Synthesis and preliminary biological evaluation of (2S,1′R,2′S) - and (2S,1′S,2′R)-2-(2′-phosphonocyclopropyl)glycines, two novel conformationally constrained L-AP4 analogues

Article abstract of DOI:10.1016/j.bmcl.2005.09.014

Two novel constrained L-AP4 analogues, (2S,1′R,2′S)- and (2S,1′S,2′R)-2-(2′-phosphonocyclopropyl)glycines (7) and (8), were synthesized and evaluated as mGluR ligands. Compound 7 showed to be a group III mGluRs agonist with micromolar activity.

Full text of DOI:10.1016/j.bmcl.2005.09.014

Bioorganic & Medicinal Chemistry Letters 16 (2006) 196–199
Synthesis and preliminary biological evaluation of (2S,10R,20S)-
and (2S,10S,20R)-2-(20-phosphonocyclopropyl)glycines, two novel
conformationally constrained L-AP4 analogues
Laura Amori,a Michaela Serpi,a Maura Marinozzi,a Gabriele Costantino,a
Monica Gavilan Diaz,a, Mette Brunsgaard Hermit,b Christian Thomsenb
and Roberto Pellicciaria,*
a
`
Dipartimento di Chimica e Tecnologia del Farmaco, Universita di Perugia, Via del Liceo, 1-06123 Perugia, Italy
bH. Lundbeck A/S, 9 Ottilavej, DK-2500 Valby, Copenhagen, Denmark
Received 18 October 2004; revised 5 September 2005; accepted 8 September 2005
Available online 5 October 2005
Abstract—Two novel constrained L-AP4 analogues, (2S,10R,20S)- and (2S,10S,20R)-2-(20-phosphonocyclopropyl)glycines (7) and
(8), were synthesized and evaluated as mGluR ligands. Compound 7 showed to be a group III mGluRs agonist with micromolar
activity.
Ó 2005 Elsevier Ltd. All rights reserved.
Among the strategies aimed at finding new potent and
selective ligands for glutamate receptors, the bioisosteric
replacement of L-glutamate x-carboxylate function has
been a particularly fruitful one.1 The substitution of this
moiety for the phosphonate one in compounds charac-
terized by an appropriate intermediate chain and by
the D-configuration at the amino acidic moiety, such
as in D-AP5 (1, Chart 1) and D-AP7 (2), has given access
to the first generation of potent and selective NMDA
competitive antagonists.2 More recently, another phos-
phonate analogue of L-glutamate, L-AP4 (3), has proved
to be a potent agonist for the group III family of metab-
otropic glutamate receptors (mGluRs) and as such an
important tool for its pharmacological characteriza-
tion.3 The lack of selectivity of L-AP4 (3) towards the
array of group III subtypes (mGluR4, 6, 7, and 8),
however, gives interest to the search for new potent,
subtype selective group III ligands. Conformational
freezing of L-AP4 (3), in this regard, is a route that
has been followed with some success.
(+)-PPG (4), an L-AP4 analogue in which the phospho-
nate and the glycine moieties are kept in a collinear
*
Corresponding author. Tel.: +39 0755855120; fax: +39
0755855124; e-mail: rp@unipg.it
Chart 1. Selected L-Glu analogs as excitatory amino acid receptors
ligands.
Present address: Departamento de Quimica Organica, Facultad de
Farmacia, Campus de Cartuja s/n, 18071 Granada, Spain.
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.09.014

