Surprising bacterial nucleotidyltransferase selectivity in the conversion of carbaglucose-1-phosphate

The drive to understand the molecular determinants of carbohydrate binding as well as the search for more chemically and biochemically stable sugar derivatives and carbohydrate-based therapeutics has led to the synthesis of a variety of analogues that replace the glycosidic oxygen with sulfur or carbon. In contrast, the effect of substitution of the ring oxygen on the conformations and enzymatic tolerance of sugars has been largely neglected, in part because of the difficulty in obtaining these analogues. Herein we report the first synthesis of the carbocyclic version of the most common naturally occurring sugar-1-phosphate, glucose-1-phosphate, and its evaluation with bacterial and eukaryotic sugar nucleotidyltransferases. In contrast to results with the eukaryotic enzyme, the carbaglucose-1-phosphate serves as a substrate for the bacterial enzyme to provide the carbocyclic uridinediphosphoglucose. This result demonstrates the first chemoenzymatic strategy to this class of glycosyltransferase inhibitors and stable activated sugar mimics for cocrystallization with glycosyltransferases and their glycosyl acceptors. This difference in turnover between enzymes also suggests the possibility of using sugar nucleotidyltransferases in vivo to convert prodrug forms of glycosyltransferase inhibitors. In addition, we report several microwave-assisted reactions, including a five minute Ferrier rearrangement with palladium, that accelerate the synthesis of carbocyclic sugars for further studies.