An efficient approach to Aspidosperma alkaloids via [4 + 2] cycloadditions of aminosiloxydienes: Stereocontrolled total synthesis of (±)-tabersonine. Gram-scale catalytic asymmetric syntheses of (+)-tabersonine and (+)-16-methoxytabersonine. Asymmetric syntheses of (+)-aspidospermidine and (-)-quebrachamine

Article abstract of DOI:10.1021/ja017863s

Described is a concise, highly stereocontrolled strategy to the Aspidosperma family of indole alkaloids, one that is readily adapted to the asymmetric synthesis of these compounds. The strategy is demonstrated by the total synthesis of (±)-tabersonine (rac-1), proceeding through a 12-step sequence. The basis for this approach was provided by a highly regio- and stereoselective [4 + 2] cycloaddition of 2-ethylacrolein with 1-amino-3-siloxydiene developed in our laboratory. Subsequent elaboration of the initial adduct into the hexahydroquinoline DE ring system was accomplished efficiently by a ring-closing olefin metathesis reaction. A novel ortho nitrophenylation of an enol silyl ether with (o-nitrophenyl)phenyliodonium fluoride was developed to achieve an efficient, regioselective introduction of the requisite indole moiety. The final high-yielding conversion of the ABDE tetracycle into pentacyclic target rac-1 relied on intramolecular indole alkylation and regioselective C-carbomethoxylation. Our approach differs strategically from previous routes and contains built-in flexibility necessary to access many other members of the Aspidosperma family of indole alkaloids. The versatility of the synthetic strategy was illustrated through the asymmetric syntheses of the following Aspidosperma alkaloids: (+)-aspidospermidine, (-)-quebrachamine, (-)-dehydroquebrachamine, (+)-tabersonine, and (+)-16-methoxytabersonine. Of these, (+)-tabersonine and (+)-16-methoxytabersonine were synthesized in greater than 1 -g quantities and in enantiomerically enriched form (～95% ee). The pivotal asymmetry- introducing step was a catalyzed enantioselective Diels-Alder reaction, which proceeded to afford the cycloadducts in up to 95% ee. Significantly, the synthetic sequence was easy to execute and required only four purifications over the 12-step synthetic route.

Full text of DOI:10.1021/ja017863s

Published on Web 04/06/2002
An Efficient Approach to Aspidosperma Alkaloids via [4 + 2]
Cycloadditions of Aminosiloxydienes: Stereocontrolled Total
Synthesis of (()-Tabersonine. Gram-Scale Catalytic
Asymmetric Syntheses of (+)-Tabersonine and
(+)-16-Methoxytabersonine. Asymmetric Syntheses of
(+)-Aspidospermidine and (-)-Quebrachamine
Sergey A. Kozmin, Tetsuo Iwama, Yong Huang,† and Viresh H. Rawal*
Contribution from the Department of Chemistry, The UniVersity of Chicago,
Chicago, Illinois 60637
Received December 21, 2001
Abstract: Described is a concise, highly stereocontrolled strategy to the Aspidosperma family of indole
alkaloids, one that is readily adapted to the asymmetric synthesis of these compounds. The strategy is
demonstrated by the total synthesis of (()-tabersonine (rac-1), proceeding through a 12-step sequence.
The basis for this approach was provided by a highly regio- and stereoselective [4 + 2] cycloaddition of
2-ethylacrolein with 1-amino-3-siloxydiene developed in our laboratory. Subsequent elaboration of the initial
adduct into the hexahydroquinoline DE ring system was accomplished efficiently by a ring-closing olefin
metathesis reaction. A novel ortho nitrophenylation of an enol silyl ether with (o-nitrophenyl)phenyliodonium
fluoride was developed to achieve an efficient, regioselective introduction of the requisite indole moiety.
The final high-yielding conversion of the ABDE tetracycle into pentacyclic target rac-1 relied on intramolecular
indole alkylation and regioselective C-carbomethoxylation. Our approach differs strategically from previous
routes and contains built-in flexibility necessary to access many other members of the Aspidosperma family
of indole alkaloids. The versatility of the synthetic strategy was illustrated through the asymmetric syntheses
of the following Aspidosperma alkaloids: (+)-aspidospermidine, (-)-quebrachamine, (-)-dehydroquebra-
chamine, (+)-tabersonine, and (+)-16-methoxytabersonine. Of these, (+)-tabersonine and (+)-16-
methoxytabersonine were synthesized in greater than 1-g quantities and in enantiomerically enriched form
( ∼95% ee). The pivotal asymmetry-introducing step was a catalyzed enantioselective Diels-Alder reaction,
which proceeded to afford the cycloadducts in up to 95% ee. Significantly, the synthetic sequence was
easy to execute and required only four purifications over the 12-step synthetic route.
Introduction
veloped a catalyzed asymmetric DA reaction of aminosiloxy-
dienes that affords the cycloadducts in high yields and with
The field of alkaloid synthesis, pioneered more than half a
century ago, remains an active and exciting area of chemistry
and provides a fertile ground for innovation. The pharmacologi-
cal significance of naturally occurring alkaloids along with the
challenge posed by their diverse and intricate structures has
inspired much new chemistry. Efficient and elegant routes to
these often daunting molecular frameworks have been devel-
oped. The unique problems incurred during the studies have,
in turn, fueled the development of new synthetic methods that
have proven to be of general use.
excellent enantiomeric excesses.1i,k Overall, the broad scope of
the DA reaction of these dienes provides an attractive foundation
for the development of new strategies to natural products,
especially alkaloids. We recognized that the product of the DA
reaction with an R-substituted acrolein, cycloadduct II, incor-
porates and perfectly positions an amine, a silyl enol ether, and
an aldehyde functionality for further elaboration into tetrahy-
droindoline (III) and hexahydroquinoline (IV) substructures,
which are found in many natural alkaloids (Scheme 1). We
(1) (a) Kozmin, S. A.; Rawal, V. H. J. Org. Chem. 1997, 62, 5252-5253. (b)
Kozmin, S. A.; Rawal, V. H. J. Am. Chem. Soc. 1997, 119, 7165-7166.
(c) Kozmin, S. A.; Janey, J. M.; Rawal, V. H. J. Org. Chem. 1999, 64,
3039-3052. (d) Kozmin, S. A.; Green, M. T.; Rawal, V. H. J. Org. Chem.
1999, 64, 8045-8047. (e) Kozmin, S. A.; Rawal, V. H. J. Am. Chem. Soc.
1999, 121, 9562-9573. (f) Kozmin, S. A.; He, S.; Rawal, V. H. Org. Synth.
2000, 78, 160-168. (g) Janey, J. M.; Iwama, T.; Kozmin, S. A.; Rawal,
V. H. J. Org. Chem. 2000, 65, 9059-9068. (h) Huang, Y.; Rawal, V. H.
Org. Lett. 2000, 2, 3321-3323. (i) Huang, Y.; Iwama, T.; Rawal, V. H. J.
Am. Chem. Soc. 2000, 122, 7843-7844. For preparation of 1-(dimethyl-
amino)-3-(tert-butyldimethylsilyloxy)-1,3-butadiene, see: (j) Kozmin, S.
A.; He, S.; Rawal, V. H. Org. Synth. 2000, 78, 152-159. (k) Huang, Y.;
Iwama, T.; Rawal, V. H. Org. Lett. 2002, 4, 1163-1166.
Over the past few years, we have been actively studying the
chemistry of 1,3-butadienes that have a nitrogen attached at the
1-position and an oxygen at the 3-position (I).1 Such aminosil-
oxydienes are highly reactive and readily undergo Diels-Alder
(DA) cycloadditions with a variety of dienophiles. The cycload-
ditions proceed with complete regiocontrol and in some cases
with exceptional diastereoselectivity. Indeed, recently we de-
† Recipient of Abbott Laboratories Graduate Fellowship.
9
4628
J. AM. CHEM. SOC. 2002, 124, 4628-4641
10.1021/ja017863s CCC: $22.00 © 2002 American Chemical Society

Aspidosperma Alkaloids
A R T I C L E S
Scheme 1
Scheme 2
describe here the realization of this principle through the
development of a conceptually novel strategy to the Aspi-
dosperma family of alkaloids. The strategy is illustrated through
the concise, stereocontrolled synthesis of (()-tabersonine (rac-
1), as well as the asymmetric syntheses of (+)-tabersonine (ent-
1), (+)-16-methoxytabersonine, (+)-aspidospermidine, (-)-
quebrachamine, and (-)-dehydroquebrachamine.2
Background and Biological Significance
The Aspidosperma family represents one of the largest groups
of indole alkaloids, with more than 250 compounds isolated
from various biological sources.3 The basic skeletal features of
these compounds, particularly the complex pentacyclic ABCDE
framework, can be seen in the namesake of the family,
aspidospermine. In our development of a new strategy to these
alkaloids, we selected as the target the slightly more elaborate
and significant member of this family, tabersonine (1), which
the fact that its coupling with catharanthine (4) leads to the
pharmacologically significant Vinca alkaloids, such as vinblas-
tine (5) and vincristine (6).8 It is interesting to note that
catharanthine, a member of the Iboga family of alkaloids, also
appears to arise from an intramolecular DA reaction of dehy-
drosecodine 2, but by an alternate mode of cyclization.9
Vinblastine, vincristine, and some closely related derivatives
display potent anticancer activity, and these compounds are used
currently for the treatment of several types of tumors. Vincris-
tine, for example, when used in combination with other
anticancer agents, is remarkably effective for the treatment of
pediatric leukemia, with a success rate of >90%. Tabersonine
is important not only for its biosynthetic relationship to the Vinca
alkaloids but also because it is the chemical progenitor of these
alkaloids. The chemical conversion of tabersonine to vindoline
and then to different bisindole alkaloids is well documented.9
For these reasons, tabersonine represents an attractive target for
synthesis.10 A stereocontrolled, concise, and high-yielding route
to tabersonine would open up the possibility of synthesizing
not only vinblastine and vincristine but also numerous structural
analogues that would be difficult to obtain via the natural
compounds. Finally, since the biosynthesis and most chemical
syntheses of vindoline go through 16-methoxytabersonine (7)
plays a central role in the biosynthesis and the synthetic chem-
istry of Aspidosperma alkaloids.
Tabersonine was first isolated in 1954 from Amsonia taber-
naemontana by Le Men et al. Soon after the initial report, the
alkaloid was isolated from several other natural sources,
indicating its relative biological abundance.4 Wenkert had
proposed in 1962 that tabersonine could arise biosynthetically
via the intramolecular DA reaction of dehydrosecodine 2
(Scheme 2).5 Support for this hypothesis came from subsequent
labeling experiments and by the isolation of secodine from
natural sources.6 Tabersonine serves as the biosynthetic prede-
cessor to several members of the Aspidosperma family, most
notably vindoline (3).7 The importance of vindoline stems from
(2) Preliminary communication: Kozmin, S. A.; Rawal, V. H. J. Am. Chem.
Soc. 1998, 120, 13523-13524.
(6) (a) Evans, D. A.; Smith, G. F.; Smith, G. N.; Stapleford, K. S. J. Chem.
Soc., Chem. Commun. 1968, 859-861. (b) Sakai, S.; Aimi, N.; Kato, K.;
Ido, H.; Watanabe, Y.; Haginiva, J. Chem. Pharm. Bull. 1971, 19, 1503-
1505. (c) Battersby, A. R.; Bhatnagar, A. K. J. Chem. Soc., Chem. Commun.
1970, 193-194.
(3) (a) Saxton, J. E. Indoles, Part 4: The Monoterpenoid Indole Alkaloids;
Wiley: Chichester, 1983. (b) Herbert, R. B. In The Monoterpenoid Indole
Alkaloids; supplement to Vol. 25, part 4, of The Chemistry of Heterocyclic
Compounds; Saxton J. E., Ed.; Wiley: Chichester, 1994; Chapter 1. (c)
Saxton, J. E. The Monoterpenoid Indole Alkaloids; supplement to Vol. 25,
part 4, of The Chemistry of Heterocyclic Compounds; Saxton J. E., Ed.;
Wiley: Chichester, 1994; Chapter 8. (d) Saxton, J. E. In The Alkaloids;
Cordell, G. A., Ed.; Academic Press: New York, 1998; Vol. 50. (e) Saxton,
J. E. In The Alkaloids; Cordell, G. A., Ed.; Academic Press: New York,
1998; Vol. 51, Chapter 1. (f) Toyota, M.; Ihara, M. Nat. Prod. Rep. 1998,
327-340 and references therein.
(7) (a) Danieli, B.; Lesma, G.; Palmisano, G.; Riva, R. J. Chem. Soc., Chem.
Commun. 1984, 909-911. (b) Danieli, B.; Lesma, G.; Palmisano, G.; Riva,
R. J. Chem. Soc., Perkin Trans. 1 1987, 155-161.
(8) (a) Potier, P.; Langlois, N.; Langlois, Y.; Gueritte, G. J. Chem. Soc., Chem.
Commun. 1975, 670-671. (b) Kutney, J. P.; Ratcliffe, A. H.; Treasurywala,
A. M.; Wunderly, S. Heterocycles 1975, 3, 639-649. (c) Langlois, N.;
Gueritte, G.; Langlois, Y.; Potier, P. J. Am. Chem. Soc. 1976, 98, 7017-
7024. (d) Mangeney, P.; Andriamialisoa, R. Z.; Langlois, N.; Langlois,
Y.; Potier, P. J. Am. Chem. Soc. 1979, 101, 2243-2245.
(4) Janot, M.-M.; Pourrat, H., Le Men, J. Bull. Soc. Chim. Fr. 1954, 707-
708.
(9) (a) Scott, A. I. Acc. Chem. Res. 1970, 3, 151-157. (b) Atta-Ur-Rahman,
Basha, A. Biosynthesis of Indole Alkaloids; Clarendon Press: Oxford, 1983.
(5) Wenkert, E. J. Am. Chem. Soc. 1962, 84, 98-102.
9
J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002 4629

