Total synthesis of (+)-haplophytine

Full text of DOI:10.1002/anie.200904588

Communications
DOI: 10.1002/anie.200904588
Natural Product Synthesis
Total Synthesis of (+)-Haplophytine**
K. C. Nicolaou,* Stephen M. Dalby, Shuoliang Li, Takahiro Suzuki, and David Y.-K. Chen*
Despite the many impressive accomplishments in the field of
total synthesis in recent years,[1] a number of natural products
have proven stubbornly resistant to its advances.[2] Among
them is haplophytine (1, Scheme 1a), which has only very
recently succumbed to synthesis following the elegant work of
Fukuyama, Tokuyama and co-workers.[3] Haplophytine was
first isolated by Snyder and co-workers in 1952, and identified
as the principle bioactive component of the wild flower
Haplophyton cimicidum,[4] valued for centuries by the Aztecs
and subsequent settlers of Central America for its insecticidal
properties. A heterodimeric indole alkaloid, haplophytine
features a particularly complex polycyclic array of ten rings,
six stereocenters (five of which are quaternary) and a highly
ꢀ
congested carbon carbon bond adjoining the two distinct
halves of the molecule. The tetracyclic left-hand domain
features a unique bridged ketone structure, while the right-
hand domain consists of the naturally occurring aspidosperma
alkaloid, aspidophytine (2, Scheme 1b).[4,5]
A complete
appreciation of haplophytineꢀs molecular structure was only
reached some 21 years subsequent to its isolation, following
extensive chemical degradation, spectroscopic, and X-ray
crystallographic studies from the groups of Cava, Yates, and
Zacharias,[6] which included identification of the dihydrobro-
mide derivative 3 (Scheme 1a). As depicted, this compound is
formed through a unique acid-mediated skeletal rearrange-
ment of the left-hand domain involving the 1,2-shift of an
ꢀ
aminal C N bond. Under basic conditions, however, this
process can be reversed such as to return haplophytine
through a complementary semi-pinacol type mechanism. As
[*] Prof. Dr. K. C. Nicolaou, Dr. S. M. Dalby, Dr. S. Li, Dr. T. Suzuki,
Dr. D. Y.-K. Chen
Chemical Synthesis Laboratory @ Biopolis, Institute of Chemical
and Engineering Sciences (ICES), Agency for Science, Technology
and Research (A*STAR)
Fax: (+65)687-45870
E-mail: kcn@scripps.edu
Scheme 1. Structures of haplophytine (1, a), haplophytine dihydrobro-
mide (3, a), aspidophytine (2, b), synthetic left-hand fragment 4 (b)
and our retrosynthetic analysis of haplophytine (1, c). Ac=acetyl,
Bn=benzyl, Cbz=benzyloxycarbonyl, TBS=tert-butyldimethylsilyl,
TMSE=trimethylsilylethyl.
david_chen@ices.a-star.edu.sg
Prof. Dr. K. C. Nicolaou
Department of Chemistry and The Skaggs Institute for Chemical
Biology, The Scripps Research Institute, 10550 North Torrey Pines
Road, La Jolla, CA 92037 (USA)
and
Department of Chemistry and Biochemistry, University of California,
San Diego, 9500 Gilman Drive, La Jolla, CA 92093 (USA)
part of a program directed toward the total synthesis of
haplophytine, we have previously demonstrated the suitabil-
ity of this rearrangement as a means to construct the left-hand
domain fragment 4 (Scheme 1b).[7] A similar approach has
been reported independently by Fukuyama and co-workers.[8]
Our approach to aspidophytine,[5f] meanwhile, was specifi-
cally designed to complement that for 4, in order to facilitate
the total synthesis of haplophytine (1). Herein we wish to
report the culmination of this work with our own asymmetric
total synthesis of this historic synthetic target.
