Synthesis, pharmacological characterization, and molecular modeling of heterobicyclic amino acids related to (+)-2-aminobicyclo[3.1.0]hexane-2,6- dicarboxylic acid (LY354740): Identification of two new potent, selective, and systemically active agonists for group II metabotropic glutamate receptors

Article abstract of DOI:10.1021/jm980616n

As part of our ongoing research program aimed at the identification of highly potent, selective, and systemically active agonists for group II metabotropic glutamate (mGlu) receptors, we have prepared novel heterobicyclic amino acids (-)-2-oxa-4-aminobicyclo[3.1.0]hexane-4,6- dicarboxylate (LY379268, (-)-9) and (-)-2-thia-4-aminobicyclo[3.1.0]hexane- 4,6-dicarboxylate (LY389795, (-)-10). Compounds (-)-9 and (-)-10 are structurally related to our previously described nanomolar potency group II mGlu receptor agonist, (+)-2-aminobicyclo[3.1.0]hexane-2,6-dicarboxylate monohydrate (LY354740 monohydrate, 5), with the C4-methylene unit of 5 being replaced with either an oxygen atom (as in (-)-9) or a sulfur atom (as in (- )-10). Compounds (-)-9 and (-)-10 potently and stereospecifically displaced specific binding of the mGlu2/3 receptor antagonist ([3H]LY341495) in rat cerebral cortical homogenates, displaying IC₅₀ values of 15 ± 4 and 8.4 ± 0.8 nM, respectively, while having no effect up to 100 000 nM on radioligand binding to the glutamate recognition site on NMDA, AMPA, or kainate receptors. Compounds (-)-9 and (-)-10 also potently displaced [³H]LY341495 binding from membranes expressing recombinant human group II mGlu receptor subtypes: (-)-9, K(i) = 14.1 ± 1.4 nM at mGlu2 and 5.8 ± 0.64 nM at mGlu3; (-)-10, K(i) = 40.6 ± 3.7 nM at mGlu2 and 4.7 ± 1.2 nM at mGlu3. Evaluation of the functional effects of (-)-9 and (-)-10 on second-messenger responses in nonneuronal cells expressing human mGlu receptor subtypes demonstrated each to be a highly potent agonist for group II mGlu receptors: (-)-9, EC₅₀ = 2.69 ± 0.26 nM at mGlu2 and 4.58 ± 0.04 nM at mGlu3; (-)-10, EC₅₀ = 3.91 ± 0.81 nM at mGlu2 and 7.63 ± 2.08 nM at mGlu3. In contrast, neither compound (up to 10 000 nM) displayed either agonist or antagonist activity in cells expressing recombinant human mGlu1a, mGlu5a, mGlu4a, or mGlu7a receptors. The agonist effects of (-)-9 and (-)-10 at group II mGlu receptors were not totally specific, however, as mGlu6 agonist activity was observed at high nanomolar concentrations for (-)-9 (EC₅₀ = 401 ± 46 nM) and at micromolar concentrations (EC₅₀ = 2 430 ± 600 nM) for (-)-10; furthermore, each activated mGlu8 receptors at micromolar concentrations (EC₅₀ = 1 690 ± 130 and 7 340 ± 2 720 nM, respectively). Intraperitoneal administration of either (-)-9 or (-)-10 in the mouse resulted in a dose-related blockade of limbic seizure activity produced by the nonselective group I/group II mGluR agonist (1S,3R)-ACPD ((-)-9 ED₅₀ = 19 mg/kg, (-)-10 ED₅₀ = 14 mg/kg), indicating that these molecules effectively cross the blood-brain barrier following systemic administration and suppress group I mGluR-mediated limbic excitation. Thus, heterobicyclic amino acids (-)-9 and (-)-10 are novel pharmacological tools useful for exploring the functions of mGlu receptors in vitro and in vivo.

Full text of DOI:10.1021/jm980616n

J . Med. Chem. 1999, 42, 1027-1040
1027
Syn th esis, P h a r m a cologica l Ch a r a cter iza tion , a n d Molecu la r Mod elin g of
Heter obicyclic Am in o Acid s Rela ted to (+)-2-Am in obicyclo[3.1.0]h exa n e-
2,6-d ica r boxylic Acid (LY354740): Id en tifica tion of Tw o New P oten t, Selective,
a n d System ica lly Active Agon ists for Gr ou p II Meta botr op ic Glu ta m a te
Recep tor s
J ames A. Monn,* Matthew J . Valli, Steven M. Massey, Marvin M. Hansen,‡ Thomas J . Kress,‡
J ames P. Wepsiec,‡ Allen R. Harkness,‡ J ohn L. Grutsch, J r.,‡ Rebecca A. Wright, Bryan G. J ohnson,
,†
,†
Sherri L. Andis, Ann Kingston, Rosemarie Tomlinson, Richard Lewis,† Kelly R. Griffey,§
J oseph P. Tizzano,§ and Darryle D. Schoepp
Discovery Chemistry, Process Research and Development, Neuroscience, and Toxicology Research Divisions, Eli Lilly and
Company, Indianapolis, Indiana 46285, and Lilly Research Centre, Ltd., Windelsham, Surrey GU20 6PH, U.K.
Received November 2, 1998
As part of our ongoing research program aimed at the identification of highly potent, selective,
and systemically active agonists for group II metabotropic glutamate (mGlu) receptors, we
have prepared novel heterobicyclic amino acids (-)-2-oxa-4-aminobicyclo[3.1.0]hexane-4,6-
dicarboxylate (LY379268, (-)-9) and (-)-2-thia-4-aminobicyclo[3.1.0]hexane-4,6-dicarboxylate
(LY389795, (-)-10). Compounds (-)-9 and (-)-10 are structurally related to our previously
described nanomolar potency group II mGlu receptor agonist, (+)-2-aminobicyclo[3.1.0]hexane-
2,6-dicarboxylate monohydrate (LY354740 monohydrate, 5), with the C4-methylene unit of 5
being replaced with either an oxygen atom (as in (-)-9) or a sulfur atom (as in (-)-10).
Compounds (-)-9 and (-)-10 potently and stereospecifically displaced specific binding of the
mGlu2/3 receptor antagonist ([3H]LY341495) in rat cerebral cortical homogenates, displaying
IC50 values of 15 ( 4 and 8.4 ( 0.8 nM, respectively, while having no effect up to 100 000 nM
on radioligand binding to the glutamate recognition site on NMDA, AMPA, or kainate receptors.
Compounds (-)-9 and (-)-10 also potently displaced [3H]LY341495 binding from membranes
expressing recombinant human group II mGlu receptor subtypes: (-)-9, Ki ) 14.1 ( 1.4 nM
at mGlu2 and 5.8 ( 0.64 nM at mGlu3; (-)-10, Ki ) 40.6 ( 3.7 nM at mGlu2 and 4.7 ( 1.2 nM
at mGlu3. Evaluation of the functional effects of (-)-9 and (-)-10 on second-messenger
responses in nonneuronal cells expressing human mGlu receptor subtypes demonstrated each
to be a highly potent agonist for group II mGlu receptors: (-)-9, EC50 ) 2.69 ( 0.26 nM at
mGlu2 and 4.58 ( 0.04 nM at mGlu3; (-)-10, EC50 ) 3.91 ( 0.81 nM at mGlu2 and 7.63 (
2.08 nM at mGlu3. In contrast, neither compound (up to 10 000 nM) displayed either agonist
or antagonist activity in cells expressing recombinant human mGlu1a, mGlu5a, mGlu4a, or
mGlu7a receptors. The agonist effects of (-)-9 and (-)-10 at group II mGlu receptors were not
totally specific, however, as mGlu6 agonist activity was observed at high nanomolar concentra-
tions for (-)-9 (EC50 ) 401 ( 46 nM) and at micromolar concentrations (EC50 ) 2 430 ( 600
nM) for (-)-10; furthermore, each activated mGlu8 receptors at micromolar concentrations
(EC50 ) 1 690 ( 130 and 7 340 ( 2 720 nM, respectively). Intraperitoneal administration of
either (-)-9 or (-)-10 in the mouse resulted in a dose-related blockade of limbic seizure activity
produced by the nonselective group I/group II mGluR agonist (1S,3R)-ACPD ((-)-9 ED50 ) 19
mg/kg, (-)-10 ED50 ) 14 mg/kg), indicating that these molecules effectively cross the blood-
brain barrier following systemic administration and suppress group I mGluR-mediated limbic
excitation. Thus, heterobicyclic amino acids (-)-9 and (-)-10 are novel pharmacological tools
useful for exploring the functions of mGlu receptors in vitro and in vivo.
In tr od u ction
system (CNS), participating in an expansive list of
physiological and pathophysiological processes.1,2 The
neuronal effects of (S)-Glu are mediated by two hetero-
geneous families of cell membrane-associated receptors,
the ion-channel-linked or ionotropic glutamate (iGlu)
receptors (iGluRs) and the G-protein-coupled or me-
tabotropic glutamate (mGlu) receptors (mGluRs).3,4
The mGlu receptors are highly heterogeneous with
respect to their structure, function, and localization
within the CNS. There are currently eight known mGlu
(S)-Glutamic acid ((S)-Glu) is the primary excitatory
neurotransmitter in the mammalian central nervous
* Corresponding author: Dr. J ames A. Monn, Eli Lilly and Co., Lilly
Corporate Center, Mail Drop Code 0510, Indianapolis, IN 46285. Tel:
(317) 276-9101. Fax: (317) 277-7600. E-mail: Monn@Lilly.com.
‡ Process Research and Development Division.
Neuroscience Division.
§ Toxicology Research Division.
† Lilly Research Centre, Ltd.
10.1021/jm980616n CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/02/1999

