Abstract
Despite their fundamental biological relevance, structure-property relationships in N-glycans are fundamentally lacking, and their highly multidimensional compositional and conformational phase-spaces remain largely unexplored. The torsional flexibility of the glycosidic linkages and the ring dynamics result in wide, rugged free-energy landscapes that are difficult to sample in molecular dynamics simulations. We show that a novel enhanced-sampling scheme combining replica-exchange with solute and collective-variable tempering, enabling transitions over all relevant energy barriers, delivers converged distributions of solvated N-glycan conformers. Several dimensionality-reduction algorithms are compared and employed to generate conformational free-energy maps in two-dimensions. Together with an originally developed conformation-based nomenclature scheme that uniquely identify glycan conformers, our modelling procedure is applied to reveal the effect of chemical substitutions on the conformational ensemble of selected high-mannose-type and complex glycans. Moreover, the structure-prediction capabilities of two commonly used glycan force fields are assessed via the theoretical prediction of experimentally available NMR J-coupling constants. The results confirm the key role of especially ω and ψ torsion angles in discriminating between different conformational states, and suggest an intriguing correlation between the torsional and ring-puckering degrees of freedom that may be biologically relevant.
Competing Interest Statement
The authors have declared no competing interest.
Footnotes
↵* E-mail: colombi{at}hmi.uni-bremen.de
Graphical abstract added; deleted a sentence in the result section about a putative (not confirmed) correlation between peaks in the Sketch-Map pairwise distance distribution and the number of omega angles in a glycan; added a sentence in the result section commenting on details of the free-energy maps in the space of the puckering coordinates.