A R T I C L E S
Kozmin et al.
Scheme 3
Scheme 4
membered ring containing the C-6-C-7 double bond could be
constructed via ring-closing metathesis reaction. Conversion of
10 into tetracycle 11 would rely on indolization. In this context,
it is noteworthy that there was no good precedent for ac-
complishing this transformation in a regiocontrolled manner.
We planned either to use the known methods of indole
synthesessi.e., Gassman’s or Fischer’s protocolssor to develop
a new method for regioselective indolization.
The assembly of the spiroindolenine portion of the pentacyclic
skeleton would be accomplished by an intramolecular alkylation.
Whereas this construction had precedent, the reported yields
were low, so a more effective solution would have to be
developed. Regioselective incorporation of carbomethoxy group,
also an unsolved problem in the alkaloids field, would then
furnish tabersonine (1).
as an intermediate, the development of a direct synthesis of this
alkaloid was an important objective.11
Synthetic Strategy. The highly compact but intricate pen-
tacyclic skeleton of tabersonine (1) represents a significant
synthetic challenge. The central cyclohexene ring (E) is
substituted at five of the six carbons and possesses two
quaternary chiral centers. The cis relationship of the three
contiguous stereocenters at C-12, C-19, and C-5 is responsible
for the overall conformational rigidity of the molecule. The
combination of these structural challenges and the potent
pharmacological properties exhibited by several members has
made these alkaloids attractive targets for chemical synthesis.10
Our strategy to the Aspidosperma pentacycle is based on the
recognition that control of the cis relationship of the two
stereogenic centers at C-5 and C-19 is tantamount to solving
the stereochemical problems posed by these alkaloids (Scheme
3). A particularly attractive solution was possible through a
Diels-Alder reaction of our aminosiloxydienes.1 The Diels-
Alder reaction of an appropriately functionalized aminosiloxy-
diene 8 with ethylacrolein would give cycloadduct 9 with
complete regioselectivity and excellent endo selectivity. This
simple transformation would orient the amino group and the
aldehyde functionality in a cis relationship and position the enol
ether for formation of the indole ring, thus addressing all the
issues associated with these alkaloids (vide infra). The six-
Results and Discussion
(I) Total Synthesis of (()-Tabersonine. The Pivotal Diels-
Alder Reaction. The synthesis of racemic tabersonine began
with the preparation of 1-amino-3-siloxy-1,3-butadiene 16,
which was equipped with the carbomethoxy group, to temper
the reactivity of the amino group, and the allyl group, chosen
to provide the necessary handle for constructing the piperidine
ring. We expected to prepare diene precursor 15 by the reported
acid-catalyzed condensation of a secondary amide with the
commercially available monoacetal 13.12 The required N-
allylcarbamate 14 was prepared in 96% crude yield by the
reaction of 2 equiv of allylamine with methyl chloroformate in
dichloromethane. For the condensation reaction, we utilized a
slightly modified protocol, which included the use of chloro-
form, as a higher boiling solvent, to facilitate the product
formation (Scheme 4). Under these conditions, the reaction
reached completion within 24 h and furnished the desired
vinylogous imide 15 in 90% yield.13 On large scales, although
the condensation of 14 with acetal 13 to provide vinylogous
imide 15 proceeded cleanly, the isolated yield of the product
(10) (a) Ziegler, F. E.; Bennett, G. B. J. Am. Chem. Soc. 1971, 93, 5930-5931.
(b) Ziegler, F. E.; Bennett, G. B. J. Am. Chem. Soc. 1973, 95, 7458-7464.
(c) Takano, S.; Hatakeyama, S.; Ogasawara, K. J. Am. Chem. Soc. 1979,
101, 6414-6420. (d) Le´vy, J.; Laronze, J.-Y.; Laronze, J.; Le Men, J.
Tetrahedron Lett. 1978, 1579-1580. (e) Kuehne, M. E.; Okuniewitcz, F.
J.; Kirkemo, C. L.; Bohnert, J. C. J. Org. Chem. 1982, 47, 1335-1343. (f)
Kuehne, Martin E.; Podhorez, David E. J. Org. Chem. 1985, 50, 924-
929. (g) Bornmann, W. G.; Earley, W. G.; Marko, I. J. Org. Chem. 1986,
51, 1, 2913-2927. (h) Kuehne, M. E.; Podhorez, D. E.; Mulamba, T.;
Bornmann, W. G. J. Org. Chem. 1987, 52, 347-353. (i) Kuehne, M. E.;
Hugel, G.; Cartier, D.; Levy, J. Tetrahedron Lett. 1989, 30, 4513-4516.
(j) Kalaus, G.; Greiner, I.; Kajtar-Peredy, M.; Brlik, J.; Szabo, L.; Szantay,
C. J. Org. Chem. 1993, 58, 1434-1442. (k) Kuehne, M. E.; Wang, T.;
Seaton, P. J. J. Org. Chem. 1996, 61, 6001-6008. (l) Kuehne, M. E.;
Bandarage, U. K.; Hammach, A.; Li, Y.-L.; Wang, T. J. Org. Chem. 1998,
63, 2172-2183. (m) Kobayashi, S.; Peng, G.; Fukuyama, T. Tetrahedron
Lett. 1999, 40, 1519-1522.
(12) Khokhlov, P. S.; Savenkov, N. F.; Sokolova, G. D.; Strepikheev, Yu. A.;
Kolesova, V. A. J. Org. Chem. USSR 1982, 875.
(13) Initially, we attempted the preparation of the diene precursor 15 by a two-
step sequence exploiting the acylation of secondary vinylogous amide 56.
While the reaction of allylamine with methoxybutenone occurred to give
56, the second step turned out to be troublesome. The desired vinylogous
imide 15 was typically obtained in 20-40% yield under the different
conditions examined.
(11) (a) Overman, L. E.; Sworin, M.; Burk, R. M. J. Org. Chem. 1983, 48,
2685-2690. (b) Cardwell, K.; Hewitt, B.; Ladlow, M.; Magnus, P. J. Am.
Chem. Soc. 1988, 110, 2242-2248. (c) Kobayashi, S.; Ueda, T.; Fukuyama,
T. Synlett 2000, 883-886. See also ref 10h.
9
4630 J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002

Aspidosperma Alkaloids
A R T I C L E S
Scheme 5
Table 1. Ring-Closing Metathesis of Carbamate 19
solvent
(M)
catalyst
(mol %)
temp
(°C)
time
(h)
conversiona
(%)
yieldb
(%)
entry
1
2
3
4
5
C6H6 (0.03)
C6H6 (0.03)
PhMe (0.09)
CH2Cl2 (0.10)
C6H6 (0.09)
Ae (5)
Ae (10)
Ae (5)
Ae (7)
Be (5)
20
60
80
40
60
40
20
10
2
60
80
ndc
67
58
75
88
60d
85
1
100
after vacuum distillation was lower (45%, 31.3 g). The lower
yield is ascribed to the thermal instability of the vinylogous
imide. Since a large amount of imide 15 was easily prepared,
further effort was not expended on optimizing the large-scale
preparation.
a The extent of conversion was determined by H NMR analysis of the
crude reaction mixture. b Isolated yield after chromatographic purification.
c The isolated yield was not determined in this case. d The other 40%
corresponded to the byproduct formed by isomerization of the allyl group.
e Catalysts employed:
1
Vinylogous imide 15 was converted to the desired diene 16
via a modification of our established protocol. Initially, the
presence of the allyl group in 15 proved to be problematic for
the diene formation. The use of 1.1 equiv of KHMDS (0.5 M
in toluene), according to our standard protocol,2 resulted in
partial isomerization of the double bond to form 17, especially
when the reaction was quenched at room temperature. The
isomerization was less of a problem at lower temperature, but
the silylation was incomplete. Two solutions to this problem
were found. The double bond isomerization was avoided by
using the vinylogous amide in slight excess. An even better
result was obtained when we used the base NaHMDS, which
is available in a more convenient form, as a 1.0 M solution in
THF. The deprotonation was accomplished using a slight excess
of NaHMDS followed by treatment of the resulting enolate with
TBSCl. The reaction was then quenched at -78 °C to afford a
quantitative crude yield of diene 16, which was sufficiently pure
for use in the cycloaddition step.
using a ring-closing metathesis (RCM) reaction.15 The aldehyde
was first converted into the diene 19 via a standard Wittig
olefination protocol. No hydrolysis of the silyl enol ether moiety
was observed under these conditions.
The RCM reaction was investigated next using both Grubbs’
ruthenium16 catalyst and Schrock’s molybdenum17 catalyst. The
results of this study are summarized in Table 1. The reaction
was carried out in benzene at room temperature in the presence
of Grubbs catalyst. NMR analysis after 16 h indicated clean
formation of the cyclized product 20, but the reaction only
progressed to ∼60% conversion (entry 1). The amount of the
desired product did not significantly increase over the next 24
h, suggesting that the catalyst had decomposed. At higher
temperature, the reaction progressed to ∼80% conversion and
afforded 20 in 67% isolated yield (entry 2). The mildest
conditions for the RCM using this catalyst involved the use of
dichloromethane as the solvent (40 °C, 2 h) and gave a 75%
yield of 20 (entry 3). The reaction proceeded even more cleanly
using Schrock’s catalyst, although greater care was required to
exclude oxygen while the catalyst was handled and during the
reaction. The cyclization was essentially complete after 1 h at
60 °C in benzene and afforded the desired product 20 in 88%
yield (entry 4). Dimeric byproducts were not observed in any
of the olefin metathesis experiments described above.18
To simplify the NMR spectra of carbamate 20, which exists
as a 1:1 mixture of rotomers at ambient temperature, the
carbamate was reduced to the corresponding tertiary amine 21.
The structure of 21, established by a combination of the DQF
Diene 16 was expected to display diminished reactivity
compared to the previously described 1-(dialkylamino)-3-siloxy-
1,3-dienes, since the electron-donating ability of the nitrogen
atom is diminished by the presence of the carbomethoxy group.
In accord with this expectation, no reaction occurred at room
temperature between 16 and ethylacrolein. However, the cy-
cloaddition proceeded readily upon heating the reaction mixture
to 65-85 °C in toluene for 48 h. Despite the higher temperature,
the reaction was completely regioselective and very highly endo
selective,14 affording the desired adduct 18 in >95% isolated
yield (Scheme 5). The relative stereochemistry of 18 was
1
tentatively assigned at this stage from its high-temperature H
NMR spectrum. The compound displayed a very characteristic
pattern, in which both H1 and H2 were observed as doublets
with a vicinal coupling constant of 5.5 Hz, consistent with a
conformation in which the nitrogen substituent occupies a
pseudoaxial position and the aldehyde functionality, an equato-
rial orientation. This arrangement of the substituents was in a
good agreement with previously described conformational
analysis of endo adducts of aminosiloxydienes.1c
(15) Recent reviews on the application olefin metathesis to organic synthesis:
(a) Schuster, M.; Blechert, S. Angew. Chem., Int. Ed. Engl. 1997, 36, 2037.
(b) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413. (c) Armstrong,
S. K. J. Chem. Soc., Perkin Trans. 1 1998, 371-388. (d) Phillips, A. J.;
Abell, A. D. Aldrichimica Acta 1999, 32, 75-89. (e) Schrock, R. R.
Tetrahedron 1999, 55, 8141-8153. (f) Wright, D. L. Curr. Org. Chem.
1999, 3, 211-240. (g) Jorgensen, M.; Hadwiger, P.; Madsen, R.; Stutz, A.
E.; Wrodnigg, T. M. Curr. Org. Chem. 2000, 4, 565-588. (h) Furstner, A.
Angew. Chem., Int. Ed. 2000, 39, 3013-3043.
Ring-Closing Metathesis. Having achieved the cis relative
stereochemistry required for Aspidosperma alkaloids, we then
proceeded to construct the hexahydroquinoline (DE) ring system
(16) Fu, G.; Nguyen, S. T.; Grubbs, R. R. J. Am. Chem. Soc. 1993, 115, 9856-
9857.
(17) (a) Schrock, R. R.; Murdzek, J. S.; Bazan, G. C.; Robbins, J.; DiMare, M.;
O’Regan, M. J. Am. Chem. Soc. 1990, 112, 3875-3886. (b) Bazan, G. C.;
Oskam, J. H.; Cho, H.-N.; Park, L. Y.; Schrock, R. R. J. Am. Chem. Soc.
1991, 113, 6899-6907.
(14) A small amount (<5%) of a diastereomeric adduct, tentatively assigned as
exo, was observed in the NMR spectrum of the crude product mixture.
9
J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002 4631

