Stereospecific synthesis and absolute configuration of the (2S,3S,4S)-isomer of 2-methyl-2-(carboxycyclopropyl)glycine (MCCG)

Article abstract of DOI:10.1016/S0957-4166(03)00049-1

The conformationally restricted metabotropic glutamate receptor antagonist (2S,3S,4S)-2-methyl-2-(carboxycyclopropyl)glycine 1 (MCCG) has been synthesized in a stereoselective manner (>99% ee) with the (2S,3S,4S) absolute configuration of this molecule be

Full text of DOI:10.1016/S0957-4166(03)00049-1

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 14 (2003) 593–596
Stereospecific synthesis and absolute configuration of the
(2S,3S,4S)-isomer of 2-methyl-2-(carboxycyclopropyl)glycine
(MCCG)
Hassan Pajouhesh,a,* Kenneth Curry,b Hossein Pajouhesh,a Mahmonir H. Mereshta and
Brian Patrickc
aPrecision Biochemicals Inc., 303-2386 East Mall, Vancouver, BC, Canada V6T 1Z3
bCNS Research Division, Prescient Neuropharma Inc., 311-2386 East Mall, Vancouver, BC, Canada V6T 1Z3
cDepartment of Chemistry, University of British Columbia, Vancouver, BC, Canada V6T 1Z3
Received 27 November 2002; accepted 3 January 2003
Abstract—The conformationally restricted metabotropic glutamate receptor antagonist (2S,3S,4S)-2-methyl-2-(carboxycyclo-
propyl)glycine 1 (MCCG) has been synthesized in a stereoselective manner (>99% ee) with the (2S,3S,4S) absolute configuration
of this molecule being confirmed by X-ray crystallographic analysis. Subsequent physico–chemical studies were undertaken and
the data are at odds with those of the commercially available product. © 2003 Elsevier Science Ltd. All rights reserved.
1. Introduction
geometry and stereochemistry have also been used to
differentiate between mGluRs and iGluRs and then to
further differentiate the receptor sub-types contained
within each family. To date, eight distinct metabotropic
glutamate receptor proteins (mGluR1–8) have been iden-
tified and divided into three subgroups according to
sequence homology, signal transduction mechanism and
pharmacology.4,5 Group I is linked to the activation of
phospholipase C that subsequently increases the hydrol-
ysis of phosphatidyl inositol (PI), while groups II and III
are both negatively coupled to adenylate cyclase activity.
The mGluRs have been implicated in both normal and
pathological neurotransmission in the mammalian cen-
tral nervous system and are thus considered novel targets
for therapeutic intervention.
Many rigid or conformationally restricted analogs of the
excitatory neurotransmitter
L
-glutamate 2 (L-Glu, Fig. 1)
have now been synthesized for the elucidation of the
pharmacology of glutamate receptors located in the
central nervous systems of many species.1–3 Glutamate
receptors are now divided into two broad categories,
those associated with the neurotransmission of excitatory
responses mediated through ion channel-coupled recep-
tors (ionotropic glutamate receptors (iGluRs) and those
associated with neuroregulatory processes mediated
through G-protein-coupled receptors, known as metabo-
tropic glutamate receptors (mGluRs).
containing pharmacophoric functional groups of known
L-Glu analogs
Figure 1.
* Corresponding author. Present address: NeuroMed Technologies 301-2389 Health Sciences Mall, UBC Vancouver, BC, Canada V6T 1Z4. Fax:
+1-604-822-9978; e-mail: hpajouhesh@neuromedtech.com
0957-4166/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0957-4166(03)00049-1

