Total synthesis of piericidin A1 and B1 and key analogues

Article abstract of DOI:10.1021/ja0632862

Full details of the total synthesis of piericidin A1 and B1 and its extension to the preparation of a series of key analogues are described including ent-piericidin A1 (ent-1), 4′-deshydroxypiericidin A1 (58), 5′-desmethylpiericidin A1 (73), 4′-deshydroxy-5′- desmethylpiericidin A1 (75), and the corresponding analogues 51, 59, 76, and 77 bearing a simplified farnesyl side chain. The evaluation of these key analogues, along with those derived from their further functionalizations, permitted a scan of the key structural features providing new insights into the role of the substituents found in both the pyridyl core as well as the side chain. A strategic late stage heterobenzylic Stille cross-coupling reaction of the pyridyl core with the fully elaborated side chain permitted ready access to the analogues in which each half of the molecule could be systematically and divergently modified. The pyridyl cores were assembled enlisting inverse electron demand Diels-Alder reactions of N-sulfonyl-1-azabutadienes, while key elements of side chain syntheses include an anti selective asymmetric aldol to install the C9 and C10 relative and absolute stereochemistry (for natural and ent-1) and a modified Julia olefination for formation of the C5-C6 trans double bond with convergent assemblage of the side chains.

Full text of DOI:10.1021/ja0632862

Published on Web 08/18/2006
Total Synthesis of Piericidin A1 and B1 and Key Analogues
Martin J. Schnermann,† F. Anthony Romero,† Inkyu Hwang,†
Eiko Nakamaru-Ogiso,‡ Takao Yagi,‡ and Dale L. Boger*,†
Contribution from the Departments of Chemistry and Molecular and Experimental Medicine,
and the Skaggs Institute for Chemical Biology, The Scripps Research Institute,
10550 North Torrey Pines Road, La Jolla, California 92037
Received May 10, 2006; E-mail: boger@scripps.edu
Abstract: Full details of the total synthesis of piericidin A1 and B1 and its extension to the preparation of
a series of key analogues are described including ent-piericidin A1 (ent-1), 4′-deshydroxypiericidin A1 (58),
5′-desmethylpiericidin A1 (73), 4′-deshydroxy-5′-desmethylpiericidin A1 (75), and the corresponding
analogues 51, 59, 76, and 77 bearing a simplified farnesyl side chain. The evaluation of these key analogues,
along with those derived from their further functionalizations, permitted a scan of the key structural features
providing new insights into the role of the substituents found in both the pyridyl core as well as the side
chain. A strategic late stage heterobenzylic Stille cross-coupling reaction of the pyridyl core with the fully
elaborated side chain permitted ready access to the analogues in which each half of the molecule could
be systematically and divergently modified. The pyridyl cores were assembled enlisting inverse electron
demand Diels-Alder reactions of N-sulfonyl-1-azabutadienes, while key elements of side chain syntheses
include an anti selective asymmetric aldol to install the C9 and C10 relative and absolute stereochemistry
(for natural and ent-1) and a modified Julia olefination for formation of the C5-C6 trans double bond with
convergent assemblage of the side chains.
Introduction
Piericidin A1 (1) is the prototypical member of an important
class of biologically active natural products isolated from
Streptomyces mobaraensis and pactam (Figure 1).1 The most
widely recognized biological target of 1 and members of this
natural product class is the mitochondrial electron transport chain
protein NADH-ubiquinone reductase (Complex I), and the
piericidins are among the most potent inhibitors identified to
date (Ki ) 0.6-1.0 nM). Complex I mediates the transfer of
protons into the intermembrane space of the mitochondria
through a NADH driven reduction of ubiquinone (3) to its
hydroquinone,2 and the piericidins have contributed extensively
to the elucidation of the properties of this highly complex
enzyme.3 Although initially investigated as an insecticide, the
piericidins have been found to exhibit a range of promising
biological activity. Some piericidins have been shown to exhibit
antimicrobial and antifungal activity with minimum inhibitory
concentrations as low as 6 µg/mL, while others have been shown
to block antibody formation (inhibit the immune response to
antigen exposure) at the remarkably low concentration of 100
pg/mL.4-6 Most recently, 1 was identified as a potent inhibitor
Figure 1. Structure of Piericidin A1 and B1.
of interleukin-2 (IL-2) production in a cellular functional assay.7
IL-2 is a regulator of many essential immune responses, and
its overexpression is linked to malignancy, autoimmune diseases,
organ transplant rejection, and renal allograft rejection. Addi-
tionally, and of the most significance to our own interests, 1
was identified as a highly selective antitumor agent with greater
selectivity and potency than those of the comparison standard
mitomycin C.8 In fact, a number of Complex I inhibitors from
the rotenoid class of natural products, of which rotenone (4) is
the prototypical member, have received considerable attention
as anticancer agents and have exhibited efficacious in vivo
activity in animal models.9 Although the precise mechanism
responsible for the selective antitumor activity of such Complex
† Department of Chemistry and the Skaggs Institute for Chemical
Biology.
‡ Department of Molecular and Experimental Medicine.
(1) Takahashi, N.; Suzuki, A.; Tamura, S. J. Am. Chem. Soc. 1965, 87, 2066.
(2) Nelson, D. L.; Cox, M. M. Lehninger Principles of Biochemistry; Worth:
New York, 2000; Chapter 19.
(6) Matsumoto, M.; Mogi, K.; Nagaoka, K.; Ishizeki, S.; Kawahara, R.;
Nakashima, T. J. Antibiot. 1987, 40, 149.
(3) Esposti, M. D. Biochim. Biophys. Acta 1998, 1364, 222.
(4) Yoshida, S.; Yoneyama, K.; Shiraishi, S.; Watanbe, A.; Takahashi, N. Agric.
Biol. Chem. 1977, 41, 849.
(5) Kimura, K.; Takahashi, H.; Miyata, N.; Yoshihama, M.; Uramoto, M. J.
Antibiot. 1996, 49, 697.
(7) Ahn, S. C.; Kim, B. Y.; Oh, W. K.; Kang, D. O.; Heo, G. Y.; Kim, M. S.;
Lee, M. S.; Ahn, J. S. J. Antibiot. 2002, 55, 1013.
(8) Kitaura, N.; Shirata, K.; Niwano, M.; Mimura, M.; Takahara, Y. Japanese
Patent 05339156 A2, 1993; Chem. Abstr. 1994, 120, 208590.
