dried to give 64 (1.94 g, 93%)
(S)-1-(2-Amino-2-carboxyethyl)-3-(2-carboxythiophene-3-yl-
methyl)-5-methylpyrimidine-2,4-dione (55). Following a proce-
dure similar to that described for 67, a 60% suspension of sodium
hydride in mineral oil (0.33 g, 8.3 mmol), 3-(2-carboxythiophene-
3-yl-methyl)-5-methylpyrimidine-2,4-dione (52) (1.10 g, 4.14 mmol),
and (S)-3-(t-butoxycarbonylamino)oxetan-2-one (33) (0.77 g, 4.1
mmol) yielded 55 (0.56 g, 39%) as a white solid; mp 226-228 °C
(dec); [R]25D ) -7.6 (c 0.345, 6 M HCl); ee >99% as determined
1
as a pink solid. mp >300 °C; H NMR (DMSO-d6, 270.17 MHz)
δ 1.84 (s, 3H, HCCCH3), 5.26 (s, 2H, CH2Ar), 7.06 (s, 1H,
HCCCH3), 7.38-7.46 (m, 4H, Ph), 7.66-7.69 (m, 2H, Ar + Ph),
11.06 (d, J ) 5.9 Hz, 1H, NH); MS (electrospray) 343 [M + H]+;
341 [M - H]-.
1
by chiral HPLC; H NMR (D2O [+DCl, pH 1], 399.78 MHz) δ
(S)-1-(2-Amino-2-carboxyethyl)-3-(2-carboxy-5-phenylth-
iophene-3-yl-methyl)-5-methylpyrimidine-2,4-dione (67). A 60%
suspension of sodium hydride in mineral oil (0.42 g, 11 mmol)
was added to a solution of 3-(2-carboxy-5-phenylthiophene-3-yl-
methyl)pyrimidine-2,4-dione (64) (1.64 g, 4.79 mmol) in dry DMF
(170 mL) and stirred at room temperature for 18 h. (S)-3-(t-
Butoxycarbonylamino)oxetan-2-one (33) (0.90 g, 4.8 mmol) was
added, and the mixture was stirred at room temperature for 48 h.
The mixture was evaporated under reduced pressure (1 mmHg,
60 °C), and 2 M aqueous HCl was added (50 mL). The mixture
was evaporated under reduced pressure, and the residue was
suspended in water (10 mL) and applied to Dowex 50WX8-400
ion-exchange resin (H+ form) (0.25 mmol of cation/mL of resin;
100 mL) with stirring for 30 min. The mixture was then applied to
a column containing an equivalent volume of Dowex 50WX8-400
resin and eluted with water until no ninhydrin positive fractions
were observed, THF/water (1:1), until the eluate was colorless, and
then elution was continued with 1 M aqueous pyridine. The
ninhydrin positive fractions of the 1 M aqueous pyridine eluate
were combined and evaporated to dryness under reduced pressure.
Crystallization of the residue from water yielded 67 (0.84 g, 38%)
as a white solid. To further purify the final product, it was twice
stirred in hot dioxane, filtered, and air-dried. mp 275-278 °C (dec);
1.92 (d, J ) 1.1 Hz, 3H, HCCCH3), 4.35 (dd, JBA ) 15.0 Hz, JBX
) 6.6 Hz, 1H, CH2CH), 4.45 (dd, JAB ) 15.0 HZ, JAX ) 5.1 Hz,
1H, CH2CH), 4.56 (dd, JAX + JBX ) 11.7 Hz, 1H, CH2CH), 5.40
(s, 2H, CH2Ar), 6.82 (d, J ) 5.5 Hz, 1H, Ar), 7.58 (d, J ) 1.1 Hz,
HCCCH3), 7.64 (d, J ) 5.5 Hz, 1H, Ar); 13C NMR (D2O [+DCl,
pH 1], 100.54 MHz) δ 29.0, 36.7, 39.8, 98.9, 115.7, 115.6, 120.6,
128.8, 128.9, 132.3, 140.7, 153.2, 153.3, 156.8; MS (electrospray)
354 (M + H)+; 376 (M + Na)+; 352 (M - H)-.
(S)-1-(2-Amino-2-carboxyethyl)-5-bromo-3-(2-carbox-
ythiophene-3-yl-methyl)pyrimidine-2,4-dione (56). Following a
procedure similar to that described for 67, 5-bromo-3-(2-carbox-
ythiophene-3-yl-methyl)-5-pyrimidine-2,4-dione (53) (3.00 g, 9.06
mmol), a 60% suspension of sodium hydride in mineral oil (0.72
g, 18 mmol), and (S)-3-(t-butoxycarbonylamino)oxetan-2-one (33)
(1.52 g, 8.15 mmol) gave 56 (2.23 g, 59%) as a white solid; mp
192-196 °C (dec); [R]25 ) +3.2 (c 0.538, 6 M HCl); ee >98%
D
as determined by chiral HPLC; 1H NMR (TFA-d, 270.17 MHz) δ
4.70 (dd, JBA ) 15.8 Hz, JBX ) 6.3 Hz, 1H, CHCH2), 4.83 (dd,
JAB ) 15.8 Hz, JAX ) 3.0 Hz, 1H, CHCH2), 5.02 (dd, JAX + JBX
) 8.6 Hz, 1H, CHCH2), 5.76 (s, 2H, CH2Ar), 6.93 (d, J ) 5.3 Hz,
1H, Ar), 7.67 (d, J ) 5.3 Hz, 1H, Ar), 8.06 (s, 1H, HCCBr); MS
(electrospray) 418 [M + H]+; 420 [M + H]+; 416 [M - H]-; 418
[M - H]-.