A R T I C L E S
Kozmin et al.
Scheme 6
Figure 1. Structural assignment of 21.
COSY and NOESY experiments, is shown in Figure 1, in what
is believed to be the preferred conformation. The pseudoequa-
torial orientation assigned for the amine is consistent with the
small coupling observed between H1 and H2 (J
) 2.0
H1-H2
Hz). The cis stereochemistry of the hexahydroquinoline (21)
was evident from the strong NOESY cross-peaks between the
proton adjacent to the nitrogen and the methylene unit of the
ethyl group. This observation confirmed the endo selectivity of
the key Diels-Alder reaction.
Attempted Gassman Indolization. The next major hurdle
in the synthesis was the conversion of bicyclic compound V to
the “angular” tetracyclic intermediate VI (eq 1). Formation of
assigned the R-stereochemistry for the thiomethyl substituent
based on the large diaxial coupling between H1 and H2 (J ) 10
Hz). Disappointingly, the Gassman indolization of 22 proved
problematic. Neither the desired imine nor the corresponding
anilinoketone were detected under the standard Gassman
indolization protocols examined. The lack of success of this
reaction can be understood in light of the congested steric
environment around the sulfur atom, as well as the possibility
of competing [1,2] rearrangement of the intermediate ylide, a
known side reaction of the Gassman method.20
Fischer Indolization. We next turned our attention to the
venerable Fisher indolization method for the synthesis of the
desired “angular” tetracycle. The hope was that either there
would be inherent preference for the desired, angular tetracycle
or that we could exploit a modification of the Fisher reaction
to accomplish the desired indole formation.22 Treatment of the
silyl enol ether 20 with dilute hydrochloric acid accomplished
a clean hydrolysis to bicyclic ketone 23, with the cis stereo-
chemistry intact. This conversion, of course, sacrificed the
double bond position that was achieved through the initial
cycloaddition (Scheme 6).
In preparation for Fischer indolization, ketone 23 was
converted to the phenylhydrazone by heating it with phenyl-
hydrazine hydrochloride in the presence of sodium carbonate.
The crude hydrazone was then heated at 95 °C in glacial acetic
acid, a weakly acidic medium known to favor idolization toward
the more substituted carbon.22 Under these conditions, the
indolization proceeded cleanly and in high yield (93% overall)
but afforded both possible indole isomers, in roughly equal
the indole ring in a regiocontrolled manner was critical. To this
end, it is noteworthy that the Diels-Alder reaction had not only
positioned the amine and the sp2 hybridized carbon in the
required cis arrangement, but it had also correctly set up the
enol ether for formation of the desired indole. The problem was
that despite the plethora of methods for indole synthesissand
numerous applications of thesesthere did not appear to be a
way to convert an enol ether regiospecifically to the corre-
sponding indole.19
A solution to the regiocontrolled indole synthesis was
envisioned through the Gassman indole synthesis.20 The plan
was to convert the enolsilyl ether to the corresponding thio-
methyl ketone, which would lock the system into producing
the desired regioisomer of the final indole. The required
thiomethyl ketone 22 was prepared directly from the silyl enol
ether 20 upon its treatment with methylthiomethanesulfonate
and TBAF at -78 °C (Scheme 6).21 As expected, this reaction
gave the thiomethyl ketone 22, which was isolated in 80% yield.
The thiomethyl compound was present as a single diastereomer,
(19) Recent reviews of indole synthesis: (a) Gribble, G. W. J. Chem. Soc., Perkin
Trans. 1 2000, 1045-1075. (b) Sundberg, R. J. Indoles; Academic Press:
London, 1996. (c) Sundberg, R. J. In ComprehensiVe Heterocyclic
Chemistry II; Katritzky, A. R., Rees, C. W., Scriven, E. F. V., Eds.; Elsevier
Science: Oxford, 1996; Vol. 2, pp 119-206. (d) Do¨pp, H.; Do¨pp, D.;
Langer, U.; Gerding, B. In Methoden der Organischen Chemie; Kreher,
R., Ed.; Georg Thieme Verlag: Stuttgart, 1994; Hetarene I, Teil 2, pp 546-
1336 and references therein.
(18) We were interested in assessing the ring-closing metathesis capability of a
tertiary amine, such as 57, compared to carbamate 19. The required amine
(57) was prepared by reduction of 19 with lithium aluminum hydride.
Although no reaction occurred when 57 was treated with Grubbs catalyst
A, a clean cyclization occurred in the presence of the Schrock carbene B,
affording bicyclic amine 21 in 75% yield. This compound was determined
to be identical to the one previously obtained from reduction of carbamate
20.
(20) Gassman, P. G.; van Bergen, T. J.; Gilbert, D. P.; Berkeley, W. C. J. Am.
Chem. Soc. 1974, 96, 5495-5508.
(21) Takeuchi, K. In Encyclopedia of Reagents for Organic Synthesis; Paquette,
L. A., Ed.; Wiley: New York, 1995; Vol. 5, pp 3538-3541.
(22) Fischer Indole synthesis: (a) Robison, B. The Fischer Indole Synthesis; J.
Wiley & Sons: New York, 1982. (b) Hughes, D. L. Org. Prep. Proc. Int.
1993, 25, 607-632. See also: (c) Bonjoch, J.; Catena, J.; Valls, N. J. Org.
Chem. 1996, 61, 7106-7115.
9
4632 J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002

Aspidosperma Alkaloids
A R T I C L E S
Scheme 7
Scheme 10
Scheme 8
Scheme 9
sponding cross-coupling product 29 in 52% yield. Subsequent
treatment of 29 with dimethylamine afforded vinylogous amide
26 in nearly quantitative yield.
amounts. The two isomers were readily separated by column
chromatography and assigned the structures indicated. Some
variations of the above reaction conditions were examined, but
these did not significantly alter the ratio of the two isomeric
indoles.
Incorporation of Aromatic Fragment into Aminosiloxy-
dienes. Given the difficulties encountered with the Gassman
and Fischer indolization methods, we decided to tackle the
problem by a different strategy, one in which the indole aromatic
ring was incorporated in the starting material. To this end, we
considered the use of appropriately functionalized 2-aryl-
substituted aminosiloxydiene VII (Scheme 7). According to this
plan, an endo-selective cycloaddition of VII with ethylacrolein
would furnish VIII; unmasking of aniline and hydrolysis of silyl
enol ether would be followed by imine formation and indoliza-
tion to give tetracycle IX.
Since 2-aryl-substituted aminosiloxydiene had not been
prepared previously, we decided to first construct the 2-phenyl-
substituted derivative 27 in order to test its chemical reactivity.
The diene precursor 26 was prepared as reported in 96% yield
simply by heating phenylacetone and DMF-dimethylacetal at
80-90 °C (Scheme 8).23 Vinylogous amide 26 was then
subjected to our standard silylation procedure, which afforded
the corresponding diene 27 in 96% isolated yield. This diene
was found to be relatively stable in solution in either benzene
or toluene, however, attempts to dissolve it in CDCl3 led to
significant hydrolysis to the vinylogous amide 26.
A second route to 2-aryl-substituted aminosiloxydienes was
developed, one based on the Suzuki cross-coupling reaction24
between 2-iodo-3-methoxybutenone (28) and the appropriate
arylboronic acid (Scheme 9). This method promised to be more
versatile given the ready availability of a variety of substituted
arylboronic acids. Indeed, it was found that the reaction between
iodide 2825 and phenylboronic acid proceeded well under the
conditions developed by Johnson et al.,26 to give the corre-
A brief study was carried out to assess the reactivity of diene
27 with activated dienophiles. Diene 27 reacted readily with
N-phenylmaleimide at -20 °C to afford the endo adduct 30 in
88% yield (Scheme 10). The reaction of diene 27 with
diethylacetylene dicarboxylate also proceeded smoothly to afford
in 81% yield the tetra-substituted benzene derivative 31, wherein
the cycloadduct had aromatized through the loss of dimethyl-
amine.27 The reaction conditions used for the phenyl-substituted
diene are comparable to that used earlier for the parent
dimethylaminosiloxydienes, indicating a similar reactivity profile
for the two dienes.
Cycloaddition of the phenyl-substituted diene 27 with meth-
acrolein was particularly significant, since it allowed us to test
not only the efficiency but also the diastereoselctivity of this
reaction. To our delight, the reaction produced a single
cycloadduct diastereomer, assigned the endo stereochemistry
(32) based on its spectroscopic characteristics (Scheme 10).
Treatment of 32 with an aqueous solution of HCl in THF
accomplished the desired hydrolysis to enone 33 in good yield.
The cycloaddition of 27 with methyl acrylate afforded a 1:1
mixture of the readily separable endo and exo isomers. Each
diastereomer was efficiently converted to 4-(hydroxymethyl)-
2-phenylcylohexenone (35) using lithium aluminum hydride
reduction and an HF-mediated elimination sequence. These
examples demonstrate the synthetic utility of 1-amino-2-phenyl-
3-siloxydiene (27) for the assembly of substituted 2-phenyl-2-
cyclohexenones, which are valuable intermediates for complex
molecule synthesis. The present method represents a novel, more
direct alternative to the currently available methods to such
compounds, which involve a multistep arylation of a cyclohex-
enone precursor.28
Unfortunately, we were unable to extend the chemistry
outlined above to the preparation of the dienes 36 and 37, which
(26) Ruel, F. C.; Braun, M. P.; Johnson, C. R. Org. Synth. 1997, 75, 69-77.
(27) Dimethylaminomaleate was obtained as a byproduct in this reaction when
2 equiv of diethylacetylene dicarboxylate was employed. Presumably, the
byproduct results from the Michael addition of the dimethylamine, which
is produced by the elimination of the initially formed cycloadduct under
the reaction conditions (see Supporting Information).
(23) (a) Bennett, G. B.; Mason, R. B., Org. Prep. Proc. Int. 1978, 10, 67-72.
(b) Haefliger, W.; Hauser, D. Synthesis 1980, 236-238.
(24) For reviews, see: (a) Miyaure, N., Suzuki, A. Chem. ReV. 1995, 95, 2457-
2483. (b) Suzuki, A. Pure Appl. Chem. 1991, 63, 419-422.
(25) (a) Hodgson, D. M.; Witherington, J.; Moloney, B. A.; Richards, I. C.;
Brayer, J.-L. Synlett. 1995, 32-34. (b) Campos, P. J.; Tan, C.-Q.;
Rodriguez, M. A. Tetrahedron Lett. 1995, 36, 5257-5260.
(28) Wender, P. A.; Erhardt, J. M.; Letendre, L. J. J. Am. Chem. Soc. 1981,
103, 2114-2116.
9
J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002 4633

