Accurate Structure Prediction for Cyclic Peptides Containing Proline Residues with High-Temperature Molecular...

Area of Science:

• Computational chemistry
• Biochemistry
• Drug discovery

Background:

• Cyclic peptides (CPs) are promising drug candidates, with many natural CPs containing essential peptidyl cis-prolyl bonds.
• Conventional molecular dynamics simulations face challenges in accurately sampling cis/trans isomerization due to high rotational barriers.

Purpose of the Study:

• To develop and validate a novel computational protocol for accurate structure prediction of proline-containing cyclic peptides.
• To enhance the study of cyclic peptide-protein complexes, particularly those involving cis-prolines.

Main Methods:

• Combining high-temperature molecular dynamics (high-T MD) with the residue-specific force field (RSFF2C).
• Utilizing a probability density-based reweighting method for enhanced sampling.
• Applying the protocol to predict structures of 23 cyclic peptides and analyze the SFTI-1/trypsin complex.

Main Results:

• The proposed method accurately predicted the structures of 19 out of 23 cyclic peptides (backbone RMSD < 1.0 Å) compared to X-ray data.
• Demonstrated applicability to CP-complexes by studying the sunflower trypsin inhibitor (SFTI-1)/trypsin complex.
• The conformation of SFTI-1 in solution was found to be consistent with its bound conformation.

Conclusions:

• The developed protocol offers a robust and accurate approach for predicting the structures of proline-containing cyclic peptides.
• This method facilitates the study of cyclic peptide-protein interactions, aiding in drug design and development.
• The findings suggest that the solution conformation of SFTI-1 may pre-organize it for binding to trypsin.