[**] We thank Prof. E. J. Corey for kindly providing us with an authentic
sample of natural haplophytine. We also thank Doris Tan (ICES) for
assistance with high resolution mass spectroscopy (HRMS), and
Dr. Tommy Wang Chern Hoe [Singapore Immunology Network
(SIgN-A*STAR)], Dr. Aitipamula Srinivasulu (ICES), and Chia Sze
Chen (ICES) for X-ray crystallographic analysis. Financial support
for this work was provided by A*STAR, Singapore.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200904588.
7616
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7616 –7620

Angewandte
Chemie
As shown in Scheme 1c, the right-hand aspidophytine
domain was retrosynthetically excised to reveal indole 5 and
vinyl iodide 6, which would be annulated[5f] sequentially
through Suzuki–Miyaura coupling,[9] a reductive Vilsmeier–
Haack reaction,[10] and radical cyclization.[11] The left-hand
domain of indole 5 would arise through the oxidative skeletal
rearrangement of bis-enamine 7,[7] which itself would derive
from the oxidative coupling of enantiopure tetrahydro-b-
carboline 8 and diphenol 9.[7]
The preparation of 8 commenced with the conversion of
commercially available 7-benzyloxyindole (10) to nitroalkene
11 (Scheme 2, 50%).[12] Reduction with LiAlH4 to the
Scheme 2. Construction of tetrahydro-b-carboline 8. Reagents and
conditions: a) 1-dimethylamino-2-nitroethylene (1.0 equiv), TFA
(2.0 equiv), CH2Cl2, 238C, 0.5 h, 50%; b) LiAlH4 (5.0 equiv), THF,
08C!reflux, 3 h; c) succinic anhydride (1.04 equiv), CH2Cl2, 238C, 2 h;
d) amberlyst-15 (20% wt/wt), MeOH, reflux, 4 h, 92% over three
steps; e) Pd/C (10% wt/wt, 0.05 equiv), H2 (1 atm), MeOH/THF (1:1),
238C, 8 h; f) AcCl (1.4 equiv), Et3N (2.0 equiv), THF, ꢀ78!238C,
2.5 h, 79% over two steps; g) POCl3 (8.0 equiv), DMPU, 238C, 2 h;
h) 15 (0.03 equiv), HCO2H/Et3N (5:2), DMF, 238C, 0.5 h; i) CbzCl
(2.8 equiv), Et3N (3.4 equiv), CH2Cl2, ꢀ78!238C, 9 h, 35% over three
steps. DMF=N,N’-dimethylformamide DMPU=1,3-dimethyl-3,4,5,6-
tetrahydro-2(1H)-pyrimidinone, TFA=trifluoroacetic acid, THF=tetra-
hydrofuran.
Scheme 3. Construction of indole fragment 5 and ORTEP drawing of
compound 19. Reagents and conditions: a) 8 (2.0 equiv), 9 (1.0 equiv),
PIFA (1.1 equiv), MeCN, ꢀ308C, 36 h, then 238C, 1 h, 23% [based on
25% conversion of 8 (the more valuable component), or 11.5% based
on 9 (the less valuable component)]; b) Cs2CO3 (3.0 equiv), MeI
(5.0 equiv), DMF, 238C, 2 h, 76%; c) K2CO3 (2.0 equiv), MeOH, 08C,
20 min; d) Cs2CO3 (3.0 equiv), BnBr (1.2 equiv), DMF, 238C, 1 h, 70%
over two steps; e) Cs2CO3 (5.0 equiv), MeI (30 equiv), DMF, 238C,
12 h, 65%; f) LiOH (2.0m aq.)/EtOH (1:1), 238C, 20 min; g) (COCl)2
(5.0 equiv), DMF (cat.), PhH, 08C, 0.5 h; h) iPr2NEt (10 equiv), PhH,
0!238C, 1 h, 60% over three steps; i) mCPBA (1.5 equiv), NaHCO3
(5.0 equiv), CH2Cl2, ꢀ58C, 12 h, 78%; j) DDQ (5.0 equiv), PhH, 758C,
12 h, 63%. PIFA=phenyliodine-bis-trifluoroacetate, PhH=benzene,
mCPBA=meta-chloroperoxybenzoic acid, DDQ=2,3-dichloro-5,6-
dicyano-1,4-benzoquinone.
