(+)-Rimocidin synthetic studies. Construction of an advanced C(1-18) polyol fragment

Article abstract of DOI:10.1021/ol034989g

(Matrix presented) The synthesis of an appropriately functionalized advanced C(1-18) polyol fragment of the mycosamine-glycosylated polyene macrolide, (+)-rimocidin (1), has been achieved in a highly efficient manner. Highlights of the strategy include th

Full text of DOI:10.1021/ol034989g

ORGANIC
LETTERS
2003
Vol. 5, No. 15
2751-2754
(+)-Rimocidin Synthetic Studies.
Construction of an Advanced
C(1−18) Polyol Fragment
Amos B. Smith, III,* Suresh M. Pitram, and Michael J. Fuertes
Department of Chemistry, Monell Chemical Senses Center and
Laboratory for Research on the Structure of Matter,
UniVersity of PennsylVania, Philadelphia, PennsylVania 19104
smithab@sas.upenn.edu
Received June 3, 2003
ABSTRACT
The synthesis of an appropriately functionalized advanced C(1−18) polyol fragment of the mycosamine-glycosylated polyene macrolide, (+)-
rimocidin (1), has been achieved in a highly efficient manner. Highlights of the strategy include the SN2/SN2′ addition of dithiane anions to
vinyl epoxides and the multicomponent linchpin union of 2-TBS-1,3-dithiane with two advanced epoxides.
Rimocidin (1), an architecturally interesting antifungal poly-
ene macrolide glycosylated at C(17) with (-)-mycosamine,
was isolated from Streptomyces rimosus in 1951.1 Early
structural analysis by Cope and co-workers2a led to the
assignment of the absolute configuration of the C(27)
stereocenter. However, it was not until 1995 that the complete
stereostructure was shown to comprise a 28-membered
aglycon, termed rimocidinolide, possessing an all-trans
tetraene system and nine stereogenic centers.2f Importantly,
the C(9-18) segment is common to a number of clinically
effective antifungal agents, including (+)-amphotericin B and
(+)-nystatin A1.3
In this Letter, we report an efficient, stereocontrolled
synthesis of an advanced C(1-18) polyol fragment for (+)-
rimocidin (1), appropriately functionalized for future incor-
poration into the macrolide skeleton. The synthetic strategy
calls upon both our one-flask, multicomponent linchpin union
of silyl dithianes with epoxides,4 employed to great advantage
in our recent spongistatin syntheses,5 and the highly chemose-
lective SN2 and SN2′ reactions of simple and sterically
encumbered dithianes with vinyl epoxides, also recently
disclosed by our laboratory.6 The overall strategy is outlined
in Scheme 1.
Final assembly of the advanced C(1-18) fragment 2 was
envisioned to involve union of epoxides 3 and 5 with the
anion derived from 2-TBS-1,3-dithiane (4) via the three-
(1) Davisson, J. W.; Tanner, F. W., Jr.; Finlay, A. C.; Solomons, I. A.
Antibiot. Chemother. 1951, 1, 289-290.
(2) (a) Cope, A. C.; Axen, U.; Burrows, E. P. J. Am. Chem. Soc. 1966,
87, 4221-4227. (b) Cope, A. C.; Burrows, E. P.; Derieg, M. E.; Moon, S.;
Wirth, W. D. J. Am. Chem. Soc. 1965, 87, 5452-5460. (c) Falkowski, L.;
Golik, J.; Zielinski, J.; Borowski, E. J. Antibiot. 1976, 29, 197-198. (d)
Falkowski, L.; Zielinski, J.; Golik, J.; Bylec, E.; Borowski, E. J. Antibiot.
1978, 31, 742-749; (e) Pandey, R. C.; Rinehart, K. L., Jr. J. Antibiot. 1977,
30, 146-157. (f) Sowinski, P.; Pawlak, J.; Borowski, E.; Gariboldi, P. J.
Antibiot. 1995, 48, 1288-1291.
(3) Omura, S. Macrolide Antibiotics: Chemistry, Biology, Practice;
Academic Press: New York, 1984.
(4) Smith, A. B., III; Boldi, A. M. J. Am. Chem. Soc. 1997, 119, 6925-
6926.
(5) Smith, A. B., III; Doughty, V. A.; Sfouggatakis, C.; Bennett, C. S.;
Koyanagi, J.; Takeuchi, M. Org. Lett. 2002, 4, 783-786.
(6) Smith, A. B., III; Pitram, S. M.; Gaunt, M. J.; Kozmin, S. A. J. Am.
Chem. Soc. 2002, 124, 14516-14517.
10.1021/ol034989g CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/01/2003