9
10.1021/ja0632862 CCC: $33.50 © 2006 American Chemical Society
J. AM. CHEM. SOC. 2006, 128, 11799-11807
11799

A R T I C L E S
Schnermann et al.
a N-sulfonyl-1-aza-1,3-butadiene24 with tetramethoxyethene25
followed by Lewis acid promoted aromatization for assemblage
of the functionalized pyridine core.26 Based on prior work with
such systems, an azadiene bearing an additional strong electron-
withdrawing group at C-2 would be expected to facilitate the
Diels-Alder reaction such that the fully substituted and sensitive
tetramethoxyethene could be anticipated to react under mild
conditions. Additional key elements of the approach include
the use of an asymmetric anti-aldol reaction to install the C9
and C10 relative and absolute stereochemistry, a modified Julia
olefination for formation of the C5-C6 trans double bond with
convergent assemblage of the side chain, and a surprisingly
effective penultimate heterobenzylic Stille cross-coupling reac-
tion of the pyridyl core with the fully elaborated side chain.
Notably, the strategic late stage coupling of the fully function-
alized pyridyl core with the fully elaborated side chain was
anticipated to permit access to analogues in which either half
of the molecule could be systematically and divergently27
altered.
Thus, in addition to the natural products, we were especially
interested in the synthesis of a series of analogues that would
define the key features of this class of molecules responsible
for their biological activity. Of particular interest was the
suggestion that the 4-hydroxypyridine of 1, as the pyridone
tautomer, mimics the quinone of ubquinone (coenzyme Q, 3)
with the side chain of 1 mimicking the prenyl chain. This
apparent structural relationship has been confirmed through their
competitive binding against Complex I. However, competitive
substrate binding may not fully account for the inhibitory activ-
ity, as 1 binds to at least two sites on Complex I: one compet-
itive with coenzyme Q and one competitive with rotenone,
neither of which are competitive with each other.3,28,29 Moreover
and although piericidin A1 competitively blocks the binding of
endogenous ubiquinone (the most upstream inhibition site of
the inhibitor-binding domain of Complex I), the majority of
specific Complex I inhibitors exhibit noncompetitive or un-
competitive behavior against the artificial substrate n-decylubi-
quinone.30 Thus, the absence of mutual exclusivity between the
inhibitor and the ubiquinone binding sites presents a major
challenge in identifying the actual Complex I sites of action of
these potent agents and their relation to the ubquinone reduction
sites.31 Thus, the exact role of the apparent structural homology
between 1 and 3 has not been fully established. It was envisioned
that a carefully constructed series of analogues could be used
to probe such questions.
Figure 2. Retrosynthetic Analysis.
I inhibitors has not been fully elucidated, several explanations
have been advanced.10-13
Despite this array of potentially useful biological activity,
no total synthesis of a member of this large class of natural
products had been described prior to our initial disclosure.14
Shortly following this preliminary disclosure, Phillips and Keaton
reported a complementary total synthesis of (-)-7-demethyl-
piericidin A1, a natural product closely related to 1.15 In earlier
efforts directed at this family, a racemic synthesis of analogues
incorporating the full side chain, but a substituted benzene in
place of the substituted pyridine of 1, has been disclosed, and
an asymmetric synthesis of the C6-C13 segment of the side
chain bearing the most recent stereochemical assignment
(9R,10R) has been reported.16,17 In addition, the preparation of
several analogues bearing simplified side chains has been
disclosed.18-20 Notably, and at the time that our work and that
of Phillips were conducted, the originally assigned absolute
stereochemistry21 of the side chain substituents of 1 had been
challenged, reassigned, and remained to be unambiguously
established.22,23
Herein we provide full details of this synthesis of piericidin
A1 and B1 by an approach that, like that of Phillips, established
the absolute stereochemistry of the natural products and report
its subsequent extension to the synthesis and evaluation of a
series of key analogues (Figure 2). Central to the approach is
the use of an inverse electron demand Diels-Alder reaction of
(9) Udeani, G. O.; Gerhauser, C.; Thomas, C. F.; Moon, R. C.; Kosmeder, J.
K.; Kinghorn, A. D.; Moriarty, R. M.; Pezzuto, J. M. Cancer Res. 1997,
57, 3424 and references therein.
(10) Fang, N.; Casida, J. E. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 3380.
(11) Li, N.; Ragheb, K.; Lawler, G.; Sturgis, J.; Rajwa, B.; Melendez, J. A.;
Robinson, J. P. J. Biol. Chem. 2003, 278, 8516.
(12) Kushnareva, Y.; Murphy, A. N.; Andreyev, A. Biochem. J. 2002, 368, 545.
(13) Sherer, T. B.; Betarbet, R.; Testa, C. M.; Seo, B. B.; Richardson, J. R.;
Kim, J. H.; Miller, G. W.; Yagi, T.; Matsuno-Yagi, A.; Greenamyre, J. T.
J. Neurosci. 2003, 34, 10756.
(14) Schnermann, M. J.; Boger, D. L. J. Am. Chem. Soc. 2005, 127, 15704.
(15) Keaton, K. A.; Phillips, A. J. J. Am. Chem. Soc. 2006, 128, 408.
(16) (a) Ono, M.; Yoshida, N.; Kokubu, Y.; Sato, E.; Akita, H. Chem. Pharm.
Bull. 1997, 45, 1428. (b) Ono, M.; Yoshida, N.; Akita, H. Chem. Pharm.
Bull. 1997, 45, 1745.
(24) (a) Boger, D. L.; Corbett, W. L.; Wiggins, J. M. J. Org. Chem. 1990, 55,
2999. (b) Boger, D. L.; Corbett, W. L.; Curran, T. T.; Kasper, A. M. J.
Am. Chem. Soc. 1991, 113, 1713. (c) Boger, D. L.; Kasper, A. M. J. Am.
Chem. Soc. 1989, 111, 1517. (d) Boger, D. L.; Curran, T. T. J. Org. Chem.
1990, 55, 5439. (e) Clark, R. C.; Pfeiffer, S. S.; Boger, D. L. J. Am. Chem.
Soc. 2006, 128, 2587. (f) Reviews: Boger, D. L. Tetrahedron 1983, 39,
2869. Boger, D. L. Chem. ReV. 1986, 86, 781. Boger, D. L. Chemtracts:
Org. Chem. 1996, 9, 149.
(25) Scheeren, J. W.; Staps, R. J. F. M.; Nivard, R. J. F. Recl. TraV. Chim.
Pays-Bas 1973, 92, 11.
