The polyol domain of amphidinol 3. A stereoselective synthesis of the entire C(1)-C(30) sector

Article abstract of DOI:10.1021/ol0511833

(Chemical Equation Presented) The richly oxygenated C(1)-C(30) polyol segment of amphidinol 3 has been synthesized in protected form. Incorporated in this long chain are 10 of the 25 stereogenic centers housed in the target. The asymmetric pathway that has been developed is based on the efficient union of three independently prepared subunits.

Full text of DOI:10.1021/ol0511833

ORGANIC
LETTERS
2005
Vol. 7, No. 14
3111-3114
The Polyol Domain of Amphidinol 3. A
Stereoselective Synthesis of the Entire
C(1)−C(30) Sector
Leo A. Paquette* and Shuh-Kuen Chang
EVans Chemical Laboratories, The Ohio State UniVersity, Columbus, Ohio 43210
paquette.1@osu.edu
Received May 19, 2005
ABSTRACT
The richly oxygenated C(1)−C(30) polyol segment of amphidinol 3 has been synthesized in protected form. Incorporated in this long chain are
10 of the 25 stereogenic centers housed in the target. The asymmetric pathway that has been developed is based on the efficient union of
three independently prepared subunits.
The amphidinols (AMs) comprise a small group of archi-
tecturally unique marine metabolites that exhibit potent
hemolytic and antifungal properties. AM1 was discovered
in the primitive alga Amphidinium klebsii during the routine
screening of dinoflagellates and was immediately recognized
to be the first member of an exciting new class of long-
chain polyhydroxypolyene natural products.1 The isolation
of AM2 was reported in 1995,2 to be followed by the
complete structural characterization of AM3 (1) at the hands
of Murata and co-workers in 1999.3 Among the particularly
notable features of this actinomycetous polyketide are the
25 stereogenic centers distributed over much of the flexible
backbone and two extensively oxygenated tetrahydropyrans.4
The skipped acyclic polyolefin segment, C(52)-C(67), that
forms the northern sector is also a distinctive facet of the
global ensemble.
As fascinating as the biosynthetic lineage of 1 is, its
structure and promising biological activity are the elements
that engaged our interest in its total synthesis. The successful
generation of three fragments of AM3 have previously been
reported by others. BouzBouz and Cossy devised a route to
C(1)-C(14) based on iterative enantioselective allyltitana-
tions and chemoselective cross-metathesis reactions.5
A
double-allylboration sequence was applied by Flamme and
Roush to arrive at a protected C(1)-C(25) subunit.6 Very
recently, the Rychnovsky group has successfully imple-
mented a C-glycosidation approach to the C(39)-C(52)
sector of this polyketide.7 To us, adaptation of the Julia-
Kocienski olefination8 was considered to offer a serviceable
means for assembling subunits 2 and 3 in Scheme 1. These
targeted intermediates have in turn proven to be amenable
to merging with 4 so as to arrive at the largest fragment of
1, C(1)-C(30), yet yielding to synthesis.
The synthesis of building block 2 was initiated by the
homologation of crotonaldehyde to carbinol 5a. The ensuing
(1) Satake, M.; Murata, M.; Yasumoto, T.; Fujita, T.; Naoki, H. J. Am.
Chem. Soc. 1991, 113, 9859.
(2) Paul, G. K.; Matsumori, N.; Murata, M.; Tachibana, K. Tetrahedron
Lett. 1995, 36, 6279.
(3) Murata, M.; Matsuoka, S.; Matsumori, N.; Paul, G. K.; Tachibana,
K. J. Am. Chem. Soc. 1999, 121, 870.
(4) For biosynthetic and other studies, consult: (a) Houdai, T.; Matsuoka,
S.; Murata, M.; Satake, M.; Ota, S.; Oshima, Y.; Rhodes, L. L. Tetrahedron
2001, 57, 5551. (b) Paul, G. K.; Matsumori, N.; Konoki, K.; Murata, M.;
Tachibana, K. J. Mar. Biotech. 1997, 5, 124.
(5) BouzBouz, S.; Cossy, J. Org. Lett. 2001, 3, 1451.
(6) Flamme, E. M.; Roush, W. R. Org. Lett. 2005, 7, 1411.
(7) deVicente, J.; Betzemeier, B.; Rychnovsky, S. D. Org. Lett. 2005,
7, 1853.
(8) Blakemore, P. R.; Cole, W. J.; Kocienski, P. J.; Morley, A. Synlett
1998, 26.
10.1021/ol0511833 CCC: $30.25
© 2005 American Chemical Society
Published on Web 06/10/2005

