GArNet: A Genetic Algorithm-Based Protein Redesign Approach to Optimize Mutation Combinations Informed by Network...

Area of Science:

• Biochemistry
• Bioengineering
• Computational Biology

Background:

• Protein redesign is crucial for enzyme engineering, aiming to improve enzyme properties through targeted mutations.
• A significant challenge in protein redesign is identifying optimal mutation combinations efficiently and reproducibly.

Purpose of the Study:

• To introduce GArNet, a Genetic Algorithm-based protein redesign approach informed by Network theory, for optimizing mutation combinations.
• To evaluate GArNet's effectiveness in identifying beneficial mutations and enhancing enzyme properties.

Main Methods:

• GArNet employs a two-phase optimization process using virtual evolution to generate mutation networks.
• Phase I involves creating complete network data from homologous sequences, representing mutations as nodes and co-occurrence as edges.
• Phase II converts the mutation network into a scale-free network to select high-centrality mutations.

Main Results:

• GArNet demonstrated >73.5% mutational reproducibility for S-hydroxynitrile lyase (S-HNL) design, significantly outperforming conventional methods (25%).
• Computational analysis and experimental tests confirmed GArNet's ability to generate active S-HNL and NAD+-dependent l-threonine 3-dehydrogenase (TDH) mutants.
• The generated mutants exhibited enhanced thermostability and soluble expression compared to native enzymes.

Conclusions:

• Representing mutations and their co-occurrence as a network is an effective strategy for identifying beneficial mutations in protein redesign.
• GArNet offers a reproducible and efficient method for enzyme engineering, leading to improved enzyme variants.
• The GArNet tool is publicly available, facilitating its application in protein redesign research.