Total synthesis of (±)-herbindole B and (±)-cis-trikentrin B

Article abstract of DOI:10.1021/ol047498k

(Chemical Equation Presented) Herbindole B and cis-trikentrin B are naturally occurring indoles having the unusual and synthetically challenging pattern of carbon substitution at the 4-7 and 5-7 positions, respectively, with no substitution at the 1-3 positions. The total syntheses of these polyalkylated indoles have been achieved in 19 and 12 steps, respectively. The synthesis of herbindole B relies on two iterations of a quinine monoimine Diels-Alder reaction, while cis-trikentrin B uses a single cycloaddition of a suitable quinone monoimine. Indolization of the adducts provides suitably substituted benzopyrrole nuclei for elaboration to the natural products.

Full text of DOI:10.1021/ol047498k

ORGANIC
LETTERS
2005
Vol. 7, No. 7
1215-1218
Total Synthesis of (
)-cis-Trikentrin B
±)-Herbindole B and
(
±
Stephen K. Jackson, Scott C. Banfield, and Michael A. Kerr*
Department of Chemistry, UniVersity of Western Ontario, London,
Ontario, Canada N6A 5B7
makerr@uwo.ca
Received December 6, 2004
ABSTRACT
Herbindole B and cis-trikentrin B are naturally occurring indoles having the unusual and synthetically challenging pattern of carbon substitution
at the 4 7 and 5 7 positions, respectively, with no substitution at the 1 3 positions. The total syntheses of these polyalkylated indoles have
been achieved in 19 and 12 steps, respectively. The synthesis of herbindole B relies on two iterations of a quinine monoimine Diels Alder
−
−
−
−
reaction, while cis-trikentrin B uses a single cycloaddition of a suitable quinone monoimine. Indolization of the adducts provides suitably
substituted benzopyrrole nuclei for elaboration to the natural products.
The herbindoles are a series of structurally similar polyalky-
lated cyclopent[g]indoles which were isolated from the
orange Western Australian sponge Axinella sp. (Figure 1).1
co-workers.2 The trikentrins differ only in the nature of the
substitution at the 4- and 5-positions (indole numbering). The
herbindoles exhibit cytotoxicity against KB cells and are
active as a general fish antifeedant whereas the trikentrins
show growth inhibitory activity against the Gram-positive
bacteria Bacillus subtilis.
The herbindoles and the trikentrins bear extensive substi-
tution on the benzenoid portion of the indole ring system.
This well-known synthetic challenge coupled with the
stereochemical issues related to the relative stereochemistry
of the two benzylic stereogenic centers make these com-
pounds synthetic targets whose relatively small size belies
their complexity. Indeed, the only synthesis of any of the
herbindoles comes from the laboratory of Natsume, whose
complex synthetic scheme3 is evidence of the challenge posed
by such a visually unassuming target.
The trikentrins, on the other hand, have received more
attention from the synthetic community. Successful ap-
(1) Herb, R.; Carroll, A. R.; Yoshida, W. Y.; Scheuer, P. J.; Paul, V. J.
Tetrahedron 1990, 46, 3089.
Figure 1. Herbindole and trikentrin indoles.
(2) Capon, R. J.; Macleod, J. K.; Scammells, P. J. Tetrahedron 1986,
42, 6545.
(3) (a) Muratake, H.; Mikawa, A.; Seino, T.; Natsume, M. Chem. Pharm.
Bull. 1994, 42, 854. (b) Muratake, H.; Mikawa, A.; Seino, T.; Natsume,
M. Chem. Pharm. Bull. 1994, 42, 846.
They are related to the better known trikentrins, isolated from
the marine sponge Trikentrion flabelliforme by Capon and
10.1021/ol047498k CCC: $30.25
© 2005 American Chemical Society
Published on Web 03/08/2005

