Stereoselective synthesis of (+)-aspidofractinine

Article abstract of DOI:10.1021/jo9009497

(Chemical Equation Presented) We describe the synthesis of (+)-aspidofractinine, the enantiomer of a naturally occurring alkaloid of the kopsane family. Key features of the synthesis include a stereospecific cyanate to isocyanate rearrangement on a chiral scaffold, a ring-closing alkene metathesis to cleave the chiral auxiliary, and a chemoselective cyclopropanation to introduce the quaternary carbon at position 7 of aspidofractinine.

Full text of DOI:10.1021/jo9009497

pubs.acs.org/joc
Stereoselective Synthesis of (þ)-Aspidofractinine
David Gagnon and Claude Spino*
ꢀ ꢀ ꢀ ꢀ
Departement de Chimie, Universite de Sherbrooke, 2500 Boulevard Universite, Sherbrooke, Quebec, Canada,
J1K 2R1
claude.spino@usherbrooke.ca
Received May 12, 2009
We describe the synthesis of (þ)-aspidofractinine, the enantiomer of a naturally occurring alkaloid of
the kopsane family. Key features of the synthesis include a stereospecific cyanate to isocyanate
rearrangement on a chiral scaffold, a ring-closing alkene metathesis to cleave the chiral auxiliary, and
a chemoselective cyclopropanation to introduce the quaternary carbon at position 7 of aspidofracti-
nine.
Introduction
reverse multidrug resistance in vincristine-resistant KB
cells.5
Aspidofractinine is a complex alkaloid1 isolated from the
leaves of Pleiocarpa tubicana and Aspidosperma refractum,
and its structure was elucidated in 1963 by Schmid and co-
workers (Figure 1).2 It has since been isolated from Aspidos-
perma pyrifolium3 and other Apocynaceae species.4 Aspido-
fractinine is considered the parent compound of a large
family of related indole alkaloids often isolated from the
genus Kopsia that are widely found throughout Southeast
Asia (Figure 1). Many members of this family were found to
The synthesis of aspidofractinine was first completed by
Ban and his collaborators in 1976, who obtained aspido-
fractinine as a racemic mixture.6 Since then, it has been
synthesized four times as a racemic mixture, including a
second synthesis from the Ban group.7 The kopsane alka-
loids have an unusually strained carbon skeleton with three
all-carbon quaternary centers, two of which are contiguous.
Retron 6 is central in our strategy, and it contains all the
carbon atoms needed to construct rings A-E of aspidofrac-
tinine (Scheme 1). We planned to access retron 6 rapidly and
stereoselectively from our chiral auxiliary p-menthane-3-car-
boxaldehyde 8 and use to advantage the fact that the latter
could be cleaved via a ring-closing metathesis reaction all the
while producing ring E of the target molecule. Cyclopropane
5 turned out to be a crucial intermediate in solving a nagging
problem of aromatization. As in the syntheses of Ban and
(1) Hesse, M. The Alkaloids, Nature’s Curse or Blessing? Wiley-VCH:
Zurich, 2002; 413 pp. (b) Fattorusso, E., Taglilatela-Scafati, O., Eds.
€
Modern Alkaloids; Wiley-VCH: Weinheim, Germany, 2008; 665 pp. (c) Saxton,
J. E. In The Alkaloids; Cordell, G. A., Ed.; Academic Press: New York, 1998; Vol.
51, pp 2-197.
(2) (a) Djerassi, C.; Budzikiewicz, H.; Owellen, R. J.; Wilson, J. M.;
Kump, W. G.; Le Count, D. J.; Battersby, A. R.; Schmid, H. Helv. Chim.
Acta 1963, 46, 742–751. (b) Guggiesberg, A.; Gorman, A.; Bycroft, B. W.;
Schmid, H. Helv. Chim. Acta 1969, 52, 76–89. (c) Patel, M. B.; Schumann, D.;
Bycroft, B. W.; Schmid, A. Helv. Chim. Acta 1964, 47, 1147–1153.
(3) Xavier de Araujo, J. J.; Antheaume, C.; Trindade, R. C. P.; Schmitt,
M.; Bourguignon, J. J.; Goulart, S. A. E. Phytochem. Rev. 2007, 6, 183–188.
(4) Lim, K.-H.; Hiraku, O.; Komiyama, K.; Koyano, T.; Hayashi, M.;
Kam, T.-S. J. Nat. Prod. 2007, 70, 1302–1307.
(6) Ban, Y.; Homna, Y.; Oishi, T. Tetrahedron Lett. 1976, 14, 1111–1114.
(7) (a) Ban, Y.; Ohnuma, T.; Oishi, T.; Kinoshita, H. Chem. Lett. 1986,
ꢀ
927–930. (b) Cartier, D.; Ouahrani, M.; Levy, J. T. Tetrahedron Lett. 1989,
30, 1951–1954. (c) Dufour, M.; Gramain, J. C.; Husson, H. P.; Sinibaldi, M.
E.; Troin, Y. Tetrahedron Lett. 1989, 30, 3429–3432. (d) Wenkert, E.; Liu, S.
