Concise total synthesis of (+)-WIN 64821 and (-)-ditryptophenaline

Article abstract of DOI:10.1002/anie.200704960

(Chemical Equation Presented) On a fast track: The secondary metabolites (+)-WIN 64821 and (-)-ditryptophenaline have been synthesized in six and seven steps, respectively, from amino acid derivatives in a concise and enantioselective manner. The gram-scale synthesis of key intermediates and the simultaneous introduction of vicinal quaternary stereocenters are described. The synthesis and structural confirmation of (-)-1′-(2-phenylethylene) ditryptophenaline is also reported.

Full text of DOI:10.1002/anie.200704960

Angewandte
Chemie
DOI: 10.1002/anie.200704960
Natural Product Synthesis
Concise Total Synthesis of (+)-WIN 64821 and
(ꢀ)-Ditryptophenaline**
Mohammad Movassaghi,* Michael A. Schmidt, and James A. Ashenhurst
The structurally fascinating and biologically active secondary
metabolites (+)-WIN 64821 (1) and (ꢀ)-ditryptophenaline
(2), isolated from Aspergillus flavus cultures, are members of
the dimeric diketopiperazine alkaloid family (Scheme 1).[1]
methyltransferase,[1e] and (+)-11,11’-dideoxyverticillin A
(Scheme 1), a tyrosine kinase inhibitor with potent antitumor
activity.[1f] Based on the pioneering work of Hino,[4a] Naka-
gawa et al. reported the first synthesis of (ꢀ)-2 through a
thallium(III)-promoted oxidative dimerization reaction (in
3% yield).[5] In 2001, Overman and Paone reported an
elegant total synthesis of (ꢀ)-ent-WIN 64821 and (ꢀ)-2 in
which alkylation reactions were employed for the introduc-
tion of the quaternary stereocenters.[6] Herein we describe a
concise enantioselective total synthesis of naturally occurring
alkaloids (+)-1 and (ꢀ)-2 in six and seven steps, respectively,
from commercially available amino acid derivatives. Addi-
tionally, we report the conversion of (ꢀ)-2 into N-styrenyl
derivatives as well as the structural confirmation of (ꢀ)-3.
The retrosynthetic analysis of (+)-WIN 64821 (1) illus-
trates our planned approach to preparing these dimeric
diketopiperazine alkaloids (Scheme 2). We envisioned simul-
Scheme 1. Representative dimeric diketopiperazine alkaloids.
Many of these alkaloids, including the closely related (ꢀ)-N1-
(2-phenylethylene)ditryptophenaline (3),[2] contain vicinal
quaternary stereocenters[3] that connect two hexahydropyr-
roloindole substructures (Scheme 1).[4] Bioactivity-guided
studies led to the identification of (+)-1 as a potent
competitive substance P antagonist with submicromolar
potency for the human neurokinin 1 and the cholecystoki-
nin B receptors,[2] whereas alkaloids (ꢀ)-2 and (ꢀ)-3 were
found to be weaker inhibitors for the former receptor.[1] Many
closely related and potently biologically active epidithiodike-
topiperazine derivatives[1d] are known, including (+)-chaeto-
cin (Scheme 1), the first inhibitor of a lysine-specific histone
Scheme 2. Retrosynthetic analysis of (+)-WIN64821 ( 1).
taneously securing the imposing vicinal quaternary stereo-
centers of (+)-1 by a reductive dimerization[7,8] of a C3-
halogenated diketopiperazine 5 (Scheme 2). While diketopi-
perazine 6 could be readily accessed from l-tryptophan and
l-phenylalanine, the strategic positioning of an electron-
[*] Prof. Dr. M. Movassaghi, M. A. Schmidt, Dr. J. A. Ashenhurst
Department of Chemistry
withdrawing group(E) on the indolyl nitrogen atom of
6
Massachusetts Institute of Technology
Cambridge, MA 02139 (USA)
E-mail: movassag@mit.edu
could allow the preparation of the desired C3-halogenated
derivative 5. Inspired by the pioneering reports by the
research groups of Hino,[4a] Crich,[4c] and Danishefsky[9] on
the synthesis and chemistry of C3a-functionalized hexahy-
dropyrroloindoles, we envisioned that a C3- halogenated
diketopiperazine 5 would serve as a versatile precursor to a
short-lived intermediate 4 en route to (+)-1.