[R]25 ) +1.0 (c 0.455, 6 M HCl); ee >99% as determined by
D
chiral HPLC; 1H NMR (TFA-d, 270.17 MHz) δ 2.12 (s, 3H,
HCCCH3), 4.66-4.84 (m, 2H, CHCH2), 5.05-5.02 (m, 1H,
CHCH2), 5.73 (s, 2H, CH2Ar), 7.09 (s, 1H, Ar), 7.41-7.59 (m,
6H, Ph + HCCCH3); MS (electrospray) 452 [M + Na]+.
(S)-1-(2-Amino-2-carboxyethyl)-3-(2-carboxythiophene-3-yl-
methyl)-5-trifluoromethylpyrimidine-2,4-dione (57). Following
a procedure similar to that described for 67, 3-(2-carboxythiophene-
3-yl-methyl)-5-trifluoromethylpyrimidine-2,4-dione (54) (2.00 g,
6.25 mmol), a 60% suspension of sodium hydride in mineral oil
(0.52 g, 13 mmol), and (S)-3-(t-butoxycarbonylamino)oxetan-2-one
(33) (1.17 g, 6.25 mmol) gave 57 (1.25 g, 49%) as a white solid.
mp 187-190 °C (dec); [R]25D ) -4.5 (c 0.51, 6 M HCl); ee >99%
as determined by chiral HPLC; 1H NMR (TFA-d, 270.17 MHz) δ
4.77 (dd, JBA ) 15.8 Hz, JBX ) 6.3 Hz, 1H, CH2CH), 4.87 (dd,
JAB ) 15.8 Hz, JAX ) 3.5 Hz, 1H, CH2CH), 5.05 (dd, JAX + JBX
) 9.6 Hz, 1H CH2CH), 5.73 (s, 2H, CH2Ar), 6.96 (d, J ) 5.3 Hz,
1H, Ar), 7.66 (d, J ) 5.3 Hz, 1H, Ar), 8.23 (s, 1H, HCCCF3); MS
(electrospray) 408 [M + H]+; 406 [M - H]-.
(S)-1-(2-Amino-2-carboxyethyl)-3-(2-carboxybenzyl)-5-meth-
ylpyrimidine-2,4-dione (34). Following a procedure similar to that
described for 67, a 60% suspension of sodium hydride in mineral
oil (0.44 g, 11 mmol), 3-(2-carboxybenzyl)-5-methylpyrimidine-
2,4-dione (31) (1.43 g, 5.51 mmol), and (S)-3-(t-butoxycarbony-
lamino)oxetan-2-one (33) (1.03 g, 5.51 mmol) yielded 34 (1.11 g,
58%) as a white solid. mp 217-221 °C (dec); [R]25 ) -7.2 (c
D
1
0.572, 6 M HCl); ee >99% as determined by chiral HPLC; H
NMR (D2O + NaOD [pH ) 12], 270.17 MHz) δ 3.60 (dd, JAX
+
JBX ) 14.2 Hz, 1H, CHCH2), 3.89 (dd, JBA ) 13.9 Hz, JBX ) 7.6
Hz, 1H, CHCH2), 4.05 (dd, JAB ) 13.9 Hz, JAX ) 5.9 Hz, 1H,
CHCH2), 5.34 (s, 2H, CH2Ar), 6.96-6.99 (m, 1H, Ph), 7.32-7.35
(m, 2H, Ph), 7.49 (s, 1H, HCCCH3), 7.53-7.56 (m, 1H, Ph); 13C
NMR (D2O + NaOD [pH ) 12], 67.94 MHz) δ 14.9, 46.5, 56.3,
58.0, 112.7, 127.5, 129.9, 130.6, 132.3, 135.6, 140.7, 144.6, 155.5,
168.9, 180.0, 182.4; MS (electrospray) 348 [M + H]+; HRMS m/z
(M + Na)+ calculated for C16H18N3O6, 348.1190; found, 348.1188.
(S)-1-(2-Amino-2-carboxyethyl)-3-(2-carboxybenzyl)-5-fluo-
ropyrimidine-2,4-dione (35). Following a procedure similar to that
described for 67, 3-(2-carboxybenzyl)-5-fluoropyrimidine-2,4-dione
(32) (0.48 g, 1.8 mmol), a 60% suspension of sodium hydride in
mineral oil (0.14 g, 3.6 mmol), and (S)-3-(t-butoxycarbonylamino)-
oxetan-2-one (33) (0.34 g, 1.8 mmol) gave 35 (0.22 g, 34%) as
(S)-1-(2-Amino-2-carboxyethyl)-3-(2-carboxy-4,5-dibro-
mothiophene-3-yl-methyl)-5-methylpyrimidine-2,4-dione (68).
Following a procedure similar to that described for 67, 3-(2-carboxy-
4,5-dibromothiophene-3-yl-methyl)-5-methylpyrimidine-2,4-di-
one (65) (4.67 g, 11 mmol), a 60% suspension of sodium hydride
in mineral oil (0.97 g, 24.2 mmol), and (S)-3-(t-butoxycarbony-
lamino)oxetan-2-one (33) (2.06 g, 11 mmol) gave 68 as a white
solid (0.68 g, 12%); mp 210-213 °C (dec); [R]25D ) +3.8 (c 0.495,
1
6 M HCl); ee >99% as determined by chiral HPLC; H NMR
(TFA-d, 399.78 MHz) δ 2.02 (s, 3H, HCCCH3), 4.62-4.70 (m,
1H, HCCH2), 4.79-4.86 (m, 1H, HCCH2), 4.96-5.06 (m, 1H,
HCCH2), 5.71 (s, 2H, CH2Ar), 7.50 (s, 1H, HCCCH3); MS
(electrospray) 510 [M - H]-.

