Transition State-Based Computational Enzyme Design.

Area of Science:

• Computational biology
• Biochemistry
• Protein engineering

Background:

• Traditional knowledge-based models have limitations in designing proteins with novel functionalities.
• Physics-based approaches offer greater flexibility for incorporating unusual chemical entities and complex energetic landscapes.

Purpose of the Study:

• To present a physics-based computational approach for protein design using the Proteus software.
• To demonstrate the utility of this approach in enzyme engineering, specifically for altering substrate stereospecificity.

Main Methods:

• Development and application of the Proteus software for physics-based energy evaluation and sequence-conformation exploration.
• Utilizing adaptive landscape flattening for direct free energy difference sampling.
• Applying the model to the tyrosyl-tRNA synthetase system to investigate stereospecificity inversion.

Main Results:

• The Proteus model successfully identified the native sequence for L-tyrosine recognition.
• The model predicted specific mutations capable of altering the enzyme's specificity towards D-tyrosine.
• Demonstrated the capability of physics-based methods to guide rational enzyme design.

Conclusions:

• Physics-based computational protein design, as implemented in Proteus, is a powerful tool for enzyme engineering.
• This approach facilitates the design of enzymes with altered substrate specificities by considering detailed energetic factors.
• The methodology provides a framework for exploring enzyme functional modifications and designing novel biocatalysts.