J. Org. Chem. 1994, 59, 7677–7682.
(5) Subramaniam, G.; Hiraku, O.; Hayashi, M.; Koyano, T.; Komiyama,
K.; Kam, T.-S. J. Nat. Prod. 2008, 71, 53–57.
DOI: 10.1021/jo9009497
r
Published on Web 07/15/2009
J. Org. Chem. 2009, 74, 6035–6041 6035
2009 American Chemical Society

Gagnon and Spino
JOCArticle
SCHEME 2
FIGURE 1. Three alkaloids from the kopsane family.
SCHEME 1. Retrosynthesis of (þ)-Aspidofractinine
of its double bond. Fortunately, a one-pot sequence, invol-
ving the elimination of HBr with NaHMDS at -95 °C to
yield 12c, the generation of the alkynyllithium with n-BuLi,
and the addition of p-menthane-3-carboxaldehyde 8 to this
mixture, gave propargyl alcohol 13a. The two diastereomers
thus formed were more easily separable after reduction of the
triple bond with Red-Al, which gave a 71% yield of pure
allylic alcohol 14a after two steps, along with 20% of its pure
diastereomer 14b.
Transposing the chirality of the carbinol in 14a was
achieved using a cyanate-to-isocyanate rearrangement.11
The most frequently used method to generate the cyanate
is to dehydrate the corresponding carbamic acid, which in
this case is compound 15a (Scheme 3). Here, reaction time,
temperature, and acidity of the milieu were crucial to avoid
decomposition as alcohol 14a, carbamate 15a, and isocya-
nate 16a are all prone to decomposition. Fortunately, an
excellent yield for the sequence 14a f 15a f 16a was
obtained after careful optimization (Scheme 3). The diaster-
eomeric ratio indicated for 16a was determined from the
Troc derivative 17a obtained by treatment with Cl3CCH2OH
and Ti(O-i-Pr)4.12 We later found that it was possible to add
2,2,2-trichloroethanol directly to the reaction flask after
dehydration of carbamate 15a, and without the need to
add a Lewis acid, to obtain a good yield of 17a in a single
step.
Wenkert and their co-workers, we elected to use a Diels-
Alder cycloaddition on intermediate 4 to introduce the C18-
C19 bridge and complete the synthesis.6,7d
Results and Discussion
To prepare one of the key intermediates in the synthesis,
namely, compound 7, we needed to stereoselectively add the
appropriate vinylmetal to aldehyde 8. The addition of
AlMe3, in catalytic amount or in excess, to a mixture of a
vinyllithium and a chiral aldehyde leads to a significant
increase in the expected Felkin-Ahn selectivity.8 We use
this methodology, whenever possible, to add vinyllithiums
to p-menthane-3-carboxaldehyde 8 and typically get ratios of
diastereomeric alcohols such as 7 in the range of 50-100:1.
However, preparing the requisite vinylbromide 11b
(Scheme 2) from aldehyde 109 proved impractical under a
variety of conditions, and its conversion to a vinyllithium
was problematic, giving several unidentified products after
addition of aldehyde 8. We speculated that the problem
stemmed from deprotonation elsewhere in the molecule,
particularly at the allylic position. In such a case, the addi-
tion of the corresponding alkynyllithium can be a suitable
alternative, giving ratios in the range of 5-10:1.
Having successfully introduced the C-N bond with con-
trol of its absolute stereochemistry, the chiral auxiliary could
now be cleaved, concomitantly forming ring E of aspido-
fractinine. Ring-closing metathesis (RCM) using 5 mol % of
Grubbs’ second generation catalyst provided compound 18
in 92% (Scheme 4).13 Byproduct 19 is volatile and can be
easily removed from the crude reaction mixture under re-
duced pressure. It can be recovered and converted back to
auxiliary 8 in good yield by ozonolysis.13
We thus set out to prepare alkyne 12b. The direct Sey-
ferth-Gilbert alkynylation method10 from 10 did not work,
and adding 2 equiv of n-BuLi to 11a (made from 10 using the
Corey-Fuchs procedure) resulted in a multitude of pro-
ducts, many of which had a propen-1-yl unit at the 2-position
of the indole moiety. It appears that the strong base can
deprotonate the allyl fragment, which allows the migration
(11) For a recent account, see (a) Ichikawa, Y. Synlett 2007, 2927–2936.
(b) Roy, S.; Spino, C. Org. Lett. 2006, 8, 939. (c) Ichikawa, Y.; Yamauchi, E.;
Isobe, M. Biosci. Biotechnol. Biochem. 2005, 69, 939–943. (d) Ichikawa, Y.;
Ito, T.; Isobe, M. Chem.;Eur. J. 2005, 11, 1949–1957. (e) Matsukawa, Y.;
Isobe, M.; Kotsuki, H.; Ichikawa, Y. J. Org. Chem. 2005, 70, 5339–5341. (f)
Nishiyama, T.; Isobe, M.; Ichikawa, Y. Angew. Chem., Int. Ed. 2005, 44,
4372–4375.
(8) Spino, C.; Granger, M.-C.; Tremblay, M.-C. Org. Lett. 2002, 4, 4735–
4737.