1568 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Dolman et al.
(S)-1-(2-Amino-2-carboxyethyl)-3-(2-carboxyfuran-3-yl-meth-
yl)-5-methylpyrimidine-2,4-dione (69). Following a procedure
similar to that described for 67, 3-(2-carboxyfuran-3-yl-methyl)-
5-methylpyrimidine-2,4-dione (66) (1.03 g, 4.12 mmol), a 60%
suspension of sodium hydride in mineral oil (0.33 g, 8.2 mmol),
and (S)-3-(t-butoxycarbonylamino)oxetan-2-one (33) (0.77 g, 4.1
mmol) gave 69 (0.48 g, 35%) as a white solid; mp 225-227 °C
(dec); [R]25D ) -9.5 (c 0.595, 6 M HCl); ee >99% as determined
(30 min preincubation), to block NMDA receptors. Results are
expressed as mean ( SEM, n ) 3.
To further characterize the AMPA receptor antagonist activity
of compounds 55 and 67, experiments were performed on AMPA
receptors expressed on motoneurones in the hemisected neonatal
rat spinal cord preparation in the presence of tetrodotoxin (TTX,
10 µM for 2 min, then 0.1 µM continuously) to block action-
potential-dependent release. No electrical stimulation was applied,
thus allowing measurement of depolarizations evoked by an
exogenously applied AMPA receptor agonist.26 Noncumulative
concentration-response curves were obtained for the selective
AMPA receptor agonist 9 (1 min applications) in the absence and
presence of 200 µM 55 or 67 (30 min preincubation), allowing
dose ratios to be calculated. The KD values for antagonism of 9 by
55 or 67 were calculated from the dose ratios using the Gaddum-
Schild equation.
Characterization of 55 and 67 on NMDA and Group I
Metabotropic Glutamate Receptors Expressed on Neonatal Rat
Motoneurones. Experiments performed to investigate the effect
of 55 and 67 on receptors expressed on motoneurones in the
hemisected neonatal rat spinal cord preparation were carried out
in the presence of tetrodotoxin (TTX, 10 µM for 2 min, then 0.1
µM continuously) to block action-potential-dependent release. No
electrical stimulation was applied, thus allowing measurement of
depolarizations evoked by exogenously applied agonists.26 In
experiments to determine the antagonist selectivity of 55 and 67,
medium containing approximately equi-effective concentrations of
either 1 (0.7 µM), NMDA (10 µM), or DHPG (20 µM) was applied
for 1 min, in the absence and presence of 55 (10 µM; 30 min
preincubation) or 67 (1 µM; 30 min preincubation).
Calcium Fluorescence Assays Using Recombinant Human
AMPA and Kainate Receptor Subtypes. AMPA Receptor
Assays. HEK293 cells stably expressing human AMPA receptors
were seeded into poly-D-lysine-coated 96-well plates (Becton
Dickinson Labware, Bedford, MA) 1 or 2 days prior to experiments
at 60 000 cells/well (1 day) or 30 000 cells/well (2 day). Cells were
washed three times with 100 µL of assay buffer composed of Hanks
Balanced Salt Solution without phenol red (Invitrogen) with 20 mM
HEPES and 3.7 mM CaCl2 added (final [CaCl2] ) 5 mM). Plates
were then incubated for 2-3 h at room temperature in 40 µL of
assay buffer with 8 µM Fluo3-AM dye (Molecular Probes Inc.,
Eugene, OR). Following dye incubation, cells were rinsed once
with 100 µL of assay buffer. Finally, 50 µL of assay buffer, which
included the AMPA receptor potentiator LY392098 (10 µM; to
prevent desensitization of AMPA receptors), was added to wells,
and fluorescence was measured using a fluorometric imaging plate
reader (FLIPR; Molecular Devices, Sunnyvale, CA). The FLIPR
added a first addition of 50 µL of LY392098-containing assay
buffer, followed by a second addition of 100 µL of LY392098-
containing buffer 3 min later. 55 or 67 was added in the absence
of agonist during the first addition, and in the presence of 100 µM
glutamate during the second addition.
1
by chiral HPLC; H NMR (TFA-d, 399.78 MHz) δ 2.10 (s, 3H,
HCCCH3), 4.66 (dd, JBA ) 15.6 Hz, JBX ) 6.3 Hz, 1H, CH2CH),
4.78 (dd, JAB )15.6 Hz, JAX ) 2.9 Hz, 1H, CH2CH), 5.00 (dd, JAX
+ JBX ) 8.8 Hz, 1H, CH2CH), 5.60 (s, 2H, CH2Ar), 6.50 (s, 1H,
Ar), 7.54 (s, 1H, HCCCH3), 7.64 (s, 1H, Ar); MS (electrospray)
336 [M - H]-; 337 [M]-; 338 [M + H]+; 360 [M + Na]+.
(S)-1-(2-Amino-2-carboxyethyl)-5-bromo-3-(2-carboxybenzyl)-
pyrimidine-2,4-dione (36). To a solution of (S)-1-(2-amino-2-
carboxyethyl)-3-(2-carboxybenzyl)pyrimidine-2,4-dione (11) (0.15
g, 0.43 mmol) in water/acetic acid (1:1; 10 mL) was added a 2 M
solution of bromine in acetic acid (0.11 mL), and the mixture was
stirred at room temperature for 30 min. The reaction had not gone
to completion, as deduced by TLC, and a further aliquot of bromine
solution was added (0.33 mL). The mixture was stirred for a further
18 h and then concentrated under reduced pressure. The remaining
residue was suspended in water (5 mL) and applied to Dowex
50WX8-400 ion-exchange resin (H+ form, 0.25 mmol of cation/
mL of resin; 5 mL) with stirring for 30 min. The mixture was then
applied to a column containing an equivalent volume of Dowex
50WX8-400 resin (H+ form) and eluted with water until no
ninhydrin positive fractions were observed, THF/water (1:1), until
the eluate was colorless, and then elution was continued with 1 M
aqueous pyridine. The ninhydrin positive fractions of the 1 M
aqueous pyridine eluate were combined and evaporated to dryness
under reduced pressure.
