38
B. Schmidt
in three routine steps. The same enantiomerically pure starting material 116 was also
used for the synthesis of the C-5–C-9 fragment: ring opening of 116 with the Liacetylide of 1-octyne, protection of the resulting secondary alcohol as a TBS-ether,
hydrogenation of the triple bond and concomitant cleavage of the PMB-ether, oxidation of the primary alcohol to an aldehyde and finally a Wittig-olefination furnished
the unsaturated ester 119. The stereocenters C-6 and C-7 were installed in the required
configuration using Sharpless asymmetric dihydroxylation to yield 120, which was
protected as an acetonide, followed by a one-pot reduction of the ester to an aldehyde
and Wittig olefination, cleavage of the TBS-ether, and Yamaguchi esterification with
118. Ring-closing metathesis of the product, diene 121, was investigated but required
a very high catalyst loading of 20 mol% of second-generation Grubbs’ catalyst B1.
Ring-closing metathesis of a fully deprotected precursor (obtained after cleavage of
the TBS ether and the acetonide) gave a complex mixture of products. The best result
was obtained if the TBS-ether was cleaved prior to the RCM step, and the acetonide
was removed after RCM. Via this route the catalyst loading could be lowered to 10
mol% and the product 114 was obtained with high (E)-selectivity. In contrast to the
original report by Hiep et al. and the two simultaneously published total syntheses
(Table 2, entries 1 and 2) compound 114 was found to have not only a significantly
different value of specific rotation but also a positive sign of specific rotation, whereas
all other analytical data (in particular NMR data) matched those previously reported
for seimatopolide A very well. From these observations the authors conclude that
the absolute configuration of seimatopolide A was erroneously assigned and that the
actual structure is ent-114, with a (3S,6S,7S,9R)-configuration of the stereocenters.
The proposed revision of absolute configuration was shortly afterwards confirmed
by total syntheses of both enantiomers of seimatopolide A using chiral-pool-derived
starting materials (Table 2, entries 4 and 5) [129]. Starting from l-(+)-tartrate the
C 2 -symmetric diol (S,S)-111 was synthesized in four steps following a previously
published route [139]. The C 2 -symmetry of this enantiomerically pure building block
was exploited for differentiation of the C=C-double bonds: selective monoprotection was accomplished with the sterically demanding OH-protecting group triphenylmethyl (step 1). The nonyl substituent at C-9 was installed by cross-metathesis
with 1-undecene (step 2), catalyzed by the phosphine-free Umicore-M 51 catalyst
B3, which is particularly well suited for CM reactions. Sharpless epoxidation of the
newly generated C=C double-bond (step 3), followed by regioselective reductive
epoxide cleavage with sodium bis(2-methoxyethoxy)aluminum hydride (Red-Al®)
and acid-catalyzed cleavage of the triphenylmethyl-protecting group (step 4) established stereocenter C-9. Finally, the C-6–C-7-diol was regioselectively protected as
an acetonide. In the next step, the C-5–C-9-fragment 122 was coupled with carboxylic
acid (S)-118, the C-1–C-4 fragment, using Shiina’s conditions (cf. Scheme 6). RCM
of diene 123 proceeded with very high (E)-selectivity. Simultaneous deprotection of
the C-6–C-7-acetonide and the C-3-TBS-ether was accomplished with trifluoroacetic
acid. All analytical data obtained for the product ent-114 matched those reported for
natural seimatopolide A, including the sign of specific rotation. The non-natural enantiomer of seimatopolide A, ent-seimatopolide A (114), was synthesized through the
B. Schmidt
in three routine steps. The same enantiomerically pure starting material 116 was also
used for the synthesis of the C-5–C-9 fragment: ring opening of 116 with the Liacetylide of 1-octyne, protection of the resulting secondary alcohol as a TBS-ether,
hydrogenation of the triple bond and concomitant cleavage of the PMB-ether, oxidation of the primary alcohol to an aldehyde and finally a Wittig-olefination furnished
the unsaturated ester 119. The stereocenters C-6 and C-7 were installed in the required
configuration using Sharpless asymmetric dihydroxylation to yield 120, which was
protected as an acetonide, followed by a one-pot reduction of the ester to an aldehyde
and Wittig olefination, cleavage of the TBS-ether, and Yamaguchi esterification with
118. Ring-closing metathesis of the product, diene 121, was investigated but required
a very high catalyst loading of 20 mol% of second-generation Grubbs’ catalyst B1.
Ring-closing metathesis of a fully deprotected precursor (obtained after cleavage of
the TBS ether and the acetonide) gave a complex mixture of products. The best result
was obtained if the TBS-ether was cleaved prior to the RCM step, and the acetonide
was removed after RCM. Via this route the catalyst loading could be lowered to 10
mol% and the product 114 was obtained with high (E)-selectivity. In contrast to the
original report by Hiep et al. and the two simultaneously published total syntheses
(Table 2, entries 1 and 2) compound 114 was found to have not only a significantly
different value of specific rotation but also a positive sign of specific rotation, whereas
all other analytical data (in particular NMR data) matched those previously reported
for seimatopolide A very well. From these observations the authors conclude that
the absolute configuration of seimatopolide A was erroneously assigned and that the
actual structure is ent-114, with a (3S,6S,7S,9R)-configuration of the stereocenters.
The proposed revision of absolute configuration was shortly afterwards confirmed
by total syntheses of both enantiomers of seimatopolide A using chiral-pool-derived
starting materials (Table 2, entries 4 and 5) [129]. Starting from l-(+)-tartrate the
C 2 -symmetric diol (S,S)-111 was synthesized in four steps following a previously
published route [139]. The C 2 -symmetry of this enantiomerically pure building block
was exploited for differentiation of the C=C-double bonds: selective monoprotection was accomplished with the sterically demanding OH-protecting group triphenylmethyl (step 1). The nonyl substituent at C-9 was installed by cross-metathesis
with 1-undecene (step 2), catalyzed by the phosphine-free Umicore-M 51 catalyst
B3, which is particularly well suited for CM reactions. Sharpless epoxidation of the
newly generated C=C double-bond (step 3), followed by regioselective reductive
epoxide cleavage with sodium bis(2-methoxyethoxy)aluminum hydride (Red-Al®)
and acid-catalyzed cleavage of the triphenylmethyl-protecting group (step 4) established stereocenter C-9. Finally, the C-6–C-7-diol was regioselectively protected as
an acetonide. In the next step, the C-5–C-9-fragment 122 was coupled with carboxylic
acid (S)-118, the C-1–C-4 fragment, using Shiina’s conditions (cf. Scheme 6). RCM
of diene 123 proceeded with very high (E)-selectivity. Simultaneous deprotection of
the C-6–C-7-acetonide and the C-3-TBS-ether was accomplished with trifluoroacetic
acid. All analytical data obtained for the product ent-114 matched those reported for
natural seimatopolide A, including the sign of specific rotation. The non-natural enantiomer of seimatopolide A, ent-seimatopolide A (114), was synthesized through the