Development of new glycosylation methodologies for the synthesis of archaeal-derived glycolipid adjuvants

To commercialize the production of glycolipid adjuvants, their synthesis needs to be both robust and inexpensive. Herein we describe a semi-synthetic approach where the lipid acceptor is derived from the biomass of the archaeon Halobacterium salinarum, and the glycosyl donors are chemically synthesized. This work presents some preliminary results using the promoter system N-iodosuccinimide (NIS) and a stable 0.25 M solution of boron trifluoride-trifluoroethanol (BF(3)center dot TFE(2)) in dichloromethane. This promoter system allows for the use of peracetyl alkyl(aryl)thioglycosides that are available in high yield from inexpensive disaccharide starting materials by promoting clean glycosylation reactions from which pure product can be easily isolated. Conventional glycosylation using NIS-silver trifluoromethanesulfonate (AgOTf) leads to extensive acetyl transfer to the archaeol acceptor and numerous byproducts that make purification complicated. As part of preliminary structure-adjuvant activity studies, we describe the one-pot synthesis of a gentiobiose beta-Glcp-(1 -> 6)-Glcp-SEt donor with an O-2 benzoyl group, which can be used to prepare a disaccharide attached to archaeol in 85% overall yield, and the related glycolipid trisaccharide beta-Glcp-(1 -> 6)-beta-Glcp-(1 -> 6)-beta-Glcp-(1 -> O)-archaeol. The synthesis of the isomeric beta-Glcp-(1 -> 6)-alpha-Glcp-(1 -> O)-archaeol featuring a >10:1 alpha/beta alpha-selective glycosylation using the promoter system N-phenylselenylphthalimide-trifluoromethanesulfonic acid (TfOH) is also presented, along with the adjuvant properties of the corresponding archaeosomes (liposomes comprised entirely of combinations of isoprenoid archaeal-like lipids). These new vaccine formulations extend previous observations that glycolipids are integral to the activation of MHC type I pathways via CD8(+) antigen-specific T-cells. The beta-Glcp-(1 -> 6)-beta-Glcp-(1 -> 6)-beta-Glcp-(1 -> O)-archaeol trisaccharide is shown to be more active than the Glcp-(1 -> 6)-beta-Glcp-(1 -> O)-archaeol disaccharide. Crown Copyright (C) 2009 Published by Elsevier Ltd. All rights reserved.