Crystallization of the residue from water yielded 36 (0.11 g, 42%)
as a white solid. mp 184-185 °C (dec); [R]25D ) 6.9 (c 0.11, 6 M
1
HCl); ee 92% as determined by chiral HPLC; H NMR (TFA-d,
270.17 MHz) δ 4.72 (dd, JBA ) 15.8 Hz, JBX ) 5.9 Hz, 1H, CH2-
CH), 4.85 (dd, JAB ) 15.8 Hz, JAX ) 3.0 Hz, 1H, CH2CH), 5.05
(dd, JAX + JBX ) 9.2 Hz, 1H, CH2CH), 5.83 (s, 2H, CH2Ph), 7.14
(d, J ) 7.6 Hz, 1H, Ph), 7.62-7.48 (m, 2H, Ph), 8.08 (s, 1H,
HCCBrCO), 8.24 (dd, J ) 1.3 Hz, J ) 7.6 Hz, 1H, Ph); MS
(electrospray) 412.3 [M+].
Electrophysiology. Antagonism of Kainate Responses on
Dorsal Root C-Fibers by Novel Willardiine Derivatives. Experi-
ments to test the antagonist effect of the novel compounds at
GLUK5-containing kainate receptors were conveniently carried out
on kainate-induced responses on isolated dorsal roots from non-
anaesthetized 1-5-day-old rats, as described in detail previously.12
To prevent desensitization of kainate receptors, the dorsal root was
superfused with 1 mg mL-1 Concanavalin A for 20 min after a 20
min exposure to glucose-free superfusion medium. Standard su-
perfusion medium was then applied throughout the experiments.
This allowed measurement of depolarizations evoked by the
exogenously applied agonist, kainate (1 min applications). Non-
cumulative, nonsequential concentration-response curves were
constructed for kainate in the absence and presence of the antagonist
(30 min preincubation). The EC50 values (10-15 µM) for kainate-
induced depolarization of the dorsal root determined in assays of
antagonists did not differ significantly from previously reported
values.12,24a,25
Characterization of AMPA Receptor Antagonists. Hemisected
spinal cords from non-anaesthetized 1-5-day-old rats killed by
cervical dislocation were prepared and used according to the
reported method.38 To assess AMPA receptor antagonist activity,
the ability of the compounds to block the fast component of the
dorsal root-evoked ventral root potential (fDR-VRP) in the neonatal
rat hemisected spinal cord preparation was measured, as described
in detail previously.12,26 Concentration-response curves were
constructed for test antagonists (5 min applications), in the presence
of 2 mM MgSO4/50 µM (R)-2-amino-5-phosphonopentanoic acid
Kainate Receptor Assays. All receptor clones were stably
expressed in HEK293 cells. The GLUK5(Q)/GLUK2 cell line was
created by retroviral infection of cDNA coding for the human
GLUK2 subunit (EAA2; Allelix Biopharmaceuticals) into the
GLUK5(Q)-expressing cell line using the pMNLZRS/IB retroviral
expression vector. HEK293 cell lines stably expressing a cloned
GLUK5(Q)39 or GLUK6(Q) receptor subunit,40 or coexpressing
GLUK5(R) and GLUK6(Q)5c, or GLUK6(Q) and GLUK2 41 have been
,
previously described. Kainate receptor expression levels for all
transfected cell lines have been previously determined in our
laboratory by saturation binding of [3H]kainate to intact cells. Bmax
values for specific [3H]kainate binding are GLUK5(Q), 1.7 ( 0.5
pmol/mg; GLUK5(R)/6(Q), 8 ( 2 pmol/mg; GLUK5(Q)/GLUK2, 0.6
( 0.1 pmol/mg; GLUK6(Q), 2.7 ( 0.3 pmol/mg; GLUK6(Q)/GLUK2
,
1.7 ( 0.3 pmol/mg.42 To address the possibility that changes in
receptor expression levels over time could influence experimentally
determined IC50 values, the EC50 value for glutamate and the IC50
values for routinely used control antagonists such as 4 (Figure 1)
were continuously monitored, and no significant drift in these values

N3-Substituted Willardiine DeriVatiVes
Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7 1569
Fantaske, R.; Ogden, A.; Gates, M.; Ornstein, P. L.; Lodge, D.;
Bleakman, D.; Collingridge, G. L. Kainate receptors are involved in
synaptic plasticity. Nature 1999, 402, 297-301. (d) Filla, S. A.;
Winter, M. A.; Johnson, K. W.; Bleakman, D.; Bell, M. G.; Bleisch,
T. J.; Castano, A. M.; Clemens-Smith, A.; Del Prado, M.; Dieckman,
D. K.; Dominguez, E.; Escribano, A.; Ho, K. H.; Hudziak, K. J.;
Katofiasc, M. A.; Martinez-Perez, J. A.; Mateo, A.; Mathes, B. M.;
Mattiuz, E. L.; Ogden, A. M. L.; Phebus, L. A.; Stack, D. R.;
Stratford, R. E.; Ornstein, P. L. Ethyl (3S,4aR,6S,8aR)-6-(4-ethoxy-
carbonylimidazol-1-ylmethyl)decahydroisoq uinoline-3-carboxylic
ester: a prodrug of a GluR5 kainate receptor antagonist active in
two animal models of acute migraine. J. Med. Chem. 2002, 45, 4383-
4386. (e) Dominguez, E.; Iyengar, S.; Shannon, H. E.; Bleakman,
D.; Alt, A.; Arnold, B. M.; Bell, M. G.; Bleisch, T. J.; Buckmaster,
J. L.; Castano, A. M.; Prado, M. D.; Escribano, A.; Filla, S. A.; Ho,
K. H.; Hudziak, K. J.; Jones, C. K.; Martinez-Perez, J. A.; Mateo,
A.; Mathes, B. M.; Mattiuz, E. L.; Ogden, A. M. L.; Simmons, R.