L. Amori et al. / Bioorg. Med. Chem. Lett. 16 (2006) 196–199
197
fashion by the phenyl spacer, is a potent mGluR4 ago-
nist (EC50 = 3.2 lM), also active at mGluR7
(EC50 = 48 lM),4 reported to be endowed with neuro-
protective properties5 and anxiolytic- and antidepres-
sant-like effects.6 The introduction of a cyclopropyl
ring in 2,3 position of the skeleton of L-AP4 (3) as in
(Z)- and (E)-( )-2-amino-2,3-methano-4-phosphonobu-
tanoic acids (5 and 6, respectively) has been shown to
give group III agonists in which the potency of the par-
ent is retained.7 Compound 5 with an EC50 of 0.58 lM is
the most potent mGluR4a agonist so far reported.
Although these derivatives have been instrumental in
producing pharmacophore models for group III
agonists,8,9 a clear picture of the steric and conforma-
tional requirements for the selective interaction of phos-
phonate analogue with individual group III mGluR
subtypes is still missing.
In the frame of our continuing interest in the develop-
ment of chemical tools for the characterization of gluta-
matergic pathways, we have undertaken the synthesis of
two novel conformationally constrained analogues of L-
AP4 (3), namely trans-(2S,10R,20S)- and (2S,10S,20R)-2-
(20-phosphonocyclopropyl) glycines (PCG-1, 7) and
(PCG-2, 8) in which the pharmacophoric groups, the
glycine moiety and the phosphonate group, are partially
rigidified into extended disposition according to the
pharmacophore models for group III agonists.8,9 The
two new compounds are phosphono analogs of carboxy-
cyclopropylglycines (L-CCGs), conformationally con-
strained analogs of L-glutamate. (2S,10S,20S)-L-CCG-I
(15), in particular, is a potent mGluR agonist with sig-
nificant activity also at group III subtypes. It must be
mentioned that compounds 7 and 8 have previously
and independently been prepared with a different
chemistry at Novartis (Basel) and tested at the human
mGlu4 receptor. The biological activity is consistent
with our present results. The NovartisÕs data are still
unpublished.18
Scheme 1. Reagents and conditions: (a) LDA, THF, ꢀ78–0 °C; (b)
PCC, CH2Cl2, rt; (c) (i) R-(ꢀ)-a-phenylglycinol, MeOH, rt; (ii)
TMSCN; (iii) mplc; (d) (i) Pb(OAc)4, CH2Cl2-MeOH, 0 °C; (ii) 6 N
HCl, reflux; (iii) Dowex 1X8-200.
12 and 13 thus obtained were then submitted to oxida-
tive cleavage with lead tetraacetate (CH2Cl2–MeOH,
0 °C, 20 min), acidic hydrolysis (6 N HCl), and ion-ex-
change resin chromatography (Dowex 1X8-200, 1.5 N
acetic acid) to afford the corresponding amino acids
(2S,10R,20S)-PCG-1 (7) and (2S,10S,20R)-PCG-2 (8) in
42% and 46% yields, respectively.13
The absolute stereochemistry of the (phosphonocyclo-
propyl)glycines 7 and 8 was tentatively assigned on the
basis of NMR analyses. 1H–1H coupling constants
0
(J2-1 ), which are similar in both distereoisomers 7 and
8 (J = 9.86 Hz and J = 9.98 Hz, respectively), suggested
that the relative orientation of H(2)–H(10) was antiperi-
The synthesis and the preliminary biological properties
of 7 and 8 are reported herein (Scheme 1). trans-Diethyl
2-(hydroxymethyl) cyclopropyl phosphonate (10) was
obtained from the readily available diethyl c,d-epox-
ybutylphosphonate10 (9) via a basic catalyzed intramo-
lecular epoxide opening reaction.11 While in the
original report nBuLi was used as the base, in our hands
LDA (1.1 equiv in THF, ꢀ78 °C) proved to be the base
of choice affording the desired 10 in 59% yield. Oxida-
tion of 10 by PCC in dichloromethane at room temper-
ature led in 63% yield to the racemic aldehyde 11 which
was submitted to the condensation with R-(ꢀ)-a-phenyl-
glycinol (MeOH, rt, 24 h). The subsequent addition of
trimethylsilyl cyanide to the Schiff base thus formed
(TMSCN, 0 °C then rt, 24 h) afforded the mixture of
(2S,10R,20S)- and (2S,10S,20R)-aminonitriles 12 and 13
along with minor amounts of the (2R,10R,20S)- and
(2R,10S,20R)-diastereoisomers
(2S/2R = 75:25
by
HPLC).12 Separation of the two more abundant a-
aminonitriles by flash chromatography (ethyl acetate/
methanol, 95:5) and subsequent mplc (acetone/n-hexane,
70:30) afforded the desired isomers 12 and 13 in 10% and
15% yields, respectively. The two (2S)-a-aminonitriles
Figure 1. Newman projections of the proposed preferred conforma-
tions of L-CCG-I (15), L-CCG-II (14), PCG-1 (7) and PCG-2 (8).

