Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Â
L. Somsak, V. Nagy / Tetrahedron: Asymmetry 11 (2000) 1719±1727
1720
analogue inhibitors of muscle GPb known to date is a glucopyranosylidene-spiro-hydantoin4,5,11
9 (Ki=3.1 mM11 or 4.2 mM12 for muscle GPb). Several synthetic routes4,5,11 have been suggested
for the preparation of 9. Unfortunately, all of them consist of rather many steps (8±10 from an
available starting material or even more from the appropriate free sugar) or are stereochemically
inecient, producing the C-6 epimer of 9 as the major product which is a much weaker inhibitor
(Ki=320 mM4 or 105 mM12). These problems clearly hinder more sophisticated biochemical or
biological investigations with inhibitor 9.
Recently we have shown12 that a slight modi®cation of 9, namely changing the C-8 carbonyl to
a thiocarbonyl group as in 8, brings about practically no change in the inhibition of muscle GPb
(Ki=5.1 mM12). Similar inhibitor constants for 8 and 9 have been obtained for muscle GPa and
liver GPa and b enzymes12 as well, rendering these two compounds equipotent inhibitors. In
contrast to 9, compound 8 can advantageously be synthesized in a simple, six-step, highly
stereoselective procedure starting from d-glucose. This method could open the way for preparing
larger amounts of 8 required for more elaborate biological investigations, in order to understand
further the mechanism of action of GPs and validate the concept of their inhibition as a potential
therapy of NIDDM. However, to have a really practical synthesis several modi®cations have still
been needed to improve the ꢀ2% overall yield obtained for 8 from d-glucose by the published
protocol. Thus, the seemingly unavoidable formation of hydroxy-amide 6 (Ac instead of Bz)
accompanied in the ring closing step has been signi®cantly reduced, as will be described in a
forthcoming paper.13 Next, the preparation of the key starting material b-d-glucopyranosyl
cyanide (like 2) has had to be amended, because the low yield (11%) of its acetyl-protected derivative
has been the main reason for the very low overall yield of the sequence. The use of benzoyl protecting
groups, as disclosed in this paper, has also reduced the number of chromatographic separations
and facilitated the preparation of the important enzyme inhibitor 8 in gram quantities.
2. Results and discussion
The preparation of acetylated d-glycopyranosyl cyanides was studied with extreme care by
Myers and Lee.14,15 The best procedure proposed by them for 2,3,4,6-tetra-O-acetyl-b-d-gluco-
pyranosyl cyanide, i.e. fusion of acetobromo-glucose with mercury(II) cyanide15 gave in our hands
generally 10±12% of the target compound isolated by several consecutive crystallizations and/or
chromatographic puri®cation. There is no information in the literature to explain why the
cyanation of acetobromo-glucose is so unselective under various conditions. The most important
by-products are 1,2-O-(1-cyano-ethylidene) derivatives formed by an attack of the cyanide on the
probable 1,2-acetoxonium ion intermediate and penta-O-acetyl-d-glucopyranoses.14,15 Another
possibility for the preparation of the acetylated b-d-glucopyranosyl cyanide could have been the
procedure of Koll and Fortsch starting from the corresponding acetylated C-glycosyl nitro-
methane.16 However, chromatographic separations would have been required even in this case.
Therefore, we decided to use benzoyl protecting groups with the expectation of also promoting a
higher tendency for crystallization of the benzoylated sugar derivatives. Another expected
advantage was that in certain cases a lower tendency for formation of 1,2-orthoester derivatives
from the intermediate 1,2-acyloxonium ion has been reported for 2-O-benzoylated sugars relative
to their 2-O-acetylated counterparts.17
Benzobromo-glucose 1 obtained by a known procedure from penta-O-benzoyl-d-gluco-
pyranoses with HBr in acetic acid18 was treated with mercury(II) cyanide in dry nitromethane at