Synthesis of a Tri- and Tetrasaccharide Fragment Specific for the Shigella flexneri Serotype 5a O-Antigen. A Reinvestigation

Abstract

ABSTRACT Stereocontrolled, stepwise synthesis of methyl α-L-rhamnopyranosyl-(1→2)-[α-D-glucopyranosyl-(1→3)]-α-L-rhamnopyranoside (A(E)B, 1) and methyl 2-acetamido-2-deoxy-β-D-glucopyranosyl-(1→2)-α-L-rhamnopyranosyl-(1→2)-[α-D-glucopyranosyl-(1→3)]-α-L-rhamnopyranoside (DA(E)B, 2) is described; these constitute the methyl glycosides of fragments of the O-specific polysaccharide of Shigella flexneri serotype 5a. Two routes to trisaccharide 1 were considered. Route 1 involved the coupling of a precursor to residue A and a disaccharide EB, whereas route 2 was based on the condensation of a precursor to residue E and a disaccharide AB. Rather surprisingly, the latter afforded the β-anomer of 1, namely methyl α-L-rhamnopyranosyl-(1→2)-[β-D-glucopyranosyl-(1→3)]-α-L-rhamnopyranoside as the major product. Route 1 was preferred. Overall, several observations made during this study suggested that, for the construction of higher fragments, a suitable precursor to rhamnose A would require protecting groups of low bulkiness at position 3 and 4. Therefore, the 2-O-acetyl-3,4-di-O-allyl-α-L-rhamnopyranosyl trichloroacetimidate (35) was the precursor of choice to residue A in the synthesis of the tetrasaccharide 2. The condensation product of 35 and methyl 2,3,4,6-tetra-O-benzyl-α-D-glucopyranosyl-4-O-benzyl-α-L-rhamnopyranoside was selectively deacylated and condensed to 2-trichloroacetamido-3,4,6-tri-O-acetyl-2-deoxy-α-D-glucopyranosyl trichloroacetimidate to afford the corresponding fully protected tetrasaccharide 45. Controlled stepwise deprotection of the latter proceeded smoothly to afford the target 2. It should be emphasised that the preparation of 45 was not straightforward, several donors and coupling conditions that were tested resulted only in the complete recovery of the acceptor. Distortion of several signals in the 13C NMR spectra of the fully or partially protected tetrasaccharide intermediates suggested that steric hindrance, added to the known low reactivity of HO-2 of rhamnosyl acceptors, probably played a major role in the outcome of the glycosidation attempts.