Total synthesis of pederin

Article abstract of DOI:10.1002/anie.200701677

(Chemical Equation Presented) Blisteringly fast: The potent cytotoxic blistering agent pederin has been synthesized (see scheme). The synthesis is diastereoselective and concise (just 12 steps for the longest linear sequence), and features a formal hetero

Full text of DOI:10.1002/anie.200701677

Communications
DOI: 10.1002/anie.200701677
Natural Products
Total Synthesis of Pederin**
John C. Jewett and Viresh H. Rawal*
Dedicated to Professor Hisashi Yamamoto
Pederin (1, Figure 1) has a long and rich history among
bioactive natural products.[1] Present in the haemolymph of
rove beetles of the genus Paederus,[2] this compound was first
the pederamine part on the right—connected through an
amide bond to a labile aza acetal. The chemical synthesis of
pederin has been pursued by many groups, and with successes
having been reported by the groups of Matsumoto, Kocienski,
and Nakata.[13,14] Despite years of work, fundamental issues
about pederin, such as its cellular target or its mode of action,
remain unresolved. An efficient synthesis of pederin that
proceeds with good diastereoselectivity would enable the
preparation of sufficient quantities of the natural product and
analogues for such chemical biology studies. In connection
with our interest in this family of natural products, we recently
reported the total synthesis of mycalamide A (2).[15,16] We
detail below a synthesis of pederin that is concise, stereocon-
trolled, and high yielding.
Our synthetic strategy was devised to provide rapid access
to pederin and its analogues, and to address the difficult
stereochemical issues that plagued earlier syntheses. The plan
was to take advantage of a diastereoselective hetero-Diels–
Alder reaction (HDA) or a Mukaiyama aldol/cyclization
sequence[17] for the convergent assembly of two comparably
sized pieces, and thereby generate a functionalized dihydro-
pyran-4-one, in which the stereochemistry at the C15 position
would be set as syn relative to that at C17. At the outset, the
HDA was expected to be complicated by the fact that the
required Danishefsky-type diene 4 has a terminal gem-
dimethyl moiety, and it would be reacting with an aldehyde
in which the existing chirality resides on a carbon atom b to
the carbonyl group. Adding to these difficulties was the
recognition that the undesired diastereomer would be favored
based on both chelation and nonchelation models.[18] In the
event, the reaction of 4 and 3a, catalyzed by traditional Lewis
acids, gave the expected pyranone product (5a, R = Me;
Table 1, entries 1 and 2) in, at best, modest diastereoselectiv-
ity, in favor of the undesired anti diastereomer.[19] The
insensitivity of dimethoxy aldehyde 3a to even bulky Lewis
acids, such as the MAPH complex reported by Yamamoto
and Saito (Table 1, entry 3),[20] prompted us to examine a
derivative in which the b-alkoxy substituent possessed
significantly altered steric and electronic properties.
Figure 1. Two members of the pederin family.
obtained in crystalline form in 1919 by Netolitzky and was
shown to be a potent vesicant.[3] Pavan and Bo led a massive
isolation effort, using 25 million field-collected Paederus
fuscipes beetles, to secure a sizable quantity of pure pederin,[4]
to allow its chemical formula and provisional structure to be
determined by Quilico and co-workers in 1965.[5,6] Independ-
ently, Matsumoto et al. proposed a slightly modified structure
for pederin,[7] one that was subsequently corroborated
through X-ray analysis.[8]
Over the years much attention has been accorded to
pederin, due in no small part to its impressive and varied
biological activities. In addition to its blistering and necrotic
properties,[9] pederin displays nanomolar toxicity against a
broad range of cancer cell lines.[10] Although the details of its
mode of action remain unclear, pederin has been shown to
inhibit protein and DNA synthesis without significantly
affecting RNA synthesis,[11] and to suppress entry of cells
into prophase, thereby blocking mitosis in normal and tumor
