Artificial glycoconjugate polymers (glycopolymers) and other multivalent carbohydrate ligands constitute a new class of biomimetic supramolecules. [1] They have shown many biological applications in, for example, cultivation, tumor diagnosis and detection, and the trapping of viruses and bacterial toxins. [2] Their utility is ascribed mainly to their strong and species-specific interactions with the receptor proteins as a result of multivalent binding and/or carbohydrate cluster effects. Although most of the glycopolymers so far prepared and applied were made up of simple mono-and disaccharides, more intense interest has been directed to glycopolymers carrying cell-surface oligosaccharides such as sialyl Lewis X (sLe X ) [3] and globosyl oligosaccharides [4] that have potentially greater biological significance. From a practical viewpoint, however, the synthesis and application of these glycoconjugates are seriously restricted by the difficulty in preparing target oligosaccharides in sufficient amounts prior to their incorporation into the multivalent models.In the course of our synthetic study of glycopolymers carrying the mimics of cell-surface oligosaccharides, [5] we found recently that an acrylamide copolymer carrying a-l-fucopyranoside and 3-sulfo-b-d-galactopyranoside as a side chain shows strong activity in blocking L-selectin/sLe X tetrasaccharide adhesion. [6] As judged from the observation that none of the acrylamide copolymers carrying only one of the two glycosides showed a notable activity, the observed activity was ascribed to the cooperative binding of the a-lfucoside and 3-sulfo-b-d-galactoside to L-selectin. This finding indicated the potential utility of the module approach and prompted us to generalize it as a ™carbohydrate module method∫.For the facile understanding of our ™carbohydrate module method∫, let us assume a model consisting of the binding of a branching pentasaccharide with the receptor protein (Figure 1). In this model the two glycoside residues (A1 and A2) at the nonreducing terminal provide key binding interactions with the receptor protein at the r(A1) and r(A2) binding sites. Conventional mimic syntheses may target a branching trisaccharide carrying A1, A2, and B residues, thus developing the multivalent model (polymer-(A1 þ A2 þ B)) in Figure 1 b. Our approach targets a copolymer-(A1/A2) species carrying the key interactive sugars (Figure 1 c) to circumvent the difficulty in preparing the branching saccharide. The copolymer-(A1/A2) species is assumed to have a certain probability of occupying both of the binding sites and, thus, is expected to show a higher binding activity than the polymers carrying only one of the key sugars (poly-(A1) and poly-(A2)). This situation means that copolymerization of two key interactive sugars (A1 and A2 in the present case) provides a facile way to mimic the biologically active structures of oligosaccharides. The glycopolymers thus derived may show potent biological activity in blocking the binding of the oligosaccharide to receptor proteins...