Synthesis of key fragments of leiodelide A

Article abstract of DOI:10.1021/ol201183f

The synthesis of all key fragments of the marine macrolide leiodelide A is described. The polyoxygenated northern subunit is derived from d-xylose, while the southern subunit is rapidly assembled via an aldol reaction and Horner-Wadsworth-Emmons olefination. This highly convergent approach will allow for rapid modification and assembly of several isomers of leiodelide A, which may be necessary considering the assignment of leiodelide B has been previously shown to be incorrect.

Full text of DOI:10.1021/ol201183f

ORGANIC
LETTERS
2011
Vol. 13, No. 12
3246–3249
Synthesis of Key Fragments of
Leiodelide A
Mathieu F. Chellat, Nicolas Proust, Matthew G. Lauer, and James P. Stambuli*
Department of Chemistry, The Ohio State University, 100 West 18th Avenue, Columbus,
Ohio 43210, United States
stambuli@chemistry.ohio-state.edu
Received May 4, 2011
ABSTRACT
The synthesis of all key fragments of the marine macrolide leiodelide A is described. The polyoxygenated northern subunit is derived from
D-xylose, while the southern subunit is rapidly assembled via an aldol reaction and HornerÀWadsworthÀEmmons olefination. This highly
convergent approach will allow for rapid modification and assembly of several isomers of leiodelide A, which may be necessary considering the
assignment of leiodelide B has been previously shown to be incorrect.
The leiodelides are a family of biologically active
macrolides produced by the deep-water marine sponge
Leiodermatium and first isolated by Fenical and co-
workers (Figure 1).1 Structures of these two 19-mem-
bered macrolides were proposed as 1 and 2 based on
spectral analysis, and both currently have at least
€
one unassigned stereogenic center. Recently, Furstner
et al. published the synthesis of the proposed structure
of leiodelide B (2) along with three diastereomers
and concluded that none of these isomers matched
the reported spectroscopic data enough to claim
identity.2
Leiodelide A (1) exhibits cytotoxic activity against
HL-60 leukemia and OVCAR-3 ovarian cancer cell lines.
It features an oxazole-containing 19-membered macrolide,
seven stereocenters, and a 10-carbon side chain with an
R,R-disubstituted carboxylic acid terminus similar to that
Figure 1. Proposed structures of and biosynthetic relationship
between leiodelide A (1) and leiodelide B (2).
(1) Sandler, J. S.; Colin, P. L.; Kelly, M.; Fenical, W. J. Org. Chem.
2006, 71, 7245.
ꢀ
(2) Larivee, A.; Unger, J., B.; Thomas, M.; Wirtz, C.; Dubost, C.;
of okadaic acid.3 The combined factors of low natural
supply, biological activity, and unusual structure made it
an attractive target for total synthesis. Moreover, a total
synthesis and absolute configuration determination of
€
Handa, S.; Furstner, A. Angew. Chem., Int. Ed. 2011, 50, 304.
(3) (a) Tachibana, K.; Scheuer, P. J.; Tsukitani, Y.; Kikuchi, H.; Van
Engen, D.; Clardy, J.; Gopichand, Y.; Schmitz, F. J. J. Am. Chem. Soc.
1981, 103, 2469. (b) Forsyth, C. J.; Sabes, S. F.; Urbanek, R. A. J. Am.
Chem. Soc. 1997, 119, 8381.
r
10.1021/ol201183f
Published on Web 05/25/2011
2011 American Chemical Society

