New strategies towards proline derivatives as conformationally constrained arginine analogues

Article abstract of DOI:10.1016/S0040-4039(02)00424-0

The cyano derivatives of 3-functionalized dehydroprolines led to arginine semi-constrained analogues.

Full text of DOI:10.1016/S0040-4039(02)00424-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 3243–3246
New strategies towards proline derivatives as conformationally
constrained arginine analogues
Nadia Pellegrini, Martine Schmitt,* Se´bastien Guery and Jean-Jacques Bourguignon
Laboratoire de Pharmacochimie de la Communication Cellulaire, UMR 7081 CNRS/ULP, Universite´ Louis Pasteur,
Faculte´ de Pharmacie, 74 Route du Rhin, 67401 Illkirch Cedex, France
Received 21 December 2001; accepted 28 February 2002
Abstract—The cyano derivatives of 3-functionalized dehydroprolines led to arginine semi–constrained analogues. © 2002 Elsevier
Science Ltd. All rights reserved.
The need for selective ligands acting at neuropeptides
receptors constitutes an important challenge for medici-
nal chemists. Starting from the structures of the
endogenous ligand, several studies highlighted the spe-
cific role played by an arginine residue (neuropeptide
FF, neuropeptide Y, thrombin). In addition, substi-
tuted arginine derivatives have been described as potent
ligands of these receptors (Scheme 1).1–3
of other amino acids, its presence restricts its conforma-
tional flexibility, which may efficiently modify binding
to its target. In addition, this peptidomimetic com-
pound is expected to be more stable towards metabolic
enzymes. The use of proline as a template in order to
introduce conformational restrictions has already been
described for norleucine, phenylalanine, tyrosine, aspar-
tic acid, glutamic acid and glutamine.4,5
Several ornithine and arginine semi-rigid analogues
have been described in the literature6–9 in particular
trans-3-substituted proline.10 In this latter case, the
authors accomplished the synthesis in a seven-step
sequence starting from a non commercially available
a-unsaturated lactam. In addition, only the trans iso-
mer could be obtained.
Proline is a unique natural a-aminoacid inducing spe-
cific electronic and geometric features, when compared
to alanine. When such a ring is included in the structure
We now describe a convenient and mild preparation of
cis- and trans-N-acyl 3-substituted prolines 1 as semi-
rigid BIBP 32362 or RF21 derivatives (Scheme 1). This
method is based on a catalytic hydrogenation of the
double bond of a 3-substituted N-acyl pyrrole 2 as
outlined in Scheme 2. This general strategy has been
earlier successfully used by us for the preparation of
3-acetic acid indoline-2-carboxylic acid and of 3-phenyl
indoline-2-carboxamides, as semi-rigid glutamic acid
and phenylalanine mimetics, respectively.11
Retrosynthetically, the dihydropyrrole 2 could be pre-
pared from 3-pyrrolidinone 3 by a two-step procedure
including successively a Wittig reaction and a reduction
of the corresponding ester into an alcohol (Scheme 2).
Scheme 1.
Keywords: arginines; constrained analogues; Mitsunobu; proline;
pyrrolidine.
* Corresponding author. Tel.: +33 390 24 42 31; fax: +33 390 24 43
10; e-mail: schmitt@pharma.u-strasbg.fr
Following a literature procedure, the b-keto pyrrolidine
3 was prepared by intramolecular Dieckman-type
cyclizations of several N-acyl glycinates (Scheme 3).
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)00424-0