A R T I C L E S
Kozmin et al.
Scheme 12
o-nitrophenylation/reduction sequence represents a new syn-
thesis of indoles, one that allows, for the first time, the
regiospecific synthesis of indoles from the corresponding enol
silyl ethers.30 The generality of the sequence was evaluated and
the results were recently published.30 A clear advantage of this
method is that it allows for the regiocontrolled synthesis of
various substituted indoles by simply using the requisite
aryliododonium salts. The formation of the 6-methoxy-
substituted indole is illustrated through the asymmetric synthesis
of 16-methoxytabersonine (vide infra).
Pentacycle Assembly and End Game. Having accomplished
the stereocontrolled synthesis of the angular tetracycle, what
remained was the development of an efficient end game to the
Aspidosperma skeleton. Completion of the Aspidosperma pen-
tacycle required the seemingly simple task of introducing a two-
carbon unit to form the pyrrolidine ring that is spiro-fused to
the indole (eq 2). This alkylation route to the spiroindolenine
possess, in a masked form, the amino group in the ortho position
required for the indolization. The precursors to the diene were
found to be incompatible with the silylation protocol.
(o-Nitrophenyl)phenyliodonium Fluoride (NPIF): A New
Reagent for Regiocontrolled Indole Synthesis. The difficulties
encountered in attempting to selectively introduce the indole
unit using available methodology prompted us to develop a new
reagent capable of carrying out this transformation with high
efficiency and complete regioselectivity. The objective was to
find a reagent that would allow introduction of the o-nitrophenyl
groupsor some equivalent thereofsdirectly to an enol ether.
The arylated ketone could then be converted to the desired indole
in a regiospecific manner. Aryliodonium salts were particularly
attractive as arylating agents, especially since Koser had shown
that simple phenylation of ketone enol silyl ethers was pos-
sible.29 The specific arylating agent that was required is NPIF
(39),30 which was synthesized by the two-step sequence outlined
in Scheme 11. The oxidation of o-iodonitrobenzene with K2S2O8
in benzene afforded a mixed diaryliodonium salt. Treatment of
the crude hydrogen sulfate salt with aqueous KI gave (o-
nitrophenyl)phenyliodonium iodide (38) as an orange solid in
76-87% yield. The exchange of the iodide counterion with a
fluoride using aqueous AgF afforded the desired unsymmetrical
nitrophenylating reagent (39) in 66-78% yield. These two
compounds have been prepared on >25- and >10-g scales,
respectively, without any complications. Iodonium iodide 38
can also be prepared using CrO3 as the oxidant.31
unit had been demonstrated early in the Aspidosperma alkaloids
area through the groundbreaking work of Ziegler et al.,34 but
the reported efficiency of the process was variable. Indeed, to
circumvent the difficulties associated with the alkylative
introduction of the two-carbon unit, Magnus developed an
elegant, Pummerer rearrangement-based sequence.11b Ultimately,
since the most concise route to tabersonine was through the
intramolecular SN2 alkylation at the â-position of the indole,
we examined variations of the Ziegler strategy. Additional
precedent for this route came from the work of Natsume et al.,34b
who had constructed the spiropyrrolidine ring by alkylative
displacement of a mesylate upon deprotonation of indole with
potassium tert-butoxide in DMSO or DMF. The reported yield
for the process was, again, quite low (26-32%). The most
encouraging precedent for this sequence was provided by
Rubiralta and co-workers,34d who had accomplished the pyrro-
lidine ring closure in a related system in 70% yield by
displacement of an in situ generated tosylate.
Scheme 11
On the basis of the work of Beringer et al.,32 the more electron
deficient o-nitrophenyl group was expected to transfer during
the arylation. To our delight, treatment of enol ether 20 with
NPIF regiospecifically produced the desired R-arylated ketone
40 in 94% yield, as a single diastereomer (Scheme 12). The
nitro group was then cleanly reduced with TiCl3,33 wherein the
intermediate aniline underwent spontaneous condensation to
yield the angular tetracyclic indole 24 in 89% yield. This
In preparation for the spiroindolenine synthesis, the methyl
carbamate group on 24 was first removed with trimethylsilyl
iodide in refluxing CH2Cl2.35,36 Alkylation of the resulting
secondary amine 41 with a 10-fold excess of bromoethanol in
the presence of sodium carbonate in refluxing ethanol for 18 h
afforded a quantitative yield of the desired tertiary amine 42,
poised for construction of the spiroindolenine (Scheme 13).37
(29) Chen, K.; Koser, G. F. J. Org. Chem. 1991, 56, 5764-5767 and references
therein.
(30) The scope of this two-step indole synthesis has been examined: Iwama.
T.; Birman, V. B.; Kozmin, S. A.; Rawal, V. H. Org. Lett, 1999, 1, 673-
676.
(31) (a) Kazmierczak, P.; Skulski, L. Synthesis 1995, 1027-1032. (b) Kaz-
mierczak, P.; Skulski, L. Bull. Chem. Soc. Jpn. 1997, 70, 219-224.
(32) (a) Beringer, F. M.; Gindler, E. M.; Rapoport, M.; Taylor, R. J. J. Am.
Chem. Soc. 1959, 81, 351-361. (b) Beringer, F. M.; Forgione, P. S.; Yudis,
M. D. Tetrahedron 1960, 8, 49-63.
(33) (a) Ho, T.-L.; Wong, C. M. Synthesis 1974, 45. (b) Moody, C. J.;
Rahimtoola, K. F. J. Chem. Soc., Perkin Trans. 1 1990, 673-679.
(34) (a) Ziegler, F. E.; Spitzner, E. B. J. Am. Chem. Soc. 1973, 95, 7146-
7149. (b) Natsume, M.; Utsunomiya, I. Heterocycles 1982, 17, 111-115.
(c) Ogawa, M.; Kitagawa, Y.; Natsume, M. Tetrahedron Lett. 1987, 28,
3985-3986. (d) Forns, P.; Diez, A.; Rubiralta, M. J. Org. Chem. 1996,
61, 7882-7888 and references therein.
(35) Olah, G. A.; Narang, S. C. Tetrahedron 1982, 38, 2225-2277.
9
4634 J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002

Aspidosperma Alkaloids
A R T I C L E S
at C1 and C2, the central C ring is proposed to exist in the boat
conformation. The selected NOESY cross-peaks are in full
agreement with the proposed structure and cis stereochemistry
Scheme 13
at C12, C19, and C5. The 500-MHz 1H NMR and 125-MHz 13
C
NMR spectra of our synthetic sample of 43 were identical to
those recently reported in the literature.39g Significantly, a one-
pot modification of this two-step procedure was subsequently
developed (vide infra).
The selective introduction of the carbomethoxy group at C1
was necessary at this stage in order to complete the total
synthesis of tabersonine (1). As was true for the spiroindolenine
formation, the introduction of the carbomethoxy group by direct
acylation of the aza enolate had precedent,11a but the reaction
was reported to give ∼1:1 mixture of the C-acylation and
N-acylation products. Indeed, following Overman’s protocol,
when imine 43 was treated with an excess of LDA, followed
by addition of methyl chloroformate at -78 °C, tabersonine
(rac-1) was obtained in 37% yield, along with a significant
amount of the expected N-acylation product. The C versus N
selectivity issue was solved by using cyanomethyl formate
(Mander’s reagent),40 which is known to be a “softer” acylating
agent and to promote C-acylation over O-acylation. Deproto-
nation of 43 with LDA in THF followed by a quick injection
of Mander’s reagent40 cleanly and selectively produced the
C-acylated product, (()-tabersonine (rac-1), isolated in 70-
1
80% yield (eq 3). The 500-MHz H NMR and 125-MHz 13C
The cyclization protocol reported by Rubiralta and co-workers
was examined first, but with unsatisfactory results. Treatment
of 42 with t-BuOK followed by addition of p-toluenesulfonyl
chloride afforded the desired pentacycle, (()-didehydroaspi-
dospermine (43), but in a variable yield (30-50%, Scheme
13).34d The poor reproducibility of the reaction was attributed
to the competing tosylation of the indole nitrogen, which
prevented the desired alkylative cyclization from taking place.
To improve the reliability of the protocol, we examined a
two-step alternative, wherein first the alcohol would be selec-
tively activated and then the ring closure would be ac-
complished. Treatment of amino alcohol 42 with Et3N and
methanesulfonyl chloride in CH2Cl2 resulted in clean formation
of a single compound, which was found not to be the expected
mesylate but the chloroamine 44, isolated in 90% yield (Scheme
13). The formation of 44 is understandable based on related
precedents. Displacement of the initially formed mesylate by
the proximate nucleophilic nitrogen atom of the piperidine ring
can produce aziridinium ion, which is then opened by the
chloride ion.38
NMR spectra of our synthetic sample of tabersonine were
identical to those previously reported in the literature.10d,e,41
Overall, this exercise established a novel, concise, and high-
yielding route to Aspidosperma alkaloids, culminating in the
synthesis of (()-tabersonine, wherein the overall regio- and
stereocontrol was predicated on the initial Diels-Alder reaction.
In the course of validating the strategy, a new method for
regioselective indole formation was developed, as were new
protocols for spiroindolenine construction and carbomethoxy-
lation R to imines.
(II) Gram-Scale Asymmetric Synthesis of (+)-Taberso-
nine. Installation of Chirality via Catalytic Asymmetric
Diels-Alder Reaction. The success of the tabersonine synthesis
motivated us to pursue our original, more ambitious goal, the
synthesis of vinblastine and vincristine. However, the develop-
ment of a meaningful synthesis of these bisindole alkaloids,
which arise from the coupling of two chiral components, requires
Since chloride 44 also represented a cyclization precursor, it
was treated with a solution of t-BuOK in deuterated THF. The
1H NMR analysis of the crude reaction mixture indicated a clean
formation of (()-didehydroaspidospermidine (43) as the sole
reaction product.39 The structure of 43 was determined by a
combination of COSY, NOESY, and HMQC experiments. On
the basis of the observed coupling constants between protons
(39) (a) Plat, M.; Men, J. L.; Janot, M. M.; Wilson, J. M.; Budzikiewicz, H.;
Durham, L. J.; Nakagawa, Y.; Djerassi, C. Tetrahedron Lett. 1962, 271-
276. (b) Zsadon, B.; Otta, K. Acta Chim. (Budapest) 1971, 69, 87-95. (c)
Zsadon, B.; Horvath-Otta, K. Herba Hung. 1973, 12, 133-138. (d)
Henriques, A.; Husson, H. P. Tetrahedron Lett. 1981, 22, 567-570. (e)
Henriques, A.; Kan, C.; Chiaroni, A.; Riche, C.; Husson, H. P.; Kan, S.
K.; Lounasmaa, M. J. Org. Chem. 1982, 47, 803-811. (f) Crippa, S.;
Danieli, B.; Lesma, G.; Palmisano, G.; Passarella, D.; Vecchietti, V.
Heterocycles 1990, 31, 1663-1667. (g) Dupont, C.; Guenard, D.; Tcher-
tanov, L.; Thoret, S.; Gueritte, F. Bioorg. Med. Chem. 1999, 7, 2961-
2969.
(36) (a) Rawal, V. H.; Michoud, C.; Monestel, R. F. J. Am. Chem. Soc. 1993,
115, 3030-3031. (b) Rawal, V. H.; Iwasa, S. J. Org. Chem. 1994, 59,
2685-2686.
(37) Rubiralta, M.; Diez, A.; Bosch, J.; Solans, S. J. Org. Chem. 1989, 54, 5591-
5597.
(38) Similar observations have been reported: (a) Lee, J. W.; Son, H. J.; Lee,
J. H.; Yoon, G. J.; Park, M. H. Synth. Commun. 1996, 26, 89-94. (b)
Springer, C. J.; Antoniw, P.; Bagshawe, K. D.; Searle, F.; Bisset, G. M.
F.; Jarman, M. J. Med. Chem. 1990, 33, 677-681.
(40) (a) Mander, L. N. In Encyclopedia of Reagents for Organic Synthesis;
Paquette, L. A., Ed.; Wiley: New York, 1995; Vol. 5, pp 3466-3469. (b)
Crabtree, S. R.; Mander, L. N.; Sethi, S. P. Org. Synth. 1991, 70, 256-
264 and references therein.
9
J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002 4635

