Glycosylation reactions with alpha- and beta- gluco and galactopyranosyl acceptors: Experimental and theoretical study

DEL VIGO, E. A.; MARINO, C.; STORTZ, C. A.

New Orleans

Simposio; 28º International Carbohydrates Symposium,; 2016

ICO

The knowledge and control of the glycosylation regioselectivity allow the development of simpler reaction sequences for the synthesis of oligosaccharides and glycoconjugates, minimizing the use of protecting groups. Regioselectivity responds to multiple steric and electronic factors operating in both the donor and acceptor and are characteristic to each particular sugar.In this study we aimed to compare the differential reactivity of HO-3 and HO-4 of glycosyl acceptors 1. Such acceptors were synthesized by a sequence of simple reactions from α- and β-anomers of D-Galp and D-Glcp. In order to compare the regioselectivity of different methods, donors 2-4 were synthesized and the corresponding glycosylations were performed. The product ratios in each case were determined by 1H NMR of the crude reactions and by column chromatography. The results were compared with molecular modeling studies conducted following the methodology developed by Colombo et al.1 A conformational study using simplified analogues of 1-4 was performed. Methyl and acetyl groups at O-2 and O-6 were considered instead of benzyl or benzoyl groups, respectively. Then, starting from the lowest energy conformers, stability of cationic intermediates in O-3 and O-4 were studied using a methyl cation as a simplified glycosyl donor. Modeling studies showed that the cationic intermediary charged in O-4 is the most stable for both benzylated Galp and Glcp acceptors, so, in opposition with previous results with D-GlcNAc,1 disaccharides with the 1→4 glycosidic linkage should be expected to be the major products. On the other hand, experimentally such acceptors preferably reacted via the O-3, in all the cases. Contrasting with previous studies the experimental and theoretical results were not coincident, suggesting that some factor, probably a hydrogen bonding, is overrated in the calculations against the actual experimental effect.