Computational Method for the Detection of Communication Pathways in Enzymes that Correlate with Experimentally Defined Thermal Activation Networks.

Enzyme catalysis has been shown to depend on distal pathways that channel thermal energy from solvent to the active site. Here, we develop a method to compute these pathways using soybean lipoxygenase (SLO), an extensively studied prototype for dynamical initiation of C-H activation. Experiments with SLO have previously identified a cone-shaped network connecting a surface loop residue (Gln322) to a buried active site residue, Leu546, positioned adjacent to the reactive bond of substrate. We introduce microsecond molecular dynamics and a correlated motion-based protocol to obtain an atomistic analysis of this long-range communication. The developed approach also enables systematic screening of communication between active site-specific residues that directly contact bound substrate and surface-exposed residues on the protein-solvent interface. In addition to providing deeper molecular insight into the experimentally mapped thermal energy network in SLO, this methodology has the potential to discover communication trends across diverse enzyme families. The simulations recover the experimentally demonstrated thermal initiation loop and the Leu546-directed cone in SLO, exclude the negative-control Ser596, and explain the preference for Leu546 over Leu754, a second active site residue in contact with bound substrate. Mutational analysis further reveals the impact of single-site mutations on the network preference between Leu546 and Leu754. These results unify experiments and computation, corroborate an anisotropic channeling of thermal energy in SLO, and establish a general framework for computing site-specific distal intraprotein pathways capable of the thermal initiation of enzyme function.