(9) 2-Allylindole 9 was prepared according to (a) Kondo, Y.; Yoshida,
A.; Sakamoto, T. J. Chem. Soc., Perkin Trans. 1 1996, 2331–2332. (b)
Saulnier, M. G.; Gribble, W. J. Org. Chem. 1982, 47, 757–761. Its acylation
to give 10 was carried out using the protocol of (c) Bennasar, M.-L.; Zulaica,
E.; Tummers, S. Tetrahedron Lett. 2004, 45, 6283–6285.
(10) Seyferth, D.; Marmor, R. S.; Gilbert, P. J. Org. Chem. 1971, 36,
1379.
(12) Spino, C.; Joly, M.-A.; Godbout, C.; Arbour, M. J. Org. Chem.
2005, 70, 6118–6121.
(13) (a) Spino, C.; Boisvert, L.; Douville, J.; Roy, S.; Lauzon, S.;
Minville, J.; Gagnon, D.; Chabot, C. J. Organomet. Chem. 2006, 691,
5336–5355. (b) Boisvert, L.; Beaumier, F.; Spino, C. Can. J. Chem. 2006,
84, 1290–1293.(c) Grubbs, R. H., Ed. Handbook of Metathesis; Wiley-VCH:
Weinheim, Germany, 2003; Vols. 1-3.
6036 J. Org. Chem. Vol. 74, No. 16, 2009

Gagnon and Spino
JOCArticle
SCHEME 3
SCHEME 4
Carbamate 18 and many homologous compounds aroma-
tize rather easily to give the known carbazole 21.14 This fact
dictated our strategy for the ensuing sequence of reactions.
Indeed, removal of the Troc protecting group gave a high
yield of amine 20, which simply could not be derivatized
under pain of aromatization to carbazole 21. Mancini as well
as Magnus and their respective co-workers have also experi-
enced a similar challenge, and the latter resorted to conceal-
ing one double bond in the structure until later in their
sequence when the system could no longer aromatize.15 We
instead elected to carry out the RCM cleavage after installing
the required carbon chains on the nitrogen, temporarily
postponing the aromatization problem.
In the event, the amine 22 was produced in quantitative
yield from 17a, and it was first alkylated and then acylated to
give 23a or 23b without difficulty (Scheme 5). RCM on either
compound proceeded uneventfully, at room temperature, to
give the central intermediates 24a or 24b. Note that RCM
could not be performed later in the sequence because any
attempt to make the C6-C7 bond on an open structure like
23a or 23b (cf. Scheme 5) would most likely lead to the
product with the wrong stereochemistry at C-7 or at best to
mixtures of stereoisomers.
Our initial design called for a biscyclization to convert 24
to 26, as shown in Scheme 6. Rather appealing, direct, and, in
principle, feasible, this strategy was nevertheless short-lived,
and we quickly realized that the C6-C7 bond could not be
prepared in this way. Aromatization was again the culprit,
and any attempt to effect the first cyclization (24a f 25)
failed, giving mostly carbazole 21 or its desulfonylated
analogue. Magnus,16 Rodriguez,17a Banwell,17b Heath-
cock,17c Rawal,17d and their respective research teams were
able to effect such an electrophilic cyclization on substrates
that could not aromatize. Like them, we varied the leaving
group in 24 (X=Cl, Br, OMs, OH), added silver triflate to
promote its departure, and tried removing the sulfone group
under reducing or basic conditions to augment the nucleo-
philicity of the indole, all to no avail.
Thinking that the aromatization problem could be cur-
tailed if a less potent leaving group replaced the amide, we
tried the same cyclization on tertiary amine 27. However,
aromatization to give carbazole 21 took place just as easily,
and in hindsight, it is probable that the formation of the
aziridinium 28 is fast, thus creating an even better leaving
group than the amide. The aziridinium functionality might
(16) Gallagher, T.; Magnus, P.; Huffman, J. C. J. Am. Chem. Soc. 1983,
105, 4750–4757.
(14) (a) Tsang, W. C. P.; Munday, R. H.; Brasche, G.; Zheng, N.;
Buchwald, S. L. J. Org. Chem. 2008, 73, 7603–7610. (b) Bennasar, M.-L.;
Zulaica, E.; Tummers, S. Tetrahedron Lett. 2004, 45, 6283–6285.
(15) (a) Biolatto, B.; Kneeteman, M.; Mancini, P. Tetrahedron Lett. 1999,
40, 3343–3346. (b) Magnus, P.; Gallagher, T.; Brown, P.; Huffman, J. C. J.
Am. Chem. Soc. 1984, 106, 2105–2114.
(17) (a) Rodriguez, J. G.; Urrutia, A. Tetrahedron 1999, 55, 11095–11108.
(b) Banwell, M. G.; Lupton, D. W. Org. Biomol. Chem. 2005, 3, 213–215. (c)
Toczko, M. A.; Heathcock, C. H. J. Org. Chem. 2000, 65, 2642–2645. (d)
Kozmin, S. A.; Iwama, T.; Huang, Y.; Rawal, V. H. J. Am. Chem. Soc. 2002,
124, 4628–4641.
