A new metabotropic glutamate receptor agonist with in vivo anti-allodynic activity

Article abstract of DOI:10.1016/j.bmc.2010.06.051

As part of the vital search towards improved therapeutic agents for the treatment of neuropathic pain, the central nervous system glutamate receptors have become a major focus of research. Outlined herein are the syntheses of two new biologically active 3′-cycloalkyl-substituted carboxycyclopropylglycines, utilizing novel synthetic chemistry. The reaction between substituted 1,2-dioxines and an aminophosphonate furnished the cyclopropane core in a single step with all required stereochemistry of pendant groups. In vitro binding assays at metabotropic glutamate receptors revealed selective activity. In vivo testing in a rodent model of neuropathic pain indicated one amino acid significantly and dose-dependently decreased mechanical allodynia.

Full text of DOI:10.1016/j.bmc.2010.06.051

Bioorganic & Medicinal Chemistry 18 (2010) 6089–6098
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
A new metabotropic glutamate receptor agonist with in vivo
anti-allodynic activity
Nathan J. Stanley a,b, Mark R. Hutchinson b, Trine Kvist c, Birgitte Nielsen c, Jesper Mosolff Mathiesen c,
Hans Bräuner-Osborne c, Thomas D. Avery a, Edward R. T. Tiekink d, Daniel Sejer Pedersen c,
Rodney J. Irvine b, Andrew D. Abell a, Dennis K. Taylor e,
*
a Discipline of Chemistry, The University of Adelaide, Adelaide 5005, Australia
b Discipline of Pharmacology, The University of Adelaide, Adelaide 5005, Australia
c Department of Medicinal Chemistry, University of Copenhagen, 2100 Copenhagen, Denmark
d Department of Chemistry, University of Malaya, 50603 Kuala Lumpur, Malaysia
e School of Agriculture, Food and Wine, The University of Adelaide, Glen Osmond 5005, Australia
a r t i c l e i n f o
a b s t r a c t
Article history:
As part of the vital search towards improved therapeutic agents for the treatment of neuropathic pain, the
central nervous system glutamate receptors have become a major focus of research. Outlined herein are
the syntheses of two new biologically active 30-cycloalkyl-substituted carboxycyclopropylglycines, utiliz-
ing novel synthetic chemistry. The reaction between substituted 1,2-dioxines and an aminophosphonate
furnished the cyclopropane core in a single step with all required stereochemistry of pendant groups. In
vitro binding assays at metabotropic glutamate receptors revealed selective activity. In vivo testing in a
rodent model of neuropathic pain indicated one amino acid significantly and dose-dependently
decreased mechanical allodynia.
Received 6 May 2010
Revised 16 June 2010
Accepted 17 June 2010
Available online 25 June 2010
Keywords:
1,2-Dioxines
Carboxycyclopropylglycines
Metabotropic glutamate receptors
In vivo
Ó 2010 Elsevier Ltd. All rights reserved.
Neuropathic pain
Allodynia
1. Introduction
sequence homology, signal transduction mechanisms and ago-
nist/antagonist interactions.4 It has previously been shown that
Chronic neuropathic pain affects a significant proportion of the
population worldwide and decreases quality of life. Sufferers expe-
rience increased sensations of pain, known as hyperalgesia as well
as allodynia where painful sensations arise from innocuous touch
or pressure. Many of the current medications used in its treatment
do not give adequate and effective pain relief in all cases1,2 thus,
much research has been focused on finding alternatives. The excit-
drugs targeting specific subtypes of the metabotropic glutamate
receptors show efficacy in various pain models including those
where allodynia and hyperalgesia are present.5,6 Over the past 10
years, many new mGluR ligands have been synthesized both as
experimental probes and potential therapeutic agents.7 One potent
class of competitive mGluR2/3 ligands, the carboxycyclopropylgly-
cines (CCGs, Fig. 1) show potential for development as a new range
of therapeutic agents.8,9 These analogues of the neurotransmitter
atory neurotransmitter L-glutamate (1), plays a significant role in
the modulation of pain signalling.3 There are two main classes of
glutamate receptors, the ionotropic and the metabotropic. The
ionotropic glutamate receptors (iGluRs) are ligand-gated ion chan-
O
O
H2N
H2N
H
2
nels divided into three broad types: N-methyl-D-aspartate (NMDA),
OH
OH
a-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)
3'
and 2-carboxy-3-carboxymethyl-4-isopropenylpyrrolidine (kai-
nate). In contrast, the metabotropic glutamate receptors (mGluRs)
are G-protein coupled receptors (GPCRs), having eight known sub-
types, mGluR1–mGluR8, divided into three groups based on
HO
HO
R
O
O
2
2a
R = H, F, Me ( ), CH2OH ( ),
alkyl, aryl
1
L-glutamate
(2S)-carboxycyclopropylglycines
* Corresponding author. Tel.: +61 8 8303 7239; fax: +61 8 8303 7116.
E-mail address: dennis.taylor@adelaide.edu.au (D.K. Taylor).
Figure 1. Structural comparison between glutamate and carboxycyclopropylgly-
cines.
0968-0896/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2010.06.051

6090
N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
glutamate (1) are conformationally constrained by a constituent
cyclopropane ring that contains an additional substituent (R) that
can influence receptor specificity and selectivity of one mGluR sub-
type over another.
Recent advances in synthetic chemistry have provided a general
route for synthesis of 30-substituted CCGs, allowing exploration of
their therapeutic potential.10 Herein, we report the synthesis and
biological activity of two new cycloalkyl-substituted CCGs (( )-3
and ( )-4, Fig. 2), one of which (( )-3) shows anti-allodynic activity
in a rodent model of neuropathic pain. Based on known structure–
activity relationships of previously tested CCGs, it was hypothe-
sized that the activity of these compounds may switch from
agonist to antagonist with increasing cycloalkyl ring size.
c
-hydroxyenone formed by base catalysed ring-opening of a 1,2-
dioxine (Scheme 1). Michael addition of the phosphonate nucleo-
phile to the -hydroxyenone and intramolecular ring-closure
c
formed the protected cyclopropane amino acids as a mixture of
diastereoisomers, that were separated by flash chromatography.10
This synthetic route is significantly shorter than other reported
methods with the added advantage that the 30-substitution can
be altered simply by preparing the appropriate 1,2-dioxine.8,11
The starting 1,3-butadienes (5a, 5b) were prepared in excellent
yields, 90% and 97%, respectively, by Wittig reaction of cycloalkyl-
carboxaldehydes with the ylide derived from cinnamyl triphenyl-
phosphonium chloride on reaction with potassium tert-butoxide.
Photolysis (dye sensitised photo-oxidation) of the 1,3-butadienes
employing rose bengal bis(triethylammonium) salt in the presence
of oxygen, afforded the desired 3-cycloalkyl-6-phenyl-3,6-dihy-
dro-1,2-dioxines (6a, 6b) in good yields of 69% and 59%, respec-
tively. The protected cyclopropane amino acids were obtained by
2. Results and discussion
2.1. Synthesis
reaction with ( )-Cbz-a-phosphonoglycine trimethylester under
basic conditions followed by separation of the diastereoisomers
by flash chromatography to provide the pure isomers (7a, 7b).
The relative stereochemistry of the cyclopropane substituent could
not be determined by 1H NMR or 2D NMR due to overlap of the
cyclopropyl and cycloalkyl signals and since identical ROESY inter-
actions were observed for both diastereoisomers. However, the rel-
ative stereochemistry of cyclopropane 7a ( Fig. 3) and 7b was
determined by single crystal X-ray crystallography and shown to
match that for the previously reported pharmacologically active
CCGs.8,9
The key step in the synthesis of the desired CCG analogues ( )-3
and ( )-4 involved reaction of an aminophosphonate and a cis
O
3'
O
3'
H2N
H
H2N
H
OH
OH
HO
HO
O
O
A Baeyer–Villiger oxidation of 7b was performed using trifluor-
operacetic acid produced in situ from trifluoroacetic anhydride and
hydrogen peroxide.34 This gave the desired phenyl ester, 9b in 73%
( )-3
( )-4
Figure 2. Two new carboxycyclopropylglycines.
Scheme 1. Reagents and conditions: (a) PhCH2CH@CHPPh3Cl, KOBut, rt, Et2O; (b) O2, h
m, rose bengal, 0 °C, DCM; (c) (i) ( )-Cbz-a-phosphonoglycine trimethylester, LDA mono
THF complex in cyclohexane, ꢀ78 °C to rt, THF; (ii) chromatographic separation of diastereoisomers; (d) for 9a: m-CPBA, rt, CHCl3; for 9b: TFAA, H2O2, 0 °C, DCM; (e) MeOH,
H2SO4, reflux; (f) LiOH(aq), rt, THF; (g) H2, Pd/C (5 wt %), MeOH.

