1H-cyclopentapyrimidine-2,4(1H,3H)-dione-related ionotropic glutamate receptors ligands. structure-activity relationships and identification of potent and selective iGluR5 modulators

Article abstract of DOI:10.1021/jm800865a

(S)-CPW399 ((S)-1) is a potent and excitotoxic AMPA receptor partial agonist. Modifying the cyclopentane ring of (S)-1, we developed two of the most potent and selective functional antagonists (5 and 7) for kainate receptor (KA-R) subunit iGluR5. Derivatives 5 and 7, with their unique pharmacological profile, may lead to a better understanding of the different roles and modes of action of iGluR1-5 subunits, paving the way for the synthesis of new potent, subunit selective iGluR5 modulators.

Full text of DOI:10.1021/jm800865a

6614
J. Med. Chem. 2008, 51, 6614–6618
1H-Cyclopentapyrimidine-2,4(1H,3H)-dione-Related Ionotropic Glutamate Receptors Ligands.
Structure-Activity Relationships and Identification of Potent and Selective iGluR5 Modulators
Stefania Butini,¶, Darryl S. Pickering,‡ Elena Morelli,§, Salvatore Sanna Coccone,¶, Francesco Trotta,¶, Meri De Angelis,¶,
Egeria Guarino,¶, Isabella Fiorini,¶, Giuseppe Campiani,*,¶, Ettore Novellino,§, Arne Schousboe,‡ Jeppe K. Christensen,# and
Sandra Gemma¶,
European Research Centre for Drug DiscoVery and DeVelopment (NatSynDrugs), Banchi di Sotto 55, 53100 Siena, Italy, Dipartimento
Farmaco Chimico Tecnologico (DFCT), UniVersita` degli Studi di Siena, Via Aldo Moro, 53100 Siena, Italy, Dipartimento di Chimica
Farmaceutica e Tossicologica (DCFT), UniVersita` di Napoli Federico II, Via D. Montesano 49, 80131 Napoli, Italy, Department of
Pharmacology and Pharmacotherapy, Faculty of Pharmaceutical Sciences, UniVersity of Copenhagen, 2 UniVersitetsparken,
DK-2100 Copenhagen, Denmark, and NeuroSearch A/S, PederstrupVej 97, Ballerup Denmark
ReceiVed July 14, 2008
(S)-CPW399 ((S)-1) is a potent and excitotoxic AMPA receptor partial agonist. Modifying the cyclopentane
ring of (S)-1, we developed two of the most potent and selective functional antagonists (5 and 7) for kainate
receptor (KA-R) subunit iGluR5. Derivatives 5 and 7, with their unique pharmacological profile, may lead
to a better understanding of the different roles and modes of action of iGluR1-5 subunits, paving the way
for the synthesis of new potent, subunit selective iGluR5 modulators.
Introduction
of the ligand-binding cores of AMPA-R and KA-R subunits
recently boosted the development of subtype selective iGluRs
modulators.4-7 It is now accepted that among the cloned AMPA
and KA iGluRs, subunits iGluR1-4 are AMPA sensitive, while
the KA-Rs are composed of iGluR5-7 and KA1,2 subunits.
iGluR5-7 in combination with KA1 and KA2 subunits can form
functional homomeric or heteromeric ion channels. KA-Rs are
widely distributed in the CNS, including the dorsal horn and
dorsal root ganglion (DRG) neurons. The iGluR5 subunit is the
most readily detectable at presynaptic sites in the DRG. iGluRs
play a principal role in spinal cord nociceptive transmission,
and activation of iGluRs in central endings of DRG neurons
depolarizes primary afferents and may sustain presynaptic
inhibition of C fibers. Thus, presynaptic KA-Rs may play an
important role in the modulation of nociception, being a novel
target for therapeutic strategies. However, the role of KA-Rs
in brain function is not fully understood yet; therefore, potent
and highly selective agonists and antagonists are still needed.
L-Glutamate (Glu) is the major excitatory neurotransmitter
in the mammalian central nervous system (CNS), and it
modulates the functions of most neuronal circuits in the CNS.