3244
N. Pellegrini et al. / Tetrahedron Letters 43 (2002) 3243–3246
Scheme 2.
Scheme 3.
Table 1.
Entry
Cpd 4
X
R2
R3
3 (%)
5 (%)
a
a
1
2
3
4
a
b
c
OEt
SEt
-N(OMe)Me
OMe
Boc
COOEt
Boc
Et
Et
Et
t-Bu
77
73
45
–
–
b
d
Bz
a In our hands, regioisomers leading from 4a could not separated.
b Recovery of significant amount of N-benzoyl glycinate as the result of retro-Michael addition.
Rapoport et al.12 described two separable regioisomers
(3 and 5) which could be obtained by a similar
intramolecular cyclization, but in our hands, the sepa-
ration failed (entry 1 in Table 1). The use of a semi-
thiol ester,13 or of the Weinreb amide,14 as more
electrophilic ester equivalents, was found to be more
interesting (entries 2 and 3). We have observed that the
introduction of a more bulky ester (R3=t-Bu, R2=Bz,
entry 4) altered the regiochemical selectivity. In this
case, 3 was obtained in 45% yield accompanied by 55%
of starting N-benzoyl glycinate as the result of retro-
Michael addition.
electron-withdrawing group, to favor migration of the
double bond into the pyrrolidine ring (Scheme 4).
The cyano ester 712 was first protected as a N-Boc
derivative. The enol obtained by cyclization using tBu-
OK was trapped with TMSCl. After removal of TMS,
the resulting 3-pyrrolidinone reacted with [(methoxycar-
bonyl)methylene]triphenylphosphorane to afford the
expected dihydropyrrole 9 in good yield. It is notewor-
thy that both the N-Boc acyl and cyano groups
decreased H-2 proton acidity, which led to the endo-
cyclic migration of the double bond, as shown by the
NMR spectrum.15 In addition, as endocyclic migration
could be efficiently controlled, this approach can be
generalized to other 3-substituted proline groups as
original semi-constrained a-aminoacids.
Treatement of the b-ketoester 3 with Wittig reagent led
to the diester 6 in 70% yield. In contrast to the indoline
series,11 this compound still presents an exocyclic dou-
ble bond, which could not allow a C-2 and C-3 stereo-
control (Scheme 3). Therefore, we planned to replace
the ester in position 2 by a cyano group as a more
Reduction of the methyl ester using LiBH4 and cata-
lytic hydrogenation of the dihydropyrrole ring afforded

N. Pellegrini et al. / Tetrahedron Letters 43 (2002) 3243–3246
3245
Scheme 4. (i) (Boc)O2, (ii) 1. t-BuOk, toluene then TMSCl, 2. H2O, 3. Ph3PCHCOOMe, toluene, (iii) LiBH4, THF, (iv)
H2/Pd/C/MeOH, (v) 1. HCl (g), MeOH, 2. PhCOCl/TEA/CH2Cl2, (vi) MeOH, HCl, quantitative.
the cis-3-substituted proline 11, characterized by an AB
system in its 1H NMR, and only one C-2 methine signal
in its 13C NMR spectrum. Simultaneous acidic methanol-
ysis of the cyano group and N-Boc deprotection followed
by direct N-acylation gave a mixture of two compounds
in 67% overall yield; they were identified as the alcohol
ester 13a and the corresponding lactone 12. This latter
compound generated quantitatively the open derivative
13a by methanolysis. NOE experiments confirmed the
already established cis configuration for 13a. A weak
NOE was also observed between H-2 and H-3.
anion as catalyst.18 The resulting amides 14a and 14b
have been separated by chromatography on silica.
Guanidylation of 14a and 14b with guanidine tri-Boc
using Mitsunobu conditions19 followed by removal of the
protecting groups yielded the guanidines 16a and 16b
[MS m/z 394.24 (M+H)+].
In conclusion, motivated by the pharmacological interest
we may expect for either the cis and the trans stereomers
16a and 16b as arginine semi-rigid derivatives, we devel-
oped an efficient method for their preparation. In addi-
tion functionalization of position 3 of proline combined
with an endocyclic double bond as an efficient method
for controlling the configuration at C-2 and C-3 of the
pyrrolidine ring will be very useful in the future for
building interesting alkaloid analogues.
Compound 13a could be epimerized in a 9:1 ratio into
the trans isomer 13b by means of 3.5 equiv. of LDA at
−78°C for 3 h16 (Scheme 5).17 Conversion of the methyl
ester into a N-benzylamide was carried out with cyanide
Scheme 5. (i) LDA/THF, (ii) PhCH2NH2/NaCN/MeOH, (iii) tri-Boc guanidine/Ph3P/DEAD/THF, (iv) TFA.