1028 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Monn et al.
receptor subtypes (mGlu1-8) that have been classified
into three subfamilies, or groups, based on their amino
acid sequence homology, signal transduction mecha-
nisms, and responses to prototypic agonists.3-8 Group
I mGluRs (mGlu1, mGlu5, and splice variants) are
positively coupled through Gs to phospholipase C (PLC).
When activated, these receptors induce phosphoinositide
hydrolysis and subsequent calcium mobilization within
the target cell. Group I mGluRs are nonselectively
activated by micromolar concentrations of (1S,3R)-
ACPD (1)8 and selectively activated by 3,5-dihydroxy-
phenylglycine (2).8-11 More recently, rac-2-chloro-5-
hydroxyphenylglycine (3) has been reported as an
mGluR5 subtype-selective agonist.12 Group II mGluRs
(mGlu2 and mGlu3) are negatively coupled through Gi
to adenylyl cyclase (AC). Agonist activation of group II
mGluRs results in the inhibition of 3′,5′-cyclic adenosine
monophosphate (c-AMP) formation. Compound 1 acti-
vates group II mGluRs but only at concentrations that
also produce agonist effects at the group I receptors.8,13
Selective agonists for the group II mGlu receptors
include two molecules previously described by our
laboratory, (2R,4R)-4-aminopyrrolidine-2,4-dicarbox-
ylate ((2R,4R)-APDC, 4), a compound that activates
mGlu2 and mGlu3 receptors at high nanomolar to low
micromolar concentrations,14,15 and (1S,2S,5R,6S)-2-
aminobicyclo[3.1.0]hexane-2,6-dicarboxylate (LY354740,
5), a compound possessing low nanomolar potency at
these target proteins.16,17 Group III mGluRs (mGlu4,
mGlu6, mGlu7, and mGlu8) are also negatively coupled
to AC and are activated by (S)-2-amino-4-phosphono-
butyrate ((S)-AP4, 6) and (S)-serine O-phosphate ((S)-
SOP, 7).8,18-22 More recently, mGluR6 subtype-selective
agonist activity has been reported for (S)-homo-AMPA
(8).23
Ch em istr y
The preparation of racemic and nonracemic 9 is
depicted in Scheme 1. Carboxycyclopropanation of furan
was accomplished by way of rhodium-catalyzed conden-
sation with ethyl diazoacetate and afforded bicyclic
adduct 11 in consistent, though low (20-40%), overall
yields.24 The relative stereochemistry about the cyclo-
1
propane ring of 11 was confirmed by H NMR analysis
which revealed a 5.7-Hz couple between the ring fusion
protons H1 and H5 and a 2.5-Hz couple between each
of the ring fusion protons and H6. Regiospecific incor-
poration of a hydroxyl functionality at the C4-position
was achieved in 87% yield by hydroboration of 11
followed by in situ conversion of the intermediate
organoborane to the corresponding carbinol 12 by treat-
ment with buffered hydrogen peroxide or sodium per-
borate.25 Further oxidation of 12 under Swern condi-
tions afforded heterobicyclic ketone 13 in 84% yield.
Conversion of the ketone functionality of 13 to dia-
stereomerically pure 5′-spirohydantoin 14 was effected
by treatment of this ketone with ammonium carbonate
and potassium cyanide in methanol. Compound 14 was
obtained in 79% yield by direct crystallization from this
reaction mixture.26 Hydrolysis of 14 then afforded (()-9
in 62% yield. Selective saponification of the C6-ester
functionality of 14 afforded rac-carboxylic acid 15 which
was conveniently resolved by selective crystallization of
either the (R)- or (S)-2-phenylglycinol salts (-)-16 or
(+)-16, respectively. The so resolved hydantoin carbox-
ylates (-)-15 and (+)-15 were isolated from aqueous
HCl, and the enantiomeric excess of each isomer was
then established by chiral chromatography to be >98%.27
Exhaustive alkaline hydrolysis of each of the hydantoin
enantiomers (-)-15 or (+)-15 then afforded the non-
racemic amino acid products (-)-9 and (+)-9, respec-
tively. Single-crystal X-ray analysis of (-)-9 (Figure 1)
firmly established the relative stereochemistry of this
molecule.28
In a strictly analogous manner, the preparation of
racemic and nonracemic 10 was achieved in similar
overall chemical yield and enantiomeric purity by
employing thiophene as the heteroaromatic partner in
the initial carboxycyclopropanation reaction (Scheme
2).24-27
P h a r m a cologica l Meth od s
R a t Br a in Glu t a m a t e R ecep t or Affin it y. Glu-
tamate receptor affinity was determined in rat forebrain
homogenates by examining the ability of test ligands
to displace [3H]CGP39653, [3H]AMPA, [3H]kainic acid,
and [3H]LY341495 from NMDA, AMPA, kainate, and
group II mGlu receptors, respectively, as previously de-
scribed.29-32 In those cases where an IC50 value of less
than 100 µM could be determined, experiments were
performed in triplicate. Data are expressed as the mean
IC50 ( standard error.
Affin ity F or Recom bin a n t Hu m a n m Glu Recep -
tor Su btyp es. Test compounds were evaluated for their
ability to displace [3H]LY341495 from recombinant
human mGlu2 and mGlu3 receptor subtypes individu-
ally expressed in RGT cells.33,34 The Ki values ((
standard error) were calculated from the IC50 values
As part of our overall research program directed
toward the preparation of highly potent and selective
ligands for mGlu receptor subtypes, we have continued
to focus our efforts on the identification of group II
mGluR agonists. In this account, we report the prepara-
tion of two novel heterobicyclic amino acids related to
5, (-)-2-oxa-4-aminobicyclo[3.1.0]hexane-4,6-dicarbox-
ylate ((-)-9, LY379268) and (-)-2-thia-4-aminobicyclo-
[3.1.0]hexane-4,6-dicarboxylate ((-)-10, LY389795), and
their pharmacological characterization as potent group
II mGlu receptor agonists.

Heterobicyclic mGluR Agonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 1029
Sch em e 1a
a
Reagents and conditions: (a) ethyl diazoacetate, Rh2(OAc)4, 10 °C, 2 h; (b) thexylborane, THF, 0 °C, 2 h; (c) 30% H2O2, phosphate
buffer (pH 7), room temperature, overnight; (d) oxalyl chloride, DMSO, Et3N, CH2Cl2, -65 °C; (e) (NH4)2CO3, KCN, MeOH; (f) crystallization
at pH 7; (g) 2 N NaOH, reflux, overnight; (h) anion-exchange chromatography; (i) 2 N NaOH, room temperature, 0.5 h; (j) (R)-(-)-2-
phenylglycinol, EtOH/H2O (5:1); (k) 1 N HCl, EtOAc extraction; (l) (S)-2-phenylglycinol, EtOH.
performed as previously described.37 Briefly, male NIH
swiss mice were physically restrained to allow uni-
lateral intracerebral (ic) injections of 1, which were
made using a 10-µL Hamilton microsyringe. Animals
were pretreated at the time indicated with either sterile
water (5 µL, ic) or test compound at the dose indi-
cated. Animals were observed for 30 min following
administration of 1. Limbic seizures in treated mice
were characterized by the presence of at least one
episode of clonic forelimb contractions followed by hind
limb rearing to a praying stance, then loss of balance,
F igu r e 1. Thermal elipsoid representation of (-)-9.28 The
non-H atom elipsoids enclose the 50% probability limits.
and falling. The onset of these behaviors was within 1
min following injection; behaviors lasted approximately
20 min.
employing the Cheng-Prusoff equation35 and represent
Molecu la r Mod elin g. All initial models were derived
from the X-ray structure of 5.17 The acid functions were
modeled in their deprotonated states and the amines
as either the neutral or protonated form, to represent
the major forms present in aqueous solution at physi-
ological pH 7.4. The suffix -NH2 indicates the neutral
amine form and the suffix -NH3 the protonated amine
form of a compound. Structures were optimized by a
direct Hartree-Fock approach, as implemented in
Spartan v5.0. The basis set used was 6-31G**. Charges
were derived by fitting to the electrostatic potential.
Molecular similarity calculations were performed using
the carbo index, a 3-component Gaussian function,
exhaustive rigid search at 15° intervals followed by a
simplex optimization, as implemented in ASP v3.21.
the mean of at least three separate experiments.
Secon d -Messen ger Resp on ses in Cells Exp r ess-
in g Recom bin a n t m Glu Recep tor s. Test compounds
were evaluated for their ability to influence the produc-
tion of second messengers in “RGT” cells. RGT cells are
nonneuronal AV12-664 cells coexpressing both GLAST
(a recombinant glutamate transporter to minimize
constituent glutamate activity) and a recombinant hu-
man group I (mGlu1a or mGlu5a), group II (mGlu2 or
mGlu3), or group III (mGlu4a, mGlu6, mGlu7a, or
mGlu8) receptor employing methods previously de-
scribed.16,36
Mou se Lim bic Seizu r e Assa y. Evaluation of the
effects of test compounds on limbic seizure activity
produced by intracerebral administration of 1 was

1030 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Monn et al.
Sch em e 2a
a
Reagents and conditions: (a) ethyl diazoacetate, Rh2(OAc)4, 70 °C, 3 h; (b) thexylborane, THF, room temperature, 2 h; (c) NaBO3‚4H2O,
phosphate buffer (pH 7), room temperature, 1 h; (d) oxalyl chloride, DMSO, Et3N, CH2Cl2, -40 °C; (e) (NH4)2CO3, KCN, EtOH/H2O
(2.5:1), 35 °C; (f) crystallization from 2-propanol; (g) 2 N NaOH, reflux; (h) anion-exchange chromatography; (i) 2 N NaOH, room temperature,
0.5 h; (j) (R)-(-)-2-phenylglycinol, EtOH/H2O (5:1); (k) aqueous HCl, 0 °C; (l) (S)-2-phenylglycinol, EtOH/H2O (5:1); (m) pH 3, 0 °C.
Resu lts
Glu t a m a t e R ecep t or R a d ioliga n d Bin d in g in
1). As was observed in the rat brain homogenate binding
experiments, a high degree of stereoselectivity was also
evident for binding of the individual enantiomers to the
glutamate site on cloned group II mGlu receptors, with
the (-)-isomers displaying high-affinity binding to
mGlu2 ((-)-9, Ki ) 14.1 ( 1.4 nM; (-)-10, Ki ) 40.6 (
3.7 nM) and mGlu3 ((-)-9, Ki ) 5.8 ( 0.6 nM; (-)-10,
Ki ) 4.7 ( 1.2 nM) receptors relative to the (+)-isomers,
which displaced [3H]LY341495 binding only at high
ligand concentrations ((+)-9, Ki ) 1 995 ( 34.3 nM at
mGlu2 and 1 037 ( 89 nM at mGlu3; (+)-10, Ki ) 818.8
( 81.3 nM at mGlu2 and 169.7 ( 20.2 nM at mGlu3).
In comparison, compound 5 displayed significantly lower
affinity for both mGlu2 (Ki ) 74.9 ( 9.1 nM) and mGlu3
(Ki ) 93.3 ( 2.6) receptors than that observed for either
(-)-9 or (-)-10.
Secon d-Messen ger Respon ses. Heterobicyclic amino
acids (-)-9 and (-)-10 potently inhibited forskolin-
stimulated c-AMP production in RGT cells expressing
recombinant human mGluR2 ((-)-9, EC50 ) 2.69 ( 0.26
nM; (-)-10, EC50 ) 3.91 ( 0.81 nM) or mGluR3 ((-)-9,
EC50 ) 4.58 ( 0.04 nM; (-)-10, EC50 ) 7.63 ( 2.08 nM).
The data are shown in Figure 3A,B and summarized in
Table 2. Agonist activity was also evinced by these
molecules in the mGlu6 receptor-expressing cells ((-)-
9, EC50 ) 401 ( 46 nM; (-)-10, EC50 ) 2 430 ( 600
nM) and in the mGlu8 receptor-expressing cells ((-)-9,
Vitr o. Racemic heterobicyclic amino acids (()-9 and
(()-10 each potently displaced [3H]LY341495 binding
from rat forebrain homogenates (IC50 ) 29 ( 4 nM for
(()-9 and 26.1 ( 1.5 nM for (()-10) (Figure 2, Table
1). This effect was determined to be highly stereoselec-
tive, with the mGlu receptor-active enantiomers (-)-9
(IC50 ) 15 ( 4 nM) and (-)-10 (IC50 ) 8.4 ( 0.8 nM)
displaying approximately 300- and 90-fold higher af-
finity for [3H]LY341495-labeled sites than the corre-
sponding inactive enantiomers (+)-9 and (+)-10. Fur-
thermore, the potencies of (-)-9 and (-)-10 in this assay
compare favorably to that observed for the related
carbobicyclic amino acid 5 (IC50 ) 254 ( 86 nM). In
contrast, at concentrations up to 100 000 nM, neither
racemic nor nonracemic 9 or 10 displaced radioligand
binding to NMDA, AMPA, or kainate receptors ex-
pressed in rat forebrain homogenates.
As expected based on the foregoing results and
the known mGlu2/3 receptor selectivity for [3H]-
LY341495,32-34 heterobicyclic amino acids 9 and 10
potently displaced this radioligand from recombinant
human mGlu2 ((()-9, Ki ) 36.5 ( 1.3 nM; (()-10, Ki )
67.2 ( 4.8 nM) and mGlu3 ((()-9, Ki ) 14.3 ( 0.4 nM;
(()-10, Ki ) 9.4 ( 0.3 nM) receptors expressed in
membranes prepared from RGT cells (Figure 2, Table