M. A.; Stack, D. R.; Stratford, R. E.; Winter, M. A.; Wu, Z.; Ornstein,
P. L. Two prodrugs of potent and selective GluR5 kainate receptor
antagonists active in three animal models of pain. J. Med. Chem.
2005, 48, 4200-4203. (f) Weiss, B.; Alt, A.; Ogden, A. M.; Gates,
M.; Dieckman, D. K.; Clemens-Smith, A.; Ho, K. H.; Jarvie, K.;
Rizkalla, G.; Wright, R. A.; Calligaro, D. O.; Schoepp, D.; Mattiuz,
E. L.; Stratford, R. E.; Johnson, B.; Salhoff, C.; Katofiasc, M.;
Phebus, L. A.; Schenck, K.; Cohen, M.; Filla, S. A.; Ornstein, P. L.;
Johnson, K. W.; Bleakman, D. Pharmacological characterization of
the competitive GLUK5 receptor antagonist decahydroisoquinoline
LY466195 in vitro and in vivo. J. Pharmacol. Exp. Ther. 2006, 318,
772-781.
was ever observed. Additionally, cells were never passaged more
than 20 times.
Cell growth and ion influx studies using a FLIPR were carried
out exactly as described previously, in the presence of concanavalin
A.12 The antagonist 55 or 67 was added in the absence of agonist
during the first addition, and in the presence of 100 µM glutamate
during the second addition. Concentration-response curves for 55
and 67 were analyzed using GraphPad Prism 3.02 software (San
Diego, CA), with slope factor fixed at 1, and top and bottom fixed
at 100% and 0% inhibition, respectively. The dissociation constant
(Kb) was calculated according to the Cheng-Prusoff equation43 from
the IC50 value for inhibiting 100 µM glutamate-induced calcium
influx.
Acknowledgment. We would like to thank the following
organizations for funding this work: BBSRC, UK and the Brain
Research Center of the 21st Century Frontier Research Program
funded by the Korean Ministry of Science and Technology. We
also would like to thank Louise C. Argent and Anna Thom from
Tocris Bioscience for providing chiral HPLC and mass spec-
trometry data. We thank Dr. Richard Sessions (Department of
Biochemistry, University of Bristol) for access to the docking
software and the Centre for Computational Chemistry (Univer-
sity of Bristol) for access to their computing resources. O.T.P.
was supported by the Academy of Finland. This research was
funded in part by the Intramural Research Program of the NIH,
National Institute of Child Health and Human Development.
(6) (a) O’Neill, M. J.; Bond, A.; Ornstein, P. L.; Ward, M. A.; Hicks,
C. A.; Hoo, K.; Bleakman, D.; Logde, D. Decahydroisoquinolines:
novel competitive AMPA/kainate antagonists with neuroprotective
effects in global cerebral ischaemia. Neuropharmacology 1998, 37,
1211-1222. (b) Simmons, R. M. A.; Li, D. L.; Hoo, K. H.; Deverill,
M.; Ornstein, P. L.; Iyengar, S. Kainate GluR5 receptor mediates
the nociceptive response to formalin in the rat. Neuropharmacology
1998, 37, 25-36. (c) O’Neill, M. J.; Bogaert, L.; Hicks, C. A.; Bond,
A.; Ward, M. A.; Ebinger, G.; Ornstein, P. L.; Michotte, Y.; Lodge,
D. LY377770, a novel iGluR5 kainate receptor antagonist with
neuroprotective effects in global and focal cerebral ischaemia.
Neuropharmacology 2000, 39, 1575-1588. (d) Smolders, I.; Bor-
tolotto, Z. A.; Clarke, V. R. J.; Warre, R.; Khan, G. M.; O’Neill, M.
J.; Ornstein, P. L.; Bleakman, D.; Ogden, A.; Weiss, B.; Stables, J.
P.; Ho, K. H.; Ebinger, G.; Collingridge, G. L.; Lodge, D.; Michotte,
Y. Antagonists of GLUK5-containing kainate receptors prevent
pilocarpine-induced limbic seizures. Nat. Neurosci. 2002, 5, 796-
804.
(7) Sanders, J. M.; Ito, K.; Settimo, L.; Pentikainen, O. T.; Shoji, M.;
Sasaki, M.; Johnson, M. S.; Sakai, R.; Swanson, G. T. Divergent
pharmacological activity of novel marine-derived excitatory amino
acids on glutamate receptors. JPET 2005, 314, 1068-1078.
(8) Sasaki, M.; Maruyama, T.; Sakai, R.; Tachibana, K. Synthesis and
biological activity of dysiherbaine model compound. Tetrahedron
Lett. 1999, 40, 3195-3198.
(9) (a) Christensen, J. K.; Varming, T.; Ahring, P. K.; Jørgensen, T. D.;
Nielsen, E. Ø. In vitro characterization of 5-carboxyl-2,4-di-benza-
midobenzoic Acid (NS3763), a noncompetitive antagonist of GLUK5
receptors. J. Pharmacol. Exp. Ther. 2004, 309, 1003-1010. (b)
Valgeirsson, J.; Nielsen, E. Ø.; Peters, D.; Kristensen, A. S.; Madsen,
U. Bioisosteric Modifications of 2-arylureidobenzoic acids: selective
noncompetitive antagonists for the homomeric kainate receptor
subtype GluR5. J. Med. Chem. 2004, 47, 6948-6957.