N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
6091
M)
Table 1
Metabotropic glutamate receptor binding data
Compound
mGluR1 (
l
M) mGluR2 (lM)
mGluR4 (
l
( )-2 (R = Me)
( )-3 (R = c-C3H5)
( )-4 (R = c-C6H11
>100
>100
>100
0.01a
>100
0.05 [7.395 0.152]a,b >100
)
62c
>100
Plain and bold text refers to agonist (EC50
)
and antagonist (IC50) potencies,
respectively.
a
n = 4.
b
Values in brackets indicate pEC50 SEM.
n = 3.
c
down regulation of adenylyl cyclase and a consequent decrease
in cAMP levels. Antagonist activity was measured as an increase
in cAMP levels in the presence of the endogenous agonist, gluta-
mate. The methyl-substituted cyclopropyl amino acid ( )-2 was
used as a positive control and had a measured EC50 value of
0.01 lM at mGluR2, which is comparable to that reported previ-
ously by Collado et al.11 Test compound ( )-3 showed agonist
activity at mGluR2 which was approximately fivefold lower than
Figure 3. Molecular structure and crystallographic numbering scheme for com-
pound 7a (C24H25NO5). Diagram drawn with 50% displacement ellipsoids.
the positive control, with a measured EC50 of 0.05 lM. Overall,
the in vitro data indicates that compound ( )-3 is potent and selec-
tive for mGluR2 over mGluR1 and mGluR4, as well as having no
activity at the ionotropic receptors. Thus cyclopropyl amino acid
( )-3 is a good candidate for further investigation in vivo. Finally,
test compound ( )-4 showed weak antagonist activity at mGluR2
yield. The use of this methodology on 7a gave rise to some decom-
position and only gave a low yield of 9a (24%). However, the use of
m-CPBA produced 9a in good yield (77%). Formation of the phenyl
ester was confirmed by 1H NMR with an absence of aromatic and
cyclopropane resonances associated with the protons adjacent to
the ketone moiety and the appearance of phenyl ester resonance
at ca. 7.1 ppm. After conversion to the ester, a shift in cyclopropyl
carbonyl resonances was also observed in the 13C NMR from
approximately 198 ppm to ca. 172 ppm. Direct hydrolysis of com-
pounds 9a and 9b to give carboxylic acids 11a and 11b was com-
plicated by the release of phenol, making purification of the
desired product very difficult. To circumvent this problem, 9a
and 9b were first transesterified to give 10a and 10b in excellent
(96%) and good (76%) yields, respectively, after purification by flash
column chromatography. Ester hydrolysis gave almost quantitative
yields of 11a and 11b (92% and 97%, respectively).
The final step in the synthesis involved removal of the Cbz pro-
tecting group of 11a and 11b by catalytic hydrogenation, to give
the desired racemic amino acids 3 and 4. These were purified by
conversion to the corresponding hydrochloride salts by addition
of concentrated HCl and removal of excess acid in vacuo. The de-
sired free amino acids ( )-3 and ( )-4 were re-isolated by stirring
with anhydrous propylene oxide followed by isolation of the pre-
cipitate by filtration.11 It should be noted that the very low yield
(16%) of ( )-4 was thought to result from decomposition whilst
converting to the hydrochloride salt. By comparison, compound
( )-3 appeared more stable to the acidic conditions.
with an IC50 of 62 lM. The in vitro assays confirm the initial
hypothesis where it was expected that cyclopropyl-substitution,
as in compound ( )-3 would create an agonist, whereas cyclohexyl
substitution, as in compound ( )-4, would create an antagonist.
Compound ( )-3 is structurally quite similar to the positive control,
however the cyclopropyl-substitution is slightly more bulky than a
simple methyl group, which may account for the loss of potency. It
has been shown that if the methyl group is exchanged with an
ethyl group, then binding activity decreases 20-fold. However, sub-
stitution with a CH2OH group (Fig. 1, 2a) in place of the ethyl (Et)
group increases agonist activity by more than 30-fold relative to an
Et group. This suggests the region of the receptor where the cyclo-
propyl substituent lies is a polar region and so the presence of a
hydrophobic cyclopropyl group creates a less favourable interac-
tion. On the other hand, compound ( )-4 is substituted with a com-
paratively bulky cyclohexyl group which seems to convey
antagonist activity. Substitution by other bulky, hydrophobic
groups, such as phenyl and xanthenyl has been reported previously
to consistently produce receptor antagonists, thus this result is
consistent with steric size being important for determining agonist
or antagonist activity at this position.12,13 The new compounds
were only tested at three of the eight subtypes of the metabotropic
glutamate receptors; mGluR1, mGluR2 and mGluR4, representing
the three groups of receptors. It is possible that these compounds
may have also shown some activity at the five receptor subtypes
which were not investigated.
3. Pharmacology
3.1. Receptor binding assays
3.2. In vivo neuropathic pain model
The in vitro drug screening at native rodent NMDA, kainate and
AMPA ionotropic glutamate receptor subtypes showed a lack of
activity for compounds ( )-3 and ( )-4 at any of these receptor
subtypes as indicated by IC50 >100 lM. Compound ( )-2 was used
as a negative control as it has not previously been reported to show
The cyclopropane amino acid ( )-3 showed potent and selective
activity in the in vitro binding assays. Based on this, ( )-3 was
tested in vivo in a neuropathic pain model, the chronic constriction
injury of the sciatic nerve in Sprague-Dawley rats. Von Frey fila-
ments were employed in order to test for neuropathic pain, as ob-
served through the presence of mechanical allodynia. During this
testing procedure, fine filaments of increasing diameter and bend-
ing force are touched to the underside of the hind paw until the
animal shows a withdrawal response. Illustrated in Figure 4, prior
to surgery, all rats responded on the von Frey test at approximately
9.0 g, both ipsilaterally and contralaterally as indicated at baseline
activity at the ionotropic receptors.
Referring to Table 1, the data obtained from functional assays at
the metabotropic glutamate receptors indicates that both new
compounds exhibited functional binding activity at mGluR2. The
second messenger assays to measure mGluR2 activity involved
the radioimmunoassay detection of forskolin-stimulated cAMP
concentrations. The binding of agonists to the receptor causes a

6092
N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
(BL). Following chronic constriction injury of the sciatic nerve, sub-
sequent von Frey testing on days 3, 10 and 14 revealed a decrease
in response threshold to the minimum 0.4 g bilaterally, indicating
that all rats were allodynic. On the day of drug dosing, a pre-drug
baseline (PDBl) measurement was also made to ensure the reliabil-
ity of inter-day test results. After 1 h an intrathecal dose of (2R,4R)-
APDC (500 nmols), von Frey response thresholds were significantly
increased bilaterally, compared to animals receiving vehicle. After
3 h intrathecal drug dosing, response thresholds were similar to
the 1 h post-drug time point. After 1 h an equimolar intrathecal
dose of compound ( )-3, von Frey response thresholds of drug-
treated animals were significantly increased bilaterally, compared
to animals receiving vehicle. After 3 h drug dosing, von Frey re-
sponse thresholds of drug-treated animals were significantly in-
creased bilaterally, compared to animals receiving vehicle. After
24 h, all animals had returned to pre-drug allodynia bilaterally,
with positive control and drug-treated animals no longer signifi-
cantly different from vehicle-treated controls. No other behav-
ioural effects of either drug were noted.
Mechanical Thresholds
15
10
5
***
**
**
0
Treatment
In order to investigate any possible dose dependency of the
anti-allodynic effect of ( )-3, two lower doses, 50 nmols and
250 nmols, were also administered. For clarity, only the data relat-
ing to the ipsilateral 3 h time point is shown in Figure 5. The data
strongly indicates a dose dependency of ( )-3 with 50 nmols caus-
ing an increase in response threshold to 5.0 g, 250 nmols causing a
significant increase to 7.5 g and 500 nmols causing a significant in-
crease in response threshold to 10.0 g compared to vehicle.
This work was carried out to ascertain the in vivo effects of the
novel, selective Group II metabotropic glutamate receptor (mGluR)
agonist ( )-3 in a rodent model of neuropathic pain. The model em-
ployed in this study was chronic constriction of the sciatic nerve,
where the nerve bundle undergoes loose ligation with chromic
gut at the mid-thigh level. The outcome is a neuropathic pain state
that develops over a period of two weeks and results in reliably
measurable allodynia, resembling that reported and measured in
human neuropathic pain states. Although the surgery manipulates
the sciatic nerve on only one side of the animal, allodynia develops
on both the operated side (ipsilateral) and the non-operated side
(contralateral) as can be seen in Figure 4.14 This phenomena is
known as mirror-image allodynia and there is mounting evidence
to suggest that it is caused by glial activation and the associated
action of pro-inflammatory cytokines.14 Due to a lack of test com-
pound and the lack of pharmacokinetic data, it was opted to
administer the drugs via the intrathecal route, by lumbar puncture
between the L5 and L6 vertebrae. Drugs that act at the spinal level
have the advantage of avoiding side effects possibly encountered
when high systemic doses are required to achieve the needed
Figure 5. Comparison of the changes in ipsilateral mechanical threshold 3 h after
intrathecal administration of (2R,4R)-APDC (500 nmols), ( )-3 (50, 250 and
**
P
***
P
500 nmols) or vehicle. Data expressed in grams SEM, n = 6.
<0.01,
<0.001 compared to vehicle.
