Stereoselective Glycosylations

Carbohydrates are the most abundant natural organic compounds on Earth and their importance can be understood by the fact that over 20% of bacterial natural products are glycosylated and that the sugar core is important for their biological activity. The configuration at the anomeric centre is crucial for the biological activity of oligosaccharides and glycoconjugates, and even a change in a single linkage can affect the activity of such molecules. Hence, controlling the anomeric selectivity in chemical glycosylation is arguably the most important and challenging aim in the field of carbohydrate chemistry. It is also important for chemists to have access to building blocks that would be able to furnish either a- or ß-linkages in a regio- and stereoselective manner. ß-Selective Mannosylations and Rhamnosylations from Glycosyl Hemiacetals: Herein, a highly selective synthesis of ß-mannosides and ß-rhamnosides from glycosyl hemi-acetals is reported, following a one-pot chlorination, iodination, glycosylation sequence employing cheap oxalyl chloride, phosphine oxide and LiI. 14 ß-Rhamnosides and two ß-mannosides were synthesised, mechanistic investigations were carried out and a 1-mmol scale reaction for ß-mannosylation was performed. The present protocol works excellently with a wide range of glycosyl acceptors and with armed glycosyl donors, without requiring conformationally-restricted donors or directing groups. Stereoselective Galactosidations: Galactose is present as a ‘capping’ motif in the mammalian glycome with both the a- and ß-linkages. However, no universal building block has been found for either an a- or ß-linkage for the solid-phase oligosaccharide automated synthesis. Herein, I report efforts towards the synthesis of the thiogalactosidic donor precursors which bear an azide group at positions 4-OH and 6-OH, that can undergo a ‘click’ switch between being a donor for a- and ß-linkages. Thus, a pyridine directing group was installed via a [3+2] Huisgen’s cycloaddition in order to make use of Demchenko’s H-bond-mediated aglycone delivery concept. Furthermore, other transformations of the azide functional group, such as Staudinger ligation and ‘traceless’ Staudinger, were explored. Apart from the pyridine directing group, other aromatic moieties were tested in order to examine electronic effects on the glycosylation outcome. In total, 26 novel galactosyl donors were synthesised and examined in the glycosylation reaction leading to glycosides with varying levels of a- or ß-selectivity. Donors that give excellent a- and ß-selectivity were successfully identified and it is proposed that electronic and conformational effects play a role in the observed selectivities (e.g., the formation of ß-triflate under glycosylations conditions with galactosyl donors bearing EWG at position 4, is responsible for the excellent a-selectivity obtained in these reactions). Ferrier Rearangements: Herein, it is reported that 2-thiouracil can catalyse the Ferrier rearrangement reactions of glycals. The reaction has been optimised and benzyl protected glycals, such as glucal, gulal, allal and rhamnal gave the Ferrier product with satisfactory a-selectivity (a/ß 67:33 to a-only). Furthermore, making use of the bulkier TBS protecting group, the reaction of glycals with a glycosyl acceptor gave two easily separable glycosides, the Ferrier rearrangement product as minor and the desired disaccharide as major. I also discovered that when using Boc protecting groups on the glycal, the glycosylation reaction with sugar, non-sugar and amino-acid glycosyl acceptors led to the desired products after an intermolecular Ferrier rearrangement. Another interesting find was that when thiols were used as acceptors and peracetylated glucal as the donor, there was no need to use the 2-thiouracil catalyst in order to form the Ferrier rearranged product in nitromethane.