Estimation of Hydrogen-Exchange Protection Factors from MD Simulation Based on Amide Hydrogen Bonding Analysis

Area of Science:

• Biophysics
• Computational Biology
• Structural Biology

Background:

• Hydrogen exchange (HX) studies, particularly HX mass spectrometry (HX-MS), are vital for understanding protein folding, structure, and dynamics.
• Interpreting the extensive data from HX-MS and other HX methods requires reliable computational predictions based on protein structures or models.
• Previous computational HX modeling often relied on solvent-accessible surface area metrics, with limited success in interpreting experimental data.

Purpose of the Study:

• To develop and demonstrate a novel computational method for predicting hydrogen exchange mass spectrometry (HX-MS) data.
• To validate the prediction method by comparing computational results with experimental HX-MS data for various proteins.
• To assess the method's ability to improve experimental data resolution and provide residue-level insights into protein dynamics.

Main Methods:

• A computational HX-MS prediction method was developed, classifying amide hydrogen bonding modes to mimic local unfolding.
• Molecular dynamics (MD) simulation snapshots were analyzed to determine NH bonding configurations and calculate protection factors (PFs).
• A logistics growth function was used to map bonding states to PFs, which were then used to predict peptide deuteration values.

Main Results:

• High correlation was observed between predicted and experimental deuteration values for fragment peptides in fatty acid synthase thioesterase (FAS-TE) and other benchmarking systems.
• The method successfully decomposed in-exchange curves into rate classes that correlated with MD predictions, indicating accurate capture of physical processes.
• Residue-resolved protection factor predictions for staphylococcal nuclease showed strong agreement with NMR data, outperforming other literature algorithms.

Conclusions:

• The developed MD-based HX prediction approach provides a transferable and scalable tool for interpreting HX-MS data using protein structures or models.
• This method significantly enhances the analysis of HX-MS experiments by offering accurate predictions and improving experimental resolution.
• The approach correctly captures underlying physical processes at the single-residue level, advancing the field of protein dynamics analysis.