Structural Flexibility of an enzyme upon binding of ligand at an active site: using estimate of entropy over ensemble of contact matrices

Abstract

Contact matrices have been widely used to represent proteins structures. They are transformed to protein contact networks to determine key functional residues, analyze effects of ligand binding, and protein protein interaction. We extend the approach of contact matrices to the aggregation of contact matrices across homologous samples of proteins by defining the frequency contact matrix (FCM). A FCM encodes the frequency of a contact between the side chains of structurally aligned sites. Using this approach, we analyzed the general sitewise contact entropy of a set of protein structures with and without a ligand-bound for residues close to the active site (≤ 10) and farther from the active site. Dataset comprised of enzymes from enolase, mandelate racemase, and muconate cycloisomerase subgroups within the enolase superfamily were constructed. The results show that the median of contact entropy of hydrophobic residues is higher than that for hydrophilic residues. A significant relationship was also observed between the mean contact entropy of residues and their respective properties such as size and hydrophobicity, which indicated that as the hydrophobicity or size of residues increases, the contact entropy of residues also increased. On comparing the contact entropy of residues in the ligand-bound and ligand-free structure, no change was observed for the enolase superfamily. It was also observed that the information obtained from the residue contact entropy values has significant relationship with sequence entropy values. Sequence entropy values indicate the uncertainty of residue type at structurally aligned positions. For some datasets, the residue contact entropy values were found to be sensitive to distance from the active site. The same entropy value distribution for ligand-bound and ligand-free datasets may be due to the dataset collection method and choice of sub-optimal parameters.