The total synthesis of eleutherobin

Article abstract of DOI:10.1021/ja990215c

The total synthesis of the title compound (1), starting with (R)-( - )-α-phellandrene (6), has been accomplished. The synthesis rigorously proves the relative stereochemical relationship of the diterpenoid and carbohydrate domains of eleutherobin. Key reactions included a Nozaki - Kishi ring closure to produce a furanophane (see 37 → 38), a pyranose to furanose transposition (see 50 → 47), and a novel oxycarbaglycosidation (cf. 58 → 87) for joining the two domains.

Full text of DOI:10.1021/ja990215c

J. Am. Chem. Soc. 1999, 121, 6563-6579
6563
The Total Synthesis of Eleutherobin
Xiao-Tao Chen,† Samit K. Bhattacharya,† Bishan Zhou,† Clare E. Gutteridge,†,§
Thomas R. R. Pettus,† and Samuel J. Danishefsky*,†,‡
Contribution from the Department of Chemistry, Columbia UniVersity, HaVemeyer Hall,
New York, New York 10027, and the Laboratory for Bioorganic Chemistry,
Sloan-Kettering Institute for Cancer Research, 1275 York AVenue, New York, New York 10021
ReceiVed January 21, 1999
Abstract: The total synthesis of the title compound (1), starting with (R)-(-)-R-phellandrene (6), has been
accomplished. The synthesis rigorously proves the relative stereochemical relationship of the diterpenoid and
carbohydrate domains of eleutherobin. Key reactions included a Nozaki-Kishi ring closure to produce a
furanophane (see 37 f 38), a pyranose to furanose transposition (see 50 f 47), and a novel oxycarbagly-
cosidation (cf. 58 f 87) for joining the two domains.
The demonstration that paclitaxel (taxol) (2) is a valuable
resource at the clinical level in the treatment of human ovarian
and breast tumors has had significant consequences in medi-
cine.1,2 The use of paclitaxel continues to expand as there is
growing evidence that it may provide patients benefit against a
host of other cancers.3 As a consequence of these life-extending
advances, the clinical success of paclitaxel has inspired searches
for other drugs which might operate through related modalities
of action. Central to the quest for paclitaxel-inspired new agents
is the seminal recognition of Horwitz and associates that the
mode of action of the parent drug involves the inhibition of
disassembly of microtubules.4 In the absence of findings to the
contrary, it is assumed that the in vitro mode of action
established by Horwitz is also germane to the clinical efficacy
of the agent. Hence, the search for other substances operating
in the paclitaxel tradition involves, in the first instance, screening
at the level of microtubule assembly stabilization concurrently
with evaluations of cytotoxicity.
extensive, multidisciplinary research effort including fermenta-
tion, structure proof, total synthesis,7 and pharmacology.8
As interest in the epothilones was growing at a variety of
levels, another natural product series substantially sharing the
paclitaxel mode of action was discovered.9 Thus, several
structurally related natural products, classifiable as “eleuthe-
sides”, were identified. These compounds were shown to exhibit
potent antitumor properties. Two compounds, identified some
time ago, which bear the eleutheside skeleton, are valdivone10
and sarcodictyin.11 Of the newer structures in this general family,
the one which interested us the most was eleutherobin. The
habitat from which eleutherobin was isolated is a rare alcynacean
identified as an Eleuthorobia species in marine soft corals found
in Western Australia.9 The isolation of natural products from
these coral sources is not a simple matter. This being the case,
there seemed to be a possible opportunity for total synthesis to
play a role in providing access not only to analogues but also
to eleutherobin itself (vide infra).
From a diverse array of natural sources, other agents have
been discovered. These include discodermolide (3)5 and
epothilones (4),6 which have subsequently been proven to
display a paclitaxel-like mechanism of action. The study of these
potential drugs, particularly the epothilones, has inspired an
Furthermore, the structure of eleutherobin, shown to be 1,
contains (as will be seen) an assortment of challenges to the
science of chemical synthesis. In addition to an aglycon sector
(7) (a) Meng, D.; Bertinato, P.; Balog, A.; Su, D.-S.; Kamenecka, T.;
Sorensen, E. J.; Danishefsky, S. J. J. Am. Chem. Soc. 1997, 119, 10073.
(b) Nicolaou, K. C.; Ninkovic, S.; Sarabia, F.; Vourloumis, D.; He, Y.;
Vallberg, H.; Finlay, M. R. V.; Yang, Z. J. Am. Chem. Soc. 1997, 119,
7974. (c) Schinzer, D.; Limberg, A.; Bauer, A.; Bohm, O. M.; Cordes, M.
Angew. Chem., Int. Ed. Engl. 1997, 36, 523. (d) May, S. A.; Grieco, P. A.
Chem. Commun. 1998, 1597. (e) Mulzer J.; Mantoulidis, A.; O¨ hler, E.
Tetrahedron Lett. 1998,39, 8633. (f) Sinha, S. C.; Barbas, C. F., III; Lerner,
R. A. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 14603.
(8) (a) Bollag, D. M.; McQueney, P. A.; Zhu, J.; Hensens, O.; Koupal,
L.; Liesch, J.; Goetz, M.; Lazarides, E.; Woods, C. M. Cancer Res. 1995,
55, 2325. (b) For desepothilone B, see: Chou, T.-C.; Zhang, X.-G.; Balog,
A.; Su, D.-S.; Meng, D.; Savin, K.; Bertino, J. R.; Danishefsky, S. J. Proc.
Natl. Acad. Sci. U.S.A. 1998, 95, 9642.
(9) (a) Fenical, W. H.; Jensen, P. R.; Lindel, T. (UC) U.S. Pat. 5473057,
1995; Chem. Abstr. 1996, 102, 194297z]; . (b) Lindel, T.; Jensen, P. R.;
Fenical, W.; Long, B. H.; Casazza, A. H.; Carboni, J.; Fairchild, C. R. J.
Am. Chem. Soc. 1997, 119, 8744. (c) Long, B. H.; Carboni, J. M.;
Wasserman, A. J.; Cornell, L. A.; Casazza, A. M.; Jensen, P. R.; Lindel,
T.; Fenical, W.; Fairchild, C. R. Cancer Res. 1998, 58, 1111.
(10) (a) Kennard, O.; Watson, D. G.; di Sanseverino, L. R.; Tursch, B.;
Bosmans, R.; Djerassi, C. Tetrahedron Lett. 1968, 2879. (b) Lin, Y.; Bewley,
C. A.; Faulkner, D. J. Tetrahedron 1993, 49, 7977.
† Columbia University.
‡ Sloan-Kettering Institute for Cancer Research.
§ Current address: Merck Research Laboratories, Rahway, NJ 07065.
(1) (a) Favin, V. The Chemistry and Pharmacology of Paclitaxel and its
DeriVatiVes; Elsevier: New York, 1995. (b) Taxane Anticancer agents:
Basic Science and Current Status; Georg, G. I., Chen, T. I., Ojima, I., Vyas,
D. M., Eds.; ACS Symposium Series 583; American Chemical Society:,
Washington, DC, 1995. (c) Rowinsky, E. K.; Eisenhauer, E. A.; Chaudhary,
V.; Arbuck, S. G.; Donehower, R. C. Seminars Oncol. 1993, 20, 1. (d)
Rose, W. C. Anti-Cancer Drugs 1992, 3, 311.
(2) (a) Gan, Y.; Wientjes, M. G.; Au, J. L. S. Clin. Cancer Res. 1998,
4, 2949. (b) Blagosklonny, M. V.; Schulte, T.; Nguyen, P.; Trepel, J.;
Neckers, L. M. Cancer Res. 1996, 56, 1851.
(3) (a) Yoo, D. Y.; Park, J. K.; Choi, J. Y.; Lee, K. H.; Kang, Y. K.;
Kim, C. S.; Shin, S. W.; Kim, Y. H.; Kim, J. S. Clin. Cancer Res. 1998, 4,
3063. (b) Yeo, W.; Leung, T. W. T.; Chan, A. T. C.; Chiu, S. K. W.; Yu,
P.; Mok, T. S. K.; Johnson, P. J. Eur. J. Cancer 1998, 34, 2027.
(4) (a) Horwitz, S. B. Trends Pharmacol. Sci. 1992, 13, 134. (b) Schiff,
P. B.; Fant, J.; Horwitz, S. B. Nature 1979, 277, 665.
(5) (a) Gunasekara, S. P.; Gunasekara, M.; Longley, R. E. J. Org. Chem.
1990, 55, 4912. (b) Hung, D. T.; Chen, J.; Schreiber, S. L. Chem. Biol.
1996, 3, 287.
(6) Ho¨fle, G.; Bedorf, N.; Steinmetz, H.; Schomburg, D.; Gerth, K.;
Reichenbach, H. Angew. Chem., Int. Ed. Engl. 1996, 35, 1567.
(11) (a) D’Ambrosio, M.; Guerriero, A.; Pietra, F. HelV. Chim. Acta 1987,
70, 2019. (b) D'Ambrosio, M.; Guerriero, A.; Pietra, F. HelV. Chim. Acta
1988, 71, 964.
10.1021/ja990215c CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/30/1999

6564 J. Am. Chem. Soc., Vol. 121, No. 28, 1999
Chen et al.
Figure 1.
which had hitherto not been assembled12 in the laboratory,
eleutherobin displays a somewhat novel urocanic ester linkage
as well as an arabinose domain.
to generate enough of the agent for early in vivo evaluations.
While more drug could eventually be amassed from its natural
sources, this difficult and costly venture would most likely be
undertaken only if the biological performance of eleutherobin
would justify such a re-isolation. For these reasons, and in
keeping with our long-standing interest in natural products which
appear to owe their activity to tubulin binding, we undertook a
total synthesis of eleutherobin.13
Synthetic Planning and Strategy. Since, as discussed above,
we could not at the time be fully confident about the absolute
stereochemistry of eleutherobin, a scheme which could be
readily adapted to deliver either antipode of the natural agent
would be desirable. Similarly, since the relative relationships
of the arabinosidal and terpenoid domains were not known with
acceptable rigor, we looked forward to the possibility of
permuting the independently synthesized domains. Hopefully
in that way we could evaluate which hybrid structure in fact
corresponds to the “real” eleutherobin. Ever mindful of the goal
of producing enough eleutherobin for early in vivo evaluations,
we focused on schemes which held promise to be fairly concise
in their execution and would lead to significant rather than
symbolic amounts of final product.
While expression 1 was suggested by Fenical and co-workers
to properly represent the structure of eleutherobin, the degree
of rigor in the assignment was not uniform.9b Thus, NMR
experiments solidly defined the relative stereochemistry within
the diterpenoid and carbohydrate segments. However, as the
authors pointed out, there were no through space correlations
to relate the stereochemistry of the aglycon and the arabinose
domains. Thus, in principle, either segment might in fact be
enantiomeric with that shown in expression 1. One such
possibility is considered in some detail below (see neoeleuth-
erobin, 94). Still another possibility could be one in which both
sectors would be enantiomeric with those shown. In that case
the issue would be that of the absolute configuration of
eleutherobin itself. Here again, the authors did not provide fully
convincing data as to the absolute configuration of the natural
product.9b
Regardless of these unaddressed issues of structure, the
biological data for eleutherobin already rendered it of consider-
able interest.9c Thus, at the level of gross cytotoxicity, eleuth-
erobin exhibited an IC50 (10-15 nM) value which puts it
squarely in the taxol range. Moreover, eleutherobin was
competitive with paclitaxel in terms of microtubule assembly
stabilization and was shown to competitively bind in the
paclitaxel-binding domain.
To begin the total synthesis program, we started with the
operational proposition that its terpenoid sector is as set forth
in structure 1. Accordingly, an intriguing choice presented itself.
The starting material we came to favor was (R)-(-)-R-
phellandrene.14 We note that the particular choice of (R)-(-)-
Given the difficult availability of eleutherobin from natural
sources as noted above, it would remain for chemical synthesis
(13) Some of the results contained herein were published as preliminary
communications shortly after the publication of the first total synthesis of
eleutherobin (ref 12b): (a) Chen, X.-T.; Zhou, B.; Bhattacharya, S. K.;
Gutteridge, C. E.; Pettus, T. R. R.; Danishefsky, S. J. Angew. Chem., Int.
Ed. Engl. 1998, 37, 789. (b) Chen, X.-T.; Gutteridge, C. E.; Bhattacharya,
S. K.; Zhou, B.; Pettus, T. R. R.; Hascall, T.; Danishefsky, S. J. Angew.
Chem., Int. Ed. Engl. 1998, 37, 185.
(14) The proof of the absolute configuration of (R)-(-)-R-phellandrene
involves the convergence of several independent lines of experimentation
and is secure. See: (a) Klyne, W.; Buckingham, J. Atlas of Stereochemistry;
Oxford University Press: New York, 1974; p 78. (b) For a crystallographic
determination of a carbonylative ring-enlarged product of (R)-(-)-R-
phellandrene complexed to Fe(CO)3, see: Eilbracht, P.; Hittinger, C.;
Kufferath, K.; Henkel, G. Chem. Ber. 1990, 123, 1071.
(12) Near the end of the course of our studies (see ref 13), the total
syntheses of sarcodictyins, eleutherobin, and eleuthosides were described
by Professor K. C. Nicolaou and colleagues. (a) For the total synthesis of
sarcodictyin A, see: Nicolaou, K. C.; Xu, J. Y.; Kim, S.; Pfefferkorn, J.;
Ohshima, T.; Vourloumis, D.; Hosokawa, S. J. Am. Chem. Soc. 1998, 120,
8661. (b) For the total syntheses of eleutherobin and eleuthosides, see:
Nicolaou, K. C.; van Delft, F.; Ohshima, T.; Vourloumis, D.; Xu, J. Y.;
Hosokawa, S.; Pfefferkorn, J.; Kim, S.; Li, T. Angew. Chem., Int. Ed. Engl.
1997, 36, 2520. Nicolaou, K. C.; Ohshima, T.; Hosokawa, S.; van Delft,
F.; Vourloumis, D.; Xu, J. Y.; Pfefferkorn, J.; Kim, S. J. Am.Chem. Soc.
1998, 120, 8674.