(26) (a) Boger, D. L.; Cassidy, K. C.; Nakahara, S. J. Am. Chem. Soc. 1993,
115, 10733. (b) Boger, D. L.; Hu¨ter, O.; Mbiya, K.; Zhang, M. J. Am.
Chem. Soc. 1995, 117, 11839. (c) Boger, D. L.; Hong, J. J. Am. Chem.
Soc. 1998, 120, 1218. (d) Boger, D. L.; Blagg, B. S. J. Tetrahedron 2002,
58, 6343.
(17) Cox, C. M.; Whiting, D. A. J. Chem. Soc., Perkin Trans. 1 1991, 1901.
(18) (a) Schmidtchen, F. P.; Rapoport, H. J. Am. Chem. Soc. 1977, 99, 7014.
(b) Gutman, M.; Klaitchko, S. FEBS Lett. 1976, 67, 348.
(19) Yoshida, S.; Takahashi, N. Heterocycles 1978, 10, 425.
(20) Chung, K. H.; Cho, K. Y.; Asami, Y.; Takahashi, N.; Yoshida, S. Z.
Naturforsch 1989, 44c, 609.
(27) Boger, D. L.; Brotherton, C. E. J. Org. Chem. 1984, 49, 4050.
(28) Friedrich, T.; Van Heek, P.; Leif, H.; Ohnishi, T.; Forche, E.; Kunze, B.;
Jansen, R.; Trowitzsch-Kienast, W.; Hofle, G.; Reichenbach, H.; Weiss,
H. Eur. J. Biochem. 1994, 219, 691.
(21) Takahashi, N.; Suzuki, A.; Kimura, Y.; Miyamoto, S.; Tamura, S.
Tetrahedron Lett. 1967, 8, 1961.
(29) Okun, J. G.; Lummen, P.; Brandt, U. J. Biol. Chem. 1999, 274, 2625.
(30) Ramsay, R. R.; Krueger, M. J.; Youngster, S. K.; Gluck, M. R.; Casida, J.
E.; Singer, T. P. J. Neurochem. 1991, 56, 1184.
(22) Yoshida, S.; Shiraishi, S.; Fujita, K.; Takahashi, N. Tetrahedron Lett. 1975,
16, 1863.
(23) Jansen, R.; Hofle, G. Tetrahedron Lett. 1983, 24, 5485.
(31) Lummen, P. Biochim. Biophys. Acta 1998, 1364, 287.
9
11800 J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006

Piericidin A1 and B1 and Key Analogues
A R T I C L E S
Scheme 1
Figure 3. X-ray crystal structures of 13 and 17.
Et3N, CH2Cl2, 0 °C, 20 min).34 Direct conversion of the ketone
10 to 12 (MeSO2NH2, TiCl4, 0 °C, 1 h, 40%) was also possible
but was not pursued due to lower initial conversions. Treatment
of 12 with 8 (toluene, 50 °C, 18 h, 64% for two steps from 11)
smoothly afforded the [4 + 2] cycloadduct 13. As anticipated,
the electron-deficient N-sulfonyl-1-azadiene 12 substituted with
the additional C2 electron-withdrawing substituent26a,b partici-
pated in the Diels-Alder reaction with the electron-rich
dienophile 8 at or near room temperature even though 8 is
tetrasubstituted and sterically demanding. Despite its potential
hydrolytic lability, the cycloadduct 13 (mp 116-118 °C) proved
remarkably robust being isolated by conventional chromatog-
raphy on SiO2 and its structure was confirmed by single-crystal
X-ray diffraction (Figure 3).35 Although efforts to induce
aromatization of 13 under a variety of basic conditions were
surprisingly unsuccessful, having been anticipated to result from
a well-precedented C4 deprotonation and elimination of a C3
methoxy group followed by methanesulfonyl deprotonation and
loss of sulfene concurrent with C2 methoxy elimination,36 the
use of the Lewis acid BF3‚OEt2 cleanly affected this transforma-
tion (CH2Cl2, 0 °C, 1 h, 88%). Reduction of the ethyl ester 14
with DIBALH (CH2Cl2, 0 °C, 1 h, 92%) was followed by
protection of the resulting alcohol 15 as its TIPS ether provided
16 (TIPSCl, imidazole, DMF, 25 °C, 18 h, 95%). We had
envisioned that directed lithiation of the pyridyl C4 position
followed by a borylation/oxidation sequence would serve to
introduce the remaining oxygen substituent.37 Thus, treatment
of 16 with excess base followed by trimethylborate (5 equiv of
BuLi, 6 equiv of B(OMe)3, -78 °C, 1 h, 88%), and oxidative
cleavage of the resulting aryl boronate ester provided the
C-silylated product 17 resulting from both the desired C4′
hydroxylation and a competitive reverse Brook rearrangement
of the benzylic TIPS ether under the strongly basic conditions.38
This unexpected result was confirmed with an X-ray analysis
of 17 (Figure 3).35 The use of fewer equivalents of base resulted
in the product arising only from silyl migration indicating that
the primary site of the deprotonation under these conditions was
the benzylic center, and the use of alternative silyl ether protect-
ing groups (TBS and TBDPS) or the free alcohol 15 itself
resulted in lower yields of the diol 19. The deprotection of 17
proceeded cleanly and, interestingly, through the O-silylated
intermediate 18 resulting from an initial Brook rearrangement
to give 19 (Bu4NF, THF, 25 °C, 36 h, 96%). Thus, if the
desilylation reaction was quenched after 30 min versus 36 h,
the intermediate O-silyl ether 18 was isolated in yields as high
Scheme 2
Total Synthesis of Piericidin A1 and Piericidin B1
Our initial objective was to use the Diels-Alder reaction of
tetramethoxyethene with N-sulfonyl-1-azadiene 632 bearing a
C4 benzyloxy substituent permitting direct access to the fully
substituted pyridine upon base-catalyzed aromatization (Scheme
1). Conversion of the ketone 5 to the requisite azadiene 6
(PhSO2NH2, TiCl4, 0 °C, 1 h, 20%) proved problematic pro-
viding 6 in low yields. Similarly, the alternative two-step
approach enlisting the oxime 7 did not prove viable. More
significantly, N-sulfonyl-1-azadiene 6 failed to react with either
825 or ethyl vinyl ether in an initial survey of reaction conditions.
