Total syntheses of (+)-zampanolide and (+)-dactylolide exploiting a unified strategy

Article abstract of DOI:10.1021/ja020635t

The first total syntheses of (+)-zampanolide (1) and (+)-dactylolide (2), members of a new class of tumor cell growth inhibitory macrolides, have been achieved. Key features of the unified synthetic scheme included the stereocontrolled construction of the cis-2,6-disubstituted tetrahydropyran via a modified Petasis-Ferrier rearrangement, a highly convergent assembly of the macrocyclic domain, and, in the case of zampanolide, a Curtius rearrangement/acylation tactic to install the N-acyl hemiaminal. The complete relative and absolute stereochemistries for both (+)-zampanolide and (+)-dactylolide were also assigned, albeit tentatively in the case of (+)-zampanolide (1).

Full text of DOI:10.1021/ja020635t

Published on Web 08/23/2002
Total Syntheses of (+)-Zampanolide and (+)-Dactylolide
Exploiting a Unified Strategy
Amos B. Smith, III,* Igor G. Safonov, and R. Michael Corbett
Contribution from the Department of Chemistry, Monell Chemical Senses Center, and
Laboratory for Research on the Structure of Matter, UniVersity of PennsylVania,
Philadelphia, PennsylVania 19104
Received May 3, 2002
Abstract: The first total syntheses of (+)-zampanolide (1) and (+)-dactylolide (2), members of a new class
of tumor cell growth inhibitory macrolides, have been achieved. Key features of the unified synthetic scheme
included the stereocontrolled construction of the cis-2,6-disubstituted tetrahydropyran via a modified Petasis-
Ferrier rearrangement, a highly convergent assembly of the macrocyclic domain, and, in the case of
zampanolide, a Curtius rearrangement/acylation tactic to install the N-acyl hemiaminal. The complete relative
and absolute stereochemistries for both (+)-zampanolide and (+)-dactylolide were also assigned, albeit
tentatively in the case of (+)-zampanolide (1).
In 1996 Tanaka and Higa disclosed the isolation, partial
structure elucidation, and biological activity of the architecturally
novel macrolide (-)-zampanolide (1), obtained from Fascio-
spongia rimosa,1 an Okinawan sponge that previously yielded
significant quantities of (+)-latrunculin A and (-)-laulimalide,2
as well as the related, but new metabolites (+)-latrunculin S
and (-)-neolaulimalide.3 The structure of (-)-zampanolide (1),
including the relative stereochemistry between C(11), C(15),
and C(19) stereogenic centers, was secured via extensive NMR
analysis. The stereochemical disposition of the C(20) hydroxyl
group and the absolute stereochemistry, however, were not
defined.
upenamide.8 Herein we describe, in full, an account of the design
and execution of the first total synthesis and tentative stereo-
chemical assignment of the non-naturally-occurring antipode,
(+)-zampanolide (1). In addition, we will present the total
synthesis and structure assignment of the related macrolide (+)-
dactylolide (2) exploiting a unified synthetic strategy.9
The extreme scarcity of (-)-zampanolide and the impressive
cytotoxicity displayed against the P388, HT29, A549, and
MEL28 cell lines (IC50 1-5 ng/mL), in conjunction with our
longstanding interest in the construction of complex marine
natural products, prompted us to launch a synthetic venture
targeting this metabolite.4 From the synthetic perspective, the
challenging features of 1 include the highly unsaturated 20-
membered macrolactone incorporating a cis-2,6-disubstituted
tetrahydropyran and the unusual N-acyl hemiaminal side chain,
a structural feature associated with several highly bioactive
marine metabolites, such as the echinocandins B and C,5 the
mycalamides A and B,6 spergualin,7 and the recently reported
Synthetic Analysis of Zampanolide. Our synthetic strategy
for zampanolide (1) called for initial construction of the
macrocyclic domain, with installation of the N-acyl hemiaminal
structural unit late in the synthetic sequence, due to the
anticipated fragile nature of this reactive moiety. Since neither
the relative stereochemistry at C(20) nor the absolute config-
uration of zampanolide (1) had been defined, we arbitrarily set
1 as our initial target (Scheme 1). Disconnection at the amide
linkage, followed by application of a Curtius rearrangement
transform,10 suggested an activated 2(Z),4(E)-hexadienoic acid
(fragment D) and R-alkoxy acid 3, wherein the C(7) ketone
* Address correspondence to this author. E-mail: smithab@sas.upenn.edu.
(1) Tanaka, J.; Higa, T. Tetrahedron Lett. 1996, 37, 5535.
(2) Jefford, C. W.; Bernardinelli, G.; Tanaka, J.; Higa, T. Tetrahedron Lett.
1996, 37, 159.
(3) Tanaka, J.; Higa, T.; Bernardinelli, G.; Jefford, C. W. Chem. Lett. 1996,
255.
(4) Smith, A. B., III; Safonov, I. G.; Corbett, R. M. J. Am. Chem. Soc. 2001,
123, 12426.
(5) (a) Benz, F.; Knu¨sel, F.; Nu¨esch, J.; Treichler, H.; Voser, W.; Nyfeler, R.;
Keller-Schierlein, W. HelV. Chim. Acta 1974, 57, 2459. (b) Traber, R.;
Keller-Jusle´n, C.; Loosli, H.-R.; Kuhn, M.; von Wartburg, A. HelV. Chim.
Acta 1979, 62, 1252.
(7) Isolation: Takeuchi, T.; Iinuma, H.; Kunimoto, S.; Masada, T.; Ishizuka,
M.; Takeuchi, M.; Hamada, M.; Naganawa, H.; Kondo, S.; Umezawa, H.
J. Antibiot. 1981, 34, 1619. Structure: Umezawa, H.; Kondo, S.; Iinuma,
H.; Kunimoto, S.; Ikeda, Y.; Iwazawa, H.; Ikeda, D.; Takeuchi, T. J.
Antibiot. 1981, 34, 1622.
(8) Jime´nez, J. I.; Goetz, G.; Mau, C. M. S.; Yoshida, W. Y.; Scheuer, P. J.;
Williamson, R. T.; Kelly, M. J. Org. Chem. 2000, 65, 8465.
(6) (a) Perry, N. B.; Blunt, J. W.; Munro, M. H. G.; Pannell, L. K.. J. Am.
Chem. Soc. 1988, 110, 4850. (b) Perry, N. B.; Blunt, J. W.; Munro, M. H.
G.; Thompson, A. M. J. Org. Chem. 1990, 55, 223.
(9) Smith, A. B., III; Safonov, I. G. Org. Lett. 2002, 4, 635.
(10) (a) Roush, W. R.; Marron, T. G. Tetrahedron Lett. 1993, 34, 5421. (b)
Hoffmann, R. W.; Schlapbach, A. Tetrahedron Lett. 1993, 34, 7903.
9
11102
J. AM. CHEM. SOC. 2002, 124, 11102-11113
10.1021/ja020635t CCC: $22.00 © 2002 American Chemical Society