Total Synthesis of Eleutherobin
J. Am. Chem. Soc., Vol. 121, No. 28, 1999 6565
double bonds of 12 from the trisubstituted olefin (C11-C12)
(see Figure 3). The furanophane 12 could be derived from 10
via a two-stage interpolation of 11 (wherein groups W and Z
are not specified). To generate the dihydrofuran sector (14), the
plan thus called for delivery of a methyl nucleophile (possibly
as a methyl Grignard or methyllithium agent) to C7 of the
enedione 13. Of course, provision to direct the desired nucleo-
philic methylation to this site rather than to C4 would be
necessary.
Fortunately, an attractive solution, which simultaneously
addressed both of the issues of chemoselectivity, presented itself.
This strategy involved the prospect that the substrate for
epoxidation would contain a free hydroxyl at C8 (see structure
15). Hopefully, this hydroxyl center would direct the oxidant
to the proximal unsaturation.17 In this way, the vulnerability of
the furan to oxidation relative to that of the C11-C12 olefinic
linkage would be enhanced. Moreover, the C8 hydroxyl in
structure 17 would be expected to engage the resulting C4
ketone in a hemiacetal linkage thereby accomplishing chemod-
ifferentiation between carbons 4 and 7 (see structure 18). We
hoped that at a strategic point, the pyranose resulting from this
site specific nucleophilic methylation could be opened (cf. 19)
and re-closed to provide furanose 20. The latter might eventually
be converted to 1. Undefined, for the moment, is the functional-
ity at C3 as the synthesis unfolds. The key point is that C3 be
maintained in a form where it would not compete with the
important directing effect, discussed above, which would be
confined to the C8 hydroxyl group (see structure 15). Integration
of these considerations led us to the program for synthesis
contemplated in Figure 4, wherein R′ is not specifically defined.
We now describe the translation of these conjectures to practice
in the context of a total synthesis of eleutherobin.
Our synthesis commenced (Scheme 1), as projected, with the
reaction of dichloroketene (7) (generated as shown by zinc
induced reductive elimination of trichloroacetyl chloride) with
the commercially available (R)-(-)-R-phellandrene (6).15 In our
early experiments we utilized pure samples of 6. Under these
conditions we could readily identify two products which were
the desired 8 and, surprisingly, its stereoisomer 21 in an 8:1
ratio. Apparently, no regioisomers of these products were
produced. Compounds 21 and 8 were separated by silica gel
chromatography and the structures assigned through spectro-
scopic means.18
Figure 2.
R-phellandrene (6) would not be in keeping with our preference
for a single synthesis which could accommodate the absolute
configuration (to be determined eventually) of eleutherobin, by
drawing upon either enantiomer of the starting material.
Apparently, R-phellandrene is only available commercially in
the R- configuration as shown in expression 6. However, this
antipode, already containing the trisubstituted double bond,
seemed to be sufficiently promising in terms of our ultimate
destination as to warrant special consideration. In embarking
on this course, we envisioned that if the aglycon section of
eleutherobin were enantiomeric with that which was formulated
by Fenical,9 it would be possible to use other entries from the
chiral pool to gain access to the ent versions of the series that
would be forthcoming from (R)-(-)-R-phellandrene.
Another key supposition in our planning exercise was that
the disubstituted double bond of R-phellandrene, though some-
what more proximal to the potentially hindering isopropyl group,
could be distinguished as the site of chemical reactivity. In this
regard, we favored reaction types which would be responsive
to the degree of substitution of the double bond itself and might
be responsive to directing effects imposed on the double bond
through its conjugation with the trisubstituted linkage. An
attractive subtarget would be cyclobutanone 9, which would
have to be introduced anti to the proximal isopropyl function.
A candidate reaction for gaining access to the desired bicyclo
[4.2.0] ring system would be a 2 + 2 cycloaddition of
R-phellandrene with dichloroketene (7).15 It was anticipated that
this type of reaction would be quite sensitive to the degree of
substitution on the double bond. Furthermore, we expected the
regiochemical course of the cycloaddition step to be controlled
by the conjugating effects of the trisubstituted double bond.
Hence it was projected that in such a cycloaddition process,
initial bond formation would occur through electrophilic attack
of the ketene carbonyl group at the olefinic center adjacent to
the isopropyl function. Accordingly, 8 would be the expected
product. Reduction of the chlorine groups would afford 9.
From this point, we envisioned a fragmentation to produce a
structure written as 10. The vagueness implied in notations X
and Y in structure 10 is deliberate. At this relatively unstructured
level of planning, we avoid commitment to any particular pattern
of terminal functionality in this expression. We shall address
all the critical particulars encompassed in this broad conceptual
framework as the plan gains coherence.
As the projected route gained continuing experimental
validation, we had need for substantial quantities of 8. Accord-
ingly, we took recourse to relatively inexpensive bulk com-
mercial offerings of crude phellandrenes.19 Purification was
deferred to the stage of compound 9, obtained by reductive bis-
dechlorination of 8.
Considerable thought was given to the way in which the
cyclobutanone could be exploited. We were committed at the
outset to the concept generalized in the expression 10 + 11 f
12. Of course, it would be critical to obtain the equivalent of
(16) (a) Adger, B. M.; Barett, C.; Brennan, J.; McKervey, M. A.; Murray,
R. W. Chem. Commun. 1991, 1553. Lefebvre, Y. Tetrahedron Lett. 1972,
133. (b) Achmatowicz, O.; Bukowski, P.; Szechner, B.; Zwierzchowska,
Z.; Zamojski, A. Tetrahedron 1971, 27, 1973. (c) Cavill, G. W.; Laing, D.
G.; Williams, P. J. Aust. J. Chem. 1969, 22, 2145.
(17) (a) Henbest, H. B., Wilson, R. A. L. J. Chem. Soc. 1957, 1958. (b)
For a previous instance of hydroxyl-directed epoxidation utilizing dimeth-
yldioxirane, see: Chow, K.; Danishefsky, S. J. J. Org. Chem. 1990, 55,
4211.
Upon studying the matter further, we considered the pos-
sibility that the 2,5-dihydrofuran sector (14) could arise from a
formal enedione such as 13 which could in turn be derived by
epoxidation of a furan moiety (12) and subsequent bond
reorganization.16 We were not unmindful, however, that dif-
ficulties could be encountered in distinguishing the furanoid
(15) (a) For a ketene cycloaddition with a cyclohexadiene, see the
following: Greenlee, M. L. J. Am. Chem. Soc. 1981, 103, 2425. (b) For a
diastereoselective ketene cycloaddition, see: Kanazawa, A.; Delair, P.;
Pourashraf, M.; Greene, A. E. J. Chem. Soc., Perkin Trans. 1 1997, 13,
1911.
(18) The fact the two isomers in question were indeed stereoisomers
and not regioisomers was proven by HMBC assignments on both 8 and
21.
(19) (R)-(-)-R-Phellandrene, 50% purity, is available from Fluka Chemi-
cal Corp., 1001 West St. Paul Ave., Milwaukee, WI 53233.

6566 J. Am. Chem. Soc., Vol. 121, No. 28, 1999
Chen et al.
Figure 3.
10 in a fashion where the terminii symbolized as X and Y were
cleanly differentiated in their chemical proclivities. Furthermore,
access to the specific version of 10 had to be uncomplicated. It
was from these perspectives that we turned to the elegant
chemistry of Trost20 involving the dimethylaminomethylenation
of a cyclobutanone with a Bredereck reagent.21 This reaction
accomplishes the functional equivalent of R-formylation of a
strained ketone. In the case at hand, subjection of 9 to the action
of Bredereck reagent (22) afforded 23 in 75% yield. Pursuing
the “formyl” equivalency still further, compound 23 was treated
under conditions (p-TsOH‚H2O-MeOH at 60 °C) which were
expected to favor methanolysis. This step was followed by an
exchange reaction with acetone which accomplished de-acetal-
ization of the first formed 24, providing 25 in 60% overall yield
from 23.
predicting the stereochemical outcome of this coupling. Least
predictable was the conformation of the aldehyde group entering
in the reaction. While some rotamers held out the prospect of
face selectivity in the addition step, in other hypothetical
conformers, factors favoring the two possible diastereotopic
modalities of attack seemed to be quite balanced. In a “worst
case” scenario, it was hoped that the wrong (8R) facial
stereoisomer could be rehabilitated in the scheme through direct
inversion or via oxidation followed by reduction to provide the
desired (8S) alcohol.
In the event, coupling did occur giving rise to a 1.3:1 mixture
of diastereomers.23 The “major” product, isolated in 57% yield,
was shown to be the required 28 on the basis of arguments
discussed below (vide infra). Its hydroxyl function could be
protected as a tert-butyl diphenylsilyl ether derivative (31) (97%
yield). Small and variable amounts of another isomer formulated
as the undesired 29 were also obtained. However, the bulk of
the (8R) product suffered lactonization, emerging as 30 (30-
40%). Unlike the situations with 28 and 29, the stereochemistry
of this compound could be rigorously assigned.24 When the
lactone 30 was subjected to the action of Triton-B and the
resultant salt treated with methyl iodide, the “minor” hydroxy
ester was obtained. It was in this way that we confirmed that
the latter has the stereochemistry shown in 29. By inference,
the major hydroxy ester is shown to be 28. The rigorous but
somewhat indirect structural assignment of 28 was subsequently
supported by a crystallographic determination of advanced
intermediate 45 (vide infra) in which the stereochemistry at C8
had been encoded.25 Though the configurational assignments
were secure, the synthesis suffered badly from a lack of stereo-
control in the first stage of connecting the phellandrene-derived
domain with the furan-derived nucleophile (cf. 10 + 11).
Figure 4.
As a specific version of the generalized system 11, we studied
the usefulness of readily available 2,5-dibromofuran (26).22
Monometalation of this compound was accomplished via
reaction with n-BuLi in THF. Lithio derivative 27, thus
generated, reacted with aldehyde ester 25. This coupling step
turned out to be the worst phase of the total synthesis
undertaking. From the beginning we had no clear rationale for
(23) The 1.3:1 ratio reflects a ratio of 28:[29 + 30].
(24) 1D-selective NOESY on the lactone, 30, led to the assignment of
the critical C8 center. The C8, C10, and C1 H's showed mutual NOE
enhancement upon irradiation of each individual one. Furthermore, the C1
H and the isopropyl methyls showed enhancement, thus proving that the
C1 H and therefore C10 and C8 as well were all â.
(20) Trost, B. M.; Preckel, M.; Leichter, L. M. J. Am. Chem. Soc. 1975,
97, 2224.
(21) (a) Bredereck, H.; Effenberger, F.; Simchen, G. Chem. Ber. 1963,
96, 1350. (b) For a review, see: Abdulla, R. F.; Brinkmeyer, R. S.
Tetrahedron 1979, 35, 1675.
(22) (a) For preparation, see: Keegstra, M. A.; Klomp, A. J. A.;
Brandsma, L. Synth. Commun. 1990, 20, 3371. (b) For use, see: Wellmar,
U.; Ho¨rnfeldt, A.-B.; Gronowitz, S. J. Heterocycl. Chem. 1995, 32, 1159.
(c) For 5-iodo-2-furylmagnesium iodide see: Gilman, H.; Mallory, H. E.;
Wright, G. F. J. Am. Chem. Soc. 1932, 54, 733. (d) For 5-bromo-2-
thienyllithium, see: Shiao, M.-J.; Shih, L.-H.; Chia, W.-L.; Chau, T.-Y.
Heterocycles 1991, 32, 2111.
(25) We considered the possibility of integrating the C8-R products (29
and 30) into the main synthetic stream. As shown in Scheme 1, 30 could
be converted to 29. Indeed the oxidation of 29 could be conducted but the
resultant ketone gave a mixture of streosiomers at C8 upon attempted
reduction under several conditions.

Total Synthesis of Eleutherobin
J. Am. Chem. Soc., Vol. 121, No. 28, 1999 6567
Scheme 1a
a (a) trichloroacetyl chloride, Zn, Et2O, sonication, 0 °C, 65%; (b) Zn, MeOH, NH4Cl, 87%; (c) 22, 60 °C, 75%; (d) p-TsOH‚H2O, MeOH, 60
°C; (e) p-TsOH‚H2O, acetone, 60% for (d and e); (f) 2,5-dibromofuran (26) + n-BuLi, THF, -78 °C f (27), 27 + 25 in THF f 28, 57%; (g)
TBDPSCl, imidazole, DMAP, 0 °C, 97%; (h) Triton B, THF; MeI.
Scheme 2a
a (a) t-BuLi, THF:pentane:Et2O (4:1:1), -110 °C; Bu3SnCl, ∼80%; (b) i DIBAL-toluene, CH2Cl2, -78 °C, >95%; ii Dess-Martin periodinane,
∼90%; (c) i DIBAL-toluene CH2Cl2 -78 °C, >95%; ii MsCl, pyridine, DMAP, 0 °C, >95%; iii KCN, 18-c-6 ether, CH3CN, 80 °C, 96%; (d)
DIBAL-toluene, toluene, -78 to 0 °C, 84%.
Similarly frustrating were the results of a series of attempted
olefination reactions conducted on aldehdye 32, derived from
ester 31, as shown (Scheme 2).26 Our original plan contemplated
extension of this aldehyde through hetero-branched olefination
to give rise to a system of the type 33. The thought was that
the tri-n-butylstannyl function in 33 would facilitate formation
of a furanocycle corresponding to 12 (see 32 f 33 f 34).
Unfortunately, the Wittig olefination phase failed. Another
failure was sustained in the attempted olefination of aldehyde
32 with a view to reaching 35, which also seemed to be of some
potential.27 Once again, this modified Wittig reaction failed. A
series of such projected olefinations using compounds 32 itself
or closely related congeners on related systems where the
functionality on the furan was somewhat different were equally
unsuccessful. Depending on the phosphorus-based reagent used,
reactions failed either for a lack of reactivity of the aldehyde
or due to surprising vulnerability of the furan. A listing of these
failed reactions is given in the Supporting Information.
At this stage, we returned to ester 31 and successfully carried
out a one-carbon extension to reach nitrile 36 and thence the
aldehyde 37, as shown. In keeping with the governing concep-
tion of 10 + 11 f 12, all efforts were now directed to achieving
carbon-carbon bond formation between the bromine-bearing
carbon of the furan (C4) and C3 of the elongated constructs 36
or 37.
(26) The bromo ester 31 was converted to the stannylated ester 31a, as
shown in Scheme 2 by the action of t-BuLi at -110 °C followed by
immediate quenching with Bu3SnCl. A 6:1 ratio of the stannylated and the
debrominated compound resulted. These could be separated by chroma-
tography.
(27) For instance, Stille couplings of a 1,1-disubstituted vinylidene
dibromide have been carried out: Shen, W.; Li, W., private communication.
Among the possibilities which were investigated were (i) free
radical cyclization, (ii) attempted lithiation induced cyclization,
and (iii) samarium(II) iodide mediated reductive cyclization.

