Potential energy hypersurfaces of nucleotide sugars: Ab initio calculations, force-field parametrization, and exploration of the flexibility

Glycosyl esters of nucleoside di- or monophosphates, generally referred to as "sugar nucleotides", serve as a sugar donor during the biosynthesis of o ligo- and polysaccharides. Therefore, they are of primary importance in car bohydrate metabolism in the living world. Not only the molecules themselves but especially their complexes with proteins are of interest in structural glycobiology. For computational studies on these molecules, it is necessar y to have access to empirical methods with appropriate force field parametr ization. In this work, we propose a set of parameters, developed using ab i nitio calculations with the 6-31G* basis set at the SCF level on model comp ounds, for the commonly used AMBER force field. By implementation of the ne w parameter set together with the CICADA conformational search program, we have obtained a semiquantitative description of conformational space, showi ng that nucleotide sugars can adopt several conformational families. The ma jority of them exhibit a "folded" rather than an "extended" geometry due to frequent intramolecular hydrogen bonds and "stacking" interactions between the base and the six-membered sugar ring. For the sake of comparison, two molecular dynamics simulations were run in an explicit water environment. T he first simulation (3 ns) started with the semi-extended X-ray geometry an d displayed major variations of all torsional angles, allowing for the visi t of three conformational families. The second simulation (5 ns) started wi th the folded global minimum from the CICADA search. After about 3ns, a tra nsition for the ribose pucker yielded to the visit of a more extended confo rmational family. Experimental results show that in crystalline state, or i n protein/carbohydrate complexes, extended conformations which are stabiliz ed by the interaction with surrounding molecules or with the protein surfac e are more frequent.