leiodelide A would hopefully help resolve the “leiodelide B
puzzle”.2
diastereomers in the event of discrepancies between the
synthetic and natural products.
The synthesis of bromide 5 commenced with the forma-
tion of olefin 14 (Scheme 1). Starting from known alcohol
11,5 methylation provided lactone 12, which was then
opened to the corresponding Weinreb amide. Since the
chromatographic purification of the amide on silica gel
resulted in recyclization to 12, immediate conversion to the
triethylsilyl ether 13 was performed.
Scheme 1. Synthesis of the Northern Subunit Precursor 14
Reduction of amide 13 to the corresponding aldehyde
using DIBAL set the stage for the Wittig olefination. The
coupling partner, phosphonium bromide 8, was prepared
in two steps: bromination of 2-(benzyloxy)ethanol6 fol-
lowed by treatment with ethyltriphenylphosphonium bro-
mide in the presence of n-BuLi.7 The Wittig reaction
resulted in an inconsequential 1:1 mixture of E- and Z-
olefins 14, which were separated for analytical purposes.
The reduction/deprotection sequence proved to be more
difficult than expected and required some optimization.
Treatment of unsaturated benzyl ether 14 with various
sources of palladium all led to hydrogenolysis of the benzyl
ether to alkane 15 (Scheme 2). This product is presumably
formed via an olefin migration/palladium π-allyl forma-
tion/elimination mechanism. Palladium is known to have
a high isomerization activity,8 and this type of reaction
has been previously reported with both homoallylic
Figure 2. Retrosynthetic analysis of leiodelide A (1).
As outlined in Figure 2, our retrosynthetic strategy
started by disconnection at C22 to reveal macrolide 4
and side chain 3. Macrolide 4 could be accessed from
simplified fragments5, 6, and 7. First, the northern subunit
5, with its three contiguous oxygenated chiral centers, was
envisioned to originate from the opening of chiral lactone
12 (derived from D-xylose) followed by Wittig olefination
with phosphonium bromide 8. Second, the R,β-unsatu-
rated ester 7 could be generated from an Evans aldol
reaction between acylated oxazolidinone 9 and aldehyde
10 followed by a HornerÀWadsworthÀEmmons olefina-
tion reaction to install the ester moiety. Finally, oxazole 6
can be obtained in three steps from known 2-thiooxazole.4
The key steps of the synthesis would be the appending of
the northern and southern fragments onto the oxazole via
metal-catalyzed cross-coupling reactions.4 This highly
convergent strategy would provide a great deal of flex-
ibility when it comes to derivatization and SAR studies.
It would also provide us with easy access to multiple
ꢀ
(5) Han, S.-Y.; Joullie, M. M.; Fokin, V. V.; Petasis, N. A. Tetra-
hedron: Asymmetry 1994, 5, 2535.
(6) Hammerschmidt, F.; Kaehlig, H. J. Org. Chem. 1991, 56, 2364.
(7) Marinier, A.; Deslongchamps, P. Can. J. Chem. 1992, 70, 2350.
(8) Rylander, P. N. Hydrogenation Methods; Academic Press: London;
Orlando, FL, 1985; pp 31À36.
(9) Cocker, W.; Shannon, P. V. R.; Staniland, P. A. J. Chem. Soc. C
1966, 41.
(10) Nishimura, S.; Uramoto, M.; Watanabe, T. Bull. Chem. Soc.
(4) Counceller, C. M.; Eichman, C. C.; Proust, N.; Stambuli, J. P.
Adv. Synth. Catal. 2011, 353, 79.
Jpn. 1972, 45, 216.
Org. Lett., Vol. 13, No. 12, 2011
3247

