An efficient enantioselective synthesis of (-)-galanthamine

Article abstract of DOI:10.1002/1521-3773(20020802)41:15<2795::AID-ANIE2795>3.0.CO;2-2

An effective sequence: Palladium-catalyzed asymmetric allylic alkylation, Heck cyclization, and diastereoselective allylic oxidation were used in the total synthesis of (-)-galanthamine (3) in 14.8 % overall yield (from 1 and 2, Troc = 2,2,2-trichloroethoxycarbonyl) and with 96 % ee. This improved procedure provides the shortest and most efficient nonbiomimetic synthesis of the acetylcholinesterase inhibitor.

Full text of DOI:10.1002/1521-3773(20020802)41:15<2795::AID-ANIE2795>3.0.CO;2-2

COMMUNICATIONS
OH
An Efficient Enantioselective Synthesis of
H
H
O
route 1
(À)-Galanthamine**
O
H3CO
H3CO
Barry M. Trost* and Weiping Tang
NCH3
NCH3
Galanthamine (1),[1] the parent member of the galanth-
amine-type Amaryllidaceae alkaloids, is a centrally acting
competitive and reversible inhibitor of acetylcholinesterase
(Ache), which significantly enhances cognitive functions of
patients suffering from Alzheimer×s disease.[2] It was first
approved in Austria and most recently in the rest of Europe
and in the United States for the treatment of Alzheimer×s
disease.
4
1
route 2
OH
H
O
O
H3CO
CO2CH3
H3CO
Br
CN
CHO
CHO
5
6
Because of the limited supplies and the high cost of its
isolation from natural sources,[3, 4] several syntheses have been
reported which use biomimetic oxidative bisphenol cou-
pling[4, 5] to create the critical spiro quaternary carbon center
of narwedine (2), which is converted into 1 by diastereose-
lective reduction.[3] We[6] disclosed the first enantioselective
synthesis of (À)-1 in 2000 in which a sequential palladium-
catalyzed asymmetric allylic alkylation (AAA)[7] and intra-
molecular Heck cyclization were used. At the same time,
several other groups utilized similar Heck cyclizations to
construct the quaternary carbon center of (Æ)-3-deoxyga-
lanthamine[8, 9] and(Æ)-1.[10]
Scheme 1. Retrosynthetic analysis of (À)-galanthamine (1).
on the success in the synthesis of the related Amaryllidaceae
alkaloids. Attempts to directly oxidize 4 to 1 failed. Therefore,
a four-step protocol was developed to introduce the C3
oxygen group stereoselectively. It appeared that the benzylic
position is more labile towards oxidation than the allylic
position.[9]
We envisioned a second-generation strategy, which had two
goals: 1) realization of an effective general strategy for this
family of alkaloids by total synthesis of galanthamine, and
2) development of a flexible strategy to provide ready access
to galanthamine analogues by simple modification of the
synthetic scheme. Scheme 1 (route 2) illustrates the key
feature of this new strategy, which envisions closing the
hydrobenzazepine ring after the introduction of the C3 allylic
hydroxy group. Direct allylic oxidation is proposed to set the
C3 OH group for intermediate 5, whose benzylic position was
already at the oxidation state of an aldehyde.
OH
3
O
OH
H
H
H
5 O
O
O
H3CO
H3CO
H3CO
1
6
11
NCH3
NCH3
N
9
(CH2)n
R
1
2
3
As in the first-generation synthesis, aryl ether 6[6] was
prepared in good yield (72%) and with high enantiomeric
excess (87% to 88% ee) by palladium-catalyzed reaction of
2-bromovanillin (7)[17] with carbonate 8 (available in two steps
from glutaraldehyde and the Emmons Wadsworth Horner
reagent)[18] in the presence of chiral ligand 9 (Scheme 2). All
attempts to effect the intramolecular Heck reaction of aryl
ether 6 failed in our previous synthesis, resulting primarily in
ionized product phenol 7. Early results[19] suggested that the
presence of electron-withdrawing substituents on the phenol
ring favored the palladium-catalyzed ionization over the
intramolecular Heck reaction. Therefore, in our earlier work 6
was reduced with DIBAL-H and the resulting diol was
protected as the bis-silyl ether for the Heck reaction. Recent
reports[8, 9] showed that the Heck cyclization worked smoothly
in similar systems in the presence of an aldehyde on the
phenol ring. This observation suggested that the electron-
withdrawing group on the olefin favored the ionization,
whereas the electron-withdrawing group on the phenol ring
facilitated oxidative addition of the palladium to both the
In the endeavor to search for more potent inhibitors of
Ache, there is considerable interest in derivatives that are
based on (À)-galanthamine as a lead structure,[11, 12] since (À)-
galanthamine is less toxic than other Ache inhibitors such as
physostigmine and tacrine.[11, 13, 14] Among them, galanth-
amine derivatives 3[14, 15] or their iminium salts, synthesized
by selective N-demethylation followed by N-alkylation of
galanthamine,[16] were found to be more potent (up to 70-fold)
than galanthamine in inhibiting Ache. To date, syntheses that
employ the oxidative phenol coupling step have not been used
for direct access to these derivatives.