corresponding tryptamine followed by acylation with succinic
anhydride and methylation provided methyl ester 12 (92%
over three steps). To facilitate the forthcoming oxidative
coupling with diphenol 9, the benzyl ether of 12 was replaced
with an acetate group to provide 13 (see below). Ring closure
under modified Bischler–Napieralski conditions (POCl3,
DMPU)[13] then afforded the rather labile imine 14, which
was immediately subjected to asymmetric Noyori reduction[14]
employing the ruthenium catalyst 15, and, subsequently, Cbz
protection to deliver the targeted fragment 8 in 35% over
three steps, and in > 95% ee.[15]
With tetrahydro-b-carboline 8 secured, its coupling with
diphenol 9 and formation of perhaps the singularly most
difficult bond bridging the two halves of the target molecule
Angew. Chem. Int. Ed. 2009, 48, 7616 –7620
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7617

Communications
was at hand (Scheme 3). Employing our previously described
With the left-hand substructure of haplophytine secured,
coupling of 5 to the aspidophytine domain precursor vinyl
iodide 6 by palladium-mediated cross-coupling was now
required. Extensive optimization was necessary to identify
appropriate means for this coupling, which was complicated
by the lability of a range of indole C-2 derivatives (5, for
example, R = SnMe3/B(OH)2/I, Scheme 4) and accompanying
lithiation/reaction of the left-hand domain lactam in their
preparation. Ultimately, it was found that lithiation of indole
5 with LiTMP in the presence of B-methoxy-pinacolatobor-
ane in THF at ꢀ1008C for 15 min provided the relatively
stable pinacol borane 22 (Scheme 4), which was submitted
without purification to Suzuki–Miyaura coupling[9] with vinyl
iodide 6. Accordingly, treatment of 22 with [Pd(dppf)Cl2],
Ph3As, and TlOEt in anhydrous DMSO effected the desired
coupling together with fortuitous cleavage of the methyl
carbamate, providing adduct 23 in 67% yield. In particular,
the use of TlOEt in combination with DMSO proved critical
to the success of this reaction, enabling the rate of coupling to
compete effectively with the otherwise rather facile process of
proto-deborylation.
N-Methylation of indole 23 then provided 24, a compound
in line with our previous synthesis of aspidophytine.[5f]
Accordingly, treatment of lactam 24 with Tf2O smoothly
provided the iminium salt 25, which was rapidly reduced with
NaBH4 at low temperature (ꢀ788C, 2 min) in order to
circumvent accompanying ketone reduction. This delivered
piperidine 26 as essentially a single diastereomer in good yield
(76%). Selective desilylation of 26 (HF·py) followed by
conversion of the ensuing primary alcohol (27) to xanthate
ester 28 (NaH, CS2 and MeI; 70% yield over two steps) set
method,[7] treatment of a mixture of 8 and 9 with PIFA
(1.1 equiv) in MeCN at ꢀ308C for 36 h effected the desired
coupling and provided the propellane-like hexacycle 16 in
23% yield, based on 25% conversion of 8 [the more valuable
component, or 11.5% based on 9 (the less valuable compo-
nent)].[16] Whilst the reaction could not be driven to
completion, tetrahydro-b-carboline 8 was successively recy-
cled to provide sufficient throughput of material for com-
pletion of the synthesis. The excellent diastereoselectivity
observed for acetate 8 (d.r. > 20:1) was in stark contrast to
that for the corresponding benzyl ether N-methyl carboxylate,
which proceeded to give only a 5:1 mixture of diastereomers,
presumably reflecting a subtle electronic effect (see Support-
ing Information).