6568 J. Am. Chem. Soc., Vol. 121, No. 28, 1999
Chen et al.
Scheme 3
oxidation of the furan when the C8 hydroxyl was protected,
either with m-chloroperoxybenzoic acid or with DMDO, were
unsuccessful. There were encountered significant interferences
arising from the oxidation of the trisubstituted double bond.
However, with 40 in which the C8 hydroxyl is free and DMDO
as the oxidizing agent, this competition was avoided.
We next turned to the nucleophilic methylation at C7. In our
original published approach,13 compound 41 itself was subjected
to action of methyllithium. This reaction, indeed, provided a
42% yield of 42. Although the yield was somewhat disappoint-
ing, the process was apparently stereoselective as expected. The
nucleophile approached unhindered from the R-face of the
carbonyl group as shown in the Chem 3D representation of 41
in Scheme 4.
We now had to address the matter of converting the pyranose
to the furanose. Our scheme contemplated opening of the
hemiacetal engagement between C8 and C4. The liberated C8
hydroxyl would be trapped (with the trapping electrophile, P+).
A new hemiacetal joining the tertiary alcohol at C7 and the
newly unveiled ketone function at C4 would be established (see
42 f 43 f 44). If such a valence tautomerization could be
accomplished, it would then be necessary to methylate the
hemiacetal like hydroxyl at C4 (see compound 46, Scheme 5).
Indeed, as described in our previous report,13 these steps could
be reduced to practice (Scheme 5). We elected to explore the
possibility of acetylation as a reaction to differentiate the
secondary hydroxyl at C8 and the tertiary hydroxyl at C7 in
hypothetical intermediate 43. In the event, this reaction was
successful leading to acetate 45 (44, P ) acetyl) presumably
via 43 (P ) acetyl). Fortuitously, the acetate deriVed from this
reaction was solid and an X-ray crystallographic determination
performed on crystals of 45 proVided unambiguous proof of
structure as well as all the stereochemical assignments posited
thus far.
Remarkably, it proved possible to methylate the tertiary
hydroxyl in this potentially labile system through the action of
silver oxide-methyl iodide (see compound 46). At this stage
we had need to modify the protection pattern at the C8 oxygen
to correspond to a stable arrangement that would ensure its
survival in the future reactions which were contemplated.
Toward this end, de-acylation was selectively accomplished at
C8 of 46 and the hydroxyl group in the resultant 47 was
converted to its tert-butyldimethylsilyl derivative 48.
Though a great deal of progress had been achieved, the
cumbersome and mediocre yielding nature of our route came
to be of increasing concern as we sought to bring through sub-
stantial quantities of material in anticipation of in vivo biological
evaluation. With these concerns in mind, we investigated a
bolder strategy for the pyranose f furanose interconversion.
Toward this end, we returned to the action of DMDO on
compound 40. The purified product (cf. 41) was converted to
its trimethylsilyl derivative 49. Reaction of this compound with
methyllithium provided, as expected, compound 50 in excellent
yields (unlike the poor yields encountered in substrate 41 bearing
the C4 free hydroxyl group). Remarkably, treatment of this
product with methanol in the presence of catalytic p-TsOH‚
H2O afforded the previously encountered 47. The overall yield
from 41 f 47 was ca. 80%. A mechanism to account for the
siloxy-mediated valence isomerization is implied in Scheme 6.
These attempts were all unsuccessful. Presumably, these and
related failures listed in the Supporting Information reflect the
highly strained character of the target.
Fortunately one workable ring closure method was discovered
(Scheme 3). This thrust involved formation of a m-cyclophane
derivative from a Nozaki-Kishi28 reductive cyclization of
bromoaldehyde 37, thus enabling the C3-C4 bond construction
(see compound 38). Moreover, the reaction occurred with
excellent selection in the formation of the stereogenic center at
C3.29 While this carbon was destined to become a ketone as
the synthesis unfolded, it was certainly a considerable conven-
ience to be able to obtain 38 as a single entity from the mixture
in which it was highly enriched. That the configuration at C3
is, in fact, R was rigorously proven only after crystallographic
analysis at a later point in the synthesis.
With this critical step accomplished, the next consideration
was to maintain a clear-cut chemical differentiation between
the C8 and C3 oxygen substituents such that a free hydroxyl is
uniquely exposed at C8. Toward this end, the C3 hydroxyl was
protected as a pivaloate (see compound 39). At this stage, we
could achieve deprotection of the alcohol at C8 (see compound
40). The stage for studying the viability of the program implied
in expression 15 f 18 was at hand.
Several attempts were initiated to achieve clean oxidation of
the furan sector. The most successful involved the reaction of
compound 40 with dimethyldioxirane (DMDO) (ca. 1 equiv)
in acetone at -78 °C.30,31 Under these conditions, there was
isolated a 94% yield of the desired hydroxypyranone 41 (Scheme
4). In retrospect, the advantage of keeping the C8 hydroxyl free
had been demonstrated. Thus, attempts to achieve site specific
(28) (a) Takai, K.; Kimura, K.; Kuroda, T.; Hiyama, T.; Nozaki, H.
Tetrahedron Lett. 1983, 24, 5281. Takai, K.; Tagashira, M.; Kuroda, T.;
Oshima, K.; Utimoto, K.; Nozaki, H. J. Am. Chem. Soc. 1986, 108, 6048.
(b) Jin, H.; Uenishi, J.; Christ, W. J.; Kishi, Y. J. Am. Chem. Soc. 1986,
108, 5644. Kishi, Y. Pure Appl. Chem. 1992, 64, 343. (c) For a review of
the Nozaki-Kishi reaction, see: Cintas, P. Synthesis 1992, 248. (d) See
also: Eckhardt, M.; Bru¨ckner, R. Liebigs Ann. Chem. 1996, 473.
(29) A 15:1 mixture of isomers at C3 favoring the 3R isomer was
obtained.
(30) (a) See ref 17b. (b) Sharpless directed epoxidation conditions
(Sharpless, K. B.; Michaelson, R. C. J. Am. Chem. Soc. 1973, 95, 6136)
using catalytic VO(acac)2 and t-BuOOH were also tried but DMDO31 proved
much superior.
At the time of our initial report,13 we could advance no
rigorous evidence concerning the stereochemistry of 41 or 49.
In this uncertainty, we tentatively represented the relationship
of the bridgehead substituents as out-out.32 This arrangement
corresponds to the presumed thermodynamically most stable
(31) (a) Murray, R. W.; Jeyaraman, R. J. Org. Chem. 1985, 50, 2847.
(b) Murray, R. W.; Singh, M. Org. Synth. 1997, 74, 91.

Total Synthesis of Eleutherobin
J. Am. Chem. Soc., Vol. 121, No. 28, 1999 6569
Scheme 4
Scheme 5
Scheme 6
signals in the proton NMR spectrum. Sharpening of peaks could
be observed only at temperatures around 220 K. In the first
level of analysis, low-temperature COSY and 1D NOESY
experiments might have been construed to suggest an in-out
assignment for the TMS-protected compound 49. However,
the matter was unclear. Eventually, some crystals of the lactol
enone were obtained. Crystallographic analysis (Figure 5) in-
deed established the out-out formulation shown in 41. [Although
very good diffraction patterns were obtained from crystals of
41, the structure could not be refined beyond an R factor of
0.15].
.
p
Figure 5.
configuration (Macromodel v 5.5 calculations).33 Spectroscopi-
cally, however, it was difficult to obtain persuasive proof
regarding this stereochemical question due to the conformational
fluxionality of the ring system. This resulted in broadening of
With an excellent route to 47 and thence 48 now in hand,
we focused on reaching ketone 52. This subgoal was readily
accomplished by reductive de-pivaloylation of 48 through the
agency of diisobutylaluminum hydride (see compound 51),
followed by oxidation with TPAP as shown (Scheme 7).
Several options for proceeding from ketone 52 to eleutherobin
(1) were considered. Conceptually, the most straightforward
pathway would be to accomplish a homologation at C3 (Scheme
(32) For a discussion on out-out and in-out arrangements in bridged
compounds, see: Eliel, E. L.; Wilen S. H. Stereochemistry of Organic
compounds; Wiley: New York, 1994; pp 791-793.
(33) Macromodel v 5.5 calculations on 49: Erel (kcal/mol) ) 0 (out-
out), 12.35 (H in-OTMS out); 14.33 (H out-OTMS in).

6570 J. Am. Chem. Soc., Vol. 121, No. 28, 1999
Chen et al.
Scheme 7
led to unworkably low yields of 59. Certainly, we might have
pursued this reaction in much greater detail and, perhaps,
reached the R,â-unsaturated ester in serviceable yields. Clearly,
even if methoxycarbonylation were accomplished, it would be
necessary to reduce the ester linkage to a primary alcohol to
reach 54.
Before embarking on such a yield enhancement course, it
seemed appropriate to investigate another option. The palladium-
mediated coupling of vinyl triflates with vinylstannanes is well-
known as a route to conjugate dienes.36 Hence, we came to
wonder whether a one-carbon homologation could be achieved
via a relatively rare sp3 version of the Stille coupling.37,38 Toward
this end we synthesized p-methoxybenzyl tri-n-butylstannane
following Buchwald’s procedure.39 Remarkably, coupling was
accomplished in 55% yield using tetrakis(triphenylphosphine)-
palladium(0) in the presence of lithium chloride (Scheme 9).
Since the vinyl triflate has the E configuration (the C2-C3
double bond shown in 58), it is not surprising that the structure
of the sp3 Stille cross-coupling product is as shown in 59. In
principle, the next phase would involve deprotection at C15 to
liberate the required allylic alcohol function (see system 61).
The alcohol group would serve as the glycosyl acceptor site
for attachment of a suitable arabinosyl donor (cf. coupling of
54 and 55).
8). This one-carbon fragment would correspond to C15 of
eleutherobin. It would eventually be presented as a hydroxy-
methyl group. Clearly, the homologation must be coordinated
with introduction of the C2-C3 double bond (see system 54).
Glycosylation of acceptor 54 with a suitable arabinosyl donor
(55) followed by deprotection at C8 would establish 56.
Thereafter, the properly methylated urocanic acid acylating agent
(cf. 57) would be installed. Total deprotection (either concur-
rently or in phases) would be required to reach 1.
Scheme 8
Of course, for this scheme to be successful in terms of
reaching eleutherobin, it would be necessary to achieve dif-
ferential deprotection at C15 while retaining the methyl glyco-
side at C4. A few probes in this regard (see Scheme 9) failed
to lead to the desired 61 and seemed to underscore the
vulnerability of this methyl glycoside.
Additional serious concerns as to embarking on this otherwise
logical course of deprotection of 60 and arabinosylation of 61
arose from a concurrent investigation dedicated to probing the
likely stereoselectivity margins of the impending classical
glycosylation phase of the synthesis. In anticipation of such an
approach, we had begun to cope with the synthesis of a suitable
arabinosyl donor. It would be necessary to protect the oxygens
at C3′ and C4′ of such an arabinosyl donor. While several
approaches for generating a suitable glycosyl coupling candidate
with appropriate protection patterns at the oxygens connected
to carbons 2′, 3′, and 4′ were considered and explored, we
eventually settled on the sequence shown below in Scheme 10.40
The route started with D-arabinose (62). Following per-
acetylation (see compound 63), an ethylthio group was intro-
duced at the anomeric position, as shown in structure 64.
Cleavage of the acetates through the action of sodium methoxide
in methanol gave rise to 65, in which the cis-related hydroxyl
groups at C3′ and C4′ were engaged as an isopropylidene
linkage (see compound 66). The uniquely exposed hydroxyl
group at C2′ was protected as a TBS ether (see compound 67)
or as a SEM derivative (68). Anticipating some experiments to
be described later, L-arabinose 69 was converted in an identical
fashion to enantiomeric arabinosyl donors 70 and 71 as well as
the p-methoxybenzyl derivative 72.
Focusing at the outset on the problem of reaching a version
of 53, we came to favor the intermediacy of a vinyl triflate (58).
With the C2-C3 double bond in place, a variety of reaction
types for introducing the one-carbon C15 fragment could be
explored. In the event, the triflate 58 was synthesized from the
ketone 52. In this way, the C2-C3 double bond problem,
implied in formalism 53, had been accommodated.34
The possibility of introducing the one-carbon fragment by
palladium-induced methoxycarbonylation35 of the vinyl triflate
function was considered and briefly explored. In the event, this
reaction attempted under standard conditions (see Scheme 9)
(34) At this stage we inferred that the geometry of the vinyl triflate to
be E as shown in 58. A key point in the argument was the comparison of
the C2-H in the NMR spectrum of the vinyl triflate with that of the
corresponding H in the R,â-unsaturated aldehyde i (derived from the ketone
52, see Supporting Information for details). This was accomplished in a
two-step procedure consisting of methoxymethylenation and a subsequent
Conia reaction with singlet oxygen, see: Rousseau, G.; LePerchec, P.; Conia,
J. M. Synthesis 1978, 67. Thus, an NOE enhancement between C2-H and
C5-H was observed in i but not in the vinyl triflate 58, implying that the
C2-C3 double bond in i is Z and by inference E in the vinyl triflate (see
structure 58). Rigorous proof as to the correctness of the assignment was
achieved when the total synthesis of eleutherobin was accomplished.
(35) Cacchi, S.; Morera, E.; Ortar, G. Tetrahedron Lett. 1985, 26, 1109.
(36) (a) Scott, W. J.; Crisp, G. T.; Stille, J. K. J. Am. Chem. Soc. 1984,
106, 4630. (b) Scott, W. J.; Stille, J. K. J. Am. Chem. Soc. 1986, 108, 3033.
(37) (a) Kosugi, M.; Sumiya, T.; Ogata, T.; Sano, H.; Migita, T. Chem.
Lett. 1984, 1225. (b) Majeed, A.; Antonsen, Ø.; Benneche, T.;Undheim,
K. Tetrahedron 1989, 45, 993. (c) Cook, G. K.; Hornback, W. J.; Jordan,
C. L.; McDonald, J. H., III; Munroe, J. E. J. Org. Chem. 1989, 54, 5828.
(d) Fe´re´zou, J. P.; Julia, M.; Li, Y.; Liu, W.; Pancrazi, A. Synlett 1991, 53.
(38) For selective alkyl transfer to a vinyl iodide using internal
coordination at tin, see: Vedejs, E.; Haight, A. R.; Moss, W. O. J. Am.
Chem. Soc. 1992, 114, 6556.
(39) Buchwald, S. L.; Nielsen, R. B.; Dewan, J. C. Organometallics 1989,
8, 1593.

Total Synthesis of Eleutherobin
J. Am. Chem. Soc., Vol. 121, No. 28, 1999 6571
Scheme 9
Scheme 10a
a (a) Ac2O, pyridine, DMAP, 0 °C f rt, 100%; (b) EtSH, BF3‚Et2O, CH2Cl2, rt, 76% [R:â ) 10:1]; (c) NaOMe, MeOH, rt, 100%; (d) 2,2-
dimethoxypropane, p-TsOH‚H2O, rt, 96%; (e) (t-Bu)Me2SiCl, imidazole, DMAP, CH2Cl2, rt, 93%; (f) Me3SiCH2CH2OCH2Cl (SEM-Cl), (i-Pr)2NEt,
CH2Cl2, rt, 82%; (g) p-MeOC6H4CH2Cl (PMBCl), NaH, DMSO, 91%; (h) MeOTf (3 equiv), 2,6-di-tert-butylpyridine (3.5 equiv), CH2Cl2:Et2O
[1:2], 4 Å molecular sieves, 0 °C.
Our earliest studies aimed at introducing an axial arabinose
function were actually conducted with the L-system 72 (see
Scheme 10). Two model alcohols were evaluated as L-arabinosyl
acceptors. They were cyclohexenol and cinnamyl alcohol. A
variety of conditions were screened, and representative condi-
tions are shown in Scheme 10. As these results show, we were
unable to develop a glycosidation procedure which would
provide the required â-glycoside,41 with usable stereoselectivity
employing a thioethyl donor.
We then turned to the possibility of utilizing a trichloroace-
timidate as an arabinosyl donor. For this purpose we started
with compound 73 bearing an o-nitro benzyl group at the
anomeric center of the L-arabinose system.42 Alkylation at the
C2′ hydroxyl with SEM chloride afforded 74 which was
photolytically deprotected to provide 75. Activation of the
(41) By conventions of carbohydrate nomenclature, descriptors R and â
in the arabinose series define the configurational relationship between C1
and C4 of the pyranoside. For the nomenclature of carbohydrates, see:
McNaught, A. D. Carbohydr. Res. 1997, 297, 1. In terms of the structures
presented here, R would correspond to the equatorial anomer and â to the
axial one.
(42) Nicolaou, K. C.; Hummel, C. W.; Nakada, M.; Shibayama, K.;
Pitsinos, E. N.; Saimoto, H.; Mizuno, Y.; Baldenius, K.-U.; Smith, A. L. J.
Am. Chem. Soc. 1993, 115, 7625.
(40) (a) Tri-O-acetyl thioethyl L-arabinoside has been described before:
Pakulski, Z.; Pierozynski, D.; Zamojski, A. Tetrahedron 1994, 50, 2975.
(b) For a suitably protected thiophenyl arabinoside, see: Nicolaou, K. C.;
Trujillo, J. I.; Chibale, K. Tetrahedron 1997, 53, 8751. (c) Also see ref
12b.

