A molecular dynamics strategy for CSαβ peptides disulfide-assisted model refinement.

Area of Science:

• Biochemistry
• Structural Biology
• Computational Chemistry

Background:

• Cysteine-stabilized antimicrobial peptides show promise as anti-infective agents.
• Accurate structural data is crucial for understanding peptide function and optimizing drug development, but is often experimentally unavailable.

Purpose of the Study:

• To develop and validate a protocol for predicting the structure of cysteine-rich peptides, particularly those stabilized by disulfide bridges.
• To improve the accuracy of peptide structure prediction in the challenging 'midnight zone' of modeling.

Main Methods:

• A protocol was developed using a training set of clustered cysteine-stabilized alpha-beta (CSαβ) structures.
• A structure-based disulfide predictor was employed, and cysteine distances were used as constraints in molecular dynamics simulations.
• A method for final structure selection was proposed and tested on randomly selected peptides.

Main Results:

• The developed protocol achieved a 21% mean root mean square deviation improvement on the test set.
• The study successfully demonstrated the prediction of disulfide bridge networks in cysteine-stabilized peptides.
• The accuracy of structural predictions was significantly enhanced by incorporating disulfide bridge information.

Conclusions:

• Predicting disulfide bridges is feasible and enhances the accuracy of cysteine-rich peptide structure modeling.
• This approach provides a valuable tool for structure-activity relationship studies and the optimization of antimicrobial peptides.
• The methods were successfully applied to predict the structure of the unknown cysteine-rich peptide, royalisin.