J. Org. Chem. Vol. 74, No. 16, 2009 6037

Gagnon and Spino
JOCArticle
SCHEME 5
SCHEME 6
have been an excellent electrophile for the desired cyclization
had it been faster than aromatization.
with complete chemoselectivity to give a 76% isolated yield
of 31a.21 A single-crystal X-ray diffraction analysis revealed
both the relative and absolute stereochemistry of 31a (see
Supporting Information for the crystal structure and ORTEP
diagram). Pleased, we went about the business of closing ring
D. The radical cyclization of the iodide 31b gave 92% yield of
32. It is worth noting that the rigid structure of 31b is likely
responsible for this high-yielding radical cyclization. Indeed,
6-exotrigcyclizations tomake decalines areknowntobe poor-
yielding,22 and those to make decahydroquinolines appear to
be worse.23 In the latter case, hydrogen abstraction near the
nitrogen competes. In the case of compound 31b, the con-
formational rigidity may simultaneously hamper hydrogen
abstraction and assist radical cyclization. Several atom trans-
fer radical cyclization protocols were also investigated, all of
which proceeded without success.22
We also attempted a radical cyclization from 24a, thinking
that we could initiate the reaction at the haloketone in the
presence of a primary chloride. While this assumption was
correct, only reduced product (24c, X=H) was obtained.
This unfortunate result is nonetheless in line with results
published by the Magnus group, who obtained decomposi-
tion products when attempting to close the C6-C7 bond via
a radical pathway.16
We solved our problem by carrying out a chemoselective
cyclopropanation of the indole C5-C6 double bond. Indolic
alkenes have been cyclopropanated before,18 but we are aware
of only one report19 of an intermolecular chemoselective
cyclopropanation in a similar system involving a dibromo-
carbene. However, in that case, neither the yield nor the ratio
of products obtained was given. We prepared R-diazoketone
30 in one step from R-bromoketone 24b (Scheme 7).20 To our
delight, the Cu(I)-catalyzed cyclopropanation of 30 occurred
Reductive deprotection of the indole nitrogen in 32 with
the sodium anthracene radical anion concomitantly removed
the N-protecting group and ring-opened the cyclopropane to
(18) (a) Wenkert, E.; Alonso, M. E.; Gottlieb, H. E.; Sanchez, E. L. J.
Org. Chem. 1977, 42, 3945–6949. (b) Welstead, W. J.; Stauffer, H. F.;
Sancilio, L. F. J. Med. Chem. 1974, 17, 544–547. (c) Gnad, F.; Poleschak,
M.; Reiser, O. Tetrahedron Lett. 2004, 45, 4277–4280. (d) Salim, M.;
Capretta, A. Tetrahedron 2000, 56, 8063–8069. (e) Jung, M. E.; Slowinski,
F. Tetrahedron Lett. 2001, 42, 6835–6838.
(19) (a) Ashmore, J. W.; Radlick, P. C.; Helmkamp, G. K. Synth.
Commun. 1976, 6, 399–400. See also: (b) Dhanak, D.; Kuroda, R.; Reese,
C. B. Tetrahedron Lett. 1987, 28, 1827–1829.
(21) (a) Gnad, F.; Poleschak, P.; Reiser, O. Tetrahedron Lett. 2004, 45,
4277–4280. (b) Welstead, W. J. Jr.; Stauffer, H. F. Jr.; Sancilio, L. F. J. Med.
Chem. 1974, 17, 544–547. (c) Wenkert, E.; Alonso, M. E.; Gottlieb, H. E.;
Sanchez, E. L.; Pellicciari, R.; Cogolli, P. J. Org. Chem. 1977, 42, 3945–3949.
(d) Li, C.; Vasella, A. Helv. Chim. Acta 1993, 76, 197–210.
(22) Zard, S. Z. Radical Reactions in Organic Synthesis; Oxford Uni-
versity Press: New York, 2003; p 224.
(23) Lauzon, S.; Tremblay, F.; Gagnon, D.; Godbout, C.; Chabot, C.;
ꢁ
Mercier-Shanks, C.; Perreault, S.; DeSeve, H.; Spino, C J. Org. Chem. 2008,
73, 6239–6250.
(20) Toma,T.; Shimokawa, J.;Fukuyama,T. Org. Lett.2007, 9, 3195–3197.
6038 J. Org. Chem. Vol. 74, No. 16, 2009

Gagnon and Spino
JOCArticle
SCHEME 7
give imine 33 in 96% yield. These particular conditions are
necessary as, with many other reducing conditions (e.g.,
sodium amalgam), 33 suffers an additional and unwanted
reduction of the imine.
SCHEME 8
The synthesis of aspidofractinine was completed by oxida-
tion of 33 with phenylseleninic acid to the known intermedi-
ate 34, which had been prepared in a racemic form by
Wenkert and his group (Scheme 8).7d We transformed 33
into aspidofractinine 3 using their protocol as shown in
Scheme 8. When heated in benzene, enimine 34 is in equilib-
rium with its tautomer dienamine 4, which is able to react
with phenylvinylsulfone to give the Diels-Alder adduct 35.
The rest of the synthesis proceeded uneventfully to produce
(þ)-aspidofractinine 3. All spectral data were identical with
1
(24) We wish to thank Prof. J. Levy and Dr. D. Royer for a copy of the H
those reported in the literature,3,7b,24 except for the [R]D.