A R T I C L E S
Kozmin et al.
Table 2. Multigram-Scale Catalytic Enantioselective Diels-Alder
that we have access to both components in enantiomerically
enriched form. With regard to the Aspidosperma component,
the absolute stereochemistry of this piece is determined at the
Diels-Alder step, so a method was required for carrying out
this cycloaddition in an asymmetric fashion. Inspired by this
objective, we investigated the possibility of carrying out the
Diels-Alder reactions of aminosiloxydienes using a chiral
Lewis acid.42,43 We discovered that Jacobsen’s chiral Cr(III)-
salen complex44 was very effective in catalyzing the Diels-
Alder reactions of 1-amino-3-siloxy-1,3-butadienes with a wide
range of R-substituted acroleins.1i,k In particular, the antimonate
salt of the complex afforded the cycloadducts in excellent yields
(>90%) and in high enantiomeric excesses (up to 97% ee).1i,k
As such, this process was well suited for the asymmetric
synthesis of Aspidoperma alkaloids.
Reactions
catalyst
(mol %)
temp
(°C)
time
(day)
yield
(%)a
ee
(%)b
entry
1
2
3
4
45a (1.3)
45a (1.9)
45b (1.0)
45b (2.0)
45b (5.0)
45b (5.0)
rt
rt
-40
-40
-40
-40
3
2.5
10
4.5
2
70
81
85
81
84
74
88
88.5
93
93
95
5
6c
2
94
a Isolated yield in two steps from 15. b Determined by Mosher ester
analysis of the alcohol prepared by LAH reduction followed by hydrolysis.
c 38 mmol of 15 was used.
We were pleased to find that the desired cycloaddition
proceeded well (eq 4). On a 1-mmol scale, the reaction between
Scheme 14
diene 16 and ethacrolein, catalyzed by 5 mol % of the
fluoroborate catalyst 45b, afforded the cycloadduct (ent-18) in
91% yield and 96% ee (determined by the Mosher ester analysis
of the corresponding alcohol).1i,45 Several other conditions were
examined to determine the optimum procedure for the multi-
gram-scale catalytic enantioselective Diels-Alder reactions
(Table 2). It was found that the cycloaddition could be carried
out more convenientlysat room temperature and using less than
2 mol % of the catalystsusing the fluoroborate catalyst 45a,
but with a reduction in the ee to <90% (entries 1 and 2). The
enantioselectivities were appreciably higher using the antimonate
catalyst 45b. The cycloadditions proceeded at -40 °C, in good
yields and high ee’s. The reactions were effective even at low
catalyst loadings (1-2 mol % of 45b), but the reaction times
were rather long (entries 3 and 4). The best results were obtained
using 5 mol % of catalyst 45b (entries 5 and 6). The absolute
stereochemistry of ent-18 would afford the antipode of natural
tabersonine.
The remainder of the tabersonine synthesis closely paralleled
the racemic synthesis, but practical improvements allowed
multiple steps to be performed between purifications (Scheme
14). Thus, after the Witting methylenation, addition of hexane
and filtration allowed facile removal of the insoluble triphen-
ylphosphine oxide. The crude product was subjected to ring-
closing metathesis using 4.3-7.5 mol % of Grubbs’s catalyst.16
After the metathesis was complete (by NMR and TLC), the
reaction mixture was concentrated, and the resultant crude
product (ent-20), without removal of Grubbs catalyst, was
treated with phenyl(o-nitrophenyl)iodonium fluoride (39) in
THF-DMSO. The product of this three-step sequence was
purified by column chromatography to yield multigram quanti-
ties of the desired, enantiomerically enriched o-nitroarylated
ketone (+)-ent-40 in 57-62% overall yield.
(41) Wen, R.; Laronze J.-Y.; Le´vy, J. Heterocycles 1984, 22, 1061-1065.
(42) Reviews on chiral Lewis acids in organic synthesis: (a) Noyori, R.
Asymmetric Catalysts in Organic Synthesis; Wiley: New York, 1994. (b)
Togni, A.; Venanzi, L. M. Angew. Chem., Int. Ed. Engl. 1994, 33, 497-
526. (c) Oh, T. Reilly, M. Org. Prep. Proc. Int. 1994, 26, 129-158. (d)
Corey, E. J.; Guzman-Perez, A. Angew. Chem., Int. Ed. Engl 1998, 37,
388-401. (e) Dell, C. P. J. Chem. Soc., Perkin Trans 1, 1998, 3873-
3905. (f) Yamamoto, H. Lewis Acid Reagents; Oxford University Press:
Oxford, 1999. (g) ComprehensiVe Asymmetric Catalysis; Jacobsen, E. N.,
Pfaltz, A., Yamamoto, H., Eds.; Springer: 1999; Vols. I-III. (h) Ojima, I.
Catalytic Asymmetric Synthesis, 2nd ed.; Wiley-VCH: New York, 2000.
(i) Hisashi Yamamoto, H. Lewis Acids in Organic Synthesis; Wiley-VCH:
New York, 2000.
(43) For reviews on the catalytic asymmetric Diels-Alder reactions, see: (a)
Kagan, H. B.; Riant, O. Chem. ReV. 1992, 92, 1007-1019. (b) Pindur, U.;
Lutz, G.; Otto, C. Chem. ReV. 1993, 93, 741-761. (c) Dias, L. C. J. Braz.
Chem. Soc. 1997, 8, 289-332. (d) Evans, D. A.; Johnson, J. S. In
ComprehensiVe Asymmetric Catalysis; Jacobsen, E. N., Pfaltz, A., Yama-
moto, H., Eds.; Springer: New York, 1999; Vol. III, pp 1177-1235.
(44) (a) Martinez, L. E.; Leighton, J. M.; Carsten, D. H.; Jacobsen, E. N. J.
Am. Chem. Soc. 1995, 117, 5897-5898. (b) Schaus, S. E.; Brånalt, J.;
Jacobsen, E. N. J. Org. Chem. 1998, 63, 403-405. (c) Schaus, S. E.;
Brånalt, J.; Jacobsen, E. N. J. Org. Chem. 1998, 63, 4876-4877.
(45) The enantiomeric excess was determined by 1H NMR analysis of the Mosher
ester 59, obtained through the sequence shown below. See Supporting
Information for details. The absolute stereochemistry of the adduct was
determined to be 6R by comparison of chemical shifts of the Mosher ester
in 1H NMR with the reported value.1b,e,i
As mentioned above, a strong point of the arylation protocol
is that it provides ready access to various indoles. The use of
substituted aryliodonium salts allows the preparation of indoles
that are substituted on the aromatic ring. In the context of the
present project, the synthesis of 16-methoxytabersonine was of
9
4636 J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002

Aspidosperma Alkaloids
A R T I C L E S
Scheme 15
Scheme 16
In contrast to the original, two-step sequence, a one-pot
procedure was developed for the transformation of alcohols ent-
42 and ent-51 to pentacycles ent-43 and ent-52, respectively.
The treatment of ent-42 and ent-51 with MsCl and Et3N in CH2-
Cl2 followed by addition of potassium tert-butoxide afforded
the desired pentacycles. Finally, the crude pentacycles (ent-43
and ent-52) were deprotonated with LDA, and the resultant
anions reacted with methyl cyanoformate40 to provide (+)-
tabersonine (ent-1) and (+)-16-methoxytabersonine (ent-7), the
antipodes of the natural products, in 33-39% (1.14-1.46-g
quantities) and 37% (1.11-g quantity) yields, respectively, for
the five steps from ent-40 and ent-47.46,47
As stated in the Introduction, there are scores of alkaloids in
the Aspidosperma family. One strength of our strategy is that it
provides easy access to many different members of this family
through advanced intermediates in the synthetic sequence. As
a demonstration of this, the pentacyclic intermediate ent-43 was
converted into (+)-aspidospermidine (53),34d,48 (-)-quebra-
chamine (54),10c,48r,s,49 and (-)-dehydroquebrachamine (55)
(Scheme 16).10b,49d,50 The reduction of the imine double bond
of ent-43 with NaBH4 in EtOH to dehydroaspidospermidine
interest as it provided a more direct route to vindoline (and
ultimately to vinblastine) than was possible through tabersonine.
What was required was a modified diaryliodonium salt, one that
could introduce the p-methoxy-o-nitrophenyl unit R to the
carbonyl group. The necessary reagent, p-methoxy-o-nitrophen-
yl(p-methoxyphenyl)iodonium fluoride, was prepared on a
multigram scale by the general procedure used for the prepara-
tion of 39.31 Thus, p-methoxy-o-nitroiodobenzene was oxidized
with CrO3 (the alternate procedure, vide supra) and the mixture
treated with anisole. Subsequent addition of KI afforded
iodonium iodide 48, which was then converted to the corre-
sponding iodonium fluoride 46 (eq 5).
(46) The 500-MHz 1H NMR and 125-MHz 13C NMR spectra of (+)-tabersonine
(ent-1) were identical to those of (()-tabersonine synthesized above. The
500-MHz 1H NMR and 125-MHz 13C NMR spectra of our synthetic sample
of (+)-16-methoxytabersonine (ent-7) were identical to those previously
reported in the literature. See: Lounasmaa, M, Kan, S.-K. Acta Chem.
Scand., B 1980, 34, 379-381 for 1H NMR and ref 11a for 13C NMR.
(47) These compounds are enantiomers of the natural products. The natural
compounds would be readily prepared using the other enantiomer of the
chiral salen catalyst.
The p-methoxy-o-nitrophenylation of ent-20 using reagent 46
gave the expected arylation product ent-47 in 31-35% isolated
yields for the three-step sequence starting from ent-18 (Scheme
14). The arylation with 46 was not fully optimized, and the
reaction was accompanied by a considerable amount of hydro-
lyzed ketone ent-23 (20-30%). The overall yield of ent-47 was
improved to 40% by using CH2Cl2-DMSO as cosolvents for
the arylation steps.
The enantiomerically enriched arylated products ent-40 and
ent-47 were transformed into (+)-tabersonine (ent-1) and (+)-
16-methoxytabersonine (ent-7), respectively, by the five-step
sequence shown in Scheme 15. Only the final products from
the multistep sequence were purified. The reduction of ent-40
and ent-47 with TiCl3 provided crude tetracyclic indoles ent-
24 and ent-49,33 respectively, which were treated with 2 equiv
of TMSI in refluxing CH2Cl2 to afford amines ent-41 and ent-
50.35,36 The unpurified free amines were then treated with
bromoethanol in EtOH in the presence of Na2CO3.37 After
standard workup, excess bromoethanol was removed under
reduced pressure, and the resulting crude alcohols (ent-42 and
ent-51) were used in the subsequent step without purification.
(48) Racemic aspidospermidine: (a) Camerman, A.; Camerman, N.; Kutney, J.
P.; Piers, E. Tetrahedron Lett. 1965, 637-642. (b) Kutney, J. P.;
Abdurahman, N.; Le Quesne, P.; Piers, E.; Vlattas, I. J. Am. Chem. Soc.
1966, 88, 3656-3657. (c) Harley-Mason, J.; Kaplan, M. Chem. Commun.
1967, 915-916. (d) Kutney, J. P.; Piers, E.; Brown, R. T. J. Am. Chem.
Soc. 1970, 92, 1700-1704. (e) Kutney, J. P.; Abdurahman, N.; Gletsos,
C.; Le Quesne, P.; Piers, E.; Vlattas, I. J. Am. Chem. Soc. 1970, 92, 1727-
1735. (f) Laronze, J. Y.; Laronze-Fontaine, J.; Levy, J.; Le Men, J.
Tetrahedron Lett. 1974, 491-494. (g) Ban, Y.; Yoshida, K.; Goto, J.; Oishi,
T. J. Am. Chem. Soc. 1981, 103, 6990-6992. (h) Gallagher, T.; Magnus,
P. J. Am. Chem. Soc. 1982, 104, 1140-1141. (i) Ban, Y.; Yoshida, K.;
Goto, J.; Oishi, T.; Takeda, E. Tetrahedron 1983, 39, 3657-3668. (j)
Gallagher, T.; Magnus, P.; Huffman, J. C. J. Am. Chem. Soc. 1983, 105,
4750-4757. (k) Wenkert, E.; Hudlicky, T. J. Org. Chem. 1988, 53, 1953-
1957. (l) Mandal, S. B.; Giri, V. S.; Sabeena, M. S.; Pakrashi, S. C. J.
Org. Chem. 1988, 53, 4236-4241. (m) Le Menez, P.; Kunesch, N.; Liu,
S.; Wenkert, E. J. Org. Chem. 1991, 56, 2915-2918. (n) Wenkert, E.;
Liu, S. J. Org. Chem. 1994, 59, 7677-7682. (o) Callaghan, O.; Lampard,
C.; Kennedy, A. R.; Murphy, J. A. J. Chem. Soc., Perkin Trans. 1 1999,
995-1002. (p) Toczko, M. A.; Heathcock, C. H. J. Org. Chem. 2000, 65,
2642-2645. (q) Patro, B.; Murphy, J. A. Org. Lett. 2000, 2, 3599-3601.
Asymmetric aspidospermidine syntheses: (r) Node, M.; Nagasawa, H.; Fuji,
K. J. Am. Chem. Soc. 1987, 109, 7901-7903. (s) Node, M.; Nagasawa,
H.; Fuji, K. J. Org. Chem. 1990, 55, 517-521. (t) Desmaele, D.; d’Angelo,
J. J. Org. Chem. 1994, 59, 2292-2303. (u) Schultz, A. G.; Pettus, L. J.
Org. Chem. 1997, 62, 6855-6861. (v) Iyengar, R.; Schildknegt, K.; Aube,
J. Org. Lett. 2000, 2, 1625-1627. See also ref 34d.
9
J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002 4637

