Physics-based enzyme design: predicting binding affinity and catalytic activity.

Area of Science:

• Biochemistry
• Computational Biology
• Protein Engineering

Background:

• Computational enzyme design aims to create novel enzymes with desired functions.
• Current methods face challenges in accurately and rapidly evaluating enzyme variants.
• There is a need for accessible, automated computational tools grounded in physical-chemical principles.

Purpose of the Study:

• To apply and benchmark a physics-based implicit solvent MM-GBSA scoring approach for enzyme design.
• To assess the accuracy of MM-GBSA in predicting changes in protein-ligand affinity and enzyme catalytic activity.
• To develop an automated and accessible computational framework for enzyme design.

Main Methods:

• Utilized a physics-based implicit solvent MM-GBSA scoring approach.
• Benchmarked computational predictions against experimental activity data.
• Evaluated performance on a steroid binder protein, a Kemp eliminase, and an α-Gliadin peptidase.

Main Results:

• MM-GBSA accurately ranked experimentally active enzyme variants across different protein systems.
• The computational approach demonstrated effectiveness in predicting changes in binding affinity and catalytic turnover.
• The developed framework successfully identified superior enzyme variants.

Conclusions:

• The MM-GBSA scoring approach shows significant promise for enhancing computational enzyme design.
• This method can enrich the selection of active enzyme variants, reducing experimental workload.
• The automated framework facilitates broader accessibility for enzyme design researchers.