Total synthesis of (±)-decursivine

Article abstract of DOI:10.1002/ejoc.200600922

The first preparation of the antimalarial natural product decursivine is described. A Diels-Alder/Plieninger indolization protocol allows convenient preparation of the indole 15 which, in turn is a suitable substrate for a boron-enolate aldol reaction with piperonal (16). The resulting adduct 14 is transformed efficiently to the natural product. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200600922

SHORT COMMUNICATION
DOI: 10.1002/ejoc.200600922
Total Synthesis of (؎)-Decursivine
Andrew B. Leduc[a] and Michael A. Kerr*[a]
Keywords: Decursivine / Diels–Alder / Indole / Total synthesis / Boron aldol
The first preparation of the antimalarial natural product de-
cursivine is described. A Diels–Alder/Plieninger indolization
protocol allows convenient preparation of the indole 15
which, in turn is a suitable substrate for a boron–enolate aldol
reaction with piperonal (16). The resulting adduct 14 is trans-
formed efficiently to the natural product.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
Introduction
The natural product decursivine (1) was isolated in 2002
from Rhaphidophora decursiva, a plant found in the Cuc
Phuong National Park in Vietnam after fractions contain-
ing it displayed activity against the malaria-causing parasite
Plasmodium falciparum.[1] Upon isolation of decursivine,
the spectroscopic data showed a remarkable resemblance to
the natural product serotobenine (2) which was first isolated
in 1985[2] and later in 1997[3] under the name moschamin-
dole. Serotobenine, unlike decursivine, exhibits no activity
against Plasmodium falciparum.[2] The only structural dif-
ference between decursivine and serotobenine results from
the fact that the methyl group in the latter is present in a
higher oxidation state as a methylenedioxy unit in the for-
mer. This minor difference would thus indicate the methyl-
enedioxy unit is vital to the biological activity of decursiv-
ine. Both decursivine and serotobenine also likely share a
common biosynthesis involving a cinnamide composed of
serotonin and an appropriately substituted cinnamic acid.
This is implied by the fact that moschaminindolol (3) and
moschamine (4) were co-isolated with serotobenine (mosch-
amindole). If the cinnamide (such as 4) is derived from
tryptamine rather than serotonin, the biogenetic outcome
appears to be cyclization onto the indole 2-position re-
sulting in compounds such as balasubramide (5) (isolated
from Clausena indica), see Figure 1.[4]
Figure 1. Decursivine and related alkaloids.
To date, there are no synthetic activities published
towards decursivine or related compounds. While small by
most synthetic standards, decursivine poses significant syn-
thetic challenges, namely the functionalization of the 3-, 4-
, and 5-positions of the indole, the sensitivity of the elec-
tron-rich indole to oxidation, and the stereogenic centres
on the dihydrobenzofuran. Herein, we report the first syn-
thesis of decursivine.
Our initial synthetic plan is illustrated in Scheme 1. Our
final transformation was envisaged to be a simple lactamiz-
ation which would require a one-carbon homologation of
intermediate 6. The addition of a methylenedioxyphenyl-
derived organometallic reagent to the aldehyde 7 followed
by etherification would secure 6. A Plieninger indolization[5]
involving the dihydronaphthalene 8 would in turn furnish
[a] Department of Chemistry, The University of Western Ontario,
London, Ontario, Canada N6A 5B7
Fax: +1-519-661-3022
E-mail: makerr@uwo.ca
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2007, 237–240
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
237