A short synthesis of the key diketopiperazine of the
general structure 5 is shown in Scheme 3. The direct
N sulfonylation of N-Boc-l-tryptophan (7) was achieved by
Homepage: http://web.mit.edu/movassag/www/index.htm
[**] M.M. is a Beckman Young Investigator. J.A.A. acknowledges a
postdoctoral fellowship from Fonds QuØbØcois de la Recherche sur
la Nature et les Technologies. We are grateful to Prof. L. E. Overman
for a copy of the 1H NMR spectrum of (ꢀ)-1. We acknowledge
generous support from Amgen, GlaxoSmithKline, Boehringer
Ingelheim Pharmaceutical Inc., and Merck Research Laboratories.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2008, 47, 1485 –1487
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1485

Communications
Scheme 3. Concise total synthesis of (+)-WIN64821 ( 1), (ꢀ)-ditryptophenaline (2), and (ꢀ)-N1-(2-phenylethylene)ditryptophenaline (3):
a) LiHMDS, THF, PhSO2Cl, ꢀ788C, 71%. b) EDC·HCl, HOBt, Et3N , CHCl2, 238C, 94%. c) TFA, CH2Cl2, 0!238C, 3 h; then morpholine, CH2Cl2,
2
238C, 48 h, 80%. d) Br2, MeCN, 08C, 15 min, 86%. e) [CoCl(PPh3)3] (1.8 equiv), acetone, 238C, 30 min, 48%. f) SmI2 (6.0 equiv), NMP, tBuOH,
THF, 08C, 1 h, 75%. g) MeI, K2CO3, acetone, 238C, 3 days, 93%. h) [CoCl(PPh3)3] (1.8 equiv), acetone, 238C, 15 min, 52%. i) SmI2 (6.6 equiv),
NMP, tBuOH, THF, 08C, 35 min, 79%. j) BnCHO, MeCN, 708C, 8 h, 29% or BnCH(OMe)2, CSA, 238C, 81%; then H2O, C6D6, TFA, 238C, 48%.
Boc=tert-butyloxycarbonyl, LiHMDS=lithium bis(trimethylsilyl)amide, EDC·HCl=1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride,
HOBt=1-hydroxybenzotriazole, TFA=trifluoroacetic acid. NMP=N-methyl-2-pyrrolidinone, Bn=benzyl, CSA=(ꢁ)-10-camphorsulfonic acid.
treatment with LiHMDS (3 equiv)[10] followed by benzene-
sulfonyl chloride (Scheme 3). Condensation of the tryptophan
derivative (ꢀ)-8 and l-phenylalanine methyl ester (9) pro-
vided the desired amide (ꢀ)-10. Dissolving (ꢀ)-10 in
dichloromethane and subsequent treatment with trifluoro-
acetic acid followed by morpholine resulted in precipitation
of the target diketopiperazine (ꢀ)-11 as a single diastereomer
and with greater than 99% ee.[11,12] Importantly, attempts to
directly N sulfonylate the cyclo-l-tryptophan-l-phenylala-
nine (6, E = H) were unsuccessful because of its sensitivity
toward base-promoted epimerization, which lead to a mixture
of diastereomers. The bromides endo-(+)-12 and exo-(ꢀ)-13,
which are the key precursors for (+)-WIN 64821 (1) and
(ꢀ)-ditryptophenaline (2), respectively, were prepared in a
combined yield of 86% by exposure of (ꢀ)-11 to bromine in
acetonitrile.[13] The diastereomeric bromides were easily
separated and were found to be amenable to storage on a
scale greater than 10 g.
The total synthesis of (+)-WIN 64821 was then completed
in two additional steps from the endo-bromide (+)-12
(Scheme 3). After extensive experimentation with various
reaction parameters and substrates,[14] a practical set of
reaction conditions was identified for the dimerization of
diketopiperazines of the general structure 5 (Scheme 2).