6572 J. Am. Chem. Soc., Vol. 121, No. 28, 1999
Chen et al.
Scheme 11a
a (a) Me3SiCH2CH2OCH2Cl (SEM-Cl), (i-Pr)2NEt, CH2Cl2, rt, 92%; (b) hν, THF:H2O [9:1], 0 °C, 50%; (c) Cl3CCN, K2CO3, CH2Cl2, 0 °C f
rt, 81%, R:â [1:2]; (d) TMSOTf (0.1-0.4 equiv), CH2Cl2, 4 Å molecular sieves.
anomeric hydroxyl group through the reaction of potassium
carbonate and trichloroacetonitrile gave rise to 76 as a 1:2 [R:â]
mixture. To evaluate the donor system 76, we screened benzyl
alcohol and cinnamyl alcohol as potential arabinosyl acceptors.
In the event, these couplings led to semienriched ratios (favoring
the desired â-anomer) of the glycosides albeit in poor yields
(see table in Scheme 11).
At this stage we had several layers of concern. First, it was
not at all certain that a conventional glycosylation would be
possible on an aglycon (cf. 61) bearing a potentially vulnerable
allylic methyl glycoside. Moreover, if we were to pursue the
pathway adumbrated in Scheme 8, we faced the prospect of
obtaining serious anomeric mixtures in the glycosylation reac-
tion. Further complicating matters was that the separation of
these compounds tended to be quite difficult. The attendant
consequences on material throughput would be particularly
damaging.
with both D- and L-arabinose sectors. We initially approached
the problem of appending the stannyloxymethyl functionality
to a suitable arabinose sector via a Schmidt-type alkylation of
the anomeric hydroxyl with tributylstannylmethyl iodide.43 This
route was abandoned in the face of poor conversions and unac-
ceptably low yields. We then envisaged a glycosylation of a
suitable arabinosyl donor with tri-n-butylstannylmethanol (83).44
Given the glycosylation results previously described in Schemes
10 and 11, we certainly had no grounds to anticipate significant
stereoselection in the arabinosylation of 83. Of course, we could
afford the consequences of stereorandom glycosylation since
the precious aglycon was not at risk at this stage.
Scheme 13a
It was from the context of these apprehensions that a new
possibility for reaching eleutherobin presented itself. In reaching
compound 60, we had already shown that the vinyl triflate 58
is a viable substrate to undergo a novel variation of Stille
coupling with an organostannane bound to an sp3 carbon (cf.
60). Accordingly, we posed the question as to whether the
carbon to which the tin is attached could, itself, carry an entire
glycosyloxy function, in particular, the arabinose sector of
eleutherobin. In essence, we were asking whether it would be
possible to introduce a glycoside domain not to the fully
functional aglycon 61 but rather to the vinyl triflate itself. The
question is framed in more general terms in Scheme 12.
To address this issue in the context of our total synthesis, it
would be necessary to assemble glycosyloxymethyl stannanes
a (a) Bu3SnCH2OH (83), MeOTf, DTBP, CH2Cl2:Et2O [1:2], 4 Å
molecular sieves, 0 °C, [R:â ) ∼1:1], 93%; (b) i TBAF-THF, rt,
∼98%; ii Ac2O, DMAP, CH2Cl2, rt, ∼99% [DMAP ) 4-dimethylami-
nopyridine; DTBP ) 2,6-di-tert-butylpyridine].
Scheme 12
In the event, a Lo¨nn-Garegg45 glycosylation of the ethylthio
donor 67 was carried out with 83 to give a 93% yield of a ∼1:1
mixture of glycosides (84) followed by removal of the silyl
group. The components were then readily separated. For the
moment, only the axial (â) anomer41 was of interest. Installation
(43) (a) Schmidt, R. R.; Reichrath, M. Angew. Chem., Int. Ed. Engl.
1979, 18, 466. Schmidt, R. R. Angew. Chem., Int. Ed. Engl. 1986, 25, 212.
(b) Seitz, D. E.; Carroll, J. J.; Cartaya, M. C. P.; Lee, S.-H.; Zapata, A.
Synth. Commun. 1983, 13, 129.
(44) Seebach, D.; Meyer, N. Angew. Chem., Int. Ed. Engl. 1976, 15,
438.
(45) (a) Ferrier, R. J.; Hay, R. W.; Vethaviyasar, N. Carbohydr. Res.
1973, 27, 5. (b) Garegg, P. J.; Henrichson, C.; Norberg, T. Carbohydr.
Res. 1983, 116, 162. (c) Lo¨nn, H. Carbohydr. Res. 1985, 139, 105, 115.

Total Synthesis of Eleutherobin
J. Am. Chem. Soc., Vol. 121, No. 28, 1999 6573
Scheme 14a
a (a) Pd(PPh3)4 (cat.), LiCl (40 equiv), THF, 130 °C, ∼40-50%; (b) TBAF, THF, rt, 67%; (c) 89, DCC, DMAP, CHCl3, 50 °C, ∼80%; (d)
PPTS, MeOH, ∆, ∼70%.
of an acetate led to 85. In a similar way, the enantiomeric
ethylthio donor, 70 (vide supra), served to glycosylate 83
affording as expected, a 1:1 mixture of separable glycosides.
As before, the TBS group at the C2′ hydroxyl center was
removed. Again only the axial (â) anomer was carried forward.
The C2′ hydroxyl group was acetylated furnishing the potential
ent-“oxycarba” donor 86 (Scheme 13).
The stage was now set to test the oxycarbaglycosidation
concept set forth in Scheme 12. In the event, it turned out to be
possible to effect a union of the domains in this way. Thus,
reaction of vinyl triflate (58) with “oxycarbadonor” 85 gave
rise to a Stille coupling product formulated as 87 (Scheme 14).
It was assumed that nothing had occurred to affect the
stereochemistry of the C2-C3 double bond. Proof was secured
when compound 87 was eventually converted to eleutherobin
(vide infra).
In the earliest stages of this oxycarbaglycosylation, the yields
of the coupling were particularly modest. However, as devel-
opmental work went forward, some of the experimental
parameters were defined. In particular, significant concentrations
of lithium chloride were helpful.46 Careful maintenance of the
temperature and solvent conditions which are described in the
Experimental Section are critical. Happily the yields seemed to
be improving as the scale of the reaction was increased.
With this critical merger step accomplished, a clear route to
eleutherobin could now be identified for the first time. Thus,
the hydroxyl group at C8 in compound 87 was liberated by
desilylation (see compound 88). Several attempts were made
to achieve acylation of the C8-hydroxyl with the acid chlorides
derived from N-methyl urocanic acid 89.47 All of these methods
were unsuccessful though it was never fully clarified if the
failures resulted from failure to synthesize the acid chloride or,
perhaps less likely, failure to achieve coupling of the C8
hydroxyl group to the acid chloride. In the end, success was
achieved by coupling of the free acid to the C8 alcohol through
the agency of dicyclohexylcarbodiimide in the presence of
DMAP. This reaction afforded an 80% yield of compound 90.
At this stage there remained to be accomplished only the
deprotection of the oxygens at C3′ and C4′ of the arabinose
unit. This goal was attained through the action of PPTs in
methanol as shown. The acetonide was cleaved as the last step
(46) For discussions on role of LiCl as an additive in Stille reactions,
see: (a) Farina, V.; Krishnamurthy, V.; Scott, W. J. Org. React. 1997, 50,
54. (b) Farina, V.; Krishnan, B. J. Am. Chem. Soc. 1991, 113, 9585. (c)
Reference 33b.
(47) (a) See ref 10a. (b) Mawlawi, H.; Monje, M. C.; Lattes, A.; Rivie´re,
M. J. Heterocycl. Chem. 1992, 29, 1621. (c) Viguerie, N. L.; Sergueeva,
N.; Damiot, M.; Mawlawi, H.; Rivie´re, M.; Lattes, A. Heterocycles 1994,
37, 1561.

6574 J. Am. Chem. Soc., Vol. 121, No. 28, 1999
Chen et al.
of the synthesis, and the product was presumed to be fully
synthetic eleutherobin.
Professor K. C. Nicolaou.12b We can therefore confidently assert
that the relative and absolute stereochemistry of eleutherobin
is as implied in structure 1. Furthermore, the total synthesis of
this compound has thus been accomplished. Similarly, the total
synthesis of neoeleutherobin, 94, has been accomplished.
Unfortunately, we did not have an “authentic” sample of
eleutherobin available to us. While extensive representative
spectra of eleutherobin were provided by Professor Fenical, a
direct material comparison was not possible. The comparisons
were quite encouraging in terms of their virtual spectral
congruence.
Summary
Several key facets of the venture merit recapitulation. The
use of the readily available (R)-(-)-R-phellandrene (6) turned
out to be a profitable one in that it provided a properly
configured and fortuitously functionalized matrix for extensive
investigations. The use of the Nozaki-Kishi ring closure to
furanophane 38 provided a productive step in paving the way
for the critical oxidative furfuryl alcohol f pyranose transposi-
tion 40 f 41. The pyranose system served as a reliable platform
for nucleophilic methylation 41 f 42 followed by controlled
valence isomerization to a furanose (see 50 f 47). Finally, the
basis for a new method (oxycarbaglycosidation) for merging
carbohydrate and lipid domains has been demonstrated (see 58
f 87). In the course of these studies, the full structure of
eleutherobin was clarified in detail.
It was at this stage that we had to face an important issue of
intellectual rigor in using the total synthesis to support the
Fenical structure of eleutherobin. Implicit in such an application
would be the assumption that were the real eleutherobin not to
correspond to structure 1, its spectral properties would be
materially different from those of fully synthetic 1 of unambigu-
ous structure. It must be recognized that this type of surmise is
inherent in all “proofs of structure” which accrue from apparent
congruence of spectra of “authentic” and “synthetic” material.
Given the sophisticated resolving power of modern spectroscopic
measurements, this kind of assumption would seem to be
warranted. However, in the case at hand, there were two sources
of apprehension. First, we could not make a direct comparison.
We were comparing in house measurements on our synthetic
sample with data accrued elsewheresnecessarily under non-
identical conditions.
As discussed earlier, one of the goals of our program in this
total synthesis was to secure enough eleutherobin for biological
investigation in xenografts. This was indeed accomplished. The
results of both in vitro and in vivo studies with eleutherobin as
well as SAR profiles will be disclosed soon.
Furthermore, there was additional concern arising from the
bidomainal nature of the structure. Our primary focus was that
of differentiating between formulations 1 and 94 (neoeleuther-
obin, vide infra). The assumption that the spectroscopic proper-
ties of synthetic eleutherobin of unambiguous structure 1 would
be materially different from those of 94 in which the aglycon
sector is identical with 1 but the carbohydrate is enantiomeric
implies the presumption that the two domains are measurably
interactive at the level of spectroscopic readout. This need not
necessarily be the case. If each sector is fully autonomous, the
overall spectroscopic displays of eleutherobin and neoeleuth-
erobin might very well be quite similar. Given the absence of
authentic material, it seemed well to pursue this matter to a
rigorous conclusion. We were well prepared to do so since, as
described above, we had already assembled the L-arabinose
donor construct 86.
Experimental Section
General Methods. All commercial materials were used without
purification unless otherwise noted. The following solvents were
distilled under positive pressure of dry argon: tetrahydrofuran (THF),
diethyl ether (Et2O) from sodium-benzophenone ketyl, methylene
chloride (CH2Cl2) from CaH2. All the reactions were performed under
an inert (Ar) atmosphere. Spectra (1H, 13C) were recorded on AMX-
300 MHz, AMX-400 MHz, and DRX-500 MHz Bruker instruments
referenced to CDCl3 (1H NMR ) δ 7.24 and 13C NMR ) δ 77.0)
peaks unless stated otherwise. Line broadening ) 1.0 Hz was used
before Fourier transformation in 13C NMR spectra. Infrared spectra were
recorded with a Perkin-Elmer Paragon 1000 FTIR spectrometer, and
optical rotations were measured with a Jasco DIP-30 instrument using
a 10 cm length cell. Mass spectral analyses were performed with JEOL
JMS-DX-303 and NERMAG R10-10 spectrometers. X-ray crystal-
lographic analysis was performed on a Bruker P4 diffractometer using
a SMART CCD detector. Analytical thin-layer chromatography was
performed using E. Merck silica gel 60 F254 coated 0.25 mm plates.
Compounds were visualized by dipping the plates in a cerium sulfate-
ammonium molybdate solution followed by heating. Flash column
chromatography was performed using the indicated solvent on E. Merck
silica gel 60 (40-63 µm). Unless otherwise indicated all isolated
intermediates were >98% pure according to 1H and 13C NMR analysis.
Dichlorocyclobutanones 8 and 21. In a 250 mL three-necked flask,
equipped with a mechanical stirrer and an addition funnel, were placed
ether (50 mL), (R)-(-)-R-phellandrene 6 (3.5 mL, 100% purity, 21.6
mmol), and acid-washed zinc (2.82 g, 43.2 mmol) forming a gray
suspension. The apparatus was placed in a sonication bath at 0 °C.
The addition funnel was charged with a mixture of trichloroacetyl
chloride (2.89 mL, 25.89 mmol) and ether (17 mL). The contents of
the additional funnel were added in a dropwise fashion to the suspension
over 2 h with stirring and sonication. After an additional hour of
agitation at 0 °C, the reaction mixture was filtered through Celite 545
to reveal a clear pale yellow solution. The organics were concentrated
in vacuo, and the residue was taken up in ether, refiltered, and washed
with saturated solutions of NaHCO3 and NaCl. The organic layer was
dried over anhydrous Na2SO4, filtered, concentrated, and chromato-
graphed (99:1 hexanes:ether) to furnish the dichlorocyclobutanones 8
(3.4 g, 65%) and 21 (425 mg, 8%). In a similar manner the reaction
was performed on larger scales (∼100 mL of (R)-(-)-R-phellandrene,
50% purity) to yield the cyclobutanone in 60-83% yields.
In the event, reaction of compound 58 with ent-donor 86
proceeded, as was described for the D-series donor 85 (see
compound 91). The coupling yields were quite comparable. The
protocol for further progress had now been well established.
Once again, cleavage of the TBS group at position 8 liberated
the required hydroxyl group (92). Carbodiimide-mediated acy-
lation with urocanic acid derivative 89 produced 93. Finally,
the acetonide group was cleaved as before with PPTs producing
94. Needless to say, the pertinent spectra for compound 94 were
eagerly measured and the readouts carefully scrutinized. These
measurements on the L-arabinose-derived synthetic product
revealed small, but unmistakable differences with the exhibited
data provided by Fenical for “authentic” eleutherobin. The
differences, while subtle, were far more substantive than the
very slight differences between the data of the fully synthetic
compound and that reported by Fenical for authentic 1.
Moreover, with the two pure compounds in hand we could
conduct an important polarimetric measurement. Thus, [R]23
D
of fully synthetic 1 was -73.2 (c ) 0.17, MeOH). By contrast,
the fully synthetic neoeleutherobin (94) has an [R]23 under
D
comparable conditions of +44.5. It was previously reported that
[R]25D for authentic eleutherobin was -49.3 (c ) 3.0, MeOH).9a
An [R]25 of -67.0 (c ) 0.2, MeOH) had been registered for
D
the fully synthetic eleutherobin arising from the laboratories of