proaches have employed aryl radical cyclizations,4 an in-
tramolecular Heck coupling,5 a sequential heteroaromatic
azadiene Diels-Alder approach,6 indolization of pyrrole
derivatives,7 and an intramolecular Diels-Alder reaction of
allenic dienamides.8
In recent years, we have studied the Diels-Alder reactions
of quinoid imines9 and have developed a practical method
of converting the adducts to indoles which bear a high degree
of substitution.10 The process involves the Diels-Alder
reaction of a quinoid imine and a subsequent Pleininger
indole formation.11 We have been able to prepare a variety
of indoles with diverse substitution patterns. Herein we report
the application of a variant of these methods to the prepara-
tion of cis-trikentrin B and herbindole B.
Diels-Alder reaction (with appropriate reoxidation) of a
suitable dienophile 7 or 8 with butadiene and cyclopenta-
diene.
Our synythesis of cis-trikentrin B commenced with the
preparation of a suitable dienophile for the key Diels-Alder
reaction. Synthesis of the desired quinone imine 10 (Scheme
2) was straightforward from the p-aminophenol12 9 via
Scheme 2. Synthesis of cis-Trikentrin B
The retrosyntheses we had envisaged from the beginning
are shown in Scheme 1. For both targets, the required cis-
Scheme 1. Retrosynthesis of cis-Trikentrin B and Herbindole
B
tosylation and phenyliodo bis-acetate (PIBA)13 oxidation of
the resultant sulfonamide. Reaction of quinone imine 10 with
cyclopentadiene at 0 °C resulted in clean formation of the
desired cycloadduct. We were gratified to see that treatment
of the cycloadduct with acid not only catalyzed the aroma-
tization to the desired dihydronaphthalene but also facilitated
the in situ indolization to afford the desired 5-hydroxyindole
11 in excellent yield (90% over two steps). Protection of
the hydroxyl group as a triflate and dihydroxylation gave
the diol 12. It was found at this stage that the triflate
functionality was incompatible with elaboration of the
dimethylcyclopentane ring. This suggested that installation
of the 1-butenyl group should precede formation of the
dimethylcyclopentane ring.
It was felt that Stille coupling with the organotin reagent
15 would be the most efficient method for installation of
the 1-butenyl group since it required no subsequent functional
group manipulation. 1-Butenyltributylstannane 15 was formed
from 1-butyne via hydroalumination and iodonolysis,14
followed by metal-halogen exchange and trans-metalation,
which afforded the desired trans-organotin reagent 15 (trans/
cis ) 12: 1).15 Stille coupling proceeded smoothly with
nearly complete control of the olefin geometry, generating
dimethylcyclopentane unit would be revealed via oxidative
cleavage of the benzonorbornene system in 1, 5, or 6. For
cis-trikentrin B, the key intermediate 1 results from indoliza-
tion of aldehyde 2, which in turn is the product of a Diels-
Alder reaction of 3 with cyclopentadiene. In contrast to the
strategy for cis-trikentrin B, where the aldehyde necessary
for indolization is built into the dienophile, for herbindole
B the aldehyde necessary for indolization (see compound 4)
would arise by oxidative cleavage of a cyclohexenyl moiety
in 5 or 6. Both 5 and 6 are products of two iterations of a
(4) Macleod, J. K.; Monaham, L. C. Tetrahedron Lett. 1988, 29, 391.
(5) Blechert, S.; Wiedenau, P.; Monse, B. Tetrahedron 1995, 51, 1167.
(6) Boger, D. L.; Zhang, M. J. Am. Chem. Soc. 1991, 113, 4230.
(7) (a) Natsume, M.; Muratake, H. Tetrahedron Lett. 1989, 30, 5771.
(b) Natsume, M.; Muratake, H.; Seino, T. Tetrahedron Lett. 1993, 34, 4815.
(8) Kanematsu, K.; Lee, M.; Ikeda, I.; Kawabe, T.; Mori, S. J. Org. Chem.
1996, 61, 3406.
(9) Banfield, S. C.; Kerr, M. A. Can. J. Chem. 2004, 82, 131. (b) Kerr,
M. A. Synlett 1995, 1165.
(10) (a) Banfield, S. C.; England, D. B.; Kerr, M. A. Org. Lett. 2001, 3,
3325. (b) Zawada, P. V.; Banfield, S. C.; Kerr, M. A. Synlett 2003, 971.
(11) (a) Pleininger, H.; Suhr, K. Chem. Ber. 1956, 89, 270. (b) Pleininger,
H.; Volkl, A. Chem. Ber. 1976, 109, 2121. For applications of the Pleininger
indolization, see: (c) Kraus, G. A.; Yue, S.; Sy, J. J. Org. Chem. 1985, 50,
284. (d) Maehr, H.; Smallheer, J. J. Am. Chem. Soc. 1985, 107, 2943. (e)
Tichenor, M. S.; Kastrinsky, D. B.; Boger D. L. J. Am. Chem. Soc. 2004,
26, 8396.
(12) For an excellent preparative procedure of 9 from 3-methyl-4-
nitrophenol, see: Todd, M. H.; Oliver, S. F.; Abell, C. Org. Lett. 1999, 1,
1149.
(13) For the preparation of PIBA, see: Smith, J. G. Fieser and Fieser’s
Reagents for Organic Synthesis; Wiley: Toronto, 1990; Vol. 1, p 508.
(14) Alexakis, A.; Duffault, J. M. Tetrahedron 1988, 29, 6243.
(15) Analyses of the integration values for the well-separated vinyl signals
1
of the H NMR were used to determine the trans/cis ratio.
1216
Org. Lett., Vol. 7, No. 7, 2005