CNS levels. Intrathecal administration also confines drug action
to the spinal cord and thus limits the possible mechanisms respon-
sible for drug effects, thus simplifying interpretation of the data
obtained. Although only in vitro receptor binding data relating to
mGluR2 receptors was obtained, it is reasonable to assert that
compound ( )-3 would also activate mGluR3 receptors, given that
all known ligands in this compound class are approximately equi-
potent at mGluR2 and R3.8,9
It is widely reported that agonists at these receptors can signif-
icantly decrease the behavioural signs associated with neuropathic
pain, including thermal and mechanical hyperalgesia and mechan-
ical allodynia.15–17 Several in situ hybridization studies have iden-
tified mRNA encoding for the mGluR3 receptor in the spinal dorsal
horn, however, levels of mRNA encoding for the mGluR2 receptor
were found to be very low.18,19 Jia et al. used light and electron
microscopic immunocytochemistry to show the presence of
mGluR2/3 in the inner part of lamina II and also the presence of
these receptors pre-synaptically on GABAergic neurons within
the dorsal horn.20 These results were confirmed by Azkue and col-
leagues using an immunogold technique.21 Finally, both Gerber et
al. and Carlton et al. have provided evidence that mGluR2 and R3
receptors were also located pre-synaptically on glutamatergic pri-
mary afferent neurons within the dorsal horn.18,22
Ipsilateral
Contralateral
( )-3
( )-3
10
8
6
4
10
APDC (positive control)
APDC (positive control)
Vehicle
8
Vehicle
***
***
6
4
**
**
**
2
1
2
1
*** ***
**
0.6
0.4
0.6
0.4
D3 D10 D14 PDBl 1 hr 3 hr 24 hr
D3 D10 D14 PDBl 1 hr 3 hr 24 hr
BL
Timecourse (Hours post injection)
BL
Timecourse (Hours post injection)
Figure 4. Effects of an acute dose of ( )-3 (500 nmols i.t.) and a positive control, APDC (500 nmols i.t.) on mechanical allodynia following chronic constriction injury of the
** ***
sciatic nerve in Sprague-Dawley rats. Data are expressed as mean SEM, n = 6. BL = baseline, PDBl = pre-drug baseline. P <0.01,
P <0.001 compared to vehicle.

N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
6093
The known selective Group II mGluR agonist, (2R,4R)-APDC
(aminopyrrolidine-2,4-dicarboxylate), used here as a positive con-
trol, has previously been shown by Fisher et al. to prevent the
development of mechanical allodynia, following loose ligation of
the sciatic nerve chronic constriction injury, and using a repeated
intrathecal dosing regime.16 Similarly, we observed that this com-
pound is capable of producing a significant reversal of mechanical
allodynia with the greatest effect apparent at the 1 h time point.
There was no possibility to use receptor antagonism to verify that
the anti-allodynic effects observed were indeed due to Group II
mGluR activation, as receptor blockade would be pro-nociceptive,
which was undesirable.23 The test compound ( )-3 was found to
have greater in vivo anti-allodynic activity than the positive con-
trol, (2R,4R)-APDC. This result was expected given the EC50 value
of amino acid ( )-3 at mGluR2 is eight times higher, indicating
the higher in vitro potency of ( )-3.7 Interestingly, the effects on
withdrawal threshold were maximal for test drug ( )-3 when mea-
sured at the 3 h time point, compared to 1 h post drug dosing for
the positive control, (2R,4R)-APDC. This observation is difficult to
explain, apart from potential differences in the polarizability and
lipophilicity of the two compounds, which may have affected their
distribution and absorption.
glutamate agonist, LY354740, is able to reverse opioid tolerance
in vivo; an effect which is quite feasibly due to the drug’s action
at glial mGlu receptors.32
In the present study, it is possible that the selective mGluR2/3
receptor agonist, ( )-3, is acting to reverse mechanical allodynia
by interactions with both neurons and glia. The pharmacological
action of compound ( )-3 on metabotropic glutamate receptors 2
and 3 expressed both on glutamatergic and GABAergic neurons
and on glial cells within the dorsal horn of the spinal cord, most
likely mediated the reversal of mechanical allodynia observed in
this study.
4. Conclusions
A concise synthesis of a new, potent and selective Group II metab-
otropicglutamate receptoragonist is reported which exhibits in vivo
anti-allodynic efficacy in an animal model of neuropathic pain.
Overall, the therapeutic potential of targeting metabotropic gluta-
mate receptors for treatment of neuropathic pain is exemplified
and the need for further research into this class of compounds
highlighted.
Endogenous glutamate within the dorsal horn is known to act
pro-nociceptively on glutamatergic neurons which project to the
brain stem and higher brain centres.24 Hence, pre-synaptic modu-
lation of these projecting neurons is capable of attenuating pain
signaling. Since mGluR2 and R3 have been identified on glutama-
tergic primary afferent neurons, it is reasonable to assume that
mGluR2/3 ligands could modulate signaling in these projecting
neurons, thus resulting in anti-nociception.18,22 It is well known
that there exists descending inhibitory pathways which serve to
dampen and attenuate pain signaling at the spinal dorsal horn le-
vel.25 Research by Zhou et al. has highlighted the action of endog-
enous glutamate, acting by stimulation of Group II mGluRs in the
spinal dorsal horn, in causing a decrease in GABAergic neuron
activity and a subsequent decrease in GABA inhibition due to noci-
ceptive input.26 Their data suggests that GABA inhibition in the
dorsal horn is anti-nociceptive and hence is evidence for disinhibi-
tion of descending anti-nociceptive neurons which would result in
analgesia. Metabotropic glutamate receptors 2 and 3 have been
identified both on glutamatergic primary afferent neurons and
GABAergic axon terminals. Binding to excitatory glutamatergic
neurons would result in decreased ascending nerve transmission
via ascending projecting neurons, whereas binding to inhibitory
GABAergic neurons likely causes disinhibition of descending inhib-
itory neurons. Thus, it is suggested that these tandem mechanisms
could partly be responsible for the observed reversal of mechanical
allodynia.
5. Experimental section
5.1. Chemistry
Diethyl ether and dichloromethane were dried over 4 Å sieves.
Methanol was dried over 3 Å sieves. THF was dried over sodium
wire with benzophenone as indicator and distilled just prior to
use. All organic extracts were dried over anhydrous magnesium
sulfate. Thin layer chromatography was carried out using alumin-
ium sheets coated with Silica Gel 60 F254 (40 ꢁ 80 mm) and visual-
ized under 254 nm light, or developed in vanillin dip. Flash
chromatography was accomplished using Silica Gel 60 (230–400
mesh). All yields reported refer to isolated material judged to be
>95% homogenous by TLC and NMR spectroscopy.
1H and 13C NMR spectra were obtained using either a 300 MHz
or 600 MHz instrument. NMR spectra were recorded in CDCl3 solu-
tion using TMS (0 ppm) and CDCl3 (77.0 ppm) as internal standards
and D2O (4.87 ppm) using t-butanol as external zero for 13C NMR.
NMR spectra collected in CD3OD were calibrated to CD3OD (3.31
and 49.0 ppm). Infrared spectra were recorded between solid
plates (NaCl) as a nujol mull or thin film. Melting points are uncor-
rected. Electron impact (EI) mass spectra were recorded using a
mass spectrometer operating at 70 eV. Accurate mass measure-
ments were performed at the School of Chemistry, Monash Univer-
sity, Victoria, Australia using electrospray ionisation (ESI).
Elemental analysis was conducted in the Department of Chemistry,
University of Otago, Dunedin, New Zealand. X-ray crystallography
was performed by Edward R. T. Tiekink at The University of Texas,
San Antonio, USA. The following compounds were purchased and
used without further purification: rose bengal bis(triethylammo-
nium)salt, trans cinnamyl chloride, cyclopropanecarboxaldehyde,
Glial cells are also known to be present in the spinal dorsal horn
and accumulating evidence points to the presence of mGluR3
receptors on glia.20,21,27 Much research has demonstrated the
important contribution of glial cell activation and consequent
changes to glial and neuronal signaling to the development and
maintenance of neuropathic pain.14 Glial cell activation leads to in-
creased release of pro-inflammatory cytokines such as tumour
cyclohexanecarboxaldehyde and ( )-Cbz-
methyl ester.
a-phosphonoglycine tri-
necrosis factor alpha (TNFa), interleukin 1 beta (IL-1b) and inter-
leukin 6 (IL-6).28 It has been shown that levels of pro-inflammatory
cytokines are increased both centrally and spinally due to chronic
constriction injury of the sciatic nerve.29 These cytokines cause
inflammation responses that result in changes to nerve signaling
function which contributes to the allodynia and hyperalgesia
which is evident in the neuropathic pain state.30 The activation
of Group II metabotropic glutamate receptors has been demon-
strated to modulate glial cell activation and play a role in altera-
tions to levels of pro-inflammatory cytokines.31 Work by Popik et
al. has revealed that the potent, selective Group II metabotropic
5.2. General procedure for the preparation of 1-cycloalkyl-4-
phenyl-1,3-butadienes
Cinnamyl triphenylphosphonium chloride (36 mmol) was
added to anhydrous diethyl ether or anhydrous THF (150 mL).
Whilst the slurry was stirred under N2, potassium tert-butoxide
(39 mmol) was added in one portion. After stirring for 15 min,
the cycloalkanecarboxaldehyde (30 mmol), dissolved in anhydrous
ether (50 mL), was added dropwise over approximately 20 min.