Total Synthesis of Eleutherobin
J. Am. Chem. Soc., Vol. 121, No. 28, 1999 6575
The following data describe the major isomer 8: 1H NMR [300 MHz,
CDCl3] δ 5.73 (br m, 1H), 3.69 (dd, J ) 10.1, 7.8 Hz, 1H), 3.25 (d, J
) 10.4 Hz, 1H), 2.06 (m, 1H), 1.90-1.70 (m, 5H), 1.67 (m, 1H), 0.95
(d, J ) 5.7 Hz, 3H), 0.88 (d, J ) 5.7 Hz, 3H); 13C NMR [75 MHz,
CDCl3] δ 197.1, 129.6, 125.7, 87.3, 56.9, 49.7, 38.4, 30.4, 24.7, 22.4,
20.7, 19.3; IR [ν max cm-1] 3020, 2965, 1803, 1523; [R]23D ) -18.4°
(c 1.00, CHCl3); HRMS (EI) calcd for [C12H16Cl2O]+ 246.0579, found
246.0590.
methanol followed by pH 7 buffer, and the reaction was permitted to
slowly warm to room temperature. After regular extractions with ether,
the organic layer was washed with brine and then dried over anhydrous
Na2SO4. The solvents were removed under reduced pressure, and the
residue was immediately chromatographed (95:5 f 90:10 hexanes:
ether) to furnish the following compounds and some recovered aldehyde
25, which was recycled. The yields are based on recovered starting
material and varied (38-57% for 28) with the scale of the reaction.
1
1
Data for the minor isomer 21: H NMR [300 MHz, CDCl3] δ 5.83
Alcohol 28 (0.88 g, 57%): H NMR [300 MHz, CDCl3] δ 6.21 (d,
(d, J ) 7.3 Hz, 1H), 4.29 (ddd, J ) 10.4, 5.5, 1.6 Hz, 1H), 3.40 (d, J
) 10 Hz, 1H), 2.25 (m, 1H), 1.91 (d, J ) 1.6 Hz, 3H and overlapping
m, 1H), 1.70 (m, 1H), 1.22 (m, 1H), 1.05 (d, J ) 6.6 Hz, 3H), 0.95 (d,
J ) 6.6 Hz, 3H); 13C NMR [75 MHz, CDCl3] δ 197.5, 131.1, 127.8,
J ) 3.3 Hz, 1H), 6.18 (d, J ) 3.3 Hz, 1H), 5.38 (br s, 1H), 4.44 (m,
1H), 3.67 (s, 3H), 2.65 (m, 1H), 2.41 (m, 1H), 2.25 (m, 2H), 1.9 (m,
5H), 1.55 (s, 3H), 0.87 (d, J ) 6.6 Hz, 3H), 0.71 (d, J ) 6.6 Hz, 3H);
13C NMR [75 MHz, CDCl3] δ 175.6, 158.9, 135.3, 121.6, 120.9, 111.8,
108.4, 66.3, 51.5, 46.8, 37.4, 35.9, 35.3, 27.7, 23.4, 22.0, 20.9, 15.7;
86.4, 56.9, 51.4, 40.9, 30.7, 27.3, 24.2, 21.5, 21.4; IR [ν max cm-1
]
IR [ν max cm-1] 3721, 3580, 1725; [R]23 ) +53.9° (c 1.5, CHCl3);
2964, 2870, 1799, 1473.
D
HRMS (EI) calcd for [C18H25BrO4]+ 384.0937, found 384.0929.
Dechlorinated Cyclobutanone 9. To a suspension of zinc (24 g,
367 mmol) and ammonium chloride (20 g, 374 mmol) in methanol
(100 mL) at 0 °C was added a solution of the dichlorocyclobutanone
8 (17.3 g, 70.0 mmol) in methanol (100 mL). The gray suspension
was permitted to warm to room temperature (∼25 °C), stirred for about
8 h, and then filtered to remove solids. After the solvent was
concentrated in vacuo, the residue was taken up in ether and washed
with water and brine and then dried over anhydrous Na2SO4. The solvent
was again removed, and the residue was chromatographed (99:1
hexanes:ether) to furnish the cyclobutanone 9 (10.8 g, 87%): 1H NMR
[300 MHz, CDCl3] δ 5.55 (br s, 1H), 3.39 (m, 1H), 3.25 (m, 1H), 2.71
(m, 2H), 2.10 (m, 1H), 1.81 (m, 1H), 1.67 (m, 2H), 1.65 (s, 3H), 0.95
(d, J ) 6.5 Hz, 3H), 0.88 (d, J ) 6.5 Hz, 3H); 13C NMR [75 MHz,
CDCl3] δ 211.3, 134.4, 122.3, 60.7, 51.0, 37.7, 30.5, 27.5, 24.9, 21.1,
Alcohol 29 (0.38 g, 24%): 1H NMR [300 MHz, CDCl3] δ 6.25 (m,
2H), 5.47 (br s, 1H), 4.51 (m, 1H), 3.68 (s, 3H), 2.67 (m, 1H), 2.31
(m, 2H), 1.96 (m, 4H), 1.83 (m, 2H), 1.71 (s, 3H), 0.91 (d, J ) 6.6
Hz, 3H), 0.73 (d, J ) 6.6 Hz, 3H); 13C NMR [75 MHz, CDCl3] δ
175.6, 158.4, 136.2, 121.2, 121.1, 111.8, 108.9, 67.8, 51.4, 47.2, 38.2,
36.5, 35.2, 27.7, 23.4, 22.2, 20.9, 15.5; IR [ν max cm-1] 3424, 2959,
1725; [R]23 ) +71.5° (c 0.41, CHCl3).
D
Lactone 30 (0.25 g, 18%): 1H NMR [400 MHz, CDCl3] δ 6.35 (d,
J ) 3.3 Hz, 1H), 6.27 (d, J ) 3.3 Hz, 1H), 5.43 (br s, 1H), 5.25 (dd,
J ) 12.1, 2.7 Hz, 1H), 2.80 (t, J ) 7.3 Hz, 1H), 2.63 (m, 1H), 2.35
(ddd, J ) 14.2, 6.3, 2.7 Hz, 1H), 2.18 (m, 1H), 1.82 (m, 3H), 1.69 (m,
1H), 1.65 (s, 3H), 0.93 (d, J ) 4.2 Hz, 3H), 0.91 (d, J ) 4.2 Hz, 3H);
13C NMR [75 MHz, CDCl3] δ 172.7, 153.3, 132.9, 122.6, 122.0, 112.1,
111.0, 72.6, 41.4, 37.2, 34.6, 30.7, 27.4, 23.7, 21.2, 21.0, 18.5; IR [ν
max cm-1] 1732; [R]23D ) +33.8° (c 0.50, CHCl3); HRMS (FAB) calcd
for [C17H22BrO3]+ 353.0753, found 353.0755.
20.8, 19.6; IR [ν max cm-1] 2963, 1769; [R]23 ) -159.2° (c 0.96,
D
CHCl3); MS (EI) 178 [M+], 136.
Vinylogous Amide 23 via Bredereck Condensation. To a round-
bottom flask equipped with a stir bar were added the cyclobutanone 9
(6.72 g, 37.7 mmol) and (t-BuO)(Me2N)2CH (50 mL, 242 mmol). The
mixture was warmed to 60 °C for 3 h, before the excess reagent was
recovered by distillation and the residue chromatographed (70:30
hexanes:EtOAc) on Florisil to furnish the vinylogous amide 23 (5.78
g, 66%) as a 2.5:1 mixture (based on integration of vinyl protons) of
geometric isomers which was taken through to the next step. Selected
1H NMR [500 MHz, CDCl3] δ 6.94 (d, J ) 1.2 Hz, 0.29H), 5.9 (d, J
) 0.6 Hz, 0.71H), 5.47 (br unresolved m, 0.29H), 5.40 (br unresolved
m, 0.71H), 3.16 (br unresolved m, 4.3H), 2.99 (s, 1.7H), 1.81 (d, J )
1.6 Hz, 0.87H), 1.72 (d, J ) 1.4 Hz, 2.13H).
Silylated Alcohol 31. A solution of alcohol 28 (8.66 g, 22.5 mmol)
in methylene chloride (50 mL) at 0 °C was treated with imidazole (9.17
g, 134.7 mmol) and TBDPSCl (17.5 mL, 67.3 mmol), and the mixture
was stirred at 25 °C for 4 h. The reaction mixture was diluted with
methylene chloride (200 mL), washed successively with H2O and brine,
dried over anhydrous Na2SO4, filtered, and concentrated in vacuo.
Chromatography (99:1 hexanes:ether) provided the TBDPS ether 31
(12.0 g, 97%): 1H NMR [500 MHz, CDCl3] δ 7.66 (m, 2H), 7.46 (m,
2H), 7.38 (m, 4H), 7.27 (m, 2H), 5.91 (d, J ) 3.2 Hz, 1H), 5.59 (d, J
) 3.2 Hz, 1H), 5.31 (br s, 1H), 4.37 (dd, J ) 8.4, 5.2 Hz, 1H), 3.44 (s,
3H), 2.56 (dd, J ) 10.6, 4.4 Hz, 1H), 2.38 (m, 1H), 2.14 (m, 1H), 2.03
(m, 1H), 1.87 (m, 3H), 1.78 (m, 1H), 1.70 (s, 3H), 1.02 (s, 9H), 0.85
(d, J ) 6.6 Hz, 3H), 0.70 (d, J ) 6.6 Hz, 3H); 13C NMR [75 MHz,
CDCl3] δ 174.7, 157.6, 136.2, 136.0, 135.4, 133.4, 133.1, 129.6, 129.1,
127.6, 127.1, 120.8, 120.4, 111.3, 109.6, 68.2, 51.2, 47.2, 37.5, 37.4,
34.8, 27.6, 26.8, 23.2, 22.3, 20.8, 19.3, 15.4 (6 unresolved); IR [ν max
Aldehyde Ester 25. The vinylogous amide 23 (12.8 g, 54.9 mmol)
and methanol (90 mL) were added to a round-bottom flask, and
p-TsOH‚H2O (10.5 g, 55.2 mmol) was added. The mixture was stirred
for 4 h at 60 °C and cooled to room temperature, and the solvent
removed in vacuo. The residue was dissolved in acetone (120 mL),
and an additional portion of p-TsOH‚H2O (10.5 g, 55.2 mmol) was
added. After stirring for 1 h, the contents were concentrated in vacuo
and the residue was dissolved in ether and washed sequentially with
saturated solutions of NaHCO3 and brine. The organic layer was then
dried over anhydrous Na2SO4, filtered, concentrated, and chromato-
graphed (96:4 f 94:6 hexanes:ether) to yield the aldehyde ester 25
(7.8 g, 60%): 1H NMR [400 MHz, CDCl3] δ 9.68 (s, 1H), 5.41 (br s,
1H), 3.57 (s, 3H), 2.99 (m, 2H), 2.65 (m, 1H), 2.33 (m, 1H), 1.85 (m,
4H), 1.62 (s, 3H), 0.88 (d, J ) 6.6 Hz, 3H), 0.71 (d, J ) 6.6 Hz, 3H);
13C NMR [75 MHz, CDCl3] δ 201.1, 175.7, 134.1, 122.3, 51.5, 46.3,
44.5, 34.9, 34.8, 27.6, 23.2, 21.7, 20.9, 15.2; IR [ν max cm-1] 2958,
cm-1] 2959, 1732; [R]20 ) -30.1° (c 0.84, CHCl3); HRMS (FAB)
D
calcd for [C34H42BrO4Si]+ 621.2036, found 621.2029.
Alcohol. To a solution of the ester 31 (11.4 g, 18.3 mmol) in toluene
(anhydrous, 100 mL) at -78 °C was added dropwise DIBAL (1.0 M
in hexane solution, 54.9 mL). After ∼1 h, acetone (3 mL), ethyl acetate
(3 mL), and aqueous buffer (pH 7, 3 mL) were sequentially added, the
cooling bath was removed, and the solution was stirred vigorously.
After 20 min, anhydrous Na2SO4 was added and the reaction mixture
was stirred vigorously for a further 30 min. The mixture was then
filtered through a pad of anhydrous Na2SO4 in a sintered glass funnel
(medium) and the solvent removed under reduced pressure. Chroma-
tography (95:5 hexanes:ether) provided the primary alcohol (10.8 g,
99%): 1H NMR [500 MHz, CDCl3] δ 7.68 (dd, J ) 7.9, 1.3 Hz, 2H),
7.49 (dd, J ) 7.9, 1.3 Hz, 2H), 7.41 (m, 1H), 7.37 (m, 3H), 7.29 (t, J
) 7.5 Hz, 2H), 5.99 (d, J ) 3.3 Hz, 1H), 5.65 (d, J ) 3.3 Hz, 1H),
5.23 (br s, 1H), 4.78 (t, J ) 7.0 Hz, 1H), 3.65 (dt, J ) 11.1, 5.6 Hz,
1H), 3.47 (dt, J ) 10.7, 4.5 Hz, 1H), 2.12 (m, 1H), 2.06 (m, 1H), 1.86
(m, 2H), 1.70 (m, 3H), 1.49 (d, J ) 1.3 Hz, 3H), 1.43 (m, 1H), 1.28
(t, J ) 5.4 Hz, 1H), 1.04 (s, 9H), 0.82 (d, J ) 6.8 Hz, 3H), 0.76 (d, J
) 6.8 Hz, 3H); 13C NMR [75 MHz, CDCl3] δ 157.7, 137.3, 136.0,
135.6, 133.3, 129.7, 129.3, 127.5, 127.2, 121.1, 120.5, 111.5, 110.2,
68.6, 62.0, 41.7, 37.0, 36.0, 34.6, 26.9, 24.1, 22.8, 21.0, 19.2, 16.0 (8
1727; [R]23 ) +59.3° (c 0.50, CHCl3); HRMS (FAB) calcd for
D
[C14H22O3]+ 238.1569, found 238.1562.
Lithiofuran Addition. A dry round-bottom flask (100 mL), equipped
with a stir bar, was charged with tetrahydrofuran (10 mL). The aldehyde
25 (1.31 g, 5.5 mmol) was added, and the mixture was cooled to -78
°C. A separate flask (100 mL), equipped with a stir bar, was charged
with tetrahydrofuran (20 mL). The dibromomofuran 26 (3.72 g, 16.5
mmol) was added and the mixture cooled to -78 °C. n-BuLi (2.5 M,
1 equiv, 6.2 mL) was added in a dropwise fashion, and after stirring
for 10 min, the contents of this flask were cannulated in dropwise into
the flask that contained the aldehyde. The resulting mixture was allowed
to stir for ∼1 h at -78 °C, then the reaction was quenched with

