Network Dynamics as Fingerprints of Thermostability in an In Silico-Engineered DyP-Type Peroxidase

Stabilizing industrial enzymes is crucial for advancing environmentally responsible bioprocesses; however, the structural basis of thermostability remains incompletely understood. Here, we engineered thermostable variants of a tetrameric dye-decolorizing peroxidase (DyP) using two independent open-source design algorithms, yielding enzymes with significantly improved thermal performance and prolonged activity at elevated temperatures. Subsequent recombination strategies minimize the mutational burden while maintaining or enhancing stability. Structural and dynamic analyses of the thermostable variants revealed convergent features, including increased compactness, rigidity, and an enriched network of hydrogen bonds and hydrophobic interactions. Despite differing mutation profiles, stabilizing substitutions clustered in similar structural regions. Notably, the integration of dynamic modeling with protein correlation network analysis uncovered a previously unrecognized fingerprint of stabilization: highly connected structural networks characterized by denser and more persistent intra- and intermonomer interactions, greater internal cohesion, and enhanced cooperative dynamics. Tetramers exhibit long-range communication pathways and redundant routes, supporting coordinated motions that can hinder local unfolding and tetramer dissociation. These findings identify dynamic interaction networks as hypothetical new indicators of protein stability and offer a previously unexplored framework for rational enzyme design.