Efficient synthesis of okadaic acid. 2. Synthesis of the C1-C14 domain and completion of the total synthesis

Article abstract of DOI:10.1021/ja973286p

Described here are the full details of the preparation of a synthetic intermediate representing carbons 1' 14 (C1-C14) of the marine natural product okadaic acid (1), the coupling of this fragment with the previously prepared C15-C38 domain, and the completion of an efficient total synthesis of 1. The C1-C14 intermediate was prepared in 11 steps and ~20% overall yield from a functionalized δ-valerolactone derivative representing C3-C8 of 1. This featured a classic spiroketalization strategy to construct the highly substituted 1,7-dioxaspiro-[5.5]undec-4-ene system, followed by incorporation of the intact C1-C2 α-hydroxyl, α-methyl carboxylate moiety using cis-(S)- lactate pivalidene enolate. The complete C1-C14 intermediate was converted into 1 in five additional steps. Coupling of the C1-C14 fragment with the C15-C38 domain of 1 via C14 aldehyde and C15 β-keto phosphonate moieties provided the complete carbon skeleton of 1 bearing a ketone at C16 and a mixed-methyl acetal at C19. Reduction of the C16 ketone using Corey's (S)- CBS/BH₃ system and subsequent acid-triggered spiroketalization formed the Central 1,6-dioxaspiro[4.5]decane ring system. Saponification of the C1-C2 pivalidene group and final reductive cleavage of the three benzyl ethers using lithium di-tert-butylbiphenylide in THF provided 1 in 48% yield from the C1-C14 aldehyde, and in 26 steps and ~2% overall yield in the longest linear sequence from the C22-C27 synthon methyl 3-O-benzyl-α-D- altropyranoside.

Full text of DOI:10.1021/ja973286p

2534
J. Am. Chem. Soc. 1998, 120, 2534-2542
Efficient Synthesis of Okadaic Acid. 2. Synthesis of the C1-C14
Domain and Completion of the Total Synthesis
Steven F. Sabes, Rebecca A. Urbanek, and Craig J. Forsyth*
Contribution from the Department of Chemistry, UniVersity of Minnesota, Minneapolis, Minnesota 55455
ReceiVed September 18, 1997
Abstract: Described here are the full details of the preparation of a synthetic intermediate representing carbons
1-14 (C1-C14) of the marine natural product okadaic acid (1), the coupling of this fragment with the previously
prepared C15-C38 domain, and the completion of an efficient total synthesis of 1. The C1-C14 intermediate
was prepared in 11 steps and ∼20% overall yield from a functionalized δ-valerolactone derivative representing
C3-C8 of 1. This featured a classic spiroketalization strategy to construct the highly substituted 1,7-dioxaspiro-
[5.5]undec-4-ene system, followed by incorporation of the intact C1-C2 R-hydroxyl, R-methyl carboxylate
moiety using cis-(S)-lactate pivalidene enolate. The complete C1-C14 intermediate was converted into 1 in
five additional steps. Coupling of the C1-C14 fragment with the C15-C38 domain of 1 via C14 aldehyde
and C15 â-keto phosphonate moieties provided the complete carbon skeleton of 1 bearing a ketone at C16 and
a mixed-methyl acetal at C19. Reduction of the C16 ketone using Corey’s (S)-CBS/BH3 system and subsequent
acid-triggered spiroketalization formed the central 1,6-dioxaspiro[4.5]decane ring system. Saponification of
the C1-C2 pivalidene group and final reductive cleavage of the three benzyl ethers using lithium di-tert-
butylbiphenylide in THF provided 1 in 48% yield from the C1-C14 aldehyde, and in 26 steps and ∼2%
overall yield in the longest linear sequence from the C22-C27 synthon methyl 3-O-benzyl-R-D-altropyranoside.
Introduction
diverse natural products potently inhibit PP1 and PP2A by
competitive binding to the okadaic acid receptor.15 Members
of this okadaic acid class of phosphatase inhibitors have been
the objects of several recent total syntheses. These include
calyculin A,16 tautomycin,17-19 microcystin-LA,20 and motu-
porin.21
As the archetypal member of this varied array of natural
products, okadaic acid, as well as its analogues hold consider-
able, if not unique, potential for delineating the precise structural
requirements for selective phosphatase binding and inhibition.
This is due to okadaic acid’s combination of dense function-
alization, well-defined molecular topology,22,23,14 and preferential
inhibition of PP2A (Ki ) 32 pM) over PP1 (Ki ) 147 ( 34
nM).24 Among the naturally occurring inhibitors of these
enzymes, okadaic acid displays the highest degree of differential
inhibition.15 As detailed structural information of the active
site and possible ligand binding domains of the serine/threonine
Okadaic acid (1) is a marine natural product with a rich
modern history. Originally isolated from the Pacific sponge
Halichondria okadai in the 1970s as a potential anticancer
agent,1 1 was subsequently identified as a diarrhetic toxin that
accumulates in shellfish,2,3 and a potent inhibitor of protein
serine/threonine phosphatases 1 and 2A (PP1 and PP2A,
respectively).4 Thus, 1 has emerged as both a contemporary
public health concern and a valuable tool in modern biomedical
research.5-7 Okadaic acid has also become the focus of renewed
synthetic interest8-11 since Isobe’s original total synthesis was
reported in 1986.12,13 This is due to its unique and challenging
polyether structure,14 as well as the potential of compounds
based upon the okadaic acid architecture to contribute further
to our understanding of the structural basis of protein phos-
phatase inhibition. A number of additional and structurally
(1) Scheuer, P. J. Nat. Prod. 1995, 58, 335.
(2) Murata, M.; Shimatani, M.; Sugitani, H.; Oshima, Y.; Yasumoto, T.
Bull. Jpn. Soc. Sci. Fish. 1982, 48, 549.
(3) Yasumoto, T.; Murata, M.; Oshima, Y.; Sano, M.; Matsumoto, G.
K.; Clardy, J. Tetrahedron 1985, 41, 1019.
(15) Takai, A.; Sasaki, K.; Nagai, H.; Mieskes, G.; Isobe, M.; Isono,
K.; Yasumoto, T. Biochem. J. 1995, 306, 657.
(16) Evans, D. A.; Gage, J., R.; Leighton, J. L. J. Am. Chem. Soc. 1992,
114, 9434.
(4) Bialojan, C.; Takai, A. Biochem. J. 1988, 256, 283.
(5) Fujiki, H. Mol. Carcinog. 1992, 5, 91.
(6) Cohen, P.; Holmes, C. F. B.; Tsukitani, Y. Trends Biochem. Sci. 1990,
15, 98.
(17) Oikawa, M.; Ueno, J.; Oikawa, H.; Ichihara, A. J. Org. Chem. 1995,
60, 5048.
(18) Jiang, Y.; Isobe, M. Tetrahedron 1996, 52, 2877.
(19) Sheppeck, J. E.; Liu, W.; Chamberlin, A. R. J. Org. Chem. 1997,
62, 387.
(7) Schonthal, A. New Biol. 1992, 4, 16.
(8) Dankwardt, J. W. Ph.D. Thesis, University of Rochester, 1991.
(9) Forsyth, C. J.; Sabes, S. F.; Urbanek, R. A. J. Am. Chem. Soc. 1997,
119, 8381.
(10) Marko, I. E.; Dobbs, A. P.; Scheirmann, V.; Chelle, F.; Bayston,
D. J. Tetrahedron Lett. 1997, 38, 2899.
(11) Marko, I. E.; Chelle, F. Tetrahedron Lett. 1997, 38, 2895.
(12) Isobe, M.; Ichikawa, Y.; Goto, T. Tetrahedron Lett. 1986, 27, 963.
(13) Isobe, M.; Ichikawa, Y.; Bai, D.-L.; Masaki, H.; Goto, T. Tetra-
hedron 1987, 43, 4767.
(14) Tachibana, K.; Scheuer, P. J.; Tsukitani, Y.; Kikuchi, H.; Engen,
D. V.; Clardy, J.; Gopichand, Y.; Schmitz, F. J. J. Am. Chem. Soc. 1981,
103, 2469.
(20) Humphrey, J. M.; Aggen, J. B.; Chamberlin, A. R. J. Am. Chem.
Soc. 1996, 118, 8, 11759.
(21) Schreiber, S. L.; Valentekovich, R. J. J. Am. Chem. Soc. 1995, 117,
9069.
(22) Norte, M.; Gonzalez, R.; Fernandez, J. J.; Rico, M. Tetrahedron
1991, 47, 7437.