The reaction was stirred overnight at which point the excess base

6094
N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
was quenched with half-saturated NH4Cl(aq) and the volatiles re-
moved in vacuo. The residue was diluted with water (100 mL)
and extracted with dichloromethane (3 ꢁ 100 mL). After drying
over anhydrous MgSO4(s), filtration and removal of the solvent in
vacuo, the residue was taken up in hexanes and filtered through
a pad of silica (4 cm ꢁ 5 cm ID) to remove all triphenylphosphine
oxide. Removal of the hexanes in vacuo afforded the desired crude
1,3-butadienes.
(25), 105 (12), 79 (10); HRMS (ESI, [M+4H]+) calcd for C16H24O2:
248.1776, found 248.1647.
5.4. General cyclopropanation procedure
( )-Cbz-a-phosphonoglycine trimethyl ester (15 mmol) was
dissolved in anhydrous, freshly distilled THF, under N2. The solu-
tion was cooled to ca. ꢀ78 °C and freshly prepared, titrated LDA
mono-tetrahydrofuran complex in cyclohexane (13 mmol) was
added dropwise. After ca. 30 min, the 1,2-dioxine (14 mmol) was
added. After 15 min the reaction was warmed to ca. ꢀ15 to
ꢀ10 °C. The reaction was maintained at this temperature for >4 h
and then allowed to warm to rt overnight. The reaction was
quenched with half-saturated NH4Cl(aq), extracted with diethyl
ether, dried over MgSO4(s), filtered and the solvents removed in va-
cuo to yield a 1:1 mixture of (2RS)-diastereoisomers. Flash column
chromatography (20% ethyl acetate in hexanes, v/v) afforded the
desired pure diastereoisomers.
5.2.1. [(1E,3E)-4-Cyclopropylbuta-1,3-dien-1-yl]benzene (5a)
Colourless oil (17.0 g, 90%); Rf: 0.79 (35% dichloromethane in
hexanes, v/v); 1H NMR (300 MHz, CDCl3) d 0.43–0.48 (m, 2H),
0.77–0.90 (m, 2H), 1.42–1.54 (m, 1H), 1.83–1.93 (m, 1H), 4.85–
4.92 (m, 1H), 5.27–5.39 (m, 2H), 6.07–6.15 (m, 1H), 6.25–6.55
(m, 3H), 6.68–6.76 (m, 1H), 7.14–7.44 (m, 10H). The crude product
was used without further purification.
5.2.2. [(1E,3E)-4-Cyclohexylbuta-1,3-dien-1-yl]benzene (5b)
Colourless oil (22.0 g, 97%); Rf 0.81 (35% dichloromethane in
hexanes, v/v). All other physical and chemical properties were
identical to those previously reported.33 The crude product was
used without further purification.
5.4.1. ( ) Methyl (2S)-[(1R,2S,3S)-3-benzoyl-1,10-bi(cyclopropyl)-
2-yl]{[(benzyloxy)carbonyl]amino}ethanoate (7a)
Colourless needles (500 mg, 27%); mp: 84–85 °C; Rf 0.25 (20%
ethyl acetate in hexanes, v/v); IR (nujol): 3384, 3073, 2924, 2727,
1740, 1704, 1662, 1597, 1580, 1520, 1456, 1377, 1294, 1284,
5.3. General procedure for the preparation of 1,2-dioxines
1230 cmꢀ1 1H NMR (600 MHz, CDCl3) d 0.28–0.31 (m, 1H), 0.39–
;
1,2-Dioxines were prepared by the rose bengal bis(triethylam-
monium)salt sensitized [4p+2p] cycloaddition of singlet oxygen
0.42 (m, 1H), 0.61–0.68 (m, 2H), 1.01–1.06 (m, 1H), 1.67 (ddd,
J = 6.0, 3.6, 1.8 Hz, 1H), 2.02 (ddd, J = 6.0, 4.8, 3.0 Hz, 1H), 2.88
(dd, J = 4.8, 4.8 Hz, 1H), 3.82 (s, 3H), 4.51 (dd, J = 9.0, 1.8 Hz, 1H),
5.10 (dd, J = 12.6, 12.6 Hz, 1H), 5.57 (d, J = 8.4 Hz, 1H), 7.26–7.30
(m, 4H), 7.34–7.37 (m, 1H), 7.42–7.45 (m, 2H), 7.54–7.57 (m,
1H), 7.96–7.97 (m, 2H); 13C NMR (150 MHz, CDCl3) d 4.8, 8.8,
28.1, 34.2, 34.3, 52.7, 53.0, 67.1, 127.9, 128.2, 128.5, 128.6, 133.0,
136.1, 137.4, 156.1, 171.6, 198.3; MS (EI) (C24H25O5N) m/z (%):
(M+ 407, <1%), 240 (6), 185 (100), 167 (6), 157 (6), 144 (12), 129
(8), 115 (9), 105 (100), 91 (9), 77 (70), 59 (4), 51 (8); HRMS (ESI,
[M+H]+) calcd for C24H26O5N: 408.1811, found 408.1813; Anal.
Calcd for C24H25O5N: C, 70.74; H, 6.18; N, 3.44. Found: C, 70.74;
H, 6.13; N, 3.48.
onto the corresponding 1,3-butadiene. The 1,3-butadiene (3.0 g,
15 mmol) and rose bengal bis(triethylammonium)salt (100 mg,
0.09 mmol) were dissolved in dichloromethane (100 mL) and the
reaction vessel semi-immersed in an ice bath so that the reaction
mixture was maintained at a temperature of ca. 5–10 °C. A stream
of oxygen was then passed through the solution, whilst irradiating
with two to three tungsten halogen lamps (500 W) at a distance of
10 cm from the reaction vessel for 6–12 h. Upon consumption of
the diene, as indicated by TLC (35% dichloromethane in hexanes,
v/v), the volatiles were then removed in vacuo and the residue sub-
jected to flash column chromatography (10% ethyl acetate in hex-
anes, v/v), furnishing the pure 1,2-dioxine.
5.4.2. ( ) Methyl (2S)-[(1S,2S,3R)-2-benzoyl-3-cyclohexyl-
cyclopropyl]{[(benzyloxy)carbonyl]amino}ethanoate (7b)
White solid (642 mg, 19%); mp: 150–151 °C; Rf 0.62 (30% ethyl
acetate in hexanes, v/v); IR (nujol): 3387, 2926, 2727, 2669, 1737,
5.3.1. ( ) (3S,6S)-3-Cyclopropyl-6-phenyl-3,6-dihydro-1,2-
dioxine (6a)
Recrystallization from hot hexanes gave compound 6a as col-
ourless needles (15.4 g, 69%); mp: 54–55 °C; Rf 0.49 (10% ethyl ace-
tate in hexanes, v/v); IR (nujol): 2922, 2852, 2725, 2672, 1715,
1708, 1659, 1595, 1580, 1519, 1455, 1377, 1292 cmꢀ1 1H NMR
;
(600 MHz, CDCl3) d 1.08–2.13 (m, 13H), 2.74–2.76 (m, 1H), 3.79 (s,
3H), 4.29–4.32 (m, 1H), 5.07 (d, J = 12.0 Hz, 1H), 5.12 (d, J = 12.0 Hz,
1H), 5.48 (d, J = 8.4 Hz, 1H), 7.26–7.57 (m, 8H), 7.98–7.99 (m, 2H);
13C NMR (150 MHz, CDCl3) d 25.8, 26.1, 26.2, 28.0, 32.9, 33.6, 34.3,
37.0, 39.0, 52.6, 52.9, 67.1, 128.0, 128.1, 128.2, 128.5, 128.6, 133.0,
136.0, 137.3, 156.0, 171.4, 198.6; MS (EI) (C27H29O5N) m/z (%): (M+
450, <1%), 227 (100), 209 (20), 145 (30), 105 (90), 91 (30); HRMS
(ESI, [M+NH4]+) calcd for C27H35O5N2: 467.2546, found 467.2548.
1455, 1378, 1305, 1256, 1190, 1160 cmꢀ1 1H NMR (300 MHz,
;
CDCl3) d 0.31–0.67 (m, 4H), 1.02–1.13 (m, 1H), 3.87–3.92 (m,
1H), 5.48 (dd, J = 1.8, 2.1 Hz, 1H), 6.09–6.20 (m, 2H), 7.30–7.43
(m, 5H); 13C NMR (75 MHz, CDCl3) d 1.7, 3.4, 13.2, 79.9, 82.7,
137.8; MS (EI) (C13H14O2) m/z (%): (M+ 202, <1%), 173 (10), 160
(20), 133 (10), 118 (20), 105 (100), 97 (5), 77 (40), 69 (35), 51
(10); HRMS (ESI, [M+Na]+) calcd for C13H14O2Na1: 225.0894, found
225.0884; Anal. Calcd for C13H14O2: C, 77.20; H, 6.98. Found: C,
77.48; H, 6.88.