We measured a value of þ12 (c=0.27, CHCl3), which is close
in value but of opposite sign to that reported in the literature
[-14 (c=0.28, CHCl3)].2c There can be no doubt as to the
absolute stereochemistry of our synthetic material,25 and it is
the same as the one assumed by Schmid and co-workers for
natural 3 in 1964.2c,26
ꢀ
NMR spectrum of racemic 3 for comparison. See ref 7b.
(25) Our assignment of the stereochemistry of synthetic 3 is based on (a) a
single-crystal X-ray analysis of 31a using the Flack method for absolute
stereochemistry and (b) the expected stereochemistry from the sequence of
reactions that transformed alcohol 14a to isocyanate 16a as reported in ref
11b. The diastereomeric purity of alcohol 14a and isocyanate 16a was
checked against authentic diastereomers made from 14b. In any case, a loss
in enantiomeric purity would not account for a change in the sign of optical
rotation.
In conclusion, we have prepared for the first time optically
active (þ)-aspidofractinine 3 with the absolute configuration
shown in Scheme 8. The synthesis proceeded in 21 steps from
cheaply available indole. The overall yield is 2.1% with an
average yield of 83% per step (Scheme 9). Notable are the
highly stereoselective cyanate-to-isocyanate rearrangement,
the mild RCM cleavage of the chiral auxiliary, the chemo-
selective cyclopropanation of 30, and the high-yielding
6-exotrig radical cyclization of 31b.
(26) Their assignment of the absolute stereochemistry of (-)-aspidofrac-
tinine 3 is based on comparison of the OCD spectra of N-acetylaspidofracti-
nine (þ)-38 obtained from the degradation of minovincine (-)-2 and from
the acetylation of natural aspidofractinine 3 (see scheme below). However,
the assignment of the absolute stereochemistry of (-)-2 is based on a
similarity of structure with vindicafformine (-)-37. The absolute stereo-
chemistry of natural (-)-37 has been confirmed by total synthesis but not
that of natural (-)-2. It is important to note that both (þ)-2 and (-)-2 exist in
nature. For syntheses of (-)-37: (a) Kuehne, M. E.; Bandarage, U. K.;
Hammach, A.; Li, Y.-L.; Wang, T. J. Org. Chem. 1998, 63, 2172–2183. (b)
Kuehne, M. E.; Podhorez, D. E. J. Org. Chem. 1985, 50, 924–929. For the
isolation of (-)-2: (c) Douzoua, L.; Mansour, M.; Bebray, M. M.; Le Men-
Olivier, L.; Le Men, J. Phytochemistry 1974, 13, 1994–1995. (d) Meisel, H.;
Doepke, W. Tetrahedron Lett. 1970, 749–751. For the isolation of (þ)-2: (e)
Cava, M. P.; Tjoa, S. S.; Ahmed, Q. A.; Da Rocha, A. I. J. Org. Chem. 1968,
33, 1055–1059. (f) Liu, G.; Liu, X.; Feng, X. Planta Med. 1988, 54, 519–521.
Experimental Section
(þ)-Aspidofractinine (3). A suspension containing LiAlH4
(14 mg, 0.369 mmol) and 36 (6.0 mg, 0.020 mmol) in 2.5 mL
of tetrahydrofuran was stirred for 2 h at reflux. The reaction
mixture was cooled to 0 °C, then 2.0 mL of diethyl ether, 14 μL
of an aqueous solution of NaOH 15%, and 42 μL of water were
added. After 15 min of agitation at 0 °C, the mixture was
warmed to rt, and anhydrous magnesium sulfate was added.
After 15 min of agitation, the mixture was filtered and rinsed
with 50 mL of ethyl acetate. The filtrate was concentrated under
reduced pressure, and the crude product was purified by flash
chromatography on a silica gel column with acetone and
methanol (9:1) as eluant to give product 3 as a white amorphous
J. Org. Chem. Vol. 74, No. 16, 2009 6039

Gagnon and Spino
JOCArticle
SCHEME 9
solid (4.0 mg, 70%): 1H NMR (300 MHz, CDCl3) δ (ppm) 7.30
(d, 1H, J=7.6 Hz), 7.00 (t, 1H, J=7.6 Hz), 6.78 (t, 1H, J=7.6
Hz), 6.64 (d, 1H, J =7.6 Hz), 3.24 (q, 1H, J =8.8 Hz), 3.11 (s,
1H), 3.15-2.60 (m, 4H), 2.72 (ddd, 1H, J =13.7, 8.8, 3.3 Hz),
2.28-2.10 (m, 2H), 1.92-1.61 (m, 4H), 1.52 (br d, 1H, J=13.7
Hz), 1.45-1.39 (m, 1H), 1.34-1.18 (m, 5H); NMR spectrum
was identical to those previously published;3,7b,24 [R]20D = þ12
(c=0.27, CHCl3).
30.2 (t), 26.4 (d), 24.3 (t), 22.8 (q), 21.6 (q), 15.5 (q); IR (neat)
ν (cm-1) 3642-3241, 3069, 1450, 1367; LRMS (m/z, relative
intensity) 491 (Mþ, 20), 352 (100); HRMS calcd for
C30H37NO3S 491.2494, found 491.2489; [R]20D = -11.3 (c =
1.95, CHCl3); %de pure product >99% (HPLC).