Based on this initial finding, it was decided that the C4′
hydroxyl could be installed following the Diels-Alder and
aromatization sequence permitting the use of the less substituted
and more electron-deficient N-sulfonyl-1-azadiene 12 (Scheme
2). This key N-sulfonyl-1-aza-1,3-butadiene was accessed
through a two-step sequence initiated by oxime formation of
the ketone 1033 (NH2OH-HCl, EtOH, 25 °C, 18 h, 96%). Treat-
ment of the oxime 11 with Et3N and methylsulfinyl chloride
led to initial formation of the O-methanesulfinate and its in situ
homolytic rearrangement to give sulfonylimine 12 (CH3SOCl,
(33) Rambaud, M.; Bakasse, M.; Duguay, G.; Villieras, J. Synthesis 1988, 564.
(34) Brown, C.; Hudson, R. F.; Record, K. A. F. J. Chem. Soc., Perkin Trans.
2 1978, 822.
(35) The coordinates of the X-ray structures of 13 (CCDC 600039) and 17
(CCDC 600040) have been deposited with the Cambridge Crystallographic
Data Centre.
(36) Boger, D. L.; Nakahara, S. J. Org. Chem. 1991, 56, 880.
(37) Trecourt, F.; Mallet, M.; Mongin, F.; Queguiner, G. J. Org. Chem. 1994,
59, 6173.
(32) Synthesis described in the Supporting Information.
(38) Wright, A.; West, R. J. Am. Chem. Soc. 1974, 96, 3214.
9
J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006 11801

A R T I C L E S
Scheme 3
Schnermann et al.
Scheme 5. Synthesis of Piericidin A1 and B1
Scheme 4
chiral enolate derived from 24 providing 26 (84% ee, 23% for
three steps to 27) from which the undesired diastereomers
proved difficult to separate. Consequently, we adopted a more
recent preparation of aldehyde 30 disclosed by Nakao42 utilizing
a procedure developed by Heathcock44 enlisting 2 equiv of Bu2-
BOTf to promote an anti-selective aldol addition of the Evans
optically active N-acyl oxazolidinone 28. As our work pro-
gressed, we also briefly examined the conditions recently dis-
closed by Evans45 for promoting such direct anti-aldol additions
of the optically active N-acyl oxazolidinones. Thus, the reaction
of 31 with 25 provided 32 (MgCl2, TMSCl, Et3N, EtOAc, 25 °C,
24 h, 49%) in good yield and excellent diastereoselectivity.
Protection of 32 (TBSCl, imidazole, DMF, 25 °C, 16 h) followed
by reduction with LiBH4 (THF, MeOH, 0 to 25 °C, 76% for
two steps) provided an alternative and comparable approach to
alcohol 27 which was oxidized to the aldehyde 30.17 The
aldehyde 30 was converted to alcohol 34 in two steps as
previously disclosed17 and oxidized under Swern conditions to
give 35 ((COCl)2, DMSO, Et3N, -78 °C, 99%), Scheme 5. The
modified Julia coupling between 35 and 23 cleanly provided
36 (KHMDS, DME, -78 °C, 18 h, 60%) with the trans alkene
isomer as the only detected product. Lithium-halogen exchange
of the vinyl iodide 36 upon treatment with n-BuLi, followed
by treatment of the vinyllithium with Bu3SnCl, provided the
vinyl stannane 37 (n-BuLi, -78 °C, 20 min; Bu3SnCl, 20 min,
-78 to 25 °C) for the Stille coupling with 20.
as 80%, and its resubjection to the reaction conditions for a
longer period resulted in clean conversion on to 19. Conversion
of 19 to the heterobenzylic bromide 20 (1 equiv of CBr4, 1 equiv
of Ph3P, CH2Cl2, 25 °C, 30 min, 84%) provided a surprisingly
stable partner for the Stille cross-coupling given that the parent
2-bromomethyl pyridine is unstable to storage.
We anticipated that a Julia-Kocienski olefination39 coupling
would allow introduction of the requisite trans C5-C6 alkene
while convergently assembling the side chain. Such an approach
necessitated the preparation of the vinyl iodide 23 that consti-
tutes the C2-C5 segment. This was accomplished through the
reaction of alcohol 2140 with 1-phenyl-1H-tetrazole-5-thiol
(PTSH) under Mitsunobu conditions (DEAD, Ph3P, THF, 0 °C,
30 min, 71%) to provide thioether 22, which was cleanly
oxidized to the sulfone 23 with H2O2-ammonium molybdate
(EtOH-THF, 25 °C, 6 h, 89%), Scheme 3.41
With this sulfone bearing the vinyl iodide in hand, the stage
was set for its coupling with the C6-C13 side chain aldehyde
35, accessible from the known alcohol 27.17 Limitations in the
reported synthesis of the alcohol 27 (problematic diastereomer
separation)17 ultimately resulted in our use of a recently dis-
closed and more effective asymmetric anti-aldol reaction for
the preparation of 30 (Scheme 4).42
Thus, Whiting et al. reported17 the preparation of the aldehyde
30 in their efforts directed at the piericidin A1 side chain
utilizing an asymmetric anti-aldol43 of the aldehyde 25 and the
In initial studies on this final coupling reaction, it was found
that only the Pd2(dba)3/t-Bu3P catalyst system employed by Fu
and co-workers was able to affect the desired reaction between
the bromide 20 with the simplified vinyl stannane 50 (vide
infra).46 In these studies, it was found that coupling required
elevated temperatures, high catalyst loading, and LiCl as an
additive to achieve good conversions, which when applied to
(39) Blakemore, P. R.; Cole, W. J.; Kocienski, P. J.; Morley, A. Synlett 1998,
26.
(40) Ndibwani, A.; Lamothe, S.; Guay, D.; Plante, R.; Soucy, P.; Goldstein, S.;
Deslongchamps, P. Can. J. Chem. 1993, 71, 695.
(41) Ahmed, A.; Hoegenauer, E. K.; Enev, V. S.; Hanbauer, M.; Kaehlig, H.;
Ohler, E.; Mulzer, J. J. Org. Chem. 2003, 68, 3026.
(42) Nakao, Y.; Yoshida, W. Y.; Takada, Y.; Kimura, J.; Yang, L.; Mooberry,
S. L.; Scheuer, P. J. J. Nat. Prod. 2004, 67, 1332.
(43) Gennari, C.; Bernardi, A.; Colombo, L.; Scolastico, C. J. Am. Chem. Soc.
1985, 107, 5812.
(44) Danda, H.; Hansen, M. M.; Heathcock, C. H. J. Org. Chem. 1990, 55,
173.
(45) Evans, D. A.; Tedrow, J. S.; Shaw, J. T.; Downey, C. W. J. Am. Chem.
Soc. 2002, 124, 392.
(46) Littke, A. F.; Schwarz, L.; Fu, G. C. J. Am. Chem. Soc. 2002, 124, 6343.