Our previous approach to 1 relied on a direct allylic
oxidation of deoxygalanthamine 4 (Scheme 1, route 1) based
[*] Prof. B. M. Trost, W. Tang
Department of Chemistry, Stanford University
Stanford, CA 94305-5080 (USA)
Fax : (‡1)650-725-0259
E-mail: bmtrost@stanford.edu
[**] We thank the National Science Foundation and the National Institute
of Health, General Medical Sciences (GM13598), for their generous
support of our programs. Mass spectra were provided by the Mass
Spectrometry Facility of the University of California, San Francisco,
supported by the NIH Division of Research Resources.
À
À
C O and C Br bonds. We decided to keep the aldehyde
functionality while removing the electron-withdrawing sub-
stituent on the olefin by homologation of the a,b-unsaturated
ester to a b,g-unsaturated nitrile.
To this end, aldehyde 6 was protected as its dimethylacetal
by treatment with trimethyl orthoformate and catalytic
Supporting information for this article is available on the WWW under
http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2002, 41, No. 15
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4115-2795 $ 20.00+.50/0
2795

COMMUNICATIONS
perfectly aligned with the p system.
Gratifyingly, treatment of olefin 12 with
SeO2 in dioxane at 1508C in the presence
of NaH2PO4 and quartz sand in a sealed
tube provided alcohol 5 in 57% yield
(64% based on recovered starting ma-
terial; dr 10:1) with 7% recovered start-
ing material.[22] Other additives, such as
HCO2H,[23] InCl3, YbCl3 and Na2HPO4,
gave less of the desired product. At-
tempts to fully convert all the starting
material by extending the reaction time
or increasing the temperature resulted
in more decomposition. Only trace
amounts of product was found after 8 h
when the reaction was carried out in
refluxing dioxane (1108C). This repre-
sents the first successful allylic oxida-
tion[6, 9] of the galanthamine skeleton.
The stereochemistry of 5 was tentatively
assigned based on the coupling constants
of Ha/Heq and Hb/Heq (J < 5 Hz). The
Ph
NH
Ph
9
O
O
a)
HN
O
OH
CO2CH3
Br
PPh2
72%,
Ph2P
H3CO
H3CO
Br
+
TrocO
CO2CH3
CHO
b, c
CHO
87% to 88% ee
8
6
7
85%
H
O
O
O
d, e
f
91%
H3CO
H3CO
H3CO
CN
OH
CH(OCH3)2
Br
Br
76%
96% ee
after recrystallization
CN
CHO
CHO
11
12
10
Scheme 2. Convergent synthesis of the carbon skeleton of galanthamine. a) 3% 9, {[h3-C3H3PdCl]2}
(1%), Et3N, CH2Cl2, room temperature, b) TsOH (5%), CH(OMe)3, MeOH; c) DIBAL-H/toluene,
À788C; d) Ph3P, acetonecyanohydrin, DIAD, Et2O; e) TsOH (20%), THF, H2O; f) Pd(OAc)2 (15%),
dppp (15%), Ag2CO3, PhCH3, 1078C, 24 h. Troc ˆ 2,2,2-trichloroethoxycarbonyl, DIAD ˆ diiso-
propyl azodicarboxylate, TsOH ˆ p-toluenesulfonic acid, DIBAL-H ˆ diisobutylaluminum hydride,
dppp ˆ bis(diphenylphosphanyl)propane.
p-toluenesulfonic acid. Selective reduction of the a,b-unsat-
urated ester with DIBAL-H at À788C afforded allylic alcohol
10. The b,g-unsaturated nitrile 11 was prepared in 84% yield
by using a modified Mitsunobu protocol[20] followed by acid
hydrolysis. The enantiomeric excess of 11 was improved to
96% ee by recrystallization from diethyl ether and petroleum
ether with 90% mass recovery.