13 predominantly as the trans-isomer in 72% yield (trans/
cis > 20:1).15 In an attempt to minimize epimerization of
the dialdehyde, the oxidative cleavage of 13 with NaIO4 was
immediately followed by the in situ reduction with NaBH4
in a one-pot procedure. Caution: this is an extremely
Vigorous reaction and must be conducted with slow por-
tionwise addition of NaBH4. Treatment with MsCl produced
the bis-mesylate 14 in excellent yield (97% over three steps).
Reductive deoxygenation via the Fujimoto16 procedure
followed by detosylation with TBAF17 completed the dias-
tereoselective synthesis of (()-cis-trikentrin B.
Scheme 4. Revised Approach to Herbindole B
Our approach to herbindole B began with the cycloaddition
of dienophile 8 with 1,3-butadiene (Scheme 3). In the event,
Scheme 3. Initial Approach to Herbindole B
cycloaddition with cyclopentadiene to give adduct 23. At
this stage, the usual conditions for aromatization via eno-
lization of the ketone and sulfonylimino groups (DBU) failed
to produce the desired product, giving instead 25 in excellent
yield.
As an alternative to base-promoted aromatization of the
adduct, attention was turned to nucleophilic addition of
methyllithium to the carbonyl group in the hopes of preparing
24. Acid treatment of 24 would produce 5, an ideal
intermediate for conversion to the natural product. To our
surprise, rather that behaving nucleophilically, the methyl-
lithium behaved as a base and promoted the aromatization
which had failed with DBU. In this way, the phenolic
compound 6 was prepared in superb yield.
The conversion of 6 to the target compound, herbindole
B, is shown in Scheme 5. Triflation of the phenolic hydroxyl
group and chemoselective dihydroxylation (presumably due
to the strain energy release in the benzonorbornenyl system)
produced the diol 26. Acetonide formation generated 27 in
excellent yield, and dihydroxylation of the remaining olefin
afforded the intermediate diol. Oxidative cleavage of the diol
with NaIO4 gave a dialdehyde which upon acid treatment
underwent indolization to 28. Reduction and mesylation gave
29 which was a suitable substrate for reduction under the
conditions of Fujimoto. The reduction product 30 underwent
smooth Stille coupling with tetramethyltin to install the
required 5-methyl substituent. Conversion of 31 to the bis-
mesylate 33 was accomplished by acetonide hydrolysis,
periodate cleavage of the diol, reduction to the diol 32 and
treatment with mesyl chloride. Fujimoto reduction as before
and detosylation with TBAF yielded herbindole B which was
identical in all respects, except for optical rotation, with the
natural product.
a 1:1 separable mixture of adducts 16 and the desired 17
were obtained in an overall yield of 80%. The production
of 16 presumably arises via a Lewis acid-catalyzed migration
of the methoxy group in the initial adduct. Dihydroxylation
of 17 proceeded without incident to produce 18, which was
reoxidized with PIBA to the dienimine 19. Despite our best
efforts and under a variety of reaction conditions, cycload-
dition of 19 with cyclopentadiene proceeded to give 20 in a
best yield of 17% when BCl3 was used as a catalyst. Other
dienophiles similar to 19 were prepared and failed to undergo
cycloaddition to give the desired product.
The desire to use a dienophile such as 8 in our synthetic
sequence was born in the fact that the methyl group in 8
would end up being the required 5-methyl group in the
natural product. The failure of this sequence to produce
satisfactory results prompted us to reevaluate our choice of
dienophile. The quinone monoimine 7 when carried through
a similar sequence of reactions would result in a 5-hydroxy
indole (or derivative thereof) which would allow for instal-
lation of the required methyl substituent via cross-coupling.
Moreover, 7 was expected to exhibit greater dienophilicity
due to the presence of the carbonyl moiety.
We were delighted when our expectations met with reality
and the cycloaddition of 7 with butadiene proceeded (after
aromatization with DBU) to produce dihydronaphthalene 21
in quantitative yield (Scheme 4). Reoxidation with NaIO4
supported on silica18 gave 22, which underwent a second
In conclusion, new diastereoselective total syntheses of
(()-cis-trikentrin B and (()-herbindole B have been achieved.
(16) Fujimoto, Y.; Tatsuno, T. Tetrahedron Lett. 1976, 37, 3325.
(17) Yasuhara, A.; Sakamoto, T. Tetrahedron Lett. 1998, 39, 595.
(18) Zhong, Y. L.; Shing, T. K. M. J. Org. Chem. 1997, 62, 2622.
Org. Lett., Vol. 7, No. 7, 2005
1217