Heterobicyclic mGluR Agonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 1031
F igu r e 2. Displacement of [3H]LY341495 binding to rat forebrain homogenates,32 human mGluR2,33,34 and human mGluR333,34
by racemic and nonracemic 9 (top three panels) and 10 (bottom three panels). Legend: (()-9 (4), (-)-9 (9), (+)-9 (0), (()-10 (2),
(-)-10 (b), (+)-10 (O).
Ta ble 1. Effects of 5, 9, and 10 on Radioligand Binding to Glutamate Receptors in Rat Brain and to Recombinant Human mGlu2
and mGlu3 Receptors Expressed in RGT Cells
IC50 (nM) ( SEM
[3H]CGP39653b
Ki (nM) ( SEM
[3H]LY341495a
rat brain
[3H]AMPAc
rat brain
[3H]KAd
rat brain
[3H]LY341495
[3H]LY341495
compd
rat brain
mGlu2e
mGlu3e
5
254 ( 86
29 ( 4
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
>100 000
74.9 ( 9.1
36.5 ( 1.3
14.1 ( 1.4
2 000 ( 34
67.2 ( 4.8
40.6 ( 3.7
819 ( 81
93.3 ( 2.6
14.3 ( 0.4
5.8 ( 0.6
1 040 ( 89
9.4 ( 0.3
4.7 ( 1.2
170 ( 20
(()-9
(-)-9
(+)-9
(()-10
(-)-10
(+)-10
15 ( 4
5 040 ( 260
26.1 ( 1.5
8.4 ( 0.8
792 ( 27.6
a
b
d
See ref 32. See ref 29. c See ref 30. See ref 31. e See refs 33 and 34.
EC50 ) 1 690 ( 130 nM; (-)-10, EC50 ) 7 340 ( 2 720
nM). In contrast, at concentrations up to 10 000 nM (in
most instances up to 100 000 nM) neither (-)-9 nor (-)-
10 produced either agonist or antagonist effects in cells
expressing recombinant group I (mGlu1a, mGlu5a) or
other recombinant group III (mGlu4a, mGlu7a) recep-
tors (Table 2).
Mou se Lim bic Seizu r e Mod el. Intracerebral ad-
ministration of 1 (400 nmol) produced a characteristic
limbic seizure behavior in NIH Swiss mice which
persisted throughout the ensuing 20-min observation
period.37 Intraperitoneal administration of (-)-9 at
either 1, 2, or 4 h prior to 1 resulted in a dose-related
decrease in the number of mice exhibiting seizures
(Table 3). The calculated ED50 values for 1-, 2-, and 4-h
preadministration times of (-)-9 were 19, 6, and 26 mg/
kg, respectively. In a similar fashion, (-)-10 blocked
seizure behaviors with ED50 values of 14, 15, and 14
mg/kg when administered by the ip route at 1, 2, or 4 h
prior to 1.
5-NH2 (nonprotonated amine form) was superimposed
onto its X-ray structure,17 and the rms deviation over
all atoms was determined to be 0.4 Å. In fact, all the
deviations occurred in the R-amino acid group, with the
carboxylate groups being oriented at an angle of 103°
with respect to each other and the amine being more
pyramidal in the optimized structure. In all other
respects, the structures were identical so that the effect
of omitting solvent on the calculated molecular geometry
of 5-NH2 is small. Similarity calculations were used to
establish whether the heteroatom for methylene sub-
stitutions has a measurable effect on the overall shape
and/or electrostatic profile of the compounds. Thus,
similarity calculations relating the protonated (-NH3)
and nonprotonated (-NH2) forms of 5, (-)-9, and (-)-
10 (Table 4) were performed using a 1:1 weighting of
the shape and electrostatic components of similarity.
These data indicate a high degree of similarity between
protonated and nonprotonated forms of each of the
individual molecules (r2 g 0.93 in each case), as well as
between different analogues of similar (r2 g 0.90) or
dissimilar (r2 g 0.87) protonation states. An overlay of
the nonprotonated forms of 5, (-)-9, and (-)-10 is
shown in Figure 4 (left).
Molecu la r Mod elin g. The optimization calculations
were carried out in the gas phase, with no solvent or
dielectric corrections. This can lead to structures which
are considerably different from those in solution, espe-
cially for charged species. The optimized structure of

1032 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Monn et al.
F igu r e 3. Forskolin-stimulated c-AMP responses to (-)-9 (9), (-)-10 (b), and their inactive enantiomers (+)-9 (0) and (+)-10
(O) in RGT cells expressing recombinant human mGlu2 (left) or mGlu3 (right) receptors. Methods are described in detail in ref
16.
Ta ble 2. Effects of 5, (-)-9, and (-)-10 on Second-Messenger Responses in RGT Cells Expressing Recombinant Human mGlu
Receptor Subtypes
EC50 (nM) ( SEM
compd
mGlu1aa
mGlu2a,b
mGlu3a,c
mGlu4aa,d
>100 000
21 100 ( 8 600g
>100 000
mGlu5aa
mGlu6e
mGlu7aa,d
mGlu8d
5
>100 000
>100 000
>100 000
11.1 ( 2.5
2.69 ( 0.26
3.91 ( 0.81
38.0 ( 3.07
4.58 ( 0.04
7.63 ( 2.08
>100 000
>100 000
>100 000
2 990 ( 120
401 ( 46
2 430 ( 600
>100 000
>100 000
>100 000
11 500 ( 200f
1 690 ( 130h
7 340 ( 2 720i
(-)-9
(-)-10
a
b
c
See ref 16. Bmax ) 29.2 pmol/mg of protein; Emax for all compounds tested g 90%. Bmax ) 35 ( 1.2 pmol/mg of protein; Emax for all
d
e
f
g
compounds tested g 90%. See ref 36. Bmax ) 1.58 pmol/mg of protein; Emax for all compounds tested > 95%. Emax ) 50%. Emax
46%. Emax ) 91%. Emax ) 80%.
)
h
i
Ta ble 3. Effect of Systemically Administered (-)-9 or (-)-10
on Limbic Seizures Induced by 1 (400 nmol, ic) in NIH Swiss
Micea
Ta ble 4. Matrix of Combined Shape and Electrostatic
Potential Energy Similarity Coefficients (1:1 weighting) for
Protonated (-NH3) and Nonprotonated (-NH2) Forms of 5,
(-)-9, and (-)-10
(-)-9
(-)-10
(-)-9- (-)-9- (-)-10- (-)-10-
dose (mg/kg, ip)
1 h
2 h
4 h
1 h
2 h
4 h
compd
5-NH2
5-NH2 5-NH3 NH2
NH3
NH2
NH3
controlb
controlc
3
10
30
0/10
9/10
0/10
9/10
6/10
4/10
3/10
2/10
0/10
9/10
0/10
9/10
7/10
5/10
4/10
3/10
0/10
9/10
0/10
9/10
1
5-NH3
0.93
0.99
0.93
0.90
0.93
1
(-)-9-NH2
(-)-9-NH3
(-)-10-NH2
(-)-10-NH3
0.93
0.90
0.91
0.97
1
7/10
3/10
2/10
7/10
5/10
2/10
7/10
2/10
0/10
6/10
3/10
2/10
0.92
0.90
0.93
1
0.87
0.93
1
0.94
100
1
ED50 (mg/kg)
19
6
26
14
15
14
a
Compound (-)-9 or (-)-10 was administered by the ip route
at the indicated time point prior to ic administration of 1. Data
are expressed as the number of mice observed to have exhibited
limbic seizures/number treated at any time during the 15-min
for the NH2 species (Figure 5, right panel) show a
marked difference between (-)-10-NH2 when compared
with either (-)-9-NH2 or 5-NH2 and no significant
difference when comparing (-)-9-NH2 to 5-NH2. The
LUMO in (-)-10-NH2 seems to be strongly localized
around the soft sulfur atom, whereas the LUMO in both
5 and (-)-9 is primarily localized on the amino face of
these compounds.
b
observation period following administration of 1.37 Mice were
treated with H2O (ip) at the indicated time point prior to ic
administration of H2O. c Mice were treated with H2O (ip) at the
indicated time point prior to ic administration of 1.
Quantum mechanical calculations were used to probe
more subtle effects, in particular the charge and poten-
tial distributions about the ionizable functional groups
in these molecules. Figure 4 (right) shows the electro-
static potential map projected onto an isoelectron sur-
face. As can be seen, there is very little difference
between the maps generated for 5-NH2, (-)-9-NH2, and
(-)-10-NH2. The charges on the oxygen atoms of the
two carboxylate groups are essentially identical for all
three molecules. The properties that provide the great-
est discrimination between the NH2 form of these
compounds are the HOMO and LUMO orbitals (Figures
5). As shown in Figure 5, the HOMO coefficients (left
panel) are larger at the ring heteroatom position for (-)-
10-NH2 and, to a lesser extent, (-)-9-NH2 when com-
pared with 5-NH2. The coefficients on the LUMO orbital
Com p a r ison of -NH3 Sp ecies. The protonated
(-NH3) species show a greater variation between the
three compounds, with (-)-10 being clearly differenti-
ated from (-)-9 and 5. The potential maps for the acid
and the amine faces of these molecules (Figure 6 top)
show clear differences at both of the acid positions and
on the amine. This is mirrored in the HOMO and LUMO
orbitals (Figure 6, bottom) in which it is evident that
the sulfur atom is having a major effect on both HOMO
and LUMO coefficients in compound (-)-10-NH3. In
5-NH3 and (-)-9-NH3, the HOMO is focused on the C6-
carboxylic acid and the LUMO around the amine; in (-)-
10-NH3, both of these effects are diluted by the sulfur
atom.

Heterobicyclic mGluR Agonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 1033
F igu r e 4. Left: Overlay of energy-minimized structures of 5-NH2 (cyan), (-)-9-NH2 (green), and (-)-10-NH2 (purple) showing
the close match of the acids and bases and the slight differences in ring geometry at the 2-position. Right: Map of the electrostatic
potential projected on the surface defined by a constant electron density of 0.002 electron/au3 and contoured between -110 and
-225 au. The molecules are from left to right: (-)-10-NH2, (-)-9-NH2, and 5-NH2.
F igu r e 5. HOMO (left) and LUMO (right) orbital coefficients of the -NH2 species. The molecules are from left to right: (-)-
10-NH2, (-)-9-NH2, and 5-NH2.
Discu ssion
substantially higher concentrations, where isomeric
contamination by the active enantiomer cannot be
entirely ruled out. As has been shown for 5, neither
(-)-9 nor (-)-10 have any measurable affinity at native
rat brain NMDA, AMPA, or kainate receptors up to
100 000 nM.
Metabotropic glutamate research has been greatly
facilitated with the appearance of highly potent agonists
and antagonists capable of discriminating between the
individual receptor subtypes. Our laboratory has focused
on the development of ligands directed toward the group
II (mGlu2/3) receptors, with particular attention to the
development of novel agonists. The characterization of
5 as a highly potent and selective agonist for group II
mGluRs has provided important information regarding
the conformation that the naturally occurring agonist,
glutamate, likely assumes when binding to and activat-
ing these receptors. With this in mind, we set out to
prepare molecules that maintain the conformational
features of 5 while simultaneously avoiding the incor-
poration of additional steric bulk about the core bicyclic
ring system that might interfere with high-affinity
receptor recognition and/or group II selectivity.38
The affinity of (-)-9 and (-)-10 for native and
recombinant group II mGluRs compares favorably to
that observed for the parent carbocyclic amino acid 5,
with each being slightly greater than 1 order of mag-
nitude more potent than this compound in binding to
both native group II and recombinant mGlu3 receptors.
While enhanced affinity was also observed for these
analogues relative to 5 at recombinant mGlu2 receptors,
it was of a more modest (approximately 2-5-fold)
nature. As a consequence, (-)-9 and (-)-10 possess
significantly different mGlu2/mGlu3 receptor selectivity
ratios (Ki ratio for mGlu2/mGlu3 ) 2.4 for (-)-9 and
8.6 for (-)-10) compared to the parent amino acid (Ki
ratio for mGlu2/mGlu3 ) 0.8). Thus, while 5 possesses
approximately equal affinity for both mGlu2 and mGlu3
receptors, the heterocyclic analogues are considerably
more mGlu3 receptor-selective.
Both (-)-9 and (-)-10 bind with high (IC50 ∼ 10 nM)
affinity to group II mGlu receptors expressed in native
rat brain homogenates and to cell membranes express-
ing recombinant human mGlu2 and mGlu3 receptor
subtypes as measured by specific displacement of the
group II-selective radioligand [3H]LY341495. Affinity of
these compounds for the group II mGluRs was found to
be highly stereoselective, with the corresponding (+)-
enantiomers displacing [3H]LY341495 binding only at
Functionally, both (-)-9 and (-)-10 potently sup-
pressed forskolin-stimulated c-AMP formation in cells
expressing recombinant human mGlu2 and mGlu3
receptors, indicative of group II mGluR agonist activity.