O-Carbamate 15a. Allylic alcohol 14a (3.47 g, 7.60 mmol) was
dissolved in THF (75 mL), and the solution was cooled to 0 °C.
Trichloroacetylisocyanate (2.00 g, 10.6 mmol) was added drop-
wise. The reaction was stirred at 0 °C for 1 h. THF was then
evaporated from the reaction mixture, and methanol (35 mL)
was added to the residue. The mixture was cooled to 0 °C; an
aqueous saturated solution of K2CO3 (5 mL) was added por-
tionwise, and the mixture was then stirred for 30 min at 0 °C and
at rt for 1 h. The layers were separated, and the aqueous layer
was extracted twice with Et2O. The organic layers were com-
bined, dried over anhydrous magnesium sulfate, and concen-
trated under reduced pressure to give a colorless oil. The crude
product was purified by flash chromatography on a silica gel
column eluting with hexanes and ethyl acetate (7:3) to give
allylic carbamate as a colorless oil 15a (3.61 g, 96%): 1H NMR
(300 MHz, CDCl3) δ (ppm) 8.21 (dd, 1H, J=6.9, 2.5 Hz), 7.74
(d, 2H, J=7.7 Hz), 7.69-7.64 (m, 1H), 7.56-7.48 (m, 1H), 7.39
(t, 2H, J=7.7 Hz), 7.34-7.24 (m, 2H), 6.54 (dd, 1H, J=16.2, 1.1
Hz), 6.19 (dd, 1H, J=16.2, 5.8 Hz), 5.99 (ddt, 1H, J=16.9, 10.2,
6.0 Hz), 5.53 (d, 1H, J=5.8 Hz), 5.06 (dd, 1H, J=10.2, 1.4 Hz),
5.02 (dd, 1H, J=16.9, 1.4 Hz), 4.60 (br s, 2H), 3.87 (d, 2H, J =
6.0 Hz), 2.15 (m, 1H), 1.83-1.63 (m, 3H), 1.63-1.49 (m, 1H),
1.38-1.17 (m, 3H), 1.09-0.75 (m, 2H), 0.91 (d, 3H, J=7.1 Hz),
0.87 (d, 3H, J=6.6 Hz), 0.79 (d, 3H, J=6.6 Hz); 13C NMR (75
MHz, CDCl3) δ (ppm) 157.0 (s), 138.9 (s), 136.7 (s), 135.9 (s),
134.8 (d), 133.7 (d), 130.8 (d), 129.2 (d), 128.6 (s), 126.4 (d), 124.6
(d), 123.8 (d), 121.3 (d), 119.7 (d), 119.0 (s), 116.5 (t), 115.0 (d),
75.5 (d), 43.7 (d), 43.2 (d), 35.6 (t), 34.9 (t), 32.7 (d), 30.3 (t), 26.4
(d), 24.1 (t), 22.7 (q), 21.6 (q), 15.4 (q); IR (neat) ν (cm-1) 3550-
3100 (br), 1723, 1367; LRMS (m/z, relative intensity) 534 (Mþ,
15), 473 (95), 352 (45), 194 (100); HRMS calcd for C31H38N2O4S
534.2552, found 534.2542; [R]20D=-19.7 (c=1.57, CHCl3).
Trichloroethoxycarbamate 17a. Allylic O-carbamate 15a
(7.70 g, 14.4 mmol) was dissolved in DCM (500 mL), and the
solution was cooled to 0 °C. Then, Et3N (6.7 mL, 43.2 mmol)
followed by TFAA (2.2 mL, 15.8 mmol) were added over a
period of 30 min. The reaction mixture was stirred at 0 °C for 10
min, after which time 2,2,2-trichloroethanol (13.8 mL, 14.4
mmol) was added and the reaction was stirred for 3 h while
removing the ice bath and allowing the solution to warm to rt. A
saturated solution of aquous NaHCO3 was added, and the
phases were separated. The aqueous phase was extracted three
times with DCM, the organic layers were combined and dried
over anhydrous magnesium sulfate, and the solvent was evapo-
rated under reduced pressure. The crude product was purified by
Allylic Alcohol 14a. 2-Allyl-1-benzenesulfonyl-3-(2,2-dibro-
movinyl)indole 11a (4.80 g, 9.97 mmol) was dissolved in THF
(100 mL), and the solution was cooled to -95 °C. NaHMDS
(1 M in THF, 15.0 mL, 15.0 mmol) was added dropwise. The
reaction was stirred at -95 °C during 10 min. A solution of
n-BuLi in hexane (2.35 M, 8.48 mL, 19.9 mmol) was added
dropwise. The reaction was stirred at -95 °C during 10 min.
p-Menthane-3-carboxaldehyde 8 (2.01 mL, 12.0 mmol) was
dissolved in THF (40 mL) and added dropwise. The mixture
was stirred at -95 °C during 10 min and then quenched with
methanol (5.0 mL) at that temperature. After allowing the
mixture to warm to rt, saturated NH4Cl solution and the
mixture were warmed to rt. The layers were separated, and the
aqueous layer was extracted twice with EtOAc. The organic
layers were combined, washed once with brine, dried over
anhydrous magnesium sulfate, and concentrated under reduced
pressure to give yellow oil. The crude mixture of propargylic
alcohols was dissolved in THF (150 mL), and the solution was
cooled to 0 °C. Red-Al (65% w/w in toluene, 6.82 g, 21.9 mmol)
was added dropwise. The reaction was stirred at rt for 15 min.