The physiological and pathological actions of Glu are mediated
by activation of a range of excitatory amino acid transporters,
ionotropic glutamate receptors (iGluRs, ligand-gated ion chan-
nels), and the metabotropic glutamate receptors (mGluRs,
G-protein-coupled). Binding of Glu to iGluRs, which are
abundantly expressed in the brain, is a key step in the
predominant mechanism of rapid excitatory synaptic transmis-
sion in the CNS. Even if the complex roles of the iGluRs are
far from being understood in detail, it is generally accepted that
these receptors are implicated in learning and memory formation
and are associated with a number of psychiatric and neurological
disorders such as Alzheimer’s, Parkinson’s, and Huntington’s
diseases, schizophrenia, amyotrophic lateral sclerosis, and
epilepsy.1 iGluRs in the mammalian brain are encoded by a
family of 18 genes that coassemble to form the kainate,
N-methyl-D-aspartate and (S)-2-amino-3-(5-methyl-3-hydroxy-
isoxazol-4-yl)propanoic acid receptors (KA-R, NMDA-R, and
AMPA-R, respectively). Coassembly of these ion channel
proteins within the families generates several receptor subtypes.
iGluRs form tetrameric ligand-gated ion channels, and one copy
of the ligand-binding core, as a discrete domain, is present in
each subunit.2
We have previously reported the synthesis of (S)-CPW399
((S)-1, Chart 1),4 a iGluRs ligand characterized by a bicyclic
scaffold and structurally related to the naturally occurring (S)-
willardiine (2a). Derivative (S)-1 was found to be a selective,
partial agonist for iGluR1/iGluR2 subunits that, unlike (S)-2a,
stimulated an increase in [Ca2+] in mouse cerebellar granule
cells and induced neuronal cell death. Recently we provided
the X-ray structure of (S)-1 in complex with the ligand binding
core of the iGluR2 subunit and with the (Y702F)GluR2-S1S2J
mutant, enabling an understanding of the molecular mechanisms
underlying the iGluR1,2 vs iGluR3,4 subtype selectivity of
(S)-1.8
To improve the structure-activity relationships (SARs) of
the class of bicyclic analogues of willardiine, typified by (S)-1,
a new set of related analogues was developed on the basis of
the X-ray structures of (S)-1 in complex with the ligand-binding
core of iGluR2 (GluR2-S1S2J) and (Y702F)GluR2-S1S2J and
on the basis of the crystal structures of the iGluR5 and iGluR6
KA-R ligand binding cores in complexes with Glu (the
molecular modeling studies will be reported elsewhere). A
number of SAR trends was identified in the series of the bicyclic
analogues of (S)-1, and two of the most potent iGluR5 selective
A major focus in studying the mechanism of excitotoxicity
and excitatory amino acid induced neuronal cell death has
involved the role of the NMDA subtype of iGluRs because of
its ability to flux large amounts of potentially lethal Ca2+ ions
into the cell.3 Less attention has been directed to AMPA-Rs
and KA-Rs, since for many years selective agonists and
antagonists were not available. The successful structural analysis
* To whom correspondence should be addressed. Phone: 0039-0577-
234172. Fax: 0039-0577-234333. E-mail: campiani@unisi.it.
¶
Universita` degli Studi di Siena.
European Research Centre for Drug Discovery and Development.
University of Copenhagen.
Universita` di Napoli Federico II.
NeuroSearch A/S.
‡
§
#
10.1021/jm800865a CCC: $40.75
2008 American Chemical Society
Published on Web 09/24/2008

Brief Articles
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 20 6615
Chart 1. Reference and Title Compounds
and 50 and 112 nM for 7 (Table 1), while on iGluR3-iGluR4
subtypes showed an affinity in the submicromolar range. When
tested against KA-R subunits, iGluR5 and iGluR6 expressed in
Sf9 cells, compounds 5 and 7 showed a very high affinity. Both
compounds, displaying a Ki of 5 and 5.3 nM at iGluR5,
respectively, and a very weak occupancy of the iGluR6 subunit,
may be considered two of the most potent and selective iGluR5
ligands known to date. They are around 8 times more potent
than (S)-1 and 15-fold more potent than kainate. By comparison
of the affinity for subtypes iGluR1-4 and for iGluR5,6, it
appears that the analogue (S)-7 is a potent and selective ligand
for iGluR5 KA-R subtype.
2. Electrophysiological Assays. The potencies (EC50) of (S)-
1, 5, and 7 at recombinant rat AMPA-R was measured on
receptors expressed in Xenopus laeVis oocytes (Table 2). The
efficacies of (S)-1, 5, and 7 were also determined in comparison
to 1 mM Glu, all in the presence of 100 µM cyclothiazide (CTZ)
to block receptor desensitization. The three compounds behave
as partial agonists at AMPA-Rs, and their efficacy at iGluR2(Q)i
is as follows (mean ( SEM): (S)-1, 0.432 ( 0.036 (n ) 21); 5,
0.376 ( 0.025 (n ) 12); 7, 0.329 ( 0.033 (n ) 10). These
efficacy values are not statistically significantly different from
one another (P ) 0.136, one-way ANOVA). As was observed
with the Ki values, 5 and 7 are more potent at iGluR1,2 than at
iGluR3,4, with derivative 7 showing better selectivity for
iGluR1-iGluR2 than 5. It is noteworthy that while derivative
5 has a similar potency for all AMPA-R subtypes (EC50 ranging
from 11 to 34 µM), 7 is about 10-fold more potent at iGluR1,2
than at iGluR3,4. This difference in selectivity between deriva-
tives 5 and 7 at iGluR3,4 could indicate that the efficacy and/
or desensitization properties of 7 at these subunits are different
from those of 5.
