zylated trifluoroacetyl glucose.8 Martin et al. developed a
synthesis of 3 based on an intramolecular C-glycosylation
of a 2-(3′,4′,5′-trimethoxy)benzyl n-pentenyl glucoside fol-
lowed by oxidation of the benzylic methylene group. 8,10-
Di-O-methylbergenin 3 was prepared in 12.1% yield over
eight steps from peracetylated glucosyl bromide.9 Apart from
the modest overall yield, these syntheses require numerous
protecting group manipulations, thus rendering them unsuit-
able for the synthesis of an extended set of bergenin
derivatives.
Scheme 2. Installation of the C-Aryl Linkage
Here we describe a short and high-yielding total synthesis
of 3 based on an O-to-C rearrangement with 3,4,5-trimethox-
yphenol. A common route to fashion C-aryl glycosidic
linkages involves the initial installation of an O-glycosidic
linkage with an electron-rich phenol, followed by a Fries-
like O-to-C rearrangement (Scheme 1). A glycosyl donor is
of this hydroxyl group, the addition of DMAP as an acylation
catalyst proved to be essential.
Scheme 1 O-to-C Rearrangement
Scheme 3. Completion of the Bergenin Structure
activated to generate an electrophilic anomeric species that
couples to an aromatic phenol to afford an O-glycoside. The
initial O-glycoside then rearranges to the C-aryl bond under
Lewis acidic conditions. The O-to-C conversion exclusively
affords the sterically favored â-C-aryl phenolic glucoside
product.10 We envisioned that the O-to-C rearrangement
product of 3,4,5-trimethoxyphenol could be further func-
tionalized via a carbonylation and subsequent lactone forma-
tion to give the desired bergenin scaffold.
Palladium(0)-catalyzed carbonylation13 at ambient pressure
under careful exclusion of water and oxygen yielded 68%
of C-glucosyl benzoic acid derivative 7. Attempts to convert
6 directly into the C-glycosyl methyl benzoate failed,14 as
did attempts to obtain the methyl ester of 7 by direct
C-glycosylation with methyl 3,4,5-trimethoxybenzoate.
Debenzylation of 7 by hydrogenation with Pearlman’s
catalyst in methanol provided the tetrahydroxyl C-glucoside
in quantitative yield. Regioselective lactonization of the C2
hydroxyl group of the glucose scaffold was achieved by
treatment with DCC/DMAP in DMF to furnish 3 in 71%
yield. A more efficient procedure was the treatment of
deprotected 7 with SOCl2 in methanol to provide cleanly
8,10-di-O-methylbergenin in 90% yield.
Upon activation with trimethylsilyl trifluoromethane-
sulfonate (TMSOTf), 2,3,4,6-tetra-O-benzyl glucopyranosyl
trichloroacetimidate 4a11 and perbenzylated glucosyl diphenyl
phosphate 4b12 reacted with 3,4,5-trimethoxyphenol to give
the â-configured C-glycoside 5 in 67 and 57% yields,
respectively (Scheme 2). Treatment of the phenolic hydroxyl
group of 5 with triflic anhydride/lutidine resulted in the
formation of triflate 6 (Scheme 3). Due to steric hindrance
In conclusion, we have developed a short synthesis of the
bergenin scaffold that provides the target structure 3 in 33%
(8) Frick, W.; Schmidt, R. R. Carbohydr. Res. 1991, 209, 101.
(9) Rousseau, C.; Martin, O. R. Tetrahedron: Asymmetry 2000, 11, 409.
(10) Mahling, J.-A.; Schmidt, R. R. Synthesis 1993, 325.
(11) Nakanishi, N.; Nagafuchi, Y.; Koizumi, K. J. Carbohydr. Chem.
1994, 13, 981.
(13) Cacchi, S.; Lupi, A. Tetrahedron Lett. 1992, 33, 3939.
(14) Dolle, R. E.; Schmidt, S. J.; Kruse, L. I. J. Chem. Soc., Chem.
Commun. 1987, 904.
(12) Sebasan, S.; Neira, S. Carbohydr. Res. 1992, 223, 169.
2966
Org. Lett., Vol. 4, No. 17, 2002