594
H. Pajouhesh et al. / Tetrahedron: Asymmetry 14 (2003) 593–596
The use of cyclopropyl glycine derivatives in pharmaco-
logical studies on mGluRs is well documented7–9 and
examples within this family of compounds can be found
which interact with both individual and groups of
iGluRs and mGluRs.9 One particular compound, 2-
(carboxycyclopropyl)glycine (CCG), has been synthe-
sized in its eight individual isomeric forms and
subsequently tested in pharmacological systems using
mGluRs.6 A number of the analogs of this compound
show pharmacological activity, but of particular impor-
tance to this current study is the observation that the
unsaturated adduct 5 in 62% yield. In this case the alkyl
group reacted at the Re face with high selectivity. This
process has been called ‘self-reproduction of chirality’
by Seebach.13 This compound can then be cyclo-
propanated with diazomethane in the presence of palla-
dium acetate, to give one isomer only of 6. Two
subsequent deprotection and hydrolysis steps yield 1 as
a single isomer [h]D=+77 (c 0.5, H2O) {commercial
MCCG, [h]D=−54.5 (c 0.12, H2O)} and in an enan-
tiomeric excess of 99% as judged by chiral HPLC
(Chirex column,
D-penicillamine stationary phase, 1
(2S,3S,4S)-isomer of CCG ( -CCG-I) 3 is a potent
L
mM CuSO4 mobile phase).
group II mGluR agonist.9 Replacement of the a-amino
acid proton of 3 with a methyl group produces MCCG
1, transforming this molecule from an agonist into an
antagonist.7 The original communication concerning
the synthesis (without X-ray crystallography valida-
tion), assigned the configurations of the (2S,3S,4S)-
and (2S,3R,4R)-isomers with some detail of physical
data.10 Since that time the widespread use of the sup-
posed (2S,3S,4S)-isomer has become standard in the
field both as a pharmacological tool and as a model for
continuing SAR studies and new compound synthesis.
In our continuing program to elucidate mGluR func-
tion we had occasion to synthesise 1 by a route slightly
different from the published procedure.10 We were puz-
zled to find that NMR and chiroptical data agree with
the published figures of the supposed (2S,3R,4R)-iso-
mer and proceeded to produce crystals for X-ray deter-
mination of configuration. This analysis confirmed that
our compound has the (2S,3S,4S)-configuration. Using
expressed mGluR2 receptors we have compared the
actions of our compound with that of the commercially
available product. Both are antagonists, however our
compound is about three times more potent than the
commercially available one.
3. X-Ray studies
A crystal of 1 with dimensions 0.15×0.10×0.12 mm was
mounted on a glass fiber. Room temperature data were
collected on a Rigaku/ADSC CCD area detector in two
sets of scans (ƒ=0.0 to 190.0°, =0°; and ꢀ=−18.0 to
23.0°, =−90°) using 1.00° oscillations with 35.0-s
exposures. The crystal-to-detector distance was 39.94(3)
mm with a detector swing angle of −5°. The data was
processed using the d*TREK program and corrected
for Lorentz and polarization effects.
A primitive orthorhombic unit cell was found, while the
space group was determined on the basis of systematic
absences. The determination of the configuration has
been performed as a relative configurational assigment
in relation to the absolute configuration12 at C(2) and
was found to exist as zwitterionic and contain an
additional water molecule in the lattice. The absolute
configurations of C(4) and C(6) determined to be S
(Fig. 2). All non-hydrogen atoms were refined
anisotropically, while all hydrogens involved in hydro-
gen-bonding were refined isotropically. All other hydro-
gens were included in calculated positions. A list of
crystallographic data appears in Table 1. All calcula-
tions were performed using the teXsan12 crystallo-
graphic software package of Molecular Structure
Corporation.
2. Chemistry
The route used for the synthesis of (2S,3S,4S)-MCCG
1 (Scheme 1) was developed in order to avoid a com-
plex cyclopropanation step in the published method.10
Starting with the known11 (2R,4R) oxazolidinone
{[h]20=+27.0} 4, alkylation with methyl-(E)-3-bromo-
propenoate took place stereospecifically to give the
Comparative analyses (specific rotation value, 1H
NMR) were performed on the commercial and synthe-
Scheme 1.