Total Syntheses of (+)-Zampanolide and (+)-Dactylolide
A R T I C L E S
Scheme 1
Scheme 2
requisite olefin geometry needed for zampanolide. Thus,
DIBAl-H reduction of (-)-6 and (+)-7, either separately or
as the mixture, gave diol (+)-8 as a single product in near
quantitative yield. Exhaustive silylation (TBSOTf, 2,6-lutidine)
then proceeded smoothly, to provide the fully protected triol
(-)-9 in 88% yield. Reductive removal of the benzyl group
with lithium di-tert-butylbiphenylide (LDBB),17 followed by
Swern oxidation18 of the liberated hydroxyl [e.g., (-)-10],
completed construction of aldehyde (+)-A in 88% yield for the
final two steps. Fragment (+)-A was thus available in five steps
and 58% overall yield from (-)-5.
Construction of Sulfone B: The C(9-17) Fragment. To
assemble sulfone B we elected to utilize the Petasis-Ferrier
rearrangement,19 recently established in our laboratory as a
powerful, stereocontrolled entry to cis-2,6-disubstituted tetrahy-
dropyrans.20 Success here would represent the first example of
an R,â-unsaturated oxonium intermediate (e.g., 22a; Scheme
4) in the Petasis-Ferrier rearrangement. Toward this end, Brown
asymmetric allylation21 of known aldehyde 1122 installed the
C(11) stereogenic center (zampanolide numbering), both in high
yield and with excellent enantiomeric excess, the latter deter-
mined by Mosher ester analysis (Scheme 3).23 Protection of the
derived homoallylic alcohol (-)-12 (TESCl, imidazole) and
ozonolysis of the terminal olefin delivered aldehyde (+)-14 in
70% yield (two steps). Oxidation (buffered NaClO2)24 and
would be masked as a TBS ether to forestall potential complica-
tions associated with the acidity of the C(6) position. Continuing
with this analysis, three strategic disconnections of the macrolide
ring led to fragments A, B, C, and the commercially available
diethylphosphonoacetic acid (4). In the synthetic direction, we
envisioned macrolide construction via a Kocienski-Julia ole-
fination11 of aldehyde A with sulfone B, followed in turn by
nucleophilic opening of epoxide C with a mixed cyano-Gilman
cuprate,12 derived from vinyl bromide AB, incorporation of an
acyl phosphonate via Mitsunobu inversion at C(19),13 and a
Horner-Wadsworth-Emmons (HWE) macrocyclization.14 This
highly convergent design would obviate the necessity to protect
and then unmask the C(19) hydroxyl, a measure unavoidable
involving approaches to the macrocycle via Mitsunobu macro-
lactonization.
Preparation of Aldehyde A: The C(3-8) Fragment. Our
point of departure for construction of A was the known
alkynoate (-)-5, prepared via the procedure of Ogasawara et
al.15 Michael addition of lithium dimethylcuprate (Me2CuLi)16
to the acetylenic ester furnished the expected enoate (-)-6 and
the unusual macrodiolide (+)-7 in 78% and 8% yield, respec-
tively (Scheme 2). Fortuitously, (+)-7 also possessed the
(17) (a) Freeman, P. K.; Hutchinson, L. L. J. Org. Chem. 1980, 45, 1924. (b)
Ireland, R. E.; Noreck, D. W.; Mandell, G. S. J. Am. Chem. Soc. 1985,
107, 3285.
(18) Mancuso, A. J.; Swern, D. Synthesis 1981, 165.
(19) Petasis, N. A.; Lu, S.-P. Tetrahedron Lett. 1996, 37, 141.
(20) (a) Smith, A. B., III; Verhoest, P. V.; Minbiole, K. P.; Lim, J. J. Org. Lett.
1999, 1, 909. (b) Smith, A. B., III; Minbiole, K. P.; Verhoest, P. V.;
Beauchamp, T. J. Org. Lett. 1999, 1, 913. (c) Smith, A. B., III; Verhoest,
P. V.; Minbiole, K. P.; Schelhaas, M. J. Am. Chem. Soc. 2001, 123, 4834.
(d) Smith, A. B., III; Verhoest, P. V.; Minbiole, K. P.; Schelhaas, M. J.
Am. Chem. Soc. 2001, 123, 10942.
(11) Blakemore, P. R.; Cole, W. J.; Kocienski, P. J.; Morley, A. Synlett 1998,
26.
(12) (a) Bartlett, P. A.; Meadows, J. D.; Ottow, E. J. Am. Chem. Soc. 1984,
106, 5304. (b) Lipshutz, B. H.; Kozlowski, J. A.; Parker, D. A.; Nguyen,
S. L.; McCarthy, K. E. J. Organomet. Chem. 1985, 285, 437.
(13) Mitsunobu, O. Synthesis 1981, 1.
(14) (a) Masamune, S.; Bates, G. S.; Corcoran, J. W. Angew. Chem. 1977, 89,
608. (b) Burry, K. F.; Cardone, R. A.; Chen, W. Y.; Rosen, P. J. Am. Chem.
Soc. 1978, 100, 7069.
(15) Oizumi, M.; Takahashi, M.; Ogasawara, K. Synlett 1997, 1111.
(16) Corey, E. J.; Katzenellenbogen, J. J. Am. Chem. Soc. 1969, 91, 1851.
(21) (a) Brown, H. C.; Jadhav, P. K. J. Am. Chem. Soc. 1983, 105, 2092. (b)
Brown, H. C.; Ramachandran, P. V. Pure Appl. Chem. 1991, 63, 307.
(22) Boeckman, R. K., Jr.; Charette, A. B.; Asberom, T.; Johnston, B. H. J.
Am. Chem. Soc. 1987, 109, 7553.
(23) (a) Dale, J. A.; Mosher, H. S. J. Am. Chem. Soc. 1973, 95, 512. (b) Sullivan,
G. R.; Dale, J. A.; Mosher, H. S. J. Org. Chem. 1973, 38, 2143. (c) Ohtani,
I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am. Chem. Soc. 1991, 113,
4092.
(24) Kraus, G. A.; Taschner, M. J. J. Org. Chem. 1980, 45, 1175.
9
J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002 11103