A R T I C L E S
Kozmin et al.
Concentration of the volatiles in vacuo gave 44.1 g (96%) of N-allyl
carbamate (14) as a colorless oil, which was used without further
purification.
Vinylogous Imide 15. Method A. A solution of methyl allylcar-
bamate (14; 230 mg, 2.0 mmol) and acetylacetaldehyde dimethylacetal
(13; 0.57 mL, 4 mmol) in 5.0 mL of chloroform containing 20 mg of
TsOH was heated at gentle reflux for 27 h. Concentration of the reaction
mixture followed by direct flash chromatography on silica gel (50%
EtOAc/hexanes) afforded 330 mg (90%) of the vinylogous imide 15
as a colorless oil.
Method B. A solution of 14 (44.1 g, 383 mmol) and acetal 13 (52
mL, 392 mmol) in 1 L of chloroform containing 2 g of TsOH was
heated at gentle reflux for 44 h, during which time MeOH was removed
using a Soxhlet apparatus packed with 4-Å molecular sieves. The
reaction mixture was washed with saturated aqueous NaHCO3 (200
mL) and brine (200 mL), dried over MgSO4, and concentrated in vacuo.
The residue was distilled twice to give 31.3 g (45%) of vinylogous
imide 15 as a yellow oil.
Aminosiloxydiene 16. Method A. A solution of KHMDS in toluene
(0.5 M, 33.0 mL, 16.5 mmol) was diluted with THF (35 mL) and cooled
to -78 °C. To the resulting solution was added the vinylogous imide
15 (3.15 g, 17.2 mmol) in THF (20 mL) over 30 min. The reaction
was warmed to -50 °C over 2 h, cooled to -78 °C, and treated with
TBSCl (2.64 g, 17.5 mmol) in THF (15 mL). The reaction mixture
was warmed to room temperature, stirred for 30 min, diluted with ether
(350 mL), filtered through dry Celite, and concentrated in vacuo to
give essentially pure aminosiloxydiene 16 (∼100%) as an oil.
Method B. A solution of NaHMDS in THF (1 M, 22 mL, 22 mmol)
was diluted with THF (40 mL) and cooled to -78 °C. To this stirred
solution was added dropwise a solution of vinylogous imide 15 (3.66
g, 20 mmol) in THF (8.0 mL). After stirring at -78 °C for 1 h, the
reaction mixture was treated with TBSCl (3.62 g, 24 mmol) in THF
(4.0 mL). After 2 h, the cold reaction mixture was poured into ether
(400 mL), filtered through Celite, and concentrated in vacuo. The
residue was filtered again through Celite with hexanes (2 × 100 mL)
to afford essentially pure aminosiloxydiene 16 (6.00 g, (∼100%) as a
yellow oil, which was used without further purification for the next
step.
Diels-Alder Adduct 18. A solution of the diene 16 (4.56 g, 15.4
mmol) and ethylacrolein (2.0 mL, 20 mmol) in toluene (10 mL) was
heated first at 65 °C for 15 h and then at 85 °C for 33 h at which point
the NMR analysis of the reaction mixture indicated complete consump-
tion of the diene. Evaporation of volatiles under high vacuum afforded
5.69 g (97%) of the cycloadduct 18 in good purity as a colorless oil.
Wittig Methylenation of Aldehyde 18. To a suspension of
methyltriphenylphosphonium bromide (6.8 g, 19 mmol) in THF (40
mL) was added dropwise at 0 °C n-butyllithium in hexane (2.1 M, 8.0
mL, 17 mmol), and stirring was continued for additional 30 min. The
resulting yellow ylide solution was cooled to -78 °C and treated with
the aldehyde 18 (5.46 g, 14.3 mmol) in THF (30 mL). The resulting
slurry was warmed to room temperature and diluted with ether (150
mL) and water (100 mL). The aqueous layer was back-extracted with
ether (50 mL). The combined organic extracts were dried over
anhydrous MgSO4 and concentrated. The residue was dissolved in a
minimum amount of EtOAc and purified by flash column (20% EtOAc/
hexanes) to afford 4.6 g (85%) of 19 as a colorless oil.
followed by hydrogenation of the olefin with Pt2O in EtOH
afforded natural (+)-aspidospermidine (53) in 73% yield.
Treatment of ent-43 with NaBH3CN in AcOH accomplished
the fragmentation of the C-7-C-21 bond and the chemoselective
reduction of the resulting iminium ion to yield (-)-dehydro-
quebrachamine (55) in 68% yield. Hydrogenation of ent-43 with
Pt2O in AcOH promoted the same fragmentation as well as
reduction of both double bonds in the putative dihydropyri-
dinium intermediate and afforded (-)-quebrachamine (54) in
69% yield.
Summary and Conclusions
We have described here a novel, highly stereocontrolled route
to the Aspidosperma family of alkaloids. The strategy hinges
on a highly regio- and stereoselective [4 + 2] cycloaddition of
2-ethylacrolein with 1-amino-3-siloxydienes. The effectiveness
of the strategy was illustrated through the total synthesis of (()-
tabersonine, which was achieved by a concise sequence in
∼30% overall yield, the highest reported to date. The sequence
demonstrated (a) the use of an olefin metathesis reaction to
construct the cis-hexahydroquinoline ring system, having the
double bond correctly positioned for these alkaloids, (b) a novel
indole synthesis based on the regiocontrolled o-nitrophenylation
of an enol silyl ether using (o-nitrophenyl)phenyliodonium
fluoride, and (c) the high-yielding conversion of the ABDE
tetracycle into the pentacyclic target (()-1 via an intramolecular
indole alkylation and regioselective C-carbomethoxylation.
The strategy was readily adapted to the gram-scale asym-
metric synthesis of Aspidosperma alkaloids. The pivotal asym-
metry-introducing step was an enantioselective Diels-Alder
reaction catalyzed by a chiral Cr(III)-salen complex. The
cycloadduct was formed in high yield and with ∼95% ee, and
it was then carried forward to (+)-tabersonine and (+)-16-
methoxytabersonine, both of which were synthesized in >1-g
quantities. The synthetic sequence was easy to execute and
required only four purifications over the 12-step synthetic route.
The versatility of the strategy was further illustrated through
the asymmetric syntheses of (+)-aspidospermidine, (-)-dehy-
droquebrachamine, and (-)-quebrachamine.47
Experimental Section51
Methyl N-Allylcarbamate (14). Methyl chloroformate (30.9 mL,
0.40 mol) was added slowly to a chilled (0 °C) solution of allylamine
(66 mL, 0.88 mol) in CH2Cl2 (600 mL). The reaction mixture was
allowed to reach room temperature, stirred for 2 h, then washed
successively with 2 M aqueous HCl (200 mL), saturated aqueous
NaHCO3 (200 mL), and brine (200 mL), and dried over MgSO4.
(49) Racemic quebrachamine syntheses: (a) Stork, G.; Dolfini, J. E. J. Am.
Chem. Soc. 1963, 85, 2872-2873. (b) Kuehne, M. E.; Bayha, C.
Tetrahedron Lett. 1966, 1311-1315. (c) Ziegler, F. E.; Kloek, J. A.; Zoretic,
P. A. J. Am. Chem. Soc. 1969, 91, 2342-2346. (d) Hoizey, M. J.; Olivier,
L.; Levy, J.; Le Men, J. Tetrahedron Lett. 1971, 1011-1014. (e) Takano,
S.; Hatakeyama, S.; Ogasawara, K. J. Am. Chem. Soc. 1976, 98, 3022-
3023. (f) Takano, S.; Hirama, M.; Araki, T.; Ogasawara, K. J. Am. Chem.
Soc. 1976, 98, 7084-7085. (g) Giri, V. S.; Ali, E.; Pakrashi, S. C. J.
Heterocycl. Chem. 1980, 17, 1133-1134. (h) Takano, S.; Murakata, C.;
Ogasawara, K. Heterocycles 1981, 16, 247-249. (i) Wenkert, E.; Halls,
T. D. J.; Kwart, L. D.; Magnusson, G.; Showalter, H. D. H. Tetrahedron
1981, 37, 4017-4025. See also refs 39a and 48b,e,g,i,k,l. Asymmetric
quebrachamine syntheses: (j) Takano, S.; Chiba, K.; Yonaga, M.; Ogasawara,
K. J. Chem. Soc., Chem. Commun. 1980, 616-617. (k) Takano, S.; Yonaga,
M.; Ogasawara, K. J. Chem. Soc., Chem. Commun. 1981, 1153-1155. (l)
Asaoka, M.; Takei, H. Heterocycles 1989, 29, 243-244. (m) Temme, O.;
Taj, S.-A.; Andersson, P. G. J. Org. Chem. 1998, 63, 6007-6015. See
also refs 10c, and 48r,s.
Representative Ring-Closing Metathesis of 19. To a degassed
solution of 19 (130 mg, 0.34 mmol) in benzene (4.0 mL) under nitrogen
atmosphere at 20 °C was added at Schrock’s molybdenum catalyst in
benzene (0.13 mL of 100 mg of the catalyst in 1.0 mL of benzene,
0.02 mmol, 5 mol %). The solution was heated to 60 °C for 1 h, at
which point, TLC analysis indicated the reaction was complete.
(51) The experimental procedures that lead to the target molecules are shown
here. The remaining procedures, general experimental guidelines, and all
spectral data and copies of actual spectra can be found in the Supporting
Information.
(50) (a) Ziegler, F. E.; Bennett, G. B. Tetrahedron Lett. 1970, 2545-2547. (b)
Wenkert, E.; Hagaman, E. W.; Wang, N.-Y.; Kunesch, N. Heterocycles
1979, 12, 1439-1443. See also refs 10b and 49d.
9
4638 J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002