cells at concentrations of 1–10 ngmLÀ1.[12] Pederin also slows
the growth of sarcoma 180 tumors in mice.[12]
From a chemical perspective, pederin presents a formi-
dable challenge for total synthesis. Its intricate structure,
decorated with nine chiral centers, is composed of two
tetrahydropyran rings—the pederic acid unit on the left and
[*]J. C. Jewett, Prof. Dr. V. H. Rawal
Department of Chemistry
A solution to the diastereoselectivity problem was found
by using TIPS-protected aldehyde 3b. The reaction of 3b
catalyzed by BF3·OEt2 proceeded with excellent diastereose-
lectivity, albeit to provide the undesired diastereomer of the
pyranone (Table 1, entry 4). Quenching the reaction mixture
with HF removed both silyl groups and promoted the
cyclization of the Mukaiyama aldol product to provide 5b
(R = H). We were pleased to find that when same reaction
was catalyzed with the Yamamoto–Saito Lewis acid,[20] it
produced the desired pyranone diastereomer in good yield
and selectivity (Table 1, entry 6). Activation of the aluminum
University of Chicago
5735 South Ellis Avenue
Chicago, IL 60637 (USA)
Fax: (+1)773-702-0805
E-mail: vrawal@uchicago.edu
Homepage: http://rawalgroup.uchicago.edu
[**]Generous financial support from the National Cancer Institute of
the NIH (R01 CA101438) is gratefully acknowledged. We thank
Merck & Co. for additional support.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
6502
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 6502 –6504

Angewandte
Chemie
Table 1: Formal HDA diastereoselectivity.[a]
axial methyl group is expected to disfavor approach of the
nucleophile from the bottom face (A and B, Figure 2). Of the
remaining two possibilities, transition state C presents fewer
steric interactions, as it places the carbonyl carbon atom with
its alkoxy groups farther away from the olefinic carbon atoms
of the pyranone. Various conditions were examined for the
reduction of pyranone 8, and the best result was obtained with
L-selectride, which predominantly afforded the desired
diastereomer (9, 12:1). We have found that reduction of the
ketone functionality at a later stage to be less selective. The
required C10-nitrogen atom was expected to be revealed
through a Curtius rearrangement, which proceeds stereospe-
cifically. In preparation for this transformation, ester 9 was
saponified (NaOH, MeOH) and the resulting acid was
subjected to the rearrangement protocol,[22] which gave the
Teoc-protected pederamine 10, with complete retention of
the stereochemistry at C10. Silation of the hydroxy group
gave 11.
Entry
R
Lewis acid (equiv)
Ratio
Yield
syn (desired)/anti [%]
1
2
3
4
5
6
Me BF3·OEt2 (2)
Me TiCl4 (2)
Me Al(OAr)2Me (1), TMSOTf (1)
TIPS BF3·OEt2 (2)
TIPS TiCl4 (2)
22:78
51:49
22:78
6:94
75
62
60
52
nd
65
40:60
TIPS Al(OAr)2Me (0.2), TMSOTf (2) 92:8
[a]OAr =2,6-diphenylphenol, TMS=trimethylsilyl, Tf=trifluorometha-
nesulfonyl, TIPS=triisopropylsilyl, TBS=tert-butyldimethylsilyl, nd=
not determined.
alkoxide with TMSOTf is essential for a successful reaction.
The full sequence of reactions leading to pyranone 5b is
shown in Scheme 1. The TIPS-protected aldehyde 3b was
available in three steps from commercially available (S)-(+)-
glycidyl methyl ether (6). Diene 4 was prepared from a
vinylogous ester precursor under thermodynamic silation
conditions. The free hydroxy group on the pyranone product
was methylated to provide 5a.
Control of the stereochemistry of the C10-methoxy group
has plagued the previous routes to pederin. For example,
reduction of the methoxyimine of a fully assembled pederin
skeleton provided the required aza acetal with around 2:1
diastereoselectivity, in favor of the undesired diastereomer.[13]
Our plan was to introduce both the stereocenters at C10 and
C11 simultaneously by taking advantage of a diastereoselec-
tive Mukaiyama–Michael reaction, in which a carboxylate
functionality would serve as the masked form of the C10-
nitrogen atom. Thus the conjugate addition reaction of silyl
ketene acetal 7 onto pyranone 5a, when catalyzed by
Sc(OTf)3,[21] produced ester 8 in excellent diastereoselectivity.
