Due to their overfunctionality the direct a) regioselective esterification or b) stereoselective glycosidation of native, i.e. underivatised sugars (Fig. 1.1. Enzymic derivatisations of D-glucose) poses a particularly challenging problem to synthetic organic chemists. In most cases the use of suitable protected groups and/or multistep synthetic sequences - all methods of low atom efficiency - provided in the past reasonably practical solutions.
Hydrolytic enzymes such as esterhydrolases (lipases) or glycosidases are known for their capability of such derivatisations with high regio- and stereoselectivity, respectively. However, successful applications of these transformations were largely limited by the
a) low solubility of native sugars in non-polar solvents, where lipasesdisplay their highest esterification activity
b) low activity (or inactivity) of glycosidases in organic solvents.
In the past the problem was solved partially by using modified sugars like alkyl glycosides and protected sugars, solvents like DMF or pyridine or solubilising agents like phenyl boronic acid. In this work we provide satisfactory solutions for both problems using native sugars throughout in non-toxic solvents. Inspired by the discovery of Prasad and Sørensen  that 2-deoxy-D-ribose and D-ribose can be acylated by Novozym SP 435™ in THF, we were able to demonstrate that also hexoses – which are practically insoluble in such organic solvents – can be esterified with high regioselectivity and excellent yields.
Similarly, direct glycosylations via reverse hydrolysis of native, unprotected D-glucose were achieved by compromising the low solubility of these materials in organic solvents with the need of glucosidases for a microaqueous environment for activity. This way we were able to produce b-glycosides in stoichiometric reactions with up to 50 % yield. Yields of this magnitude were previously only achieved with excess of one reaction partner and the use of activated glycosides.