The pharmacology of 5 and 7 was further examined at human
iGluR1i and human iGluR5(Q)1b expressed in the mammalian
cell line CHO-K1 and using whole-cell patch-clamp electro-
physiology. Both compounds were weak partial agonists at
human iGluR1i, and responses could only be determined in the
presence of CTZ (30 µM) to block receptor desensitization
(Figure 1): 5, EC50 ) 18.5 µM (95% CI, 4.3-78 µM); 7, EC50
) 4.1 µM (95% CI, 2.3-7.3 µM). These potencies of 5 and 7
are comparable to those measured at rat iGluR1i expressed in
X. laeVis oocytes (Table 2), although the efficacies (0.16 and
0.27, respectively) are somewhat lower than those measured at
rat iGluR2(Q)i, indicating a species or iGluR subtype difference
in desensitization. Both 5 and 7 were also very weak partial
agonists at human iGluR5(Q)1b and exhibited rapid desensitiza-
tion kinetics such that complete desensitization was obtained
within 10 ms (Figure 2). Because of their high affinity, extremely
slow washout, and their weak responses, EC50 could not be
determined for 5 and 7 at iGluR5(Q)1b. However, efficacy
estimates were made using saturating concentrations of 5 and
7 with 10 mM Glu as the full agonist control (ε ) 1): 5, ε )
0.12; 7, ε ) 0.35 (Figure 2).
antagonists, the sulfur-bridged analogues 5 and 7, were identi-
fied. These compounds represent novel pharmacological tools
for investigation of the role of the iGluR5 subunit in neurotoxic
and neurodegenerative diseases.
Results and Discussion
AMPA-R and KA-R subtype selectivity was investigated by
examining the binding affinity of 1-8 and 11 (9 and 10 could
not be tested, being insoluble in the test conditions) at the
recombinant rat AMPA-Rs (GluR1-4) and KA-Rs (GluR5,6)
expressed in Sf9 cells (Table 1); a comparison was made with
(R,S)-AMPA, (L)-Glu, KA, (S)-1, (R,S)-2a, and its halogenated
analogues (2b-e). Table 2 summarizes the potencies (EC50) of
willardiine analogues and compounds (S)-1, 5, and 7 at
recombinant AMPA-Rs expressed in Xenopus laeVis oocytes.
In Vitro Pharmacology. 1. Radioligand Binding Assays. All
the compounds tested can be divided into (i) derivatives having
the same cyclopentyl ring of (S)-1 and the pyrimidinedione
portion modified or (ii) compounds in which the pyrimidinedi-
one portion of (S)-1 is preserved and modifications are
introduced at the cycloalkyl-fused system. In general, all
modification on the pyrimidinedione portion gave a decrease
in the affinity toward all receptor subtypes; in particular, a drastic
reduction was observed when a longer amino acidic chain was
introduced ((S)-1 vs 3) or when a substituent is presented on
N3 position ((S)-1 vs 4). Low affinity was also observed for
(R)-1, being around 100 times less active than its enantiomer
counterpart. A totally different behavior was found for com-
pounds in which changes were introduced at the cycloalkyl-
fused system; in general these derivatives exhibit a broad range
of affinity/selectivity, some of them binding preferably
iGluR1-iGluR2 subunits, while others showing excellent af-
finity for the KA-R subunit iGluR5. The introduction of two
methyl groups at the cyclopentyl level of (S)-1 generated 6,
which was, among AMPA-R subunits, selective toward
iGluR1-iGluR2. By comparison of AMPA-R versus KA-R
affinity, 6 was found 22 times less active on iGluR5, indicating
that 6 was endowed of significant iGluR1/iGluR2 selectivity,
being 9 times more potent with respect to iGluR5. The sulfur-
bridged analogues 5 and 7 were tested against AMPA-R and
KA-R. At iGluR1-iGluR2 these analogues proved to be 2-fold
more potent than (S)-1, with an affinity of 58 and 80 nM for 5
The very low efficacies of 5 and 7 at iGluR5(Q)1b and their
rapid desensitization rates suggested that 5 and 7 could have
antagonistic properties at this receptor. Indeed, both compounds
were able to inhibit the responses to 3 mM Glu (Figure 3) with
extremely low IC50 values (5 ) 3.8 nM, 7 ) 1.9 nM)
comparable to their affinities at rat iGluR5(Q)1b measured by
radioligand binding (Table 1). Glu responses (1 mM) at iGluR1i
were also antagonized by higher concentrations of 7 (IC50
374 nM; 95% CI, 316-444 nM)), in keeping with its partial
agonist nature (Figure 3).