Only the two C11-epimers were observed and the desired
diastereomer predominated by a 20:1 ratio (Scheme 2). The
observed selectivity can be rationalized by considering the
transition states shown in Figure 2, wherein the dimethoxy-
propane side chain is held in the equatorial orientation. The
Scheme 2. Synthesis of Teoc-pederamine (11): a) Sc(OTf)3 (10 mol%),
CH2Cl2, À788C to RT; 10% HF/MeCN quench; b) L-selectride, THF,
À788C; c) 10% NaOH/MeOH, RT; d) DPPA, Et3N, 2-trimethylsilyletha-
nol, 4-Š MS, THF, reflux; e) TMSCl, imidazole, DMF, RT. DPPA=di-
phenylphosphoryl azide, DMF=N,N-dimethylformamide, MS=molec-
ular sieves, Teoc=2-(trimethylsilyl)ethoxycarbonyl.
Scheme 1. Synthesis of 5a: a) vinyl Grignard reagent, CuI, THF,
À788C to 08C; b) TIPSCl, imidazole, CH2Cl2; c) O3, CH2Cl2; PPh3
quench; d) NaOMe, Et2O, reflux; DMSO, Me2SO4; e) TBSOTf, Et3N,
Et2O, 08C to RT; f) Al(2,6-diphenyl-phenoxide)2Me (0.4 equiv), TMSOTf
(2.1 equiv), CH2Cl2, À788C to À458C; 10% HF/MeCN quench;
g) NaHMDS, MeI, THF, À788C to 08C. DMSO=dimethyl sulfoxide,
HMDS=hexamethyldisilazane.
Figure 2. Rationale for Mukaiyama–Michael selectivity. LA=Lewis acid.
Angew. Chem. Int. Ed. 2007, 46, 6502 –6504
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
6503

Communications
[1] Extracts from the beetle were used in ancient Chinese medicine
as a remedy for ringworm and other skin maladies. For a
thorough review of the natural history of pederin, see: J. H.
Frank, K. Kanamitsu, J. Med. Entomol. 1987, 24, 155 – 191.
[2] Paederus beetles, and hence pederin, are suggested to be
responsible for two of the plagues chronicled in the book of
Exodus; see: S. A. Norton, C. Lyons, Lancet 2002, 359, 1950.
[3] F. Z. Netolitzky, Z. Angew. Entomol. 1919, 5, 252 – 257.
[4] M. Pavan, G. Bo, Mem. Soc. Entomol. Ital. 1952, 31, 67 – 82.
[5] C. Cardini, D. Ghiringhelli, R. Mondelli, A. Quilico, Tetrahedron
Lett. 1965, 6, 2537 – 2545, and references therein.
[6] C. Cardini, D. Ghiringhelli, R. Mondelli, A. Quilico, Gazz.
Chim. Ital. 1966, 96, 3 – 38.
[7] T. Matsumoto, S. Yanagiya, S. Maeno, S. Yasuda, Tetrahedron
Lett. 1968, 9, 6297 – 6300.
[8] A. Furusaki, T. Watanabe, T. Matsumoto, M. Yanagiya, Tetrahe-
dron Lett. 1968, 9, 6301 – 6304.
[9] M. Pavan, G. Bo, Phys. Comp. Oec. 1953, 307 – 312.
[10] A. Richter, P. Kocienski, P. Raubo, D. Davies, Anti-Cancer Drug
Des. 1997, 12, 217 – 227.
[11] A. Brega, A. Falaschi, L. de Carli, M. Pavan, J. Cell Biol. 1968,
36, 485 – 496.
[12] M. Soldati, A. Fioretti, M. Ghione, Experientia 1966, 22, 176 –
178.
[13] a) F. Matsuda, N. Tomiyoshi, M. Yanagiya, T Matsumoto,
Tetrahedron 1988, 44, 7063 – 7080; b) P. Kocienski, R. Narqui-
zian, P. Raubo, C. Smith, L. Farrugia, K. Muir, F. T. Boyle, J.
Chem. Soc. Perkin Trans. 1 2000, 2357 – 2384; c) T. Takemura, Y.
Nishii, S. Takahashi, J. Kobayashi, T. Nakata, Tetrahedron 2002,
Scheme 3. Synthesis of pederin: a) SOCl2, pyridine, CH2Cl2, RT;
b) LiHMDS, THF, À788C to RT; c) TBAF, THF, 08C, LiOH/MeOH
quench. Bz=benzoyl, TBAF=tetra-n-butylammonium fluoride.
