Expanding molecular modeling and design tools to non-natural sidechains

Area of Science:

• Biochemistry and Molecular Biology
• Computational Chemistry
• Drug Discovery

Background:

• Protein-protein interactions are crucial for cellular signaling and disease development.
• Inhibiting these interactions with peptide derivatives offers therapeutic potential.
• Designing effective peptide inhibitors requires understanding the impact of amino acid modifications.

Purpose of the Study:

• To develop a general computational framework for handling and predicting the effects of non-natural amino acid sidechains in peptides and proteins.
• To provide tools for designing novel peptide-based inhibitors with enhanced binding affinities.
• To facilitate research in protein engineering and therapeutic development.

Main Methods:

• Generation of structural files (pdb, mol2) and molecular mechanics parameters (CHARMM, Gromacs) for numerous non-natural sidechains.
• Development of a combined knowledge and physics-based strategy for accurate rotamer probability prediction.
• Application of the framework to non-natural mutants of a BCL9 peptide targeting beta-catenin.

Main Results:

• The framework successfully handles hundreds of non-natural amino acid sidechains.
• Accurate predictions of rotamer probabilities were achieved.
• Predicted binding free-energies for non-natural BCL9 peptide mutants showed strong correlation with experimental data.
• The study provides user-friendly visualization plug-ins and available data at http://www.swisssidechain.ch.

Conclusions:

• The developed computational framework enables efficient prediction of non-natural sidechain effects on peptide/protein interactions.
• This approach can guide the design of new peptide inhibitors with improved binding affinity.
• The readily available data and tools empower researchers in developing biological and therapeutic agents.