COMMUNICATION
Chemoselective glycosylations using 2,3-unsaturated-4-keto glycosyl donors†
Shunichi Kusumi, Sainan Wang, Tatsuya Watanabe, Kaname Sasaki, Daisuke Takahashi and
Kazunobu Toshima*
Received 4th December 2009, Accepted 17th December 2009
First published as an Advance Article on the web 7th January 2010
DOI: 10.1039/b925587g
2,3-Unsaturated-4-keto glycosyl acetates were found to ex-
hibit low reactivity under several glycosylation conditions.
Chemoselective glycosylations were effectively performed
using 2,3-unsaturated glycosyl and 2,3-dideoxy glycosyl ac-
etates as armed glycosyl donors, and 2,3-unsaturated-4-keto
glycosyl acetates as disarmed glycosyl donors.
In the field of synthetic carbohydrate chemistry, significant
attention has been paid to chemoselective glycosylation for the
effective synthesis of oligosaccharides.1 The “armed–disarmed”
concept introduced by Fraser-Reid and co-workers has been one
of the most influential ideas in this field.2 Thus, the reactivity
of a glycosyl donor can be controlled by the combinational
Fig. 1 Comparison of 2,3-unsaturated and 2,3-unsaturated-4-keto glyco-
syl donors.
use of C2 electron withdrawing and donating protecting groups.
However, this approach cannot be directly applied to 2-deoxy
glycosyl donors due to their lack of a C2 substituent. Therefore, an
alternative strategy is required for the chemoselective glycosylation
from the activation of 1, was produced in high yield even under
conditions that produce insignificant amounts of the disaccharide
5 (Entries 1–5 vs. entries 6–10 in Table 1). In addition, in these
cases, the glycosyl donor 2 did not react and was recovered in
of 2-deoxy sugars. In this context, we earlier reported that a 2,3-
high yield (Entries 6–10 in Table 1). These results clearly show
unsaturated glycosyl donor exhibits much higher reactivity than
that the 2,3-unsaturated glycosyl donor is much more reactive
the corresponding 2,3-saturated (dideoxy) glycosyl donor.3 The
than the corresponding 2,3-unsaturated-4-keto glycosyl donor,
high reactivity of 2,3-unsaturated glycosyl donors is apparently
as expected. This tendency was essentially independent of the
due to the half-chair conformation in the ground state induced
glycosylation activator used. Furthermore, it was confirmed that
by the double bond, and stabilization of the oxocarbenium
when glycosylation using 1 (1.0 equiv.), 2 (1.0 equiv.) and 3
intermediate in the transition state by the allylic cation (Fig. 1-
(a)). Based on these findings, we anticipated that 2,3-unsaturated-
(1.0 equiv.) took place in the same flask, similar results were
◦
˚
obtained (for example, TMSOTf, MS 5A, CH2Cl2, -60 C, 1 h, 4:
93% (a:b = 67 : 33), 5: 3% (a:b = 64 : 36)).
4-keto glycosyl donors4 would show much lower reactivity than
the corresponding 2,3-unsaturated and/or 2,3-dideoxy glycosyl
donor(s). This hypothesis was based on the expectation that the
oxocarbenium intermediate, generated by the activation of the 2,3-
unsaturated-4-keto glycosyl donor, would be very unstable due to
the resonance effect of the a,b-unsaturated ketone system adjacent
to the C1 cation (Fig. 1-(b)). Here, we report efficient chemoselec-
tive glycosylations using 2,3-unsaturated-4-keto glycosyl acetates
as novel disarmed glycosyl donors.
We then examined chemoselective glycosylation using the 2,3-
unsaturated glycosyl acetate 1 as a glycosyl donor and the 2,3-
unsaturated-4-keto glycosyl acetate 6 as a glycosyl accept◦or. As
shown in Scheme 1, glycosylation using TMSOTf at -75 C for
0.5 h proceeded chemoselectively to give the desired disaccharide 7
in high yield. Disaccharide 7 possesses an acetate leaving group at
the C1 position, but no epimerization was observed. In contrast,
no oligosaccharide(s) resulting from the undesired activation
of 6 (which would lead to self-condensation) was detected.
Furthermore, the reaction between disaccharide 7 and acceptor
3 proceeded smoothly using TMSOTf at -40 ◦C for 0.5 h in PhMe
to afford trisaccharide 8 in a high yield with a-stereoselectivity. The
use of PhMe as a solvent in the second glycosylation reaction was
found to be highly effective in preventing the cleavage of the first
glycosidic bond, and in increasing the a-stereoselectivity. Based
on these results, the combination of the 2,3-unsaturated and the
corresponding 2,3-unsaturated-4-keto glycosyl donors can define
a new family of armed and disarmed glycosyl donors, respectively.
With these favourable results in hand, our attention next turned
to comparison of the reactivity of 2,3-dideoxy glycosyl donors
and 2,3-unsaturated-4-keto glycosyl donors, which are disarmed
glycosyl donors for 2,3-unsaturated glycosyl donors. In this case,
To confirm our hypothesis, we first performed competitive
glycosylations using either the 2,3-unsaturated glycosyl donor
1 (1.0 equiv.) or the 2,3-unsaturated-4-keto glycosyl donor 2
(1.0 equiv.) and a glycosyl acceptor 3 (1.0 equiv.) under several
conditions. The glycosylations of 1 with 3 and 2 with 3 were
separately conducted using TMSOTf, TBSOTf, BF3·OEt2, TfOH
or montmorillonite K-10 (MK-10) as activators; the results are
shown in Table 1. It was found that the disaccharide 4, resulting
Department of Applied Chemistry, Faculty of Science and Technology, Keio
University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama 223-8522, Japan. E-mail:
toshima@applc.keio.ac.jp; Fax: +81-45-566-1576; Tel: +81-45-566-1576
† Electronic supplementary information (ESI) available: Procedure for
chemoselective glycosylation reactions and characterization of glycosides.
See DOI: 10.1039/b925587g
988 | Org. Biomol. Chem., 2010, 8, 988–990
This journal is
The Royal Society of Chemistry 2010
©
Kusumi, Shunichi