1034 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Monn et al.
F igu r e 6. Top panels: Maps of the electrostatic potential contoured between -160 and 60 au with views from the acid face (left)
and amine face (right) of the molecules. Bottom: HOMO (left) and LUMO (right) orbital coefficients. The molecules are from left
to right: (-)-10-NH3, (-)-9-NH3, and 5-NH3.
However, despite the higher affinity of these molecules
for mGlu3 receptors, each was equipotent, if not slightly
more potent, as an agonist in the mGlu2 receptor-
expressing cell line (EC50 values of 2.69 and 3.91 nM
for (-)-9 and (-)-10, respectively) in comparison to their
effects in cells expressing mGlu3 receptors (EC50 values
of 4.58 and 7.63 nM, respectively). This observation
cannot be explained simply on the basis of differential
levels of receptor expression in the two cell lines, as the
Bmax values for each are similar (29.2 and 35 pmol/mg
of protein for mGlu2 and mGlu3, respectively). It is
possible that the lack of direct correlation between
binding affinity and agonist potency may relate to
differences in the relationships that exist between
expression level and functional coupling of these recep-
tors to Gi protein subunits within the RGT cell line.
In addition to their high degree of potency at mGlu2
and mGlu3 receptors, compounds (-)-9 and (-)-10 also
demonstrated a high degree of group II mGluR agonist
selectivity when compared to their effects at other
mGluR subtypes. Neither compound produced agonist
or antagonist activity at mGlu1a, mGlu5a, or mGlu7
receptors up to 100 000 nM (10 000-100 000 times the
EC50 values for mGlu2/3 receptor activation). In cells
expressing recombinant mGlu4a receptors, high con-
centrations of (-)-9 produced a partial agonist effect
(EC50 ) 21 000 nM, Emax approximately 50%), while (-)-
10 and 5 were inactive as either an agonist or antagonist
up to 100 000 nM at this group III receptor. Each
compound also suppressed forskolin-stimulated c-AMP
formation in the mGlu8 receptor-expressing cell line
with similar (micromolar) potencies. However, the
heterocyclic amino acids (-)-9 and (-)-10 could be
distinguished from 5 at this target in that they produced
a nearly full agonist response (Emax ) 80-90%) com-
pared with the partial agonist effects (Emax ) 50%)
observed for 5. Finally, compound (-)-9 could be clearly
differentiated from 5 and (-)-10 in its agonist potency
in cells expressing mGlu6, an mGlu receptor localized
primarily in the retina.39 Thus, while both 5 and (-)-
10 produced agonist responses in this cell line at
micromolar concentrations (EC50 values ) 2 990 and
2 430 nM, respectively), (-)-9 more potently suppressed
the stimulated c-AMP response with an EC50 value of
401 nM, approximately 2 orders of magnitude higher
than that required for group II activation. This level of
activity is noteworthy as (-)-9 represents one of the
more potent compounds yet described for this mGlu
receptor subtype.40 Hence, while (-)-9 and (-)-10 in
general share a common pharmacological profile with
each other and with 5, and are highly selective for the
group II receptor class, the oxabicyclic analogue (-)-9
appears to possess modestly greater potency at mGlu4a
and mGlu8 receptors and significantly greater potency
at mGlu6 receptors than either the thiabicyclic or
carbobicyclic amino acid (-)-10 or 5.
We have previously demonstrated that selective group
II mGluR agonists are capable of blocking limbic
seizures produced by intracerebral administration of 1
(acting through group I mGlu receptor activation) and
have used this assay as a means to determine functional
group II mGluR agonist activity following systemic
administration of test ligands in mice.15,17 In the current
studies, compounds (-)-9 and (-)-10, when given by
the intraperitoneal route 1, 2, or 4 h prior to 400 nmol
of 1, blocked the limbic seizure activity produced by this

Heterobicyclic mGluR Agonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 1035
LC2000 apparatus using dual silica gel PrepPAK-500 car-
tridges. Solvent systems employed are given in parentheses
for each example. Preparative centrifugal thin-layer chroma-
tography (PC-TLC) was performed on a Harrison model 7924A
chromatotron using Analtech silica gel GF rotors. The solvent
system employed is indicated in the particular example.
Cation-exchange chromatography was performed with Dowex
50 × 8-100 ion-exchange resin and anion-exchange chroma-
tography with Bio-Rad AG1-X8 anion-exchange resin (hydrox-
ide form). Chiral HPLC analyses were performed using a
Chirobiotic V 4.6- × 250-mm column utilizing 20% THF/0.1%
trifluoroacetic acid/water as the eluent system at a solvent flow
rate of 0.6 mL/min. Detection was performed at λ ) 230 nm.
(()-Eth yl 2-Oxa bicyclo[3.1.0]h ex-3-en e-6-ca r boxyla te
(11).24 A solution of ethyl diazoacetate (100 g, 877 mmol) in
furan (250 mL) was added dropwise to a solution of Rh2(OAc)4
(0.5 g, 1.1 mmol) in furan (250 mL) with stirring at 10 °C over
a period of about 2-2.5 h. A further 0.1 g (0.2 mmol) of
Rh2(OAc)4 was added about two-thirds of the way into the
addition. After HPLC analysis showed complete consumption
of ethyl diazoacetate, a solution of NaHSO3 (200 g) in water
(400 mL) was added, and the resultant two-phase mixture was
allowed to warm to ambient temperature with stirring for 1-2
h. The reaction mixture was then extracted with methyl tert-
butyl ether (500 mL) and the organic phase washed with water
(400 mL) and saturated NaCl (300 mL) and then dried over
Na2SO4. The solvent was then removed by evaporation, and
the resultant oil was vacuum-distilled (45 °C at 0.2 mmHg) to
afford 11 (54 g, 351 mmol) in 40% yield: FDMS M+ ) 154; 1H
NMR (CDCl3) δ 0.95 (d, 1H, J ) 2.5 Hz), 1.26 (t, 3H, J ) 7
Hz), 2.78 (dd, 1H, J ) 2.5 and 5.7 Hz), 4.13 (q, 2H, J ) 7 Hz),
4.85 (d, 1H, J ) 5.7 Hz), 5.48 (t, 1H, J ) 2.6 Hz), 6.39 (d, 1H,
J ) 2.5 Hz); 13C NMR (CDCl3) δ 14.24, 22.10, 31.54, 60.59,
67.01, 106.31, 147.21, 172.82.
molecule in a dose-related manner (Table 3). The dose-
response in this assay was similar (10-100 mg/kg) for
both molecules, as was the influence of preadministra-
tion time, with both compounds producing a similar
degree of efficacy when administered either 1 h (ED50
) 19 and 14 mg/kg for (-)-9 and (-)-10, respectively)
or 4 h (ED50 ) 26 and 14 mg/kg for (-)-9 and (-)-10,
respectively) prior to 1.
In an attempt to understand the underlying physical
properties of (-)-9 and (-)-10 that give rise to the
enhanced affinity at group II (especially mGlu3) recep-
tors compared to 5, we have examined the influence of
the ring heteroatoms in (-)-9 and (-)-10 on molecular
conformation, electrostatic potential energies, and
HOMO/LUMO localization and compared these results
to those derived from 5. As shown in Table 4, hetero-
bicyclic amino acids (-)-9 and (-)-10 are highly similar
with each other and with 5 in terms of both electronic
and conformational properties. Furthermore, the simi-
larity calculations do not shed light on whether the
charge on the amino group is important to binding, as
the similarity coefficients between any nonprotonated
(-NH2) and protonated (-NH3) pair are high. Ab initio
calculations were subsequently performed on (-)-9, (-)-
10, and 5 in order to evaluate more closely the effects
of the ring heteroatoms in (-)-9 and (-)-10 on the
distribution of the highest occupied and lowest unoc-
cupied molecular orbitals relative to those of the parent
carbocycle 5 in similarly charged species. While clear
distinctions can be observed in comparing either the
HOMO or LUMO of thiabicyclic analogue (-)-10 to
carbocycle 5 in either the NH2 (Figure 5) or NH3 (Figure
6) forms, no such differences could be observed between
5 and oxabicyclic amino acid (-)-9. Hence, the enhanced
group II mGluR potency of (-)-9 and (-)-10 relative to
5 cannot be explained by changes in the distribution of
these molecular orbitals in similarly charged species.
(()-Eth yl 4-Hyd r oxy-2-oxa bicyclo[3.1.0]h exa n e-6-ca r -
boxyla te (12). A solution of thexylborane was prepared by
adding a solution of 2,3-dimethyl-2-butene (4 M, 53.0 mL) in
THF via a syringe to borane-dimethyl sulfide complex (10 M,
21.2 mL) in a dry flask under nitrogen at approximately 0 °C.
The solution was stirred for 2 h at this temperature before
use. Compound 11 (32.73 g, 212.30 mmol) was dissolved in
150 mL of THF under N2. The resultant solution was cooled
with stirring to -0.5 °C. While the stirring solution was
cooling, the system was evacuated and purged twice with N2.
The entire thexylborane solution prepared above was added
via cannula over 40 min, maintaining the temperature be-
tween 0 and 5 °C. After the mixture stirred for 2 h at this
temperature, 87 mL of 30% H2O2 was added slowly, over 70
min, to maintain the temperature at <30 °C. Following the
peroxide addition, 15 mL of pH 7 phosphate buffer (1 M
KH2PO4 and 1 M in K2HPO4) was added, and the mixture was
allowed to stir for 14 h while warming to ambient temperature.
The mixture was cooled to <5 °C, and 25 mL of saturated
aqueous Na2S2O3 was added slowly; 75 mL of EtOAc was then
added, followed slowly by 75 mL of saturated aqueous Na2S2O3.
The mixture was stirred for 15 min and then partitioned
between 75 mL of EtOAc and 30 mL of saturated aqueous
Na2S2O3. The aqueous layer was back-extracted three times
with 50 mL of EtOAc. The combined organic layers were
washed with 30 mL of brine and dried over Na2SO4. The
solvent was removed to afford 54.44 g of an oil. The oil was
purified by a flash chromatography (370 g of silica gel, wet
packed with 3:2 hexanes:EtOAc) eluting with 3:2 hexanes:
EtOAc, to afford 12 (31.72 g, 16% EtOAc by 1H NMR, corrected
yield ) 26.6 g, 154.5 mmol) in 73% yield. An analytical sample
was prepared by further evaporation of solvent (with some
Con clu sion
LY379268 ((-)-9) and LY389795 ((-)-10) are the most
potent agonists yet described for group II metabotropic
glutamate receptors. The high mGlu2/3 potency and
selectivity evinced by these heterobicyclic amino acids
lend further experimental evidence supporting the
hypothesis that glutamate interacts with mGlu2/3
receptors in a fully extended conformation. Like 5,
heterocycles (-)-9 and (-)-10 effectively penetrate into
the CNS following systemic administration and block
seizures induced by group I mGlu receptor activation.
Thus, like 5, these agents may be useful for studying
the functions of mGlu2/3 receptors both in vitro and in
vivo.
Exp er im en ta l Section
Melting points were obtained using a Thomas-Hoover capil-
liary melting point apparatus and are uncorrected. 1H and 13C
NMR data were obtained at 300.15 and 75.48 MHz, respec-
tively, with TMS as an internal standard. Field desorption
mass spectroscopy (FDMS) was performed using either a VG
70SE or Varian MAT 731 instrument. Optical rotations were
obtained using the Perkin-Elmer 241 polarimeter and are
reported at the sodium D-line (589 nm), unless otherwise
noted. Preparative HPLC was performed with the Waters Prep
1
product loss): FDMS M+ ) 172; H NMR (CDCl3) δ 1.24 (t,
3H, J ) 7.1 Hz), 1.82 (dd, 1H, J ) 1.0 and 3.7 Hz), 2.25 (t, 1H,
J ) 4.2 Hz), 2.25 (br s, 1H), 3.55 (dd, 1H, J ) 4.4 and 11.1
Hz), 3.94 (d, 1H, J ) 11.1 Hz), 4.10 (q, 2H, J ) 7.1 Hz), 4.36
(dd, 1H, J ) 1.0 and 4.9 Hz), 4.45 (d, 1H, J ) 4.4 Hz); 13C
NMR (DMSO-d6) δ 14.09, 21.25, 30.94, 60.74, 64.41, 72.35,
72.39, 74.49, 170.77. Anal. (C8H12O4) C, H.

