ONE-POT SYNTHESIS OF THIO-GLYCOMIMETICS THROUGH RING OPENING REACTIONS

Glycomimetics are compounds able to mimic structural and functional aspects of the corresponding natural carbohydrates. The main goal of using these mimetics, e.g. as therapeutic agents, is the manipulation of the chemical information encoded by sugars, controlling and altering the flow of these information. Sugars, and their mimetics, are selectively recognised by certain proteins, called lectins. An example of these lectins is DC-SIGN. DC-SIGN (Dendritic Cell-Specific ICAM-3 Grabbing Non-integrin) is a tetrameric calcium dependent (C-type) lectin expressed by dermal dendritic cells (DCs) and it is known to play a crucial role in the immunological system. In particular, DC-SIGN is the primary target of Human Immunodeficiency Virus type I (HIV-1) that hijacks DCs to disseminate the infection in the human body. With the aim to develop DC-SIGN antagonists able to mimic the main natural ligand, the Bernardi’s group has developed a series of mono- and multivalent ligands that all contain a mannose unit (whose hydroxyl groups in position 3 and 4 are involved in the interaction with the calcium ion of the Carbohydrate Recognition Domain (CRD) of DC-SIGN) connected with a conformationally locked cyclohexandiol moiety (the aglycon part, that is known to exploit additional interaction with residues present in the binding site). In the course of my PhD, I have developed a straightforward synthesis for a new class of glycomimetic DC-SIGN ligands more metabolically stable, thanks to the presence of a sulfur atom in the interglycosidic position rather than the oxygen atom of the previous library of compounds. The key step of the preparation of these thio-glycomimetics is a one-pot ring opening reaction of an enantiomerically pure epoxide by an S-acetyl-α-mannose that affords the desired pseudo-thiodisaccharide as a single isomer in 77% yield after 1 hour at 60°C under micro-wave irradiation. To further improve the process, the same reaction was also performed into a flow micro-reactor at different flow rates, temperatures and residence times. Although at certain conditions a promising conversion of 69% was observed, the results were not competitive with the ones already achieved by using a MW system, which allowed also the scale up of the process. Computational studies supported by NMR analysis showed that the new thio-glycomimetic shares a very similar conformational behaviour as the corresponding O-linked pseudo-disaccharide. Furthermore, SPR inhibition assay revealed that they share also the same affinity towards DC-SIGN (IC50 value of 0.8 mM). However, a stability test, performed against an α-mannosidase enzyme, proved that the presence of the sulfur in the interglycosidic position makes the structure significantly more stable than that of the same molecule bearing instead an oxygen in that position. The low sub-millimolar activity of monovalent ligands is characteristic of oligosaccharide-protein interactions and is overcome both in Nature and in the development of biological active glycomimetics by exploiting multivalent interactions. With this aim, the free hydroxyl group of the cyclohexane moiety of the pseudo-thiodisaccharide was functionalized with an azide-terminated linker through an esterification reaction. Afterwards, this compound was employed in the construction of multivalent systems, by exploiting click chemistry between the terminal azide of the ligand and the free terminal alkyne functions of a counterpart (work carried out during my visit in Sevilla, at the “Istituto de Investigaciones Quimicas-IIQ”, under the supervision of Dr. Javier Rojo). In particular, I have synthetized two different types of multivalent scaffolds: dendrimers and glyconanogels. These structures are still under biological investigation. Since the ester linkage of the azido-linker, introduced to allow the construction of multivalents, remained a weak point of the molecule, we decided to further increase the stability of the thio-glycomimetic, developing a new synthetic strategy that involved the ring opening of an aziridine ring rather than an epoxide. After the optimization of the synthetic procedure of the starting free aziridine, it was functionalized with different acylating reagents obtaining a series of N-acylated aziridines which were finally opened in a one-pot manner again by the α-thioacetate. The biological activity of the final compounds was finally investigated through SPR inhibition assay and all of them showed IC50 values very close to each other’s but in particular very similar to the original O-linked pseudo-disaccharide. In conclusion, two new synthetic approaches were devised for the synthesis of thio-dimannoside mimics designed to be stable against enzymatic and chemical hydrolysis. The pseudo-anomeric position of the aglycon of these compounds could be functionalized with a nitrogen atom, which allows to thether the glycomimetic structure to linkers/scaffolds/peptidic chains using a non-labile amide linkage.