(23) Matsumori, N.; Murata, M.; Tachibana, K. Tetrahedron 1995, 51,
12229.
(24) These values are for the inhibition of p-nitrophenol phosphate
phosphatase activity of PP2A isolated from rabbit skeletal muscle.15 We
have determined a similar Ki value (29 pM) for PP2A isolated from bovine
myocardial tissue: Frydrychowski, V. A.; Forsyth, C. J. Unpublished results.
S0002-7863(97)03286-1 CCC: $15.00 © 1998 American Chemical Society
Published on Web 03/06/1998

Synthesis of Okadaic Acid. 2
J. Am. Chem. Soc., Vol. 120, No. 11, 1998 2535
phosphatases becomes available,25,26 the relevance, utility, and
availability of ligands designed to probe and take advantage of
these features take on greater importance. In this context, we
have designed and recently executed an efficient and flexible
total synthesis of okadaic acid.9
A major goal of this work has been to develop a modular
and convergent approach to access the okadaic acid architecture
to allow for the facile generation of analogues incorporating
specific structural modifications. Using logical skeletal discon-
nections,12 1 was divided into three fragments, corresponding
to C1-C14, C16-C27, and C28-C38 of the natural product.
The full details of the convergent assembly of the C15-C38
domain (3) from the respective fragments was reported in a
previous paper.27 The preparation of the complementary C1-
C14 domain (4) and its utilization in conjunction with 3 for an
efficient total synthesis of okadaic acid are described here.
the C1 carboxylate to be incorporated in 4. The expected
product (E)-enone 5 would then serve as a substrate for
diastereoselective ketone reduction to provide the natural
product’s 16R-configuration. Because the C16 carbonyl of 5
would be in an acyclic region of the carbon skeleton and
insulated from the nearest stereogenic center by several carbons,
reagent-based control would be required. Subsequent intramo-
lecular ketalization under equilibrating conditions was then
expected to provide predominately the required 19R configu-
ration in the 1,6-dioxaspiro[4.5]decane ring system of 6.
Previous studies on simpler compounds indicated that the natural
product’s thermodynamically favored29 19R configuration would
predominate upon acid-induced ketalization.27,30 The chair-
chair conformation of the C19-C26 trans-dioxadecalin system
in concert with the anomeric stabilization afforded by an axial
ketal oxygen at C19 should favor the formation of 6 over the
alternative 19S epimer 7.31,32 In addition to being a source of
local instability, the 19S configuration of contra-thermodynamic
ketal 7 would likely impose a severe distortion from the native
cyclic conformation14,22 of 1. It was expected that a late stage
spiroketalization-equilibration could avoid this complication.
Synthetic Design for the C1-C14 Domain. A convergent
assembly of enone 5 required synthetic access to aldehyde 4
and keto phosphonate 3. Combined, these two advanced
synthetic intermediates embody all of the functionality required
for 1. Whereas the synthesis of keto phosphonate 3 was detailed
in a preceding paper,27 the strategy used for the preparation of
4 is outlined in Scheme 1. The intact C1-C2 R-hydroxyl,
R-methyl carboxylate moiety would be installed by diastereo-
selective alkylation of Seebach’s lactate pivalidene 8 with a C3-
C14 spiroketal-bearing fragment (9). Facial selectivity in the
alkylation of 8 to set the C2 stereogenic center of 1 was expected
to be high on the basis of ample literature precedent,33 although
previous alkylations of 8 have all involved electrophiles that
were considerably simpler in structure than 9. The electrophilic
C3 carbon of 9 would bear an R-stereogenic center and be
Results and Discussion
Overall Synthetic Plan. The strategy for a facile assembly
of 1 relied upon several key elements, chief among these was
to incorporate a maximal degree of functionalization into each
advanced synthetic intermediate so as to minimize the extent
of postcoupling transformations required to complete the
synthesis. It was anticipated that the 19R-configuration of the
central C19-C23 spiroketal could be established near the end
of the synthesis by an acid-triggered spiroketalization of a C16
hydroxyl upon a masked ketone at C19. Thus, the first
retrosynthetic disconnection was to open the C16-C19 tetrahy-
drofuran of 1 to yield the γ-hydroxy mixed-methyl acetal 2.
Only acetal and benzyl ether protecting groups were chosen to
differentiate 2 from 1. Thus, the end-game strategy relied upon
a stereoselective spiroketalization of 2 followed by final
hydroxyl and carboxylate deblocking.
(28) Blanchette, M. A.; Choy, W.; Davis, J. T.; Essenfeld, A. P.;
Masamune, S.; Roush, W. R.; Sakai, T. Tetrahedron Lett. 1984, 25, 2183.
(29) Deslongchamps, P. Stereoelectronic Effects in Organic Chemistry;
Pergamon: Oxford, 1983.
(30) Only one isomer was reportedly observed upon C16-C19 spiroket-
alization in the original synthesis of 1: Ichikawa, Y.; Isobe, M.; Goto, T.
Tetrahedron Lett. 1984, 25, 5049.
(31) Kirby, A. J. The Anomeric Effect and Related Stereoelectronic
Effects at Oxygen; Springer-Verlag: New York, 1983.
(32) Fukuuyama, T.; Akasaka, K.; Karanewsky, D. S.; Wang, C.-L. J.;
Schmid, G.; Kishi, Y. J. Am. Chem. Soc. 1979, 101, 262.
(33) Seebach, D.; Naef, R.; Calderari, G. Tetrahedron 1984, 40, 1313.
The trans-C14-C15 alkene of 2 was to be obtained by
coupling 327 and 4 under conditions sufficiently mild28 to allow
(25) Egloff, M.-P.; Cohen, P. T. W.; Reinemer, P.; Barford, D. J. Mol.
Biol. 1995, 254, 942.
(26) Goldberg, J.; Huang, H.; Kwon, Y.; Greengard, P.; Nairn, A. C.;
Kuriyan, J. Nature 1995, 376, 745.
(27) Urbanek, R. A.; Sabes, S. F.; Forsyth, C. J. J. Am. Chem. Soc. 1998,
120, 2523-2533 (preceding article in this issue).

2536 J. Am. Chem. Soc., Vol. 120, No. 11, 1998
Sabes et al.
Scheme 1
Scheme 2
Scheme 3
previous okadaic acid synthesis Isobe had reported that the yield
of the desired and anticipated major diastereomer obtained via
acid-catalyzed spiroketalization to form a similar fragment was
limited to ∼30-40%.38 This indicated a need to explore
potential improvements for the formation of the targeted okadaic
acid intermediate 10.
equatorially disposed on the C4-C8 oxane ring of the 1,7-
dioxaspiro[5.5]undec-4-ene system. Hence, potential complica-
tions arising from steric hindrance and double diastereoselective
matching34 could be anticipated. However, the expediency of
incorporating the fully functionalized, stereochemically correct,
and appropriately masked C1-C2 moiety in such a direct
fashion would avoid reliance upon alternative multistep se-
quences12,13 to incorporate the biologically essential35 R-hy-
droxyl carboxylate moiety. Once installed, the protected
R-hydroxyl carboxylate moiety would be subjected to only a
few mild transformations en route to 1. Whereas the pivalidene
group would be expected to resist premature cleavage during
late-stage C16-C23 spiroketalization catalyzed by mildly acidic
conditions, it could be removed readily by saponification at the
penultimate step in the total synthesis.
A hydroxyl group at C3 could act as a versatile precursor to
a variety of potential leaving groups in 9 for lactate addition.
Differential protection of the latent hydroxyl at C14 from those
at C2, C3, and C7 would facilitate final installation of the C14
aldehyde. Hence, the (8S)-spiroketal 10 was targeted as a key
intermediate in the preparation of 4 (Scheme 1). Thermody-
namic equilibration would be relied upon to generate prefer-
entially the natural product’s 8S configuration in 10 versus the
alternative (8R)-11 diastereomer upon acid-catalyzed dehydra-
tion of δ,δ′-dihydroxy enone 12.36 The partially flattened
conformation of the C8-C12 dihydropyran ring may diminish
the benefit of positioning the C4 oxygen axially with respect
to the C8-C12 ring, but the 8S-configuration in 10 should
confer full anomeric stabilization to the C4-C8 chair oxane
ring to strongly favor formation of the natural configuration.31
Hanessian’s application of a similar strategy to form the
substituted 1,7-dioxaspiro[5.5]undec-4-ene system of the aver-
mectins was quite successful.37 However, in the context of the
Two synthetic approaches to (Z)-enone 12 were considered.