6576 J. Am. Chem. Soc., Vol. 121, No. 28, 1999
Chen et al.
unresolved); IR [ν max cm-1] 3429, 2954, 2858; [R]20 ) -5.3° (c
and degassed as well. This solution was then added via a cannula to
the mixture of CrCl2 and NiCl2 in DMF at -60 °C. After completion
of addition, the cooling bath was removed and the reaction mixture
allowed to stir at room temperature. After 6-8 h, water was added
and the reaction mixture partitioned between ether and water. The
aqueous phase was extracted three times with ether. The organic layer
was washed with brine, dried over anhydrous Na2SO4, filtered, and
concentrated in vacuo. The resultant oil was purified by chromatography
(95:5 f 80:20 hexanes:ether) to afford the furanophane 38 (3.77 g,
70%) and the C3-â epimer (215 mg, 4%).
D
1.34, CHCl3); HRMS (FAB) calcd for [C33H42BrO3Si]+ 593.2086, found
593.2070.
Mesylation. A solution of alcohol (12.2 g, 20.48 mmol) in pyridine
(50 mL) was cooled to 0 °C prior to sequential addition of DMAP
(300 mg) and CH3SO2Cl (4.75 mL, 61.37 mmol). The cooling bath
was removed after 5 min, and the mixture was stirred at 25 °C for 4 h.
The reaction was quenched with the addition of saturated aqueous
CuSO4 and diluted with EtOAc, and the layers were separated. The
aqueous portion was extracted with EtOAc, and the combined organic
portions were dried over anhydrous Na2SO4, filtered, and concentrated.
Chromatography (95:5 f 90:10 hexanes:EtOAc) furnished the mesylate
(12.68 g, 92%) as a colorless oil: 1H NMR [500 MHz, CDCl3] δ 7.65
(dd, J ) 7.8, 1.3 Hz, 2H), 7.46 (dd, J ) 7.8, 1.3 Hz, 2H), 7.40 (m,
1H), 7.36 (m, 3H), 7.28 (t, J ) 7.6 Hz, 2H), 6.00 (d, J ) 3.3 Hz, 1H),
5.76 (d, J ) 3.3 Hz, 1H), 5.22 (br s, 1H), 4.67 (dd, J ) 7.7, 6.6 Hz,
1H), 4.21 (dd, J ) 10.1, 6.2 Hz, 1H), 4.00 (dd, J ) 10.1, 8.0 Hz, 1H),
2.85 (s, 3H), 2.05 (m, 1H), 1.94 (m, 4H), 1.82 (m, 1H), 1.57 (m, 1H),
1.47 (d, J ) 0.9 Hz, 3H), 1.42 (m, 1H), 1.00 (s, 9H), 0.78 (d, J ) 6.6
Hz, 3H), 0.77 (d, J ) 6.6 Hz, 3H); 13C NMR [75 MHz, CDCl3] δ
157.3, 135.8, 135.7, 135.5, 133.3, 133.2, 129.7, 129.4, 127.6, 127.3,
121.4, 120.7, 111.5, 110.4, 69.9, 68.1, 39.1, 37.2, 37.0, 36.6, 34.7, 31.4,
27.2, 26.8, 23.8, 22.7, 22.6, 20.6, 19.1, 17.1, 14.1 (3 unresolved); IR
Data for the major isomer, alcohol 38: 1H NMR [500 MHz, CDCl3]
δ 7.71 (m, 4H), 7.38 (m, 6H), 6.04 (d, J ) 3.0 Hz, 1H), 5.95 (d, J )
3.0 Hz, 1H), 5.01 (br s, 1H), 4.60 (dd, J ) 9.9, 5.6 Hz, 1H), 4.46 (m,
1H), 2.56 (ddd, J ) 12.5, 9.9, 2.0 Hz, 1H), 2.39 (q, J ) 12.2 Hz, 1H),
2.10 (dt, J ) 13.2, 3.4 Hz, 1H), 2.02 (m, 1H), 1.95 (d, J ) 6.0 Hz,
1H), 1.82 (m, 1H), 1.63 (m, 1H), 1.45 (dt, J ) 13.1, 5.3 Hz, 1H), 1.34
(m, 1H), 1.23 (m, 1H), 1.08 (s, 12H), 0.88 (d, J ) 6.8 Hz, 3H), 0.71
(d, J ) 6.8 Hz, 3H), -0.56 (br s, 1H); 13C NMR [75 MHz, CDCl3] δ
160.5, 159.7, 139.1, 135.9, 135.7, 133.7, 133.6, 129.7, 129.6, 127.5,
127.4, 118.5, 116.5, 110.6, 70.6, 68.1, 43.3, 39.7, 38.4, 37.2, 36.8, 26.9,
26.8, 24.8, 22.0, 21.4, 19.1, 15.0 (6 unresolved); IR [ν max cm-1] 3336,
2955; [R]20 ) +28.7° (c 0.94, CHCl3); HRMS (FAB) calcd for
D
[C34H44O3Si]+ 528.3060, found 528.3050.
[ν max cm-1] 2960, 2858, 1590; [R]20 ) -14.0° (c 0.95, CHCl3);
Data for the minor isomer: 1H NMR [500 MHz, CDCl3] δ 7.71 (m,
4H), 7.40 (m, 6H), 6.00 (dd, J ) 3.0, 1.3 Hz, 1H), 5.97 (d, J ) 3.0 Hz,
1H), 5.17 (d, J ) 9.4 Hz, 1H), 5.03 (br s, 1H), 4.58 (dd, J ) 9.8, 5.6
Hz, 1H), 2.48 (ddd, J ) 11.5, 9.9, 2.1 Hz, 1H), 2.33 (m, 1H), 1.99 (m,
2H), 1.84 (m, 1H), 1.66 (m, 1H), 1.60 (d, J ) 9.5 Hz, 1H), 1.46 (dt,
J ) 13.1, 5.4 Hz, 1H), 1.38 (m, 2H), 1.07 (s, 12H), 0.88 (d, J ) 6.9
Hz, 3H), 0.71 (d, J ) 6.9 Hz, 3H), -0.43 (br s, 1H); 13C NMR [75
MHz, CDCl3] δ 163.8, 159.4, 139.2, 136.0, 135.8, 133.9, 133.8, 129.7,
129.6, 127.6, 127.5, 118.6, 116.9, 106.4, 69.9, 69.6, 43.0, 38.0, 36.6,
36.4, 35.1, 27.0, 26.8, 25.0, 22.1, 21.5, 19.2, 14.8 (6 unresolved); IR
D
HRMS (FAB) calcd for [C34H44BrO5SSi]+ 671.1862, found 671.1869.
Cyanide 36. To a solution of the mesylate (12.86 g, 18.82 mmol)
and 18-crown-6 (12.4 g, 46.91 mmol) in CH3CN (anhydrous, 85 mL)
was added potassium cyanide (6.13 g, 94.13 mmol), and the reaction
mixture was heated to 60 °C for 2.5 h. The mixture was then quenched
with ice-water and extracted with ether. The crude product was
chromatographed (97:3 hexanes:EtOAc) to give the nitrile 36 (10.87
g, 96%): 1H NMR [500 MHz, CDCl3] δ 7.65 (dd, J ) 7.8, 1.3 Hz,
2H), 7.43 (m, 3H), 7.37 (m, 3H), 7.28 (t, J ) 7.4 Hz, 2H), 6.04 (d, J
) 3.2 Hz, 1H), 5.80 (d, J ) 3.2 Hz, 1H), 5.20 (br s, 1H), 4.65 (t, J )
7.1 Hz, 1H), 2.17 (m, 2H), 1.96 (m, 1H), 1.87 (m, 4H), 1.81 (m, 1H),
1.50 (m, 1H), 1.42 (s, 3H), 1.40 (m, 1H), 1.00 (s, 9H), 0.76 (d, J ) 7.7
Hz, 3H), 0.75 (d, J ) 7.7 Hz, 3H); 13C NMR [75 MHz, CDCl3] δ156.8,
135.8, 135.5, 134.8, 133.2, 133.0, 129.8, 129.4, 127.6, 127.3, 121.2,
121.0, 119.3, 111.6, 110.7, 77.2, 68.3, 38.5, 36.6, 35.6, 27.2, 26.8, 23.6,
22.3, 20.4, 19.1, 17.9, 17.5 (6 unresolved); IR [ν max cm-1] 2960,
2894, 2248; [R]20D ) -17.5° (c 1.05, CHCl3); HRMS (FAB) calcd for
[C34H41BrNO2Si]+ 602.2090, found 602.2090.
[ν max cm-1] 3330, 2955; [R]20 ) -17.9° (c 0.64, CHCl3); HRMS
D
(FAB) calcd for [C34H44O3Si]+ 528.3060, found 528.3050.
Pivaloate 39. The alcohol 38 (3.77 g, 7.13 mmol) and DMAP (518
mg, 4.24 mmol) were dissolved in CH2Cl2 (50 mL), and the mixture
was cooled to 0 °C. Triethylamine (5.92 mL, 4.25 mmol) and pivaloyl
chloride (2.65 mL, 21.98 mmol) were sequentially added, and the
mixture was stirred for 5 h. The reaction mixture was then diluted with
CH2Cl2 (200 mL) and washed with aqueous buffer (pH 7, 30 mL)
followed by brine. The organic layer was dried over anhydrous Na2-
SO4, concentrated, and purified by chromatography (96:4 hexanes:ether)
to give the ester 39 (3.97 g, 91%): 1H NMR [500 MHz, CDCl3] δ
7.71 (m, 4H), 7.39 (m, 6H), 6.16 (d, J ) 3.0 Hz, 1H), 5.96 (d, J ) 3.0
Hz, 1H), 5.40 (dd, J ) 11.0, 3.7 Hz, 1H), 5.01 (br s, 1H), 4.59 (dd, J
) 9.9, 5.6 Hz, 1H), 2.58 (ddd, J ) 12.7, 10.1, 2.2 Hz, 1H), 2.51 (q, J
) 12.5 Hz, 1H), 1.96 (m, 2H), 1.83 (m, 1H), 1.65 (m, 1H), 1.45 (dt,
J ) 13.0, 5.3 Hz, 1H), 1.37 (m, 1H), 1.25 (m, 1H), 1.23 (m, 9H), 1.08
(s, 9H), 1.06 (s, 3H), 0.87 (d, J ) 6.8 Hz, 3H), 0.69 (d, J ) 6.8 Hz,
3H), -0.54 (br s, 1H); 13C NMR [75 MHz, CDCl3] δ 178.0, 160.9,
156.7, 139.0, 135.9, 135.8, 133.7, 133.6, 129.7, 129.6, 127.6, 127.5,
118.5, 116.8, 113.2, 70.6, 69.5, 43.3, 39.7, 38.7, 38.4, 37.0, 32.9, 27.1,
27.0, 24.9, 22.1, 21.4, 19.1, 15.0 (9 unresolved); IR [ν max cm-1] 2955,
Aldehyde 37. To a solution of the nitrile 36 (9.75 g, 16.2 mmol) in
toluene (200 mL) at -78 °C was added dropwise DIBAL (1.0 M in
hexane solution, 32.3 mL). The reaction mixture was warmed to room
temperature over 1 h before acetone (3 mL), ethyl acetate (3 mL), and
aqueous buffer (pH 7, 3 mL) were sequentially added and the solution
stirred vigorously. After 20 min, anhydrous Na2SO4 was added and
the reaction mixture stirred vigorously for a further 30 min. The mixture
was then filtered through a pad of anhydrous Na2SO4 in a sintered glass
funnel (medium). The solvents were removed under reduced pressure.
Chromatography (96:4 hexanes:ether) provided the aldehyde 37 (8.44
g, 86%): 1H NMR [500 MHz, CDCl3] 9.54 (t, J ) 1.9 Hz, 1H), 7.64
(dd, J ) 7.9, 1.3 Hz, 2H), 7.45 (dd, J ) 7.9, 1.3 Hz, 2H), 7.41 (m,
1H), 7.34 (m, 3H), 7.27 (t, J ) 7.4 Hz, 2H), 6.01 (d, J ) 3.2 Hz, 1H),
5.71(d, J ) 3.2 Hz, 1H), 5.19 (br s, 1H), 4.59 (dd, J ) 8.9, 5.7 Hz,
1H), 2.29 (ddd, J ) 16.6, 6.6, 2.5 Hz, 1H), 2.21 (ddd, J ) 16.6, 7.1,
1.5 Hz, 1H), 2.13 (m, 1H), 1.96 (m, 1H), 1.87 (m, 2H), 1.80 (m, 2H),
1.53 (m 1H), 1.41 (d, J ) 1.0 Hz, 3H), 1.23 (m, 1H), 0.99 (s, 9H),
0.77 (d, J ) 6.8 Hz, 3H), 0.68 (d, J ) 6.8 Hz, 3H); 13C NMR [75
MHz, CDCl3] δ 202.4, 157.1, 136.2, 135.8, 135.5, 133.3, 133.2, 129.7,
129.4, 127.6, 127.3, 121.0, 120.8, 111.5, 110.6, 68.5, 43.6, 38.8, 37.4,
36.2, 34.3, 27.2, 26.8, 23.9, 22.6, 20.8, 19.1, 17.2 (6 unresolved); IR
1723; [R]20 ) +74.0° (c 0.90, CHCl3); HRMS (FAB) calcd for
D
[C39H52O4SiNa]+ 635.3533, found 635.3514.
Alcohol 40. A mixture of the silyl ether 39 (353 mg, 0.57 mmol)
was taken up in THF (5 mL), TBAF (1.0 M THF, 2.87 mL) was added,
and the reaction was stirred at room temperature for 10-12 h. The
reaction mixture was diluted with ethyl acetate and aqueous buffer (pH
7) added. The organic layer was washed with brine and dried over
anhydrous Na2SO4, and the solvent was removed under reduced
pressure. The crude product was chromatographed (85:15 hexanes:ether)
to give alcohol 40 (235 mg, 95%): 1H NMR [400 MHz, CDCl3] δ
6.44 (d, J ) 3.0 Hz, 1H), 6.29 (d, J ) 3.0 Hz, 1H), 5.43 (dd, J ) 11.0,
4.0 Hz, 1H), 5.11 (br s, 1H), 4.73 (dd, J ) 9.9, 5.7 Hz, 1H), 2.42 (m,
2H), 2.17 (br s, 1H), 1.92 (m, 3H), 1.68 (m, 2H), 1.43 (d, J ) 1.5 Hz,
3H), 1.35 (m, 1H), 1.26 (m, 1H), 1.19 (s, 9H), 0.84 (d, J ) 6.8 Hz,
3H), 0.70 (d, J ) 6.8 Hz, 3H), -0.31 (br s, 1H); 13C NMR [75 MHz,
CDCl3] δ 178.0, 160.7, 157.0, 138.5, 118.8, 116.7, 113.4, 69.3, 69.2,
42.9, 39.3, 38.6, 38.5, 37.0, 32.9, 27.0, 26.9, 24.8, 22.3, 21.2, 15.0 (2
[ν max cm-1] 2959, 1724; [R]20 ) -6.2° (c 3.05, CHCl3); HRMS
D
(FAB) calcd for [C34H43BrO3SiNa]+ 629.2063, found 629.2077.
Nozaki-Kishi Cyclization. Anhydrous chromium dichloride (5.94
g, 48.33 mmol) and anhydrous nickel chloride (1.3 g, 10.03 mmol)
were placed in a dry 1 L flask and cooled to -60 °C. Anhydrous DMF
(475 mL) was added slowly to the stirred mixture of solids. Thereafter,
the mixture was degassed (via freeze-thaw, 3 cycles). Separately, the
aldehyde 37 (6.16 g, 10.14 mmol) was taken up in DMF (150 mL)

