transformed to the corresponding mixture of R,â-unsaturated
esters 8 in four steps. Despite extensive efforts, this diaster-
eomeric mixture was found to be chromatographically
inseparable at any step in the sequence. Fortunately, the
diastereomeric mixture of the primary alcohols derived from
8 was found to be readily separable with column chroma-
tography to give the stereochemically pure 9 and its
diastereomer in approximately 30% and 8% overall yield
from 6. Via correlation with the sample previously pre-
pared,12 the relative and absolute configuration of the major
alcohol 9 was established as indicated, which matched with
the major diastereomer predicted for AD-mix-R dihydroxy-
lation.10
geometric isomers by silica gel chromatography [hexanes-
ethyl acetate (6:1)].
Esterification of the mycolactone core alcohol 3 with the
unsaturated fatty acids 4 was accomplished via the Yamagu-
chi protocol to furnish the protected mycolactones 13 as an
approximately 3:2 mixture of 4′Z and 4′E isomers in 90%
yield (Scheme 5).16,17 It was hoped that all of the protecting
Scheme 5a
Employing a minor modification of the procedure reported
by Gurjar and Cherian,8 the phosphonate (2′E,4′E,6′E)-11
with a methyl ester at C1′ was prepared.13 The anion
generated from 11 was expected to exist as a mixture of
geometric isomers. Indeed, deprotonation of 11 with LDA
in THF at -78 °C, followed by quenching with 2,4,6-
trimethylphenol, gave a mixture composed of 55%
[(2′E,4′E,6′E)-11], 25% [(2′E,4′E,6′Z)-11], 14% [(2′E,4′Z,6′E)-
11], and 6% [(2′E,4′Z,6′Z)-11].14 Upon treatment (-78 °C
f rt) with aldehyde 10, the anion gave a mixture of
(2′E,4′E,6′E,8′E,10′E)-12 (73%), (2′E,4′Z,6′E,8′E,10′E)-12
(17%), (2′E,4′E,6′Z,8′E,10′E)-12 (7%), and (2′E,4′Z,6′Z,8′E,-
10′E)-12 (3%). The structure of these geometric isomers was
deduced from analysis of 1H NMR spectra.14,15 The composi-
tion of this mixture varied from experiment to experiment,
but the ratio indicated was representative. By photolysis
(acetone-d6, tungsten lamp) or during isolation under com-
mon laboratory conditions, this composition changed to a
mixture of (2′E,4′E,6′E,8′E,10′E)-12 (36%), (2′E,4′Z,6′E,
8′E,10′E)-12 (52%), (2′E,4′E,6′Z,8′E,10′E)-12 (4%),
(2′E,4′Z,6′Z,8′E,10′E)-12 (5%), and a fifth isomer (3%).
a Reagents and Conditions: (a) Cl3C6H2COCl, i-Pr2NEt, DMAP,
PhH, rt, 20 h, 90%. (b) TBAF, THF, rt, 1 h, 81%. (c) 2:2:1 THF/
HOAc/H2O, rt, 10 h, and the recovered starting material was
recycled (once), 67%.
1
Close examination of the H NMR spectrum revealed that
the fifth isomer was a mixture of two 10′Z geometric isomers
corresponding to the two major isomers.15 As noted by
others,3,8 4′Z and 4′E geometric isomers were interconvert-
ible, and the 3:2 ratio appeared to represent the ratio at the
steady state under photochemical conditions and common
laboratory conditions. Although these geometric isomers
were chromatographically inseparable at the ester stage, the
groups in 13 could be removed in a single step. Indeed,
treatment of 13 with HF‚Py in MeCN gave the synthetic
mycolactones, but only in 5-10% yield. The low yield was
attributed largely to the instability of the products under the
deprotection conditions employed.18 To eliminate or suppress
the undesired side reactions, deprotection was carried out in
two separate steps. The three TBS groups were removed
under standard conditions (TBAF/THF/rt), to yield the triol
14 as an approximately 3:2 mixture of 4′Z and 4′E isomers
in 81% yield. The cyclopentylidene ketal of 14 was then
hydrolyzed by treatment with aqueous acetic acid [AcOH/
H2O/THF (2:1:2)] at room temperature. It should be noted
that, even under these conditions, the side reactions on the
corresponding
acids
(2′E,4′E,6′E,8′E,10′E)-4
and
(2′E,4′Z,6′E,8′E,10′E)-4 could be separated from the other
(12) The details for structural correlation of 9 with the authentic sample
reported in ref 4b is included in Supporting Information.
(13) 1H NMR (CDCl3, 500 MHz) of (2′E,4′E,6′E)-11: δ 7.35 ppm (1H,
d, J ) 16.0 Hz), 6.28 (1H, s), 5.89 (1H, d, J ) 15.5), 5.52 (1H, m), 4.12
(4H, m), 3.77 (3H, s), 2.72 (2H, dd, J ) 22.5, 7.5), 1.95 (3H, s), 1.86 (3H,
d, J ) 4.0), 1.33 (6H, t, J ) 7.0).
(14) The stereochemistry of geometric isomers was assigned via NOE
experiments on the olefinic protons.
(15) The H3′ signal in the 1H NMR spectrum (acetone-d6) is diagnostic
to differentiate these geometric isomers: δ 7.35 ppm (d, J ) 15.5 Hz) for
(2′E,4′E,6′E,8′E,10′E)-12; 7.92 (d, J ) 15.5) for (2′E,4′Z,6′E,8′E,10′E)-
12; 7.36 (d, J ) 15.5) for (2′E,4′E,6′Z,8′E,10′E)-12; 7.43 (d, J ) 15.5) for
(2′E,4′Z,6′Z,8′E,10′E)-12. The fifth isomer observed after photolysis seemed
to be a mixture of two geometric isomers, corresponding to (2′E,4′E,6′E,
8′E,10′Z)-12 and (2′E,4′Z,6′E,8′E,10′Z)-12. In addition to the H3′ proton,
the H11′ resonance is diagnostic to assign the structure for these isomers:
δ 5.58 ppm (d, J ) 8.5 Hz) for (2′E,4′E,6′E,10′Z)-12 and 5.62 (d, J ) 8.5)
for (2′E,4′Z,6′E,10′Z)-12; 5.67 (d, J ) 8.5) for (2′E,4′E,6′E,10′E)-12 and
5.66 (d, J ) 8.5) for (2′E,4′Z,6′E,10′E)-12.
(16) (a) Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M.
Bull. Chem. Soc. Jpn. 1979, 52, 1989. (b) Hikota, M.; Sakurai, Y.; Horita,
K.; Yonemitsu, O. Tetrahedron Lett. 1990, 31, 6367.
(17) No esterification was observed with EDCI/DMAP or BOP/DMAP.
(18) Byproduct formation appeared to be initiated by a Michael addition
of a hydroxyl group, followed by a secondary Michael addition(s), and
resulted in a complex mixture of the side products. Attempted base- and
acid-treatments yielded only a trace amount of the desired products.
Org. Lett., Vol. 4, No. 4, 2002
649