The reaction mixture was quenched with brine. The layers were
separated, and the aqueous layer was extracted twice with
EtOAc. The organic layers were combined, washed once with
brine, dried over anhydrous magnesium sulfate, and concen-
trated under reduced pressure to give a colorless oil. The crude
product was purified by flash chromatography on a silica gel
column eluting with hexanes and ethyl acetate (9:1) to give
allylic alcohols 14a (3.47 g, 71%) and 14b (977 mg, 20%) both as
a colorless oil. Major isomer 14a: 1H NMR (300 MHz, CDCl3)
δ (ppm) 8.25 (dd, 1H, J=7.2, 2.4 Hz), 7.76 (d, 2H, J=7.6 Hz),
7.72 (dd, 1H, J=7.2, 2.4 Hz), 7.49 (t, 1H, J=7.6 Hz), 7.37 (t, 2H,
J=7.6 Hz), 7.33-7.27 (m, 2H), 6.63 (dd, 1H, J=16.0, 1.1 Hz),
6.32 (dd, 1H, J=16.0, 5.1 Hz), 6.01 (ddt, 1H, J=17.0, 9.6, 5.8
Hz), 5.06 (d, 1H, J =9.6 Hz), 5.02 (dd, 1H, J =17.0, 1.65 Hz),
4.66 (d, 1H, J=5.1 Hz), 3.91 (d, 2H, J=5.8 Hz), 2.20 (dqi, 1H,
J =6.6, 2.8 Hz), 1.85-1.66 (m, 2H), 1.72 (d, 2H, J =11.6 Hz),
1.47 (q, 1H, J=11.2 Hz), 1.43-1.22 (m, 2H), 1.13-0.78 (m, 2H),
0.97 (d, 3H, J=6.6 Hz), 0.88 (d, 3H, J=6.6 Hz), 0.82 (d, 3H, J =
7.1 Hz); 13C NMR (75 MHz, CDCl3) δ (ppm) 138.9 (s), 136.7 (s),
135.7 (d), 135.6 (s), 134.8 (d), 133.7 (d), 129.1 (d), 128.8 (s), 126.4
(d), 124.6 (d), 123.8 (d), 119.8 (d), 119.6 (d), 119.3 (s), 116.4 (t),
115.0 (d), 71.9 (d), 44.8 (d), 43.1 (d), 35.1 (t), 34.1 (t), 32.8 (d),
6040 J. Org. Chem. Vol. 74, No. 16, 2009

Gagnon and Spino
JOCArticle
flash chromatography on a silica gel column eluting with
hexanes, DCM, and ethyl acetate (17:2:1) to give allylic carba-
mate 17a as a colorless oil (9.4 g, 87%): H NMR (300 MHz,
(d), 133.4 (d), 129.4 (d), 128.4 (d), 128.4 (s), 127.7 (d), 124.8 (d),
123.9 (d), 1.3 (d), 115.4 (d), 51.8 (s), 50.3 (d), 42.3 (t), 40.5 (s),
38.9 (t), 32.2 (d), 30.1 (t), 28.9 (t); IR (neat) ν (cm-1) 3055, 1682,
1362; LRMS (m/z, relative intensity) 440 (Mþ, 5), 299 (75), 180
(100); HRMS calcd for C23H21N2O3SCl 440.0961, found
440.0970; [R]20D = -96.2 (c=1.25, CHCl3).
1
CDCl3) δ (ppm) 8.21 (d, 1H, J=7.6 Hz), 7.74 (d, 2H, J=7.6 Hz),
7.53 (t, 1H, J=7.6 Hz), 7.47 (d, 1H, J=7.6 Hz), 7.40 (t, 2H, J =
7.6 Hz), 7.30 (t, 1H, J =7.6 Hz), 7.22 (t, 1H, J =7.6 Hz), 5.98
(ddt, 1H, J =16.8, 9.9, 5.2 Hz), 5.65-5.53 (m, 2H), 5.65-5.53
(m, 2H), 5.40 (d, 1H, J = 7.1 Hz), 5.31 (dd, 1H, J=14.2, 9.4 Hz),
5.03 (d, 1H, J = 9.9 Hz), 4.99 (d, 1H, J = 16.8 Hz), 4.76 (d, 1H, J
=12.1 Hz), 4.59 (d, 1H, J=12.1 Hz), 4.00 (dd, 1H, J=16.5, 5.2
Hz), 3.90 (dd, 1H, J=16.5, 5.2 Hz), 1.98-1.84 (m, 1H), 1.78-
1.50 (m, 3H), 1.41-1.18 (m, 2H), 1.02-0.71 (m, 2H), 0.85
(d, 3H, J =6.6 Hz), 0.76 (d, 3H, J =7.1 Hz), 0.61 (d, 3H, J =
7.1 Hz); 13C NMR (75 MHz, CDCl3) δ (ppm) 153.7 (s), 138.9 (s),
137.9 (d), 136.8 (s), 136.1 (s), 135.2 (d), 133.7 (d), 129.2 (d), 127.9
(s), 126.4 (d), 126.4 (d), 124.5 (d), 123.4 (d), 120.3 (s), 119.7 (d),
116.6 (t), 115.4 (d), 95.6 (s), 74.4 (t), 49.1 (d), 47.0 (d), 44.3 (d),
42.8 (t), 35.1 (t), 32.4 (d), 30.0 (t), 28.0 (d), 24.0 (t), 22.8 (q), 21.5
(q), 15.5 (q); IR (neat) ν (cm-1) 3475-3208 (br), 1734, 1721;
LRMS (m/z, relative intensity) 664 (Mþ, 1), 474 (40), 377 (50),
230 (70), 194 (100); HRMS calcd for C33H39N2O4SCl3 664.1696,
found 664.1702; [R]20D = -40.2 (c = 1.10, CHCl3); dr 11:1
(HPLC).