)

6616 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 20
Brief Articles
Table 1. Binding Affinities at Recombinant Rat AMPA-R and KA-R Expressed in Sf9 Cells
Ki (nM)a
compd
iGluR1o
iGluR2o(R)
iGluR3o
iGluR4o
40 ( 20
354 ( 178
3570 ( 700
2090 ( 495b
14200 ( 3500
123 ( 12
561 ( 105
656 ( 48
738 ( 115
>100000
>100000
>100000
288 ( 124
1590 ( 590
804 ( 38
15700 ( 7400
>100000
iGluR5(Q)1b
iGluR6 (V, C, R)
(R,S)-AMPA
L-Glu
kainate
(S)-1
(R,S)-2a
(S)-2b
(S)-2cb
(R,S)-2d
(S)-2e
(R)-1
(S)-3
(S,S)-4
(S)-5
(S)-6
(S)-7
21.9 ( 4.4
169 ( 27
477 ( 76
109 ( 5
410 ( 55
2.9 ( 1.6
37 ( 2
16.8 ( 2.9
282 ( 90
3690 ( 980
223 ( 24
1220 ( 305
4.0 ( 2.7
45 ( 6
20.6 ( 2.6
249 ( 12
1980 ( 340
1890 ( 580
19000 ( 480
70 ( 8
451 ( 163
434 ( 97
976 ( 287
>100000
>100000
>100000
311 ( 22
1510 ( 570
936 ( 114
6910 ( 663
>100000
1150 ( 114
140 ( 7
76 ( 27
44 ( 7
>100000
1380 ( 137
47 ( 19
14.8 ( 2.2
2.4 ( 0.2
NT
NA
332 ( 21
12.7 ( 3.5
NA
NA
>100000
>100000
>100000
>100000
NT
NA
>100000
NA
NA
NA
NA
NA
35 ( 5
83 ( 8
61 ( 3
105 ( 27
17800 ( 7350
>100000
10400 ( 4250
>100000
>100000
58 ( 18
102 ( 30
50 ( 7
NA
NA
>100000
80 ( 30
5.2 ( 0.5
976 ( 470
5.3 ( 1.3
490 ( 110
2980 ( 1050
126 ( 8
112 ( 26
3390 ( 496
43600 ( 8800
(S)-8
(S)-11
2390 ( 172
35200 ( 10200
a Mean values ( SD from at least three experiments conducted in triplicate. NT ) not tested. NA ) not active (Ki > 1 mM). b Pooled values from
iGluR4o and iGluR4co.
Table 2. Potencies (EC50) of Willardiine Analogues and Compounds 1, 6, and 8 at Recombinant AMPA-Rs Expressed in Xenopus læVis Oocytes
EC50 (µM)a
compd
iGluR1
iGluR2(Q)
iGluR3
iGluR4
(S)-willardiine (2a)b,e
11.5 ( 1.8
0.38 ( 0.07 (6)
NT
NT
NT
NT
(S)-2bc,f
0.46 ( 0.04 (10)
2.11 ( 0.20 (8)
NT
20.9 ( 1.9 (8)
NT
NT
NT
224 ( 20
33.9 ( 3.4 (8)i
120 ( 14 (9)i
11.9 ( 1.8 (5)
NT
(S)-2cf
(S)-2db,e
(S)-2eb,e
2.8 ( 0.4
NT
NT
33.6 ( 6.5
24.9 ( 3.6
11.8 ( 1.6 (8)
7.2 ( 0.8 (6)h
NT
(S)-(1)d,e
(S)-5f
13.9 ( 1.4
12.6 ( 0.6 (6)
11.8 ( 1.1 (5)h
34.3 ( 2.1g
24.2 ( 2.4 (10)i
90 ( 15 (14)i
(S)-7f
a Values are the mean ( SEM. Number of oocytes N in parentheses. N ) 3-5 for compound (S)-1. NT ) not tested. b From ref 10 (mean ( SD). c From
ref 11. d From ref 4. e EC50 at flop receptors. f EC50 at flip receptors. g EC50 at iGluR4co. h Statistically significantly different from EC50 at iGluR3 and
iGluR4 (P < 0.05, one-way ANOVA with Bonferroni t test). i Statistically significantly different (P < 0.05, one-way ANOVA with Bonferroni t test).