A direct assembly of the desired spiroketal 10 via (Z)-enone
12 might involve the monoaddition of a (Z)-vinyl nucleophile
(16) to the lactone 13 (Scheme 2). For this, the homopropar-
gylic alcohol of an alkyne (cf. 14) might be used to anchor a
net trans-methyl metalation39,40 to provide oxavinyl metallocycle
16. By analogy to Collum’s direct annulation approach to the
synthesis of the phyllanthocin spiroketal,41 the addition of 16
to 13, followed by acidic workup could result in the direct
formation of 10. A reliable alternative route to the (Z)-enone
12 would involve methyl cuprate addition upon the correspond-
ing ynone (15),38,42 which, in turn, could be obtained from the
monoaddition of alkyne 14 to lactone 13.43 Subsequent acid-
catalyzed spiroketalization would complete the synthesis of 10.
The projected synthesis of the C1-C14 domain of 1 was thus
focused on three stages: the formation of (Z)-enone 12 from
lactone and alkyne precursors, thermodynamic spiroketalization
to provide 10, and installation of the R-hydroxyl, R-methyl
carboxylate moiety using lactate 8.
Synthesis of the C1-C14 Aldehyde (4). Lactone 13,
representing carbons 3-8 of 1, was obtained from the known44
isopropyl glycoside 17 (Scheme 3). To avoid cleavage of the
silyl ether of 17, the anomeric hydroxyl was liberated in two
steps. Trans-glycosidation with thiophenol and BF3‚OEt2
yielded the R-thiophenyl glycoside 18. Oxidative cleavage of
the thioglycoside was effected cleanly using a heterogeneous
(38) Isobe, M.; Ichikawa, Y.; Bai, D.-L.; Goto, T. Tetrahedron Lett. 1985,
26, 5203.
(39) Jousseaume, B.; Duboudin, F. J. Organomet. Chem. 1975, 91, C1.
(40) Duboudin, J. G.; Jousseaume, B. J. Organomet. Chem. 1979, 168,
1.
(41) McGuirk, P. R.; Collum, D. B. J. Am. Chem. Soc. 1982, 104, 4496.
(42) Corey, E. J.; Katzenellenbogen, J. A. J. Am. Chem. Soc. 1969, 91,
1851.
(34) Masamune, S.; Choy, W.; Petersen, J. S.; Sita, L. R. Angew. Chem.,
Int. Ed. Engl. 1985, 24, 1.
(35) Nishiwaki, S.; Fujiki, H.; Suganuma, M.; Furuya-Suguri, H.;
Matsushima, R.; Iida, Y.; Ojika, M.; Yamada, K.; Uemura, D.; Yasumoto,
T.; Schmitz, F. J.; Sugimura, T. Carcinogenesis 1990, 11, 1837.
(36) Deslongchamps, P.; Rowan, D. R.; Pothier, N.; Sauve, T.; Saun-
ders: J. K. Can. J. Chem. 1981, 59, 1105.
(37) Hanessian, S.; Ugolini, A.; Dube, D.; Hodges, P. J.; Andre, C. J.
Am. Chem. Soc. 1986, 108, 2776.
(43) Chabala, J. C.; Vincent, J. E. Tetrahedron Lett. 1978, 937.
(44) Isobe, M.; Ichikawa, Y.; Goto, T. Tetrahedron Lett. 1985, 26, 5199.

Synthesis of Okadaic Acid. 2
J. Am. Chem. Soc., Vol. 120, No. 11, 1998 2537
Scheme 4
the assembly of 10 are competitive formation of the (E)-enone;
generation of the kinetic spiroketal 11; epimerization of the C7
stereogenic center; and premature loss of the C3 hydroxyl
protecting group, which might lead to formation of a bridged
ketal (cf. 30, Scheme 5) involving the C3 and C4 hydroxyls.
Therefore, minimal protection of the secondary hydroxyls, as
well as stable protection of the latent primary alcohols of
intermediates leading to 10 was desired to minimize some of
these possible side reactions.
A brief survey was made to define useful protecting groups
for the C4 and C12 hydroxyls for each of the three discrete
steps: acetylide addition to 13, cuprate addition to the derived
ynone, and deprotection-spiroketalization. The lithium acetyl-
ide dianion derived from 24, bearing no protecting group at the
homopropargylic hydroxyl, added effectively to 13. However,
the resultant dihydroxy ynone 26 was a poor substrate for
dimethylcuprate addition.
Capping the C4 and C12 hydroxyls as tert-butyldimethylsilyl
(TBS) ethers at the ynone stage allowed for near quantitative
yield in the conjugate addition step, but subsequent one-pot,
acid-induced TBS removal and bicyclodehydration was unsat-
isfactory. An improved sequence involved blocking of both
the C4 and C12 hydroxyls as trimethylsilyl (TMS) ethers, which
could be removed rapidly upon acid treatment in the deprotec-
tion-spiroketalization step. In the event, addition of the lithium
acetylide derived from 14 to lactone 13 gave ynone 27 (Scheme
5). Without purification of 27, the newly formed hydroxyl at
C4 was silylated to provide bis-TMS ether 15. Conjugate
addition of dimethylcuprate led to the â-methyl enone 28, which,
without purification, was treated with TsOH‚H2O in benzene
at room temperature to remove the TMS groups and promote
spiroketalization. Although TLC indicated rapid loss of the
TMS groups to generate 12, subsequent spiroketalization
appeared to be relatively slow, and only modest yields of the
spiroketal 10 were obtained. The 8S configuration of 10 was
confirmed at this stage by the observation of reciprocal NOE
enhancements between the C7 methine and C9 vinyl protons.
Although the overall yield of 10 from 15 was only moderate,
no direct evidence for loss of the C3-hydroxyl protecting group
was observed. The stable tert-butyldiphenylsilyl (TBDPS) ether
was specifically chosen to prevent the primary alcohol of the
hypothetical triol 29 from participating in an acid-catalyzed
ketalization process (Scheme 5).50 Instead, it is likely that the
yield of 10 reflects an unfavorable Z-E selectivity in the
generation of enone 28 via the conjugate addition of dimeth-
ylcuprate upon ynone 15.51 While the present route provides
reasonable access to 10, continuing studies are aimed at
improving this sequence.
mixture of I2, acetone, and aqueous NaHCO3.45 The resultant
lactol was oxidized cleanly to 13 using PCC.46
Carbons 11-14 of 1 were derived from the known47
pentylidene protected triol 19 (Scheme 4). Conversion of the
free alcohol into the p-methoxybenzyl ether 20, followed by
methanolytic removal of the pentylidene group yielded diol 21.
The vicinal diol was transformed into oxirane 22 in one pot by
treatment with N-p-toluenesulfonyl imidazole and NaH.48 BF3-
assisted nucleophilic opening49 of 22 with lithium trimethylsi-
lylacetylide proceeded smoothly to give secondary alcohol 23.
A two-step transposition of the trimethylsilyl group from the
alkyne terminus to the newly formed secondary hydroxyl then
provided terminal alkyne 14.
The direct addition of a (Z)-vinyl metallic species bearing a
free homoallylic alkoxide group (16) to lactone 13 was
investigated as a potentially expedient approach to the bicyclic
ketal 10. Because dihydroxy (Z)-enone 12 would result from
monoaddition of 16 to 13, the product could be dehydratively
ketalized with a minimum of handling. This would avoid the
necessity of an intervening hydroxyl group deprotection step
and minimize the exposure of the isomerizable (Z)-enone 12 to
acidic conditions. Encouraged by literature precedents for
effecting the net trans-carbometalation of propargylic alcohols
using simple Grignard reagents and catalytic CuI,39,40 an attempt
to extend this methodology to the homopropargylic alcohol 24
was made. However, trans-carbometalation of 24 could not
be effected with MeMgBr/10% CuI. Attempts to generate39,40
magnesium metallocycle 25 and add the resultant nucleophile
to lactone 13 or an aldehyde were unsuccessful. No evidence
for alternative electrophilic trapping of an in situ generated vinyl
nucleophile was observed.
Because preliminary attempts to form and add (Z)-vinyl
nucleophilic species derived from 24 to lactone 13 were
unsuccessful, installation of the C10 methyl group of okadaic
acid was deferred until after C8-C9 bond formation. Acetylide
addition to lactone 1343 followed by conjugate addition of
dimethylcuprate to the resultant ynone was expected to give
selectively the corresponding (Z)-enone.38,42 Among the po-
tential complications that may limit the utility of this route for
(45) Russell, G. A.; Ochrymowycz, L. A. J. Org. Chem. 1969, 31, 3618.