diene was the key step allowing for preparation of the natural
product in 12 steps and 14% overall yield. For herbindole
B, the key sequential Diels-Alder/oxidation/Diels-Alder/
aromatization provided an adduct which contained all the
necessary components for the tricyclic herbindole B skeleton.
In this manner, the natural product was obtained in 19 steps
and in 4% overall yield. In both cases a consequence of the
strategy was the presence of a hydroxyl group at the indole
5-position, allowing for installation by cross-coupling of the
required trans-butenyl (trikentrin) and methyl (herbindole)
substituents. The above strategies effectively accomplished
the synthesis of indoles bearing complex substitution on the
benzenoid positions of the benzopyrrole ring system while
leaving the more reactive 1-3 positions unsubstituted. A
more detailed account of the syntheses and cross-coupling
reactions of 5-triflyloxyindoles will be presented in due
course.
Scheme 5. Synthesis of Herbindole B
Acknowledgment. We thank the Natural Sciences and
Engineering Research Council of Canada, Medmira Labo-
ratories, Inc., the Ontario Ministry of Science and Technol-
ogy, and GlaxoSmithKline for funding. Acknowledgment is
made to the donors of the American Chemical Society
Petroleum Research Fund for partial support of this research.
We are grateful to Mr. Doug Hairsine for performing MS
analyses.
Supporting Information Available: Full experimental
procedures and spectroscopic data for compounds 10-14,
(()-cis-trikentrin B, 6, 7, 21, 22, 26-31, 33, and (()-
herbindole B. This material is available free of charge via
the Internet at http://pubs.acs.org.
The two complementary strategies successfully address the
challenge of substitution on the benzenoid porition of the
indole ring system. For (()-cis-trikentrin B, a Diels-Alder
reaction of a suitable quinone monoimine with cyclopenta-
OL047498K
1218
Org. Lett., Vol. 7, No. 7, 2005