Assembling and Unraveling Carbohydrates Structures Conformational analysis of synthesized branched oligosaccharides

Sammanfattning: Advances in the elaboration of vaccines and enzyme inhibitors rely on acquiring more knowledge about protein-carbohydrate binding events. Furthermore, the relationships between biological function and the three-dimensional properties of large glycans can be studied by focusing on the structural components they contained, namely, by scaling down the system under analysis. Chemical methods are useful assets as they allow the isolation and determination of epitopes; these small and recognizable fragments that lead to very specific interactions. In this thesis, biologically relevant saccharides were obtained using recently developed concepts in carbohydrate synthesis and NMR spectroscopy was used to unravel their conformational preferences.In paper I, the convergent synthesis of the tetrasaccharide found in the natural product solaradixine is described. Reactivity enhanced disaccharide glycosyl donors were coupled to a disaccharide acceptor in a 2 + 2 fashion. The computer program CASPER was subsequently used to verify the synthesized structure.The conformation arming concept employed in paper I was further investigated in paper II. An NMR-based methodology enabled the determination of the ring conformations of a set of donors. Subsequently, glycosylation reactions were performed and yields were correlated to donors ring shapes. Perturbations in the rings shape caused by bulky silyl ether protective groups were sufficient to boost the potency of several donors. As a matter of fact, complex branched oligosaccharides could be obtained in good to excellent yields.In paper III, NMR spectroscopy observables were measured to elucidate the ring shape, the mutual orientation of the rings across the glycosidic bond and the positions of the side chains of 5 trisaccharides found in larger structures. With the aid of molecular dynamics simulations, their overall conformational propensities were revealed.Finally, the software CASPER prediction skills were improved by adding, inter alia, NMR information of synthesized mono- and disaccharides to its database. Unassigned chemical shifts from polysaccharides served as input to challenge its ability to solve large carbohydrate structures.