(46) Corey, E. J.; Suggs, J. W. Tetrahedron Lett. 1975, 2647.
(47) Guindon, Y.; Yoakim, C.; Gorys, V.; Ogilvie, W. W.; Delorme,
D.; Renaud, J.; Robinson, G.; Lavallee, J.-F.; Slassi, A.; Jung, G.; Rancourt,
J.; Durkin, K.; Liotta, D. J. Org. Chem. 1994, 59, 1166.
Completion of the synthesis of the C1-C14 intermediate 4
required elaboration of the 1,7-dioxaspiro[5.5]undec-4-ene 10
to incorporate the C1-C2 R-hydroxyl, R-methyl carboxylate
(48) Cink, R. D.; Forsyth, C. J. J. Org. Chem. 1995, 60, 8122.
(49) Yamaguchi, M.; Hirao, I. Tetrahedron Lett. 1983, 24, 391.
(50) Ichikawa, Y.; Isobe, M.; Masaki, H.; Kawai, T.; Goto, T. Tetrahe-
dron 1987, 43, 4759.

2538 J. Am. Chem. Soc., Vol. 120, No. 11, 1998
Sabes et al.
Scheme 5
Scheme 6
selectivity of lactate alkylation was much lower (2.2:1.0) than
originally anticipated. That 34 and 35 were diastereomeric at
C2 was confirmed by removal of the C3 stereogenic center via
Barton deoxygenation.53 Although it was clear that the resultant
3-deoxy derivatives 36 and 31 (Scheme 6) were diastereomeric
to one another, the configuration at C2 of the latter was only
tentatively assigned to be R, in accordance with the expectation
that lactate alkylation would occur predominately anti to the
tert-butyl substituent of 8. No definitive NOE results could be
obtained to secure the C2-configurational assignment at the stage
of 31. Deeming it unlikely that a complete reversal from the
anticipated sense of diastereoselectivity would have occurred,
31 was selected for advancement toward 1. This ultimately
demonstrated that 31 and the natural product share the same
2R-configuration. The unexpected modest level of stereose-
lectivity in the addition of the lithium enolate obtained from 8
to aldehyde 33 may be due to unfavorable double diastereose-
lection between the mismatched pair 8 and 33. In particular,
secondary steric interactions in the transition state leading to
31 may serve to erode the degree of facial selectivity expected
on the basis of the tert-butyl substituent alone.
and C14 aldehyde moieties. Cleavage of the C3-silyl ether of
10 with TBAF yielded primary alcohol 32, which served as a
precursor to a variety of potential leaving groups (Scheme 6).
Numerous attempts to displace potential halide and sulfonate
leaving groups derived from 32 with enolates arising from lactate
pivalidene 8 were unsuccessful. Instead of the desired alkylation
product 31, unreacted electrophilic species (9) were generally
recovered. This may well reflect the crowded steric environ-
ment about the C3 carbon of 9.
With the installation of C1-C2 successfully completed, only
conversion of C14 into an aldehyde remained to complete the
synthesis of 4. This seemingly straightforward task was
accompanied by an unexpected acid-catalyzed isomerization.
Oxidative cleavage of the p-methoxybenzyl ether of 31 using
DDQ predictably gave primary alcohol 37, which could be
oxidized efficiently to aldehyde 4 with NaHCO3-buffered Dess-
Martin periodinane (Scheme 6). However, exposure of 37 to
acidic chloroform revealed the propensity of the δ-alkenyl
alcohol to undergo intramolecular addition of the alcohol across
the alkene to irretrievably generate tricyclic ether 38. This
presumably involves activation of the latent Michael acceptor
by protonation of a spiroketal oxygen followed by 1,4-addition
In contrast to primary halides and sulfonates, the aldehyde
(33) obtained by treatment of 32 with Dess-Martin periodi-
nane52 participated readily in carbon-carbon bond formation.
Addition of 33 to an excess of the preformed lithium enolate
derived from (S)-lactate pivalidene 8 gave alkylation products
34 and 35 in good combined yield. However, the facial
(51) Dounay, A. B.; Forsyth, C. J. Unpublished results.
(52) Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
Ireland, R. E.; Liu, L. J. Org. Chem. 1993, 58, 2899.
(53) Barton, D. H. R.; McCombie, S. W. J. J. Chem. Soc., Perkin Trans.
1 1975, 1574.

Synthesis of Okadaic Acid. 2
J. Am. Chem. Soc., Vol. 120, No. 11, 1998 2539
Scheme 7
alkene stereoselectively.54 Only assembly of the central C16-
C23 spiroketal and deprotection remained as the final synthetic
tasks. Diastereoselective reduction of the C16 carbonyl was a
necessary prelude to spiroketal formation. Corey’s CBS reagent
system55 seemed ideal for the diastereoselective reduction of
enone 5, given that the ketone was flanked by unbranched
aliphatic carbons on one side and an (E)-alkene on the other.
Thus, regio- and stereoselective reduction of 5 was accomplished
using the (S)-CBS/BH3 combination to give the (16R)-allylic
alcohol 2.56 Because partial spiroketalization occurred upon
acidic work up of the CBS reaction, 2 was not fully character-
ized. Instead, the reduction product mixture was subjected
directly to acid-catalyzed spiroketalization without rigorous prior
purification. This provided spiroketal 6 in 81% yield from 5.
As noted previously,9 a minor spiroketal stereoisomer was
generated along with 6 in this two-step sequence. Although it
has not been confirmed, it seems likely that this minor
diastereomer arises from the ketone reduction step, rather than
spiroketalization process, because the latter has been demon-
strated to be stereospecific in earlier synthetic intermediates.27
Only removal of the carboxylate and hydroxyl protecting
groups was required to complete the synthesis of 1. After the
C1-C2 pivalidene protecting group withstood the previous
coupling, reduction, and acid-induced spiroketalization reactions,
it was easily detached from 6 by mild aqueous base-induced
saponification to free both the C2 hydroxyl and C1 carboxylate
moieties. Hence, the present synthesis intercepts the previous
one12,13 at the stage of 7,24,27-tri-O-benzyl okadaic acid (40).
Saponification prior to dissolving metal cleavage of the benzyl
ethers allowed the carboxylate to be protected from reduction
as its carboxylate salt. Benzyl ethers were selected early in
the synthetic planning as useful protecting groups for the C7,
C24, and C27 hydroxyls. However, Isobe and co-workers
reported that the reductive cleavage of the same protecting
groups in the final step of the original synthesis of 1 was
complicated by overreduction using lithium in liquid am-
monia.12,13,57 Similar difficulties were encountered in the present
effort upon treatment of 40 with lithium in ammonia-ethanol.
In contrast, the use of lithium di-tert-butylbiphenylide58 (LiDBB)
solutions in THF for controlled debenzylation59 was far less
problematic. The final deprotection could be carefully opti-
mized even on small scale using stock solutions of known
concentration of LiDBB. Thus, the three benzyl ethers of 40
were reductively cleaved without substantial complications from
the other potentially sensitive functional groups to deliver 1
reproducibly in ∼70% yield after isolation and purification. The
synthetic okadaic acid obtained in this fashion was chromato-
graphically and spectroscopically indistinguishable from a
sample of 1 that was isolated from Halichondria okadai. Aside
from the initial difficulties associated with liberating the three
secondary hydroxyl groups of 1, the end-game strategy for a
rapid completion of the total synthesis played out splendidly.
of the proximal C14 hydroxyl group to C10 of the allylically
stabilized oxonium species. No evidence for formation of the
alternative ketal (39) was observed. Formation of the triox-
atricycle 38 was prevented by simply avoiding exposure of 31
to acidic conditions. Hence, the C1-C14 intermediate 4 could
be prepared in 11 steps and ∼20% overall yield from lactone
13.
Conclusion
An efficient total synthesis of the potent protein serine-
threonine phosphatase inhibitor okadaic acid has been developed
Total Synthesis of Okadaic Acid. The complete carbon
skeleton of okadaic acid was assembled by joining aldehyde 4
with keto phosphonate 327 under Masamune-Roush conditions
(Scheme 7).28 This mildly basic Horner-Wadsworth-Emmons
reaction allowed the fully functionalized coupling partners to
be joined reliably to give (E)-enone 5. In addition to conjoining
all of the functionality required for 1 that is resident in 3 and 4,
this convergent coupling cleanly provided the trans-C14-C15
(54) Wadsworth, D. H.; Schupp, O. E.; Seus, E. J.; Ford, J. A. J. J. Org.