9
11802 J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006

Piericidin A1 and B1 and Key Analogues
A R T I C L E S
Scheme 6. Synthesis of Analogues 43, 44, 45, and 46
the coupling of 20 with 37 provided 38 in superb conversions
(Pd2(dba)3, t-Bu3P, LiCl, dioxane, 70 °C, 18 h, 74%) without
protection of the pyridyl phenol. Little or no reaction was
observed at room temperature, (Ph3P)4Pd and (PhCN)2PdCl2
failed to support the reaction, DMF (100 °C, 20%) was much
less effective as a solvent than dioxane, and the addition of base
(NaHCO3, i-Pr2NEt) did not affect or improve the conversions.
A final deprotection of the TBS ether 38 (Bu4NF, THF, 50 °C,
12 h, 93%) provided piericidin A1 identical in all respects with
properties reported for the natural product (Scheme 5). However
the magnitude of the optical rotation for synthetic 1 and that
reported for natural 1 ([R]D25 +1.8 (c 0.1, MeOH) for synthetic
1 vs lit47 [R]D25 +1.0 (c 0.1, MeOH)) was not sufficient for an
unambiguous assignment of the absolute configuration. To more
confidently address this assignment, the conversion of 1 to
piericidin B1 (2), which exhibits a larger optical rotation, was
conducted.48 Thus, selective protection of the pyridine hydroxyl
of 1 as its pivolate ester 39 (PivCl, HSO4NBu4, aqueous 5 N
NaOH-CH2Cl2, 30 min, 25 °C, 92%), followed by O-methyla-
tion of the remaining secondary alcohol 40 (NaH, MeI, DMF,
25 °C, 1 h, 78%), and pivolate hydrolysis of 40 (t-BuONa,
MeOH, 60 °C, 3 h, 88%) provided 2. The properties of synthetic
piericidin B1 proved identical in all respects to properties
reported for natural 2, including its optical rotation ([R]D25 -7.3
of the oxazolidinone 31. The synthesis of the C10 hydroxy
diastereomer 43, as well as the C10 ketone 46, required protec-
tion of the C4 pyridyl phenol after it was established that direct
oxidation (Dess-Martin, PCC) of 1 itself as well as Mitsunobu
inversion of the C10 hydroxy group of 1 or 39 were not
straightforward. Thus, Dess-Martin oxidation (CH2Cl2, 25 °C,
30 min, 96%) of the pivolate ester 39, enlisted as an intermediate
in the total synthesis of piericidin B1, provided the ketone 41
(Scheme 6). Luche reduction of 41 (4 equiv of NaBH4, 6 equiv
of CeCl3‚7H2O, EtOH, -78 to 0 °C) afforded a readily separable
1:1 mixture of the desired C10 diastereomer 42 (38%) and
regenerated 39 (38%) without competitive conjugate reduction.
Hydrolysis of the pivolate ester 42 (t-BuONa, MeOH, 60 °C, 3
h, 79%) provided 43, the C10 diastereomer of piericidin A1.
The corresponding ketone 46 was not accessible from the
pivolate ester 41 requiring hydrolysis conditions that were too
harsh for 41 to withstand. Consequently, piericidin A1 was
converted to acetate 44 (AcCl, HSO4NBu4, aqueous 5 N
NaOH-CH2Cl2, 25 °C, 30 min, 93%), a key analogue of 1 in
its own right, which was oxidized to the ketone 45 (Dess-
Martin periodinane, CH2Cl2, 25 °C, 30 min, 89%). Mild
hydrolysis of 45, also an interesting analogue of 1, provided
the ketone 46 (NaHCO3, MeOH, 4 days, 71%).
Finally, two analogues containing simplified side chains were
prepared for comparison. The first, 51, incorporates the farnesyl
side chain found in coenzyme Q (3). Thus, the vinyl stannane
48, accessed from the known vinyl iodide 4749 (n-BuLi, -78 °C,
20 min; Bu3SnCl, 20 min, -78 to 25 °C), was coupled with 20
using the conditions developed for 1 (Pd2(dba)3, t-Bu3P, LiCl,
dioxane, 70 °C, 18 h) to provide 51 (76%), an analogue first
described in the studies of Rapoport (Scheme 7).18 In addition,
53 incorporating a truncated side chain was also prepared.
Coupling of the vinyl stannane 50, derived from the known vinyl
iodide 4950 (n-BuLi, -78 °C, 20 min; Bu3SnCl, 20 min, -78
to 25 °C) afforded 52 (Pd2(dba)3, t-Bu3P, LiCl, dioxane, 70 °C,
18 h, 68%) which was deprotected (Bu4NF, THF, 25 °C, 8 h,
83%) to afford 53.
18
(c 0.2, MeOH) vs lit48 [R]D -6.5 (c 3.2, MeOH)), thereby
confirming the absolute configuration assignment for 1 and 2.
Key Analogues
Following the completion of the total synthesis of 1 and 2,
our efforts turned to the preparation and evaluation of a series
of key analogues. Throughout our studies and the handling of
1, 2, 17, 19, 20, and 38, only the 4-hydroxypyridine tautomer,
and not the 4-pyridone tautomer, was observed even in protic
solvents. Provocatively, this suggests that the ability of 1 to
bind and inhibit NADH-ubiquinone reductase (Complex I) may
result from mimicry of reduced coenzyme Q (hydroquinone)
rather than 3 itself. Thus, modifications of the C4′ hydroxyl
could directly address such questions. Similarly, modifications
to the side chain of the piericidins would permit the role of its
substituents to be more clearly defined.
Key Side Chain Analogues. Early studies on the piericidins
illustrated that simplifications in the side chain, mimicking that
found in coenzyme Q, provide derivatives that maintain much
of the potency of the natural products.18,19 However, all such
work was conducted only with an assessment of Complex I
inhibition and without regard for comparisons in subsequent
cellular functional assays that would more specifically address
their potential antitumor activity. The results of prior studies
would suggest that removal of C9 and C10 substituents and/or
alterations of the stereochemistry might not have a significant
impact.18,19 Consequently, to further define the role of the side
chain C9 and C10 stereocenters, ent-1, the inverted C10 hydroxy
diastereomer 43, and the C10 ketone 46, along with 2, were
viewed as key structures. The unnatural enantiomer of piericidin
A1, ent-1 in which both the C10 and C9 stereocenters are
inverted, was accessed through a sequence identical to that
described for the total synthesis of 1 enlisting the enantiomer
(47) Urakawa, A.; Sasaki, T.; Yoshida, K.; Otani, T.; Lei, Y.; Yun, W. J. Antibiot.