Total Synthesis of Eleutherobin
J. Am. Chem. Soc., Vol. 121, No. 28, 1999 6577
unresolved); IR [ν max cm-1] 3424, 2955, 1723, 1475; [R]20D ) +29.3°
(c 0.92, CHCl3); HRMS (EI) calcd for [C23H34O4]+ 374.2457, found
374.2453.
1.49 (s, 3H), 1.23 (s, 9H), 0.94 (d, J ) 6.7 Hz, 3H), 0.86 (d, J ) 6.5
Hz, 3H), 0.15 (s, 9H); 13C NMR [75 MHz, CDCl3] δ 179, 134, 132,
131, 122, 94, 80, 79, 78, 68, 47, 39, 39, 36, 35, 30, 28, 28, 24, 22, 22,
21, 3 (4 unresolved); IR [ν max cm-1] 3474, 2959, 1725, 1479; [R]22
Enone Lactol 41. To a round-bottom flask were added the hydroxy
furan 40 (191 mg, 2.45 mmol) and CH2Cl2 (20 mL), and the mixture
was cooled to -78 °C. Dimethyldioxirane (20.3 mL, 0.116 M in
acetone, 2.35 mmol) was then syringed in slowly and the reaction
mixture stirred for 10 min. The solvent was then removed by a strong
flow of Ar. The residue was chromatographed (95:5 f 90:10 hexanes:
EtOAc) to furnish the enone 41 (903 mg, 94%): 1H NMR [300 MHz,
CDCl3] δ 7.08 (d, J ) 10.5 Hz, 1H), 6.19 (d, J ) 10.5 Hz, 1H), 5.35
(br s, 1H), 4.89 (d, J ) 7.6 Hz, 1H), 4.67 (t, J ) 4.4 Hz, 1H), 3.29 (br
s, 1H), 2.50-2.32 (m, 1H), 2.12-1.71 (m, 7H), 1.63-1.38 (m, 2H),
1.58 (s, 3H), 1.26 (s, 9H), 0.92 (d, J ) 6.9 Hz, 3H), 0.76 (d, J ) 6.9
Hz, 3H); 13C NMR [75 MHz, CDCl3] δ 196, 178, 152, 127, 123, 93,
79, 78, 40, 38, 33, 27, 27, 25, 22, 22; IR [ν max cm-1] 3442, 2960,
1711, 1691, 1479; [R]20D ) -60.3° (c 0.69, CHCl3); HRMS (EI) calcd
for [C23H34O5]+ 390.2406, found 390.2399.
D
) -31.1° (c 0.46, CHCl3); HRMS (EI) calcd for [C27H46O5Si]+
478.3115, found 478.3120.
Acid-Mediated Pyranose-Furanose Rearrangement. The alcohol
50 (936 mg, 1.96 mmol) was taken up in methanol (70 mL). p-TsOH‚
H2O (56 mg, 0.29 mmol) was added and the solution allowed to stir at
room temperature. TLC analysis after ∼30 min indicated disappearance
of starting material. Et3N (a few drops) was added to quench the
reaction, and after concentration in vacuo, the crude material was
chromatographed with (90:9:1 hexanes:EtOAc:Et3N) to give the desired
47 (742 mg, 90%): 1H NMR [500 MHz, CDCl3] δ 6.25 (d, J ) 6.0
Hz, 1H), 5.88 (d, J ) 6.0 Hz, 1H), 5.19 (br s, 1H), 4.98 (d, J ) 9.2
Hz, 1H), 3.85 (d, J ) 10.2 Hz, 1H), 3.18 (s, 3H), 2.53 (br s, 1H),
1.82-1.86 (m, 2H), 1.73 (s, 3H), 1.64-1.74 (m, 4H), 1.50 (s, 3H),
1.45 (dd, J ) 15.1, 3.1 Hz, 1H), 1.17 (s, 9H), 1.14-1.23 (m, 3H),
0.82 (d, J ) 6.8 Hz, 3H), 0.68 (d, J ) 6.8 Hz, 3H); 13C NMR [75
MHz, CDCl3] δ 177.1, 138.7, 136.5, 127.9, 119.4, 114.3, 93.4, 77.2,
73.6, 49.9, 38.9, 38.6, 38.1, 37.8, 36.9, 32.0, 26.9, 26.7, 26.0, 24.3,
23.0, 22.2, 15.1 (2 unresolved); IR [ν max cm-1] 3491, 2962, 1723,
Alcohol Lactol 42. The crude lactol enone 41 (110 mg, 0.28 mmol)
was dissolved in THF (5 mL). The solution was cooled to -78 °C,
and a solution of CH3Li (1.4 M, 0.76 mL, 1.07 mmol) was added
slowly. The reaction was stirred at -78 °C for 30 min before it was
quenched by methanol and pH 7.0 buffer. The aqueous solution was
extracted with EtOAc (3 times). The organic layer was washed with
brine and dried over anhydrous Na2SO4. After evaporation of the
solvent, the crude material was purified by chromatography (85:15 f
80:20 hexanes:EtOAc) to give the desired product 42 (46 mg, 42%):
1H NMR [400 MHz, CDCl3] δ 6.22 (d, J ) 6.0 Hz, 1H), 5.91 (d, J )
6.0 Hz, 1H), 5.18 (br s, 1H), 4.88 (d, J ) 8.6 Hz, 1H), 3.85 (d, J ) 9.6
Hz, 1H), 3.26 (br s, 1H), 2.54 (br s, 1H), 1.69 (s, 3H), 1.51 (s, 3H),
1.56-1.84 (m, 8H), 1.18 (s, 9H), 1.17-1.24 (m, 2H), 0.82 (d, J ) 6.8
Hz, 3H), 0.67 (d, J ) 6.8 Hz, 3H); 13C NMR [75 MHz, CDCl3] δ
178.8, 138.7, 136.2, 128.5, 119.3, 111.4, 93.6, 77.1, 76.8, 39.2, 38.7,
38.0, 37.9, 36.8, 31.8, 29.7, 27.1, 26.9, 26.6, 24.3, 23.0, 21.2 (2
1455, 1361, 1281; [R]20 ) +22.9° (c 1.50, CHCl3); HRMS (FAB)
D
calcd for [C25H40O5Na]+ 443.2773, found 443.2763.
Silylated Alcohol 48. To a round-bottom flask, equipped with a
stir bar, was added the alcohol 47 (715 mg, 1.70 mmol). Methylene
chloride (50 mL) was added and the solution cooled to 0 °C. Thereafter,
2,6-lutidine (1.15 mL, 9.87 mmol) was added followed by TBSOTf
(1.13 mL, 4.92 mmol). The cooling bath was then removed and the
mixture stirred for 10 min before aqueous buffer (pH 7) was added.
The reaction mixture was diluted with CH2Cl2, and the organic layer
was washed with brine and dried over anhydrous Na2SO4. The solvent
was removed under reduced pressure, and the residue was chromato-
graphed (99:1 f 96:4 hexanes:ether) to furnish the silylated compound
48 (754 mg, 83%): 1H NMR [300 MHz, CDCl3] δ 6.25 (d, J ) 6.0
Hz, 1H), 5.88 (d, J ) 6.0 Hz, 1H), 5.22 (br s, 1H), 5.03 (d, J ) 9.0
Hz, 1H), 3.87 (d, J ) 9.6 Hz, 1H), 3.22 (s, 3H), 2.57 (br s, 1H), 1.95-
1.40 (m, 9H), 1.76 (s, 3H), 1.46 (s, 3H), 1.21 (s, 9H), 0.91 (s, 9H),
0.87 (d, J ) 6.8 Hz, 3H), 0.72 (d, J ) 6.8 Hz, 3H), 0.11 (s, 3H) 0.10
(s, 3H); 13C NMR [75 MHz, CDCl3] δ 178, 139, 138, 128, 120, 115,
94, 78, 74, 50, 40, 39, 38, 38, 37, 32, 27, 27, 27, 26, 25, 24, 22, 18,
unresolved); IR [ν max cm-1] 3438, 2959, 1709; [R]20 ) +31.2° (c
D
2.09, CHCl3); MS (CI, NH3) 389 [M + H - H2O]+.
Enone Silylated Lactol 49. A round-bottom flask (equipped with a
stir bar) was charged with the enone 41 (997 mg, 2.55 mmol). CH2Cl2
(80 mL) was added and the solution cooled to -78 °C following
which 2,6-lutidine (0.89 mL, 7.65 mmol) and trimethylsilyl triflate (0.68
mL, 3.76 mmol) were added slowly and the mixture was stirred for 10
min whence TLC analysis showed disappearance of the starting
material. The cooling bath was then removed and aqueous buffer (pH
7) added. The reaction mixture was diluted with CH2Cl2, and the organic
layer was washed with brine and dried over anhydrous Na2SO4. The
solvents were removed under reduced pressure, and the crude material
was chromatographed [94:5:1 hexanes:ether:triethylamine] to furnish
the protected enone 49 (1.08 g, 92%): 1H NMR [300 MHz, CDCl3] δ
7.14 (d, J ) 10.5 Hz, 1H), 6.10 (d, J ) 10.5 Hz, 1H), 5.32 (br s, 1H),
4.71 (d, J ) 7.1 Hz, 1H), 4.60 (t, J ) 4.4 Hz, 1H), 2.41-2.27 (m,
1H), 2.10-1.68 (m, 7H), 1.62-1.38 (m, 5H), 1.23 (s, 9H), 0.93 (d, J
) 6.9 Hz, 3H), 0.77 (d, J ) 6.7 Hz, 3H), 0.16 (s, 9H); 13C NMR [75
MHz, CDCl3] δ 196, 179, 154, 125, 122, 94, 78, 39, 39, 27, 27, 25,
22, 22, 2; IR [ν max cm-1] 2959, 1727, 1691, 1480; [R]23D ) +14.2°
(c 0.45, CHCl3); HRMS (EI) calcd for [C26H42O5Si]+ 462.2802, found
462.2801.
Alcohol 50. To a round-bottom flask, equipped with a stir bar, were
added the enone 49 (1.08 g, 2.33 mmol) and ether (70 mL), and the
mixture was cooled to -78 °C. Methyllithium (1.4 M in ether, 16.6
mL, 23.24 mmol) was added slowly and the mixture stirred for 10 min
whence TLC analysis showed very little starting material left. The
reaction mixture was quenched at -78 °C with methanol and aqueous
buffer (pH 7) and the cooling bath removed. The organic layer was
washed with brine and dried over anhydrous Na2SO4, and the solvent
was removed under reduced pressure. The residue was chromatographed
[90:10:1 hexanes:EtOAc:Et3N] to furnish the alcohol 50 (963 mg, 86%;
99% based on recovered starting material): 1H NMR [300 MHz, CDCl3]
δ 5.74 (d, J ) 10.4 Hz, 1H), 5.62 (dd, J ) 10.4, 1.0 Hz, 1H), 5.40 (br
s, 1H), 4.88 (d, J ) 7.3 Hz, 1H), 4.10 (dd, J ) 12.1, 4.2 Hz, 1H),
2.69-2.55 (m, 3H), 2.38-2.33 (m, 1H), 2.19-2.05 (m, 2H), 1.96-
1.82 (m, 1H), 1.82-1.70 (m, 4H), 1.62 (s, 1H), 1.55-1.42 (m, 2H),
16, -3, -4 (4 unresolved); IR [ν max cm-1] 2957, 1728; [R]23
)
D
+25.3° (c 1.21, CHCl3); HRMS (FAB) calcd for [C31H54O5Si]+
534.3741, found 534.3761.
Alcohol 51. To a round-bottom flask, equipped with a stir bar, were
added the pivaloate 48 (812 mg, 1.52 mmol) and methylene chloride
(40 mL), and the mixture was cooled to -78 °C. DIBAL (1.0 M in
hexane, 4.44 mL) was added slowly, and the mixture was stirred for
10 min before acetone (3 mL), ethyl acetate (3 mL), and aqueous buffer
(pH 7, 3 mL) were sequentially added. The cooling bath was removed
and the resulting solution stirred vigorously for 20 min. Thereafter,
anhydrous Na2SO4 was added and the reaction mixture was stirred
vigorously for a further 30 min. The mixture was then filtered through
a pad of anhydrous Na2SO4 in a sintered glass funnel (medium), and
the solvents were removed under reduced pressure. The residue was
chromatographed (90:10 f 85:15 hexanes:ether) to furnish the alcohol
51 (600 mg, 88%): 1H NMR [300 MHz, CDCl3] δ 6.26 (d, J ) 6.1
Hz, 1H), 5.84 (d, J ) 6.0 Hz, 1H), 5.19 (br s, 1H), 3.88 (d, J ) 11.4
Hz, 1H), 3.72 (d, J ) 7.6 Hz, 1H), 3.28 (s, 3H), 2.95 (s, 1H), 2.49 (br
s, 1H), 2.00-1.52 (m, 7H), 1.75 (s, 3H), 1.50 (s, 3H), 1.30-1.24 (m,
1H), 1.24-1.13 (m, 1H), 0.89 (s, 9H), 0.86 (d, J ) 6.8 Hz, 3H), 0.75
(d, J ) 6.8 Hz, 3H), 0.10 (s, 3H), 0.07 (s, 3H); 13C NMR [75 MHz,
CDCl3] δ 140, 139, 130, 120, 117, 94, 78, 76, 51, 40, 38, 38, 37, 32,
27, 26, 26, 25, 24, 22, 18, 15, -3, -5 (2 unresolved); IR [ν max cm-1
]
3568, 2957, 2859, 1461; [R]22 ) +50.4° (c 0.55, CHCl3); HRMS
D
(FAB) calcd for [C26H45O4Si]+ 449.3087, found 449.3090.
Ketone 52. The alcohol 51 (1.09 g, 2.42 mmol) was taken up in
CH2Cl2 (60 mL) and stirred at room temperature under Ar. Ground 4
Å molecular sieves (1.3 g) were added, and the solution was stirred
for ∼30 min before addition of NMO (528 mg, 4.51 mmol) followed
by TPAP (266 mg, 0.76 mmol). TLC after 3 h showed disappearance