We examined the Heck reaction of the acetonitrile-
substituted cyclohexene 11 (Scheme 2). The Heck product
12 was obtained in high yield (91%) by using a catalytic
amount of dppp and [Pd(OAc)2] in refluxing toluene in the
presence of excess silver carbonate. Other solvent (DMF) or
ligands (dppe, dppf) gave lower yields. Once we had
established the core structure of the galanthamine, we turned
our attention to the introduction of the C3 allylic hydroxy
group. This can be realized by oxidation of the olefin to enone
followed by known diastereoselective 1,2-reduction or more
efficiently, direct oxidation to the allylic alcohol, which raises
the question regarding the diastereoselectivity. Usually, the
electrophile would approach the olefin from the less hindered
convex face. We envisioned that SeO2 would react with the
olefin from the more hindered concave face through an ene
mechanism[21] (Scheme 3), because the axial proton Hax is
high diastereoselectivity reflects the stereoelectronic require-
ment for the ene reaction.
Allylic alcohol 5 together with its epimer (10:1) was then
converted into the natural product in a one-pot process
(Scheme 4). Aldehyde 5 was treated with methylamine in
OH
H
OH
O
H
a)
H3CO
O
H3CO
CN
CN
CHO
N
R
5
13
b)
OH
OH
H
H
O
O
c)
H3CO
H3CO
OH
NR
NR
(–)-1: R=Me 62%.
(–)-15 (SPH-1339): R=nPr 56%.
14
Scheme 4. Completion of the synthesis by a one-pot procedure. a) RNH2,
MeOH; b) DIBAL-H (4 equiv), then aqueous NaH2PO4; c) NaCNBH3.
R ˆ Me: 62% of 1 and 6% of epi-1; R ˆ nPr: 56% of 15.
Hb
Ha
H
OH
Heq
H
O
O
a
H3CO
H3CO
methanol solution. Excess methylamine and methanol was
removed in vacuo. Concomitant reduction of the imine and
nitrile by DIBAL-H followed by acid quenching presumably
gave the seven-membered ring of hemiaminal 14. The
resulting solution was treated directly with sodium cyanoboro-
hydride, and (À)-galanthamine (1) and epi-1 were isolated in
62% and 6% yields, respectively. All spectral data (1H NMR,
13C NMR, IR) are identical to those of a sample of the natural
product and to data from previous synthesis, [a]D ˆ À123.1
(c ˆ 0.4, EtOH), lit[11]: [a]D ˆ À131.4 (c ˆ 0.6, EtOH), lit[6]:
57%
CN
CHO
CN
CHO
dr 10:1
5
12
OCH3
O
O
O
Se
Hax
Heq
OHC
NC
Scheme 3. Diastereoselective allylic oxidation. a) SeO2, NaH2PO4, diox-
ane, 1508C 3 h, 64% based on recovered starting material.
2796
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4115-2796 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 15

COMMUNICATIONS
[a]D ˆ À112.8 (88% ee, c ˆ 0.5, EtOH). SPH-1339 (15),[15]
a
[10] C. Guillou, J.-L. Beunard, E. Gras, C. Thal, Angew. Chem. 2001, 113,
4881 4882; Angew. Chem. Int. Ed. 2001, 40, 4745 4746.
[11] S. Y. Han, J. E. Sweeney, E. S. Bachman, E. J. Schweiger, G. Forloni,
J. T. Coyle, B. M. Davis, M. M. Joullie, Eur. J. Med. Chem. 1992, 27,
673 687.
slightly more potent inhibitor of Ache than (À)-galanthamine
1, was prepared in 56% yield by this one-pot process, by using
propylamine instead of methylamine.
This new synthesis of galanthamine (8 steps, 96% ee, 14.8%
overall yield from 7 and 8) is a significant improvement over
and successfully addresses many of the shortcomings of the
previous synthesis (14 steps, 88% ee, 1.5% overall yield).