5.4.3. ( ) Methyl (2R)-[(1R,2S,3S)-3-benzoyl-1,10-bi(cyclopro-
pyl)-2-yl]{[(benzyloxy)carbonyl]amino}ethanoate (8a)
5.3.2. ( ) (3S,6S)-3-Cyclohexyl-6-phenyl-3,6-dihydro-1,2-
dioxine (6b)
Colourless needles (500 mg, 27%); mp: 104–105 °C; Rf 0.18 (20%
ethyl acetate in hexanes, v/v); IR (nujol): 3342, 3181, 2924, 2727,
2671, 1749, 1725, 1716, 1648, 1596, 1577, 1531, 1462, 1377, 1340,
Colourless solid (2.05 g, 59%); mp <30 °C; Rf 0.84 (30% ethyl ace-
tate in hexanes, v/v); IR (nujol): 3082, 3054, 3032, 2932, 2862,
1270, 1202 cmꢀ1 1H NMR (600 MHz, CDCl3) d 0.22–0.26 (m, 1H),
;
1491, 1453, 1337, 1316, 1292, 1272, 1254 cmꢀ1
;
1H NMR
0.29–0.33 (m, 1H), 0.53–0.57 (m, 1H), 0.59–0.63 (m, 1H), 0.81–
0.82 (m, 1H), 1.54 (ddd, J = 8.4, 7.2, 4.8 Hz, 1H), 1.95 (ddd, J = 4.8,
4.2, 2.4 Hz, 1H), 2.80 (dd, J = 4.8, 4.2 Hz, 1H), 3.69 (s, 3H), 4.33–
4.36 (m, 1H), 5.11 (d, J = 12.0 Hz, 1H), 5.16 (d, J = 12.0 Hz, 1H), 5.45
(d, J = 7.8 Hz, 1H), 7.31–7.59 (m, 8H), 7.98–8.00 (m, 2H); 13C NMR
(150 MHz, CDCl3) d 4.9, 5.1, 8.3, 29.5, 31.9, 33.1, 52.5, 53.3, 67.1,
128.0, 128.1, 128.2, 128.5, 128.6, 133.1, 136.2, 137.4, 155.6, 172.1,
(300 MHz, CDCl3) d 1.30–1.90 (m, 10H), 2.18 (ddd, J = 8.7, 8.4,
8.1 Hz, 1H), 4.34–4.39 (ddd, J = 2.1, 2.1 Hz, 1H), 5.48 (dd, J = 2.4,
2.1 Hz, 1H), 6.08–6.18 (m, 2H), 7.35–7.47 (m, 5H); 13C NMR
(75 MHz, CDCl3) d 25.3, 25.4, 28.4, 29.4, 42.7, 80.1, 82.3, 126.6,
128.0, 128.4, 128.6, 128.6, 137.8; MS (EI) (C16H20O2) m/z (%): (M+
244, <1%), 198 (100), 169 (15), 156 (30), 142 (30), 129 (73), 115

N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
6095
197.9; MS (EI) (C24H25O5N) m/z (%): (M+ 407, <1%), 240 (3), 185 (100),
5.5.4. ( ) Phenyl (1S,2S,3R)-2-[(1S)-1-{[(benzyloxy)carbonyl]
amino}-2-methoxy-2-oxoethyl]-3-cyclohexylcyclopropan-
ecarboxylate (9b)
167 (6), 157 (6), 144 (15), 129 (8), 115 (8), 105 (100), 91 (9), 77 (65),
59 (4), 51 (6); HRMS (ESI, [M+H]+) calcd for C24H26O5N: 408.1811,
found 408.1816; Anal. Calcd for C24H25O5N: C, 70.74; H, 6.18; N,
3.44. Found: C, 70.66; H, 6.20; N, 3.46.
Prepared via method A: colourless needles (415 mg, 73%); mp:
143–145 °C; Rf 0.67 (30% ethyl acetate in petroleum spirit, v/v);
IR (nujol): 3387, 3069, 2922, 2727, 2671, 1740, 1702, 1656, 1590,
5.4.4. ( ) Methyl (2R)-[(1S,2S,3R)-2-benzoyl-3-cyclohexylcyclo-
propyl]{[(benzyloxy)carbonyl]amino}ethanoate (8b)
1520, 1499, 1460, 1377, 1346, 1293 cmꢀ1 1H NMR (300 MHz,
;
CDCl3) d 0.88–2.03 (m, 14H), 3.79 (s, 3H), 4.16–4.23 (m, 1H), 5.15
(dd, J = 10.2, 12.3 Hz, 2H), 5.51 (d, J = 8.4 Hz, 1H), 7.03–7.07 (m,
2H), 7.18–7.38 (m, 8H); 13C NMR (75 MHz, CDCl3) d 23.4, 25.7,
26.0, 26.1, 30.9, 32.7, 33.3, 36.1, 36.7, 52.4, 52.7, 67.2, 121.4,
125.7, 128.1, 128.3, 128.6, 129.3, 136.0, 150.6, 155.9, 171.5,
171.8; MS (EI) (C27H31O6N) m/z (%): (M+ 466, <1%), 372 (30), 344
(8), 328 (25), 310 (23), 268 (15), 251 (15), 223 (20), 91 (100); HRMS
(ESI, [M+H]+) calcd for C27H32O6N1: 466.2230, found 466.2233;
Anal. Calcd for C27H31O6N1: C, 69.66; H, 6.71; N, 3.01. Found: C,
69.40; H, 6.98; N, 2.97.
White solid (638 mg, 19%); mp: 156–158 °C; Rf 0.44 (30% ethyl
acetate in hexanes, v/v); IR (nujol): 3349, 2924, 2854, 2727, 2668,
1774, 1742, 1696, 1670, 1598, 1581, 1518, 1456, 1378, 1286 cmꢀ1
;
1H NMR (600 MHz, CDCl3) d 1.07–1.28 (m, 6H), 1.06–1.76 (m, 6H),
1.93–1.97 (m, 1H), 2.70–2.72 (m, 1H), 3.65 (s, 3H), 4.17–4.21 (m,
1H), 5.08 (d, J = 12.6 Hz, 1H), 5.18 (d, J = 12.6 Hz, 1H), 5.36 (br d,
J = 8.4 Hz, 1H), 7.31–7.37 (m, 5H), 7.47–7.50 (m, 2H), 7.56–7.59
(m, 1H), 8.00–8.01 (m, 2H); 13C NMR (150 MHz, CDCl3) d 25.8,
26.0, 26.1, 28.8, 32.3, 33.3, 33.6, 36.7, 37.1, 52.5, 53.2, 67.2,
128.0, 128.1, 128.2, 128.5, 128.6, 133.0, 136.2, 137.4, 155.5,
172.2, 198.3; MS (EI) (C27H29O5N) m/z (%): (M+ 450, <1%), 390 (7),
358 (5), 346 (7), 329 (8), 297 (7), 265 (5), 239 (15), 227 (75), 209
(25), 145 (30), 131 (15), 105 (100), 91 (60); HRMS (ESI,
[M+NH4]+) calcd for C27H35O5N2: 467.2546, found 467.2547.
5.6. General transesterification procedure
The phenyl ester was dissolved in anhydrous methanol and con-
centrated H2SO4 (two drops) was added. The solution was heated
under reflux for 16 h and solid NaHCO3 (120 mg) was added and
the methanol removed in vacuo until a volume of ca. 5 mL re-
mained. Dichloromethane was added and the solution extracted
with satd aqueous NaHCO3. The aqueous extract was extracted
with further dichloromethane, the organic extracts pooled, dried
over MgSO4, filtered and the volatiles removed in vacuo. Flash col-
umn chromatography (5% diethyl ether in dichloromethane, v/v) of
the residue afforded the pure methyl ester.
5.5. General Baeyer–Villiger oxidation procedures
5.5.1. Method A34
Trifluoroacetic anhydride (92 mmol) was added dropwise to a
stirred mixture of 30% H2O2 (29 mmol) in dichloromethane
(10 mL) and cooled to 0 °C. The temperature was maintained at
0 °C with the following dropwise addition of the phenyl ketone
(0.44 mol) dissolved in dichloromethane (10 mL). The reaction
was stirred under nitrogen and allowed to warm to rt overnight.
The reaction was quenched by pouring into 2% K2CO3(aq) and
extracting with dichloromethane. The organic extracts were
pooled, dried over MgSO4, filtered and the volatiles removed in va-
cuo. Flash column chromatography (30% ethyl acetate in hexanes,
v/v) afforded the pure phenyl ester.
5.6.1. ( ) Methyl (1R,2S,3S)-3-[(1S)-1-{[(benzyloxy)carbonyl]
amino}-2-methoxy-2-oxoethyl]-1,10-bi(cyclopropyl)-2-carbox-
ylate (10a)
Colourless, viscous oil (385 mg, 96%); Rf 0.43 (5% diethyl ether
in dichloromethane, v/v); IR (film): 3345, 3066, 3034, 3006, 2954,
2903, 2846, 1724, 1709, 1587, 1526, 1455, 1395 cmꢀ1 1H NMR
;
(600 MHz, CDCl3) d 0.26–0.28 (m, 1H), 0.32–0.35 (m, 1H), 0.57–
0.63 (m, 2H), 0.90–0.91 (m, 1H), 1.49 (ddd, J = 6.0, 3.0, 1.8 Hz,
2H), 1.70 (t, J = 4.2 Hz, 1H), 1.76 (dddd, J = 4.8, 4.8, 3.0, 1.2 Hz,
1H), 3.64 (s, 3H), 3.79 (s, 3H), 4.31 (t, J = 9.6 Hz, 1H), 5.12 (dd,
J = 14.4, 12.0 Hz, 2H), 5.44 (d, J = 8.4 Hz, 1H), 7.34 (m, 5H); 13C
NMR (150 MHz, CDCl3) d 0.2, 4.7, 4.8, 8.4, 23.2, 30.5, 31.0, 52.1,
52.8, 67.3, 128.3, 128.4, 128.7, 136.3, 156.1, 171.9, 173.5; MS
(EI) (C19H23O6N) m/z (%): (M+ 361, <1%), 258 (5), 226 (3), 194
(3), 166 (4), 139 (70), 107 (24), 91 (100), 79 (45), 59 (10);
HRMS (ESI, [M+Na]+) calcd for C19H23O6N: 384.1423, found
384.1421.
