Title - L`Information scientifique au service de la recherche

advertisement
Natural Product Synthesis
DOI: 10.1002/anie.200123456
Formal Synthesis of (+)-Catharanthine
Lionel Moisan, Pierre Thuéry, Marc Nicolas, Eric
Doris, and Bernard Rousseau
Since the discovery of vinblastine and vincristine in the late fifties,
Vinca alkaloids have turned out to be among the most powerful
drugs currently in use for the clinical treatment of cancer.1 Over the
past four decades, synthetic efforts1c,d led to the discovery of the
major hemi-synthetic drugs vinorelbine2 1 and, more recently,
vinflunine3 2. The industrial synthesis of 1 and 2 relies on the
biomimetic coupling of catharanthine 3 and vindoline 4,1 two
alkaloids that are extracted from the leaves of Madagascan
periwinkle (Catharanthus roseus, L. Don). Catharanthine is found in
minute amounts in the plant (about 0.0003 % of the dried leaves
mass)4 and no enantioselective synthesis has yet been reported,
although numerous racemic approaches have been developed.5 (+)Catharanthine has been prepared only once via resolution.6 In the
present paper we disclose the first formal synthesis of (+)catharanthine.
the isoquinuclidine. We chose as the target compound alcohol 24 in
which each stereocenter is irreversibly defined (Scheme 1).
Furthermore, in his elegant total synthesis of (±)-catharanthine,
Büchi demonstrated that 24 could easily be transformed into
catharanthine by introducing the carbomethoxy group in C-16 and
effecting dehydration of the tertiary alcohol.5a
In that context, our retrosynthetic plan was first to design a route to
optically active isoquinuclidine 17 and to couple it to the indolic
part. Since previously reported attempts to obtain optically active
isoquinuclidine by [4+2] cycloaddition were unsuccessful, 7 we
envisioned that the azabicyclo[2.2.2] system could be prepared by
ring closing metathesis of cis-2,5-dialkenyl-N-acylpiperidine 16.
Center C-14 would derive from the stereocontrolled Michael
addition of a vinyl group on cyclohexenone 10, the latter being
prepared in 5 steps starting from protected L-serine which is our
source of chirality.
O
N
N
H
N
N
H
CO2Me
(+)-Catharanthine
MeO2C
O
MeO2C
O
N H
Vinorelbine 1
OH
OCOMe
N
H CO2Me
Me
Vinflunine 2
16
N
H
21
MeO
CO2Me
(+)-Catharanthine 3
F
F
N H
OH
OCOMe
N
H CO2Me
Me
N H
N
(-)-Vindoline 4
OH
OCOMe
N
H CO2Me
Me
Catharanthine has three stereocenters C-14, C-16 and C-21 which
are integrated in the isoquinuclidine skeleton. Its structure is further
characterized by a seven membered C ring which links the indole to
[]
19
O
O
14
N
O
O
O
O
N
10
O
N
N
H
MeO2C
MeO
OH
O
N
17
N
H
MeO2C
MeO
N
H
21
16
Büchi's intermediate 24
H
N
N
14
Dr. L. Moisan, Dr. E. Doris, Dr. B. Rousseau
Service de Marquage Moléculaire et de Chimie Bioorganique
DSV/DBJC, CEA/Saclay
91191 Gif-sur-Yvette cedex (France)
Fax: (+33)169-08-79-91
E-mail: eric.doris@cea.fr
Dr. P. Thuéry
Service de Chimie Moléculaire
DSM/DRECAM, CEA/Saclay
91191 Gif-sur-Yvette cedex (France)
Dr. M. Nicolas
Laboratoires Pierre Fabre
16 rue Jean Rostand
81600 Gaillac (France)
[] This work is part of a collaboration between "CEA/Direction des
Sciences du Vivant" and "Pierre Fabre Médicament".
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
HO
OH
L-serine
NH2
Scheme 1. Retrosynthetic approach to Büchi’s intermediate.