Furthermore, by exchanging methylamine for other alkyl
amines in the last step various galanthamine derivatives are
easily accessible, as demonstrated by the synthesis of 15. This
is the shortest and most efficient nonbiomimetic total syn-
thesis of (À)-galanthamine to date. The sequential palladium-
catalyzed AAA and intramolecular Heck reaction followed
by a diastereoselective allylic oxidation provided the key
intermediate 5 with all the functionality installed, except the
hydrobenzazepine ring. The one-pot reductive cyclization
represents a simple and efficient strategy to form the latter
and to access many galanthamine analogues. All the stereo-
chemistry emanates from the palladium-catalyzed AAA. This
strategy should allow for entry into a variety of Amaryllida-
ceae alkaloids related to galanthamine.
[12] G. M. Bores, R. W. Kosley, Jr., Drugs Future 1996, 21, 621 635.
[13] S. Y. Han, S. C. Mayer, E. J. Schweiger, B. M. Davis, M. M. Joullie,
Bioorg. Med. Chem. Lett. 1991, 1, 579 580, and references therein.
[14] A. Mary, D. Z. Renko, C. Guillou, C. Thal, Bioorg. Med. Chem. 1998,
6, 1835 1850; C. Guillou, A. Mary, D. Z. Renko, E. Gras, C. Thal,
Bioorg. Med. Chem. Lett. 2000, 10, 637 639.
[15] U. Jordis, J. Frˆhlich, M. Treu, M. Hirnschall, L. Czollner, B. K‰lz, S.
Welzig, PCT Int. Appl., WO 0174820, 2001.
[16] A. Mary, D. Z. Renko, C. Guillou, C. Thal, Tetrahedron Lett. 1997, 38,
5151 5152.
[17] J. E. Toth, P. R. Hamann, P. L. Fuchs, J. Org. Chem. 1988, 53, 4694
4708.
[18] H. Amri, M. Rambaud, J. Villieras, Tetrahedron 1990, 46, 3535 3546.
[19] R. C. Larock, D. E. Stinn, Tetrahedron Lett. 1988, 29, 4687 4690.
[20] B. K. Wilk, Synth. Commun. 1993, 23, 2481 2484; M. C. Aesa, G.
¬
¬
¬
Baan, L. Novak, C. Szantay, Synth. Commun. 1995, 25, 1545 1550.
[21] K. C. Nicolaou, N. A. Petasis, Selenium in Natural Products Synthesis,
CIS, Philadelphia, 1984.
[22] B. M. Trost, D. L. Van Vranken, J. Am. Chem. Soc. 1993, 115, 444
458.
[23] K. Shibuya, Synth. Commun. 1994, 24, 2923 2941.
Received: April 4, 2002 [Z19040]
[1] For reviews see: O. Hoshino in The Alkaloids, Vol. 51 (Ed.: G. A.
Cordell), Academic Press, New York, 1998, pp. 323 424; S. F. Martin
in The Alkaloids, Vol. 30 (Ed.: A. Brossi), Academic Press, New York,
1987, pp. 251 376.
[2] M. Weinstock, CNS Drugs 1999, 12, 307 323; U. Thatte, Curr. Opin.
Cent. Peripher. Nerv. Syst. Invest. Drugs 1999, 1, 357 372; K. Unni,
CNS Drugs 1998, 10, 447 460; A. Nordberg, A.-L. Svensson, Drug
Saf. 1998, 19, 465 480; M. Rainer, Drugs Today 1997, 33, 273 279;
H. A. M. Mucke, Drugs Today 1997, 33, 251 257.
Monomeric Compounds Containing the
‡
ˆ
cis-[V( O)(OH)] Core
Evangelos J. Tolis, Manolis J. Manos,
Anastasios J. Tasiopoulos, Catherine P. Raptopoulou,
Aris Terzis, Michael P. Sigalas,* Yiannis Deligiannakis,*
and Themistoklis A. Kabanos*
[3] W.-C. Shieh, J. A. Carlson, J. Org. Chem. 1994, 59, 5463 5465.
[4] B. K¸enburg, L. Czollner, J. Frˆhlich, U. Jordis, Org. Process Res. Dev.
1999, 3, 425 431.
Monomeric metal compounds with terminal hydroxy li-
gands are important functional units in metalloproteins.[1]
[5] D. H. R. Barton, G. W. Kirby, J. Chem. Soc. 1962, 806; T. Kametani, K.
Yamaki, H. Yagi, K. Fukumoto, J. Chem. Soc. D 1969, 425 426; T.