3246
N. Pellegrini et al. / Tetrahedron Letters 43 (2002) 3243–3246
10. Mamai, A.; Hughes, N. E.; Wurthmann, A.; Madalen-
References
goitia, J. S. J. Org. Chem. 2001, 66, 6483–6486.
11. Collot, V.; Schmitt, M.; Marwah, P.; Bourguignon, J.-J.
Heterocycles 1999, 51, 2823–2847.
1. Bourguignon, J.-J.; Collot, V.; Didier, B.; Laulin, G.;
Simonet, G. Proceedings of the XIVth International Sym-
posium on Medicinal Chemistry; Elsevier Scientific:
Amsterdam, 1997; pp. 35–44.
2. Aiglstorfer, I.; Hendrich, I.; Moser, C.; Bernhardt, G.;
Dove, S.; Buschauer, A. Bioorg. Med. Chem. Lett. 2000,
10, 1597–1600.
3. Matsoukas, J. M.; Panagiotopoulos, D.; Keramida, M.;
Mavromoustakos, T.; Yamdagni, R.; Wu, Q.; Moore, G.
J.; Saifeddine, M.; Hollenberg, M. D. J. Med. Chem.
1996, 39, 3585–3591.
4. Damour, D.; Pulicani, J.-P.; Vuilhorgne, M.; Mignani, S.
Synlett 1999, 6, 786–788.
5. Carpes, M. J. S.; Miranda, P. C. M. L.; Correia, C. R. D.
Tetrahedron Lett. 1997, 38, 1869–1872.
6. Webb, T. R.; Eigenbrot, C. J. Org. Chem. 1991, 56,
3009–3016.
7. Wang, Q.; Sasaki, N. A.; Potier, P. Tetrahedron 1998, 54,
15759–15780.
8. Zhang, R.; Mamai, A.; Madalengoitia, J. S. J. Org.
Chem. 1999, 64, 547–555.
12. Blake, J.; Willson, D. C.; Rapoport, H. J. Am. Chem.
Soc. 1964, 86, 5293–5299.
13. Yamada, Y.; Ishii, T.; Kimura, M.; Hosaka, K. Tetra-
hedron Lett. 1981, 22, 1353–1354.
14. Sibi, M. P.; Christensen, J. W.; Kim, S. G.; Eggen, M.-J.;
Stessman, C.; Oien, L. Tetrahedron Lett. 1995, 36, 6209–
6212.
15. 1H NMR 9: 3.39 (s, 2H, CH2COOMe); 6 (R2=Boc,
R3=Et) mixture of rotamers 4.74 (s, 0.63H, NCH-
COOEt), 4.82 (s, 0.37H, NCHCOOEt), 6.00 (s, 0.63H,
CꢀCH), 6.05 (s, 0.37H, CꢀCH).
16. Maeda, H.; Kraus, G. A. J. Org. Chem. 1997, 62, 2314–
2315.
17. Compounds 13a and 13b could not be separated. HPLC:
13a Rt1=14.43 min, 13b Rt2=14.76 min (Vydac, gradient
H2O/MeOH: 30 min [98/2-0/100], flow: 1 mL/min). 1H
NMR 13b: 2.5 (1H, m, H-3), 4.39 (1H, d, J=7.5 Hz,
H-2); 13a: 2.7 (1H, m, H-3), 4.72 (1H, d, J=8.1 Hz, H-2).
18. Ho¨gberg, T.; Stro¨m, P.; Ebner, M.; Ra¨msby, S. J. Org.
Chem. 1987, 52, 2036–2039.
19. Dodd, D. S.; Kozikowski, A. P. Tetrahedron Lett. 1994,
35, 977–980.
9. Zhang, R.; Mamai, A.; Madalengoitia, J. S. J. Org.
Chem. 1999, 64, 547–555.