ABSTRACT
Experimental .......................................................................................................... 218 General Methods ............................................................................................... 218 Ethyl (2,3,4,6-tetra-O-acetyl-β-d-galactopyranosyl)-(1→3)-4,6-Obenzylidene-2-deoxy-1-thio-2-trichloroacetamido-β-d-glucopyranoside (3) ... 219
References .............................................................................................................. 221
The 2-deoxy-2-trichloroacetamido group is a powerful stereocontrolling auxiliary in the synthesis of 1,2-trans-linked 2-aminosugar-containing oligosaccharides.1,2 Therefore, the title thioglycoside is a glycosyl donor useful for introduction of the β-d-Gal-(1→3)-d-GlcNAc sequence into many target glycoconjugates important in the life sciences. The compound has been originally prepared by Sherman and coworkers2 within their study of glycosylation with N-trichloroacetyl-protected D-glucosamine derivatives. Side reactions and formation of byproduct were observed, including but not limited to cleavage of the benzylidene group. The highest yield of the desired disaccharide was 60%, when 2,3,4,6-tetra-O-acetyl-αd-galactopyranosyl bromide (acetobromogalactose, 2)3 was condensed with ethyl 4,6-O-benzylidene-2-deoxy-1-thio-2-trichloroacetamido-β-d-glucopyranoside (1) in CH2Cl2 in the presence of molecular sieves and sym-collidine (0.5 equiv. relative to 2). We recently reported that higher yield of the compound (90%)4 could be obtained by using a larger relative amount of base during the glycosylation step. However, when the same protocol was later repeated in our own laboratory using different batches of molecular sieves, we observed that the yield of 3 was inconsistent. We hypothesized that the basicity of molecular sieves from different batches/suppliers varied, which made it virtually impossible to rationalize the optimum amount of base to be used. To ensure isomerization of the intermediate orthoester,5,6 the reaction had to be conducted at base-decient conditions, while the conditions had to be mild to keep the acid-labile benzylidene group intact. The revised protocol (Scheme 25.1) presented here, which we have performed successfully many times, also on multi-gram scale, has consistently afforded disaccharide 3 in ~90% yield. It employs 1,1′,3,3′-tetramethylurea as base5 and does not require use of molecular sieves. Disaccharide 3 has now been obtained in crystalline form for the rst time.