198
L. Amori et al. / Bioorg. Med. Chem. Lett. 16 (2006) 196–199
Table 1. EC50 (lM) values for PCG-1 (7) and PCG-2 (8)
Compound
Group I
Group II
Group III
mGluR1
mGluR5
mGluR2
mGluR4
mGluR6
13
mGluR7
PCG-1 (7)
PCG-2 (8)
L-Glutamate
L-AP4 (3)
>1000
>1000
7.3 1.0
>1000
>1000
>1000
1.2 0.3
>1000
>1000
>1000
9.4 3.0
49 15
3
700 140
>1000
>1000
46 23
7.6 0.5
0.34 0.07
26
>1000
2
6.1 1.5
0.51 0.14
188 79
Functional data were obtained from CHO cells expressing rmGluR1, rmGluR5 or rmGluR6 or from BHK cells expressing hmGluR2, hmGluR7 or
rmGluR4 using previously described methods.3,12b,16 Potencies were determined by: (i) measuring intracellular calcium concentration (group I
mGluRs), (ii) GTP(c)35S stimulation assay (mGluR2) and (iii) measuring cAMP formation (group III mGluRs). The data shown are means SEM
of at least three independent experiments performed in triplicate.
experiments (rmGluR4 and rmGluR8).16 As summarized
in Tables 1 and 2, the new derivatives 7 and 8 show activ-
Table 2. Binding affinity values expressed as Ki (lM), for PCG-1 (7)
and PCG-2 (8)
ity as group III mGluRs agonists albeit with significant
differences in potency and subtype. In particular, PCG-
1 (7) able to displace labeled L-AP4 from rat mGluR4
with a Ki of 7.4 lM can be considered an acceptably po-
tent agonist for this receptor subtype. Reduction of the
conformational flexibility of L-AP4 (3) through insertion
of a cyclopropyl ring in 3,4 positions yielded the two
trans-cyclopropyl derivatives 7 and 8. Evaluated for their
ability to interact with a set of glutamate receptors, both
derivatives 7 and 8 showed to be inactive at ionotropic
glutamate receptors (NMDA, KA, and AMPA) and at
prototypic group I and group II mGluR subtypes. On
the contrary, both compounds activated the four group
III mGluR subtypes with different potencies, with
PCG-1 (7) always more potent than PCG-2 (8). The most
active isomer PCG-1 (7), however, is about 10-fold less
potent than the parent L-AP4 (3) or the other cyclopro-
pyl-AP4 derivative 5 at all the group III mGluR sub-
types. While confirming the need for an extended
disposition of pharmacophoric moieties for group III
recognition, our results indicated that the cyclopropyl
ring in 3,4 position of the L-AP4 skeleton, induces a con-
formational disposition which is accepted but not opti-
mal for interacting with the conserved group III
mGluR binding pockets. It can also be speculated that
there is a need for the phosphonate moiety of a certain
degree of conformational flexibility for a proper interac-
tion with group III receptors. This evidence is in contrast
with the experience gained for other glutamate receptors
(NMDA-Rs and group II mGluRs, in particular), where
restriction of the L-glutamate flexibility by insertion of
cyclopropyl moiety yielded very potent and highly selec-
tive compounds.
Compound
mGluR4
mGluR8
63 16
PCG-1 (7)
PCG-2 (8)
L-Glutamate
L-AP4 (3)
7.4 2.7
53 15
54
4
3.4 0.5
0.16 0.02
3.4 1.1
1.9 0.3
[3H]L-AP4 and [3H]LY34149517 binding experiments were performed
on rmGluR4 and rmGluR8 receptor-expressing cell membranes (BHK
cells) using a SPA assay.16 Data are means SEM of at least three
independent experiments performed in triplicate.
planar in both diastereoisomers. This evidence was also
confirmed by NOESY experiments, which did not show
any NOE between these two protons in both diastereo-
isomers 7 and 8. Taking into account the structural anal-
ogy between the (phosphonocyclopropyl)glycines 7 and
8, and the two trans-2-(20-carboxycyclopropyl)glycine
diastereoisomers [L-CCG-II (14) and L-CCG-I (15)],14
whose absolute configurations were previously unam-
biguously defined,15 a comparative study on the evalua-
tion of Ôc-gaucheÕ effects played by the amino- and
carboxylic groups, on C(20) and C(30) in the 2-(20-carb-
oxycyclopropyl)glycines (14 and 15) and 2-(20-phospho-
nocyclopropyl)glycines (7 and 8), may be instrumental
in disclosing the absolute stereochemistries of our new
compounds. In particular, assuming also in the case of
14 and 15, an antiperiplanar disposition between H(2)
and H(10) and considering that an amino group has a
larger shielding or Ôc-gaucheÕ effect on the corresponding
carbon atom than does the carboxylic group,15 the
diastereoisomer showing the C(20) chemical shift
more shielded is endowed with 2S,10S,20S configuration
as depicted in Figure 1A. On the contrary, the
2S,10R,20R-isomer, as depicted in Figure 1B, having
the amino group in a gauche relationship with C(30),
shows the chemical shift of this carbon atom more
shielded than in the other isomer. Analogously, the dia-
stereoisomer 7, showing C(20) up-fielded and C(30)
down-fielded in comparison with compound 8, can be
assigned with the 2S,10R,20S-configuration (Fig. 1C),
in analogy with L-CCG-I (15). On the contrary, PCG-
2 (8) will be endowed with 2S,10S,20R-configuration
(Fig. 1D).
References and notes
1. For a review, see: Brauner-Osborne, H.; Egebjerg, J.;
Nielsen, E.; Madsen, U.; Krogsggaard-Larsen, P. J. Med.
Chem. 2000, 43, 2609.
2. (a) Watkins, J. C.; Evans, R. H. Neurosci. Lett. 1981, 21,
77; (b) Olverman, H. J.; Jones, A. W.; Mewett, K. N.;
Watkins, J. C. Neuroscience 1988, 26, 17.
3. Thomsen, C.; Kristensen, P.; Mulvihill, E.; Haldeman, B.;
Suzdak, P. D. Eur. J. Pharmacol. 1992, 227, 361.
4. Gasparini, F.; Inderbitzin, W.; Francotte, E.; Lecis, G.;
Richert, P.; Gragic, Z.; Kuhn, R.; Flor, P. J. Bioorg. Med.
Chem. Lett. 2000, 10, 1241.
The new derivatives 7 and 8 were evaluated as potential
mGluRs ligands by functional assays (rmGluR1, rmG-
luR5, hmGluR2, rmGluR4, and hmGluR7) and binding