5.5.2. Method B
The phenyl ketone (3 mmol) and meta-chloroperbenzoic acid
(20 mmol), were dissolved in chloroform and the solution left in
the dark at rt for one month. After quenching the reaction with sat-
urated aqueous sodium thiosulfate, the resulting solution was ex-
tracted with dichloromethane. The organic extracts were washed
with saturated NaHCO3(aq) followed by saturated NaCl(aq). The or-
ganic extracts were pooled, dried over MgSO4, filtered and the vol-
atiles removed in vacuo. Flash column chromatography (30% ethyl
acetate in hexanes, v/v) afforded the pure phenyl ester.
5.5.3. ( ) Phenyl (1R,2S,3S)-3-[(1S)-1-{[(benzyloxy)carbonyl]
amino}-2-methoxy-2-oxoethyl]-1,10-bi(cyclopropyl)-2-carbox-
ylate (9a)
5.6.2. ( ) Methyl (1S,2S,3R)-2-[(1S)-1-{[(benzyloxy)carbonyl]
amino}-2-methoxy-2-oxoethyl]-3-cyclohexylcyclopropanecarb-
oxylate (10b)
Prepared via method A: pale yellow oil (75 mg, 24%); prepared
via method B: colourless oil (954 mg, 77%); Rf 0.45 (5% diethyl
ether in dichloromethane, v/v); IR (nujol): 3351, 3066, 3033,
Colourless solid (258 mg, 76%); mp: 129–130 °C; Rf 0.64 (30%
ethyl acetate in petroleum spirit, v/v); IR (nujol): 3345, 2916,
2728, 2671, 2360, 1751, 1718, 1706, 1523, 1461, 1377, 1364,
3004, 2954, 2850, 1732, 1593, 1494, 1456, 1376 cmꢀ1 1H NMR
;
1344, 1282, 1231 cmꢀ1 1H NMR (600 MHz, CDCl3) d 1.07–1.25
;
(300 MHz, CDCl3) d 0.35–0.43 (m, 2H), 0.65–0.68 (m, 2H), 0.96–
0.98 (m, 1H), 1.64 (ddd, J = 5.4, 3.6, 1.5 Hz, 1H), 1.87–1.96 (m,
2H), 3.82 (s, 3H), 4.41 (t, J = 9.6 Hz, 1H), 5.15 (dd, J = 12.0, 9.9 Hz,
2H), 5.53 (d, J = 8.1 Hz, 1H), 7.04–7.07 (m, 2H), 7.19–7.39 (m,
3H); 13C NMR (75 MHz, CDCl3) d 4.6, 4.8, 8.3, 23.3, 31.1, 31.7,
52.6, 52.7, 67.2, 121.4, 125.8, 128.1, 128.2, 128.6, 129.3, 136.1,
150.6, 155.9, 171.5, 171.6; MS (EI) (C24H25O6N) m/z (%): (M+ 423,
<1%), 201 (50), 162 (6), 134 (5), 119 (6), 107 (20), 91 (100), 79
(25), 65 (10), 51 (5); HRMS (ESI, [M+NH4]+) calcd for C24H29O6N2:
441.2026, found 441.2027.
(m, 6H), 1.39–1.41 (m, 1H), 1.55–1.91 (m, 7H), 3.65 (s, 3H), 3.77
(s, 3H), 4.08–4.11 (m, 1H), 5.09–5.15 (m, 2H), 5.37 (d, J = 7.8 Hz,
1H), 7.32–7.38 (m, 5H); 13C NMR (150 MHz, CDCl3) d 0.0, 23.2,
25.8, 26.0, 26.2, 30.2, 32.8, 33.3, 35.2, 36.6, 51.9, 52.6, 67.2,
128.1, 128.3, 128.6, 136.1, 155.8, 171.7, 173.7; MS (EI)
(C21H27O6N) m/z (%): (M+ 403, <1%), 344 (6), 300 (5), 268 (5), 236
(4), 204 (5), 181 (33), 149 (16), 133 (6), 121 (6), 108 (26), 91
(100), 79 (32), 67 (10), 55 (7); HRMS (ESI, [M+H]+) calcd for
C22H30O6N1: 404.2073, found 404.2076.

6096
N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
5.7. General ester hydrolysis procedure
34.3, 37.1, 37.2, 37.8, 53.6, 53.8, 172.2, 178.5; HRMS (ESI,
[M+H]+) calcd for C12H20O4N1: 242.1392, found 242.1387.
The methyl ester (1.0 mol) was dissolved in THF (5 mL) and stir-
red with 2.5 M aqueous LiOH (10 mL, 25 mmol) at rt overnight.
Brine (10 mL) was added followed by concentrated HCl (two
drops). The solution was extracted with ether (2 ꢁ 15 mL), dried
over MgSO4, filtered and the solvents removed in vacuo.
6. Pharmacology
6.1. In vitro methods
6.1.1. Binding assays at native iGlu receptors
5.7.1. ( ) (1R,2S,3S)-3-[(S)-{[(Benzyloxy)carbonyl]amino}
(carboxy)methyl]-1,10-bi(cyclopropyl)-2-carboxylic acid (11a)
White solid (298 mg, 92%); mp: 130–132 °C; IR (nujol): 3303,
All binding assays were performed using rat brain synaptic
membranes of cortex and the central hemispheres from adult male
Sprague-Dawley rats with tissue preparation as earlier described.35
Affinities for native AMPA, KAIN and NMDA receptors were deter-
mined using 5 nM [3H]AMPA,36 5 nM [3H]KAIN,37 and 2 nM
[3H]CGP39653.38 On the day of experiments, frozen membranes
were quickly thawed and homogenised in 40 volumes of ice-cold
buffer (pH 7.4) (30 mM Tris–HCl containing 2.5 mM CaCl2,
50 mM Tris–HCl, or 50 mM Tris–HCl containing 2.5 mM CaCl2, for
[3H]AMPA, [3H]KAIN, or [3H]CGP39653 binding, respectively), and
centrifuged (48,000 g for 10 min). This step was repeated four
times. In [3H]AMPA binding experiments, 100 mM KSCN was
added to the buffer during the final wash and during incubation.
The final pellet was re-suspended in ice-cold buffer, corresponding
to approx. 0.4–0.5 mg protein/mL. [3H]AMPA, [3H]KAIN, and
[3H]CGP39653 binding were carried out in aliquots consisting of
2922, 2727, 2672, 1702, 1456, 1377, 1301, 1245 cmꢀ1 1H NMR
;
(600 MHz, CD3OD) d 0.25–0.30 (m, 2H), 0.54–0.62 (m, 2H), 0.95–
0.96 (m, 1H), 1.50–1.53 (m, 2H), 1.83–1.87 (m, 1H), 4.14 (d,
J = 10.2 Hz, 1H), 5.10 (dd, J = 12.6, 10.2 Hz, 2H), 7.28–7.36 (m,
5H); 13C NMR (150 MHz, CD3OD) d 5.5, 5.6, 9.6, 24.7, 31.7, 32.7,
68.1, 129.2, 129.4, 130.0, 138.8, 159.2, 177.4, 210.3; MS (EI)
(C17H19O6N) m/z (%): (M+ 333, 7%), 328 (14), 310 (10), 239 (14),
212 (27), 166 (45), 152 (14), 107 (100), 91 (55), 79 (36); HRMS
(ESI, [M+H]+) calcd for C17H20O6N1: 334.1291, found 334.1289.
5.7.2. ( ) (1S,2S,3R)-2-[(S)-{[(Benzyloxy)carbonyl]amino}(car-
boxy)methyl]-3-cyclohexylcyclopropanecarboxylic acid (11b)
White solid (185 mg, 97%); mp: 160–167 °C; IR (nujol): 3306,
2923, 2854, 2729, 1728, 1694, 1682, 1538, 1463, 1409, 1377,
25 l lL test solution, and 200 lL membrane sus-
L [3H]ligand, 25
1326, 1285, 1246 cmꢀ1
;
1H NMR (600 MHz, CD3OD) d 1.06–1.41
pension and incubated for 30 min, 60 min, and 60 min, respec-
tively. Binding was terminated by filtration through GF/B filters
using a 96-well Packard Filter-Mate Cell Harvester and washing
(m, 8H), 1.63–2.02 (m, 6H), 3.93–3.95 (m, 1H), 5.05–5.12 (m,
2H), 7.25–7.35 (m, 5H); 13C NMR (150 MHz, CD3OD) d 25.1, 27.4,
27.7, 27.9, 31.7, 34.4, 34.9, 37.0, 38.5, 54.6, 68.2, 129.2, 129.4,
130.0, 138.7, 159.1, 175.4, 178.1; MS (EI) (C20H25O6N) m/z (%):
(M+ 375, <1%), 357 (12), 250 (15), 240 (15), 224 (16), 194 (14),
178 (32), 131 (35), 108 (47), 91 (100), 79 (70); HRMS (ESI,
[M+H]+) calcd for C20H26O6N1: 376.1760, found 376.1760.
with 3 ꢁ 250
lL buffer. After drying, 25 lL Microscint 0 (Perkin–El-
mer) per well was added and the plate was counted on a Topcount-
er (Perkin–Elmer). Non-specific binding was determined using
1 mM (S)-Glu. The Bradford39 protein assay was used for protein
determination using bovine serum albumin as a standard, accord-
ing to the protocol of the supplier (Bio-Rad, Milan, Italy).
