Simultaneous enhancement of multiple functional properties using evolution-informed protein design.

Area of Science:

• Protein engineering and computational biology.
• Enzyme function and directed evolution.

Background:

• Designing proteins with multiple enhanced properties is challenging.
• Predicting combinations of mutations that maintain or improve protein function requires novel methods.
• Sequence co-variation models leverage homologous sequences to predict protein properties and activities.

Purpose of the Study:

• To apply EVcouplings to computationally design variants of TEM-1 β-lactamase.
• To assess the functionality, stability, and activity of designed protein variants.
• To demonstrate the efficacy of evolutionary models in guiding large-scale protein sequence alterations.

Main Methods:

• Utilized EVcouplings, a model of sequence co-variation, for computational protein design.
• Applied the method to the TEM-1 β-lactamase protein.
• Experimentally characterized 14 computationally designed variants.

Main Results:

• Nearly all 14 experimentally characterized designs were functional.
• One variant exhibited 84 mutations from the nearest natural homolog while remaining functional.
• Designed variants showed significant increases in thermostability and activity on multiple substrates.
• Protein structures of the designed variants were nearly identical to the wild type enzyme.

Conclusions:

• Evolutionary models are effective for guiding substantial sequence modifications in protein design.
• This approach can generate functional protein diversity with multiple enhanced properties.
• The study validates the use of EVcouplings for creating highly engineered, functional proteins.