1036 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Monn et al.
(()-E t h yl 4-Oxo-2-oxa b icyclo[3.1.0]h exa n e-6-ca r b ox-
yla te (13). Oxalyl chloride (25.70 g, 202.44 mmol) in CH2Cl2
(300 mL) under N2 was added dropwise over 35 min to a
solution of DMSO (28.74 g, 367.8 mmol) in CH2Cl2 (75 mL)
while keeping the temperature below -65 °C. The solution was
stirred for 10 min and cooled back to -70 °C. A solution of
31.68 g of 12 (26.29 g, 152.71 mmol, corrected for residual
EtOAc) dissolved in 100 mL of CH2Cl2 was added dropwise
over 40 min while maintaining the temperature at -67 °C.
The mixture was stirred for 5 min; then 62 mL (45.01 g, 444.83
mmol) of triethylamine was added dropwise over 15 min,
keeping the temperature below -50 °C. After stirring for 15
min, TLC indicated complete reaction, and the mixture was
allowed to warm to about -40 °C. The mixture was filtered
and washed through with 300 mL of CH2Cl2. The filtrate was
extracted two times with 150 mL of 1 N HCl. The aqueous
layer was back-extracted with 50 mL of CH2Cl2. The combined
organic layers were washed with 75 mL of brine and dried
over Na2SO4. Most of the solvent was removed by rotary
evaporation to leave 44.36 g of liquid. A few seed crystals were
added, and the flask was blanketed with N2 and stirred at
ambient temperature for 30 min while a thin slurry formed.
To the room temperature slurry was slowly added 20 mL of
hexanes. The slurry was stirred for 90 min at ambient
temperature and then for 3 h in an ice/NaCl/water bath. The
solids were filtered, washed with 25 mL of 5:1 hexanes:EtOAc,
and dried under vacuum to afford 13 (19.48 g, 114.6 mmol) as
white crystals. A second crop of crystals (2.28 g, 13.4 mmol)
was obtained. The combined yield was 84%: mp 62-65 °C;
in 2 N NaOH (63.2 mL) was stirred for 30 min at ambient
temperature. The hydrolysis was then quenched by the addi-
tion of 12 N HCl (5.27 mL, 63.2 mmol). The reaction mixture
was stirred for 3 h at 0 °C and then vacuum-filtered. The solid
collected was dried under vacuum at 50 °C overnight, affording
(()-15 (6.12 g, 28.9 mmol) in 91% yield: mp 258-261 °C;
1
FDMS M+ ) 212; H NMR (DMSO-d6) δ 2.24 (s, 1H), 2.26 (s,
1H), 3.35 (d, 1H, J ) 11 Hz), 4.05 (d, 1H, J ) 11 Hz), 4.39 (d,
1H, J ) 5 Hz); 13C NMR (DMSO-d6) δ 22.14, 30.75, 65.74,
68.32, 70.61, 156.32, 171.11, 175.63. Anal. (C8H8N2O5) C, H,
N.
4-(Spir o-5′-h ydan toin )-2-oxabicyclo[3.1.0]h exan e-6-car -
boxylic Acid , (R)-2-P h en ylglycin ol Sa lt ((-)-16). To a
solution of (R)-2-phenylglycinol (0.52 g, 3.8 mmol) in ethanol
(20 mL) and water (4 mL) was added (()-15 (0.80 g, 3.8 mmol).
The mixture was heated to reflux, and an additional 1 mL of
water was added, producing a homogeneous solution. After
approximately 30 min at reflux, the mixture was allowed to
cool to ambient temperature. After stirring overnight, the
reaction mixture was filtered, washed with 1 mL of a cold 25:5
mixture of ethanol and water, and dried under vacuum at 50
°C overnight, to afford (-)-16 (0.57 g, 1.63 mmol) in 43% yield
as a white solid: mp 198-201 °C; optical rotation RD ) -111°
1
(c ) 1, H2O); H NMR (DMSO-d6) δ 2.05 (t, 1H, J ) 3.3 Hz),
2.20 (d, 1H, J ) 3 Hz), 3.30 (d, 1H, J ) 11 Hz), 3.50 (m, 1H),
3.55 (m, 1H), 4.0 (d, 1H, J ) 11 Hz), 4.1 (m, 1H), 4.18 (d, 1H,
J ) 6 Hz), 7.25 (m, 1H), 7.30 (m, 2H), 7.35 (m, 2H). Anal.
(C16H19N3O6) C, H, N. The enantiomeric excess was determined
to be 98.8% by chiral HPLC.27
1
FDMS M+ ) 170; H NMR (CDCl3) δ 1.19 (t, 3H, J ) 7 Hz),
4-(Spir o-5′-h ydan toin )-2-oxabicyclo[3.1.0]h exan e-6-car -
boxylic Acid , (S)-2-P h en ylglycin ol Sa lt ((+)-16). Combined
filtrates from various crystallizations of the (R)-2-phenyl-
glycinol salt (-)-16 were concentrated to dryness (4.61 g, 13.2
mmol), then treated with 1 N HCl, and extracted with hot
EtOAc. The organic phases were combined, dried over MgSO4,
and concentrated to afford the optically enriched hydantoin
6-carboxylate (+)-15 (2.8 g, 13.2 mmol, ee ∼ 70%). This was
stirred in hot EtOH (100 mL) and treated in one portion with
a solution of (S)-2-phenylglycinol in hot EtOH (50 mL). The
resulting mixture was warmed to boiling as the volume was
reduced to approximately 200 mL. The solids were filtered,
and the filtrate was allowed to slowly cool to room temperature
overnight. Three crops of crystals of (+)-16 were obtained (2.35
g, 6.7 mmol) in 51% yield from enriched (+)-15. Each crop was
found to possess an ee > 98% and were thus combined: mp
217-220 °C; optical rotation RD ) 103° (c ) 1, H2O). Anal.
(C16H19N3O6‚0.5H2O) C, H, N.
(+)-4-(Sp ir o-5′-h yd a n toin )-2-oxa bicyclo[3.1.0]h exa n e-
6-ca r boxylic Acid ((+)-15). Resolved salt (+)-16 (2.30 g, 6.6
mmol) was partitioned between hot EtOAc and 1 N HCl. The
layers were separated, and the aqueous one was extracted with
hot EtOAc. The organic phases were combined, dried over
MgSO4, and concentrated to afford (+)-15 (1.3 g, 6.1 mmol) in
93% yield: mp 267-269 °C; FDMS M+ ) 212; optical rotation
RD ) 128° (c ) 1, methanol); ee ) 99%. Anal. (C8H8N2O5‚
0.4AcOH) C, H, N.
2.35 (dd, 1H, J ) 0.8 and 3.5 Hz), 2.47 (t, 1H, J ) 4.1 Hz),
3.83 (d, 1H, J ) 17.6 Hz), 4.0 (d, 1H, J ) 17.8 Hz), 4.08 (q,
2H, J ) 7 Hz), 4.75 (d, 1H, J ) 4.4 Hz); 13C NMR (CDCl3) δ
14.04, 27.25, 31.85, 61.48, 67.94, 69.97. Anal. (C8H10O4‚
0.25H2O) C, H.
(()-E t h yl 4-(Sp ir o-5′-h yd a n t oin )-2-oxa b icyclo[3.1.0]-
h exa n e-6-ca r boxyla te (14). To a slurry of ammonium car-
bonate (5.65 g, 58.8 mmol) and potassium cyanide (2.01 g, 30.9
mmol) in 25 mL of methanol at ambient temperature was
added a solution of 13 (5.0 g, 29.4 mmol) in 25 mL of methanol.
The mixture was stirred at ambient temperature and moni-
tored by HPLC. After 23 h the reaction was judged complete
by TLC. The mixture was diluted with 100 mL of water, cooled,
and seeded. The pH was adjusted from 9.6 to 7.0 with 6 N
HCl giving a white solid. The slurry was stirred at 0-5 °C for
1.5 h, filtered, and washed with 75 mL of cold water-methanol
(2:1). The white solid was dried under vacuum at 40 °C
affording 14 (5.55 g, 23.1 mmol) in 79% yield: mp 161-163
1
°C; FDMS M+ ) 240; H NMR (DMSO-d6) δ 1.19 (t, 3H, J )
7 Hz), 2.38 (dd, 1H, J ) 5.4 and 3.9 Hz), 2.42 (dd, 1H, J ) 3.9
and 1.5 Hz), 3.37 (d, 1H, J ) 10.3 Hz), 4.05 (q, 2H, J ) 7 Hz),
4.07 (d, 1H, J ) 10.3 Hz), 4.39 (dd, 1H, J ) 5.4 and 1.5 Hz),
8.14 (s, 1H), 10.80 (br s, 1H); 13C NMR (DMSO-d6) δ 14.05,
21.95, 30.90, 60.40, 65.79, 68.21, 70.57, 156.18, 169.40, 175.41.
Anal. (C10H12N2O5‚0.6H2O) C, H, N.
(()-4-Am in o-2-oxa b icyclo[3.1.0]h exa n e-4,6-d ica r b ox-
ylic Acid ((()-9). A solution of 14 (0.33 g, 1.37 mmol) in 2 N
NaOH (10 mL) was warmed under reflux for 24 h in a glass
flask. The pH was adjusted to approximately 11 with 1 N HCl,
and the resulting solids (presumably silicic acid from the
reaction vessel) were removed by filtration. After the volume
was reduced to approximately 5 mL under reduced pressure,
the solution was applied to AG1-X8 anion-exchange resin. The
product was eluted with 3 N AcOH and concentrated to
dryness. The amorphous solid was crystallized from 15 mL of
H2O (saturated solution pH ∼ 3) to afford (()-9 (0.158 g, 0.84
mmol) in 62% yield: mp > 200 °C dec; FDMS M+ (+H) ) 188;
1H NMR (D2O + KOD) δ 2.15 (br d, 1H, J ) 4 Hz), 2.31 (dd,
1H, J ) 4.1 and 4.6 Hz), 3.45 (d, 1H, J ) 9.8 Hz), 4.20 (d, 1H,
J ) 9.8 Hz), 4.39 (br d, 1H, J ) 5 Hz); 13C NMR (D2O + KOD,
63 MHz) δ 26.35, 33.61, 66.70, 75.12, 179.05 (one carbonyl not
observed). Anal. (C7H9NO5) C, H, N.
(-)-4-(Sp ir o-5′-h yd a n toin )-2-oxa bicyclo[3.1.0]h exa n e-
6-ca r boxylic Acid ((-)-15). Resolved salt (-)-16 (3.8 g, 10.9
mmol) was partitioned between hot EtOAc and 1 N HCl. The
layers were separated, and the aqueous one was extracted with
hot EtOAc. The organic phases were combined, dried over
MgSO4, and concentrated to afford (-)-15 (2.15 g, 10.1 mmol)
in 93% yield: mp 260-262 °C; FDMS M+ (+H) ) 213; optical
rotation RD ) -134° (c ) 1, methanol); ee ) 99.6%. Anal.
(C8H8N2O5) C, H, N.
(+)-4-Am in o-2-oxa b icyclo[3.1.0]h exa n e-4,6-d ica r b ox-
ylic Acid ((+)-9). A solution of (+)-15 (1.2 g, 5.6 mmol) in 2
N NaOH (15 mL) was warmed under reflux overnight. The
reaction mixture was cooled to room temperature, treated with
6 N HCl to pH 1, and concentrated to dryness. The solids were
reconstituted in H2O, and 1 N NaOH was added to adjust the
pH to 12. The solution was subjected to ion-exchange chro-
matography (Bio-Rad AG 1-X8, elute with 3 N AcOH) to afford
(()-4-(Sp ir o-5′-h yd a n toin )-2-oxa bicyclo[3.1.0]h exa n e-
ca r boxylic Acid ((()-15). A solution of 14 (7.59 g, 31.6 mmol)