A R T I C L E S
Smith et al.
Scheme 3
actions with the C(11) axial hydrogen. Initial difficulties in the
scale-up of this reaction suggested that triflic acid (TfOH) is
the actual catalyst. Adventitious water, more pronounced on
small scale, may generate TfOH in situ from TMSOTf, as well
as TMS2O. Large-scale reactions did not proceed until a catalytic
amount (5-15 mol %) of TfOH was added (see the Supporting
Information for details). Methylenation of (+)-19 with the easily
prepared and convenient to handle Petasis-Tebbe reagent (Cp2-
TiMe2)28 furnished a difficultly separable mixture of enol ethers
(+)-20 [6:1 cis/trans at C(15); 72% yield].29 The stage was thus
set for the Petasis-Ferrier rearrangement,19 the strategic trans-
formation required for construction of sulfone B. In the event,
treatment of the mixture with 1 equiv of dimethylaluminum
chloride (Me2AlCl)20 at -78 °C effected the rearrangement to
provide the desired cis-pyranone (+)-23 in 59% isolated yield
after chromatographic separation from the trans isomer (+)-24
(12%).
Completion of sulfone B was then achieved in a straightfor-
ward fashion (Scheme 5); installation of the exo-methylene
exploiting a standard Wittig reaction was followed by removal
of the BPS group (HF, MeCN) to afford alcohol (-)-26 in 95%
yield for the two steps. Incorporation of the thiotetrazole via
the Mitsunobu protocol,13 employing commercially available
1-phenyl-1H-tetrazole-5-thiol, followed by oxidation [H2O2,
(NH4)6Mo7O24‚4H2O]30 of the derived sulfide (+)-27, proceeded
smoothly to furnish sulfone (-)-B (66% yield, two steps).
Subtarget (-)-B was thus available in 13 steps and in 12%
overall yield from aldehyde 11.
removal of the TES group (aqueous HCl) then furnished the
â-hydroxy acid (-)-15, which upon bis-silylation25 (HMDS)
provided (+)-16, the last acyclic intermediate required for
sulfone B; the overall yield from (+)-14 was 84% (three steps).
Condensation of (+)-16 with 2(E)-3-bromobut-2-enal (17),26
employing TMSOTf,27 furnished an inseparable mixture of the
corresponding dioxanones [10:1 at C(15)] in 82% combined
yield, with the cis congener (+)-19 as the major product
(Scheme 4). Presumably, this transformation proceeds via
Scheme 4
Scheme 5
Synthesis of Epoxide C: The C(18-20) Fragment. The
two stereogenic centers in fragment C were envisioned to arise
from the known diethyl ketal (-)-28 (Scheme 6), conveniently
prepared in four steps (50% overall yield) from (+)-dimethyl
tartrate according to the protocols of Somfai and Yonemitsu.31
transition state 18, wherein the aldehyde side chain adopts a
pseudoequatorial orientation to avoid unfavorable steric inter-
(27) (a) Noyori, R.; Murata, S.; Suzuki, M. Tetrahedron 1981, 37, 3899. (b)
Seebach, D.; Imwinkelried, R.; Stucky, G. HelV. Chim. Acta 1987, 70, 448.
(28) Petasis, N. A.; Bzowej, E. I. J. Am. Chem. Soc. 1990, 112, 6392.
(29) The observed decay in the cis/trans ratio for this transformation was partially
attributed to the harsh reaction conditions.
(30) Schultz, H. S.; Freyermuth, H. B.; Buc, S. R. J. Org. Chem. 1963, 28,
1140.
(25) Harada, T.; Yoshida, T.; Kagamihara, Y.; Oku, A. J. Chem. Soc., Chem.
Commun. 1993, 1367.
(26) Prepared by oxidation of 2(E)-3-bromobut-2-enol with PCC in 79% yield.
For preparation of the latter, see: Corey, E. J.; Bock, M. G.; Kozikowski,
A. P.; Rama Rao, A. V.; Floyd, D.; Lipshutz, B. Tetrahedron Lett. 1978,
19, 1051.
9
11104 J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002