H. Pajouhesh et al. / Tetrahedron: Asymmetry 14 (2003) 593–596
595
sized MCCG 1 and the data clearly indicates that our
finding are consistent with the compound in the previ-
ously published investigation,10 which is assigned the
(2S,3S,4S)-configuration. The discrepancies between
our study and the published synthesis10 suggest a trans-
position of the two MCCG isomers in the originally
published structure and what is assigned as the
(2S,3S,4S)-configuration is in fact the (2S,3R,4R)-
configuration.
Many of the receptor systems associated with glutamic
acid demonstrate a clear and strong dependence on
both geometry and stereochemistry. We also believe
this to be the case for the isomers of MCCG. Using the
well documented isomers of LCCG as a template we
can see that the mGluR2 receptor demonstrates a clear
preference for the (S)-configuration at the a-amino acid
carbon and also a 50–100 times preference for the
(2S,3S,4S)-isomer over the (2S,3R,4R)-isomer.9 In this
case the stereochemistry of the ring is crucial and
indicates that the orientation determines the agonist
properties the two isomers exhibit. With the two
MCCG isomers the difference in apparent IC50 is only
about 3× between the commercial product and the
(2S,3S,4S)-isomer. In this case the rings must be able to
attain a conformation which favors antagonist activity
in both cases. It is surprising that commercial MCCG
with the (2S,3R,4R)-configuration has substantial
antagonist activity at mGluR2, since the parent com-
pound LCCG-I does not have the same tolerance
towards inversion of chiral centers at the cyclopropane
ring. Of greater importance perhaps is the fact that
much of the design work for next generation group II
agonists and antagonists is based on the erroneous
assumption that the absolute stereochemistry of the
ligand is known. Our work has demonstrated the dan-
gers of making such assumptions and the importance of
careful assignment of absolute configuration.
Figure 2. X-Ray crystal structure of (2S,3S,4S)-MCCG.
Table 1. Crystallographic data
Formula
Fw
C7H13NO5
191.18
Colour, habit
Crystal size (mm)
Crystal system
Space group
Clear, block
0.15×0.12×0.10
Orthorhombic
P212121
7.0802(9)
10.0398(8)
12.170(2)
90
,
a (A)
4. Experimental
4.1. General
,
b (A)
,
c (A)
h (°)
i (°)
90
k (°)
90
865.1(2)
4
1.468
408.00
Mo Ka
0.125
0.993–1.000
0.0–190.0
−18.0–23.0
50.2
6164
1535
142
0.089
0.076, 0.106
0.88
0.02
Melting points were obtained with a Fisher–Johns
apparatus and are uncorrected. Optical rotations were
measured using a Optical Activity AA-1000 polarime-
ter. 1H and 13C spectra were recorded on a Bruker
AC-200 spectrometer with SiMe4 as internal standard
and using CDCl3 as solvent. Elemental analysis were
performed by the Canadian Microanalytical Service
Ltd. HPLC were determined with a Varian 9012 pump
and Varian 9050 UV–vis detector. Column chromatog-
raphy was performed using silica gel 60 of 230–400
mesh.
3
,
V (A )
Z
Dcalcd (g/cm3)
F(000)
Radiation
v (mm−1
)
Transmission factors
€ Oscillation range (=0.0) (°)
ꢀ Oscillation range (=90.0) (°)
2qmax (°)
Total reflections
Unique reflections
No. of variables
4.2. (2R,4S)-2-Phenyl-3-(carbobenzyloxy)-4-methyl-4-
[(3-methoxy)carbonyl-(1E)-propenyl]oxazolidin-5-one, 5
Rmerge
R, Rw (on F2, all data)
Goodness-of-fit
To a solution of potassium hexamethyldisilazide (0.5 M
in toluene 26.7 mL, 13.36 mmol) at −78°C was added a
solution of 4 (3.78 g, 12.15 mmol) in THF (20 mL)
dropwise. After the mixture was stirred for 45 min at
Max D/| (final cycle)
Residual density (e/A )
3
,
0.40, −0.35
R=ꢀꢁꢁFo ꢁ−ꢁFc ꢁꢁ/ꢀꢁFo ꢁ, Rw=ꢀ((Fo2−Fc2)2/ꢀw(Fo2)2)0.5
.
2
2
2