Chem. 1965, 30, 0, 680.
(55) Corey, E. J.; Bakshi, R. K.; Shibata, S.; Chen, C.-P.; Singh, V. K.
J. Am. Chem. Soc. 1987, 109, 7925.
(56) The use of stoichiometric amounts of the CBS reagent was effective
and convenient for small-scale reductions, although catalytic amounts may
suffice for preparative-scale reduction of 5.
(57) Ichikawa, Y.; Isobe, M.; Goto, T. Agric. Biol. Chem. 1988, 52, 975.
(58) Freeman, P. K.; Hutchinson, L. L. J. Org. Chem. 1980, 45, 1924.
(59) Ireland, R. E.; Smith, M. G. J. Am. Chem. Soc. 1988, 110, 854.

2540 J. Am. Chem. Soc., Vol. 120, No. 11, 1998
Sabes et al.
formed. The solution was cooled to -78 °C, and a solution of 15
(234 mg, 270 µmol) in diethyl ether (2 mL) was added via cannula.
After 2.5 h, saturated aqueous NH4Cl (3 mL) was added and the mixture
allowed to warm to room temperature and stir until the aqueous phase
became bright blue. The solution was extracted with diethyl ether (3
× 25 mL) and the combined organic phases were washed with H2O
and saturated aqueous NaCl (10 mL ea), dried over Na2SO4, filtered,
and concentrated. The residue was filtered through silica gel with
hexanes-ethyl acetate (4:1, v/v) and the filtrate concentrated to give
crude enone 28 (234 mg, 265 µmol). This was dissolved in benzene
(10 mL), and p-toluenesulfonic acid monohydrate (2.5 mg, 13 µmol)
was added. After the solution was stirred at room temperature for 11
h, it was diluted with diethyl ether (30 mL), washed with H2O and
saturated aqueous NaCl (5 mL ea), dried over Na2SO4, filtered, and
concentrated. Silica gel column chromatography (hexanes-ethyl
acetate, 10:1, v/v) of the residue gave 10 (74 mg, 103 µmol, 38% from
and executed on the basis of a direct and convergent synthetic
strategy. This involved the individual syntheses and sequential
coupling of fragments representing carbons 28-38, 16-27, and
1-14 of the natural product. As detailed in the preceding paper,
the C28-C38 fragment was prepared in 10 steps from methyl
(S)-3-hydroxy-2-methylpropionate, and its C16-C27 coupling
partner was assembled from readily available methyl 3-O-
benzyl-R-D-altropyranoside in 14 steps.27 Their direct coupling
and subsequent elaboration to form a fully functionalized C16-
C38 synthetic intermediate (3) spanned an additional 5-7
steps.27 The complementary C1-C14 intermediate (4) was
prepared in 11 steps from functionalized δ-valerolactone (13)
and alkyne (14) intermediates representing carbons 3-8 and
9-14, respectively. The incorporation of a maximal degree of
functionalization into each advanced intermediate allowed a
rapid completion of the total synthesis, which was effected in
5 steps and 48% yield from 3 and 4. This featured a highly
stereoselective 2-step construction of the central 1,6-dioxaspiro-
[4.5]decane system by the tandem use of Corey’s CBS reduc-
tion55 and an acid-triggered trans-ketalization. In addition to
delivering the natural product and highlighting a number of
useful synthetic tactics and methods, the present work is notable
because minor variations of the synthetic route will provide
access to previously unavailable analogues of okadaic acid. Such
compounds will be valuable for probing the structural basis of
protein serine-threonine phosphatase inhibition.
15) as a clear, colorless oil: Rf 0.38 (hexanes-ethyl acetate, 5:1, v/v);
1
[R]25 ) -21 (c 0.4, CHCl3); IR (neat) 2929, 1512, 1427 cm-1; H
D
NMR (CDCl3, 500 MHz) δ 7.70-7.64 (m, 4H), 7.45-7.23 (m, 11H),
7.25 (d, J ) 8.5 Hz, 2H), 6.85 (d, J ) 8.5 Hz, 2H), 5.19 (s, 1H), 4.65
(d, J ) 12.5 Hz, 1H), 4.52 (d, J ) 12.5 Hz, 1H), 4.49 (d, J ) 11.5 Hz,
1H), 4.42 (d, J ) 11.5 Hz, 1H), 3.89-3.80 (m, 2H), 3.79 (s, 3H), 3.77
(dd, J ) 4.5, 9.0 Hz, 1H), 3.68 (dd, J ) 5.5, 10.0 Hz, 1H), 3.48 (dd,
J ) 6.5, 10.0 Hz, 1H), 3.43 (dd, J ) 8.0, 8.5 Hz, 1H), 3.29 (dd, J )
4.5, 11.5 Hz, 1H), 2.12-2.02 (m, 2H), 2.01-1.82 (m, 4H), 1.76 (s,
3H), 1.30 (m, 1H), 1.05 (s, 9H), 1.02 (s, 3H); 13C NMR (CDCl3, 125
MHz) δ 159.0, 138.9, 137.5, 135.7, 133.7, 130.9, 129.5, 129.2, 129.1,
128.3, 128.1, 127.6, 127.5, 123.0, 113.6, 95.9, 79.0, 72.6, 72.1, 71.4,
69.8, 69.2, 67.1, 55.2, 38.3, 33.2, 28.2, 26.9, 24.1, 23.1, 19.3, 13.9;
ΗRMS calcd for C45H56O6Si [M + H]+ 721.3925, found 721.3939.
Experimental Section27
(2R)-Carboxylate 34 and (2S)-Carboxylate 35. To a stirred -78
°C solution of diisopropylamine (102 µL, 733 µmol) in THF (8 mL)
under N2 was added n-butyllithium (261 µL of a 2.30 M solution in
hexanes, 600 µmol). After 20 min, a solution of (S)-lactate pivalidene
833 (105 mg, 667 µmol) in THF (1 mL) was added via cannula. After
20 min, a solution of 33 (32 mg, 67 µmol) in THF (1 mL) was added
via cannula. The resulting solution was allowed to stir for 15 min
before saturated aqueous NH4Cl was added. The mixture was allowed
to warm to room temperature, and diluted with diethyl ether (30 mL),
washed with H2O and saturated aqueous NaCl (5 mL ea), and dried
over Na2SO4. Filtration and concentration gave a residue that was
purified by silica gel column chromatography (hexanes-ethyl acetate,
10:1 to 7:1, v/v) to give 34 (26 mg, 41 µmol, 60%) and 35 (12 mg, 19
µmol, 28%) as clear, colorless oils. Data for 34: Rf 0.58 (hexanes-
ethyl acetate, 2:1, v/v); [R]25D ) +9.5 (c 1.2, CHCl3); IR (neat) 3469,
2962, 1797, 1515 cm-1; 1H NMR (CDCl3, 500 MHz) δ 7.32-7.27 (m,
7H), 6.85 (d, J ) 8.5 Hz, 2H), 5.40 (s, 1H), 5.16 (s, 1H), 4.61 (d, J )
12.5, 1H), 4.49 (d, J ) 9.5 Hz, 1H), 4.47 (d, J ) 9.5 Hz, 1H), 4.42 (d,
J ) 12.5 Hz, 1H), 3.87 (ddd, J ) 3.0, 5.0, 11.5 Hz, 1H), 3.84 (dd, J
) 5.0, 9.5 Hz, 1H), 3.71 (ddd, J ) 3.5, 8.0, 11.5 Hz, 1H), 3.48 (dd, J
) 2.5, 9.0 Hz, 1H), 3.36 (dd, J ) 8.5, 8.5 Hz, 1H) 3.23 (dd, J ) 4.5,
11.5 Hz, 1H), 2.71 (d, J ) 9.5 Hz, 1H), 2.10-2.04 (m, 2H), 1.93 (m,
1H), 1.83 (m, 2H), 1.76 (s, 3H), 1.63 (m, 1H), 1.49 (s, 3H), 1.03 (d, J
) 6.5 Hz, 3H), 0.90 (s, 9H); 13C NMR (CDCl3, 125 MHz) δ 173.6,
158.9, 138.8, 137.9, 131.1, 129.3, 128.2, 127.7, 127.4, 121.9, 113.6,
109.3, 96.4, 80.8, 78.3, 76.1, 72.6, 72.3, 71.3, 70.2, 67.9, 55.3, 38.3,
34.4, 32.8, 27.8, 23.5, 23.3, 23.0, 22.2, 13.9; ΗRMS calcd for C37H50O9
[M + H]+ 639.3534, found 639.3546.