(10) (a) Jane, D. E.; Pook, P. C.-K.; Sunter, D. C.; Udvarhelyi, P. M.;
Watkins, J. C. New willardiine analogs with potent stereoselective
actions on mammalian spinal neurones. Br. Pharmacol. 1991, 104,
333P. (b) Patneau, D. K.; Mayer, M. L.; Jane, D. E.; Watkins, J. C.
Activation and desensitization of AMPA/kainate receptors by novel
derivatives of willardiine. J. Neurosci. 1992, 12, 595-606. (c) Wong,
L. A.; Mayer, M. L.; Jane, D. E.; Watkins, J. C. Willardiines
differentiate agonist binding sites for kainate-versus AMPA-preferring
glutamate receptors in DRG and hippocampal neurones. J. Neurosci.
1994, 14, 3881-3897. (d) Jane, D. E.; Hoo, K.; Kamboj, R.; Deverill,
M.; Bleakman, D.; Mandelzys, A. Synthesis of Willardiine and
6-Azawillardiine analogs: Pharmacological characterization on cloned
homomeric human AMPA and kainate receptor subtypes. J. Med.
Chem. 1997, 40, 3645-3650.
Supporting Information Available: Elemental analyses for
intermediates and final compounds. Experimental and spectroscopic
data for intermediates. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
(1) (a) Bettler, B.; Mulle, C. Neurotransmitter receptors. 2. AMPA and
kainate receptors. Neuropharmacology 1995, 34, 123-139. (b)
Fletcher, E. J.; Lodge, D. New developments in the molecular
pharmacology of R-amino-3-hydroxy-5-methyl-4-isoxazolepropionate
and kainate receptors. Pharmacol. Ther. 1996, 70, 65-89. (c)
Bleakman, D.; Lodge, D. Neuropharmacology of AMPA and kainate
receptors. Neuropharmacology 1998, 37, 1187-1204. (d) Chittajallu,
R.; Braithwaite, S. P.; Clarke, V. R.; Henley, J. M. Kainate
receptors: subunits, synaptic localization and function. Trends
Pharmacol. Sci. 1999, 20, 26-35.
(2) (a) Watkins, J. C.; Jane, D. E. The glutamate story. Br. J. Pharmacol.
2006, 147, S100-S108. (b) Jane, D. E. Antagonists acting at the
NMDA receptor complex: potential for therapeutic applications. In
Glutamate and GABA Receptors and Transporters; Krogsgaard-
Larsen, P., Egebjerg, J., Schousboe, A., Eds.; Taylor and Francis:
London, 2002; pp 69-98. (c) Bra¨uner-Osborne, H.; Egebjerg, J.;
Nielsen, E. O.; Madsen, U.; Krogsgaard-Larsen, P. Ligands for
glutamate receptors: design and therapeutic prospects. J. Med. Chem.
2000, 43, 2609-2645. (d) Kew, J. N. C.; Kemp, J. A. Ionotropic
and metabotropic glutamate receptor structure and pharmacology.
Psychopharmacology 2005, 179, 4-29.
(3) Lodge, D.; Dingledine, R. Ionotropic Glutamate Receptors. The
IUPHAR Compendium of Receptor Characterization and Classifica-
tion, 2nd ed.; IUPHAR Media Ltd.: London, 2001.
(4) (a) Sheardown, M.; Nielsen, E.; Hansen, A.; Jacobsen, P.; Honore,
T. 2,3-Dihydroxy-6-nitro-7-sulphamoyl-benzo(F)quinoxaline: A neu-
roprotectant for cerebral ischemia. Science 1990, 247, 571-574. (b)
Watkins, J. C.; Krogsgaard-Larsen, P.; Honore´, T. Structure-activity
relationships in the development of excitatory amino acid receptor
agonists and competitive antagonists. Trends Pharmacol. Sci. 1990,
11, 25-33.
(5) (a) Clarke, V. R.; Ballyk, B. A.; Hoo, K. H.; Mandelzys, A.; Pellizzari,
A.; Bath, C. P.; Thomas, J.; Sharpe, E. F.; Davies, C. H.; Ornstein,
P. L.; Schoepp, D. D.; Kamboj, R. K.; Collingridge, G. L.; Lodge,
D.; Bleakman, D. A hippocampal GluR5 kainate receptor regulating
inhibitory synaptic transmission. Nature 1997, 389, 599-603. (b)
Vignes, M.; Clarke, V. R.; Parry, M. J.; Bleakman, D.; Lodge, D.;
Ornstein, P. L.; Collingridge, G. L. The GluR5 subtype of kainate
receptor regulates excitatory synaptic transmission in areas CA1 and
CA3 of the rat hippocampus. Neuropharmacology 1998, 37, 1269-
1277. (c) Bortolotto, Z. A.; Clarke, V. R.; Delany, C. M.; Parry, M.
C.; Smolders, I.; Vignes, M.; Ho, K. H.; Miu, P.; Brinton, B. T.;
(11) (a) Dolman, N. P.; Troop, H. M.; More, J. C. A.; Alt, A. J.; Ogden,
A. M.; Jones, S.; Morley, R. M.; Roberts, P. J.; Bleakman, D.;
Collingridge, G. L.; Jane, D. E. Synthesis and pharmacology of
willardiine derivatives acting as antagonists of kainate receptors. J.