Total Syntheses of (+)-Zampanolide and (+)-Dactylolide
A R T I C L E S
Scheme 6
sium bis(trimethylsilyl)amide (KHMDS) at -78 °C, followed
by addition of aldehyde (+)-A and warming to ambient
temperature. Vinyl bromide (-)-AB was reproducibly obtained
as the sole C(8,9) olefinic isomer, in 88% yield. The E geometry
was confirmed by the large NMR coupling constant (JH(8-9)
)
15.3 Hz) for the olefinic hydrogens.
Union of Fragments (-)-AB and (-)-C: A Difficult
Maneuver. We next turned our attention to addition of the
mixed cyano-Gilman cuprate (31), derived from vinyl bromide
(-)-AB and lithium 2-thienylcyanocuprate,36 to epoxide (-)-C
(Scheme 8). Due to the high reactivity of alkenyl groups and
the low transferability of the 2-thienyl group, cuprates of this
type are extremely useful in organic synthesis, especially vis-
a`-vis reactions at sp3 centers.12,37
Scheme 8
Protection of the primary hydroxyl in (-)-28 as the 3,4-
dimethoxybenzyl (DMB) ether was then followed by proteolytic
removal of the diethyl ketal to reveal diol (+)-30 (90% yield,
two steps). The best conditions to effect epoxide ring formation
proved to be the modified Mitsunobu procedure of Abushanab
et al.,32 which effectively furnished epoxide (-)-C, with
retention of the C(19) stereogenicity.33 Fragment (-)-C was
thus prepared in three steps and in 79% overall yield from
diethyl ketal (-)-28.
Union of Fragments (+)-A and (-)-B. With multigram
quantities of A and B in hand, we began assembly of the
macrocyclic ring. The first step, the one-pot Kocienski-modi-
fied11 Julia olefination34 of aldehyde (+)-A with sulfone (-)-
B, forged the trans C(8,9) double bond (Scheme 7). This tactic,
Scheme 7
Initially, modest yields (16-38%) of the (+)-ABC adduct
were obtained, leading us to conduct a detailed analysis of the
coupling process. We first established that the initial lithium-
halogen exchange occurs rapidly, even at -100 °C, and that
the derived vinyllithium is stable at 0 °C for at least 15 min, as
evidenced by reaction with benzaldehyde (81% yield). This
study supported our conjecture that the coupling reaction
proceeds poorly at the cuprate stage. Earlier observations by
Bertz et al. indicated that THF can lead to a decrease in cuprate
reactivity.38 Performing the reaction in diethyl ether (Et2O),
however, had little or no effect on the yield of (+)-ABC.
Eventually, identification of a series of byproducts 33-35
(Scheme 9), in addition to the expected side product 32 resulting
from protonation of either the derived vinyllithium or cuprate
31, suggested that the cuprate was extremely sensitive to
adventitious oxygen.
particularly attractive since the reaction leads to high trans
selectivity in the absence of such factors as R-branching or
conjugation,35 involved treatment of sulfone (-)-B with potas-
Rigorous elimination of oxygen from the argon atmosphere
via use of an OXICLEAR argon filter proved particularly
S. L.; Rojas, C. M. Angew. Chem., Int. Ed. 2000, 39, 4612. (d) Smith, A.
B., III; Brandt, B. M. Org. Lett. 2001, 3, 1685. (e) Mulzer, J.; O¨ hler, E.
Angew. Chem., Int. Ed. 2001, 40, 3842. (f) Liu, P.; Jacobsen, E. N. J. Am.
Chem. Soc. 2001, 123, 10772. (g) Lee, E.; Song, H. Y.; Kang, J. W.; Kim,
D.-S.; Jung, C.-K.; Joo, J. M. J. Am. Chem. Soc. 2002, 124, 384.
(36) Lithium 2-thienylcyanocuprate was either purchased from Aldrich Chemi-
cals or freshly prepared from commercially available 2-thienyllithium and
CuCN according to the procedure of Lipshutz (see ref 12b).
(37) (a) Smith, A. B., III; Friestad, G. K.; Duan, J. J.-W.; Barbosa, J.; Hull, K.
G.; Iwashima, M.; Qiu, Y.; Spoors, P. G.; Bertounesque, E.; Salvatore, B.
A. J. Org. Chem. 1998, 63, 7596. (b) Smith, A. B., III; Friestad, G. K.;
Barbosa, J.; Bertounesque, E.; Hull, K. G.; Iwashima, M.; Qiu, Y.; Salvatore,
B. A.; Spoors, P. G.; Duan, J. J.-W. J. Am. Chem. Soc. 1999, 121, 10468.
(38) Bertz, S. H.; Eriksson, M.; Miao, G.; Snyder, J. P. J. Am. Chem. Soc. 1996,
118, 10906.
(31) (a) Somfai, P.; Olsson, R. Tetrahedron 1993, 49, 6645. (b) Horita, K.;
Sakurai, Y.; Hachiya, S.; Nagasawa, M.; Yonemitsu, O. Chem. Pharm.
Bull. 1994, 683.
(32) Cichy, A. F.; Saibaba, R.; El Subbagh, H. I.; Panzica, R. P.; Abushanab,
E. J. Org. Chem. 1991, 56, 4653.
(33) Interestingly, application of N-tosyl imidazole for epoxidation of (+)-96
resulted in a 7:1 mixture of (-)-C and the C(19) epimer.
(34) (a) Baudin, J. B.; Hareau, G.; Julia, S. A.; Ruel, O. Tetrahedron Lett. 1991,
32, 1175. (b) Baudin, J. B.; Hareau, G.; Julia, S. A.; Lorne, R.; Ruel, O.
Bull. Soc. Chim. Fr. 1993, 130, 856.
(35) For recent applications of the Kocienski-Julia olefination in total synthesis
of complex natural products, see: (a) Williams, D. R.; Brooks, D. A.;
Berliner, M. A. J. Am. Chem. Soc. 1999, 121, 4924. (b) Smith, A. B., III;
Wan, Z. Org. Lett. 1999, 1, 1491. (c) Williams, D. R.; Cortez, G. S.; Bogen,
9
J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002 11105

A R T I C L E S
Smith et al.
Scheme 9
quantities of the (+)-ABC fragment in hand, the next step in
the proposed synthetic sequence called for introduction of an
acyl phosphonate at the C(19) secondary hydroxyl with inversion
of configuration, employing the commercially available dieth-
ylphosphonoacetic acid (4).13 The standard protocol for the
Mitsunobu reaction calls for premixing the alcohol and acid,13
followed by addition of triphenylphosphine (PPh3) and diethyl-
azodicarboxylate (DEAD). Surprisingly, incomplete consump-
tion of (+)-ABC was observed even after prolonged stirring of
this mixture at ambient temperature (12 h). Neither additional
quantities of DEAD and PPh3 nor reversing the order of reagent
addition produced an effect on the reaction outcome. Heating
the reaction mixture to 40-50 °C also proved unproductive,
producing instead almost exclusively a side product (38), which
results by elimination of the activated hydroxyl in the presumed
(+)-ABC-PPh3 complex 37 (Scheme 11).
Scheme 11
effective, not only reducing greatly byproduct formation (33-
35) but also increasing the yield of the coupled (+)-ABC
alcohol. A new byproduct 36, involving ring opening of the
tetrahydropyran ring, however, began to emerge (Scheme 9).
Use of 2 equiv of t-BuLi for the lithium-halogen exchange
led to the desired adduct (+)-ABC in 40-50% yield, with yields
of 36 ranging from 6% to 27%. Best results, however, were
obtained upon reducing the stoichiometry of t-BuLi to 1.7 equiv;
the (+)-ABC fragment was thus produced in 69-72% yield,
with only 3% of 36 (Scheme 10). The isolated yield (ca. 13-
19%) of 32 remained rather constant in all cases.
Scheme 10
Saturation of the reaction medium with a combination of 4
(15 equiv), PPh3 (6.5 equiv), and DEAD (7 equiv) at ambient
temperature eventually resulted in complete consumption of the
starting material. The reaction now occurred in less than 3 h.
To our chagrin, however, esterification of (+)-ABC had resulted
in acyl phosphonate (+)-39, with complete retention of the
C(19) stereogenicity (Scheme 12)! This stereochemical outcome,
unrecognized at the time, will be addressed in detail later. To
avoid confusion with what follows, the C(19) stereogenicity in
all relevant compounds [from (+)-39 to (-)-53] will be depicted
with their correct (uninverted) configurations.
Completion of the Macrocycle. Selective removal of the
TBS group at C(3) was next achieved with HF‚pyridine in good
yield (Scheme 13); oxidation of the derived allylic alcohol with
Dess-Martin periodinane39 then smoothly furnished aldehyde
(-)-40, substrate for the intramolecular Horner-Wadsworth-
Esterification of (+)-ABC with Diethylphosphonoacetic
Acid (4): An Anomalous Mitsunobu Reaction. With sufficient
9
11106 J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002