3-Deoxy-34 (31). To a stirred 0 °C solution of 34 (26 mg, 41 µmol)
in THF (2.5 mL) under N2 was added NaH (29 mg, 1.2 mmol). The
stirred mixture was allowed to warm to room temperature over 30 min
before CS2 (73 µL, 1.2 mmol) was added. After 10 min, methyl iodide
(76 µL, 1.2 mmol) was added and the solution became yellow. After
another 10 min, the yellow color dissipated and the solution was cooled
to 0 °C. Saturated aqueous NH4Cl (0.3 mL), diethyl ether (5 mL),
and H2O (1.5 mL) were added. The separated aqueous phase was
extracted with diethyl ether (3 mL) and the combined organic phases
were washed with saturated aqueous NaCl (2 mL), dried over MgSO4,
filtered, and concentrated. The residue was azeotropically dried from
benzene (5 mL) and then dissolved in toluene (5 mL). After the
resultant solution was deoxygenated with a stream of Ar for 15 min,
1. Synthesis of the C1-C14 Fragment. (2R,3S,8R,11S)-8-O-
Benzyl-12-O-(tert-butyldiphenylsilyl)-1-O-(4-methoxybenzyl)-2-methyl-
3,11-bis-O-(trimethylsilyl)-7-oxo-5-undecyne-1,3,8,11,12-pentaol (15).
To a stirred -78 °C solution of 14 (1.268 g, 3.96 mmol) in THF (40
mL) under N2 was added n-butyllithium (1.427 mL of a 2.50 M solution
in hexanes, 3.57 mmol). After the mixture was stirred for 20 min, a
solution of 13 (1.030 g, 2.17 mmol) in THF (10 mL) was added via
cannula. After 40 min, saturated aqueous NH4Cl (5 mL) was added,
followed by H2O (25 mL), and the mixture allowed to warm to room
temperature. The mixture was extracted with diethyl ether (2 × 150
mL) and the combined organic extracts were washed with H2O and
saturated aqueous NaCl. The organic fraction was dried over Na2SO4,
filtered, and concentrated. The residue was chromatographed on silica
gel (hexanes-ethyl acetate, 6:1) and the resultant 27 was dissolved in
CH2Cl2 (40 mL). To this solution was added chlorotrimethylsilane (300
µL, 2.39 mmol) followed by imidazole (259 mg, 4.34 mmol). After
stirring for 45 min, the mixture was diluted with diethyl ether (100
mL) and washed with H2O and saturated aqueous NaCl (15 mL ea).
The organic phase was dried over over Na2SO4, filtered, and concen-
trated. The residue was purified by silica gel column chromatography
(hexanes-ethyl acetate, 10:1 then 3:1, v/v) to give 15 (1.529 g, 1.77
mmol, 82%) as a clear, colorless oil: Rf 0.32 (hexanes-ethyl acetate,
5:1, v/v); [R]25 ) +16 (c 0.5, CHCl3); IR (neat) 2958, 2210, 1674,
D
1515, 1423 cm-1; 1H NMR (CDCl3, 500 MHz) δ 7.68-7.66 (m, 4H),
7.44-7.24 (m, 13H), 6.87 (m, 2H), 4.73 (d, J ) 11.5 Hz, 1H), 4.42 (d,
J ) 11.5 Hz, 1H), 4.39 (d, J ) 11.5 Hz, 1H), 4.37 (d, J ) 11.5 Hz,
1H), 3.89 (m, 2H), 3.80 (s, 3H), 3.69 (m, 1H), 3.55 (dd, J ) 5.5, 10.5
Hz, 1H), 3.45 (dd, J ) 6.0, 9.5 Hz, 1H), 3.43 (dd, J ) 6.0, 10.5 Hz,
1H), 3.33 (dd, J ) 5.5, 9.0 Hz, 1H), 2.62 (dd, J ) 5.5, 17.5 Hz, 1H),
2.52 (dd, J ) 7.0, 17.5 Hz, 1H), 2.04-1.89 (m, 3H), 1.70 (m, 1H),
1.42 (m, 1H), 1.05 (s, 9H), 0.92 (d, J ) 12.0 Hz, 3H), 0.13 (s, 9H),
0.03 (s, 9H); 13C NMR (CDCl3, 125 MHz) δ 189.3, 159.0, 137.6, 135.6,
135.5, 133.6, 133.5, 130.6, 129.6, 129.1, 128.4, 127.9, 127.8, 127.7,
113.7, 95.4, 85.2, 80.4, 72.8, 72.7, 72.3, 72.1, 71.4, 67.9, 55.3, 38.9,
30.1, 28.4, 26.8, 25.8, 19.2, 13.8, 0.33, 0.28; ΗRMS calcd for C50H70O7-
Si3 [M + H]+ 867.4508, found 867.4459.
Spiroketal 10. To a -78 °C supension of CuI (205 mg, 1.08 mmol)
in diethyl ether (8 mL) under N2 was added methyllithium (1.54 mL
of a 1.40 M solution in diethyl ether, 2.16 mmol). The mixture was
allowed to warm slowly to ∼-40 °C until a clear and colorless solution

Synthesis of Okadaic Acid. 2
J. Am. Chem. Soc., Vol. 120, No. 11, 1998 2541
2,2′-azobisisobutyronitrile (27 mg, 0.16 mmol) and tri-n-butyltin hydride
(88 µL, 0.33 µmol) were added, and the mixture was heated at 80 °C
for 2 h under Ar. After the solution was cooled to room temperature,
it was diluted with ethyl acetate (5 mL), washed with H2O and saturated
aqueous NaCl (2 mL ea), dried over Na2SO4, filtered, and concentrated.
Silica gel column chromatography (hexanes-ethyl acetate, 1:0 to 7:1,
v/v) of the residue gave 31 (22 mg, 35 µmol, 88%) as a clear, colorless
became turbid after 10 min and was stirred for an additional 20 h. The
mixture was diluted with diethyl ether (5 mL), washed with H2O and
saturated aqueous NaCl (0.5 mL ea), dried over Na2SO4, filtered, and
concentrated. Silica gel column chromatography (hexanes-ethyl
acetate, 5:1, v/v) of the residue gave 5 (12 mg, 10 µmol, ∼86%) as a
clear, colorless oil: Rf 0.54 (hexanes-ethyl acetate, 2:1, v/v); [R]25
D
) +7 (c 0.2, CHCl3); IR (neat) 2948, 1788, 1682 cm-1 1H NMR
;
oil: Rf 0.74 (hexanes-ethyl acetate, 2:1, v/v); [R]25 ) +11 (c 0.2,
(CDCl3, 500 MHz) δ 7.37-7.24 (m, 15H), 7.04 (dd, J ) 8.5, 16.0 Hz,
1H), 6.19 (d, J ) 16.0 Hz, 1H), 5.45 (s, 1H), 5.35 (s, 1H), 5.19 (s,
1H), 5.05 (s, 1H), 4.87 (d, J ) 12.0 Hz, 1H), 4.76 (d, J ) 12.0 Hz,
1H), 4.73 (d, J ) 11.0 Hz, 1H), 4.60 (d, J ) 12.5 Hz, 1H), 4.57 (d, J
) 11.0 Hz, 1H), 4.45 (d, J ) 12.5 Hz, 1H), 4.26 (d, J ) 8.0 Hz, 1H),
3.94-3.81 (m, 4H), 3.71-3.60 (m, 3H), 3.41 (dd, J ) 10.0, 10.0 Hz,
1H), 3.27 (dd, J ) 4.5, 11.5 Hz, 1H), 3.23 (dd, J ) 2.0, 10.0 Hz, 1H),
3.21 (s, 3H), 2.70 (ddd, J ) 6.5, 10.5, 16.5 Hz, 1H), 2.61-2.55 (m,
2H), 2.06-1.98 (m, 2H), 1.92-1.77 (m, 11H), 1.75 (s, 3H), 1.70-
1.36 (m, 13H), 1.44 (s, 3H), 1.29 (m, 1H), 1.17 (d, J ) 6.5 Hz, 3H),
0.94 (d, J ) 6.5 Hz, 3H), 0.89 (s, 9H), 0.88 (d, J ) 7.0 Hz, 3H); 13C
NMR (CDCl3, 125 MHz) δ 199.9, 175.1, 150.1, 143.5, 138.6, 137.0,
130.9, 128.3, 128.2, 128.1, 127.7, 127.6, 127.5, 127.4, 122.6, 112.7,
108.5, 102.8, 99.1, 96.0, 95.7, 85.3, 85.3, 78.6, 78.1, 77.2, 75.1, 74.4,
73.6, 73.1, 71.1, 70.7, 70.5, 65.4, 60.4, 47.5, 43.2, 42.0, 35.9, 34.7,
34.1, 33.6, 32.3, 31.2, 30.4, 29.2, 27.4, 26.3, 25.5, 25.4, 23.7, 23.4,
22.9, 18.8, 16.6, 16.0, 10.8; ΗRMS (MALDI) calcd for C71H96NaO14
[M+ Na]+ 1195.6701, found 1195.6660.
