Glycorandomization

Chemoenzymatic glycorandomization was inspired by the early pathway engineering work of Hutchinson and coworkers that suggested natural product glycosyltransferases were capable of utilizing non-native sugar nucleotide donors.[6] The initial platform for chemoenzymatic glycorandomization was based upon a set of two highly permissive sugar activation enzymes (a sugar anomeric kinase and sugar-1-phosphate nucleotidyltransferase) to afford sugar nucleotide libraries as donors for these promiscuous glycosyltransferases where the permissivity of the corresponding sugar kinase[7] and nucleotidyltransferase[8][9] was expanded by enzyme engineering and directed evolution. The first application of this three enzyme (kinase, nucleotidyltransferase and glycosyltransferase) strategy enabled the product of a set of >30 differentially glycosylated vancomycins, some members of which were further diversified chemoselectively by virtue of the installation of sugars bearing chemoselective handles.[10][11][12] This enzymatic platform has been further advanced through glycosyltransferase evolution[13] and capitalizing upon the discovery of the reversibility of glycosyltransferase-catalyzed reactions first discovered in the context of calicheamicin biosynthesis.[14][15]

Neoglycorandomization is a chemoselective glycodiversification method inspired by the alkoxyamine-based ‘neoglycosylation’ reaction first described Peri and Dumy.[16] This reaction proceeds via an oxy-iminium intermediate to ultimately provide the more thermodynamically-favored closed ring neoglycoside. The neoglycosylation reaction is compatible with a wide range of saccharide and aglycon functionality where neoglycoside anomeric stereospecificity is a thermodynamically-driven. Importantly, structural and functional studies reveal neoglycosides to serve as good mimics of their O-glycosidic comparators. The first neoglycorandomization proof of concept focused upon digitoxin where the rapid generation and cancer cell line cytotoxicity screening of 78 digitoxigenin neoglycosides revealed unique analogs with improved anticancer activity and reduced potential for cardiotoxicity.[17] This platform has since been automated and used as an effective medicinal chemistry tool to modulate the properties of a range of natural products and pharmaceutical drugs.[18]

Both chemoenzymatic glycorandomization and neoglycorandomization use free reducing sugars and unprotected aglycons and are thereby a notable advance over classical glycosylation methods. A notable advantage of the enzymatic approach is the use of the corresponding genes encoding for the permissive kinases, nucleotidyltransferases and/or glycosyltransferases for in vivo synthetic biology applications to afford in vivo glycorandomization.[19] However, it is important to note the enzymatic platform is dependent upon the permissivity of the enzymes employed. In contrast, the main hurdle to chemoselective neoglycorandomization is installation of the alkoxylamine handle. Unlike the enzymatic approach, the anomeric stereoselectivity of the chemoselective method depends upon the reducing sugar used and can, in some cases, lead to anomeric mixtures.