Scheme 1. Retrosynthetic Analysis of the C(1)-C(30) Subsector of Amphidinol 3
conversion to 5b resulted in critical differential blocking of
4-bromo-2-methyl-1-butene added smoothly, leading ulti-
mately to 9 after oxidation. The need to avoid alkaline
requirements persisted in 9, thereby requiring that removal
of the TBDPS functionality be brought about with TBAF in
the presence of acetic acid.10 These nonbasic conditions gave
rise to 10 in 95% yield, thereby making possible the
diastereoselective 1,3-syn reduction of this â-hydroxy ketone
with diethylmethoxyborane.11 The preparation of 2 was
completed after acetalization, PMB deprotection, and
Johnson-Lemieux cleavage of the double bond.
The conversion of D-malic acid into 13 was designed to
allow differential protection of the hydroxyl groups with TBS
at the primary position and TBDPS at the secondary (Scheme
3). Arrival at 13 allowed for the bifurcative construction of
the aldehyde ester 14 and the primary alcohol 15. Both
pathways gave cause to be concerned with competing
lactonization. For this reason, the controlled desilylation of
the three oxygenated sites (Scheme 2). The acyclic backbone
was extended by ozonolytic cleavage of the terminal double
bond in 5b prior to union of the aldehyde so formed with
the potassium salt of sulfone 12 under conventional Julia-
Kocienski conditions. The latter fragment was generated via
a Mitsunobu protocol.9 The first important observation relates
to the exclusive E configuration of 6. Another key finding
was our awareness that chemoselective desilylation of the
OTBS group in 6 in advance of asymmetric dihydroxylation
allowed for the chromatographic purification of triol 7, which
arose as the major diastereomer (4:1 dr). Aldehyde 8 was
secured by sequential acetalization with 2,2-dimethoxypro-
pane and reaction with the Dess-Martin periodinane. This
intermediate proved to be sensitive to â-elimination, thus
setting aside the possibility of its coupling to lithium-based
reagents. On the other hand, the Grignard derived from
Scheme 2. Synthesis of Hydroxy Ketone 2
3112
Org. Lett., Vol. 7, No. 14, 2005