1996, 49, 1052.
(48) Takahashi, N.; Suzuki, A.; Kimura, Y.; Miyamoto, S.; Tamura, S.; Mitsui,
T.; Fukami, J. Agric. Biol. Chem. 1968, 32, 1115.
(49) Svatos, A.; Urbanova, K.; Valterova, I. Collect. Czech. Chem. Commun.
2002, 67, 83.
(50) Andresen, G.; Eriksen, A. B.; Dalhus, B.; Gunderson, L. L.; Rise, F. J.
Chem. Soc., Perkin Trans. 1 2001, 1662.
9
J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006 11803

A R T I C L E S
Schnermann et al.
Scheme 7. Synthesis of Analogues 51 and 53
Scheme 9. Synthesis of Analogues 58 and 59
Scheme 10
Scheme 8. Synthesis of Analogues 54 and 55
Key Pyridyl Analogues. A key objective of the analogue
assessment was to probe the role of the pyridyl C4′ hydroxyl
group within 1. This was addressed both by its selective
protection (i.e., 44 and 45) and with the total synthesis of 4′-
deshydroxypiericidin A1 (59). Thus, complementing the avail-
ability of the labile C4′ acetates 45 and 46 of piericidin A1 and
its C10 ketone, the hydrolytically stable C4′ methyl ether 54
was prepared by selective methylation of piericidin A1 (MeI,
HSO4NBu4, aqueous 5 N NaOH-CH2Cl2, 25 °C, 2 h, 90%),
Scheme 8. Similarly, C4′ phenol methylation of 51, bearing the
farnesyl side chain, was also conducted providing 55 (MeI,
HSO4NBu4, aqueous 5 N NaOH-CH2Cl2, 25 °C, 2 h, 90%).
More significantly, C4′-deshydroxypiericidin A1 (58) was
prepared by total synthesis for examination. The availability of
the pyridyl alcohol 15, its one-step conversion to the Stille cross-
coupling substrate 56 (1 equiv of CBr4, 1 equiv of Ph3P, CH2-
Cl2, 25 °C, 30 min, 79%), and its convergent coupling with 37,
constituting the fully functionalized piericidin A1 side chain,
provided 57 (Pd2(dba)3, t-Bu3P, LiCl, dioxane, 70 °C, 3 h, 66%)
in two steps from materials available from the total synthesis
of 1 and 2 (Scheme 9). A final silyl ether deprotection of 57
(HF-pyridine, 25 °C, 48 h, 54%) provided 58. Similarly, the
C4′-deshydroxy analogue 59 bearing the simplified farnesyl side
chain was prepared by Stille coupling of 56 with 48 to provide
59 (Pd2(dba)3, t-Bu3P, LiCl, dioxane, 70 °C, 3 h, 59%) in a
single step (Scheme 9). Thus, an attribute of the route to
piericidin that emerged from the failure of 6 to productively
provide the fully substituted pyridyl core was the generation of
15 and the post cycloaddition introduction of the C4′ hydroxyl
group. Its availability, like that of a subsequent pyridyl core
(65, Scheme 10), provided ready access to the corresponding
C4′-deshydroxypiericidin A1 analogues.
our knowledge, been conducted.19 Since it represents a key
substituent also found on the quinone of coenzyme Q, a pro-
nounced impact of this substituent might well help establish
the apparent structural relationship between the piericidin pyridyl
core and the coenzyme Q quinone/hydroquinone subunit. Conse-
quently, the synthetic approach was modified to access this
alternatively substituted pyridine core. This entailed conversion
of diethyl oxalate to â-keto ester 60 (CH2dCHMgBr, -78 °C)
followed by oxime formation to provide 61 (NH2OH-HCl,
EtOH, 18 h, 45% for two steps).33 Treatment of the oxime 61
with methylsulfinyl chloride and Et3N (CH2Cl2, 0 °C, 20 min)
afforded the N-sulfonyl-1-azadiene 62 which smoothly partici-
pated in the Diels-Alder reaction with 8 to yield cycloadduct
63 (toluene, 50 °C, 18 h, 51% for two steps from 61).
Aromatization upon treatment of 63 with BF3‚OEt2 (CH2Cl2,
0 °C, 1 h, 88%) afforded the ester 64 which was reduced with
DIBALH (CH2Cl2, 0 °C, 1 h, 94%) to provide alcohol 65 that
was protected as the TIPS ether 66 (TIPSCl, imidazole, DMF,
25 °C, 18 h, 88%). Subjection of 66 to the borylation/oxidation
sequence afforded a mixture of the C- and O-silylated products
67 and 68 (5 equiv of BuLi, 6 equiv of B(OMe)3, 1 h, 37%
(67) and 35% (68)). This mixture of products does not seem to
be a result of partial Brook rearrangement of 68 to 67 in the
course of the basic aqueous workup as both products are present
by TLC prior to workup. Moreover extended exposure (18 h)
Preceding studies with simplified structures implicated the
C5′ methyl group as a key pyridyl substituent, although studies
to probe its role with an intact piericidin structure have not, to
9
11804 J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006

Piericidin A1 and B1 and Key Analogues
A R T I C L E S
Scheme 11. Synthesis of Analogues 73, 75, 76, and 77
Table 1. Compound Assessments
mitochondrial
complex I inhibition
IC50a (nM)
L1210
IC50b (nM)
compd
1
3.7
8.8
5.1
5
700
6
ent-1
2
4
5.0
20
43
44
45
46
51
53
54
55
58
59
73
75
76
77
8.0
24
26
6.3
75
3
90
600
95
6000
800
8000
800
5000
2500
4000
5500
20 000
3.8
10 000
130
750
10
400
4.0
9.0
110
2200
a Tests performed on bovine submitochondrial particles (SMP). Con-
centrations required for half-maximal inhibition, average of 1-2 determina-
tions. b L1210 Cytotoxic activity, average of 1-2 determinations performed
in triplicate.
to the aqueous workup conditions does not change the ratio of
the products, suggesting the product mixture arises in the course
of the reaction. Deprotection of the 67 and 68 mixture (Bu4NF,
THF, 25 °C, 18 h, 69%) afforded 70, and its subsequent
conversion to bromide 71 (1 equiv of CBr4, 1 equiv of PPh3,
CH2Cl2, 76%) proceeded cleanly. To access the corresponding
4′-deshydroxy-5′-desmethyl derivatives, 65 was similarly con-
verted to the bromide 69 (1 equiv of CBr4, 1 equiv of PPh3,
CH2Cl2, 81%).