Kametani, K. Yamaki, H. Yagi, K. Fukumoto, J. Chem. Soc. C 1969,
2602 2605; T. Kametani, K. Yamaki, T. Terui, J. Heterocycl. Chem.
1973, 10, 35; K. Shimizu, K. Tomioka, S. Yamada, K. Koga, Hetero-
cycles 1977, 8, 277 282; K. Shimizu, K. Tomioka, S. Yamada, K. Koga,
Chem. Pharm. Bull. 1978, 26, 3765 3771; D. Krikorian, R. Vlahov, S.
Parushev, M. Chinova, I. Vlahov, H. J. Schaefer, H. Duddeck, G.
Snatzke, Tetrahedron Lett. 1984, 25, 2969 2972; R. Vlahov, D.
Krikorian, G. Spassov, M. Chinova, I. Vlahov, S. Parushev, G. Snatzke,
L. Ernst, K. Kieslich, W.-R. Abraham, W. S. Sheldrick, Tetrahedron
1989, 45, 3329 3345; J. Szewczyk, J. W. Wilson, A. H. Lewin, F. I.
Carroll, J. Heterocycl. Chem. 1995, 32, 195 199; D. A. Chaplin, N.
Fraser, P. D. Tiffin, Tetrahedron Lett. 1997, 7931; L. Czollner, W.
Frantsits, B. K¸enburg, U. Hedenig, J. Frˆhlich, U. Jordis, Tetrahedron
Lett. 1998, 39, 2087 2088; Y. Kita, M. Arisawa, M. Gyoten, M.
Nakajima, R. Hamada, H. Tohma, T. Takada, J. Org. Chem. 1998, 63,
6625 6633; D. Krikorian, V. Tarpanov, S. Parushev, P. Mechkarova,
Synth. Commun. 2000, 30, 2833 2846; M. Node, S. Kodama, Y.
Hamashima, T. Baba, N. Hamamichi, K. Nishide, Angew. Chem. 2001,
113, 3150 3152; Angew. Chem. Int. Ed. 2001, 40, 3060 3062.
[6] B. M. Trost, F. D. Toste, J. Am. Chem. Soc. 2000, 122, 11262 11263.
[7] B. M. Trost, F. D. Toste, J. Am. Chem. Soc. 1998, 120, 815 816; B. M.
Trost, F. D. Toste, J. Am. Chem. Soc. 1998, 120, 9074 9075; B. M.
Trost, F. D. Toste, J. Am. Chem. Soc. 1999, 121, 3543 3544; B. M.
Trost, F. D. Toste, J. Am. Chem. Soc. 1999, 121, 4545 4554; B. M.
Trost, H. C. Tsui, F. D. Toste, J. Am. Chem. Soc. 2000, 122, 3534 3535.
[8] C. Pilger, B. Westermann, U. Flˆrke, G. Fels, Synlett 2000, 1163 1165.
[9] P. J. Parsons, M. D. Charles, D. M. Harvey, L. R. Sumoreeah, A. Shell,
G. Spoors, A. L. Gill, S. Smith, Tetrahedron Lett. 2001, 42, 2209 2211.
À
Metal OH (M OH) units are proposed as the active species
[*] Dr. T. A. Kabanos, Dr. E. J. Tolis, M. J. Manos, Dr. A. J. Tasiopoulos
Department of Chemistry
Section of Inorganic and Analytical Chemistry
University of Ioannina
45110 Ioannina (Greece)
Fax : (‡30)6510-44831
E-mail: tkampano@cc.uoi.gr
Dr. Y. Deligiannakis
Department of Environmental and Natural Resources Management
Laboratory of Physical Chemistry
University of Ioannina
Seferi 2, 30100 Agrinio (Greece)
Fax : (‡30)6410-39576
E-mail: ideligia@cc.uoi.gr
Dr. M. P. Sigalas
Department of Chemistry
Laboratory of Applied Quantum Chemistry
Aristotle University of Thessaloniki
54124 Thessaloniki (Greece)
Fax : (‡30)310-99-7739
E-mail: sigalas@chem.auth.gr
Dr. C. P. Raptopoulou, Prof. A. Terzis
NRCPS Demokritos
Institute of Materials Science
15310 Agia Paraskevi Attikis (Greece)
Angew. Chem. Int. Ed. 2002, 41, No. 15
¹ WILEY-VCH Verlag GmbH, 69451 Weinheim, Germany, 2002
1433-7851/02/4115-2797 $ 20.00+.50/0
2797