Total Syntheses of (+)-Zampanolide and (+)-Dactylolide
A R T I C L E S
Scheme 12
we turned to elaboration of the C(20) N-acyl hemiaminal,
followed by introduction of the unsaturated side chain. Selective
removal of the primary DMB protecting group (DDQ)42 in the
presence of the C(20) PMB ether readily afforded primary
alcohol (-)-42, albeit in a modest yield of 67% (Scheme 14).
The major problem associated with this reaction was incomplete
consumption of the staring material, (+)-41. Addition of more
than 1 equiv of DDQ not only failed to force the reaction to
completion but instead resulted in the generation of two
1
prominent side products. Identified by H NMR, one was the
expected C(20) alcohol 43, resulting from competitive removal
of the PMB group; the other was the p-methoxybenzylidine
acetal 44, stemming presumably from intramolecular closure
of the liberated primary hydroxyl onto the quinone methide
intermediate derived from the activated PMB ether. In retrospect,
the outcome of this transformation reveals the conceptual
shortcoming of positioning two protecting groups susceptible
to oxidative cleavage on vicinal hydroxyl groups.
Emmons reaction.14 At first, closure of the macrolactone under
Masamune-Roush conditions (LiCl, DBU, MeCN)40 led to the
desired 20-membered macrolide [(+)-41], albeit in low yield,
with little or no recovery of the starting material. However,
treatment of a dilute solution (0.006 M) of (-)-40 in THF with
1 equiv of sodium bis(trimethylsilyl)amide at -78 °C, followed
by warming of the reaction mixture to 0 °C, furnished the
desired macrolide (+)-41 as a single olefinic isomer in 66%
yield.41
Scheme 14
Scheme 13
Conversion of alcohol (-)-42 to the corresponding R-alkoxy
acid (-)-3 (Scheme 15) was next achieved in good yield via a
two-step oxidation (Dess-Martin,39 followed by buffered Na-
ClO224). We were thus ready to install the N-acyl hemiaminal
moiety via the proposed Curtius rearrangement.10 Initially we
explored the one-pot protocol of Yamada,43 utilizing diphen-
ylphosphoryl azide (DPPA). Unfortunately, this method proved
unsatisfactory, leading to significant decomposition of the
starting material. Attention was therefore shifted to the stepwise
procedure of Weinstock.44 Treatment of (-)-3 in turn with
excess Hu¨nig’s base, isobutyl chloroformate, and aqueous
sodium azide, followed by thermal rearrangement45 and capture
of the isocyanate with 2-(trimethylsilyl)ethanol,46 provided the
desired R-alkoxy Teoc-carbamate (-)-45 in 66% yield, with
complete retention (vide infra) of the C(20) stereogenicity.10
Fragment D: 2(Z),4(E)-Hexadienoic Acid Chloride. The
synthesis of fragment D, coupling partner for carbamate (-)-
(42) Horita, K.; Yoshioka, T.; Tanaka, T.; Oikawa, Y.; Yonemitsu, O.
Tetrahedron 1986, 42, 3021.
Installation of the N-Acyl Hemiaminal. With the assumption
that the macrocyclic ring of zampanolide (1) was now in hand,
(43) Shiori, T.; Ninomiya, K.; Yamada, S. J. Am. Chem. Soc. 1972, 94, 6203.
(44) Weinstock, J. J. Org. Chem. 1961, 26, 3511. Also see: Overman, L. E.;
Taylor, G. F.; Petty, C. B.; Jessup, P. J. J. Org. Chem. 1978, 43, 2164.
(45) This process was conveniently monitored by FT IR spectroscopy. The azide
absorption band (ca. 2140 cm-1) was no longer detectable after 15 min of
(39) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155.
(40) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.; Masamune,
S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
(41) Smith, A. B., III; Rano, T. A.; Chida, N.; Sulikowski, G. A.; Wood, J. L.
J. Am. Chem. Soc. 1992, 114, 8008.
refluxing. The corresponding isocyanate absorbed at ca. 2250 cm-1
.
(46) The 2250 cm-1 isocyanate absorption band disappeared 5 h after addition
of 2-(trimethylsilyl)ethanol.
9
J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002 11107

A R T I C L E S
Smith et al.
Scheme 15
Scheme 17
45, entailed three steps. Doebner condensation47 of malonic acid
with crotonaldehyde furnished the corresponding unsaturated
diacid 46, which was subjected to decarboxylation (quinoline,
130 °C) to afford 2(Z),4(E)-hexadienoic acid (47) as the sole
isomer (Scheme 16).48 Conversion to acid chloride D was then
achieved with n-butyllithium (THF) and oxalyl chloride.49 This
procedure obviated the necessity to isolate the sensitive and
volatile fragment D, since the byproducts [n-butane (g), CO2
(g), CO (g), and LiCl (solubilized in THF)] would not be
expected to interfere with the proposed carbamate acylation.
Best results (59%) were obtained by deprotonation of (-)-45
with 1 equiv of sodium bis(trimethysilyl)amide at -78 °C,
followed by careful addition of acid chloride D in THF.49
Attempted removal of the trimethylethoxycarbonyl (Teoc) and
TBS protecting groups in one step with tetrabutylammonium
fluoride (TBAF) afforded alcohol (-)-50, albeit in modest 42%
yield, with substantial amounts of decomposition, ostensibly due
to the basicity of TBAF. Exploiting the lower reactivity of the
C(7) allylic TBS ether to TBAF at 0 °C, sequential desilylation
via the intermediary amide 49 furnished (-)-50 in 74% for the
two steps. Oxidation with Dess-Martin periodinane39 then
provided the penultimate ketone (+)-51.
Scheme 16
Syntheses of C(19)-epi-Zampanolide and C(19,20)-Bis-epi-
zampanolide. Having presumably successfully arrived at the
macrocyclic framework of zampanolide, possessing the protected
N-acyl hemiaminal moiety, we focused on completion of the
synthesis. Introduction of the unsaturated amide side chain was
achieved in 59% yield via acylation of carbamate (-)-45 with
fragment D (Scheme 17). On small scale, this transformation
proved difficult, either failing to proceed in some cases or
affording low yields, presumably due to adventitious water.50
Removal of the PMB group in (+)-51 under oxidative
conditions also proved challenging. Initial treatment with DDQ
or cerium ammonium nitrate in dichloromethane failed either
at 0 °C or at ambient temperature. Other mild conditions, such
as SnCl2/TMSCl/anisole51 and BCl3‚SMe2,52 led to major
decomposition. The radical anion (LDBB)17 also proved inef-
fective, even when used in large excess. Undaunted, we revisited
the initial DDQ conditions, eventually discovering that exposure
(50) In such cases, the starting material was recovered in essentially quantitative
yield, with the C(20) stereogenicity remaining intact, as determined by 1H
NMR analysis.
(51) Akiyama, T.; Shima, H.; Shoichiro, O. Synlett 1992, 415.
(52) Congreve, M. S.; Davison, E. C.; Fuhry, M. A. M.; Holmes, A. B.; Payne,
A. N.; Robinson, R. A.; Ward, S. E. Synlett 1993, 663.
(47) Johnson, J. R. Org. React. 1942, 1, 210.
(48) Crombie, L.; Crombie, W. M. L. J. Chem. Soc., Perkin Trans. 1 1994,
1267.
(49) (a) Roush, W. R.; Pfeifer, L. A. J. Org. Chem. 1998, 63, 2062. (b) Roush,
W. R.; Pfeifer, L. A.; Marron, T. G. J. Org. Chem. 1998, 63, 2064.
9
11108 J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002