6578 J. Am. Chem. Soc., Vol. 121, No. 28, 1999
Chen et al.
of starting material. The solution was then filtered through a short plug
of SiO2. Evaporation of solvents furnished the ketone 52 (949 mg, 87%)
which was good enough to be used for the next step or could be
chromatographed: 1H NMR [500 MHz, CDCl3] δ 6.42 (br s, 0.7H),
6.2-5.82 (br unresolved m, 1.3H), 5.16 (br s, 1H), 3.92 (d, J ) 10.7
Hz, 1H), 3.3 (br s, 3H), 2.66 (br m, 1H), 2.5-2.0 (br unresolved m,
2.6H), 1.92-1.83 (br unresolved m, 1.4H), 1.7 (br unresolved m, 5H),
1.55 (s, 0.7H), 1.51 (s, 3.3H), 1.23 (br unresolved m, 2H), 0.87 (s,
12H), 0.76 (br unresolved m, 3H), 0.09 (s, 3H), 0.05 (s, 3H); 13C NMR
[75 MHz, CDCl3] δ 145.4, 138.1, 133.2, 126.5, 125.6, 121.7, 112.4,
93.7, 80.4, 50.2, 41.4, 38.4, 34.2, 33.3, 28.8, 25.7, 25.0, 24.3, 22.2,
22.1, 20.1, 17.7, -3.7, -4.9 (2 unresolved); IR [ν max cm-1] 2957,
2931, 1721, 1464; [R]23D ) +36.2° (c 1.03, CHCl3); HRMS (CI, NH3)
calcd for [C26H44O4Si]+ 448.3009, found 448.3000.
mg). Methyl triflate (343 mg, 2.1 mmol) was added subsequently, and
after 12 h of stirring at 0 °C, the reaction mixture was concentrated in
vacuo. Purification of the residue by chromatography (98:2 hexanes:
EtOAc) furnished the stannylmethyl glycosides 84 as a 1:1 mixture
(417 mg, ∼93%) which was difficult to separate. This mixture was
dissolved in THF (1 mL), and TBAF (1 M, 1.4 mL) was added. The
resulting yellow solution was stirred for 2 h at room temperature, the
solvent was removed under reduced pressure, the crude material was
chromatographed (80:20 hexanes:EtOAc), and the mixture of anomers
could be easily separated at this stage furnishing the desired â-stan-
nylmethoxy anomer (155 mg, 45%): 1H NMR [400 MHz, CDCl3] δ
4.58 (d, J ) 3.5 Hz, 1H), 4.16 (d, J ) 3.1 Hz, 1H), 4.07 (t, J ) 6.3
Hz, 1H), 4.02 (dd, J ) 16.6, 8.6 Hz, 1H), 3.85 (dt, J ) 13.4, 2.4 Hz,
2H), 3.71 (dd, J ) 3, 6.6 Hz, 1H), 3.57 (dd, J ) 16.3, 8 Hz, 1H), 2.02
(s, 1H), 1.54-1.40 (m, 9H), 1.30 (s, 3H), 1.29-1.24 (m, 6H), 0.97-
0.83 (m, 15H); 13C NMR [75 MHz, CDCl3] δ 109.6, 101.2, 76.7, 73.4,
Triflate 58. To a round-bottom flask, equipped with a stir bar, were
added the ketone 52 (465 mg, 1.04 mmol) and THF (20 mL), and the
mixture was cooled to 0 °C. LHMDS (1.0 M in THF, 1.10 mL) was
added slowly, and the mixture was stirred for 1 h. After removal of
the cooling bath, the mixture was stirred for a further 1 h and then
cooled to -78 °C. In a separate flask 2-[N,N-bis(trifluoromethylsul-
fonyl)amino]-5-chloropyridine (2.16 g, 5.50 mmol) was dissolved in
THF (20 mL), and this solution was added via cannula to the reaction
flask. After 30 min, the cooling bath was removed and the mixture
allowed to stir for 30 min before aqueous buffer (pH 7) was added.
The reaction mixture was extracted with ether, the organic layer was
washed with brine and dried over anhydrous Na2SO4, and the solvent
was removed under reduced pressure. The residue was chromatographed
(98:2 f 96:4 hexanes:ether) to furnish the triflate 58 (426 mg, 71%):
1H NMR [300 MHz, CDCl3] δ 6.26 (d, J ) 5.9 Hz, 1H,), 5.88 (d, 1H,
J ) 5.9 Hz), 5.52 (d, J ) 10.2 Hz, 1H), 5.29 (br unresolved m, 1H),
3.82 (ddd, J ) 10.2,4.0, 3.4 Hz, 1H), 3.59 (d, J ) 7.2 Hz, 1H), 3.25
(s, 3H), 2.35 (m, 1H), 2.1 (m, 2H), 1.70 (dd, J ) 13.4, 2 Hz, 1H), 1.59
(s, 3H), 1.56-1.35 (m, 2H), 1.45 (s, 3H), 1.33-1.21 (m, 2H), 1.27 (s,
3H), 1.24 (s, 3H), 1.18 (s, 9H), 0.11(s, 3H), 0.1 (s, 3H); 13C NMR [75
MHz, CDCl3] δ 145.4, 138.1, 133.2, 126.5, 125.6, 121.7, 112.4, 93.7,
80.4, 50.2, 41.4, 38.4, 34.2, 33.3, 28.8, 25.7, 25.0, 24.3, 22.2, 22.1,
20.1, 17.7, -3.7, -4.9 (3 unresolved); IR [ν max cm-1] 2961, 2860,
70.5, 59.9, 59.0, 29.5, 28.3, 27.7, 26.3, 14.1, 9.5, (8 unresolved); [R]20
D
) -91.4° (c 1.0, CHCl3); HRMS (FAB) calcd for [C21H42O5Sn + K]+
533.1694, found 533.1692.
For the ent-stannylmethyloxy glycoside, [R]25 ) +92.3° (c 1.04,
D
CHCl3).
Data for the r-anomer:1H NMR [400 MHz, CDCl3] δ 4.19 (m,
1H), 4.09 (m, 3H), 3.95 (d, J ) 6.3 Hz, 1H), 4.02 (dd, J ) 16.6, 8.6
Hz, 1H), 3.85 (dt, J ) 13.4, 2.4 Hz, 1H), 3.71 (dd, J ) 7.6 Hz, 1H),
3.73 (dd, J ) 13.2, 3.3 Hz, 1H), 3.65 (d, J ) 8 Hz, 1H), 3.50 (t, J )
7.6 Hz, 1H), 2.29 (br s, 1H), 1.49 (m, 9H), 1.32 (s, 3H), 1.32 (s, 3H),
1.24 (m, 6H), 0.87 (m, 9H); 13C NMR [75 MHz, CDCl3] δ 110.5, 105.5,
78.5, 74,3, 73.5, 63.5, 60.3, 29.6, 29.5, 29.4, 29.3, 28.4, 28.0, 27.7,
27.6, 27.3, 26.4, 14.1, 9.5, 7.3; [R]20D ) -14.7° (c 1.0, CHCl3); HRMS
(FAB) calcd for [C21H42O5Sn + K]+ 533.1694, found 533.1677.
Acetate 85. The â-stannylmethoxy glycoside (135 mg, 0.27 mmol),
obtained from the desilylation step above, was taken up in methylene
chloride (2 mL) at 0 °C. Acetic anhydride (140 mg, 1.37 mmol) and
DMAP (149 mg, 1.2 mmol) were then added. The reaction mixture
was stirred for 2 h, the contents were concentrated, and the residue
was purified by chromatography (90:10 hexanes:EtOAc) to afford the
acetate 85 (146 mg, 99%): 1H NMR [500 MHz, CDCl3] δ 4.83 (dd, J
) 8, 4.8 Hz, 1H), 4.66 (d, J ) 3.3 Hz, 1H), 4.23-4.17 (m, 2H), 4.02-
3.89 (m, 2H), 3.82 (dd, J ) 10.5, 2.8 Hz, 1H), 3.48 (dd, J ) 19, 8.5
Hz, 1H), 2.06 (s, 3H), 1.54-1.40 (m, 9H), 1.32-1.23 (m, 9H), 0.95-
0.81 (m, 15H); 13C NMR [75 MHz, CDCl3] δ 170.9, 109.7, 99.4, 74.0,
73.5, 73.0, 58.9, 58.7, 29.6, 28.4, 26.7, 21.3, 14.1, 9.4 (9 unresolved);
1419; [R]23 ) -5.64° (c 1.21, CHCl3).
D
Thioglycoside 67. To a solution of the thioglycoside 64 (730 mg,
2.3 mmol) in MeOH (10 mL) was added NaOMe (12 mg, 0.23 mmol).
The solution was stirred for 3 h and then neutralized with Dowex-50
(H+) resin and filtered. The filtrate was concentrated in vacuo. The
solid residue 65 was then dissolved in acetone (5 mL), and p-TsOH‚
H2O (47.3 mg, 0.23 mmol) and 2,2-dimethoxypropane (2 mL) were
added. This mixture was stirred for 1 h before being quenched with
saturated aqueous NaHCO3 (10 mL). The reaction mixture was extracted
with EtOAc, and the organic layer washed with brine and dried over
anhydrous Na2SO4. After evaporation of the solvents, the crude material
was purified by chromatography (50:50 hexanes:EtOAc) to afford the
acetonide 66 (515 mg, 96%). The acetonide 66 (515 mg, 2.2 mmol)
was taken up in methylene chloride (20 mL). Imidazole (748 mg, 11
mmol) and DMAP (27 mg, 0.22 mmol) were added to it. TBSCl (663
mg, 4.4 mmol) was then added and the reaction mixture allowed to
stir for ∼5 h. Water was then added to the reaction mixture and
extracted with EtOAc (3 times). The organic layer was washed with
brine and then dried over anhydrous Na2SO4. The solvent was removed
under reduced pressure, and the residue was chromatographed (95:5
f 90:10 hexanes:EtOAc) to furnish the silyl ether 67 (718 mg, 94%):
1H NMR [400 MHz, CDCl3] δ 4.33 (d, J ) 8.0 Hz, 1H), 4.23 (m, 1H),
4.19 (dt, J ) 3.4, 2.5 Hz, 1H), 3.93 (t, J ) 5.9 Hz, 1H), 3.66 (dd, J )
9.4, 3.5 Hz, 1H), 3.60 (dt, J ) 6.1, 1.8 Hz, 1H), 2.60 (m, 2H), 1.47 (s,
3H), 1.30 (s, 3H), 1.22 (t, J ) 7.4 Hz, 3H), 0.85 (s, 9H), 0.083 (s, 6H);
13C NMR [75 MHz, CDCl3] δ 109.9, 85.7, 79.7, 73.9, 73.3, 65.3, 28.3,
IR [ν max cm-1] 1731; [R]20 ) -133.7° (c 0.50, CHCl3); HRMS
D
(FAB) calcd for [C23H44O6Sn + K]+ 575.1800, found 575.1815.
For the ent-stannylmethyloxy glycoside 86, [R]20 ) +179.2° (c
D
1.15, CHCl3).
Stille Reaction. The triflate 58 (283 mg, 0.49 mmol), stannane 85
(2.00 g, 3.74 mmol), Pd(PPh3)4 (117 mg, 0.10 mmol), and LiCl (861
mg, 20.35 mmol) were placed in a vial and suspended in THF (1.4
mL). This mixture was degassed (via freeze-thaw, 3 cycles) and the
contents sealed. This was then heated to 130 °C until the mixture had
turned black (40 min) and then cooled to room temperature. The mixture
was subjected to chromatography (93:7 f 88:12 hexanes:ether),
yielding the coupling product 87 (170 mg, 52%) as a pale-yellow oil.
Although not completely pure, this was taken forward for the next step.
Alcohol 88. The silyl ether 87 (270 mg, 0.40 mmol) was taken up
in THF (5 mL), and TBAF (1 M THF solution, 1.6 mL) was added to
it. The reaction mixture was stirred at 0 °C for 2 h. Aqueous buffer
(pH 7) was added followed by methylene chloride. The organic layer
was washed with brine, dried over anhydrous Na2SO4, filtered, and
concentrated in vacuo. The crude product was chromatographed (85:
15 f 80:20 hexanes-EtOAc) to give the alcohol 88 (161 mg, 67%):
1H NMR [500 MHz, CDCl3] δ 6.07 (d, J ) 5.9 Hz, 1H), 6.01 (d, J )
5.9 Hz, 1H), 5.49 (d, J ) 9.4 Hz, 1H), 5.28 (br s, 1H), 4.88 (dd, J )
8.1, 3.2 Hz, 1H), 4.82 (d, J ) 3.2 Hz, 1H), 4.26-4.30 (m, 2H), 4.21
(m, 1H), 3.80-3.96 (m, 4H), 3.66 (d, J ) 7.4 Hz, 1H), 3.18 (s, 3H),
2.32 (m, 2H), 2.08 (s, 3H), 1.98 (m, 1H), 1.71 (br s, 1H), 1.61 (m,
1H), 1.57 (s, 3H), 1.52 (s, 3H), 1.51 (s, 3H), 1.40-1.47 (m, 2H), 1.33
(s, 3H), 1.18 (m, 1H), 0.89 (d, J ) 6.3 Hz, 3H), 0.88 (d, J ) 6.3 Hz,
3H); 13C NMR [75 MHz, CDCl3] δ 170.3, 136.9, 134.0, 133.7, 132.8,
130.5, 121.6, 115.5, 109.3, 92.8, 91.3, 80.6, 77.2, 73.6, 73.1, 72.0, 69.1,
26.6, 26.3, 24.9, 18.6, 15.4, -3.9, -4.3 (2 unresolved); [R]20
)
D
+33.86° (c 1.06, CHCl3); HRMS (CI, NH3) calcd for [C16H32O4SSi+1]+
349.1869, found 349.1884.
For the ent-thioglycoside 70, [R]20 ) -24.7° (c 0.94, CHCl3).
D
Alcohol. To a stirred mixture of the thioglycoside 67 (241 mg, 0.692
mmol) in dry CH2Cl2/Et2O (1/2, v/v, 2 mL) at 0 °C were added
tributylstannylmethyl alcohol (83) (444 mg, 1.38 mmol), 2,6-di-tert-
butylpyridine (544 mg, 2.4 mmol), and molecular sieves (4 Å, 100

Total Synthesis of Eleutherobin
J. Am. Chem. Soc., Vol. 121, No. 28, 1999 6579
58.8, 49.5, 42.3, 39.2, 34.4, 34.0, 29.2, 28.0, 26.3, 24.5, 22.2, 22.1,
6.14 (d, J ) 5.9 Hz, 1H), 6.11 (d, J ) 5.9 Hz, 1H), 5.59 (d, J ) 9.4
Hz, 1H), 5.26 (br s, 1H), 4.97 (m, 2H), 4.80 (d, J ) 7.3 Hz, 1H), 4.28
(d, J ) 12.2 Hz, 1H), 4.03 (s, 3H), 3.97 (m, 2H), 3.81 (d, 12.8 Hz,
1H), 3.70 (s and overlapping m, 3 + 1H), 3.19 (s, 3H), 2.60 (m, 1H),
2.29 (m, 1H), 2.13 (s, 3H), 1.97 (m, 1H), 1.7-1.15 (m, 6H), 1.52 (s,
3H), 1.44 (s, 3H), 0.97 (d, J ) 6.5 Hz, 3H), 0.91 (d, J ) 6.6 Hz, 3H);
21.1, 20.4; IR [ν max cm-1] 3484, 2964, 1754; [R]20 ) -76.0° (c
D
1.36, CHCl3); HRMS (FAB) calcd for C31H46O9K [M + K+] 601.2779,
found 601.2777.
Urocanic Acylation. The alcohol 88 (131 mg, 0.23 mmol), N-
methylurocanic acid47 (152 mg, 0.92 mmol), 1,3-dicyclohexylcarbo-
diimide (262 mg, 1.27 mmol), and DMAP (253 mg, 2.07 mmol) were
dissolved in chloroform (20 mL). This mixture was warmed to 50 °C
for 2 h. Aqueous buffer (pH 7) was added followed by methylene
chloride. The organic layer was washed with brine and dried over
anhydrous Na2SO4, and the solvents were removed under reduced
pressure. The residue was subjected to chromatography (99:1 CH2Cl2:
MeOH), yielding the coupling product 90 (128 mg, 80%) which was
slightly impure and was taken forward for the next step without further
purification.
[R]20 ) +44.5° (c 0.09, MeOH); MS (CI, CH4) for [C35H49O10N2]+
D
657.
Acknowledgment. This work was supported by the National
Institutes of Health (Grants HL25848 and AI 16943 (S.J.D.)
and CA-08748 (SKI Core Grant)). Graduate fellowships are
gratefully acknowledged by X.-T. C. (Kanagawa Academy of
Science and Technology) and B.Z. (Pharmacia-Upjohn). Post-
doctoral fellowships are gratefully acknowledged by C.E.G. (The
Royal Commission for the Exhibition of 1851) and T.R.R.P.
(The National Science Foundation). We are grateful to Vinka
Parmakovich and Barbara Sporer of the Columbia University
Mass Spectral Facility. We are indebted to Professor Gerard
Parkin, Mr. Tony Hascall, and Dr. Brian Bridgewater for
expeditious X-ray crystallographic determinations. We also
thank Professor Fenical for kindly providing us a copy of the
spectral data of natural eleutherobin.
Eleutherobin 1. The acetonide 90 (70 mg, 0.10 mmol) was dissolved
in methanol (15 mL), and PPTS (50 mg, 0.20 mmol) was added. This
mixture was warmed to 50 °C for 36 h. After cooling to room
temperature, aqueous buffer (pH 7) was added followed by methylene
chloride. The organic layer was washed with brine and dried over
anhydrous Na2SO4, and the solvents were removed under reduced
pressure. The residue was subjected to chromatography (95:5 CH2Cl2:
MeOH) yielding eleutherobin 1 (60 mg, 91%): 1H NMR [500 MHz,
CDCl3] δ 7.52 (d, J ) 15.6 Hz, 1H), 7.45 (s, 1H), 7.08 (s, 1H), 6.55
(d, J ) 15.6 Hz, 1H), 6.10 (d, J ) 5.8 Hz, 1H), 6.07 (d, J ) 5.8 Hz,
1H), 5.54 (d, J ) 9.4 Hz, 1H), 5.26 (br s, 1H), 4.96 (dd, J ) 9.8, 3.5
Hz, 1H), 4.90 (d, J ) 3.5 Hz, 1H), 4.80 (d, J ) 7.4 Hz, 1H), 4.28 (d,
J ) 12.2 Hz, 1H), 4.01 (dd, J ) 9.8, 3.5 Hz, 1H), 3.97 (m, 1H), 3.86
(d, J ) 12.2 Hz, 1H), 3.82 (d, J ) 12.5 Hz, 1H), 3.70 (s, 3H), 3.20 (s,
3H), 2.60 (m, 1H), 2.28 (m, 1H), 2.10 (s, 3H), 1.97 (m, 1H), 1.52-
1.61 (m, 4H), 1.51 (s, 3H), 1.43 (s, 3H), 1.35-1.42 (m, 1H), 1.22-
Supporting Information Available: Schemes for attempted
Wittig reactions on 32 and attempted cyclizations on 36 and
37 and experimental procedures and spectroscopic data for
compound 45, 46, 63, 64, 72, 76 and i and intermediates in
glycosylation reactions on model substrates. In addition, ex-
perimental copies of 1H and 13C of all new compounds described
in the Experimental Section and full crystallographic details of
41 and 45, including crystal data and structure refinement,
atomic coordinates, and bond lengths and angles and anisotropic
displacement coefficients of crystal structures (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
1.31 (m, 4H), 0.96 (d, J ) 6.6 Hz, 3H), 0.91 (d, J ) 6.6 Hz, 3H); 13
C
NMR [75 MHz, CDCl3] δ 171.4, 166.7, 139.2, 138.4, 137.4, 136.5,
134.1, 133.7, 132.8, 131.0, 122.8, 121.3, 115.9, 115.9, 93.4, 89.9, 81.5,
71.7, 69.4, 69.1, 68.1, 62.1, 49.7, 42.3, 38.7, 34.2, 33.6, 31.4, 29.0,
24.5, 24.2, 22.2, 22.0, 21.0, 20.5; IR [ν max cm-1] 3545, 3124, 2913,
2852, 1731, 1697, 1650; [R]20 ) -73.2° (c 0.17, CHCl3); HRMS
D
(FAB) calcd for [C35H49O10N2]+ 657.3387, found 657.3388.
1
Neoeleutherobin 94: H NMR [500 MHz, CDCl3] δ 7.53 (d, J )
15.6 Hz, 1H), 7.45 (s, 1H), 7.08 (s, 1H), 6.56 (d, J ) 15.7 Hz, 1H),
JA990215C