Aspidosperma Alkaloids
A R T I C L E S
Examination of the crude reaction product, after evaporation of the
solvent, showed complete conversion to the product. Chromatographic
purification (20% ether/hexanes) afforded 105 mg (88%) of 20 as a
colorless oil.
Cleavage of Carbamate 24. To a solution of the carbamate 24 (198
mg, 0.64 mmol) in CH2Cl2 (9 mL) was added under nitrogen
atmosphere iodotrimethylsilane (0.21 mL, 1.45 mmol). The reaction
mixture was heated to a gentle reflux for 1 h. Methanol (0.5 mL) was
added, and the resulting solution concentrated. The residue was
partitioned between 1.2 M aqueous HCl (20 mL) and ether (60 mL).
The ethereal layer was washed with 1.2 M aqueous HCl (2 × 10 mL).
The combined aqueous layers were neutralized at 0 °C with 30%
aqueous NaOH. The resulting aqueous phase was reextracted with CH2-
Cl2 (3 × 25 mL). The combined organic layers were dried with
anhydrous Na2SO4 and concentrated to afford 147 mg (91%) of 41 as
a pale yellow solid.
Alkylation of Amine 41 with Bromoethanol. To a solution of 41
(30 mg, 0.12 mmol) in absolute ethanol (1.0 mL) were sequentially
added freshly distilled bromoethanol (0.85 mL, 1.2 mmol) and
anhydrous sodium carbonate (127 mg, 1.2 mmol). The resulting
suspension was heated to reflux for 18 h, and the reaction was monitored
by TLC. The reaction mixture was diluted with EtOAc, filtered, and
concentrated, and the residue was purified by flash chromatography
(10% MeOH/EtOAc) to afford 35 mg (100%) of the alkylated product
42 as a white solid.
Preparation of Chloroethyl Tetracycle 44. To a CH2Cl2 (10 mL)
solution of amino alcohol 42 (72 mg, 0.24 mmol) and triethylamine
(51 mg, 70 µL, 0.50 mmol) was added methanesulfonyl chloride (34
mg, 23 µL, 0.30 mmol) at -20 °C. The reaction mixture was allowed
to reach room temperature and stirred for an additional 30 min.
Saturated aqueous solution of NaHCO3 was added, and the aqueous
layer was extracted with CH2Cl2. The combined organic layers were
dried over Na2SO4 and concentrated. Flash chromatography of the crude
product (50% ether/hexanes) yielded 68 mg (90%) of 44 as a white
solid, which slowly decomposes at room temperature.
(o-Nitrophenyl)phenyliodonium Iodide (38). Method A.30 To a
stirred solution of o-iodonitrobenzene (20.0 g, 80 mmol) in H2SO4 (450
mL) was added K2S2O8 (24.0 g, 88 mmol) in small portions followed
by benzene (100 mL) at room temperature, and the mixture was stirred
vigorously for 3 h. The reaction mixture was poured over ice (total
volume 1.6 L), and the insoluble material was removed by filtration.
The filtrate was treated with aqueous KI (66 g/100 mL), causing an
orange precipitate that was filtered and washed thoroughly with H2O
(∼2 L) followed by acetone (38 is slightly soluble in acetone). The
collected solid was dried over P2O5 under reduced pressure to afford
27.8 g (76%) of the title compound as an orange powder.
Method B. A mixture of CrO3 (5.87 g, 59 mmol), AcOH (40 mL),
and Ac2O (20 mL) was stirred at room temperature for ∼45 min. The
resulting red brown solution was treated with 2-iodonitrobenzene (19.92
g, 80 mmol) followed by dropwise addition of concentrated H2SO4
(14 mL), while maintaining the temperature below 25 °C (water bath).
After 15 min, benzene (14.2 mL, 160 mmol) was added dropwise, while
continuing to keep the temperature below 25 °C. The resulting mixture
was heated at ∼55 °C for 1.5 h, cooled in an ice bath, and then poured
into cold 30% MeOH in H2O (200 mL). The resulting precipitate was
removed by filtration, and the filtrate was treated with saturated aqueous
Na2S2O3 (3 mL) followed by aqueous KI (16.0 g, 96 mmol/45 mL of
H2O). After stirring for 1 h, the orange precipitate was filtered and
washed with cold H2O until the filtrate became colorless (∼1 L), then
with cold MeOH (∼130 mL, 38 is slightly soluble in MeOH). The
collected solid was dried over P2O5 in vacuo to afford 38 as an orange
powder (24.3 g, 67%).
Preparation of Didehydroaspidospermidine (43). To a solution
of chloramine 44 (60 mg, 0.19 mmol) in THF (6.0 mL) was added
dropwise a 0.4 M solution of t-BuOK in THF (0.71 mL, 0.28 mmol),
causing the formation of a white precipitate. The reaction mixture was
stirred for 16 h at 20 °C. Excess base was quenched by addition of
small amounts of water (0.10 mL). The resulting mixture was diluted
with ether (40 mL), dried with anhydrous Na2SO4, filtered, and
concentrated. The residue was redissolved in CH2Cl2 (7 mL), dried
further with anhydrous Na2SO4, and filtered. Concentration under
vacuum gave 46 mg (87%) of 43 as a yellow foamy solid which was
used directly in the next step. Further purification of 43 was ac-
complished using preparative TLC (CHCl3/MeOH/Et3N, 300:9:4). The
complete structural assignment was established by COSY, HMQC, and
NOESY experiments. See Supporting Information.
(()-Tabersonine (rac-1). A solution of LDA was prepared by
addition of n-BuLi (2.5 M solution in hexane, 80 µL, 0.20 mmol) to
diisopropylamine (65 µL, 0.50 mmol) in THF (0.8 mL) at -70 °C.
The resulting clear solution was stirred for 20 min at -70 °C and slowly
treated with a solution of imine 43 (28 mg, 0.10 mmol) in THF (0.7
mL). The reaction mixture was allowed to warm to -20 °C over a
1.0-h period and then cooled to -70 °C, and methyl cyanoformate (20
µL, 0.25 mmol) was added in one portion. The reaction mixture was
diluted with ether (6 mL) and 1.2 M aqueous solution of HCl (3 mL).
The aqueous phase was separated and carefully neutralized by addition
of saturated aqueous NaHCO3 solution. The product was extracted into
CH2Cl2 (2 × 5 mL), and the combined organic layers were dried over
Na2SO4 and concentrated to give 27 mg (80%) of tabersonine as a pale
yellow solid that was clean by NMR. 1H and 13C NMR spectra for this
sample were in good agreement with those described in the literature.
Catalytic Asymmetric Diels-Alder Reaction. To a chilled (-40
°C), stirred mixture of catalyst 45b (50 mg, 0.05 mmol) and oven-
dried, powdered 4-Å molecular sieves (0.8 g) in CH2Cl2 (1.0 mL) were
added ethacrolein (196 mL, 2.0 mmol) and diene 16 (296 mg, 1.0
mmol). The stirred mixture was kept at -40 °C for 2 days. The
molecular sieves were removed by filtration through Celite and washed
(o-Nitrophenyl)phenyliodonium Fluoride (NPIF, 39).30 To a
vigorously stirred solution of AgF (5.07 g, 40 mmol) in H2O (80 mL)
was added in small portions the iodonium iodide 38 (18.12 g, 40 mmol).
The reaction was moderately exothermic. After the mixture was stirred
for 3 h, the insoluble material was removed by filtration, and the filtrate
was concentrated below 20 °C under reduced pressure (<1 Torr). The
residue was suspended in MeCN (20 mL), to which was added ether
(20 mL). The mixture was chilled to 0 °C, and the resulting crystalline
product was filtered and washed with a small amount of MeCN
followed by ether to give 10.45 g (76%) of NPIF as a brown solid.
Impure NPIF gradually turns dark, but it can still be used for
o-nitrophenylation.
o-Nitrophenylation of Enol Ether 20. To a solution of silyl enol
ether 20 (250 mg, 0.71 mmol) in 2.0 mL of 1:2 THF-DMSO solution
was added in one portion the NPIF iodonium salt 3930 (287 mg, 0.83
mmol). The reaction mixture darkened and was stirred for 1.5 h at 20
°C, at which point TLC analysis revealed complete disappearance of
starting material. The reaction mixture was diluted with water (50 mL)
and extracted with ether (3 × 30 mL). The combined organic layers
were dried over anhydrous Na2SO4, filtered, and concentrated. The crude
product was purified by flash chromatography (40% EtOAc/hexanes)
to give 40 (236 mg, 94%) as a pale yellow solid.
Reductive Indolization of Nitrophenyl Ketone 40. To a solution
of ammonium acetate (1.15 g, 15 mmol) in water (6.0 mL) was added
1.44 M TiCl3 in water (3.7 mL, 5.3 mmol). The resulting dark purple
solution was diluted with acetone (6.0 mL) and slowly treated with
nitrophenyl ketone 40 (236 mg, 0.66 mmol) in acetone (6.0 mL). The
reaction mixture was stirred for 15 min at 20 °C, diluted with water
(100 mL), and extracted with EtOAc (3 × 60 mL). The combined
organic layers were washed with saturated solution of NaHCO3, dried
over anhydrous Na2SO4, filtered, and concentrated. The crude product
was purified by flash chromatography (33% EtOAc/hexanes) to give
182 mg (89%) of the desired tetracyclic carbamate 24 as a white foamy
solid.
9
J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002 4639