Under optimized reaction conditions, treatment of (+)-12
with tris(triphenylphosphine)cobalt chloride (14, 1.8 equiv)[15]
in acetone (0.1m with respect to (+)-12) at 238C provided
direct access to the N-sulfonylated dimer (ꢀ)-15 as a single
diastereomer in 43–48% yield. Importantly, this reductive
dimerization exclusively provided the required cis-5,5-fused
bicycle of the hexahydropyrroloindole substructure.[16] It
should be noted that the dimerization substrate endo-bromide
(+)-12, and to a lesser extent the diketopiperazines in the
exo series (for example 13, Scheme 3), as well as the
corresponding dimerization products were found to be
sensitive toward base-promoted epimerization and autoxida-
tive decomposition. Ultimately, reductive removal of the
N-benzenesulfonyl groups of (ꢀ)-15 under optimized reaction
conditions was achieved by using samarium diiodide
(6.0 equiv) in a mixture of anhydrous tetrahydrofuran,
N-methylpyrrolidinone, and tert-butanol to give the first
synthetic sample of the natural enantiomer (+)-WIN 64821
(1, [a]2D1 = + 230 (c = 0.15, MeOH)); lit.:[1b] [a]D = + 200 (c =
0.15, MeOH) in 75% yield.[11] Notably, these conditions did
ꢀ
not lead to significant reductive fragmentation of the C3 C3’
bond, nor the epimerization of the base-sensitive diketopi-
perazine substructure.
Similarly, the total synthesis of (ꢀ)-ditryptophenaline (2)
was completed in three steps from exo-bromide (ꢀ)-13
(Scheme 3). Treatment of (ꢀ)-13 with methyl iodide and
potassium carbonate gave the corresponding N14-Me exo-
bromide (ꢀ)-16 in 93% yield. Treatment of (ꢀ)-16 with the
cobalt(I) complex 14 in acetone at 238C afforded the dimer
(ꢀ)-17 as a single diastereomer in 47–52% yield. Reductive
removal of the benzenesulfonyl groups provided (ꢀ)-ditryp-
tophenaline (2, [a]2D1 = ꢀ292 (c = 0.97, CH2Cl2)); lit.:[1a] [a]2D4
=
ꢀ330 (c = 0.52, CH2Cl2) in 79% yield (Scheme 3).[11] Signifi-
cantly, the reaction conditions described here for the di-
merization event were directly applicable to gram-scale
synthesis (for example (+)-12!(ꢀ)-15, 43% yield on a 1-g
1486
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1485 –1487

Angewandte
Chemie
Saita, Tennen Yuki Kagobutsu Toronkai Koen Yoshishu 1994, 36,
scale, and (ꢀ)-16!(ꢀ)-17, 47% yield on a 2.5-g scale). The
successful application of this key transformation to both the
endo and exo series of diketopiperazine substrates
(Scheme 3) offers a practical route for late-stage assembly
of related derivatives.
557; d) U. Anthoni, C. Christophersen, P. H. Nielsen in Alkaloids
Chemical and Biological Perspectives, Vol. 13 (Ed.: S. W. Pellet-
ier), Pergamon, London, 1999, pp. 163 – 236; e) D. Greiner, T.
Bonaldi, R. Eskeland, E. Roemer, A. Imhof, Nat. Chem. Biol.
2005, 1, 143; f) Y.-X. Zhang, Y. Chen, X.-N. Guo, X.-W. Zhang,
W.-M. Zhao, L. Zhong, J. Zhou, Y. Xi, L.-P. Lin, J. Ding, Anti-
Cancer Drugs 2005, 16, 515.
Heating a solution of (ꢀ)-ditryptophenaline (2) at 708C
with excess phenylacetaldehyde in acetonitrile over 8 hours
provided the first synthetic sample of (ꢀ)-3 ([a]2D2 = ꢀ131.5
(c = 0.36, CHCl3); lit.:[2] [a]D = ꢀ125 (c = 0.05, CHCl3))[11] in
29% yield, accompanied by products derived from thermal
decomposition. The spectral data for our synthetic sample of
(ꢀ)-3 matched that for the natural product, thus confirming
the reported structure for this alkaloid. The thermal decom-
position of (ꢀ)-3 can be avoided by using a two-stepsequence
at ambient temperature. Condensation of (ꢀ)-2 with the
dimethoxyacetal of phenylacetaldehyde at 238C gave the
N1,N1’-bis-b-styrene derivative (ꢀ)-18 in 81% yield within
1.5 h. The partial hydrolysis of (ꢀ)-18 at 238C cleanly
produced alkaloid (ꢀ)-3 in 48% yield in 20 minutes, with
the majority of the mass balance as recovered (ꢀ)-18.