5.8. General catalytic hydrogenation procedure
6.1.2. Binding assays at recombinant mGlu receptors
All cell culture reagents were from Gibco unless stated otherwise.
The Cbz-protected amine (0.8 mmol) was dissolved in methanol
(10 mL) and 5% Pd/C (20 wt %) was added. The reaction was evac-
uated under vacuum and subsequently filled with H2(g) from a
hydrogen balloon. The reaction was stirred at rt for 48 h, filtered
through a pad of kenite using methanol and water and concen-
trated in vacuo. The crude product was dissolved in water (5 mL)
and concd HCl (two drops) was added. The solvent was removed
in vacuo and the residual HCl salt (0.5 mmol) dissolved in metha-
nol (0.6 mmol). Anhydrous propylene oxide (25 mL/mmol) was
then added and the solution stirred under an N2 atmosphere at rt
overnight. The precipitated solid was collected by filtration and
washed with ethyl acetate to give the pure free amine.
6.1.3. Inhibition of forskolin-stimulated cAMP production
CHO cells expressing the mGluR2 or mGluR4 (previously de-
scribed by Tanabe et al.40,41) were cultured in DMEM with Gluta-
MAX-I with 10% dialyzed FBS, 1% Penicillin–Streptomycin, and
2.5 g/L -proline at 5% CO2 at 37 °C. Cells were seeded at 26,000
L
cells/96-well 24 h before assaying. Prior to compound addition
cells were washed with DPBS. For agonist testing cells were first
incubated with cAMP-ground buffer (DPBS with 1 mM CaCl2,
1 mM MgCl2 and 1 mM IBMX) for 20 min at 37 °C. The buffer
was then replaced with test compounds diluted in cAMP-ground
buffer supplemented with 25 lM forskolin and incubated at
5.8.1. ( ) (1R,2S,3S)-3-[(S)-Amino(carboxy)methyl]-1,10-bi(cyclo-
propyl)-2-carboxylic acid (3)
37 °C for 10 min. For antagonist testing cells were first incubated
with test compounds diluted in cAMP-ground buffer for 20 min
at 37 °C, and then replaced with cAMP-ground buffer containing
White solid (158 mg, 95%); dec: 234 °C; IR (solid): 3013, 1693,
1656, 1612, 1562, 1508, 1470, 1457, 1393, 1343, 1321, 1300,
test compounds, 25
lM forskolin, and
L-glutamate corresponding
1239 cmꢀ1 1H NMR (600 MHz, D2O) d 0.05–0.13 (m, 2H), 0.30–
;
to the EC80 (20 M for mGluR2, 30
l
lM for mGluR4) for 10 min at
37 °C. All reactions were terminated by aspiration and addition
of ice-cold sodium acetate buffer pH 6.2 supplemented with 0.1%
triton X-100 and 0.1 mM IBMX. cAMP levels were quantified using
the Adenylyl Cyclase Activation FlashPlateÒ Assay (Perkin Elmer)
and interpolated from a cAMP standard curve. All experiments
were performed in triplicate and the results are given as mean
SEM of at least two independent experiments.
0.35 (m, 1H), 0.40–0.44 (m, 1H), 0.72–0.77 (m, 1H), 1.51–1.55
(m, 2H), 1.76 (ddd, J = 4.8, 4.2, 2.4 Hz, 1H), 3.82 (d, J = 11.4 Hz,
1H); 13C NMR (150 MHz, D2O) d5.53, 9.14, 25.06, 28.72, 32.61,
53.97, 172.37, 178.03; HRMS (ESI, [M + H]+) calcd for C9H14O4N1:
200.0923, found 200.0917.
5.8.2. ( ) (1S,2S,3R)-2-[(S)-Amino(carboxy)methyl]-3-cyclohe-
xylcyclopropanecarboxylic acid (4)
6.1.4. Inositol phosphate (IP) turnover assay
Off-white solid (16 mg, 16%); IR (solid): 2924, 2851, 1680, 1586,
1554, 1451, 1391, 1320, 1231 cmꢀ1 1H NMR (600 MHz, D2O) d
;
CHO cells expressing the mGluR1 (previously described by
Aramori and Nakanishi42) were cultured as described in the cAMP
assay above. Cells were seeded at 26,000 cells/96-well and labeled
0.80–0.93 (m, 6H), 1.27–1.43 (m, 6H), 1.64–1.73 (m, 2H), 3.60 (d,
J = 10.8, 1H); 13C NMR (150 MHz, D2O) d25.2, 27.3, 29.1, 33.8,

N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
6097
with myo-[2-3H]Inositol (4
lCi/mL, TRK911, GE Healthcare) 24 h
daily for any sign of infection. No such cases occurred in this study.
In brief, four sterile chromic gut sutures (cuticular 4–0 chromic gut,
FS-2; Ethicon, Somerville, NJ, USA) were loosely tied around the
gently isolated sciatic nerve. Drug testing was delayed until
16 days after surgery to ensure that neuropathic pain was well
established prior to the initiation of drug delivery.
before assaying. Prior to compound addition cells were washed
with DPBS. For agonist testing cells were first incubated with IP-
ground buffer (HBSS containing 20 mM HEPES, 1 mM CaCl2,
1 mM MgCl2 and 0.85 mg/mL LiCl, pH 7.4) for 30 min at 37 °C.
The buffer was then replaced with test compounds diluted in IP-
ground buffer and incubated at 37 °C for 30 min. For antagonist
testing cells were first incubated with test compounds diluted in
IP-ground buffer for 30 min at 37 °C, and then replaced with IP-
6.2.5. Von Frey testing
All testing was conducted blind with respect to group assign-
ment. Rats received at least three 60-min habituations to the test
environment prior to behavioural testing. The von Frey test was
performed within the sciatic innervation region of the hind paws
as previously described.46 Assessments were made prior to (base-
line) and at 1 and 3 h after intrathecal drug dosing. A logarithmic
series of 10 calibrated Semmes-Weinstein monofilaments (von
Frey hairs; Stoelting, Wood Dale, IL, USA) was applied randomly
to the left and right hind paws to define the threshold stimulus
intensity required to elicit a paw withdrawal response. Log stiff-
ness of the hairs was determined by log10 (milligrams ꢁ 10) and
ranged from 3.61 (4.07 g) to 5.18 (15.136 g). The behavioural re-
sponses were used to calculate absolute threshold (the 50% paw
withdrawal threshold) by fitting a Gaussian integral psychometric
function using a maximum likelihood fitting method, as described
previously.46 This fitting method allows parametric analyses that
otherwise would not be appropriate.
ground buffer containing test compounds and 30 lM L-glutamate
corresponding to the EC80 for 30 min at 37 °C. All reactions were
terminated by aspiration and addition of ice-cold 10 mM formic
acid and incubation for 30 min at 4 °C. Yttrium silicate scintillation
proximity assay beads (RPNQ0010, GE Healthcare) were used for
measuring radioactivity from generated [3H]IP as previously de-
scribed.43 Radioactivity was quantified in a Packard TopCount
microplate scintillation counter and responses read as counts per
minute (CPM).
6.2. In vivo methods
6.2.1. Animals
Experiments were carried out on 12 male, pathogen-free Spra-
gue-Dawley rats (325–500 g) housed in groups of three per cage.
Rats were given free access to food and water and maintained on
a controlled 12/12 h light/dark cycle with lights on at 0700 h.
6.2.6. Data analysis
6.2.2. Ethics
All data is reported as mean SEM. Von Frey data was analysed
as the interpolated 50% thresholds (absolute threshold) in log base
10 of stimulus intensity (monofilament stiffness in milli-
grams ꢁ 10). Pre-drug baseline measures were analysed by one-
way ANOVA. Post drug time course measures were analysed by re-
peated measures two-way ANOVAs followed by Bonferroni post
hoc tests. The 3 h time point data was compared to vehicle by
one-way ANOVA followed by Dunnett’s t test post hoc. Where
appropriate, P <0.05 was considered statistically significant.
Ethical approval was obtained from the University of Adelaide
Animal Ethics Committee for all animal tests and manipulations
and care was taken to minimize the extent of suffering and dura-
tion of pain, where doing so would not interfere with the project.
All experimental work involving animals abided by the guidelines
found in the Australian Code of Practice for the Care and Use of
Animals for Scientific Purposes.
6.2.3. Drugs
The test compound ( ) (1R,2S,3S)-3-[(S)-amino(carboxy)-
methyl]-1,10-bi(cyclopropyl)-2-carboxylic acid (( )-3) was syn-
thesized in the Discipline of Chemistry, University of Adelaide
and the positive control (2R,4R)-aminopyrrolidine-2,4-dicarboxyl-
ate (APDC) was purchased from Tocris Bioscience (Bristol, UK).
