We introduced the term "glycopolymer" to identify water-soluble polymers onto which carbohydrate haptens are covalently appended . In this respect, glycopoly-mers should differ from pseudopolysaccharides, which refer to chemically modified polysaccharides and to insoluble materials used in affinity chromatography. Curiously, we initially made the first sialylated glycopolymers  to screen anti-sialic acid antibodies obtained from a sialylated neoglycoproteins [18b,73-75]. The original aim was to generate an antigen deprived of cross-reactive hapten (Neu5Ac), which inevitably would have been produced from other protein carriers. It was then quickly realized that these copolymers offered great potential as inhibitors in cell adhesion processes. Patents for a cancer diagnostic kit consisting of a sialylated protein (vaccine) and an ELISA screening antigen (polymer) were filed several years ago . While this activity was ongoing, we reported the inhibitory potential of these novel sialylated copolymers in influenza flu virus inhibition of hemagglutina-tion. These early observations were then similarly made by several other groups [77,78].
Earlier reviews [10,11,58,79] had described numerous polymerization methods that have been used in synthesizing a wide range of glycopolymers. Amazingly, even though most methodologies can afford better organized copolymers than the one initially used (i.e., by random acrylamide copolymerization or modification), very few alternative methods have been exploited for sialosides. The syntheses and applications of glycopolymers are now covered in several reviews and book chapters [79-82], and only recent developments are highlighted.
Reducing sugars can be reductively aminated with ammonia or other amines. The resulting amine derivatives can then be transformed into ^-acrylamide monomers useful in copolymerization strategy. Unfortunately, acrylamide copolymeriza-tion affords polymers that may vary greatly in their batch-to-batch molecular weight distributions. An improved protocol consists of synthesizing polyacrylates (43) having active ester functionality NHS  or 4-NO2Ph  (Scheme 8). After aminolysis or hydrolysis, the molecular weight of the resulting polyacrylamides or polyacrylic acids can be compared against commercially available polymer standards used in HPLC. This allows the synthesis of reproducible lots of a given copolymer. Once established, the starting polyacrylates 43 can be treated with various amounts of any amine-containing sugars. Quenching the residual reactive esters with amine 1 (R1NH2) followed by amine 2 (R2NH2), or simply with ammonia or water, afforded copolymers having desired biophysical properties [77,83]. The second and third amines may include probes (e.g., biotin, fluorescamine), lipid groups, other sugars, and peptides. The strategy has also been elegantly used to generate sialopolymer libraries . Sialic acid copolyacrylamides such as 44 were obtained with O/S-aryl
spacers [32,49], C-glycosides , and other related spacers . GM3-type copolymer 45 was obtained using the foregoing procedure .
Thioaryl sialoside comonomer 46  is easier than C-glycoside  to produce in a stereocontrolled manner. It is also resistant to sialidases that are simultaneously present on flu virions. Interestingly, it could be directly incorporated onto both poly-l-lysine and proteins by 1,4-conjugate additions (Michael addition) at pH 10 to provide biocompatible random copolymer 47 (Scheme 9) . By analogy, reductively aminated a-(2,8)-polysialic acid (colominic acid) 48 can be N-acryloy-lated into compounds 49, which undergo 1,4-conjugate addition onto poly-l-lysine to provide copolymer 50, isolated as its biotin conjugate . In 1998 Wong et al.  reported an analogous strategy whereby a lysoganglioside derivative was ami-dated to poly-l-glutamic acid (DP 540) together with the fluorescent tag 4,4-difluo-5,7-dimethyl-4-bora-3a,4a-diaza-s-indacene-3-propionyl group (BODIPY) to give 51. The copolymer showed picomolar inhibition of H1N1 influenza hemagglutinin that corresponded to an improved binding of 1000-fold over gangliosides GM3 or lyso-GM3 and 105-fold relative to the monosaccharide sialyl lactose. Lipidlike copolymer 51 is thus as active as a polymerized liposome published in 1993 .
All the strategies discussed so far included sialylated copolymers having randomly distributed sialic acid residues. As these approaches failed to generate regular copolymers onto which the hapten distributions would be interspaced at constant distances, it was deemed necessary to prepare copolymers by addition polymerization . Chapter 8 by Mann and Kiessling offers an alternative strategy for the use of ring-opening metathesis polymerization (ROMP) of norbornene derivatives to generate analogous polymers. The novel strategy described herein depends on a sialic acid monomer (53) having two amine groups that can be copolymerized by a reiterative addition process onto a bisisocyanate (Scheme 10) (R. Roy, Y. Makimura, unpublished data). Thus, 4-nitrophenyl sialoside 9a was reduced and treated with
bromoacetyl chloride. After N-alkylation with Cbz-protected 3,3'-iminobis-(propylamine) , intermediate 52 was obtained in 75% yield. Hydrogenolysis provided diamine 53 quantitatively. Finally, addition polymerization of 53 with either 1,4-butanediamine or 1,6-hexanediamine bisisocyanates and ester hydrolysis gave low molecular weight copolymers 54 and 55 in ~50% yield, with a degree of polymerization of ~15.
It is worth mentioning that all known glycopolymers have shown strong inhibitory properties when used in conjunction with carbohydrate binding proteins. The exact origins of the increased associative forces conferred on glycopolymers are not clearly understood. Aside from the individual binding site's affinity constants (^D), which obviously remained unchanged, entropic penalties that occur upon binding are minimized with multivalent ligands. The ligands' local high concentrations are certainly also affecting association/dissociation's kinetics (kon/koff). Some external fac-
Was this article helpful?