cyclopropanes9 and homoallylic ethyl ethers.10 Switching
to platinum or rhodium at high pressures or in polar
solvents led to reduction of the phenyl ring. Finally, it
was found that a two-step platinum/palladium reaction
with adequate solvents and pressures afforded the desired
saturated alcohols 16 in 88% yield. This strategy has
recently been used by Nicolaou et al. for the synthesis of
englerin A.11 At this point, the two epimeric alcohols were
separated by column chromatography and were both
carried on to their respective bromides in nearly quantita-
tive yield. This late stage formation of the unassigned C13
stereocenter will give us rapid access to both epimers of the
macrolide and hopefully allow us to attribute stereochem-
istry by comparison of spectroscopic data.
Scheme 3. Synthesis of the Southern Subunit 7
organotin would be attached early on in the sequence
(Scheme 3).
Scheme 2. Synthesis of the Northern Subunit 5
The synthesis commenced with an Evans aldol reac-
tion between Davies’ SuperQuat oxazolidinone 912 and
known aldehyde 1013 to provide 18. Evans’ auxiliary
commonly requires two steps to regenerate the aldehyde:
cleavage to the alcohol followed by oxidation to the
aldehyde.14 In our case, oxidation of the intermediate
alcohol to the aldehyde was low yielding and nonrepro-
ducible. On the other hand, the SuperQuat auxiliary
afforded the desired aldehyde in one step. Therefore,
DIBAL cleavage of the chiral auxiliary followed by
HornerÀWadsworthÀEmmons olefination gave the
fully functionalized southern fragment 7 in three steps.
The assignment of the stereochemistry of this southern
half was described as “tenuous” by Fenical et al.,1 which
makes this short and convergent approach easily amen-
able to the synthesis of the other enantiomer.
Scheme 4. Synthesis of the Side Chain 3
In order to determine the absolute stereochemistry at
C13, the triethylsilyl group was removed from one of the
epimeric alcohols (16) and an X-ray structure of 17 was
obtained on the resulting crystalline solid.
The southern fragment was first envisioned to come
from a late stage hydrometalation of the corresponding
alkyne to form alkenylstannane 7. After extensive screen-
ing it became obvious that this strategy would not be viable
for steric reasons or due to side reactions with the unsatu-
rated ester. We therefore envisioned a route where the
(11) Nicolaou, K. C.; Kang, Q.; Ng, S. Y.; Chen, D. Y. K. J. Am.
Chem. Soc. 2010, 132, 8219.
(12) Bull, S. D.; Davies, S. G.; Nicholson, R. L.; Sanganee, H. J.;
Smith, A. D. Tetrahedron: Asymmetry 2000, 11, 3475.
(13) Lipshutz, B. H.; Clososki, G. C.; Chrisman, W.; Chung, D. W.;
Ball, D. B.; Howell, J. Org. Lett. 2005, 7, 4561.
(14) Bull, S. D.; Davies, S. G.; Nicholson, R. L.; Sanganee, H. J.;
Smith, A. D. Org. Biomol. Chem. 2003, 1, 2886.
(15) Liang, Z.; Ma, S.; Yu, J.; Xu, R. J. Org. Chem. 2007, 72, 9219.
3248
Org. Lett., Vol. 13, No. 12, 2011

With both the northern and southern halves of the
macrolide synthesized, we then concentrated on the con-
struction of the side chain (Scheme 4). Starting from
homopropargylic alcohol 20, Mitsunobu reaction with
2-thiobenzothiazole followed by m-CPBA oxidation af-
forded sulfone 21. Copper mediated addition onto
2-methallyl chloride15 gave us the complete carbon skele-
ton in three easily scalable steps. From there, Sharpless
asymmetric dihydroxylation16 resulted in chiral diol 23
(85% yield, 90% ee) which was converted to the acid by
TEMPO oxidation.17 Finally, protection of acid 24 as the
2-(trimethylsilyl)ethyl ester18 and subsequent palladium-
catalyzed Z-selective reduction of the alkyne delivered the
fully functionalized side chain 3. The trimethylsilylethyl
ester was chosen to allow the deprotection of all
silyl groups in one final step. A potentially difficult
saponification of the carboxylate of the side chain in the
presence of the macrolactone would therefore be
avoided.2
In summary, the synthesis of all the fragments of
leiodelide A has been accomplished in a short and very
convergent manner. Assembly of the macrolactone and
completion of the synthesis are currently underway and
will be reported in due time.
Acknowledgment. This work was generously supported
by The Ohio State University. We thank Dr. Judith
Gallucci (OSU) for the X-ray structure of 17.
Supporting Information Available. Experimental pro-
cedures and characterization of all compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(16) Jacobsen, E. N.; Marko, I.; Mungall, W. S.; Schroeder, G.;
Sharpless, K. B. J. Am. Chem. Soc. 1988, 110, 1968.
(17) Zhao, M.; Li, J.; Mano, E.; Song, Z.; Tschaen, D. M.; Grabow-
ski, E. J. J.; Reider, P. J. J. Org. Chem. 1999, 64, 2564.
(18) Hashimoto, T.; Maruoka, K. Synthesis 2008, 3703.
Org. Lett., Vol. 13, No. 12, 2011
3249