Covalent flexible peptide docking in Rosetta.

Area of Science:

• Computational chemistry
• Drug discovery
• Structural biology

Background:

• Electrophilic peptides can irreversibly bind targets, offering potential for previously undruggable proteins.
• Discovering these covalent peptide binders remains a significant challenge in molecular design.

Purpose of the Study:

• To develop and validate a computational pipeline, Rosetta CovPepDock, for identifying and designing electrophilic peptide binders.
• To enable rapid discovery of potent and selective irreversible peptide binders for therapeutic applications.

Main Methods:

• Developed Rosetta CovPepDock, a computational pipeline for peptide docking incorporating covalent binding to receptor cysteines.
• Applied the pipeline retrospectively to known disulfide-bound and electrophilic peptide datasets.
• Created a protocol for designing electrophilic binders from non-covalent complexes or protein-protein interfaces, utilizing a large PDB dataset.

Main Results:

• Rosetta CovPepDock achieved high accuracy in predicting near-native models, especially from native peptide conformations (89-100%).
• The design protocol identified 7154 potential peptide candidates from the PDB.
• A proof-of-concept study yielded seven potent and selective irreversible peptide binders, with one validated by X-ray crystallography.

Conclusions:

• Rosetta CovPepDock is an effective computational tool for discovering electrophilic peptide binders.
• The developed protocol facilitates rapid identification of novel covalent peptide binders.
• These findings hold significant potential for developing new therapeutics and chemical probes.