Scheme 1
Scheme 2
amount of silyl migration product (+)-15 were obtained in
a combined yield of 87%. Without separation, removal of
the TBS groups with TsOH, acetonide formation, and
treatment with TBAF to remove the TIPS group furnished
(+)-7 in 73% yield.
Scheme 3
component coupling protocol.4 Epoxide 3, possessing two
dithiane moieties, in turn, would arise via dithiane coupling
of 7 with an epoxide derived from 6, followed by conversion
of the acetonide to the terminal epoxide and hydrogenation
of the two olefins. Importantly, both 6 and 7 can be
assembled from vinyl epoxide (+)-9 exploiting, respectively,
the SN2 and SN2′ reaction manifolds observed upon the
addition of the lithium anions of 1,3-dithiane and 2-TIPS-
1,3-dithiane, while advanced epoxide 5 would arise from the
union of the similarly prepared ent-6 and (R)-benzyl glycidyl
ether (8). Finally, both antipodes of 9 were anticipated to be
readily available in three steps from commercial 1,4-
pentadien-3-ol.7
From the outset, we had planned to reduce the olefins
present in both (+)-7 and (+)-12 after their union. Unfor-
tunately, the adjacent allylic system in (+)-7 destabilized
the dithiane anion, preventing union with epoxide (+)-12.9
To circumvent this event, (+)-7 was hydrogenated (Wilkin-
son’s catalyst; 85%) to furnish (+)-16 (Scheme 4), which
in turn could be readily lithiated, as determined by deuterium
Scheme 4
We began the synthesis with the construction of 3.
Addition of the lithium anion of 1,3-dithiane (10) to vinyl
epoxide (+)-97 proceeded as expected, with high selectivity,
via the SN2 manifold (Scheme 2) to furnish (+)-6 in 73%
yield, along with a minor amount (15%) of (+)-11, the
product of silyl migration. Removal of the silyl groups in
both adducts and one-step Fraser-Reid epoxide formation8
[NaH, trisylimidazole (Tris-Imid), THF, 72%] then furnished
(+)-12.
To construct dithiane (+)-7 (Scheme 3), we turned to the
SN2′ manifold of vinyl epoxides, reacting lithiated 2-TIPS-
1,3-dithiane (13) with (+)-9. Adduct (+)-14 and a modest
2752
Org. Lett., Vol. 5, No. 15, 2003