Med. Chem. 2005, 48, 7867-7881. (b) Dolman, N. P.; More, J. C.

1570 Journal of Medicinal Chemistry, 2007, Vol. 50, No. 7
Dolman et al.
A.; Alt, A.; Ogden, A. M.; Troop, H. M.; Bleakman, D.; Collingridge,
G. L.; Jane, D. E. Structure-activity relationship studies on N3-
substituted willardiine derivatives acting as AMPA or kainate receptor
antagonists. J. Med. Chem. 2006, 49, 2579-2592.
(28) Word, J. M.; Lovell, S. C.; Richardson, J. S.; Richardson, D. C.
Asparagine and glutamine: using hydrogen atom contacts in the
choice of side-chain amide orientation. J. Mol. Biol. 1999, 285,
1735-1747.
(29) Lovell, S. C.; Word, J. M.; Richardson, J. S.; Richardson, D. C. The
penultimate rotamer library. Proteins 2000, 40, 389-408.
(30) Lehtonen, J. V.; Still, D. J.; Rantanen, V. V.; Ekholm, J.; Bjo¨rklund,
D.; Iftikhar, Z.; Huhtala, M.; Repo, S.; Jussila, A.; Jaakkola, J.;
Pentika¨inen, O.; Nyro¨nen, T.; Salminen, T.; Gyllenberg, M.; Johnson,
M. S. BODIL: a molecular modeling environment for structure-
function analysis and drug design. J. Comput.-Aided Mol. Des. 2004,
18, 401-419.
(12) More, J. C. A.; Nistico, R.; Dolman, N. P.; Clarke, V. R. J.; Alt, A.
J.; Ogden, A. M.; Buelens, F. P.; Troop, H. M.; Kelland, E. E.; Pilato,
F.; Bleakman, D.; Bortolotto, Z. A.; Collingridge, G. L.; Jane, D. E.
Characterisation of UBP296; a novel, potent and selective kainate
receptor antagonist. Neuropharmacology 2004, 47, 46-64.
(13) Barker, G. R. I.; Warburton, E. C.; Koder, T.; Aggleton, J. P.; Bashir,
Z. I.; Auberson, Y.; Jane, D. E.; Brown, M. W. Kainate and NMDA
glutamate receptor antagonism impairs recognition memory at
different times. J. Neurosci. 2006, 26, 3561-3566.
(31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin,
K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone,
V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G.
A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.;
Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J.
W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.;
Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghava-
chari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-
Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill,
P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople,
J. A. Gaussian 03, revision C.02; Gaussian, Inc.: Wallingford, CT,
2004.
(32) Jones, G.; Willett, P.; Glen, R. C. A genetic algorithm for flexible
molecular overlay and pharmacophore elucidation. J. Comput.-Aided
Mol. Des. 1995, 9, 532-549.
(33) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
Development and validation of a genetic algorithm for flexible
docking. J. Mol. Biol. 1997, 267, 727-748.
(34) Merritt, E. A.; Bacon, D. J. Raster3D: Photorealistic molecular
graphics. Methods Enzymol. 1997, 277, 505-524.
(35) Pentika¨inen, U.; Settimo, L.; Johnson, M. S.; Pentika¨inen, O. T.
Subtype selectivity and flexibility of ionotropic glutamate receptors
upon antagonist ligand binding. Org. Biomol. Chem. 2006, 4, 1058-
1070.
(36) Giller, S. A.; Zhuk, R. A.; Lidak, M. J.; Ziderman, A. A. GB Patent
Specification 1,168,391, 1969.
(37) Salimbeni, A.; Canevotti, R.; Paleari, F.; Poma, D.; Caliarri, S.; Fici,
F.; Cirillo, R.; Renzetti, A. R.; Subissi, A.; Belvisi, L.; Bravi, G.;
Scolastico, C.; Gichetti, A. N-3-Substituted pyrimidones as potent,
orally active, AT1 selective angiotensin II receptor antagonists. J.
Med. Chem. 1995, 38, 4806-4820.
(38) Evans, R. H.; Francis, A. A.; Jones, A. W.; Smith, D. A. S.; Watkins,
J. C. The effects of a series of ω-phosphonic R-carboxylic amino
acids on electrically evoked and excitant amino acid-induced
responses in isolated spinal cord preparations. Br. J. Pharmacol. 1982,
75, 65-75.
(39) Korczak, B.; Nutt, S. L.; Fletcher, E. J.; Hoo, K. H.; Elliott, C. E.;
Rampersad, V.; McWhinnie, E. A.; Kamboj, R. K. cDNA cloning
and functional properties of human glutamate receptor EAA3 (GluR5)
in homomeric and heteromeric configuration. Recept. Channels 1995,
3, 41-49.
(40) Hoo, K. H.; Nutt, S. L.; Fletcher, E. J.; Elliott, C. E.; Korczak, B.;
Deverill, R. M.; Rampersad, V.; Fantaske, R. P.; Kamboj, R. K.
Functional expression and pharmacological characterization of the
human EAA4 (GluR6) glutamate receptor: a kainate selective channel
subunit. Recept. Channels 1994, 2, 327-337.
(14) (a) Nomura, H.; Yoshioka, Y.; Minami, I.; Synthesis of tetrahydro-
2-furyl derivatives of 5-substituted uracils. Chem. Pharm. Bull. 1979,
27, 899-906. (b) Noell, C. W.; Cheng, C. C. Pyrimidines. XXI.
1-(Tetrahydro-2-furyl)pyrimidines (1). J. Heterocycl. Chem. 1968,
5, 25-28.
(15) Dalpozzo, R.; De Nino, A.; Maiuolo, L.; Procopio, A.; Romeo, R.;
Sindona, G. A convenient method for the synthesis of N-vinyl
derivatives of nucleosides. Synthesis 2002, 2, 172-174.