Total Syntheses of (+)-Zampanolide and (+)-Dactylolide
A R T I C L E S
of ketone (+)-51 to 2 equiv of DDQ in wet CH2Cl2 at 50 °C
yielded a mixture of two polar compounds (1.8:1), readily
separable by HPLC, in a combined yield of 56%.
Alternatively, initial activation of (+)-ABC (i.e., 55) does
indeed occur (Scheme 20); however, due to the steric bulk
surrounding the activated C(19) stereocenter, a more facile
neighboring group participation may ensue, inverting the C(19)
stereocenter and generating a highly electrophilic intermediate
56, which in turn undergoes ring opening to furnish (+)-39 with
net retention of the stereochemistry at C(19). The high level of
regioselectivity required in this proposal, however, leaves this
explanation suspect.
1
Initial H NMR and HRMS analysis indicated removal of
the PMB group in both products, suggesting that the two
products were epimeric at C(20). However, detailed comparison
1
of the H NMR spectral data of the C(20) epimers with the
literature data for (-)-zampanolide (1) clearly demonstrated that
neither was zampanolide!53 Since no olefin migration and/or
isomerization appeared to have occurred, we surmised that the
lack of identity might have been due to the Mitsunobu reaction.
Assuming this to be the case, we tentatively assigned the two
products as (-)-52 and (-)-53, both possessing the undesired
stereogenicity at C(19).
Scheme 20
The Mitsunobu Reaction Revisited. To secure the stereo-
chemical outcome of the Mitsunobu reaction, we subjected ester
(+)-39 to saponification (Scheme 18). The resulting product
1
proved identical (500 MHz H and 125 MHz 13C NMR) with
the (+)-ABC alcohol of known C(19) stereogenicity prepared
earlier. Thus, inversion at C(19) had not occurred.
Scheme 18
Construction of the Zampanolide Macrocycle: A Second
Generation Strategy (Scheme 21). With the stereochemical
outcome of the Mitsunobu reaction understood, all that would
be required to arrive at the correct zampanolide macrolide would
Scheme 21
Two rationales could explain this unexpected result. The first,
and the most likely scenario (Scheme 19), would entail failure
of the PPh3-DEAD complex to activate the C(19) hydroxyl
due to steric inaccessibility. Consequently, diethylphospho-
noacetic acid (4), present in large excess relative to (+)-ABC,
would consume the DEAD and PPh3; nucleophilic attack of the
C(19) hydroxyl on the activated 4-PPh3 complex could then
lead to (+)-39, the product with retention of C(19) configuration.
Scheme 19
9
J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002 11109

A R T I C L E S
Smith et al.
be access to the diastereomeric epoxide. We chose instead to
employ the more readily available antipode of the epoxide [i.e.,
(+)-C]. Use of (+)-C would lead to the zampanolide macrolide
possessing the â-C(20) relative configuration as our synthetic
target. Importantly, this change would have no consequence vis-
a`-vis our overall strategy for construction of the zampanolide
skeleton.
Toward this end, union of the mixed cyano-Gilman cuprate
(31), derived from vinyl bromide (-)-AB, with epoxide (+)-
C,54 led, as expected, to alcohol (-)-ABC′, now possessing the
desired relative C(19) stereogenicity (Scheme 21). Esterification
with diethylphosphonoacetic acid (4) under the Steglich condi-
tions (DCC, DMAP),55 followed by selective removal of the
C(3) silyl group and Dess-Martin oxidation,39 furnished alde-
hyde (-)-59, substrate for the HWE macrocyclization.14 Pleas-
ingly, macrocycle (+)-60 was produced in 72% yield.41
Total Syntheses of (+)-Zampanolide and (+)-C(20)-epi-
Zampanolide. With ample quantities of the fully elaborated
macrocycle (+)-60 in hand, now possessing the requisite relative
stereochemistry at C(19), we turned to complete the total
synthesis. Oxidative removal of the DMB protecting group42
in (+)-60, followed by a two-step oxidation39,24 of the derived
primary hydroxyl in (-)-61, produced (-)-62, substrate for
Curtius rearrangement (Scheme 22); as before, rearrangement
Iterative removal of the Teoc and TBS moieties, as before,
followed by oxidation of the liberated C(7) hydroxyl, provided
ketone (+)-67, a single compound in 75% overall yield for the
three steps (Scheme 23). Oxidative removal of the PMB
protecting group (DDQ, pH 7 buffer, CH2Cl2, ambient temper-
ature) next produced a mixture (1.3:1) of two polar compounds
epimeric at C(20). Best results (75% combined yield) were
obtained with freshly recrystallized DDQ. Importantly, after
HPLC separation, the major, less polar component possessed
spectral data identical in all respects to the literature data for
natural (-)-zampanolide (e.g., 500 MHz 1H and 125 MHz 13
C
NMR; 600 MHz COSY and HMQC; HRMS and IR),53 except
for chiroptic properties {natural 1, [R]D29 ) -101° (c 0.12, CH2-
25
Cl2); synthetic 1, [R]D ) +102.3° (c 0.09, CH2Cl2)}.4 The
structure of (+)-C(20)-epi-1 was secured via extensive spec-
troscopic analysis. It should be emphasized that at this point
the relative stereochemistry at C(20) in both (+)-1 and (+)-
C(20)-epi-1 remained unknown.
Scheme 23
Scheme 22
Comments on the Final Deprotection. Well aware of the
potential instability of the N-acyl hemiaminal functionality, we
had carefully selected the PMB protecting group, reasoning that
acidic or basic conditions might be incompatible with the N-acyl
hemiaminal group, as well as other functionality in the molecule.
Protecting groups susceptible to hydrogenolysis were also
excluded. The optimal choice thus appeared to be the p-
methoxybenzyl group, known to release upon DDQ oxidation
under essentially neutral conditions.56 Nonetheless, epimerization
at C(20) upon treatment with DDQ did occur presumably via
an electrophilic acyliminium ion (i.e., 69). Although the exact
origin of 69 is unclear, the most plausible pathway would entail
afforded the desired R-alkoxy carbamate (-)-63, both in a highly
stereospecific fashion and in good overall yield (75%). Acyla-
tion49 with acid chloride D then gave the Teoc-protected amide
(-)-64.
(53) We thank Professors Tanaka and Higa for providing copies of the 1H and
13C NMR, COSY, HMQC, and IR spectra for natural (-)-zampanolide
(1).
(54) Epoxide (+)-C was prepared in seven steps from commercially available
(-)-dimethyl tartrate via a synthetic sequence analogous to that shown in
Scheme 6; see the Supporting Information for details.
(55) Neises, B.; Steglich, W. Angew. Chem., Int. Ed. Engl. 1978, 17, 522.
(56) Oikawa, Y.; Yoshioka, T.; Yonemitsu, O. Tetrahedron Lett. 1982, 23, 885.
9
11110 J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002

