Limitations in the description of conformational preferences of C-disaccharides: The (1 → 3)-C-mannobiose case

Abstract Conformational preferences of two C -glycosyl analogues of Man p -(1 → 3)-Man p , were studied using a combined method of theoretical and experimental chemistry. Molecular dynamics was utilized to provide conformational behavior along C -glycosidic bonds of methyl 3-deoxy-3- C -[(α- d -mannopyranosyl)methyl]-α- d - and l -mannopyranosides. The OPLS2005 and Glycam06 force fields were used. Simulations were performed with explicit water (TIP3P) and methanol. Results were compared with a complete conformational scan at the MM4 level with the dielectric constant corresponding to methanol. In order to verify predicted conformational preferences, vicinal 3 J HH NMR coupling constants were calculated by the Karplus equation on simulated potential energy surfaces (PES). A set of new parameters for the Karplus equation was also designed. Predicted 3 J HH were compared with experimental data. We also used reverse methodology, in which the 3 J HH coupling constants were calculated at the DFT level for each family of ( ϕ, ψ )-conformers separately and then experimental values were decomposed onto calculated 3 J HH couplings in order to obtain experimentally derived populations of conformers. As an alternative method of evaluation of preferred conformers, analysis of sensitive 13 C chemical shifts was introduced. We were able to thoroughly discuss several fundamental issues in predictions of preferred conformers of C -saccharides, such as the solvent effect, reliability of the force field, character of empirical Karplus equation or applicability of NMR parameters in predictions of conformational preferences in general.