Figure 1. 5 and 7 activate human iGluR1i expressed in CHO-K1 cells. (A) Dose-response curves for the activation of iGluR1i by 5 (filled circles,
EC50 ) 18.5 µM) and 7 (filled squares, EC50 ) 4.1 µM) in the presence of 30 µM CTZ. Values were normalized to a maximal current induced by
10 mM L-Glu in the presence of 30 µM CTZ and represent the mean ( SEM of 3–4 separate experiments. (B) Representative current responses
induced by 5 (left) and 7 (right) in a iGluR1i expressing cell, which was voltage-clamped at -60 mV in the whole-cell configuration. The stimulations
were conducted in the presence of 30 µM CTZ. The cell was initially subjected to 1 s pulses of 10 mM L-Glu. After the washout of L-Glu,
increasing concentrations of 5 and 7 were applied to the cell in 1 s pulses.
N3-Substituted willardiine analogues have also been shown
to be antagonists at iGluR5, some having nanomolar affinity
and iGluR5 selectivity.9 However, no agonist activity was
reported for these compounds, suggesting that, unlike the
bicyclic compounds described herein, they are full antagonists.
Additionally, the binding mode at iGluR5 of one of the most
potent N3-substituted analogues was experimentally determined.
While the binding mode of our lead compound (S)-1 was
determined at iGluR2,8 it has not been experimentally evaluated
at iGluR5. X-ray determination of 5 in complex with the binding
core of iGluR5 subunit is in progress. Given the differences in
the structures of the two series of compounds, it would seem
unlikely that they bind to iGluR5 in the same fashion. It could
therefore be speculated that bicyclic N3-substituted willardiine
analogues could bind to iGluR5 with even higher affinity than
the currently existing antagonists.
Conclusion
In summary, we prepared novel ligands for AMPA-R and
KA-R structurally related to the bicyclic willardiine analogue
(S)-1. Analysis of the X-ray structure of (S)-1 in complex with
GluR2-S1S2J and with a (Y702F)GluR2-S1S2J mutant allowed
the development of derivative 6 endowed with improved
selectivity for iGluR1-iGluR2 subunits. Furthermore, through
a structural modification of the cyclopentane ring of (S)-1, we
developed two of the most potent and selective functional
antagonists (5 and 6) for KA-R subunit iGluR5. Derivatives 5
and 6, with their unique pharmacological profile, may lead to a

Brief Articles
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 20 6617
Figure 2. Representative current responses induced by 5 (left) or 7 (right) in iGluR5(Q)1b expressing cells, which were voltage-clamped at -60
mV in the whole-cell configuration. The cells were initially subjected to 100 ms pulses of 10 mM L-Glu. After the washout of Glu (NaRinger), a
100 ms pulse of 5 or 7 was applied to the cell.
Figure 3. Compounds 5 and 7 desensitize both human iGluR1i and iGluR5(Q)1b receptors expressed in CHO-K1 cells. (A) Dose-response curve
for the inactivation of iGluR5(Q)1b by 5 (open circles, IC50 ) 3.8 nM) and 7 (open squares, IC50 ) 1.9 nM) and of iGluR1i by 7 (filled squares,
IC50 ) 374 nM). The cells were stimulated with 100 ms control pulses of 3 mM L-Glu (iGluR5(Q)1b) or 1 mM L-Glu (iGluR1i) in 30 s intervals.
When a stable response level was obtained, the cells were pretreated with 5 and 7, before the next pulses with L-Glu. Values were normalized to
the control L-Glu currents in the absence of 5 or 7 and represent the mean ( SEM of 4-10 separate experiments. (B) Representative L-Glu (3 mM)
induced current responses at iGluR5(Q)1b receptors without (left) and in the presence of 3 nM 5 (right). (C) Representative L-Glu (3 mM) induced
current responses at iGluR5(Q)1b receptors without (left) and in the presence of 3 nM 7 (right). (D) Representative L-Glu (1 mM) induced current
responses at iGluR1i receptors without (left) and in the presence of 300 nM 7 (right).
(S,S)-1,3-Di(3′-amino-3′-carboxypropyl)-6,7-dihydro-1H-cyclo-
penta[d]pyrimidin-2,4(3H,5H)-dione (4). 4 was obtained from 15
(1.13 g, 1.41 mmol) following the procedure described for 3 but
increasing the pressure of H2 (90 psi) and the reaction time (4 days).