Enimine 34. Compound 33 (110 mg, 0.413 mmol) was dis-
solved in THF (30 mL) and pyridine (5 mL). Phenylseleninic acid
(234 mg, 1.24 mmol) was added, and the mixture was heated to
reflux for 30 min. The mixture was cooled to rt, brine was added,
and the phases were separated. The aqueous layer was extracted
twice with ethyl acetate. The organic layers were combined, dried
over anhydrous magnesium sulfate, and concentrated under
reduced pressure to give a colorless oil. The crude product was
purified by flash chromatography on a silica gel column starting
with EtOAc and going to a mixture of EtOAc and MeOH (19:1)
to give the product 34 as a white solid (67 mg, 61%): 1H NMR
(300 MHz, C6D6) δ (ppm) 7.65 (d, 1H, J=7.4 Hz), 7.08 (t, 1H,
J=7.4 Hz), 6.95 (t, 1H, J =7.4 Hz), 6.79 (d, 1H, J=7.4 Hz), 6.52
(dd, 1H, J =10.0, 3.0 Hz), 5.40 (dt, 1H, J =10.0, 1.9 Hz), 4.27
(m, 1H), 3.35 (dd, 1H, J=5.5, 1.1 Hz), 2.48 (d, 1H, J=17.9 Hz),
2.34 (d, 1H, J=17.9 Hz), 2.24 (m, 1H), 1.49-1.40 (m, 1H), 1.32-
1.05 (m, 2H), 1.03-0.79 (m, 2H); the 1H NMR spectral data in
CDCl3 were identical to that reported in the literature;7d IR
(neat) ν (cm-1) 3042, 1695, 1452; LRMS (m/z, relative intensity)
264 (Mþ, 100), 180 (30); HRMS calcd for C17H16N2O 264.1263,
found 264.1259; [R]20D=þ39.2 (c=1.06, C6H6).
Pentacyclic Compound 31a. [Cu(OTf)]2 toluene (22 mg, 0.043
3
mmol) was dissolved in DCM (20 mL). The diazo compound 30
(94 mg, 0.200 mmol) diluted in 20 mL of DCM was added to the
reaction mixture over 4 h with a syringe pump. The reaction was
stirred at rt for 30 min. The reaction mixture was concentrated,
and the crude product was purified by flash chromatography on
a silica gel column eluting with hexanes and EtOAc (1:1) to give
the product 31a as a white solid (68 mg, 76%): mp 160-161 °C;
1H NMR (300 MHz, CDCl3) δ (ppm) 7.90 (d, 2H, J=7.5 Hz),
7.75 (d, 1H, J=7.5 Hz), 7.61 (t, 1H, J=7.5 Hz), 7.52 (d, 2H, J=
7.5 Hz), 7.26 (t, 2H, J =7.5 Hz), 7.06 (t, 1H, J =7.5 Hz), 6.22
(ddd, 1H, J =9.4, 6.5, 1.9 Hz), 6.08 (ddd, 1H, J =9.4, 5.5, 3.3
Hz), 4.70 (d, 1H, J=6.5 Hz), 3.60-3.44 (m, 4H), 2.92 (ddd, 1H,
J=13.7, 7.4, 6.0 Hz), 2.44 (ddd, 1H, J=19.2, 3.3, 2.2 Hz), 2.07
(m, 1H), 1.98 (sept, 1H, J =7.1 Hz), 0.91 (s, 1H); 13C NMR
(75 MHz, CDCl3) δ (ppm) 169.3 (s), 143.2 (s), 136.8 (s), 134.0
Acknowledgment. We thank the Natural Sciences and
Engineering Council of Canada, Merck-Frosst Canada
ꢀ
Ltd., and the Universite de Sherbrooke for their financial
support.
Supporting Information Available: Experimental proce-
dures for all new compounds not included in the Experimental
1
Section, H NMR spectra for all new compounds, and X-ray
tables and data for compound 31a. This material is available free
of charge via the Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 74, No. 16, 2009 6041