Both drugs were administered intrathecally, compound ( )-3 as a
suspension (500 nmols) or solution (50 and 250 nmols) and APDC
as a solution in Milli-Q water containing 12 mM HCl. The method
of acute intrathecal drug administration was based on that de-
scribed previously.44 Briefly, intrathecal operations were con-
ducted under isoflurane anaesthesia (Phoenix Pharmaceuticals,
St. Joseph, MO, USA) by threading sterile polyethylene-10 tubing
(PE-10 Intramedic Tubing; Becton Dickinson Primary Care Diag-
nostics, Sparks, MD, USA) guided by an 18-gauge needle between
the L5 and L6 vertebrae. The catheter was inserted such that the
proximal catheter tip lay over the lumbosacral enlargement. The
catheters were pre-loaded with drugs at the distal end in a total
Acknowledgements
This work was generously funded by grants provided by the
Australian Research Council and also a commercialisation grant
awarded to D.K.T from the University of Adelaide. N.J.S is grateful
to be recipient of a University of Adelaide postgraduate divisional
scholarship. M.R.H is recipient of a National Health and Medical
Research Council CJ Martin Fellowship (ID 465423).
Supplementary data
Supplementary data (proton and carbon NMR spectra of com-
pounds 3, 4, 6a, 6b, 7a, 7b, 8a, 8b, 9a, 9b, 10a, 10b, 11a and 11b
as well as details of X-ray crystallography for compounds 7a and
7b) associated with this article can be found, in the online version,
at doi:10.1016/j.bmc.2010.06.051.
volume of no greater than 25 lL and delivered over 20–30 s once
the catheter was in position. The catheters were 17 cm in length,
References and notes
and were attached to a pre-loaded Hamilton syringe.
1. McQuay, H. J. Eur. J. Pain 2002, 6, 11.
2. Colombo, B.; Annovazzi, P. O.; Comi, G. Neurol. Sci. 2006, 27, S183.
3. Chizh, B. A. Amino Acids 2002, 23, 169.
4. Kew, J. N.; Kemp, J. A. Psychopharmacology 2005, 179, 4.
5. Neugebauer, V. Pain 2002, 98, 1.
6. Jang, J. H.; Kim, D. W.; Sang Nam, T.; Se Paik, K.; Leem, J. W. Neuroscience 2004,
128, 169.
7. Schoepp, D. D.; Jane, D. E.; Monn, J. A. Neuropharmacology 1999, 38, 1431.
8. Collado, I.; Pedregal, C.; Bueno, A. B.; Marcos, A.; Gonzalez, R.; Blanco-Urgoiti, J.;
Perez-Castells, J.; Schoepp, D. D.; Wright, R. A.; Johnson, B. G.; Kingston, A. E.;
Moher, E. D.; Hoard, D. W.; Griffey, K. I.; Tizzano, J. P. J. Med. Chem. 2004, 47,
456.
6.2.4. Chronic constriction injury (CCI)
Neuropathic pain was induced using the chronic constriction
injury (CCI) model of partial sciatic nerve injury. CCI was per-
formed at the mid-thigh level of the left hind leg as previously de-
scribed.45 Animals were anesthetized with isoflurane. The shaved
skin was treated with ethanol to cleanse and the surgery was asep-
tically performed. Animals were monitored post-operatively until
fully ambulatory prior to return to their home cage, and checked

6098
N. J. Stanley et al. / Bioorg. Med. Chem. 18 (2010) 6089–6098
9. Gonzalez, R.; Collado, I.; de Uralde, B. L.; Marcos, A.; Martin-Cabrejas, L. M.;
Pedregal, C.; Blanco-Urgoiti, J.; Perez-Castells, J.; Fernandez, M. A.; Andis, S. L.;
Johnson, B. G.; Wright, R. A.; Schoepp, D. D.; Monn, J. A. Bioorg. Med. Chem.
2005, 13, 6556.
10. Avery, T. D.; Greatrex, B. W.; Pedersen, D. S.; Taylor, D. K.; Tiekink, E. R. J. Org.
Chem. 2008, 73, 2633.
11. Collado, I.; Pedregal, C.; Mazon, A.; Espinosa, J. F.; Blanco-Urgoiti, J.; Schoepp, D.
D.; Wright, R. A.; Johnson, B. G.; Kingston, A. E. J. Med. Chem. 2002, 45, 3619.
12. Pellicciari, R.; Costantino, G.; Marinozzi, M.; Macchiarulo, A.; Amori, L.; Josef
Flor, P.; Gasparini, F.; Kuhn, R.; Urwyler, S. Bioorg. Med. Chem. Lett. 2001, 11,
3179.
13. Pellicciari, R.; Marinozzi, M.; Natalini, B.; Costantino, G.; Luneia, R.; Giorgi, G.;
Moroni, F.; Thomsen, C. J. Med. Chem. 1996, 39, 2259.
14. Hansson, E. Acta Physiol. 2006, 187, 321.
15. Bleakman, D.; Alt, A.; Nisenbaum, E. S. Semin. Cell Dev. Biol. 2006, 17, 592.
16. Fisher, K.; Lefebvre, C.; Coderre, T. J. Pharmacol. Biochem. Behav. 2002, 73, 411.
17. Simmons, R. M.; Webster, A. A.; Kalra, A. B.; Iyengar, S. Pharmacol. Biochem.
Behav. 2002, 73, 419.
27. Petralia, R. S.; Wang, Y. X.; Niedzielski, A. S.; Wenthold, R. J. Neuroscience 1996,
71, 949.
28. Watkins, L. R.; Maier, S. F. Physiol. Rev. 2002, 82, 981.
29. Garrison, C. J.; Dougherty, P. M.; Kajander, K. C.; Carlton, S. M. Brain Res. 1991,
565, 1.
30. Cao, H.; Zhang, Y. Q. Neurosci. Biobehav. Rev. 2008, 32, 972.
31. Jung, C. Y.; Lee, S. Y.; Choi, H. S.; Lim, E. J.; Lee, M. K.; Yang, G. Y.; Han, S. R.;
Youn, D. H.; Ahn, D. K. Neurosci. Lett. 2006, 409, 173.
32. Popik, P.; Kozela, E.; Pilc, A. Br. J. Pharmacol. 2000, 130, 1425.
33. Murakami, M.; Ubukata, M.; Ito, Y. Tetrahedron Lett. 1998, 39, 7361.
34. Anastasia, M.; Allevi, P.; Ciuffreda, P.; Fiecchi, A.; Scala, A. J. Org. Chem. 1985, 50,
321.
35. Ransom, R. W.; Stec, N. L. J. Neurochem. 1988, 51, 830.
36. Honoré, T.; Nielsen, M. Neurosci. Lett. 1985, 54, 27.
37. Braitman, D. J.; Coyle, J. T. Neuropharmacology 1987, 26, 1247.
38. Sills, M. A.; Fagg, G.; Pozza, M.; Angst, C.; Brundish, D. E.; Hurt, S. D.; Wilusz, E.
J.; Williams, M. Eur. J. Pharmacol. 1991, 192, 19.
39. Bradford, M. M. Anal. Biochem. 1976, 72, 248.
18. Gerber, G.; Zhong, J.; Youn, D.; Randic, M. Neuroscience 2000, 100, 393.
19. Ohishi, H.; Shigemoto, R.; Nakanishi, S.; Mizuno, N. J. Comp. Neurol. 1993, 335,
252.
40. Tanabe, Y.; Masu, M.; Ishii, T.; Shigemoto, R.; Nakanishi, S. Neuron 1992, 8, 169.
41. Tanabe, Y.; Nomura, A.; Masu, M.; Shigemoto, R.; Mizuno, N.; Nakanishi, S. J.
Neurosci. 1993, 13, 1372.
20. Jia, H.; Rustioni, A.; Valtschanoff, J. G. J. Comp. Neurol. 1999, 410, 627.
21. Azkue, J. J.; Mateos, J. M.; Elezgarai, I.; Benitez, R.; Osorio, A.; Diez, J.; Bilbao, A.;
Bidaurrazaga, A.; Grandes, P. Neurosci. Lett. 2000, 287, 236.
22. Carlton, S. M.; Hargett, G. L.; Coggeshall, R. E. Neuroscience 2001, 105, 957.
23. Fisher, K.; Coderre, T. J. Pain 1996, 68, 255.
24. Fundytus, M. E. CNS Drugs 2001, 15, 29.
25. Millan, M. J. Prog. Neurobiol. 2002, 66, 355.
26. Zhou, H. Y.; Zhang, H. M.; Chen, S. R.; Pan, H. L. J. Neurophysiol. 2007, 97, 871.
42. Aramori, I.; Nakanishi, S. Neuron 1992, 8, 757.
43. Brandish, P. E.; Hill, L. A.; Zheng, W.; Scolnick, E. M. Anal. Biochem. 2003, 313, 311.
44. Milligan, E. D.; Hinde, J. L.; Mehmert, K. K.; Maier, S. F.; Watkins, L. R. J. Neurosci.
Method 1999, 90, 81.
45. Milligan, E. D.; Maier, S. F.; Watkins, L. R. Method Mol. Med. 2004, 99, 67.
46. Milligan, E. D.; O’Connor, K. A.; Nguyen, K. T.; Armstrong, C. B.; Twining, C.;
Gaykema, R. P.; Holguin, A.; Martin, D.; Maier, S. F.; Watkins, L. R. J. Neurosci.
2001, 21, 2808.