Heterobicyclic mGluR Agonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 1037
(+)-9 (0.83 g, 4.4 mmol) in 79% yield as a white solid after
evaporation: mp > 275 °C dec; FDMS M+(+H) ) 188; optical
rotation RD ) 62° (c ) 1, H2O). Anal. (C7H9NO5‚0.3H2O) C, H,
N.
3H, J ) 7 Hz), 2.34 (dd, 1H, J ) 3 Hz and 3 Hz), 2.58-2.60
(m, 1H), 3.19-3.23 (m, 1H), 3.30 (d, 2H, J ) 2 Hz), 4.16 (q,
2H, J ) 7 Hz); 13C NMR (CDCl3) δ 14.15, 29.26, 30.80, 34.08,
34.93, 61.71, 168.83, 206.40. Anal. (C8H10O3S) C, H.
(-)-4-Am in o-2-oxa b icyclo[3.1.0]h exa n e-4,6-d ica r b ox-
ylic Acid ((-)-9). A solution of (-)-15 (2.1 g, 10 mmol) in 2 N
NaOH (35 mL) was warmed under reflux overnight. The
reaction mixture was cooled to room temperature, treated with
6 N HCl to pH 1, and concentrated to dryness. The solids were
reconstituted in H2O, and 1 N NaOH was added to adjust the
pH to 12. The solution was subjected to ion-exchange chro-
matography (Bio-Rad AG 1-X8, elute with 3 N AcOH) to afford
(-)-9 (1.51 g, 8.07 mmol) in 81% yield as a white solid after
evaporation: mp > 275 °C dec; FDMS M+(+H) ) 188; 1H NMR
(D2O + KOD) δ 1.94 (br d, 1H, J ) 4 Hz), 2.08 (dd, 1H, J )
4.1 and 4.6 Hz), 3.15 (d, 1H, J ) 9.8 Hz), 3.95 (d, 1H, J ) 9.8
Hz), 4.15 (br d, 1H, J ) 5 Hz); 13C NMR (D2O + KOD, 63 MHz)
δ 26.79, 31.40, 66.50, 73.46, 174.64, 177.57; optical rotation
RD ) -63° (c ) 1, H2O). Anal. (C7H9NO5) C, H, N.
(()-Eth yl 4-(Sp ir o-5′-h yd a n toin )-2-th ia bicyclo[3.1.0]-
h exa n e-6-ca r boxyla te (20). A solution of 19 (3.22 g, 17.3
mmol) in EtOH (25 mL) and H2O (10 mL) at ambient
temperature was treated consecutively with (NH4)2CO3 (3.37
g, 43.3 mmol) and KCN (1.41 g, 21.6 mmol) and warmed at
35 °C until the reaction was judged complete by TLC. The
reaction mixture was acidified with 6 N HCl, aqueous NaCl
was added, and the product was extracted with EtOAc. All of
the organic phases were combined, dried (MgSO4), and con-
centrated. Crystallization from 2-propanol afforded 20 (2.25
g, 8.8 mmol) in 51% yield: mp 197-200 °C; FDMS M+ ) 256;
1H NMR (DMSO-d6) δ 1.20 (t, 3H, J ) 7 Hz), 2.16 (dd, 1H, J
) 3.4 and 3.9 Hz), 2.39 (dd, 1H, J ) 3.9 and 6.9 Hz), 2.72 (dd,
1H, J ) 1 and 12.8 Hz), 2.96 (dd, 1H, J ) 3.4 and 6.9 Hz),
3.17 (dd, 1H, J ) 1 and 12.8 Hz), 4.08 (q, 2H, J ) 7 Hz), 8.07
(d, 1H, J ) 1 Hz), 10.79 (s, 1H) (also observed protons for
2-propanol, integrated ratio ) 0.75 eq); 13C NMR (DMSO-d6)
δ 13.99, 24.33, 25.44, 29.99, 33.99, 34.90, 60.55, 61.98, 70.55,
155.87, 169.93, 174.85. Anal. (C10H12N2O4S‚0.75 2-propanol)
C, H, N.
(()-4-Am in o-2-t h ia b icyco[3.1.0]h exa n e-4,6-d ica r b ox-
yla te ((()-10). A solution of 20 (0.85 g, 3.30 mmol) in 2 N
NaOH (20 mL) was warmed under reflux for 4 days. The
reaction mixture was then acidified with 6 N HCl and
concentrated to dryness. The solid was reconstituted in H2O
at pH 11, applied to AG1-X8 anion-exchange resin, eluted with
3 N AcOH, and concentrated to dryness. The product was
triturated in a mixture of hot H2O/2-propanol mixture and
filtered to afford (()-10 (0.31 g, 1.5 mmol) in 46% yield: mp
> 250 °C; FDMS M+ ) 203; 1H NMR (D2O + KOD) δ 1.55 (dd,
1H, J ) 3.1 and 3.8 Hz), 2.02 (dd, 1H, J ) 4.1 and 6.9 Hz),
2.25 (d, 1H, J ) 12.4 Hz), 2.41 (dd, 1H, J ) 3.1 and 6.9 Hz),
2.79 (d, 1H, J ) 12.4 Hz); 13C NMR (D2O + KOD) δ 24.89,
26.60, 35.16, 36.11, 67.32, 177.31, 178.10. Anal. (C7H9NO4S‚
0.5H2O) C, H, N.
(()-4-(Sp ir o-5′-h yd a n toin )-2-th ia bicyclo[3.1.0]h exa n e-
6-ca r boxyla te ((()-21). To a 1000-mL recovery flask were
added 20 (93.6 g, 365 mmol) and 365 mL (730 mmol) of 2 N
NaOH. The resulting red homogeneous solution was stirred
at ambient temperature for approximately 30 min. HPLC
analysis showed consumption of starting material and forma-
tion of 21. The reaction was quenched by the addition of 12 N
HCl (60.9 mL, 730 mmol). The resulting mixture was cooled
to 0 °C in an ice-water bath and stirred for 3 h before being
vacuum-filtered through a glass frit. The precipitated product
was dried under vacuum at 50 °C overnight, affording (()-21
(72.3 g, 317 mmol) in 87% yield as an off-white solid: mp >
280 °C dec; M+ ) 228; 1H NMR (DMSO-d6) δ 2.02 (t, 1H, J )
2 Hz), 2.30 (dd, 1H, J ) 3 and 3 Hz), 2.69 (d, 1H, J ) 11 Hz),
2.89 (dd, 1H, J ) 4 and 4 Hz), 3.15 (d, 1H, J ) 11 Hz); 13C
NMR (DMSO-d6) δ 24.62, 29.73, 34.19, 34.87, 70.70, 156.01,
171.62, 175.08. Anal. (C8H8N2O4S) C, H, N.
(()-4-(Sp ir o-5′-h yd a n toin )-2-th ia bicyclo[3.1.0]h exa n e-
6-ca r boxyla te (R)-2-P h en ylglycin ol Sa lt ((-)-22). To a 5-L
three-neck round-bottom flask were added (()-21 (70.1 g, 307
mmol) and (R)-2-phenylglycinol (42.1 g, 307 mmol) in 3500 mL
of 95% ethanol (toluene-denatured) and 700 mL of water. The
resulting brown heterogeneous mixture was heated to reflux.
At reflux all the solid material dissolved, producing a homo-
geneous dark amber solution. After refluxing for approximately
15 min, the reaction mixture was allowed to slowly cool to room
temperature. After stirring at room temperature overnight
(approximately 16 h) the mixture was vacuum-filtered through
a glass frit. The solid collected on the frit was rinsed with 2 ×
5-mL portions of a cold mixture of 95% ethanol (toluene-
denatured) and water (50:10) before being dried under vacuum
at 60 °C overnight. The product (-)-22 (40.16 g, 110 mmol)
was obtained in 36% yield as a white solid: mp 228-230 °C;
1H NMR (DMSO-d6) δ 1.95 (t, 1H, J ) 3 Hz), 2.13 (m, 1H),
2.68 (d, 1H, J ) 11 Hz), 2.71 (m, 1H), 3.11 (d, 1H, J ) 11 Hz),
(()-Eth yl 2-Th ia bicyclo[3.1.0]h ex-3-en e-6-ca r boxyla te
(17). A solution of ethyl diazoacetate (11.4 g, 100 mmol) in
thiophene (20 mL) was added dropwise to a 70 °C solution of
Rh2(OAc)4 in thiophene (100 mL). Upon complete addition, the
reaction mixture was warmed under reflux for 3 h, concen-
trated to an orange-colored oil, and purified by preparative
HPLC (10% EtOAc/hexanes) to afford 17 (6.51 g, 38.2 mmol)
1
in 38% yield: FDMS M+ ) 170; H NMR (CDCl3) δ 1.09 (dd,
1H, J ) 3 and 3 Hz), 1.26 (t, 3H, J ) 7 Hz), 3.04 (ddd, 1H, J
) 2.6, 3.3, and 7.5 Hz), 3.49 (ddd, 1H, J ) 1.8, 3.0, and 7.5
Hz), 4.15 (q, 2H, J ) 7 Hz), 5.88 (dd, 1H, J ) 5.6 and 2.6 Hz),
6.15 (dd, 1H, J ) 1.8 and 5.6 Hz); 13C NMR (CDCl3) δ 14.29,
23.92, 34.06, 38.99, 60.89, 122.94, 127.66, 174.05. Anal.
(C8H10O2S) C, H.
(()-Eth yl 4-Hyd r oxy-2-th ia bicyclo[3.1.0]h exa n e-6-ca r -
boxyla te (18). A solution of thexylborane was prepared by
dropwise addition of a solution of borane-dimethyl sulfide
complex (10 M, 5.46 mL, 54.6 mmol mL) to a solution of 2,3-
dimethyl-2-butene (6.49 mL, 54.6 mmol) in THF (10 mL) at 0
°C. The solution was stirred for 2 h at this temperature and
then was added via cannula to a solution of 17 (9.30 g, 54.6
mmol) in THF (50 mL) at 0 °C. Upon complete addition, the
cooling bath was removed and the reaction was allowed to
warm to ambient temperature and remain there for 2 h. The
reaction mixture was then cooled to 0 °C, and 71 mL of pH 7
phosphate buffer (1 M KH2PO4 and 1 M in K2HPO4) was added
dropwise, followed immediately by NaBO3‚4H2O (10.1 g, 65.5
mmol). The resulting mixture was allowed to stir at ambient
temperature for 1 h. The mixture was then partitioned
between H2O and EtOAc, and the product was extracted with
EtOAc. The organic layers were combined, dried (Na2SO4), and
concentrated. The carbinol was purified by preparative HPLC
(10% EtOAc/hexanes to 50% EtOAc/hexanes eluent) affording
18 (4.50 g, 23.9 mmol) in 44% yield: FDMS M+ ) 188; 1H NMR
(CDCl3) δ 1.19 (t, 3H, J ) 7 Hz), 1.59 (dd, 1H, J ) 3 Hz and
3 Hz), 2.29-2.32 (m, 1H), 2.64 (dd, 1H, J ) 13 and 4 Hz), 2.81-
2.88 (m, 3H), 4.07 (q, 2H, J ) 7 Hz), 4.54 (d, 1H, J ) 4 Hz);
13C NMR (CDCl3) δ 14.09, 25.10, 29.29, 35.86, 37.63, 60.90,
74.42, ester carbonyl not observed. Anal. (C8H12O3S) C, H.
(()-Eth yl 4-Oxo-2-th ia bicyclo[3.1.0]h exa n e-6-ca r box-
yla te (19). A solution of 18 (5.5 g, 29.2 mmol) in CH2Cl2 (25
mL) was added slowly to a solution of oxalyl chloride (3.8 mL,
43.8 mmol) and DMSO (4.1 mL, 58.4 mmol) in CH2Cl2 (325
mL) at -78 °C. The rate of addition of 18 was monitored so to
maintain an internal temperature below -60 °C. Upon com-
plete addition, the reaction was allowed to slowly warm to -40
°C, then was chilled to -78 °C, and treated dropwise with
triethylamine (15.18 g, 150 mmol). The reaction mixture was
allowed to warm to room temperature and then was parti-
tioned between 1 N HCl and Et2O. The product was extracted
with Et2O; the organic phases were combined, washed with
brine, dried (MgSO4), and concentrated. The product was
purified by preparative HPLC (5% EtOAc/hexanes to 75%
EtOAc/hexanes) to afford 19 (3.34 g, 18 mmol) in 61% yield:
1
mp 56-57 °C; FDMS M+ ) 186; H NMR (CDCl3) δ 1.27 (t,