D
CHCl3); IR (neat) 2926, 1738, 1549 cm-1; 1H NMR (CDCl3, 500 MHz)
δ 7.31 (d, J ) 9.0 Hz, 2H), 7.30-7.25 (m, 5H), 6.86 (d, J ) 9.0 Hz,
2H), 5.28 (s, 1H), 5.18 (s, 1H), 4.61 (d, J ) 12.5 Hz, 1H), 4.48 (d, J
) 12.5 Hz, 1H), 4.47 (d, J ) 11.0 Hz, 1H), 4.42 (d, J ) 11.0 Hz, 1H),
3.91-3.87 (m, 1H), 3.86 (dd, J ) 4.5, 9.0 Hz, 1H), 3.80 (s, 3H), 3.70
(ddd, J ) 3.5, 8.0, 11.5 Hz, 1H), 3.35 (dd, J ) 8.5, 9.0 Hz, 1H), 3.27
(dd, J ) 4.5, 11.5 Hz, 1H), 2.12-2.02 (m, 2H), 2.01-1.60 (m, 6H),
1.75 (s, 3H), 1.44 (s, 3H), 1.31 (m, 1H), 1.03 (d, J ) 7.0 Hz, 3H),
0.90 (s, 9H); 13C NMR (CDCl3, 125 MHz) δ 175.3, 157.1, 138.8, 137.5,
131.0, 129.2, 128.2, 127.8, 127.4, 122.5, 119.1, 113.6, 108.2, 96.1,
78.3, 72.6, 72.4, 71.3, 70.0, 68.2, 65.4, 55.3, 42.0, 38.3, 34.2, 32.9,
32.0, 25.2, 24.5, 23.3, 14.0; ΗRMS calcd for C37H50O8 [M + H]+
623.3585, found 623.3624.; HRMS (MALDI) calcd for C37H50O8Na
[M + Na]+ 645.3404, found 645.3407.
C14 Alcohol 37. To a mixture of 31 (27 mg, 43 µmol), CH2Cl2
(3.5 mL), an aqueous Na2PO4 buffer (pH ) 7, 2.0 mL), and tert-butyl
alcohol (0.35 mL) was added dichlorodicyanoquinone (49 mg, 0.22
mmol). The reaction flask was placed in a aqueous bath and sonicated
for 1 min and then assayed for disappearance of 31 by TLC. This
process was repeated for a total of three 1-min sonications, at which
point no 31 remained. The mixture was diluted with diethyl ether (8
mL) and washed with saturated aqueous NaHCO3 (2 mL). The aqueous
phase was extracted with diethyl ether (2 × 2 mL) and the combined
organic phases were washed with H2O and saturated aqueous NaCl
(1.5 mL ea), dried over Na2SO4, filtered, and concentrated. The residue
was purified by silica gel column chromatography (hexanes-ethyl
acetate, 5:1 to 3:2, v/v) to give 37 (17 mg, 34 µmol, 77%) as a clear,
colorless oil: Rf 0.26 (hexanes-ethyl acetate, 2:1, v/v); [R]25D +9.7 (c
0.4, CHCl3); 1H NMR (CDCl3, 500 MHz) δ 7.28-7.15 (m, 5H), 5.34
(s, 1H), 5.25 (s, 1H), 4.44 (d, J ) 12.5 Hz, 1H), 4.31 (d, J ) 12.5 Hz,
1H), 4.13-4.07 (m, 2H), 4.02 (m, 1H), 3.88 (m, 1H), 3.17 (dd, J )
5.0, 12.0 Hz, 1H), 2.01-1.94 (m, 2H), 1.85-1.77 (m, 2H), 1.68 (m,
1H), 1.66-1.56 (m, 4H), 1.55 (s, 3H), 1.46 (s, 3H), 1.08 (d, J ) 7.0
Hz, 3H), 0.86 (s, 9H); 13C NMR (CDCl3, 125 MHz) δ 175.4, 139.9,
137.5, 128.9, 128.9, 128.8, 128.7, 128.4, 128.1, 127.9, 123.4, 108.5,
96.6, 79.1, 72.6, 71.6, 65.9, 65.8, 43.5, 40.8, 34.6, 33.9, 32.3, 25.7,
25.0, 23.8, 23.3, 14.5; ΗRMS calcd for C29H42O7 [M + H]+ 503.3010,
found 503.3033.
1,2-Di-O-(S)-pivalidene-7,24,27-tri-O-benzylokadaic Acid (6). To
a stirred solution of (S)-2-methyl-CBS-oxazaborolidine55 (136 µL of a
1.3 M solution in toluene, 0.18 mmol) in THF (0.6 mL) at 0 °C and
under N2 was added borane-tetrahydrofuran complex (120 µL of a 1
M solution in THF, 120 µmol) followed by a solution of 5 (13.6 mg,
11.6 µmol) in THF (0.35 mL). After 5 min, H2O (∼200 µL) was added
and the mixture was allowed to warm to room temperature. Diethyl
ether (2 mL) was added and the mixture was washed with 5% aqueous
HCl. The aqueous phase was extracted with diethyl ether (2 × 0.5
mL), and the combined organic phases were washed with H2O and
saturated aqueous NaCl (0.5 mL ea), dried over Na2SO4, filtered, and
concentrated. The crude allylic alcohol 2 (Rf 0.28; hexanes-ethyl
aceatate, 2:1, v/v) was filtered through silica gel with ethyl acetate,
the filtrate was concentrated and then diluted with benzene (1 mL).
p-Toluenesulfonic acid monohydrate (1 mg, 5 µmol) was added and
the mixture stirred at room temperature for 2 h. Triethylamine (∼300
µL) was added, the mixture was filtered through silica gel, and the
filtrate was concentrated. Silica gel column chromatography (hexanes-
ethyl acetate, 12:1, v/v) of the residue gave 6 (10.7 mg, 9.4 µmol, 81%)
as a clear, colorless oil: Rf 0.70 (hexanes-ethyl acetate, 2:1, v/v); [R]25
D
) +18 (c 0.2, CHCl3); IR (neat) 2934, 2874, 1792, 1728, 1664 cm-1
;
1H NMR (CDCl3, 500 MHz) δ 7.39-7.23 (m, 15H), 5.78 (dd, J )
8.5, 10.5 Hz, 1H), 5.60 (dd, J ) 7.5, 10.5 Hz, 1H), 5.41 (s, 1H), 5.33
(s, 1H), 5.19 (s, 1H), 5.01 (s, 1H), 4.85 (d, J ) 13.0 Hz, 1H), 4.76 (d,
J ) 13.0 Hz, 1H), 4.75 (d, J ) 11.0 Hz, 1H), 4.62 (d, J ) 12.5 Hz,
1H), 4.58 (m, 1H), 4.56 (d, J ) 12.5 Hz, 1H), 4.47 (d, J ) 12.5 Hz,
1H), 4.24 (d, J ) 8.0 Hz, 1H), 4.02 (m, 1H), 3.91 (m, 2H), 3.72-3.61
(m, 5H), 3.28 (dd, J ) 4.0, 12.0 Hz, 1H), 3.22 (dd, J ) 2.5, 10.5 Hz,
1H), 2.40 (m, 1H), 2.22 (m, 1H), 2.09-1.93 (m, 5H), 1.87-1.77 (m,
10H), 1.75 (s, 3H), 1.69-1.36 (m, 13H), 1.45 (s, 3H), 1.30 (m, 1H),
1.08 (d, J ) 7.0 Hz, 3H), 0.91 (d, J ) 7.0 Hz, 3H), 0.89 (s,9H), 0.87
(d, J ) 7.5 Hz, 3H); HRMS (MALDI) calcd for C70H94O13Na [M +
Na]+ 1165.6595, found 1165.6569.