Total Syntheses of (+)-Zampanolide and (+)-Dactylolide
A R T I C L E S
Scheme 26
either the facile elimination of p-anisaldehyde from the quinone
methide intermediate 68 (Scheme 24) or an acid-promoted
dehydration of the liberated N-acyl hemiaminal.
Scheme 24
To suppress the observed epimerization, we explored the use
of a variety of organic buffers. Addition of bases such as 2,6-
di-tert-butyl-4-methylpyridine did not prevent epimerization.
Switching to a MeCN/pH 7 buffer system resulted only in
recovery of starting material. Other conditions, such as cerium
ammonium nitrate in MeCN/H2O, were equally unsuccessful.
Attempts to derivatize the C(20) stereocenter for Mosher ester
analysis also met with failure.
Tentative Assignment of the C(20) Relative Configuration.
Unable to prevent erosion of the stereogenicity at C(20) upon
deprotection of (+)-67 and thereby assign with certainty the
relative stereochemistry, we reasoned that reprotection of (+)-1
and (+)-C(20)-epi-1 as the PMB ether, in conjunction with
spectroscopic correlation with (+)-67, possessing the known
C(20) stereogenicity, would permit assignment of the C(20)
configuration. After considerable experimentation, the PMB
protection protocol of Hanessian57 emerged as the method of
choice, albeit not without difficulty (Scheme 25).
C(20)-epi-1 congener afforded a 3.7:1 mixture of (+)-67 and
(+)-70; (+)-zampanolide (1), on the other hand, afforded (+)-
70 and (+)-67 in a 7.6:1 ratio. Additional evidence for the
structural assignments was provided by the various electronic
spectra. Both (+)-1 and (+)-70 possessed UV absorption
maxima (λmax) at 259.2 nm, whereas (+)-C(20)-epi-1 and (+)-
67 absorbed at 262.7 nm. Taken together, these results permit
the tentative assignment of the relative and absolute stereo-
chemistry of (+)-zampanolide (1) as 11R, 15R, 19R, and 20R.4
Total Synthesis of (+)-Dactylolide (2). Shortly after comple-
tion of the synthesis of (+)-1, we became aware of the isolation,
partial structure assignment, and biological activity of (+)-
dactylolide (2), a new cytotoxic natural product derived from a
marine sponge of the genus Dactylospongia, collected off the
coast of the Vanuatu islands (Scheme 27).58 In addition to (+)-
dactylolide (2), the sponge also yielded the known marine
metabolites (-)-mycothiazole, (-)-isolaulimalide, (+)-latrun-
culin A, and (-)-laulimalide. Interestingly, the latter two
metabolites were also isolated from the Okinawan sponge F.
rimosa, the source of (-)-zampanolide (1).1
Scheme 25
The structure of 2 comprises the macrocyclic domain of
zampanolide (1), with cis relative stereochemistry between the
C(11) and C(15) stereogenic centers, and undefined, but
configurationally stable stereogenicity at C(19);59 the ab-
solute stereochemistry was not assigned.58 Interestingly, (+)-
dactylolide (2) displayed modest tumor cell growth inhibitory
activity in a number of cell lines [i.e., 63% inhibition of L1210
(lymphatic leukemia of mice) and 40% inhibition of SK-OV-3
(carcinoma of the ovaries) tumor cell lines at 3.2 µg/mL].58 The
mode of action, however, was not reported.
Via this tactic, protection of (+)-C(20)-epi-1 afforded a
mixture of products (ca. 1.5:1), the minor component displaying
1
spectroscopic properties (i.e., H NMR) identical to those of
ketone (+)-67 (Scheme 26). Extensive spectroscopic analysis
of the major product permitted assignment of structure (+)-70.
Presumably, the trifluoromethanesulfonic acid, generated in the
course of the reaction, promotes the epimerization at C(20).
Buffering the reaction medium with 2,6-di-tert-butyl-4-meth-
ylpyridine abated epimerization. Under these conditions, the (+)-
Since (+)-2 can be viewed as a simplified analogue of (-)-
1, parallel biological screening of both metabolites would
provide data on the significance of the N-acyl hemiaminal
moiety in zampanolide. With this goal, and with the desire to
(58) Cutignano, A.; Bruno, I.; Bifulco, G.; Casapullo, A.; Debitus C.; Gomez-
Paloma, L.; Riccio, R. Eur. J. Org. Chem. 2001, 775.
(59) Apparently, Riccio and co-workers were unaware of (-)-zampanolide (1).
(57) Hanessian, S.; Huynh, H. K. Tetrahedron Lett. 1999, 40, 671.
9
J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002 11111

