3
SO2Cl2-TfOH, CF3SO3CH3, dimethyl(methylthio) sulfonium
trifluoromethanesulfonate (DMTST), nitroso tetrafluoroborate
(NOBF4), and BF3.OEt2) were applied to the reaction but none of
them yielded the desired compound. With little success on the
direct pyruvylation of the trans-diol, we decided to use
acetoxyacetone as a precursor. To effect the transformation, we
activated the hydroxyl groups by protecting them with a silyl
group (Scheme 3, compound 21). With these changes, formation
of the acetal proceeded readily to give compound 22 in very good
yield. Subsequently, the acetyl protecting groups were removed
to form 23. A selective oxidation of the primary alcohol into
carboxylic acid followed by a methyl esterification of the acid
furnish the precursor 24.
addition, to confirm the absolute configuration of the CH3 on
the pyruvate ring, monomer 24 was initially subjected to 2D
experiments but no conclusive results were achieved. Increasing
the steric bulk on the ester by changing the methyl ester to benzyl
ester (Scheme 5) improved the degree of interactions, resulting in
successful 2D experiments. The proton at the C-3 position could
be derived with the use of COSY followed by NOESY, allowing
an interaction between the C-3 proton and the –CH3 of the
pyruvate group to be observed (See Supporting Information).
With the individual carbohydrate fragments synthesized, we
proceeded to glycosylate them. Compounds 13 and 24 were
glycosylated to give disaccharide 25 in good selectivity and yield
(Scheme 4). With an acetyl directing group at C-2 of compound
13, glycosylation gave only the -anomer of the disaccharide.
Thereafter, the ester groups on 25 were hydrolyzed. The
carboxylic acid groups were subsequently protected with a
benzyl ester 26 so as to enable global deprotection via catalytic
hydrogenation in the final step. Next, dibutyltin oxide in the
presence of benzyl bromide was used to benzylate either the C-2
or C-3 hydroxy group, but not both. This reaction provided
compound 27 and the C-3 benzylated product in 58% and 35%
yield respectively. Compound 27 was subsequently glycosylated
with fucose 12 to afford the protected trisaccharide 28 in good
stereoselectivity. With the protected trisaccharide 28 synthesized,
global deprotection of the remaining benzyl ether groups using
Pd/C in the presence of hydrogen gas gave trisaccharide 1.
Scheme 5 Synthesis of glucuronic acid fragment 29
Conclusions
In summary, we report the synthesis of the trisaccharide unit
of CPS from the liver abscess strain K. pneumoniae. The
synthetic transformations involved three building blocks 12, 13
and 24 which were eventually glycosylated into trisaccharide 1.
The overall efficiency of the assembly process benefitted from a
set of appropriately selected protecting groups. Complete
stereoselectivity for each glycosylation was achieved via the
orthogonality and participating effects of the blocking groups. In
the course of this synthesis, we have also demonstrated a new
methodology to effect pyruvylation of trans-diol.
Acknowledgments
The authors thank the Ministry of Education (Singapore) for
the financial support provided (MoE Tier 2 Grant: R-143-000-
589-112).
References and notes
1
2
(a) R. Podschun, U. Ullmann, Clin. Microbiol. Rev. 11 (1998) 589-603.
(b) R. Ramphal, P. G. Ambrose, Clin. Infect. Dis. 42 (2006) S164-S172.
(a) Y. Kohayagawa, K. Nakao, M. Ushita, N. Niino, M. Koshizaki, Y.
Yamamori, Y. Tokuyasu, H. Fukushima, J. Infect. Chemother. 15 (2009)
248-251.
(b) L. K. Siu, K. M. Yeh, J. C. Lin, C. P. Fung, F. Y. Chang, Lancet
Infect. Dis. 12 (2012) 881-887.
3
4
W. Chen, C. H. Chen, K. L. Chiu, H. C. Lai, K. F. Liao, Y. J. Ho, W. H.
Hsu, Crit. Care Med. 36 (2008) 1184-1188.
(a) S. Zamze, L. Martinez-Pomares, H. Jones, P. R. Taylor, R. J. Stillion,
S. Gordon, S. Y. Wong, J. Biol. Chem. 277 (2002) 41613-41623.
(b) C. T. Fang, Y. P. Chuang, C. T. Shun, S. C. Chang, J. T. Wang, J.
Exp. Med. 199 (2004) 697-705.
5
6
S. J. Jr. Cryz, E. Fürer, R. Germanier, Infect. Immun. 54 (1986) 403-407.
F. L. Yang, Y. L. Yang, P. C. Liao, J. C. Chou, K. C. Tsai, A. S Yang, F.
Sheu, T. L. Lin, P. F. Hsieh, J. T. Wang, K. F. Hua, S. H. Wu, J. Biol.
Chem. 24 (2011) 21041-21051.
7
(a) R. D. Astronomo, D. R. Burton, Nat. Rev. Drug Discov. 9 (2010)
308-324.
8
(a) L. G. Bennett, C. T. Bishop, Immunochem. 14 (1977) 693-696.
(b) K. Hiruma, J. Tamura, H. Hashinoto, Carbohyd. Res. 282 (1996) 299-
306.
Scheme 4 Synthesis of trisaccharide 1 from the individual
carbohydrate fragments.
(c) I. Bajza, J. Kerékgyárto, J. Hajkó, L. Szilágyi, A. Lipták, Carbohyd.
Res. 253 (1994) 111-120.
(d) H. Lõnn, Carbohyd. Res. 139 (1985) 105-113.
(e) P. Fugedi, P. J. Garegg, Carbohyd. Res. 149 (1986) C9-C12.
(f) V. Pozsgayand, J. Jennings, J. Org. Chem. 52 (1987) 4635-4637.
(g) T. Ziegler, E. Ekhardt, G. Herold, Tetrahedron Lett. 33 (1992) 4413-
4416.
The stereochemistry of 1 could not be determined accurately
as the compound exist as an anomeric mixture when both the α
and β form of the C1 hydroxy are present. As a result, 2D
experiments such as COSY, HSQC and HMBC (supporting
information) could only be performed on the anomerically-pure
intermediate 28 where the C1 hydroxy remained benzylated. In
9
C. L. Pereira, A. Geissner, C. Anish, P. H. Seeberger, Angew. Chem. Int.
Ed. Engl., 2015, 54, 10016-10019.
Corresponding author. e-mail: shwu@gate.sinica.edu.tw (S H Wu); email: chmlamyl@nus.edu.sg (Y Lam)