1038 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Monn et al.
3.59 (m, 2 H), 4.1 (m, 1H), 7.25 (m, 1H), 7.31 (m, 2H), 7.38 (m,
2H); 99.9% ee (by HPLC); optical rotation [R]D ) -149.56 (c )
1.0, methanol). Anal. (C16H19N3O5S‚0.75H2O) C, H; N: calcd,
11.10; found, 10.28.
solid was dried in a vacuum oven at 70 °C for 18 h to afford
crude (-)-10 (22.0 g, 108.2 mmol) in 81% yield. This material
was combined with a separate batch of similarly prepared
material (24.3 g, 120 mmol) in 231 mL of water, and the thick
slurry was heated at reflux for 2 h. The mixture was allowed
to stir at room temperature overnight and subsequently for
1.5 h in an ice/water bath. The solid was collected by filtration,
washed with 25 mL of cold water, and dried at 75 °C, affording
(-)-10 (41.69 g 205.0 mmol) in 76% overall yield: mp 273-
274 °C dec; FDMS M+ ) 203.4; 1H NMR (D2O + KOD, 300
MHz) δ 1.58 (t, 1, J ) 3.5 Hz), 2.07 (dd, 1, J ) 7.0 and 4.2
Hz), 2.29 (d, 1, J ) 12.5 Hz), 2.44 (dd, 1, J ) 7.0 and 3.0 Hz),
2.84 (d, 1, J ) 12.5 Hz); 13C NMR (D2O + KOD, 63 MHz) δ
28.71, 30.64, 36.43, 37.20, 70.04, 176.55, 179.25; optical
rotation [R]D ) -159.6 (c ) 1, H2O). Anal. (C7H9NO4S) C, H,
N.
(+)-4-Am in o-2-t h ia b icyco[3.1.0]h exa n e-4,6-d ica r b ox-
yla te ((+)-10). A solution of 1.12 g (4.9 mmol) of (+)-21 in 15
mL of 2 M NaOH (29 mmol, 6 equiv) was placed in a Teflon
sleeve inside a 50-mL Parr bomb. The mixture was heated at
97 °C for 16 h. After cooling to room temperature, the solution
was transferred to a flask and treated with 12 N HCl to lower
the pH to about 3 (2.5 mL). The resulting slurry was stirred
at room temperature for 30 min and at 0 °C for 1 h. The solid
was collected by filtration and rinsed with cold H2O. After
drying with vacuum on the filter for 30 min, the soapy solid
was dried in a vacuum oven at 70 °C for 1 h to afford 0.78 g
(78%) of (+)-10 as a white solid: mp >280 °C dec; FDMS M+
) 203; 1H NMR (D2O + KOD) δ 1.68 (t, 1H, J ) 3.7 Hz), 2.15
(dd, 1H, J ) 4.0 and 7.1 Hz), 2.39 (d, 1H, J ) 12.4 Hz), 2.55
(dd, 1H, J ) 3.3 and 7.1 Hz), 2.94 (d, 1H, J ) 12.4 Hz); 13C
NMR (D2O + KOD) δ 27.82, 29.92, 38.08, 38.98, 70.43, 180.77,
181.20; optical rotation [R]D ) 168.2 (c ) 1.0, H2O). Anal. (C7H9-
NO4S) C, H, N.
(()-4-(Sp ir o-5′-h yd a n toin )-2-th ia bicyclo[3.1.0]h exa n e-
6-ca r b oxyla t e (S)-2-P h en ylglycin ol Sa lt ((+)-22). The
filtrate from resolution of 5.5 g of (()-21 using (R)-2-phenyl-
glycinol was concentrated under reduced pressure, and the
residue was dissolved in 50 mL of H2O. As 7 mL of 6 M HCl
was added, a precipitate formed. The precipitate was collected
by filtration and washed with H2O. The solid was dried in a
vacuum oven at 80 °C for 6 h to afford 2.26 g of optically
enriched (+)-21. The acid was combined with 1.35 g of (S)-2-
phenylglycinol, 113 mL of ethanol, and 23 mL of H2O. The
mixture was heated to reflux to afford a solution and was then
allowed to cool to room temperature overnight. The resulting
solid was collected by filtration and dried in a vacuum oven
at 50 °C for 6 h to afford 2.42 g of (+)-22 (2.15 g corrected for
residual ethanol, 60% yield from enriched (+)-21) as a white
solid: 98.4% ee by chiral HPLC; mp 230-232 °C; 1H NMR
(DMSO-d6) δ 1.93 (t, 1, J ) 5 Hz), 2.11 (dd, 1H, J ) 5 and 7
Hz), 2.68 (m, 2H), 3.11 (d, 1H, J ) 12 Hz), 3.55 (m, 2H), 4.10
(dd, 1H, J ) 5 and 7 Hz), 7.25-7.45 (m, 5H), 8.12 (br s, 1H);
optical rotation [R]D ) 147.87 (c ) 1.0, methanol). Anal.
(C16H19N3O5S‚0.90H2O) C, H; N: calcd, 11.01; found, 10.49.
(-)-4-(Sp ir o-5′-h yd a n toin )-2-th ia bicyclo[3.1.0]h exa n e-
6-ca r b oxyla t e ((-)-21). To a 2-L three-neck round-bottom
flask were added (-)-22 (113 g, 308 mmol) and 675 mL of
water. The resulting heterogeneous brown mixture was stirred
at ambient temperature for approximately 15 min during
which time all the solid material was observed to dissolve
producing a homogeneous brown solution. HCl (12 N) (28.4
mL, 339 mmol) was then added dropwise over 5 min. An off-
white precipitate formed. An ice-water bath was placed
underneath the flask and its contents stirred at 0 °C for 60
min before being vacuum-filtered through a glass frit. The off-
white solid collected on the frit was rinsed with 2 × 5-mL
portions of cold water before being dried under vacuum at 60
°C overnight to afford (-)-21 (61.72 g, 270.5 mmol) in 88%
Ack n ow led gm en t. The authors gratefully acknowl-
edge Dr. Thomas Britton for helpful discussions per-
taining to the preparation of 9 and 10. In addition, the
authors thank Mr. J oseph Kennedy for performing the
chiral HPLC analyses, Dr. Gregory Stevenson for the
X-ray crystal structure of (-)-9, Dr. D. O. Calligaro and
Mr. J . S. Horng for the ionotropic glutamate receptor
binding data, and the Physical Chemistry Department
for providing NMR, mass spectral, and combustion
analyses.
1
yield as an off-white solid: mp > 280 °C dec; M+ ) 228; H
NMR (DMSO-d6) δ 2.01 (t, 1H, J ) 2 Hz), 2.29 (dd, 1H, J ) 3
and 8 Hz), 2.69 (d, 1H, J ) 12 Hz), 2.89 (dd, 1H, J ) 3 and 8
Hz), 3.15 (d, 1H, J ) 12 Hz), 8.10 (br s, 1H), 10.75 (br s, 1H),
12.52 (br s, 1H); 13C NMR (DMSO) δ 24.61, 29.70, 34.22, 34.86,
70.67, 155.98, 171.56, 175.02; optical rotation [R]D ) -202.3
(c ) 1.0, methanol). Anal. (C8H8N2O4S) C, H, N.
(+)-4-(Sp ir o-5′-h yd a n toin )-2-th ia bicyclo[3.1.0]h exa n e-
6-ca r boxyla te ((+)-21). To a slurry of 2.32 g (2.07 g corrected
for ethanol content, 98.4% ee) of (+)-22 in 18 mL of H2O was
added 2.1 mL of 3 N HCl. The resulting slurry was stirred for
5 min at room temperature and for 40 min at 0 °C. The solid
was collected by filtration and rinsed with H2O. The solid was
dried in a vacuum oven at 70 °C for 1.5 h to afford 1.23 g (95%)
of (+)-21 as a tan solid: 98.6% ee by chiral HPLC; mp >280
°C dec; FDMS M+ ) 228; 1H NMR (DMSO-d6) δ 2.01 (t, 1H, J
) 4 Hz), 2.32 (dd, 1H, J ) 3 and 8 Hz), 2.67 (d, 1H, J ) 12
Hz), 2.90 (dd, 1H, J ) 3 and 8 Hz), 3.15 (d, 1H, J ) 12 Hz),
8.12 (br s, 1H), 10.75 (br s, 1H), 12.51 (br s, 1H); 13C NMR
(DMSO-d6) δ 24.61, 29.70, 34.23, 34.86, 70.67, 155.98, 171.56,
175.02; optical rotation [R]D ) 203.9 (c ) 1.0, methanol). Anal.
(C8H8N2O4S) C, H, N.
(-)-4-Am in o-2-t h ia b icyco[3.1.0]h exa n e-4,6-d ica r b ox-
yla te ((-)-10). To 30.6 g (134 mmol) of (-)-21 in an autoclave
was added 450 mL (900 mmol) of 2 M aqueous sodium
hydroxide. The solution was heated at 100 °C for 23 h. The
reaction mixture was filtered though Celite, and the resulting
orange/brown solution was evaporated to approximately one-
half the original volume using a rotary evaporator (bath 50
°C). The solution was cooled in an ice/water bath, and 57 mL
of 12 M HCl was added over 5 min. The bath was removed,
and additional 12 N HCl was added to adjust the pH to
approximately 3. The slurry was stirred for 30 min at room
temperature and for 1 h in an ice/water bath. The solid was
collected by filtration and rinsed with cold water. The collected
Su p p or tin g In for m a tion Ava ila ble: X-ray crystallo-
graphic data pertaining to (-)-9 and tablulated combustion
analyses on new compounds. This material is available free
of charge via the Internet at http://pubs.acs.org.
Refer en ces
(1) Thomas, J . T. Excitatory amino acids in health and disease. Prog.
Geriatrics 1995, 43, 1279-1289.
(2) Lipton, S. A.; Rosenberg, P. A. Excitatory amino acids as a final
common pathway for neurologic disorders. N. Engl. J . Med.
1994, 330, 613-622.
(3) Hollmann, M.; Heinemann, S. Cloned glutamate receptors.
Annu. Rev. Neurosci. 1994, 17, 31-108.
(4) Nakanishi, S.; Masu, M. Molecular diversity and functions of
glutamate receptors. Annu. Rev. Biophys. Biomol. Struct. 1994,
23, 319-348.
(5) Schoepp, D. D.; Conn, P. J . Metabotropic glutamate receptors
in brain function and pathology. Trends Pharmacol. Sci. 1993,
14, 13-20.
(6) Bockaert, J .; Pin, J .; Fagni, L. Metabotropic glutamate recep-
tors: an original family of G protein-coupled receptors. Fundam.
Clin. Pharmacol. 1993, 7, 473-485.
(7) Pin, J .-P.; Duvoisin, R. The metabotropic glutamate receptors:
structure and functions. Neuropharmacology 1995, 34, 1-26.
(8) Conn, P. J .; Pin, J .-P. Pharmacology and functions of metabo-
tropic glutamate receptors. Annu. Rev. Pharmacol. Toxicol. 1997,
37, 205-237.

