A. N. Parker et al. / Tetrahedron Letters 54 (2013) 5322–5324
5323
Scheme 3. Synthesis of gymnoascolide A (2).
Scheme 1. Retrosynthesis of butenolides 1–3.
Scheme 4. Synthesis of eutypoid A (3).
group could be removed prior to the installation of the C3 aryl ring
(Scheme 3). Thus, hydrolysis of acetal 8, followed by methanolic
NaBH4 reduction gave 3-bromofuranone 9 in good yield.11b Suzuki
coupling of vinyl bromide 9 with phenylboronic acid afforded gym-
noascolide A (2) in 91% yield.
Scheme 2. Synthesis of racemic microperfuranone (1).
Finally, eutypoid A (3) was prepared from 3-bromofuranone 9
via a Suzuki coupling/demethylation sequence (Scheme 4). The
coupling of 4-methoxyphenylboronic acid with 3-bromofuranone
9 gave 3-(4-methoxyphenyl)furanone 12 in excellent yield. Subse-
quent demethylation of the methyl-aryl ether with BCl3 and Bu4NI
afforded eutypoid A (3) in 87% yield.16
In summary, we have developed an efficient and operationally
simple approach to the 4-benzyl-3-phenylbutenolide core utilizing
two well established methodologies: the Suzuki–Miyaura coupling
reaction and the Mannich-type reaction involving glyoxylic acid.
With this synthetic strategy we are able to report the first total
synthesis of eutypoid A (3) (8 steps, 46%), as well as an improved
synthesis of racemic microperfuranone (1) (6 steps, 56% yield)
and gymnoascolide A (2) (7 steps, 49%).
c
-Hydroxybutenolides [5-hydroxy-2(5H)-furanones] can be
readily prepared through the oxidation of furans or by Mannich-
type reactions.10,11 The Mannich-type amino alkylation involving
glyoxylic acid, an enolizable aldehyde, and a secondary amine
was first developed by Wermuth and Bourguignon in 1981.11a They
had demonstrated that 4-alkyl-5-hydroxy-2(5H)-furanones could
easily be obtained in this one-pot reaction methodology. Given
that the 4-benzylbutenolide core could be prepared in a single
step, we sought to utilize this methodology in the synthesis of
three natural products 1–3 (Scheme 1).
To date, there has been no reported synthesis of any member of
the eutypoid family. Furthermore, we felt our synthetic approach
would be applicable to the synthesis of racemic microperfuranone
(1)12 and gymnoascolide A (2),13 offering a more efficient and oper-
ationally simple strategy. The key steps in our approach involve the
Suzuki coupling14 of an aryl boronic acid with the appropriate 4-
benzyl-3-bromofuranone (8 or 9) and a Mannich-type aminoalky-
lation of commercially available 3-phenylpropionaldehyde
(Scheme 1).11
Acknowledgments
Financial support for this project was provided by AUM Faculty
Grant-In-Aid and the Physical Sciences Department of Auburn Uni-
versity Montgomery.
Initial attempts to prepare 4-benzyl-5-hydroxyfuranone 10 in a
single step via condensation of 3-phenylpropionaldehyde with gly-
oxylic acid in the presence of piperdine hydrochloride resulted in
lower than expected yields (50–53%). The low yields were due to
incomplete elimination of the initially formed amino product.
Optimization of this reaction sequence was achieved by conduct-
ing the Mannich-type aminoalkylation/elimination in two separate
steps affording furanone 10 in 78% yield (Scheme 2).
Conversion to the methoxy acetal followed by a bromination
in situ dehydrobromination sequence gave 3-bromofuranone 8 in
excellent yield. Using the reaction conditions reported by Zhang
et al., the Suzuki coupling of 3-bromofuranone 8 with phenylbo-
ronic acid proceeded smoothly, affording 3-phenylfuranone 11 in
97% yield.14c Finally, racemic microperfuranone (1) was obtained
in 85% yield following hydrolysis of the methoxy acetal.15
Gymnoascolide A (2) was prepared by two different routes.
Methanolic NaBH4 reduction of microperfuranone (1) afforded
gymnoascolide A (2) in 87% yield. Alternately, the C5 hydroxyl
Supplementary data
Supplementary data (experimental procedures, characteriza-
tion data, 1H and 13C NMR spectra for all compounds) associated
with this article can be found, in the online version, at http://
References and notes