A molecular dynamics-based algorithm for evaluating the glycosaminoglycan mimicking potential of synthetic, homogenous, sulfated small molecules

Abstract

Glycosaminoglycans (GAGs) are key natural biopolymers that exhibit a range of biological functions including growth and differentiation. Despite this multiplicity of function, natural GAG sequences have not yielded drugs because of problems of heterogeneity and synthesis. Recently, several homogenous non-saccharide glycosaminoglycan mimetics (NSGMs) have been reported as agents displaying major therapeutic promise. Yet, it remains unclear whether sulfated NSGMs structurally mimic sulfated GAGs. To address this, we developed a three-step molecular dynamics (MD)-based algorithm to compare sulfated NSGMs with GAGs. In the first step of this algorithm, parameters related to the range of conformations sampled by the two highly sulfated molecules as free entities in water were compared. The second step compared identity of binding site geometries and the final step evaluated comparable dynamics and interactions in the protein-bound state. Using a test case of interactions with fibroblast growth factor-related proteins, we show that this three-step algorithm effectively predicts the GAG structure mimicking property of NSGMs. Specifically, we show that two unique dimeric NSGMs mimic hexameric GAG sequences in the protein-bound state. In contrast, closely related monomeric and trimeric NSGMs do not mimic GAG in either the free or bound states. These results correspond well with the functional properties of NSGMs. The results show for the first time that appropriately designed sulfated NSGMs can be good structural mimetics of GAGs and the incorporation of a MD-based strategy at the NSGM library screening stage can identify promising mimetics of targeted GAG sequences.