incorporation. Coupling with epoxide (+)-12, however, again
proved unsuccessful. We reasoned that the dithiane in (+)-
12 undergoes competitive deprotonation leading to complex
mixtures. To test this hypothesis, a “capped” dithiane was
sought. Ideal for this purpose appeared to be the phenyl
moiety [e.g., 20; Scheme 5], as earlier studies6 had revealed
that the lithium anion of 2-phenyl-1,3-dithiane (17) also
furnishes only SN2 adducts with simple vinyl epoxides.
Equally important, removal of the dithiane at an advanced
stage in the synthesis would afford the phenyl ketone, a
reasonable surrogate for the required carboxyl group, requir-
ing only regioselective Baeyer-Villager oxidation.10
6). Diimide reduction of the alkene, a process known to
proceed in the presence of dithianes,11 was then followed
by removal of the acetonide, epoxide formation,8 and
protection of the resulting hydroxyl as the TBS ether
(TBSOTf, 2,6-lutidine). The result was bis-dithiane (-)-22,
available in four steps and 30% overall yield.
For advanced epoxide 5, we began with the enantiomers
of dithianes (+)-6 and (+)-11 prepared from (-)-9 (Scheme
7). Without separation, protection of the hydroxy functional-
Scheme 7
Scheme 5
Toward this end, addition of lithiated 2-phenyl-1,3-dithiane
(17) to vinyl epoxide (+)-9 led efficiently to SN2 adduct (+)-
18 (Scheme 5), again with a minor amount of (+)-19 derived
by silyl migration. Removal of the TBS groups in both
alcohols was achieved under acidic conditions (1% HCl/
MeOH, 92%); one-step Fraser-Reid epoxide formation8
(NaH, trisylimidazole, THF, 75%) then furnished (+)-20.
The overall yield for the three-step sequence was 61%.
Union of (+)-16, via the lithium anion, with epoxide (+)-
20 also proceeded in good yield to furnish (-)-21 (Scheme
ity in both alcohols as the TBS ethers and union of the
derived lithium anion with (-)-8 provided (-)-24. Removal
of the dithiane [Hg(ClO4)2, 2,6-lutidine, THF/H2O], followed
in turn by hydroxyl-directed reduction of the resulting ketone
[NaBH4, Et2BOMe],12 acetonide formation (75% over two
steps), treatment with TBAF to remove the silyl groups
(84%), and epoxide formation8 (NaH, trisylimidazole, THF,
71%) completed construction of (-)-5. The overall yield for
this eight-step sequence was 21%. With epoxides (-)-5 and
(-)-22 in hand, we explored the key multicomponent
coupling.
Scheme 6
(7) See Supporting Information for preparation of vinyl epoxides (+)-
and (-)-9.
(8) (a) Corey, E. J.; Weigel, L. O.; Chamberlin, A. R.; Lipshutz, B. J.
Am. Chem. Soc. 1980, 102, 1439-1441. (b) Hicks, D. R.; Fraser-Reid, B.
Synthesis 1974, 3, 203.
(9) On the basis of previous work from our laboratory, we reasoned that
union of (+)-7 and (+)-12 would be feasible. Unfortunately, quenching
experiments with MeOH-d4 revealed that lithiation of (+)-7 under numerous
conditions led primarily to decomposition; see: Smith, A. B., III; Condon,
S. M.; McCauley, J. A. Acc. Chem. Res. 1998, 31, 35-46.
(10) Hawthorne, M. F.; Emmons, W. D.; McCallum, K. S. J. Am. Chem.
Soc. 1958, 80, 6393-6398.
(11) (a) Cusack, N. J.; Reese, C. B.; Risius, A. C.; Roozpeikar, B.
Tetrahedron 1976, 32, 2157-2162. (b) Yamaguchi, Y.; Hayakawa, K.;
Kanematsu, K. J. Chem. Soc., Chem. Commun. 1987, 7, 515-516.
(12) Chen, K.-M.; Hardtmann, G. E.; Prasad, K.; Repic, O.; Shapiro,
M. J. Tetrahedron Lett. 1987, 28, 155-158.
Org. Lett., Vol. 5, No. 15, 2003
2753

Pleasingly, lithiation of 2-TBS-1,3-dithiane (4) in Et2O at
-78 °C (Scheme 8), followed by addition of epoxide (-)-5
(1.0 equiv, -78 °C), warming to -25 °C over a 1 h period,
and then addition of epoxide (-)-22 (2 equiv) in Et2O
containing HMPA (2-3 equiv) furnished (-)-26 in 56%
yield, accompanied by a minor amount of dithiane (+)-27
(13%).13
Scheme 9
Scheme 8
dithianes with an overall efficiency of 67%. Acetonide
removal was then achieved under mild acidic conditions to
furnish the mixed methyl acetal, along with modest loss of
the C(13) TBS ether. Protection of the C(17) hydroxyl as
the SEM ether (59%) completed construction of subtarget
(+)-28. Studies to effect conversion of (+)-28 to the C(1-
18) polyol fragment 2 of (+)-rimocidin are ongoing in our
laboratory.
In summary, we have achieved an effective, convergent
synthesis of (+)-28, an advanced C(1-18) polyol fragment
for the construction of (+)-rimocidin (1). Highlights of the
synthesis include the SN2/SN2′ addition of dithiane anions
to vinyl epoxides and the multicomponent linchpin union of
2-TBS-1,3-dithiane with two advanced epoxides. The longest
linear sequence from vinyl epoxide (+)-9 to (+)-28 is 13
steps and currently proceeds in 2% overall yield. Progress
toward the total synthesis of (+)-rimocidin (1) will be
reported in due course.
Having achieved construction of the complete C(1-18)
carbon backbone, we now faced the task of removing three
dithianes. To this end, silylation of the C(9) hydroxyl
(Scheme 9), followed by application of the Stork conditions
[PhI(O2CCF3)2, 2,6-lutidine, THF/H2O],14 removed the
Acknowledgment. Financial support was provided by the
National Institutes of Health (Institute of General Medical
Sciences) through grant GM-29028.
Supporting Information Available: Spectroscopic and
analytical data and selected experimental procedures. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(13) Reversing the order of addition of epoxides (-)-5 and (-)-22 in
the multicomponent coupling procedure also proved successful, yielding a
three-component adduct in 49% yield:
OL034989G
(14) Stork, G.; Zhao, K. Tetrahedron Lett. 1989, 30, 287-290.
2754
Org. Lett., Vol. 5, No. 15, 2003