Compound 4 was obtained as an amorphous solid in 82.1% yield.
1H NMR (200 MHz, D2O/DCl 20 wt % in D2O 9:1) δ 3.13 (m,
2H), 2.87 (m, 4H), 1.80 (t, 2H, J ) 7.3 Hz), 1.44 (t, 2H, J ) 7.6
Hz), 1.15 (m, 4H), 0.88 (m, 2H); ESI-MS m/z 353 (M - H)-;
[R]2D0 -7.4 (c 0.3, HCl 1 N). Anal. (C15H22N4O6) C, H, N.
better understanding of the different roles and modes of action
of iGluR1-6 subunits, paving the way for the synthesis of new
potent, subunit selective iGluR5 modulators.
Experimental Procedures
The synthesis of 3-11 was performed according to Schemes
1-5 in Supporting Information. We describe herein the synthesis
of the most representative analogues of the series.
(S)-1-[2′-Amino-2′-carboxyethyl]-5,7-dihydrothieno[3,4-d]pyri-
midin-2,4(1H,3H)-dione (5). A mixture of 19a (0.70 g, 1.96 mmol)
in dichloromethane (5.0 mL) and trifluoroacetic acid (5.0 mL) was
stirred for 16 h at room temperature. Then the solvent was removed
and the crude product was purified by means of ion-exchange resin
as described for derivative 3. Compound 5 was obtained as a white
amorphous solid in 74.8% yield. 1H NMR (200 MHz, CD3OD/
DCl 20 wt % in D2O 9:1) δ 4.51 (m, 1H), 4.35 (m, 4H), 3.92 (t,
2H, J ) 3.1 Hz); 13C NMR (300 MHz, D2O/DCI 20 wt % in D2O
9:1) δ 169.4, 162.5, 155.0, 153.6, 113.0, 52.4, 45.9, 36.0, 32.1;
ESI-MS m/z 513 (100) (2M - H)-, 256 (M - H)-; [R]D20 -14.1
(c 0.2, HCl 1 N). Anal. (C9H11N3O4S) C, H, N.
(R)-1-(2-Amino-2-carboxyethyl)-6,7-dihydro-1H-cyclopenta[d]-
pyrimidin-2,4(1H,3H)dione (1). 1 was prepared starting from (R)-
3-[tert-(butoxycarbonyl)amino]oxetan-2-one following the proce-
dure previously reported for (S)-1.4 Analytical data were identical
to those reported for (S)-1.4
(S)-1-(3′-Amino-3′-carboxypropyl)-6,7-dihydro-1H-cyclopenta[d]-
pyrimidin-2,4(3H,5H)-dione (3). To a solution of 14 (110.0 mg,
0.23 mmol) in methanol (10.0 mL) a catalytic amount of 10% Pd/C
was added. The suspension was shaken for 8 h at room temperature
in a Parr hydrogenator under 40 psi of H2. Then the mixture was
filtered through Celite and washed with methanol. The filtrate was
concentrated, and the resulting crude product was purified by means
of ion exchange chromatography (Dowex 50 WX 8-400 purchased
from Aldrich), using first a mixture water/ethanol 1:1 v/v as eluent
and after pyridine 1 M in water to recover the pure compound from
the resin. Compound 3 was obtained as a colorless oil in 92.0%
yield. 1H NMR (200 MHz, D2O/DCl 20 wt % in D2O 9:1) δ
3.83-3.69 (m, 3H), 2.43 (m, 2H), 2.23 (m, 2H), 1.95 (m, 2H),
1.71 (m, 2H); 13C NMR (300 MHz D2O/DCI 20 wt % in D2O 9:1)
δ 170.2, 164.3, 157.6, 153.8, 111.6, 50.1, 36.6, 31.2, 30.4, 27.7,
26.7; ESI-MS m/z 252 (M – H)-; [R]2D0 -21.0 (c 0.3, HCl 6N).
Anal. (C11H15N3O4) C, H, N.
(S)-1-(2′-Amino-2′-carboxyethyl)-6,6-dimethyl-2,3,4,5,6,7-hexahy-
drocyclopentapyrimidin-2,4-dione (6). 6 was synthesized, starting
from 19b, following the synthetic strategy reported for 5. Compound
6 was obtained as a white amorphous solid in 64% yield. 1H NMR
(200 MHz, D2O/DCl 20 wt % in D2O 9:1) δ 4.33-4.24 (m, 1H),
4.16-4.11 (m, 2H), 2.62 (s, 2H), 2.28 (s, 2H), 0.98 (s, 6H); 13C
NMR (300 MHz, D2O/DCI 20 wt % in D2O 9:1) δ 168.2, 163.5,
158.2, 153.2, 111.7, 51.4, 45.4, 44.8, 40.6, 37.1, 27.8; ESI-MS m/z
266 (M - H)-; [R]D20 -46 (c 0.5, HCl 1 N). Anal. (C12H17N3O4)
C, H, N.