A. B. Leduc, M. A. Kerr
SHORT COMMUNICATION
7.[6] The preparation of 8 would be accomplished by our benzylic alcohol, but the dihydrobenzofuran 6 directly in
recently described Diels–Alder reaction of a quinone mono- varying yields. Conditions included the aryllithium and the
imine with a suitable 1,3-butadiene.[7]
chloromagnesio derivative. The best yields involved the li-
thio species with the addition of a stoichiometric quantity
of ceric chloride.[8] The low yield was felt to be related to
the undoubtedly high enolizability of 7. The yield of 6
shown in Scheme 2 represents reactions done on a small
scale (100 mg). Attempts to scale this even to 250 mg re-
sulted in a precipitous drop in yield to about 25%. With
the inability to advance sufficient synthetic material, this
route was abandoned, a frustrating development consider-
ing the rapid formation of the required dihydrobenzofuran
unit. From this initial study, however, our spirits were
buoyed by the apparent ease of formation of the dihydrobe-
nzofuran ring from the putative dihydroxy compound.
A second generation retrosynthesis is illustrated in
Scheme 3. Again, the final step of the synthesis would be
an amide bond formation. We would also require the instal-
Scheme 1. Initial retrosynthesis of decursivine.
Our synthetic efforts commenced with the Diels–Alder
cycloaddition of the quinone monoimine 9 with the diene
10. The cycloaddition reaction proceeded to produce 11 in
an acceptable yield of 55% under hyperbaric conditions
(13 kbar). Acetylation of the phenolic hydroxy group gave
8, which was a suitable substrate for Plieninger indolization
through dihydroxylation with OsO4 and periodate cleavage.
The result was the formation of an unusually stable hemia-
minal 12 which, in contrast to our normal indolization pro-
tocol, did not undergo aromatization under acidic condi-
tions. Instead, treatment of 12 with methanesulfonyl chlo-
ride and triethylamine gave the requisite indole 7.
We were pleasantly surprised to observe that treatment of
7 with a variety of metalloaryl species gave not the expected Scheme 3. Revised retrosynthesis of decursivine.
Scheme 2. Synthesis of an advanced benzofuran intermediate. a) CH2Cl2, 13 kbar, followed by DBU (55%); b) AcCl, Et3N, THF (64%);
c) OsO4, NMO, THF, H2O; d) NaIO4, THF, H2O; e) MsCl, Et3N, THF (80%, 3 steps); f) 1-bromo-3,4-(methylenedioxy)benzene (3 equiv.),
tBuLi (6 equiv.), CeCl3 (1 equiv.), THF (40%). AcCl = acetyl chloride, MsCl = methanesulfonyl chloride.
238
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 237–240

Total Synthesis of (Ϯ)-Decursivine
SHORT COMMUNICATION
lation of an aminoethyl appendage at the indole 3-position the enolizability of the resultant molecule (or a later stage
which would lead in a disconnective fashion to 13. A facile intermediate) to ensure the required trans stereochemistry
etherification to the dihydrobenzofuran leads to 14 which of the dihydrofuran. At this stage, simple deprotection of
is the aldol product of ester 15 and piperonal 16.
the phenol in 14 met with frustration as treatment with any
Our revised synthetic strategy, illustrated in Scheme 4, acidic or basic reagent resulted in a retroaldol reaction
began with a much simpler and more scalable Diels–Alder pathway, often in near quantitative yields. Reductive re-
reaction, namely that of 9 with butadiene. Aromatization of moval of the pivaloate with LiAlH4 or LiBH4 also led to
the initially formed adduct with DBU and protection of the retroaldol products. Ultimately, it was determined that
phenolic moiety as the pivaloyl ester gave 17 in an overall treatment with diisobutylaluminum hydride in large excess
yield of 89%. Subjection of 17 to oxidative olefin cleavage resulted in removal of the pivaloate, reduction of the ester
and indolization yielded 18 in 84% overall yield. Oxi- (thereby avoiding retroaldol issues) and closure of the dihy-
dation[9] and treatment of the resulting carboxylic acid with drobenzofuran ring (on workup) to yield 13 in 70% yield.
TMSCHN2 yielded methyl ester 15 in 90% yield over two We were pleased to see that the propensity for formation of
steps. This ester was a willing participant in a boron enolate the dihydrofuran ring noted earlier was also the case in this
aldol reaction with piperonal 16 to produce 14 (76%).[10] synthetic sequence. With the dihydrobenzofuran intact (and
We were unconcerned at this time with the relative stereo- retroaldols now impossible) the ester functionality was re-
chemistry of the aldol reaction since we would be relying on stored in a two step oxidation using the Dess–Martin
Scheme 4. Final synthetic route to decursivine a) CH2Cl2, sealed tube followed by DBU (95% overall); b) PivCl, THF, Et3N (94%); c)
OsO4, NMO, THF, H2O; d) NaIO4, THF, H2O; e) H2SO4, THF (84%, 3 steps); f) NaClO2, NaH2PO4, tBuOH, H2O, 2-methyl-2-propene;
g) TMSCHN2, MeOH, benzene (90%, 2 steps); h) (Chx)2BI, Et3N, piperonal, CCl4, CH2Cl2 (76%); i) DIBAL (12 equiv.), THF (70%);
j) DMP, CH2Cl2 (83%); k) oxone, DMF; l) TMSCHN2, MeOH, benzene (80%, 2 steps); m) Mg0, NH4Cl, MeOH, THF (83%); n) 2-
(dimethylamino)-1-nitroethylene, TFA, CH2Cl2 (67%); o) TsCl, Et3N, DMF (84%); p) NaBH4, SiO2, iPrOH, CHCl3 (78%); q) Zn0, HCl,
THF; r) pyridine, microwave; s) Mg0, NH4Cl, MeOH, THF (43%, 3 steps). PivCl = trimethlyacetyl chloride, DMP = Dess–Martin
periodinane, TFA = trifluoroacetic acid, TsCl = toluenesulfonyl chloride, (Chx)2BI = dicyclohexyl(iodo)borane.
Eur. J. Org. Chem. 2007, 237–240
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
239

