Carbohydrate-protein interactions mediate a wide range of biological processes, but are challenging to study at the molecular structural level. The ability of experimental methods to obtain high quality structural information is affected by the high flexibility of carbohydrates, as well as the potential for carbohydrates to exhibit multiple binding modes with their targets. While computational methods are a valuable alternative to experimental methods to obtain structural knowledge, they are affected by the same issues. Additionally, methods such as automated molecular docking, the key computational technique for identifying structures of ligand binding to proteins, are affected by the ability of carbohydrates to form many interactions with their protein targets; often, incorrect solutions can appear to give a better fit than the correct solutions.To address the challenges faced in the structural determination of carbohydrate-protein complexes, computational mapping approaches have been developed. These methods utilize the interactions between an ensemble of carbohydrate binding modes obtained from docking and the target protein and have proven useful in characterizing a variety of carbohydrate-protein recognition scenarios.1,2These methods have now been extended to study glycosaminoglycan-protein (GAG) recognition, an area of emerging importance for the development of anticancer and antiflammatory agents. The site mapping method is demonstrated to provide improved performance for predicting the key protein residues involved in GAG recognition over the top ranked pose from molecular docking. Furthermore, incorporating both site and epitope mapping techniques facilitated virtual screening of GAG disaccharides against acidic fibroblast growth factor, specifically identifying both the known binding ligands and their correct binding modes. This study suggests the potential broad utility of the mapping techniques in virtual glycan screening.1Agostino et al., Glycobiology, 2010  , 20, 724-7352Agostino et al., J Mol Graph Model, 2013  , 40, 80-90.

Computational mapping approaches for understanding glycosaminoglycan-protein recognition

248th National Meeting (2014)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.