6618 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 20
Brief Articles
(S)-1-(2′-Amino-2′-carboxyethyl)thieno[3,2-d]pyrimidin-2,4-di-
one (7). 7 was synthesized, starting from 21, following the synthetic
strategy reported for 5. Compound 7 was obtained as an amorphous
Supporting Information Available: Chemistry, Schemes 1-5,
experimental procedures for intermediates, elemental analysis results
for final compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
1
solid in 10.0% yield. H NMR (200 MHz, CD3OD/DCl 20 wt %
in D2O 9:1) δ 8.09 (d, 1H, J ) 5.5 Hz), 7.30 (d, 1H, J ) 5.5 Hz),
4.61 (m, 2H), 4.45 (m, 1H); 13C NMR (300 MHz, D2O/DCI 20 wt
% in D2O 9:1) δ 169.6, 160.5, 153.3, 147.4, 138.1, 116.9, 113.5,
52.4, 45.4; ESI-MS m/z 509 (100) (2M - H)-, 254 (M - H)-,
167; [R]2D0 -9.7 (c 0.2, HCl 1 N). Anal. (C9H9N3O4S) C, H, N.
(S)-1-(2′-Amino-2′-carboxyethyl)quinazoline-2,4(1H,3H)-dione
(8). Derivative 24 (75.0 mg, 0.17 mmol) and a catalytic amount of
Pd/C 10% were added to solution of ammonium formate (20.0 mL,
0.4 N) in dry methanol. The mixture was heated to reflux and stirred
for 24 h. Then the mixture was filtered through Celite and the filtrate
evaporated under vacuum to give a white solid that was used in
the next step without further purification. Deprotection of the amino
function as described for 5 afforded 8 as an amorphous solid (98.7%
yield). 1H NMR (200 MHz, D2O/DCl 20 wt % in D2O 9:1) δ 7.70
(m, 1H), 7.47 (m, 1H), 7.04 (m, 2H), 4.34 (m, 2H), 4.26 (m, 1H);
13C NMR (300 MHz, D2O/DCI 20 wt % in D2O 9:1) δ 168.9,
164.0, 152.0, 139.8, 136.7, 130.5, 128.2, 124.3, 115.430, 51.3, 42.2;
ESI-MS m/z 250 (M + H)+; [R]2D0 +14.8 (c 0.3, HCl 1 N). Anal.
(C11H11N3O4) C, H, N.
References
(1) Bra¨uner-Osborne, H.; Egebjerg, J.; Nielsen, E.; Madsen, U.; Krogs-
gaard-Larsen, P. Ligands for glutamate receptors: design and thera-
peutic prospects. J. Med. Chem. 2000, 43, 2609–2645.
(2) (a) Rosenmund, C.; Stern-Bach, Y.; Stevens, C. F. The tetrameric structure
of a glutamate receptor channel. Science 1998, 280, 1596–1599. (b)
Sun, Y.; Olson, R.; Horning, M.; Armstrong, N.; Mayer, M.; Gouaux, E.
Mechanism of glutamate receptor desensitization. Nature 2002, 417, 245–
253. (c) Laube, B.; Kuhse, J.; Betz, H. J. Evidence for a tetrameric
structure of recombinant NMDA receptors. Neuroscience 1998, 18,
2954–2961.
(3) Frandsen, A.; Schousboe, A. Excitatory amino acid-mediated cyto-
toxicity and calcium homeostasis in cultured neurons. J. Neurochem.
1993, 60, 1202–1211.
(4) Campiani, G.; Morelli, E.; Nacci, V.; Fattorusso, C.; Ramunno, A.;
Novellino, E.; Greenwood, J.; Liljefors, T.; Griffiths, R.; Sinclair, C.;
Reavy, H.; Kristensen, A. S.; Pickering, D. S.; Schousboe, A.; Cagnotto,
A.; Fumagalli, E.; Mennini, T. Characterization of the 1H-cyclopentapy-
rimidine-2,4(1H,3H)-dione derivative (S)-CPW399 as a novel, potent, and
subtype-selective AMPA receptor full agonist with partial desensitization
properties. J. Med. Chem. 2001, 44, 4501–4504.