Our synthesis started from Z-L-serine 5 (Scheme 2). Upon alkaline
treatment, 5 was cleanly converted to oxazolidinone 6 in 73% yield,
and was subsequently allylated to give 7 in 61% yield. Conversion
of 7 to the corresponding Weinreb amide 8 was performed in 76%
yield using standard EDCI-mediated peptide coupling conditions.
Measurement of the enantiomeric excess of 8 indicated little erosion
(<5%) over the last three steps.8 Recrystallization from Et2O/CH2Cl2
provided material of >99% ee. Addition of vinylmagnesium
bromide followed by careful quenching of the reaction using
anhydrous acetic anhydride provided enone 9 in 81% yield.
Cyclohexenone 10 was obtained in 87% yield upon treatment with
the second-generation Grubbs catalyst (G2). Rapid filtration of
crude 10 over a short pad of silica turned out to be the most efficient
procedure to prevent epimerization of the stereogenic center C-21
during purification. With cyclohexenone 10 in hand, we then had to
set stereocenter C-14 by conjugate addition of vinyl magnesium
bromide. The 1,4-nucleophilic attack of the vinyl group proceeded
smoothly in the presence of stoichiometric amount of CuI at –78°C
to afford a 90/10 mixture of diastereomers. As the hydrogen atom
borne by C-21 was proven to be labile, the carbonyl group of 11 was
directly protected into the corresponding dioxolan 12 in 70% yield
(two steps). At this stage, chromatography over silica permitted
isolation of the major diastereomer whose cis stereochemistry was
unambiguously established by X-ray crystallographic analysis.9 The
observed stereoselectivity of the 1,4-addition might be attributed to
the coordinating effect of the nitrogen lone pair which directs syn
attack of the vinyl group. With the desired cis isomer in hand, the
second olefinic side chain was then introduced by cleavage of the
oxazolidinone under alkaline conditions followed by protection of
the resulting secondary amine as a carbomethoxycarbamate.
1
Subsequent TPAP/NMO10 alcohol oxidation provided aldehyde 15
in 75% yield. Finally, methylenation of 15 was achieved in 83%
yield using Tebbe reagent.
O
CO2H
a
NH
HO
BnO
O
O
O
O
O
O
O
O
10
X
O
R
12
N
MeO2C
13 R=H
14 R=CO2Me
i
O
O
j
HO
N
O
11
N
8 R=N(OMe)Me
9 R=vinyl
d
h
O
14
N
f
O
O
O
g
e
N
R
6 R=H
7 R=allyl
b
21
R
O
c
N
O
5
O
CO2H
CO2Me
MeO2C
N H O
H
RuLn
H
O
O

H
LnRu
16 O
R
N
k
N
l
N
15 X=O
16 X=CH2
17 R=CO2Me
18 R=H
m
O
O
Scheme 2. Synthesis of optically active isoquinuclidine 18. (a) NaOH,
H2O, rt, 73%; (b) Allyl iodide, NaH, DMF, 61%; (c) HCl.HN(OMe)Me,
EDCI, DIPEA, CH2Cl2, 76%; (d) 1) vinylmagnesium bromide, THF,
0°C; 2) acetic anhydride, 81%; (e) G2 10 mol%, CH2Cl2 reflux, 87%;
(f) vinylMgBr, CuI, THF, -78°C; (g) ethylene glycol, CH2Cl2, PTSA,
70% 2 steps; (h) NaOH, H2O/MeOH, reflux, 18h; (i) ClCO2Me,
NaHCO3, THF, 81%, 2 steps; (j) TPAP, NMO, CH2Cl2, 75%; (k) Tebbe
reagent, Et2O, -10°C, 83%; (l) G2 10 mol%, CH2Cl2, reflux 20 h, 84%;
(m) MeLi, Et2O, 0°C, 1h, 74%.
The stage was thus set for the intramolecular ene-ene metathesis.
Despite prolific use of RCM for the synthesis of bridged azabicyclic
structures11 there is no example of azabicyclo[2.2.2]alkene systems.