A R T I C L E S
Smith et al.
Scheme 27
Scheme 28
define the complete stereostructure in mind, we embarked on
the total synthesis of (+)-dactylolide (2), exploiting the same
antithetic analysis as employed in the zampanolide synthesis
(Scheme 27). Given the close structural relationship, we
envisioned (+)-dactylolide to possess the same relative [i.e., at
C(19)] and absolute stereochemistries; thus, our synthetic target
was set at 2, with the clear expectation that the use of (-)-AB
from the synthesis of (+)-zampanolide (1) would, if successful,
lead to the unnatural enantiomer of 2.
Assembly of the (+)-Dactylolide Macrocycle. As with (+)-
zampanolide (1), union of the mixed cyano-Gilman cuprate
(31),12 derived from vinyl bromide (-)-AB, now with epoxide
(+)-72, the latter prepared in one step (NaH, DMBCl) from
commercially available (S)-(-)-glycidol, furnished alcohol (-)-
73 in an unoptimized yield of 40% (Scheme 28). Acylation55
with 4 next gave phosphonoester (-)-74 in near quantitative
yield. Selective desilylation (HF‚pyridine) at C(3), and Dess-
Martin oxidation39 then afforded aldehyde (-)-71 (57% yield,
three steps), which was subjected to Horner-Wadsworth-
Emmons macrocyclization14 [NaHMDS (1 equiv), -78 °C, with
subsequent warming to 0 °C] to produce (-)-76 in 72% yield.41
Completion of the (+)-dactylolide (2) synthesis was then
achieved in four steps; unmasking the C(7) hydroxyl (TBAF),
followed by Dess-Martin oxidation,39 furnished ketone (+)-
77 in 50% yield for the two steps (Scheme 29). Oxidative
removal of the DMB protecting group (DDQ)56 and oxidation
with Dess-Martin periodinane39 then afforded (+)-dactylolide
(2) in 69% yield for the two steps. The spectral and chiroptic
data for synthetic (+)-2 proved identical in all respects with
the spectral data derived from the natural product {e.g., 1H (500
and 600 MHz) and HSQC (600 MHz) NMR and HRMS; natural
2, [R]D25 ) +30° (c 1.0, MeOH); synthetic 2, [R]D25 ) +235°
(c 0.52, MeOH)}.60
Completion of the total synthesis of (+)-dactylolide (2)
provided an interesting observation: vinyl bromide (-)-AB
gives rise both to natural dactylolide [(+)-2] and to the non-
naturally-occurring antipode of zampanolide, (+)-1. Thus,
assuming that the published chiroptic data for the two natural
products are correct, the macrocyclic domain of natural zam-
panolide is enantiomeric with that of natural dactylolide.61
Although (-)-1 and (+)-2 were isolated from two geographi-
cally widely separated, taxonomically different sponge species,
the structural similarity of the two metabolites would seem to
imply that both arise from genetically related symbiotic
(60) We thank Professor Raffaele Riccio for the 600 MHz 1H and HSQC NMR
spectra of natural (+)-dactylolide (2). Interestingly, the 125 MHz 13C NMR
data for synthetic (+)-dactylolide revealed small shifts (∆δ ) (0.1-0.9
ppm) for most carbon signals relative to the corresponding signals reported,
and presumably derived from the HSQC NMR spectrum of natural (+)-
dactylolide; the 1H NMR (i.e., 500 and 600 MHz) spectra of the synthetic
and natural materials were completely superimposable.
(61) We have verified with Professor Riccio that the observed sign of the optical
rotation of natural dactylolide is plus. For examples of enantiomeric natural
products of plant origin, see: (a) Ashworth, D. M.; Robinson, J. A. J. Chem.
Soc., Chem. Commun. 1983, 1327. (b) Spavold, Z. M.; Robinson, J. A. J.
Chem. Soc., Chem. Commun. 1988, 4. (c) Zhang, Y.; Zheng, L.; Woo,
M.-H.; Gu, Z.-M.; Ye, Q.; Wu, F.-E.; McLaughlin, J. L. Heterocycles 1995,
41, 1743 and references cited therein. (d) Melching, S.; Bu¨low, N.; Wihstutz,
K.; Jung, S.; Ko¨nig, W. A. Phytochemistry 1997, 44, 1291. (e) Fukuyama,
Y.; Yokoyama, R.; Ohsaki, A.; Takahashi, H.; Minami, H. Chem. Pharm.
Bull. 1999, 47, 454 and references cited therein.
9
11112 J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002

Total Syntheses of (+)-Zampanolide and (+)-Dactylolide
A R T I C L E S
Scheme 29
Scheme 30
identical in all respects to those of (+)-2, prepared from vinyl
bromide (-)-AB. Similarly, exposure of a 1.3:1 mixture of (+)-1
and (+)-C(20)-epi-1 to the aforementioned thermal reaction
conditions afforded (+)-2 and 78 as the only products detectable
1
by H NMR. Interestingly, permitting the reaction mixture to
stand in deuterated chloroform either at room temperature or at
60 °C did not lead to re-formation of either (+)-zampanolide
1
(1) or (+)-C(20)-epi-1, as evidenced by H NMR.
Summary. The first total syntheses of (+)-zampanolide (1)
and (+)-dactylolide (2) have been achieved. Key features of
the unified synthetic scheme included the stereocontrolled
construction of the cis-2,6-disubstituted tetrahydropyran via a
modified Petasis-Ferrier rearrangement and a highly convergent
assembly of the macrocyclic domain. Installation of the N-acyl
hemiaminal in (+)-zampanolide (1) was then achieved via a
stereospecific Curtius rearrangement. The complete relative and
absolute stereochemistries for both (+)-zampanolide and (+)-
dactylolide were also assigned, albeit tentatively in the case of
(+)-zampanolide (1).
microorganisms. That is, (+)-dactylolide might be the biosyn-
thetic precursor of (+)-zampanolide, if the latter exists in nature,
or perhaps comprises a degradation product thereof.
Chemical Correlation of (+)-Zampanolide (1) with (+)-
Dactylolide (2). To achieve a chemical correlation between the
two natural products, we reasoned that exposure of (+)-
zampanolide (1) either to base or to thermolysis in an inert
solvent would lead to elimination of the side chain by cleavage
of the C(20)-N bond, presumably via a retro-ene-like reaction,
to afford (+)-dactylolide (2) and 2(Z),4(E)-hexadienoic acid
amide (78). Initially, all attempts to promote a base-induced
fragmentation of (+)-1 either with triethylamine or 1,8-
diazobicyclo[5.4.0]undec-7-ene (DBU) at room temperature or
with sodium bis(trimethylsilyl)amide at -78 °C met with failure
(Scheme 30). Heating (+)-zampanolide (1) in benzene at 85
°C for 105 min, however, cleanly furnished (+)-dactylolide (2)
Acknowledgment. Support was provided by the National
Institutes of Health (National Cancer Institute) through Grant
CA-19033 and a postdoctoral fellowship (CA-80337) to R.M.C.
We also thank Drs. George T. Furst and Rakesh K. Kohli of
the University of Pennsylvania Spectroscopic Service Center
for assistance in securing and interpreting high-field NMR
spectra and high-resolution mass spectra, respectively.
Supporting Information Available: Experimental procedures
and analytical data for all compounds (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
1
and amide 78, as evidenced by the H NMR of the reaction
mixture. Subsequent chromatographic purification furnished (+)-
dactylolide (2), which displayed spectral and physical data
JA020635T
9
J. AM. CHEM. SOC. VOL. 124, NO. 37, 2002 11113