(5) Dolman, N. P.; Troop, H. M.; More, J. C. A.; Alt, A.; Knauss, J. L.;
Nistico, R.; Jack, S.; Morley, R. M.; Bortolotto, Z. A.; 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.
(6) 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.
(7) Mayer, M. L. Crystal structures of the GluR5 and GluR6 ligand binding
cores: molecular mechanisms underlying kainate receptor selectivity.
Neuron 2005, 45, 539–552.
(8) Frandsen, A.; Pickering, D. S.; Vestergaard, B.; Kasper, C.; Bryde
Nielsen, B.; Greenwood, J. R.; Campiani, G.; Fattorusso, C.; Gajhede,
M.; Schousboe, A.; Kastrup, J. S. Tyr702 is an important determinant
of agonist binding and domain closure of the ligand-binding core of
GluR2. Mol. Pharmacol. 2005, 67, 1–11.
(S)-1-(2′-Amino-2′-carboxyethyl)-3-benzylquinazoline-2,4(1H,3H)-
dione (9). 9 was obtained, starting from 24, following the synthetic
strategy used for 5. Compound 9 was obtained as a white amorphous
1
solid in 95.6% yield. H NMR (200 MHz, DMSO-d6) δ 8.03 (d,
1H, J ) 7.8 Hz), 7.76 (m, 1H), 7.53 (d, 1H, J ) 8.6 Hz), 7.33-7.17
(m, 6H), 5.06 (s, 2H), 4.55 (m, 2H), 4.14 (t, 1H, J ) 7.1 Hz); 13
C
NMR (300 MHz, DMSO-d6) δ 174.6, 168.9, 151.7, 142.9, 139.9,
137.3, 136.5, 129.0 (2C), 128.1, 115.8, 114.9, 98.6, 50.0, 45.1 43.1;
ESI-MS m/z 340 (M + H)+; [R]2D0 +55.7 (c 0.3, HCl 1 N). Anal.
(C18H17N3O4) C, H, N.
(S)-1-(2′-Amino-2′-carboxyethyl)-3-benzylfuro[3,4-d]pyrimidin-
2,4-dione (10). The title compound was obtained, starting from 28a,
following the synthetic strategy used for 5. Compound 10 was
obtained as a white amorphous solid in 87.5% yield. 1H NMR (200
MHz, D2O/DCl 20 wt % in D2O 9:1) δ 7.61 (s, 1H), 7.10 (s, 1H),
6.65 (m, 5H), 4.41 (s, 2H), 3.90 (m, 1H), 3.72 (m, 2H); ESI-MS
m/z 328 (M - H)-, 284, 241; [R]2D0 +31.0 (c 0.2, HCl 1 N). Anal.
(C16H15N3O5) C, H, N.
(9) Dolman, N. P.; More, J. C. A.; Alt, A.; Knauss, J. L.; Pentika¨inen, O. T.;
Glasser, C. R.; Bleakman, D.; Mayer, M. L.; Collingridge, G. L.; Jane,
D. E. 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. J. Med. Chem. 2007, 50, 1558–1570.
(10) Kizelsztein, P.; Eisenstein, M.; Strutz, N.; Hollmann, M.; Teichberg,
V. I. Mutant cycle analysis of the active and desensitized states of an
AMPA receptor induced by willardiines. Biochemistry 2000, 39,
12819–12827.
(S)-1-(2′-Amino-2′-carboxyethyl)furo[3,4-d]pyrimidin-2,4-di-
one (11). 11 was obtained, starting from 28b, following the same
synthetic strategy used for 5. Compound 11 was obtained as a white
1
amorphous solid in 95% yield. H NMR (200 MHz, D2O/DCl 20
wt % in D2O 9:1) δ 8.07 (s, 1H), 7.55 (s, 1H), 4.37-4.31 (m, 1H),
4.25-4.19 (m, 2H); ESI-MS m/z 240 (M + H)+, 223, 194; [R]D20
+38.0 (c 0.5, HCl 1 N). Anal. (C9H9N3O5) C, H, N.
(11) Greenwood, J. R.; Mewett, K. N.; Allan, R. D.; Mart´ın, B. O.;
Pickering, D. S. 3-Hydroxypyridazine 1-oxides as carboxylate bioi-
sosteres: a new series of subtype-selective AMPA receptor agonists.
Neuropharmacology 2006, 51, 52–59.
Acknowledgment. This work was partially supported by the
MIUR-PRIN, the Novo Nordisk Foundation, and the Danish
Medical Research Council. The authors thank NeuroSearch A/S
Ballerup, Denmark.
JM800865A