(16) Kametani, T.; Kigasawa, K.; Hiiragi, M.; Wakisaka, K.; Kusama,
O.; Kawasaki, K.; Sugi, H. Studies on the synthesis of chemothera-
peutics. I. Synthesis of 1-(2-tetrahydrofuryl)-5-fluorouracil [Ftorafur]
(studies on the synthesis of heterocyclic compounds. J. Heterocycl.
Chem. 1977, 14, 473-475.
(17) Love, K. R.; Andrade, R. B.; Seeberger, P. H. Linear synthesis of a
protected H-Type II pentasaccharide using glycosyl phosphate
building blocks. J. Org. Chem. 2001, 66, 8165-8176.
(18) (a) Arnold, L. D.; Drover, J. C. G.; Vederas, J. C. Conversion of
serine beta-lactones to chiral alpha amino acids by copper-containing
organolithium and organomagnesium reagents. J. Am. Chem. Soc.
1987, 109, 4649-4659. (b) Arnold, L. D.; Kalantar, T. H.; Vederas,
J. C. Conversion of serine to stereochemically pure â-substituted
R-amino acids via â-lactones. J. Am. Chem. Soc. 1985, 107, 7105-
7109. (c) Ramer, S. E.; Moore, R. N.; Vederas, J. C. Mechanism of
formation of serine â-lactones by Mitsunobu cyclization: synthesis
and use of L-serine stereospecifically labelled with deuterium at C-3.
Can. J. Chem. 1986, 64, 706-713.
(19) Charonnat, J. A.; Muchowski, J. M.; Nelson, P. H. The synthesis of
four isomeric bisthienothiepinones [1]. J. Heterocycl. Chem. 1983,
20, 1085-1087.
(20) Prugh, J. D.; Huitric, A. C.; McCarthy, W. The synthesis and proton
magnetic resonance spectra of some brominated furans. J. Org. Chem.
1964, 29, 1991-1994.
(21) Mayer, M. L.; Ghosal, A.; Dolman, N. P.; Jane, D. E. Crystal
structures of the kainate receptor GluR5 ligand binding core dimer
complex with novel selective antagonists. J. Neurosci. 2006, 26,
2852-2861.
(22) Campaigne, E.; White, R. L. Synthesis of unsymmetrical arylth-
iophenes and bithienyls via oxidative cyclisation of 1,3-butadiene-
1-thiols. J. Heterocycl. Chem. 1988, 25, 367-373.
(23) Steinkopf, W.; Hanske, W. Studien in der thiophenreihe.XXXVII.
Die Jodderivate des 3-Thiotoluns. Justus Liebigs Ann. Chem. 1937,
532, 236-249.
(24) (a) Agrawal, S. G.; Evans, R. H. The primary afferent depolarizing
action of kainate in the rat. Br. J. Pharmacol. 1986, 87, 345-355.
(b) Bettler, B.; Boulter, J.; Hermans-Borgmeyer, I.; O’Shea-Green-
field, A.; Deneris, E. S.; Moll, C.; Borgmeyer, U.; Hollman, M.;
Heinemann, S. Cloning of a novel glutamate receptor subunit,
GluR5: expression in the central nervous system during development.
Neuron 1990, 5, 583-595. (c) Wilding, T. J.; Huettner, J. E.
Functional diversity and developmental changes in rat neuronal
kainate receptors. J. Physiol. 2001, 532.2, 411-421. (d) Kerchner,
G. A.; Wilding, T. J.; Huettner, J. E.; Zhuo, M. Kainate receptor
subunits underlying presynaptic regulation of transmitter release in
the dorsal horn. J. Neurosci. 2002, 22, 8010-8017.
(41) Bleakman, D.; Ogden, A. M.; Ornstein, P. L.; Hoo, K. Pharmacologi-
cal characterization of a GluR6 kainate receptor in cultured hippoc-
ampal neurons. Eur. J. Pharmacol. 1999, 378, 331-337.
(42) Alt, A.; Weiss, B.; Ogden, A. M.; Knauss, J. L.; Oler, J.; Ho, K.;
Large, T. H.; Bleakman, D. Pharmacological characterization of
glutamatergic agonists and antagonists at recombinant human ho-
momeric and heteromeric kainate receptors in vitro. Neuropharma-
cology 2004, 46, 793-806.
(43) Cheng, Y. C.; Prusoff, W. H. Relationship between the inhibition
constant (Ki) and the concentration of inhibitor which causes 50 per
cent inhibition (IC50) of an enzymatic reaction. Biochem. Pharmacol.
1973, 22, 3099-3108.
(25) Pook, P.; Brugger, F.; Hawkins, N. S.; Clark, K. C.; Watkins, J. C.;
Evans, R. H. A comparison of the actions of agonists and antagonists
at non-NMDA receptors of C fibres and motoneurones of the
immature rat spinal cord in vitro. Br. J. Pharmacol. 1993, 108, 179-
184.
(26) More, J. C. A.; Troop, H. M.; Jane, D. E. The novel antagonist
3-CBW discriminates between kainate receptors expressed on
neonatal rat motoneurones and those on dorsal root C-fibres. Br. J.
Pharmacol. 2002, 137, 1125-1133.
(27) Hogner, A.; Greenwood, J. R.; Liljefors, T.; Lunn, M. L.; Egebjerg,
J.; Larsen, I. K.; Gouaux, E.; Kastrup, J. S. Competitive antagonism
of AMPA receptors by ligands of different classes: crystal structure
of ATPO bound to the GluR2 ligand-binding core, in comparison
with DNQX. J. Med. Chem. 2003, 46, 214-221.
JM061041U