A. B. Leduc, M. A. Kerr
SHORT COMMUNICATION
periodinane[11] (83%) followed by oxone in DMF.[12] The
resulting acid was converted into the methyl ester 19 with
TMSCHN2 (80% yield over two steps).
Acknowledgments
We thank the Natural Sciences and Engineering Research Council
In order to functionalize the indole 3-position with the (NSERC) of Canada and Boehringer Ingelheim Canada for fund-
ing. We are grateful to Mr. Doug Hairsine for performing MS
analyses. A.B. L. is the recipient of an NSERC CGSM postgradu-
ate scholarship.
necessary aminoethyl side chain, the nucleophilicity of that
position was increased by tosyl removal (83%) and treat-
ment with 2-(dimethylamino)-1-nitroethylene to yield the
nitroolefin 20 in 67% yield.[13] Retosylation followed by re-
duction of the styrenyl double bond[14] yielded 21 in % over-
all yield. The retosylation was found to be necessary in or-
der to avoid decomposition in the ensuing transformations.
Reduction of the nitro group followed by heating the re-
sulting amino ester in pyridine under microwave irradiation
produced the required lactam 22. The natural product was
then secured by removal of the tosyl group in 43% yield
over three steps. The physical data for the synthetic material
was identical in all respects with the published data except
for optical rotation. We were pleased to find that the rela-
[1] H. Zhang, S. Qiu, P. Tamez, G. T. Tan, Z. Aydogmus, N.
Van Hung, N. M. Cuong, C. Angerhofer, D. D. Soejarto, J. M.
Pezzuto, H. H. S. Fong, Pharmaceutical Biology 2002, 40, 221.
[2] H. Sato, H. Kawagishi, T. Nishimura, S. Yoneyama, Y. Yoshim-
oto, S. Sakamura, A. Furusaki, S. Katsuragi, T. Matsumoto,
Agric. Biol. Chem. 1985, 49, 2969.
[3] S. D. Sarker, T. Savchenko, P. Whiting, S. Pensri, L. N. Dinan,
Nat. Prod. Lett. 1997, 9, 189.
[4] B. Riemer, O. Hofer, G. Greger, Phytochemistry 1997, 45, 337.
[5] a) H. Plieninger, A. Voekl, Chem. Ber. 1976, 109, 2121; b) H.
Plieninger, K. Suhr, G. Werst, B. Kiefer, Chem. Ber. 1956, 89,
270.
tive stereochemistry of the vicinal stereogenic centres were [6] For a recent application of this chemistry by our group to natu-
ral products synthesis, as well as leading references, see: D. B.
identical by comparison of coupling constants with the
published data.
England, J. Magolan, M. A. Kerr, Org. Lett. 2006, 8, 2209.
[7] D. B. England, M. A. Kerr, J. Org. Chem. 2005, 70, 6519.
In summary we have succeeded in preparing for the first
[8] L. A. Paquette, J. P. Gilday, G. D. Maynard, J. Org. Chem.
time the natural product decursivine in racemic form. The
synthetic sequence involves 18 synthetic operations from 9
and produced the natural product in over 3% overall yield.
Efforts are under way to explore the use of an asymmetric
aldol for the formation of decursivine in optically pure form
and to prepare other members of this fascinating class of
natural products.
1989, 54, 5044.
[9] B. S. Bal, W. E. Childers, H. W. Pinnick, Tetrahedron 1981, 37,
2091.
[10] K. Ganesan, H. C. Brown, J. Org. Chem. 1994, 59, 2336.
[11] D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155.
[12] B. R. Travis, M. Sivakumar, G. O. Hollist, B. Borhan, Org.
Lett. 2003, 5, 1031.
[13] T. Severin, H. J. Böhme, Chem. Ber. 1968, 101, 2925.
[14] A. K. Sinhababu, R. T. Borchardt, Tetrahedron Lett. 1983, 24,
227.
Supporting Information (see also the footnote on the first page of
this article): Full experimental procedures and spectroscopic data
for all new compounds.
Received: October 20, 2006
Published Online: November 27, 2006
240
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 237–240