The enantioselective total synthesis of (+)-WIN 64821 (1)
and (ꢀ)-ditryptophenaline (2) in six and seven steps, respec-
tively, from commercially available amino acid derivatives is
described. The simultaneous introduction of the vicinal
quaternary stereocenters in these alkaloids was achieved by
a reductive homodimerization of readily available alkyl
bromides. In addition to synthesizing the first synthetic
sample of naturally occurring (+)-1, we provide structural
confirmation of the natural alkaloid (ꢀ)-3. The gram-scale
synthesis of key intermediates and dimerization of bromides
(+)-12 and (ꢀ)-16 provide a concise and preparative route to
these alkaloids. Further development and application of this
chemistry to the synthesis of other homo- and heterodimeric
alkaloids is ongoing and will be reported in due course.
[2] C. J. Barrow, D. M. Sedlock, J. Nat. Prod. 1994, 57, 1239.
[3] A. Steven, L. E. Overman, Angew. Chem. 2007, 119, 5584;
Angew. Chem. Int. Ed. 2007, 46, 5488.
[4] a) T. Hino, M. Nakagawa in The Alkaloids: Chemistry and
Pharmacology, Vol. 34 (Ed.: A. Brossi), Academic Press, New
York, 1989, pp. 1 – 75; b) U. Anthoni, C. Christophersen, P. H.
Nielsen in Alkaloids Chemical and Biological Perspectives,
Vol. 13 (Ed.: S. W. Pelletier), Pergamon, London, 1999,
pp. 163 – 236; c) D. Crich, A. Banerjee, Acc. Chem. Res. 2007,
40, 151.
[5] M. Nakagawa, H. Sugumi, S. Kodato, T. Hino, Tetrahedron Lett.
1981, 22, 5323.
[6] L. E. Overman, D. V. Paone, J. Am. Chem. Soc. 2001, 123, 9465.
[7] M. Movassaghi, M. A. Schmidt, Angew. Chem. 2007, 119, 3799;
Angew. Chem. Int. Ed. 2007, 46, 3725.
[8] Y. Yamada, D.-i. Momose, Chem. Lett. 1981, 1277.
[9] K. M. Depew, S. P. Mardsen, D. Zatorska, A. Zatorska, W. G.
Bornmann, S. J. Danishefsky, J. Am. Chem. Soc. 1999, 121,
11953.
[10] P. A. Grieco, Y. S. Hon, A. Perez-Medrano, J. Am. Chem. Soc.
1988, 110, 1630.
[11] Please see the Supporting Information for details.
[12] Under the optimized reaction conditions (ꢀ)-11 is readily
purified by crystallization, which allows the preparation of
(ꢀ)-11 on a greater than 20-gram scale with equal efficiency and
without the use of flash chromatography.
[13] This bromination reaction was more selective (12/13, 16:84) in
favor of the exo-diastereomer when performed at ꢀ408C.
Alternatively, bromination reactions conducted at 408C led to
a slight excess (12/13, 52:48) of the endo diastereomer contami-
nated with by-products arising from bromination of the aniline
ring.
[14] A variety of metal (Mn, V, and Ni) and Co(I–III) complexes,
reaction solvents (> 10), concentration, temperature, addition
rate, order of addition, and additives were examined. X = Br was
optimal compared to X = Cl or I. E = SO2Ph was most effective
compared to other sulfonyl derivatives (> 5).
[15] a) M. Aresta, M. Rossi, A. Sacco, Inorg. Chim. Acta 1969, 3, 227;
b) S. L. Baysdon, L. S. Liebeskind, Organometallics 1982, 1, 771.
[16] The mass balance for this reaction is accounted by 10% of the
corresponding C3-reduction (5, X = H, E = SO2Ph) product, as
well as products (ca. 15%) consistent with radical disproportio-
nation.
Received: October 26, 2007
Published online: January 11, 2008
Keywords: alkaloids · dimerization · enantioselectivity · indole ·
.
total synthesis
[1] a) J. P. Springer, G. Büchi, B. Kobbe, A. L. Demain, J. Clardy,
Tetrahedron Lett. 1977, 18, 2403; b) C. J. Barrow, P. Cai, J. K.
Snyder, D. M. Sedlock, H. H. Sun, R. Cooper, J. Org. Chem.
1993, 58, 6016; c) M. Hiramoto, M. Shibazaki, H. Miyata, Y.
Angew. Chem. Int. Ed. 2008, 47, 1485 –1487
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1487