Development and application of site mapping methods for the design of glycosaminoglycans

Glycosaminoglycans (GAGs) are complex polysaccharides involved in a wide range of biological signaling events, as well as being important as biological structural materials. Despite the ubiquity and importance of GAG-protein interactions in biological systems and potentially as therapeutic targets, detailed structures of such interactions are sparse in availability. Computational methods can provide detailed structural knowledge of these interactions; however, they should be evaluated against suitable test systems prior to their widespread use. In this study, we have investigated the application of automated molecular docking and interaction mapping techniques to characterizing GAG-protein interactions. A series of high-resolution X-ray crystal structures of GAGs in complex with proteins was used to evaluate the approaches. Accurately scoring the pose fitting best with the crystal structure was a challenge for all docking programs evaluated. The site mapping technique offered excellent prediction of the key residues involved in ligand recognition, comparable to the best pose and improved over the top-ranked pose. A design protocol incorporating site-and ligand-based mapping techniques was developed and applied to identify GAGs capable of binding to acidic fibroblast growth factor (aFGF). The protocol was able to identify ligands known to bind to aFGF and accurately able to predict the binding modes of those ligands when using a known ligand-binding conformation of the protein. This study demonstrates the value of mapping-based techniques in identifying specific GAG epitopes recognized by proteins and for GAG-based drug design.