Heterobicyclic mGluR Agonists
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 1039
(9) Ito, I.; Kohda, A.; Tanabe, S.; Hirose, E.; Hayashi, M.; Mitsunaga,
S.; Sugiyama, H. 3,5-Dihydroxyphenylglycine: A potent agonist
of metabotropic glutamate receptors. NeuroReport 1992, 3,
1013-1016.
(10) Birse, E. F.; Eaton, S. A.; J ane, D. E.; J ones, P. L. St. J .; Porter,
R. H. P.; Pook, P. C.-K.; Sunter, D. C.; Udvarhelyi, P. M.;
Wharton, B.; Roberts, P. J .; Salt, T. E.; Watkins, J . C. Phenyl-
glycine derivatives as new pharmacological tools for investigat-
ing the role of metabotropic glutamate receptors in the central
nervous system. Neuroscience 1993, 52, 481-488.
(11) Schoepp, D. D.; Goldsworthy, J .; J ohnson, B. G.; Salhoff, C. R.;
Baker, S. R. 3,5-Dihydroxyphenylglycine is a highly selective
agonist for phosphoinositide-linked metabotropic glutamate
receptors in the rat hippocampus. J . Neurochem. 1994, 63, 769-
772.
The reaction of diazoalkanes with thiophen. J . Chem. Soc.,
Perkin Trans. 1. 1979, 2624-2628.
(25) Brown, H. C.; Vara Prasad, J . V. N.; Zee, S.-H. Hydroboration.
71. Hydroboration of representative heterocyclic olefins with
borane-methyl sulfide, 9-borabicyclo[3.3.1]nonane, dicyclohexyl-
borane, and disiamylborane. Synthesis of heterocyclic alcohols.
J . Org. Chem. 1985, 50, 1582-1589.
(26) Minor amounts (10-20%) of the undesired C4-5′-spirohydan-
toins resulting from cyanide approach from the more hindered
(concave) face of these molecules were also observed in the crude
reaction mixtures. The relative stereochemistry about the spi-
rocyclic ring juncture of the desired rac-C4-5′-spirohydantoins
14 and 20 was determined by the observation of diagnostic NOEs
between the hydantoin 1′-NH proton, H3R, and H6 (below).
(12) Doherty, A. J .; Palmer, M. J .; Henley, J . M.; Collingridge, G. L.;
J ane, D. E. (RS)-2-Chloro-5-hydroxyphenylglycine (CHPG) ac-
tivates mGlu5, but not mGlu1, receptors expressed in CHO cells
and potentiates NMDA responses in the hippocampus. Neuro-
pharmacology 1997, 36, 265-267.
(13) Schoepp, D. D.; J ohnson, B. G.; Monn, J . A. Inhibition of cyclic
AMP formation by a selective metabotropic glutamate receptor
agonist. J . Neurochem. 1992, 58, 1184-1186.
(14) Schoepp, D. D.; J ohnson, B. G.; Salhoff, C. R.; Valli, M. J .; Desai,
M. A.; Burnett, J . P.; Mayne, N. G.; Monn, J . A. Selective
inhibition of forskolin-stimulated cyclic AMP formation in rat
hippocampus by a novel mGluR agonist, 2R,4R-4-amino-pyrrol-
idine-2,4-dicarboxylate. Neuropharmacology 1995, 34, 843-850.
(15) Monn, J . A.; Valli, M. J .; J ohnson, B. G.; Salhoff, C. R.; Wright,
R. A.; Howe, T.; Bond, A.; Lodge, D.; Griffey, K.; Tizzano, J . P.;
Schoepp, D. D. Synthesis of the four isomers of 4-aminopyrrol-
idine-2,4-dicarboxylate (APDC): identification of a potent, highly
selective, and systemically active agonist for metabotropic
glutamate receptors negatively coupled to adenylate cyclase. J .
Med. Chem. 1996, 39, 2990-3000.
(16) Schoepp, D. D.; J ohnson, B. G.; Wright, R. A.; Salhoff, C. R.;
Mayne, N. G.; Wu, S.; Cockerham, S. L.; Burnett, J . P.; Belagaje,
R.; Bleakman, D.; Monn, J . A. LY354740 is a potent and highly
selective group II metabotropic glutamate receptor agonist in
cells expressing human glutamate receptors. Neuropharmacol-
ogy 1997, 36, 1-11.
(17) Monn, J . A.; Valli, M. J .; Massey, S. M.; Wright, R. A.; Salhoff,
C. R.; J ohnson, B. G.; Howe, T.; Alt, C. A.; Rhodes, G. A.; Robey,
R. L.; Griffey, K. R.; Tizzano, J . P.; Kallman, M. J .; Helton, D.
R.; Schoepp, D. D. Design, synthesis and pharmacological
characterization of (+)-2-aminobicyclo[3.1.0]hexane-2,6-dicar-
boxylic Acid (LY354740): a potent, selective and orally active
group 2 metabotropic glutamate receptor agonist possessing
anticonvulsant and anxiolytic properties. J . Med. Chem. 1997,
40, 528-537.
(18) Kristensen, P.; Suzdak, P. D.; Thomsen, C. Expression pattern
and pharmacology of the rat type IV metabotropic glutamate
receptor. Neurosci. Lett. 1993, 155, 159-162.
(19) Tanabe, Y.; Nomura, A.; Masu, M.; Shigemoto, R.; Mizuno, N.;
Nakanishi, S. Signal transduction, pharmacological properties,
and expression patterns of two rat metabotropic glutamate
receptors, mGluR3 and mGluR4. J . Neurosci. 1993, 13, 1372-
1378.
(20) Nakajima, Y.; Iwakabe, H.; Akazawa, C.; Nawa, H.; Shigemoto,
R.; Mizuno, N.; Nakanishi, S. Molecular characterization of a
novel retinal metabotropic glutamate receptor mGluR6 with a
high agonist selectivity for L-2-amino-4-phosphonobutyrate. J .
Biol. Chem. 1993, 268, 11868-11873.
(21) Okamoto, N.; Hori, S.; Akazawa, C.; Hayashi, Y.; Shigemoto, R.;
Mizuno, N.; Nakanishi, S. Molecular characterization of a new
metabotropic glutamate receptor mGluR7 coupled to inhibitory
cyclic AMP signal transduction. J . Biol. Chem. 1994, 269, 1231-
1236.
(27) The enantiomeric purities of (-)-15, (+)-15, (-)-21, and (+)-21
were determined by chiral HPLC (see General Methods, Ex-
perimental Section, for details). Under these conditions, (-)-15
and (+)-15 exhibited elution times of 9.5 and 8.3 min, respec-
tively. (-)-21 and (+)-21 exhibited elution times of 10.5 and 9.4
min, respectively.
(28) The X-ray structure and coordinates for (-)-9 have been filed
with the Cambridge Crystallographic Database. Crystallographic
data for (-)-9 is available as Supporting Information.
(29) Sills, M. A.; Williams, M.; Wilusz, E. J .; Hurt, S. D.; Brundish,
D. E.; Angst C.; Pozza M.; Fagg, G. [3H]CGP 39653: a new
N-methyl-D-aspartate antagonist radioligand with low nano-
molar affinity in rat brain. Eur. J . Pharmacol. 1991, 192, 19-
24.
(30) Nielsen, E. O.; Madsen, U.; Schaumburg, K.; Brehm, L.; Krogs-
gaard-Larsen, P. Studies on receptor-active conformations of
excitatory amino acid agonists and antagonists. Eur. J . Med.
Chem. - Chim. Ther. 1986, 21, 433-437.
(31) Simon, J . R.; Contrera, J . F.; Kuhar, M. J . Binding of [3H]kainic
Acid, an analogue of L-glutamate, to brain membranes. J .
Neurochem. 1976, 26, 141-147.
(32) Ornstein, P. L.; Arnold, M. B.; Bleisch, T. J .; Wright, R. A.;
Wheeler, W. J .; Schoepp, D. D. [3H]LY341495, A highly potent,
selective and novel radioligand for labeling group II metabotropic
glutamate receptors. Bioorg. Med. Chem. Lett. 1998, 8, 1919-
1922.
(33) Kingston, A. E.; Ornstein, P. L.; Wright, R. A.; J ohnson, B. G.;
Mayne, N. G.; Burnett, J . P.; Belagaje, R.; Wu, S.; Schoepp, D.
D. LY341495 is a nanomolar potent and selective antagonist of
group II metabotropic glutamate receptors. Neuropharmacology
1998, 37, 1-12.
(34) J ohnson, B. G.; Wright, R. A.; Arnold, M. B.; Wheeler, W. J .;
Ornstein, P. L.; Schoepp, D. D. [3H]-LY341495 as a Novel
Antagonist Radioligand for Group II Metabotropic Glutamate
(mGlu) Receptors: Characterization of Binding to Membranes
of mGlu Receptor Subtype Expressing Cells. Neuropharmacology
1999, in press.
(35) Cheng, Y. C.; Prusoff, W. H. Relationship between the inhibition
constant (Ki) and the concentration of inhibitor which causes
fifty percent inhibition (IC50) of an enzymatic reaction. Biochem.
Pharmacol. 1973, 22, 3099-3103.
(22) Flor, P. J .; Lukic, S.; Ru¨egg, D.; Leonhardt, T.; Kno¨pfef, T.; Kuhn,
R. Molecular cloning, functional expression and pharmacological
characterization of the human metabotropic glutamate receptor
type 4. Neuropharmacology 1995, 34, 149-155.
(23) Ahmadian, H.; Nielsen, B.; Bra¨uner-Osborne, H.; J ohansen, T.
N.; Stensbøl, T. B.; Sløk, F. A.; Sekiyama, N.; Nakanishi, S.;
Krogsgaard-Larsen, P.; Madsen, U. (S)-Homo-AMPA, a specific
agonist at the mGlu6 subtype of metabotropic glutamic acid
receptors. J . Med. Chem. 1997, 40, 3700-3705.
(24) The carboxycyclopropanation of furan and thiophene with ethyl
diazoacetate has been previously reported: (a) Wenkert, E.; Guo,
M.; Lavilla, R.; Porter, B.; Ramachandran, K.; Sheu, J .-H.
Polyene Synthesis. Ready Construction of Retinol-Carotene
Fragments (()-6-(E)-LTB3 Leukotrienes and Corticrocin. J . Org.
Chem. 1990, 55, 6203-6214. (b) Gillespie, R. J .; Porter, A. E. A.
(36) Wu, S.; Wright, R. A.; Rockey, P. K.; Burgett, S. G.; Arnold, J .
S.; Rosteck, P. R., J r.; J ohnson, B. G.; Schoepp, D. D.; Belagaje,
R. Group III human metabotropic glutamate receptors 4, 7 and
8: Molecular cloning, functional expression, and comparison of
pharmacological properties in RGT cells. Mol. Brain Res. 1998,
53, 88-97.
(37) Tizzano, J . P.; Griffey, K. I.; J ohnson, J . A.; Fix, A. S.; Helton,
D. R.; Schoepp, D. D. Intracerebral 1S,3R-1-aminocyclopentane-
1,3-dicarboxylic Acid (1S,3R-ACPD) produces limbic seizures
that are not blocked by ionotropic glutamate receptor antago-
nists. Neurosci. Lett. 1993, 162, 12-16.
(38) A constrained glutamate analogue possessing a similar glutamate
conformation to that of 5, but occupying a slightly different steric
space (ABHxD-I), has been reported to display significantly lower
group II mGlu receptor agonist potency and selectivity as

1040 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Monn et al.
compared to 5: Kozikowski, A. P.; Steensma, D.; Luca-Araldi,
G.; Tu¨ckmantel, W.; Wang, S.; Pshenichkin, S.; Surina, E.;
Wroblewski, J . T. Synthesis and biology of the conformationally
restricted ACPD analogue, 2-aminobicyclo[2.1.1]hexane-2,5-di-
carboxylic acid-I, a potent mGluR agonist. J . Med. Chem. 1998,
41, 1641-1650.
high agonist selectivity for L-2-amino-4-phosphonobutyrate. J .
Biol. Sci. 1993, 268, 11868-11873.
(40) Laurie, D. J .; Schoeffter, P.; Wiederhold, K. H.; Sommer, B.
Cloning, distribution and functional expression of the human
mGlu6 metabotropic glutamate receptor. Neuropharmacology
1997, 36, 145-152.
(39) Naajima, Y.; Iwakabe, I.; Akazawa, C.; Nawa, H.; Shigemoto,
R.; Mizuno, N.; Nakanishi, S. Molecular characterization of a
novel retinal metabotropic glutamate receptor mGluR6 with a
J M980616N