7,24,27-Tri-O-benzylokadaic acid (40).12 To a solution of 6 (4.8
mg, 4.2 µmol) in THF (1 mL) was added 1 M aqueous LiOH (100 µL,
100 µmol). After the mixture was stirred for 48 h at room temperature,
the THF was removed under a stream of N2. The resulting aqueous
mixture was diluted with H2O (0.4 mL) and acidified to pH 5 with 0.5
M aqueous HCl. The mixture was extracted with diethyl ether (4 × 2
mL), and the combined ether extracts were washed with H2O and
saturated aqueous NaCl (0.5 mL ea), dried over Na2SO4, filtered, and
concentrated. The crude residue was filtered through a pad of silica
gel using hexanes-ethyl acetate-acetic acid (1:1:0.05, v/v/v) and the
filtrate was concentrated to give 40 (4.5 mg, ∼4.2 µmol) as an oil: Rf
0.69 (hexanes-ethyl acetate-acetic acid, 1:1:0.05, v/v/v); 1H NMR
(CDCl3, 500 MHz) matched previously reported data [lit.13 1H NMR
Aldehyde 4. To a stirred room-temperature solution of 37 (10 mg,
20 µmol) in CH2Cl2 (1.5 mL) was added NaHCO3 (25 mg, 0.30 mmol)
followed by the Dess-Martin periodinane reagent52 (25 mg, 60 µmol).
The mixture was stirred for 30 min before diethyl ether (4 mL) and
10% aqueous Na2S2O3 and saturated aqueous NaHCO3 (0.5 mL ea)
were added. The mixture was stirred until the organic phase became
clear and colorless. The separated aqueous phase was extracted with
diethyl ether (2 × 0.5 mL), and the combined organic fractions were
washed with H2O and saturated aqueous NaCl (1 mL ea), dried over
Na2SO4, filtered, and concentrated. Silica gel column chromatography
(hexanes-ethyl acetate, 5:1, v/v) of the residue gave 4 (9 mg, ∼0.18
µmol, 90%) as a white crystalline solid: mp 125-126 °C (hexanes-
ethyl acetate); Rf 0.57 (hexanes-ethyl acetate, 2:1, v/v); 1H NMR
(CDCl3, 500 MHz) δ 10.13 (d, J ) 3.0 Hz, 1H), 7.33-7.15 (m, 5H),
5.36 (s, 1H), 5.30 (s, 1H), 4.51 (d, J ) 12.5 Hz, 1H), 4.36 (d, J ) 12.5
Hz, 1H), 4.35 (ddd, J ) 3.5, 7.5, 11.0 Hz, 1H), 4.04 (ddt, J ) 3.5, 9.5,
12.0 Hz, 1H), 3.23 (dd, J ) 4.5, 12.0 Hz, 1H), 2.45 (ddq, J ) 3.0, 7.0,
7.0 Hz, 1H), 2.04 (m, 1H), 1.86 (m, 2H), 1.71-1.60 (m, 2H), 1.56 (s,
3H), 1.54-1.45 (m, 2H), 1.43 (s, sH), 1.31 (m, 1H), 1.03 (d, J ) 7.0
Hz, 3H), 0.92 (s, 9H).
2. Synthesis of Okadaic Acid. C14-C16 Enone 5. To a stirred
room-temperature solution of 327 (12 mg, 15 µmol) in CH3CN (0.5
mL) was added LiCl (3 mg, 0.07 mmol) followed by diisopropylethyl-
amine (4 µL, 0.02 mmol). After stirring for 10 min, a solution of 4 (6
mg, ∼12 µmol) in CH3CN (0.6 mL) was added. The resulting mixture

2542 J. Am. Chem. Soc., Vol. 120, No. 11, 1998
Sabes et al.
(500 MHz) δ 7.2-7.4 (15H), 5.74 (1H, dd, J ) 15, 8 Hz), 5.61 (1H,
dd, J ) 15, 8 Hz), 5.35 (1H, t, J ) 1 Hz), 5.17 (1H, brs), 5.14 (1H,
brs), 5.02 (1H, brs), 4.89 (1H, d, J ) 13 Hz), 4.77 (1H, d, J ) 13),
4.73 (1H, d, J ) 8 Hz), 4.61 (1H), 4.60 (1H, d, J ) 13 Hz), 4.56 (1H,
d, J ) 11 Hz), 4.48 (1H, d, J ) 13 Hz), 4.24 (1H, d, J ) 8 Hz), 4.02
(1H, tt, J ) 11, 2 Hz), 3.88-3.95 (2H), 3.55-3.7 (5H), 3.24 (1H, dd,
J ) 12, 4 Hz), 3.22 (1H, dd, J ) dd, 11, 2 Hz), 2.42 (1H, qt, J ) 7,
7 Hz), 2.23 (1H), 2.13 (1H, dd, J ) 14, 2 Hz), 1.73 (3H, s), 1.36 (3H,
s), 1.05 (3H, d, J ) 7 Hz), 0.91 (3H, d, J ) 7 Hz), 0.87 (3H, d, J )
7 Hz)]; HRMS (MALDI) calcd for C65H86O13Na [M + Na]+ 1097.5967,
found 1097.6010, calcd for C65H85O13Na2 [M + 2Na - H]+ 1119.5787,
found 1119.5837.
from H. okadai: Rf 0.27 (hexanes-ethyl acetate-acetic acid, 1:1:0.05,
v/v), 0.26 (dichloromethane-methanol, 19:1, v/v), 0.17 (diethyl ether-
acetic acid, 99.5:0.05), 0.15 (hexanes-acetone, 1:1), 0.50 (tert-butyl
methyl ether-chloroform-methanol-acetic acid, 9:1:0.5:0.05); indis-
tinguishable (Baker 7011-4 Si-HPF TLC) from naturally occurring 1
(H. okadai) upon multiple elutions in hexanes-ethyl acetate-acetic
acid (1:1:0.05), and hexanes-ethyl acetate-chloroform-acetic acid
(1:1:0.5:0.05); 1H NMR (CDCl3, 500 MHz) matched naturally occurring
1 isolated from H. okadai, and previously reported data;14,22 photocopies
of spectra are included in the Supporting Information; HRMS (MALDI)
calcd for C44H67O13Na2 [M + 2Na - H]+ 849.4377, found 849.4352.
Okadaic Acid (1). To a stirred -78 °C solution of 40 (2.5 mg, 2.3
µmol) in THF (0.2 mL) under N2 was added a solution of lithium di-
tert-butylbiphenylide59 (0.2 mL of 0.15 M solution in THF, ∼0.03
mmol). After stirring for 15 min, H2O (0.2 mL) was added to the deep
blue-green solution and the resulting colorless mixture was allowed to
warm to room temperature. The THF was removed under a stream of
N2, and the residue was diluted with H2O (0.2 mL) and washed with
hexanes (3 × 1 mL). The aqueous phase was acidified to pH 5 with
0.5 M aqueous HCl, and extracted with diethyl ether (4 × 1 mL). The
combined ether extracts were dried over Na2SO4, filtered, and concen-
trated. The residue was purified by silica gel column chromatography
(hexanes-ethyl acetate-acetic acid, 1:1:0.05, v/v/v) to give 1 (1.3 mg,
Acknowledgment. We thank Professor D. Uemura for
providing an authentic sample of 1 isolated from Halichondria
okadai. Financial support from the NIH (CA62195) and a
University of Minnesota McKnight Land-Grant Professorship
(C.J.F.) is gratefully acknowledged.
Supporting Information Available: Experimental proce-
dures and characterization data for compounds 18, 13, 14, 20-
1
24, 32, and 33; photocopies of H NMR spectra for synthetic
and natural 1 and synthetic intermediates 4-6, 10, 13-15, 18-
24, 31-34, 37, and 40 and of 13C NMR spectra for compounds
4, 5, 10, 31, and 34 (34 pages). See any current masthead page
for ordering and Web access instructions.
1.6 µmol, 69%) as a white solid: [R]25 +30.0 (c 1.3, CHCl3) [H.
D
okadai isolate lit.14 [R]20 ) +21 (c 0.33, CHCl3), H. melanodocia
D
isolate lit.14 [R]25D ) +25.4 (c 0.24, CHCl3), Prorocentrum lima isolate
lit.22 [R]D ) +16.8 (c 0.83, CHCl3)]; TLC (Baker 7011-4 Si-HPF TLC),
indistinguishably cochromatographed with naturally occurring 1 isolated
JA973286P