Advancement of 16 to the skeletal rearrangement sub-
strate 7 commenced with selective phenolic methylation to
provide 17, whose acetate group was then methanolyzed and
replaced with a benzyl group. Rupture of the superfluous
N,O-acetal and methylation of the ensuing phenol was
accomplished through treatment with an excess of Cs2CO3
(5.0 equiv) and MeI (30 equiv) in DMF, which afforded imine
18 in 65% yield over three steps. Finally, ester saponification
with LiOH, followed by acid chloride formation [(COCl)2]
and treatment with iPr2NEt, afforded the hexacyclic bis-
enamine 7 in 60% overall yield from imine 18. The bis-
enamine 7 was now set to undergo oxidation and rearrange-
ment to the appropriate haplophytine framework. Following
our previously developed procedure,[7] treatment of 7 with
mCPBA in CH2Cl2 at ꢀ58C smoothly afforded the rearranged
crystalline ketone 19 [m.p. = 182–1838C (CH2Cl2/EtOH); see
ORTEP drawing,[17] Scheme 3] in 78% yield, presumably
through intermediates 20 and 21. The targeted indole 5 was
then finally accessed in 63% yield through oxidation of
indoline 19 with DDQ.
ꢀ
the stage for the final carbon carbon bond formation through
radical cyclization. In the event, heating a mixture of xanthate
28 and nBu3SnH (3.0 equiv) in the presence of AIBN
(1.0 equiv) delivered nonacycle 29 as a single diastereoisomer
Scheme 4. Preparation of nonacycle 29. Reagents and conditions: a) LiTMP (0.67m in THF, 4.0 equiv), MeOBPin (5.0 equiv), THF, ꢀ1008C,
15 min; b) 6 (1.5 equiv), [Pd(dppf)Cl2] (0.2 equiv), Ph3As (0.5 equiv), TlOEt (3.0 equiv), DMSO, 238C, 1 h, 67% over two steps; c) NaH
(3.0 equiv), MeI (7.0 equiv), DMF, 238C, 0.5 h, 87%; d) Tf2O (2.0 equiv), DTBMP (3.0 equiv), CH2Cl2, 238C, 0.5 h; e) NaBH4 (1.5 equiv), CH2Cl2/
MeOH (2:1), ꢀ788C, 2 min, 76% over two steps; f) HF·py (excess), THF, 238C, 0.5 h; g) NaH (3.0 equiv), CS2 (34 equiv), THF, 0!238C, 1 h;
then MeI (26 equiv), 238C, 1 h, 70% for the two steps; h) nBu3SnH (3.0 equiv), AIBN (1.0 equiv), PhH, 858C, 2 h, 32%. AIBN=2,2’-
azobisisobutyronitrile, DMSO=dimethylsulfoxide, dppf=1,1’-bis(diphenylphosphino)ferrocene, DTBMP=2,6-di-tert-butyl-4-methylpyridine,
Pin=2,3-dimethylbutane-2,3-diolate, py=pyridine, Tf=trifluoromethanesulfonyl, TMP=2,2,6,6-tetramethylpiperidine.
7618
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7616 –7620

Angewandte
Chemie
Scheme 5. Completion of the total synthesis of (+)-haplophytine; a) TBAF (1.0m in THF, 5.0 equiv), THF, 238C, 2 h; then K3[Fe(CN)6] (10 equiv),
NaHCO3 (20 equiv), tBuOH/H2O (1:2), 0.5 h, 71%; b) BCl3 (1.0m in CH2Cl2, 10 equiv), pentamethylbenzene (5.0 equiv), CH2Cl2, 08C, 1 h, 58%;
c) TESOTf (5.0 equiv), 2,6-lutidine (10 equiv), CH2Cl2, 238C, 0.5 h, 68%; d) CH2O (37% aq., excess), NaBH3CN (5.0 equiv), AcOH (20 equiv),
MeOH, 238C, 0.5 h; e) K3[Fe(CN)6] (10 equiv), NaHCO3 (20 equiv), THF/tBuOH/H2O (1:1:2), 238C, 15 min; f) TBAF (1.0m in THF, 10 equiv),
THF, 238C, 0.5 h, 42% over three steps. TBAF=tetra-n-butyl ammonium fluoride, TES=triethylsilyl.
in 32% yield. The rather moderate yield for this reaction is
thought to relate to increased hindrance to cyclization relative
to aspidophytine imposed by the left-hand domain,[5f] implied
by the isolation of significant amounts of the deoxygenated,
yet uncyclized compound corresponding to 28.
methods for providing solutions to long standing problems in
natural product synthesis.