596
H. Pajouhesh et al. / Tetrahedron: Asymmetry 14 (2003) 593–596
−78°C a solution of methyl (E)-3-bromopropenoate
(2.20 g, 13.36 mmol) in THF (5 mL) at −78°C. The
resultant mixture was stirred for a further 45 min and
then quenched with sat. NH4Cl (20 mL). The aqueous
layer was extracted with Et2O (2×30 mL) and the
combined organic layers were dried over MgSO4 and
evaporated to give the crude product as an oil. The oil
was chromatographed on silica (hexanes:ethyl acetate
4:1) to give 3.0 g (62%) of 5. [h]2D5=+149 (c 0.18
ing solution was stirred at rt for 1.5 h. The solvent was
evaporated and the residue redissolved in H2O and
evaporated again. The residue was dissolved in anhy-
drous ethanol (10 mL) and propylene oxide added (5
ml). The resulting mixture was briefly heated under
reflux and the solvents removed in vacuo. The residue
was dissolved in a small amount of H2O and passed
through a reversed-phase C18 column using H2O as
eluent. The product was recrystallized from H2O to give
pure 1 (420 mg, 85%). [h]2D5=+77 (c 0.7, H2O); 1H
NMR (D2O) l 1.0–1.09 (m, 1H), 1.11–1.12 (m, 1H),
1.24 (s, 3H), 1.62–1.7 (m, 1H), 1.78–1.83 (m, 1H). Anal.
calcd for C7H11NO4·H2O: C, 43.98; H, 6.85; N, 7.33.
Found: C, 43.97; H, 6.83; N, 7.35%.
1
CHCl3); H NMR, (CDCl3) l 1.98 (s, 3H), 3.78 (s, 3H),
4.95–5.20 (br. m, 2H), 6.1 (d, J=15.9 Hz, 1H), 6.5 (s,
1H), 7.0 (d, J=15.9 Hz, 1H), 7.2–7.4 (m, 10H). Anal.
calcd for C22H21NO6: C, 66.83; H, 5.35; N, 3.54.
Found: C, 66.80; H, 5.36; N, 3.54%.
4.3. Cyclopropanation adduct, 6
References
To a solution of 5 (2.0 g, 5.06 mmol), Pd(OAc)2 (57 mg,
0.25 mmol) and ether (65 mL) was added in a dropwise
manner, a solution of CH2N2 in ether (100 mL) at 0°C.
The mixture was stirred at rt. overnight, filtered and
evaporated to give an oily residue which was chro-
matographed on silica (hexanes:ethyl acetate 4:1) to
give pure 6 as a solid (1.4 g, 67%). mp 125–127°C.
1. Curry, K.; Peet, M. J.; McLennan, H.; Magnuson, D. S.
K. J. Med. Chem. 1988, 31, 864–867.
2. Shinozaki, H.; Ishida, M.; Shimamoto, K.; Ohfune, Y.
Br. J. Pharmacol. 1989, 98, 1213–1224.
3. Lund, T. M.; Madsen, U.; Ebert, B.; Jørgensen, F. S.;
Krogsgaard-Larsen, P. Med. Chem. Res. 1991, 1, 136–
141.
4. Knopfel, T.; Kuhn, R.; Allgeier, H. J. Med. Chem. 1995,
38, 1417–1426.
5. Conn, P. J.; Pin, J. P. Ann. Rev. Pharmacol. Toxicol.
1997, 37, 205–237.
6. Shimamoto, K.; Ishida, M.; Shinozaki, H.; Ohfune, Y. J.
1
[h]2D5=+84 (c 0.1, CHCl3); H NMR (CDCl3) l 0.95
(ddd, J=8.7, 6.7, 5.1 Hz, 1H), 1.2 (dt, J=9.7, 5.1 Hz,
1H), 1.81 (s, 3H), 2.18 (dt, J=8.7, 4.8 Hz, 1H), 2.4 (br.
s, 1H), 3.67 (s, 3H), 5.1 (br. s, 2H), 6.43 (s, 1H),
7.2–7.43 (m, 10H). Anal. calcd for C23H23NO6: C,
67.47; H, 5.66; N, 3.42. Found: C, 67.44; H, 5.66; N,
3.45%.
Org. Chem. 1991, 56, 4167–4176.
7. Jane, D. E.; St. J. Jones, P. L.; Pook, P. C. K.; Salt, T.
E.; Sunter, D. C.; Watkins, J. C. Neuropharmacology
1993, 32, 725–727.
4.4. (2S,3S,4S)-N-Carbobenzyloxy-2-methyl-2-(car-
boxycyclopropyl)glycine, 7
8. Monn, J. A.; Valli, M. J.; Massey, S. M.; Wright, R. A.;
Salhoff, C. R.; Johnson, 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.
J. Med. Chem. 1997, 40, 528–537.
9. Hayashi, Y.; Tanabe, Y.; Aramori, I.; Masu, M.; Shi-
mamoto, K. Br. J. Pharmacol. 1992, 107, 539–543.
10. (a) Ma, D.; Ma, Z.; Jiang, J.; Yang, Z.; Zheng, C.
Tetrahedron: Asymmetry 1997, 8, 889–893; (b) Watkins, J.
C; Jane, D. E. Int. Pat. Applic. PCT/GB94/02690, 1995.
11. (a) Akaji, K.; Kuriyama, N.; Kiso, Y. J. Org. Chem.
1996, 61, 3350–3357; (b) Karady, S.; Amato, J. S.; Wein-
stock, M. Tetrahedron Lett. 1984, 25, 4337–4340.
12. Altomare, A.; Cascarano, M.; Giacovazzo, C.;
Guagliardi, A. J. Appl. Crystallogr. 1993, 26, 343.
13. Seebach, D.; Irmwinkelried, R.; Weber, T. Modern Syn-
thetic Methods; Springer: Berlin, 1986; Vol. IV, p. 125.
To a stirred solution of 6 (1.18 g, 2.88 mmol) in
THF/H2O (3:1, 40 mL) was added LiOH·H2O (484 mg,
11.52 mmol) at 4°C. The mixture was stirred overnight
at 25°C and then 5% aqueous NaHCO3 added (200
mL). The aqueous layer was washed with hexanes and
acidified to pH 2 with 2 N HCl. The resulting solution
was extracted with EtOAc (3×50 mL) and the organic
solution dried over MgSO4 and evaporated to give 7 as
1
an oil (0.89 g, quant.); H NMR (CDCl3) l 1.02 (m,
1H), 1.18 (dt, J=9.7, 5.1 Hz, 1H), 1.4 (s, 3H), 1.88 (br.
m, 1H), 2.0 (br. m, 1H), 5.0 (s, 2H).
4.5. (2S,3S,4S)-2-Methyl-2-(carboxycyclopropyl)glycine,
1
To a solution of 7 (0.87 g, 2.83 mmol) in CH2Cl2 (3
mL) was added 32% HBr in AcOH (5 mL). The result-