An Analysis of Central Residues Between Ligand-Bound and Ligand-Free Protein Structures Based on Network Approach

Depiction of protein structures as networks of interacting residues has enabled us to understand the structure and function of the protein. Previous investigations on closeness centrality have identified protein functional sites from three- dimensional structures. It is well recognized that ligand binding to a receptor protein induces a wide range of structural changes.

An interesting question is how central residues function during conformational changes triggered during ligand binding? The aim of this study is to comprehend at what extent central residues change during ligand binding to receptor proteins.

To determine this, we examined 37 pairs of protein structures consisting of ligand-bound and ligand-free forms. These protein structures were modelled as an undirected network and significant central residues were obtained using residue centrality measures. In addition to these, the basic network parameters were also analysed.

On analysing the residue centrality measures, we observed that 60% of central residues were common in both the ligand-bound and ligand-free states. The geometry of the central residues revealed that they were situated closer to the protein center of the mass. Finally, we demonstrated the effectiveness of central residues in amino acids substitutions and in the evolution itself. The closeness centrality was also analyzed among different protein domain sizes and the values gradually declined from single-domains to multi-domain proteins suggesting that the network has potential for hierarchical organization. Betweenness centrality measure was also used to determine the central residues and 31% of these residues were common between the holo/apo states.

Findings reveal that central residues play a significant role in determining the functional properties of proteins. These results have implications in predicting binding/active site residues, specifically in the context of drug designing, if additional information concerning ligand binding is exploited.