58, 6359– 6365; d) for
a review, see: R. Narquizian, P. J.
Coupling of pederic acid with the aza acetal fragment has
been a challenging problem in the pederin family. Indeed,
whereas carbamates such as 11 can be coupled to small acid
chlorides, their direct coupling to the full pederic acid unit has
not been reported.[22]
We were pleased to find the direct coupling strategy to be
highly effective for assembling the two halves of pederin
(Scheme 3). Pederic acid[15a] was converted into its acid
chloride in the presence of an excess of pyridine so as to
avoid degradation of the labile acetal moiety.[13a] The addition
of a toluene solution of this acid chloride to the lithium anion
of 11 gave the protected pederin 14 in 75% yield over two
steps. Complete deprotection by treatment with TBAF
followed by a hydrolytic quench gave pederin (1) in 88%
yield.
In conclusion, we have completed a concise, asymmetric
total synthesis of pederin. The synthetic route proceeds in
12 steps and 11% overall yield for the longest linear
sequence, and it features the diastereocontrolled synthesis
of pyranone 5a and construction of the amide linkage with
control of the stereochemistry of the aminal group. Impor-
tantly, the strategy is sufficiently general so as to allow the
synthesis of not only analogues of pederin, but also many
other members of the pederin family. Studies in this regard, as
well as those aimed at the identification of the biological
target of pederin, are currently underway.[24]
Kocienski in The Role of Natural products in Drug Discovery
(Eds.: J. Mulzer, R. Bohlmann), Springer, New York, 2000,
pp. 25 – 56 (Ernst Schering Research Foundation Workshop 32).
[14] It is worth noting that, despite these impressive achievements,
there has yet to be a synthesis of pederin with a longest linear
sequence of less than 20 steps.
[15] a) J.-H. Sohn, N. Waizumi, V. H. Rawal, J. Am. Chem. Soc. 2005,
127, 7290 – 7291; b) H. M. Zhong, J.-H. Sohn, V. H. Rawal, J.
Org. Chem. 2007, 72, 386 – 397; c) H. M. Zhong, PhD Disserta-
tion, The Ohio State University, Columbus, OH (USA), 1995;
d) C. M. K. Rademacher, PhD Dissertation, The Ohio State
University, Columbus, OH (USA), 1997.
[16] The remarkable structural similarity between pederin and
mycalamide A, as well as the closely related marine-sponge
isolates, led to the suggestion that these compounds may all be
produced by symbiotic microorganisms; see: a) J. Piel, D.
Butzke, N. Fusetani, D. Hui, M. Platzer, G. Wen, S. Matsunaga,
J. Nat. Prod. 2005, 68, 472 – 479; b) R. L. L. Kellner, K. Dettner,
Oecologia 1996, 107, 293 – 300.
[17] a) S. Danishefsky, Acc. Chem. Res. 1981, 14, 400 – 406; b) L. F.
Tietze, G. Kettschau, Top. Curr. Chem. 1997, 189, 1 – 120.
[18] D. A. Evans, J. L. Duffy, M. J. Dart, Tetrahedron Lett. 1994, 35,
8537 – 8540.
[19] T. Willson, P. Kocienski, K. Jarowicki, K. Isaac, P. M. Hitchcock,
A. Faller, S. F. Campbell, Tetrahedron 1990, 46, 1767 – 1782.
[20] H. Yamamoto, S. Saito, Chem. Commun. 1997, 1585 – 1592.
[21] Y. Yamashita, S. Saito, H. Ishitani, S. Kobayashi, J. Am. Chem.
Soc. 2003, 125, 3793 – 3798.
[22] W. R. Roush, T. G. Marron, Tetrahedron Lett. 1993, 34, 5421 –
5424.
[23] In the mycalamide series, it was found that the corresponding
protected amine could not be coupled to pederic acid, presum-
ably because of steric effects; see: W. R. Roush, L. A. Pfeifer,
Org. Lett. 2000, 2, 859– 862.
Received: April 16, 2007
Published online: July 23, 2007
[24] Y.-S. Lee, PhD Dissertation, University of Chicago, Chicago, IL
(USA), 2005.
Keywords: Curtius rearrangement · diastereoselectivity ·
Mukaiyama–Michael reaction · natural products · total synthesis
.
6504 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 6502 –6504