Indeed, RCM is, in most cases, performed on cis-2,6-dialkenyl-Nacylpiperidines since 1,3 allylic strain12 favors the requested axiallike orientation of the alkenyl side chains. In our case, the cis-2,5dialkenyl-N-acylpiperidine 16 requires a more challenging boat
conformation to bring the 2- and 5- substituents into an axial
orientation. Nevertheless, upon treatment with 10 mol% of the
second-generation of Grubbs catalyst in refluxing CH2Cl2,
piperidine 16 was smoothly converted to isoquinuclidine 17 in 84%
yield and >99% ee.
O
HN
N
a
O
N
H
O
18
O
c
N
N
H
X
19 X=O(CH2)2O
20 X=O
b
d
O
21
O
N
N
H
e
22 R=Vinyl
23 R=Ethyl
N
f
R
OH
N
H
OH
24
N
ref 5a
N
H
CO2Me
(+)-Catharanthine
Scheme 3. Completion of the synthesis. (a) EDCI, CH2Cl2, 94%; (b)
APTS, H2O, acetone, reflux, 96h, 88%; (c) 1) (CH3CN) 2PdCl2, AgBF4,
CH3CN, rt to 70°C, 18h, 2) MeOH, NaBH4, 0°C; (d) vinylMgBr, THF,
0°C; 48% 2 steps; (e) H2, PtO2, THF, rt, 70%; (f) AlCl3-LiAlH4, THF,
0°C, 77%.
Isoquinuclidine 17 was subsequently coupled to the indolic moiety.
The secondary nitrogen atom was deprotected using MeLi in 74%
yield and reacted with 3-carboxylic ethylindole using EDCI (94%
yield) (Scheme 3). The ketal protecting group was then removed
under acidic condition in 88% yield and the C ring closure was
performed using the PdII/AgI mixed-metal-mediated cyclization
developed by Trost and co-workers.5c Controlled reduction of the
C-Pd intermediate with NaBH4 in MeOH afforded ketone 21 as
the major product together with little amount of the corresponding
alcohol (<5%). Addition of vinyl magnesium bromide to the
carbonyl gave tertiary alcohol 22 as a single diastereomer.
Subsequent hydrogenation over PtO2 furnished the ethyl side chain
(23) in 70% yield. As previously observed by others, direct attempts
to introduce the side chain using ethyl magnesium bromide were
unsuccessful due to competing reduction of the ketone by hydride
shift from the Grignard reagent.5a Finally, reduction of the amide to
the corresponding amine was performed using AlCl3-LiAlH4 to give
the key intermediate 24. However, in our case, Büchi’s intermediate
24 is obtained in virtually optically pure form with an enantiomeric
excess higher than 99%, and can lead in 5 steps to (+)-catharanthine.
In conclusion, we disclose here the first formal synthesis of (+)catharanthine, the crucial intermediate for the industrial synthesis of
the major antitumor drug vinorelbine. Our approach starts from the
naturally occurring (L)-serine and is based on two key steps, i.e. the
stereocontrolled cis addition of a vinyl group on cyclohexenone 10
and an unprecedented RCM on a cis-2,5-dialkenyl-N-acylpiperidine
system. The approach developed toward this synthetic challenge
also opens a general route for the synthesis of optically active
isoquinuclidines.
Received: ((will be filled in by the editorial staff))
Published online on ((will be filled in by the editorial staff))
Keywords: alkaloids. metathesis. natural products. synthetic
methods.
[1]
(a) Antitumor Bisindole Alkaloids from Catharanthus Roseus (L.). The
Alkaloids, vol. 37 (Eds.: A. Brossi, M. Suffness) Academic Press Inc.: San
Diego, 1990. (b) A. U. Rahman, Z. Iqbal, H. Nasir in Studies in Natural
Products Chemistry (Ed.: Atta-ur-Rahman) Elsevier: Amsterdam, 1994, 14,
pp.805-884. (c) For recent achievements in the total synthesis of Vinca
alkaloids see: S. Yokoshima, T. Ueda, S. Kobayashi, A. Sato, T. Kuboyama,
H. Tokuyama, T. Fukuyama, J. Am. Chem. Soc. 2002, 124, 2137-2139. (d)
T. Kuboyama, S. Yokoshima, H. Tokuyama, T. Fukuyama, Proc. Natl.
Acad. Sci. USA 2004, 101, 11966-11970.