Received: August 17, 2009
Published online: September 10, 2009
With the entire carbon skeleton of haplophytine in place,
completion of the total synthesis required lactone formation,
and deprotection/N-methylation of the left-hand nitrogen
atom. The flexibility of our end-game was narrowed, however,
by an inability to remove the benzyl and Cbz groups without
concomitant removal of the TMSE ester. As such, lactone
formation was first carried out, through a sequence involving
desilylation with TBAF followed by in situ treatment of the
resulting carboxylic acid with K3[Fe(CN)6], to afford
smoothly the desired decacyclic lactone 30 (71% yield,
Scheme 5). Removal of the benzyl and Cbz protecting
groups was then achieved through treatment with BCl3 in
the presence of pentamethylbenzene,[18] which cleanly pro-
vided desmethyl haplophytine 31. Unfortunately, direct
access to haplophytine by N-methylation of 31 was hampered
by the basicity of the tertiary amine, rapid quaternization of
which was encountered with all electrophilic methylation
reagents tested.
Keywords: aminal rearrangement · natural products ·
radical cyclization · Suzuki–Miyaura coupling · total synthesis
.
[1] K. C. Nicolaou, T. Montagnon, Molecules That Changed the
World, Wiley-VCH, Weinheim, 2008, p. 366.
[2] For previous synthetic studies toward haplophytine, see: a) P.
Yates, D. A. Schwartz, Can. J. Chem. 1983, 61, 509 – 518; b) D. A.
Schwartz, P. Yates, Can. J. Chem. 1983, 61, 1126 – 1131; c) P. D.
Rege, Y. Tian, E. J. Corey, Org. Lett. 2006, 8, 3117 – 3120.
[3] H. Ueda, H. Satoh, K. Matsumoto, K. Sugimoto, T. Fukuyama,
H. Tokuyama, Angew. Chem. 2009, 121, 7736 – 7739; Angew.
Chem. Int. Ed. 2009, 48, 7600 – 7603.
[4] a) E. F. Rogers, H. R. Snyder, R. F. Fischer, J. Am. Chem. Soc.
1952, 74, 1987 – 1989; b) H. R. Snyder, R. F. Fischer, J. F. Walker,
H. E. Els, G. A. Nussberger, J. Am. Chem. Soc. 1954, 76, 2819 –
2825; c) H. R. Snyder, R. F. Fischer, J. F. Walker, H. E. Els, G. A.
Nussberger, J. Am. Chem. Soc. 1954, 76, 4601 – 4605; d) H. R.
Snyder, H. F. Strohmayer, R. A. Mooney, J. Am. Chem. Soc.
1958, 80, 3708 – 3710.
Following extensive investigation, it was found that
selective silylation of the free phenol (32) was necessary to
enable N-methylation under reductive amination conditions.
Whilst this also led to reduction of the N,O-acetal, providing
the carboxylic acid 33, in situ reoxidation with K3[Fe(CN)6]
returned lactone 34. Finally, desilylation of 34 with TBAF
provided haplophytine (1) in 42% overall yield from 32. All
physical properties (1H and 13C NMR spectra, IR, MS, and
[a]D) of synthetic haplophytine (1) were in accordance with
those obtained from a sample of the natural substance.[19]
In summary, a total synthesis of (+)-haplophytine has
been achieved, demonstrating the power of modern synthetic
[5] For total syntheses of aspidophytine, see: a) F. He, Y. Bo, J. D.