With heterobenzylic bromides 69 and 71 in hand, we were
positioned to access 5′-desmethyl and 4′-deshydroxy-5′-des-
methyl analogues bearing the piericidin side chain as well as
the simplified farnesyl side chain. As such, Stille coupling of
71 with 37 afforded 72 (Pd2(dba)3, t-Bu3P, LiCl, dioxane, 70 °C,
18 h, 62%) which was deprotected (HF-pyridine, 25 °C, 48 h,
87%) to afford 5′-desmethylpiericidin A1 (73). Similarly and
without optimization, 4′-deshydroxy-5′-desmethylpiericidin A1
(75) was accessed through coupling of 69 and 37 (Pd2(dba)3,
t-Bu3P, LiCl, dioxane, 70 °C, 3 h, 49%) followed by deprotec-
tion to give 75 (HF-pyridine, 25 °C, 48 h, 73%), and the
corresponding farnesyl side chain bearing analogues 76 and 77
were accessed through coupling 71 with 48 (Pd2(dba)3, t-Bu3P,
LiCl, dioxane, 70 °C, 18 h, 58%) and 69 with 48 (Pd2(dba)3,
t-Bu3P, LiCl, dioxane, 70 °C, 3 h, 54%), respectively.
pounds in the two assays are discussed separately below and
qualified as to their relationship.
As previously disclosed, piericidin A1 proved to be a potent
Complex I inhibitor (1, IC50 ) 3.7 nM) being slightly more
potent in our assay than piericidin B1 (2, IC50 ) 5.1 nM) and
rotenone (4, IC50 ) 5.0 nM). Small modifications in the side
chain had little impact on this activity. Thus, O-methylation of
the C10 alcohol (2, piericidin B1, IC50 ) 5.0 nM), its conversion
to a ketone (46, IC50 ) 6.3 nM), and the inversion of the C10
alcohol stereochemistry with 43 (IC50 ) 8 nM) resulted in
modest e2-fold reductions in potency. More remarkably, ent-
piericidin A1 (ent-1, IC50 ) 8.8 nM), where both the C9 methyl
and C10 hydroxy are inverted, proved to be only 2.5-fold less
potent than the natural product. Consistent with these observa-
tions and with past studies,18 replacement of the side chain with
the simplified farnesyl side chain lacking these substituents and
stereocenters provided an analogue 51 (IC50 ) 3.8 nM) that
was not distinguishable from 1 itself. Nonetheless, the impor-
tance of the presence of the side chain is clear with 53 (IC50
)
10 000 nM) where the removal of the C5-C16 segment resulted
in a 103-104 fold loss in activity. In total, these observations
are consistent with piericidin mimicking coenzyme Q leading
to predictable simplifications in the inhibitor structures.
However, as noted below, these simplifications and trends
do not similarly track with the relative cytotoxic potencies of
the analogues. Rather, the cytotoxic potency of the analogues
dropped progressively with the significance of the structural
Compound Assessments
The natural products and the key analogues were examined
for inhibition of Complex I as measured against submitochon-
drial particles51 as well as for growth inhibition in a functional,
cell-based cytotoxic assay enlisiting a mouse leukemia cell line
(L1210), Table 1. At the onset, it should be noted that the latter
structure-activity relationship (SAR) data need not follow that
of the target-based assay of Complex I inhibition because of
the additional features such cell-based assays might select for
(e.g., cell permeability). Moreover, it is not yet established that
the cytotoxic activity of the piericidins is derived from Complex
I inhibition, and as such, the data derived from the cell-based
assay may reflect other or additional unidentified targets. Conse-
quently, the trends that emerged from the evaluation of com-
changes in the side chain. Whereas piericidin B1 (2, IC50
)
6 nM) was equipotent with piericidin A1 (1, IC50 ) 5 nM), the
C10 alcohol diastereomer 43 (IC50 ) 75 nM), the C10 ketone
46 (IC50 ) 600 nM), and ent-1 (IC50 ) 700 nM) proved to be
15-150-fold less potent. Similarly, 51 (IC50 ) 95 nM) bearing
the simplified farnesyl side chain was found to be 20-fold less
potent than 1, and 53 (IC50 ) 6000 nM), lacking the C5-C16
segment of the side chain, was expectedly 1000-fold less potent.
This would suggest that either the cytotoxic activity of the
compounds is not derived solely from Complex I inhibition or
the side chain substituents and their stereochemistry contribute
(51) Matsuno-Yagi, A.; Hatefi, Y. J. Biol. Chem. 1985, 260, 14424.
9
J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006 11805

A R T I C L E S
Schnermann et al.
significantly to 1 reaching its biological target in the functional
cellular assay.
The impact of the pyridyl C4′ hydroxyl group proved just as
interesting, especially on the inhibition of Complex I and in
light of the potential coenzyme Q structural relationship.
Whereas its acetylation (44, IC50 ) 24 nM; 45, IC50 ) 26 nM)
or methylation (54, IC50 ) 130 nM) led to 7-fold and 35-fold
reductions in potency, its removal altogether with 4′-deshy-
droxypiericidin (58, IC50 ) 10 nM) had a surprisingly modest
impact reducing the Complex I inhibition only 3-fold. In marked
contrast, O-methylation of the C4′ phenol (55, IC50 ) 750 nM)
as well as the removal of the C4′ phenol (59, IC50 ) 400 nM)
in the analogues bearing the simplified farnesyl side chain
reduced Complex I inhibition >100-fold. These latter studies
support the contention that piericidin A1 mimics the coenzyme
Q hydroquinone and that the C4′ hydroxy group is central to
its activity, whereas the surprisingly potent inhibition by 4′-
deshydroxypiericidin A1 (58) suggests that the natural products
side chain substituents may somehow compensate for losses in
binding affinity that result from modifications in the pyridyl
core.
With a notable exception, this modification or removal of
the pyridyl C4′ hydroxy group substantially reduced the
cytotoxic activity of the compounds. Thus O-methylation of the
piericidin C4′ phenol (54, IC50 ) 800 nM) or its removal (58,
IC50 ) 800 nM) resulted in 200-fold reductions in cytotoxic
activity, and the impact of these changes on the simplified
analogues 55 (IC50 ) 8000 nM) and 59 (IC50 ) 5000 nM)
bearing the farnesyl side chain was an even greater 1000-fold
loss in activity. The only exceptions to these generalizations
were the C4′-O-acetyl derivatives 44 (IC50 ) 3 nM) and 45
(IC50 ) 90 nM). Thus, C4′-O-acetyl piericidin A1 (44) was
found to be equally potent, or perhaps slightly more potent, than
1 itself. Although surprising on the surface, this most likely is
the result of O-acetate hydrolysis in the cellular assay releasing
the natural product itself. Similarly, 45 was found to be roughly
6-fold more potent than the free phenol 46 but still nearly
20-fold less potent than 1.
