Abstrakt

Computer simulations of carbohydrates pose a significant challenge due to their heterogeneous chemical structure, high conformational variability, inaccuracies in interaction potentials, and computational inefficiency, which scales particularly unfavorably with chain length compared to other biopolymers. In this work, we present a coarse-grained model designed for Monte Carlo simulations of single polysaccharide chains of realistic lengths (hundreds of monomers or more). The model specifically accounts for the structural features of natural polysaccharides, which are typically built from disaccharide building blocks, and for conformational variability arising from rotations around glycosidic linkages. By extrapolating local conformational properties, thermodynamically correct configuration ensembles are generated, and high computational efficiency is achieved through rapid decorrelation of independent degrees of freedom. As an illustration of the model’s capabilities, we present a set of simulations of keratan, a natural biopolysaccharide belonging to the glycosaminoglycan family. The results indicate that sulfation has a limited impact on chain conformation. Instead, chain stiffness is primarily influenced by syn-exo ↔ anti-ψ conformational transitions in β(1 → 4) glycosidic linkages. The persistence length varies between 6 and 8 nm depending on the degree of sulfation. Moreover, for chain lengths most commonly found in nature, a linear relationship is observed between the number of monomeric units and both the radius of gyration and the end-to-end distance. Finally, we present and discuss the strategy for accounting for ionic-strength-driven alterations of the charged polysaccharide chain, which, when applied to the current case, demonstrate that, under physiological conditions, the long-range electrostatic interactions within the keratan chain can be neglected.