Altom, E. J. Corey, J. Am. Chem. Soc. 1999, 121, 6771 – 6772;
b) S. Sumi, K. Matsumoto, H. Tokuyama, T. Fukuyama, Org.
Lett. 2003, 5, 1891 – 1893; c) S. Sumi, K. Matsumoto, H.
Tokuyama, T. Fukuyama, Tetrahedron 2003, 59, 8571 – 8587;
d) J. M. Mejꢁa-Oneto, A. Padwa, Org. Lett. 2006, 8, 3275 – 3278;
e) J. P. Marino, G. Cao, Tetrahedron Lett. 2006, 47, 7711 – 7713;
f) K. C. Nicolaou, S. M. Dalby, U. Majumder, J. Am. Chem. Soc.
2008, 130, 14942 – 14943.
[6] a) I. D. Rae, M. Rosenberger, A. G. Szabo, C. R. Willis, P. Yates,
D. E. Zacharias, G. A. Jeffrey, B. Douglas, J. L. Kirkpatrick, J. A.
Weisbach, J. Am. Chem. Soc. 1967, 89, 3061 – 3062; b) D. E.
Zacharias, Acta Crystallogr. Sect. B 1970, 26, 1455 – 1464; c) P.
Angew. Chem. Int. Ed. 2009, 48, 7616 –7620
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7619

Communications
Yates, F. N. MacLachlan, I. D. Rae, M. Rosenberger, A. G.
The Cambridge Crystallographic Data Centre via www.ccdc.
cam.ac.uk/data_request/cif.
Szabo, C. R. Willis, M. P. Cava, M. Behforouz, M. V. Lakshmi-
kantham, W. Zeiger, J. Am. Chem. Soc. 1973, 95, 7842 – 7850.
[7] K. C. Nicolaou, U. Majumder, S. P. Roche, D. Y.-K. Chen,
Angew. Chem. 2007, 119, 4799 – 4802; Angew. Chem. Int. Ed.
2007, 46, 4715 – 4718.
[8] K. Matsumoto, H. Tokuyama, T. Fukuyama, Synlett 2007, 3137 –
3140.
[9] N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457 – 2483.
[10] G. Jones, S. P. Stanforth, Org. React. 1997, 49, 1 – 330.
[11] C.-C. Yang, H.-T. Chang, J.-M. Fang, J. Org. Chem. 1993, 58,
3100 – 3105.
[16] The structural identity of 16 was confirmed by its conversion to a
compound prepared independently from the related benzyl
ether 16a (m.p. = 144–1458C, EtOAc/hexanes), the major
isomer obtained from an analogous oxidative coupling reaction,
which proved sufficiently crystalline for X-ray crystallographic
analysis (see ORTEP drawing below). See the supporting
information for further details. CCDC 730565 contains the
supplementary crystallographic data for 16a.
[12] G. Buechi, C.-P. Mak, J. Org. Chem. 1977, 42, 1784 – 1786.
[13] W. M. Whaley, T. R. Govindachari, Org. React. 1951, 6, 74 – 150.
[14] N. Uematsu, A. Fujii, S. Hashiguchi, T. Ikariya, R. Noyori, J. Am.
Chem. Soc. 1996, 118, 4916 – 4917.
[15] The absolute configuration of 8 was confirmed through the
preparation (see the Supporting Information) and X-ray crys-
tallographic analysis of the Mosher amide derivative 8a (m.p. =
131–1328C, CH2Cl2/hexanes, see ORTEP drawing below).
CCDC 706349 contains the supplementary crystallographic
data for 8a. These data can be obtained free of charge from
[17] CCDC 730566 contains the supplementary crystallographic data
for 19.
[18] K. Okano, K. Okuyama, T. Fukuyama, H. Tokuyama, Synlett
2008, 1977 – 1980.
[19] We thank Professor E. J. Corey (Harvard University) for kindly
providing us with an authentic sample of natural haplophytine.
7620 www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 7616 –7620