Scheme 3. Synthesis of Phosphonium Salt 3
13 was induced with CSA in CH3OH/CH2Cl2 at -10 °C.
positioned at this site and both isomers hold comparable
synthetic importance. They were therefore processed inde-
pendently through the later steps. Excellent functional group
tolerance was operational when 21 was treated with NBS in
aqueous DMSO. Under these conditions, conversion to the
alcohol proceeded smoothly with exclusive loss of the TBS
group. Subsequent periodinane oxidation was also efficient,
leading uniquely to the targeted aldehyde 22.
Temperature control proved to be critical to the realization
of an 86% yield of the hydroxy ester. The latent potential
for ring closure during Dibal-H reduction was similarly
minimized by maintaining a temperature no greater than
-10 °C and not exceeding this upper limit during ensuing
treatment with NaBH4.
With these two subunits in hand, conditions for their union
were investigated. To evaluate the applicability of the Julia-
Kocienski olefination in this setting, 15 was transformed into
sulfone 16. Reliable conditions were soon uncovered for
efficient reaction with 14 to provide 17 in an E/Z ratio of
4:1. This isomer distribution is not of consequence since
arrival at alcohol 18 is founded on subsequent catalytic
hydrogenation, followed by hydride reduction of the ester
functionality. The resulting alcohol 18 was transformed via
the iodide into the phosphonium salt 3.
The next series of transformations involved the union of
2 to 3. The oxidation of 2 with the Dess-Martin reagent
proceeded straightforwardly to generate keto aldehyde 19,
whose Wittig reaction with 3 proceeded with involvement
of the carboxaldehyde functionality and resulted in formation
of the (Z)-olefin (1H NMR analysis), the catalytic hydrogena-
tion of which provided 20 in 77% overall yield (Scheme 4).
The preparation of subunit 21 was completed by sequential
sodium borohydride reduction and Mitsunobu reaction with
1-phenyl-1H-tetrazole-5-thiol.8 This application of the SN2
reaction expectedly required longer times (up to 2 days) to
proceed to completion in view of the secondary nature of
the seat of reaction. The two diastereomers of 21 (dr ) 5:1)
could be secured in pure form by chromatography on silica
gel. Our inability to assign configuration to C(30) in 21 is
of little consequence since a double bond is ultimately
We next undertook to explore the development of a
possible route to 4. Gratifyingly, recourse to 23 as the starting
point proved to be well suited to the task at hand. Reduction
of dimethyl (S)-malate as reported by Saito et al. with
subsequent acetonide formation provided 23.12 Once again,
the Mitsunobu reaction was employed to activate the terminal
primary carbon. Ensuing molybdate-promoted oxidation9
furnished 24, the availability of which made possible efficient
coupling to aldehyde 14, which had previously been gener-
ated. This olefination afforded 25 as a 3:1 mixture of (E)-
and (Z)-isomers (1H NMR analysis). This isomeric ratio was
increased to 6:1 by radical-induced isomerization involving
thiophenol and AIBN in refluxing benzene over a period of
1 day.13 Extension of the reaction time to 3 days led to a
still more favorable distribution of 12:1, a value that remained
unchanged at more prolonged time intervals. Although access
to alcohol 26 and sulfone 27 was realized with high
efficiency, neither product lent itself to chromatographic
separation of the major constituent. This objective was,
however, conveniently met by replacement of the acetonide
group by two tert-butyldiphenylsilyl residues as in 4.
With 4 and 22 in hand, their merger in the Julia-Kocienski
olefination was accomplished in the usual one-pot operation.
A significant bias toward trans stereoselection (dr ) 8.6:1)
was evident from the large coupling constant between H(8)
and H(9) (J ) 15.6 Hz) observed in the dominant product.
(9) Smith, A. M., III; Adams, C. M.; Kozmin, S. A. J. Am. Chem. Soc.
2001, 123, 990.
(10) Chauhan, K.; Bhatt, R. K.; Falck, J. R.; Capdevila, J. H. Tetrahedron
Lett. 1994, 35, 1825.
(11) Wanderwal, C. D.; Weller, S.; Sorensen, E. J. J. Am. Chem. Soc.
2003, 125, 5393.
(12) Saito, S.; Ishikawa, T.; Kuroda, A.; Koga, K.; Moriwake, T.
Tetrahedron 1992, 48, 4067.
(13) Schwarz, M.; Graminski, G. F.; Waters, R. M. J. Org. Chem. 1986,
51, 260.
Org. Lett., Vol. 7, No. 14, 2005
3113

Scheme 4. Assembly of the Protected C(1)-C(30) Polyhydroxylated Chain
Further oxidation was accomplished with a complex of
building blocks directly suited to the present objectives.
Effort is now being directed toward the completion of an
efficient total synthesis of natural amphidinol 3.
ammonium molybdate and hydrogen peroxide to deliver 28.
In conclusion, the convergent route to 28 documented here
offers considerable insight into an approach toward construc-
tion of a polyoxygenated chain by repeated application of
the Kocienski modification of the Julia reaction. Another key
to our success includes the economic manner in which the
antipodal malic acids can be transformed into enantiopure
Supporting Information Available: Experimental details
1
and H NMR spectra for all products. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL0511833
3114
Org. Lett., Vol. 7, No. 14, 2005