Figure 4. Key Comparisons.
nM). Significantly, this removal of the C5′ methyl group had
an even larger impact on the cytotoxic activity than the critical
C4′ hydroxy group. Incorporating these structural changes into
analogues bearing the farnesyl side chain (76 and 77) led to
even further reductions in the cytotoxic activity (IC50 ) 5500
and 20 000 nM, respectively).
Clear from these comparisons is the observation that the side
chain substituents of piericidin A1 may impact Complex I
inhibition and do so most productively and significantly when
key structural features of the pyridyl core are removed. Simi-
larly, they enhance the cytotoxic activity of the compounds,
but their impact seems to be most significant if the pyridyl core
of 1 is intact. In contrast, the pyridyl C4′ hydroxy group and
the C5′ methyl substituent appear more central to the molecule’s
properties since their modification or removal leads to reductions
in Complex I inhibition and even more significant reductions
in the cytotoxic activity where the intact pyridyl core appears
most essential. The exception to this generalization is the
unusually potent Complex I inhibition by 58, 73, and 75, where
the piericidin A1 side chain (but not farnesyl) appears to com-
pensate for their removal (Figure 4). Significantly, the series
of analogues bearing the farnesyl side chain exhibit trends
consistent with expectations of a simple coenzyme Q mimic
where both the C4′ hydroxyl group and the C5′ methyl group
each increase either the Complex I inhibition or L1210 cyto-
toxicity approximately 30-100 fold with the former target-based
(Complex I) assay exhibiting IC50’s approximately 10-fold lower
than those of the cell-based (L1210) assay. In marked contrast,
the series bearing the piericidin A1 side chain exhibited much
smaller 1-2.5-fold losses in Complex I activity with removal
of either the C4′ hydroxyl or C5′ methyl group, while experi-
The C5′ methyl group of piericidin had a similar remarkable
impact on the molecule’s properties. Its removal with 5′-
desmethylpiericidin A1 (73, IC50 ) 4.0 nM) had little or no
impact on Complex I inhibition, whereas its removal in the
analogue 76 (IC50 ) 110 nM) bearing the farnesyl side chain
reduced Complex I inhibition 30-fold relative to both 51 (IC50
) 3.8 nM) or 1 (IC50 ) 3.7 nM). Just as remarkable, removing
both the C4′ hydroxyl and C5′ methyl groups with 75 (IC50
)
9.0 nM) only reduced the Complex I inhibition ca. 2-fold relative
to piericidin A1 itself, whereas this double modification in the
analogue 77 bearing the farnesyl side chain resulted in a
compound that was >500-fold less potent than 1 or 51. As such,
the surprisingly potent activity of 73 and especially 75 suggests
that the natural product side chain substituents (vs the farnesyl
side chain) somehow compensate for losses in binding affinity
that result from modifications in the pyridyl core.
In an interesting contrast, this removal of the pyridyl C5′
methyl group led to an ca. g500-fold loss in cytotoxic potency.
Thus, 5′-desmethylpiericidin A1 (73, IC50 ) 2500 nM) proved
to be 500-fold less active than the natural product. Removing
both the C4′ hydroxy group as well as the C5′ methyl group
led to a compound that was even less potent (75, IC50 ) 4000
9
11806 J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006

Piericidin A1 and B1 and Key Analogues
A R T I C L E S
encing even greater 200-500-fold losses in cytotoxic activity
in the cellular assay. Even so, this latter series remained roughly
10-fold more potent (2-20-fold) than the farnesyl series in the
cellular assay (L1210) and was typically much more potent
(1-200-fold) against Complex I itself. Whether the lack of direct
correlation between the two series in the two assays reflects
the molecule’s ability to access Complex I in the cellular assay,
the more subtle distinctions between multiple binding sites on
Complex I and their biological consequences, or even the
intervention of additional biological targets remains to be
established.
approach that permitted access to structural analogues in which
either half of the molecule could be divergently altered at a
late stage.
Thus, the studies were extended to include total syntheses of
ent-piericidin A1 in which both the C9 and C10 stereocenters
were inverted, 4′-deshydroxypiericidin A1 (58) and 5′-desm-
ethylpiericidin (73) each lacking a key substituent of the putative
quinone mimic, 4′-deshydroxy-5′-desmethylpiericidin A1 (75),
and the simplified analogues 51, 59, 76, and 77 bearing an
unfunctionalized farnesyl side chain linked to the authentic,
4′-deshydroxypyridyl, 5′-desmethylpyridyl, or 4′-deshydroxy-
5′-desmethylpyridyl core. The evaluation of these key analogues,
along with those derived from further functionalization of the
analogues and 1, permitted a scan of the key structural features
providing new insights into the role of substituents found on
both the pyridyl core and the side chain.
Conclusion
An effective and convergent total synthesis of piericidin A1
and B1 was developed which features a facile N-sulfonyl-1-
azadiene inverse electron demand Diels-Alder reaction con-
ducted at 50 °C to access the highly functionalized pyridyl core.
The side chain was prepared enlisting an asymmetric anti-aldol
reaction to install the C9 and C10 absolute stereochemistry and
a trans-selective Julia-Kocienski C5-C6 olefination to con-
vergently assemble the side chain. A surprisingly effective and
strategic late stage Stille cross-coupling reaction of the pyridyl
core (a heterobenzylic bromide) with the fully functionalized
side chain was used to complete the synthesis. Not only did
this serve as the first disclosed total synthesis14 of a member of
this natural product class permitting an unambiguous assignment
of the unresolved absolute stereochemistry, but it did so by an
Acknowledgment. We gratefully acknowledge the financial
support of the National Institutes of Health (D.L.B. CA 42056
and T.Y. GM 33712) and the Skaggs Institute for Chemical
Biology. M.J.S. is a Skaggs Fellow. F.A.R. acknowledges the
American Cancer Society for a postdoctoral fellowship.
Supporting Information Available: Full experimental details.
This material is available free of charge via the Internet at
http://pubs.acs.org.
JA0632862
9
J. AM. CHEM. SOC. VOL. 128, NO. 36, 2006 11807