Turning a Mesophilic Protein into a Thermophilic One: A Computational Approach Based on 3D Structural Features

Abstract

In spite of considerable improvement of our understanding of factors responsible for protein thermostability, rational designing of thermostable variants of mesophilic proteins is not yet fully established. The present paper describes an effective computational strategy that we have developed to identify the most suitable mutations converting a chosen mesophilic protein into a thermophilic one starting from its 3D structure. The approach is based on the concept that stabilization of several surface residues should enhance the global stability of the protein. The method relies on the estimation of electrostatic and van der Waals interactions in computing the interaction among the side chains of individual residues and the rest of the protein. The polar or charged residues whose side chains interact weakly with the rest of the original protein are identified first. Then, for each such identified residue (A), another residue (B) in its spatial vicinity is identified. The side chain of the residue (B) is then replaced by a suitable conformer of a residue that is electrostatically complementary to the residue (A) to enhance local interactions and hence the stability of the protein. The steric effect is taken care of through van der Waals interactions. We reject the mutations that improve interactions only locally along the sequence as it is unlikely to enhance the global stability of the 3D architecture. We use the difference in self-energies (DeltaEself) as a measure of the stability difference between the original and its mutant variant. This paper presents two test cases with demonstration of the enhanced stability of such mutated proteins and validates the strategy by considering five experimentally known thermophilic-mesophilic protein pairs.