L. Amori et al. / Bioorg. Med. Chem. Lett. 16 (2006) 196–199
199
5. Sabelhaus, C. F.; Schroder, U. H.; Breder, J.; Henrich-
Noack,P.;Reymann,K.G.Br.J.Pharmacol.2000,131,655.
6. Palucha, A.; Tatarczynska, E.; Branski, P.; Szewczyk, B.;
Wieronka, J. M.; Klak, K.; Chojnacka-Wojcik, E.; Now-
ak, G.; Pilc, A. Neuropharmacology 2004, 46, 151.
7. Johansen, P. A.; Chase, L. A.; Sinor, A. D.; Koerner, J. F.;
Johnson, R. L.; Robinson, M. B. Mol. Pharmacol. 1995, 48,
140.
20-CH and 30-CH2), 1.43 (1H, m, 10-CH), 3.37 (1H, d,
J = 9.86 Hz, 2-CH); 13C NMR (D2O, 50 MHz) d 8.9 (d,
JC–P = 5.0 Hz), 12.2 (d, JC–P = 183.5 Hz), 16.9 (d,
JC–P = 3.5 Hz), 56.3, 171.1; 31P NMR (D2O, 50 MHz) d
20
24.6; ½aꢁD +81.9 (c 1.0, H2O).
Compound 8: 1H NMR (D2O, 200 MHz) d 0.94 (2H,
m, 30-CH2), 1.16 (1H, m, 20-CH), 1.55 (1H, m, 10-CH),
3.37 (1H, d, J = 9.98 Hz, 2-CH); 13C NMR (D2O,
50 MHz) d 8.7, 12.4 (d, JC–P = 184.9 Hz), 16.7, 56.0,
8. Amori, L.; Costantino, G.; Marinozzi, M.; Pellicciari, R.;
Gasparini, F.; Flor, P. J.; Kuhn, R.; Vranesic, I. Bioorg.
Med. Chem. Lett. 2000, 10, 1447.
20
170.7; 31P NMR (D2O, 50 MHz) d 25.28; ½aꢁD +11.8
(c 1.0, H2O).
9. Bessis, A. S.; Jullian, N.; Coudert, E.; Pin, J.-P.; Acher, F.
Neuropharmacology 1999, 38, 1543.
`
10. Savignac, P.; Breque, A.; Mathey, F. Synth. Commun.
1979, 2, 487.
11. Hah, H. J.; Gil, J. M.; Oh, D. Y. Tetrahedron Lett. 1999,
40, 8235.
12. (a) Chakrabotry, T. K.; Hussain, K. A.; Reddy, V. G.
Tetrahedron 1995, 33, 9179; (b) Pellicciari, R.; Marinozzi,
M.; Natalini, B.; Costantino, G.; Luneia, R.; Giorgi, G.;
Moroni, F.; Thomsen, C. J. Med. Chem. 1996, 39, 2259.
14. Pellicciari, R.; Curini, M.; Natalini, B.; Ceccherelli, P. In
IX International Symposium on Medicinal Chemistry,
Berlin, 1986; p 118.
15. Shimamoto, K.; Ishida, M.; Shinozaki, H.; Ohfune, Y.
J. Org. Chem. 1991, 56, 4167.
16. Mathiesen, J. M.; Svendsen, N.; Bra¨uner-Osborne, H.;
Thomsen, C.; Ramirez, M. T. Br. J. Pharmacol. 2003, 138,
1026.
17. Wright, R. A.; Arnold, M. B.; Wheeler, W. J.; Ornstein, P.
L.; Schoepp, D. D. Naunyn Schmiedebergs Arch. Pharma-
col. 2000, 362, 546.
13. Analytical data of the final compounds
7
and 8.
1
Compound 7: H NMR (D2O, 200 MHz) d 0.99 (3H, m,
18. Gasparini, F. Personal communication.