[2] (a) P. Mangeney, R. Z. Andriamialisoa, J.-Y. Lallemand, N. Langlois, Y.
Langlois, P. Potier, Tetrahedron 1979, 35, 2175-2179. (b) P. Mangeney, R.
Z. Andriamialisoa, N. Langlois, Y. Langlois, P. Potier, J. Org. Chem. 1979,
44, 3765-3768. (c) R. Z. Andriamialisoa, N. Langlois, Y. Langlois, P.
Potier, Tetrahedron 1980, 36, 3053-3060.
[3] (a) J. Fahy, A. Duflos, J.-P. Ribet, J.-C. Jacquesy, C. Berrier, M.-P.
Jouannetaud, F. Zunino, J. Am. Chem. Soc. 1997, 119, 8576-8577. (b) A.
Duflos, A. Kruczynski, J.-M. Barret, Curr. Med. Chem. – Anti-cancer
agents 2002, 2, 55-70.
[4] M. M. Gupta, D. V. Singh, A. K. Tripathi, R. Pandey, R. K. Verma, S.
Singh, A. K. Shasany, S. P. S. Khnuja, J. Chrom. Sci. 2005, 43, 450-453.
[5] (a) G. Büchi, P. Kulsa, K. Ogasawara, R. L. Rosati, J. Am. Chem. Soc. 1970,
92, 999-1005. (b) J. P. Kutney, F. Bylsma, Helv. Chim. Acta 1975, 58, 16721689. (c) B. M. Trost, S. A. Godleski, J. L. Belletire, J. Org. Chem. 1979,
44, 2052-2054. (d) T. Imanashi, H. Shin, N. Yagi, M. Hanaoka, Tetrahedron
Lett. 1980, 21, 3285-3288. (e) C. Marazano, M.-T. Le Goff, J.-L. Fourrey,
B. C. Das, J. Chem. Soc., Chem. Commun. 1981, 389-391. (f) S. Raucher, B.
L. Bray, J. Org. Chem. 1985, 50, 3237-3239. (g) M. Kuehne, W. G.
Bornmann, W. G. Earley, I. Marko, J. Org. Chem. 1986, 51, 2913-2927. (h)
S. Raucher, B. L. Bray, R. F. Lawrence, J. Am. Chem. Soc. 1987, 109, 442446. (i) R. J. Sundberg, J. Hong, S. Q. Smith, M. Sabat, Tetrahedron 1998,
54, 6259-6292. (j) M. T. Redding, T. Fukuyama, Org. Lett. 1999, 7, 973976.
[6] C. Szantay, H. Bölcskei, E. Gacs-Baitz, Tetrahedron 1990, 46, 1711-1732.
[7] B. L. Bray, Ph.D. Thesis, Washington University, 1987.
[8] Enantiomeric excesses were determined by chiral HPLC.
[9] For X-ray structure details see supplementary information.
[10] S. V. Ley, J. Norman, W. P. Griffith, S. P. Marsden, Synthesis 1994, 639666.
[11] A. Deiters, S. F. Martin, Chem. Rev. 2004, 104, 2199-2238.
[12] T. Itoh, N. Yamazaki, C. Kibayashi, Org. Lett. 2002, 4, 2469-2472.
2
3
Natural Product Synthesis
Lionel Moisan, Pierre Thuéry, Marc
Nicolas, Eric Doris,* and Bernard
Rousseau
Formal Synthesis of (+)-Catharanthine
The first formal synthesis of (+)catharanthine, the crucial building
block of the major antitumor agent
Vinorelbine,
is
described.
The
intermediate described by Büchi is
obtained in virtually optically pure form
starting
from
L-serine.
This
intermediate can lead in five steps to
(+)-catharanthine.
N
O
HO
N
H
OH
NH2
L-Serine
OH
Büchi's intermediate
N
known
N
H
CO2Me
(+)-Catharanthine
4
Download