DECOYS FOR DISRUPTING CARBOHYDRATE-LECTIN INTERACTIONS

Carbohydrates are biomolecules used for presentation of biochemical signals on the cell surface. The interactions between carbohydrates and lectins are involved in numerous physiological and pathophysiological processes, therefore understanding how the information is encoded in the structures of sugars, how is it read out by carbohydrate binding proteins (lectins) and furthermore how to control or alter the flow of this information is of high importance. To approach this challenge various principles are used, taking advantage of carbohydrate chemistry, supramolecular chemistry and rational protein engineering to deliver test compounds that can be glycans, glycomimetics or lectin-based. The goal of this thesis was to develop both glycomimetics and lectin constructs able to interfere with carbohydrate-lectin interactions. In the first part (at the University of Milano), the focus was set on the synthesis of glycomimetic compounds. Considering, that the major disadvantages of utilizing carbohydrates in studies are their metabolic instability and difficult synthesis, we created a new class of enzymatically and hydrolytically stable glycomimetics with the general structure of N-linked-pseudo-thio-glycosides. To simplify their synthesis, we developed a new synthetic approach in which the formation of the thiol from glycosyl thioacetate and subsequent opening of the aziridine are combined in a one-pot reaction. We then investigated the scope of this reaction on various mono- and disaccharides. Surprisingly, thio-derivatives of β-glucose, β-galactose, β-fucose, β-lactose and β-N-acetylglucosamine underwent anomeric isomerization under the reaction conditions employed. On the other hand, tiho-derivatives of α-mannose, α-rhamnose and α-Neu5Ac were configurationally stable under the reaction conditions. Therefore, to unravel the mystery of the anomeric isomerization of sugar derivatives containing anomeric sulphur we conducted an in depth literature search on this topic and performed mechanistic studies to better understand the cause of anomerization observed during the one-pot aziridine opening reactions. β-Glucosyl thioacetate was selected as a model substrate for our studies. Through a series of experiments, we confirmed that the anomerization under the experimental conditions (Et2NH, DMF) occurs through the intermediate thiolate. Furthermore, we carefully investigated the tuning of the reaction conditions to improve the anomeric ratio to synthetically useful levels. The second part of the project (at Utrecht University) was a shift from carbohydrates towards lectins. We aimed towards making new flexible multisite lectin constructs. As a model, we chose cyanovirin-N, a lectin with broad antiviral activity, due to its ability to bind to high-mannose glycans on enveloped viruses. In the first step towards making new peptide constructs we aimed to identify the minimum sequence of this lectin still able to bind carbohydrates, to be the used for multisite presentation. We did so by shortening the sequence of cyanovirin-N to domain B that bears the high affinity carbohydrate binding site. We then compared preparation processes of this peptide, synthesis versus recombinant production from bacteria. Finally, secondary structure and glycan binding ability of the new peptide constructs were investigated.