A R T I C L E S
Kozmin et al.
with CH2Cl2. The filtrate was concentrated in vacuo and the residue
purified by flash column chromatography (5% EtOAc/hexane containing
1% Et3N) to give 346 mg (91%) of cycloadduct ent-18 as a colorless
oil. The ee of ent-18 was determined to be 96% by the Mosher ester
analysis of the corresponding alcohol (see Supporting Information).
The absolute stereochemistry of ent-18 was determined to be 6R by
crude amine ent-41 as a greenish brown solid, which was used in the
next step without purification.
(3) Alkylation of ent-41. A solution of crude ent-41 (2.63 g, assumed
to be 9.63 mmol) in absolute ethanol (96 mL) was treated with
bromoethanol (2.73 mL, 38.5 mmol) followed by anhydrous sodium
carbonate (4.08 g, 38.5 mmol). The resulting suspension was heated at
reflux for 15 h. After cooling, the reaction mixture was concentrated
under reduced pressure, and the residue was suspended in ether (50
mL) and water (50 mL). The ether layer was separated, and the aqueous
layer was back-extracted with ether (50 mL). The combined ether
extracts were washed with brine (50 mL), dried over MgSO4, and
concentrated in vacuo to afford 3.37 g of crude alcohol ent-42 as a
dark brown gum, which was used without purification in the next step.
1
comparison of chemical shifts of the Mosher ester in H NMR with
the reported value.1b,e,i
Multigram-Scale Catalytic Asymmetric Diels-Alder Reaction.
The reaction was carried out as above but using 1.0 g of catalyst 45b
(83% purity, 1.0 mmol), 16 g of oven-dried powdered 4-Å sieves, 20
mL of CH2Cl2, 2.35 mL of ethacrolein (24 mmol), and 6.00 g of crude
diene 16 dissolved in 5 mL of CH2Cl2. Chromatographic purification
afforded 6.43 g (84%) of ent-18 as a yellow oil in two steps from ketone
15. The ee of ent-18 was determined, as described above, to be 95%.
(4) Cyclization of ent-42. A stirred solution of crude alcohol ent-
42 (3.06 g, assumed to be 8.75 mmol) and Et3N (1.83 mL, 13.1 mmol)
in CH2Cl2 (44 mL) was cooled at -15 °C (ice-NaCl) and treated
dropwise with methanesulfonyl chloride (0.81 mL, 10.5 mmol). After
30 min, a solution of 1 M potassium tert-butoxide in THF (21.9 mL,
21.9 mmol) was added over 5 min, and the resulting mixture was
allowed to reach room temperature. After 45 min, the reaction mixture
was poured into brine (110 mL) and extracted with CH2Cl2 (2 × 50
mL). The extracts were washed with brine (50 mL), dried over MgSO4,
and concentrated in vacuo to afford 2.75 g of crude ent-43, a dark
brown oil that was used for the next step without purification.
Gram-Scale Synthesis of o-Nitrophenylated Ketone ent-40. (1)
Wittig Reaction. To a suspension of methyltriphenylphosphonium
bromide (7.79 g, 21.8 mmol) in THF (42 mL) cooled at 0 °C was added
dropwise a 2.47 M solution of n-butyllithium in hexane (8.2 mL, 20.3
mmol). The yellow ylide solution was stirred for 1 h at 0 °C, cooled to
-78 °C, and treated dropwise over 15 min with ent-18 (6.40 g, 16.8
mmol) in THF (34 mL). The cold bath was removed, and the mixture
was allowed to warm to room temperature. After dilution with ether
(180 mL), the organic phase was washed with water (120 mL). The
aqueous phase was back-extracted with ether (60 mL). The combined
organic phases were dried over MgSO4 and concentrated in vacuo. The
crude product was suspended in hexanes (150 mL), and the precipitated
triphenylphosphine oxide was removed by filtration. Concentration of
the filtrate in vacuo afforded 6.47 g of crude ent-19 as an orange oil,
which was used without purification.
(5) Introduction of Ester Group. To a solution of LDA (17.5 mmol)
in THF (53 mL) cooled at -70 °C was added dropwise a solution of
the crude imine ent-43 (2.75 g, assumed to be 8.75 mmol) in THF (20
mL) over 25 min. After 40 min, methyl cyanoformate (1.74 mL, 21.9
mmol) was added over 5 min. The resulting mixture was stirred at
-70 °C for 20 min and then poured into brine (150 mL). The organic
layer was separated, and the aqueous layer was extracted with ether (2
× 40 mL). The combined extracts were washed with brine (70 mL),
dried over MgSO4, and concentrated in vacuo. The residue was purified
by flash column chromatography (1/10 hexanes/??????) to give 1.14 g
(39% in five steps from ent-40) of (+)-tabersonine (ent-1) as a pale
yellow gum: The 1H NMR and 13C NMR spectra of this sample were
in full agreement with those of (()-tabersonine synthesized above.46
(p-Methoxy-o-nitrophenyl)-p-methoxyphenyliodonium Iodide (48).
A mixture of CrO3 (2.20 g, 22.0 mmol), AcOH (15 mL), and Ac2O (7
mL) was stirred at room temperature for ∼45 min. To the resulting
reddish brown solution was added 2-iodo-5-methoxynitrobenzene (8.37
g, 30 mmol) in one portion. The resulting mixture was treated dropwise
with H2SO4 (5 mL), while keeping the temperature below 25 °C (water
bath). After 15 min, anisole (3.59 mL, 33 mmol) was added dropwise.
The mixture was stirred for 30 min at room temperature, cooled to 0
°C, and poured into cold 30% MeOH in H2O (70 mL). The solids were
removed by filtration, and the filtrate was treated with saturated aqueous
Na2S2O3 (1.0 mL) followed by aqueous KI (6.0 g, 36 mmol/20 mL of
H2O). After stirring for 1 h, the orange precipitate that had formed
was isolated by suction filtration and washed with cold H2O (∼200
mL, until the filtrate was colorless) followed by cold MeOH (∼10 mL;
48 is slightly soluble in MeOH). The collected solid was dried over
P2O5 in vacuo to give 10.90 g (71%) of 48.
(2) Ring-Closing Metathesis of ent-19. To a solution of crude diene
ent-19 (6.47 g, 16.8 mmol) in CH2Cl2 (340 mL) under N2 was added
Grubbs’s catalyst (1.00 g, 1.22 mmol, 7.2 mol %). The resulting solution
was heated at gentle reflux for 44 h. The reaction mixture was
concentrated in vacuo to afford 7.04 g of crude bicycle ent-20, which
was used in the next step without purification.
(3) o-Nitrophenylation. To a stirred solution of ent-20 (7.04 g, crude
material obtained above, assumed to be 16.8 mmol) in DMSO-THF
(34 mL-17 mL) was added iodonium fluoride 39 (5.80 g, 16.8 mmol)
at once with partial cooling (ice bath). The reaction was exothermic.
After stirring at room temperature for 3.5 h, the reaction mixture was
poured into water (680 mL) and extracted with ether (3 × 400 mL).
The extracts were dried over Na2SO4 and concentrated in vacuo. The
residue was purified by flash column chromatography (EtOAc/hexane,
1:4-1:2) to give 3.54 g (59% in three steps from ent-18) of ent-40 as
a yellow oil.
Gram-Scale (+)-Tabersonine (ent-1). (1) Indole Formation. To
a stirred solution of ammonium acetate (16.96 g, 220 mmol) in H2O
(85 mL) was added a purple solution of 1.32 M TlCl3 in H2O (58 mL,
76.6 mmol) followed by acetone (85 mL). The resultant dark brown
mixture was stirred vigorously and treated dropwise over 15 min with
a solution of ent-40 (3.45 g, 9.63 mmol) in acetone (85 mL). After 30
min, the reaction mixture was concentrated to about half of the original
volume and then extracted with EtOAc (3 × 100 mL). The combined
organic extracts were washed with saturated aqueous NaHCO3 (2 ×
200 mL), dried over Na2SO4, and concentrated in vacuo to give 3.03
g of crude indole ent-24 as a yellowish-green foam, which was used
without purification in the next step.
(p-Methoxy-o-nitrophenyl)-p-methoxyphenyliodonium Fluoride
(46). To a vigorously stirred solution of AgF (5.07 g, 40 mmol) in
H2O (80 mL) was added the iodonium iodide 48 (20.52 g, 40 mmol;
exothermic), and the mixture was stirred for 3 h. After removal of
insoluble material by filtration, the solution was concentrated below
20 °C under reduced pressure (<1 Torr). To the residue was added 20
mL of MeCN, casuing a precipitate to form. Ether (20 mL) was added
to the suspension, and the resulting mixture was cooled at -0 °C. The
precipitate was filtered, washed with 1:1 MeCN/ether followed by ether
to give 12.43 g (77%) of the title compound as a brown solid. Some
samples of this compound darkened with time, presumably due to the
presence of Ag residues.
(2) Cleavage of the Carbamate Group of ent-24. Trimethylsilyl
iodide (3.2 mL, 22.8 mmol) was added to a stirred solution of crude
ent-24 (3.03 g, assumed to be 9.63 mmol) in CH2Cl2 (135 mL). The
reaction mixture was heated at gentle reflux for 1 h, cooled to room
temperature, and then treated with methanol (7.5 mL). The resulting
solution was washed with 1% aqueous NaOH (100 mL) and brine (70
mL), dried over Na2SO4, and concentrated in vacuo to afford 2.63 g of
9
4640 J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002

Aspidosperma Alkaloids
A R T I C L E S
p-Methoxy-o-nitrophenylated Ketone (ent-47). ent-47 was prepared
following the procedures used for the preparation of ent-40. (1) Wittig
Reaction. A total of 6.25 g of crude ent-19 was obtained as an orange
oil from methyltriphenylphosphonium bromide (7.57 g, 21 mmol) and
aldehyde ent-18 (6.20 g, 16.2 mmol).
(2) Ring-Closing Metathesis of ent-19. A total of 6.69 g of crude
bicycle ent-20 was obtained from 6.25 g of crude ent-19 (assumed to
be 16.2 mmol) using 0.99 g (1.2 mmol, 7.5 mol %) of Grubbs’s catalyst.
(3) p-Methoxy-o-nitrophenylation. The reaction of crude enol ether
ent-20 (5.68 g, assumed to be 14.0 mmol) with iodonium fluoride 46
(5.76 g, 14.0 mmol), carried out as described above, yielded after
chromatography 0.929 g (28% in three steps from ent-18) of hydrolyzed
ketone ent-23 as the less polar product and 1.743 g (32% in three steps
from ent-18) of ent-47 as the more polar product.
(+)-16-Methoxytabersonine (ent-7). ent-7 was synthesized follow-
ing the procedures used for the preparation of (+)-tabersonine (ent-1).
(1) Indole Formation. Crude indole ent-49 (1.84 g) was obtained as
a greenish brown foam from ketone ent-47 (2.02 g, 5.2 mmol) and
1.32 M TiCl3 in H2O (32 mL, 42 mmol).
(2) Cleavage of the Carbamate Group of ent-49. Amine ent-50
(2.54 g, crude) was obtained as a dark brown foam from ent-49 (2.91
g, crude material obtained from two batches, assumed to be 8.19 mmol)
and trimethylsilyl iodide (2.68 mL, 18.8 mmol).
dissolved in EtOH (1.5 mL) treated with Pt2O (3.5 mg, 0.016 mmol)
and placed under a hydrogen atmosphere (balloon) at room temperature
for 15 h. The reaction mixture was diluted with ether, washed twice
with water, dried (MgSO4), and concentrated under reduced pressure.
The residue was purified by flash column chromatography (1/2 ether/
hexane) to give 32 mg (73%) of (+)-aspidospermidine (53) as a light
1
yellow gum. The H NMR and 13C NMR spectra of this sample were
in agreement with those reported in the literature.48o,p
Synthesis of (-)-Quebrachamine (54). A solution of the pentacyclic
intermediate ent-43 (32 mg, 0.115 mmol) in AcOH (1.5 mL) was treated
with Pt2O (2.6 mg, 0.012 mmol) placed under a hydrogen balloon at
room temperature for 2 h. The reaction mixture was diluted with ether,
washed with 1 N NaOH aqueous solution, dried (MgSO4), and
concentrated under reduced pressure. The residue was purified by flash
column chromatography (1/2 ether/hexane) to give 22.3 mg (69%) of
(-)-quebrachamine (54) as a light yellow gum. The 1H NMR and 13
C
NMR spectra of this sample were in agreement with those reported in
the literature.49m
Synthesis of (-)-Dehydroquebrachamine (55). To a solution of
the pentacyclic intermediate ent-43 (29.3 mg, 0.105 mmol) in AcOH
(1.5 mL) at room temperature was added NaBH3CN (19.8 mg, 0.315
mmol) in one portion. After 1 h, the reaction mixture was diluted with
ether, washed with 1 N aqueous NaOH solution, dried (MgSO4), and
concentrated under reduced pressure. The residue was purified by flash
column chromatography (ether/hexane, 1/10) to give 20 mg (68%) of
(-)-dehydroquebrachamine (55) as a light yellow gum. The 1H NMR
and 13C NMR spectra of this sample were in agreement with those
reported in the literature.10b,50b
(3) Alkylation of ent-50. Alcohol ent-51 (2.91 g, crude) was obtained
as a dark brown gum from ent-50 (2.54 g, crude, assumed to be 8.19
mmol) and bromoethanol (2.32 mL, 32.7 mmol).
(4) Cyclization of ent-51. The cyclization of crude alcohol ent-51
(2.91 g, assumed to be 8.19 mmol) yielded pentacyclic imine ent-52
(3.01 g, crude), which was used without purification.
Acknowledgment. This work was supported by the National
Institutes of Health (Grant R01-GM-55998). Y.H. thanks Abbott
Laboratories for a Graduate Fellowship. Merck Research
Laboratories and Pfizer Inc. are also gratefully acknowledged
for financial support. Vladimir B. Birman is thanked for valuable
input on the o-nitrophenylation chemistry (see ref 30).
(5) Introduction of Ester Group. The acylation of crude imine ent-
52 (3.01 g, assumed to be 8.19 mmol) and methyl cyanoformate (1.62
mL, 20.4 mmol), as described for (+)-tabersonine (ent-1), afforded (+)-
16-methoxytabersonine (ent-7, 1.11 g, 37% in five steps from ent-47)
1
as a light yellow gum. The H NMR and 13C NMR spectra of this
sample were in agreement with those reported in the literature.46
Synthesis of (+)-Aspidospermidine (53). To a solution of the
pentacyclic intermediate ent-43 (43.3 mg, 0.156 mmol) in EtOH (1.5
mL) was added NaBH4 (17.7 mg, 0.468 mmol) at 0 °C. After stirring
for 30 min at 0 °C followed by 1 h at room temperature, the reaction
mixture was diluted with ether, washed twice with water, dried
(MgSO4), and concentrated under reduced pressure. The residue was
Supporting Information Available: Additional information
as noted in text (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
JA017863S
9
J. AM. CHEM. SOC. VOL. 124, NO. 17, 2002 4641