How the Ligand-Induced Reorganization of Protein Internal Energies Is Coupled to Conformational Events.

Abstract

Here, we introduce a novel computational method to identify the protein substructures most likely to support the functionally oriented structural deformations that occur upon ligand-binding. To this aim, we study the modulation of protein energetics along the trajectory of a molecular dynamics simulation of different proteins in the presence and in the absence of their respective ligands, namely, human FGF, human second PDZ from human PTP1E/PTPL1, and the N terminal domain of human Hsp90. The method is based on the idea that a subset of protein residues (hotspots) may initiate the global response via the disassembly and reassembly of interactions, which is reflected in the modulation of the overall protein energetics. To identify structural hotspots and dynamic states linked to the onset of functionally relevant conformational transitions, we define an energy profile to monitor the protein energetics, based on a previously introduced approach that highlights the essential nonbonded couplings among all residues. The energy profiles are calculated along the trajectory to yield a time-dependent evolution, and their relative population in the presence and absence of the ligand is evaluated by means of a clustering procedure. It is found that interconversion between clusters, as well as their population and the density of specific energy profiles in the vicinity of structural transitions, provides specific information on the impact of the ligand in driving the protein conformational response. This analysis also highlights the hotspot residues that are most responsive to the presence of the ligand. Importantly, identified hotspots are in agreement with experimental evidence in the three considered systems. We propose that this approach can be generally used in the prediction of "allosteric hotspots" and ligand-induced conformational responses, as well as to select conformations more likely to support functional transitions (e.g., in the framework of adaptive sampling approaches).