Modeling Chromatography Binding through Molecular Dynamics Simulations with Resin Fragments.

Area of Science:

• Biophysical chemistry
• Computational chemistry
• Chemical engineering

Background:

• Accurate modeling of protein-resin interactions is crucial for optimizing biopharmaceutical purification processes.
• Current methods often require extensive experimental data, increasing time and cost.

Purpose of the Study:

• To develop and validate a molecular dynamics simulation-based approach for predicting protein interaction sites and retention times on chromatography resins.
• To establish a physics-based model for guiding the selection of purification conditions in the biotech and pharmaceutical industries.

Main Methods:

• Mimicking chromatography resins with unrestrained ligands and simulating protein-ligand interactions in explicit solvent using Replica Exchange Molecular Dynamics with Hydrogen Mass Repartitioning (REMD-HMR).
• Computing ligand interaction surfaces from simulation trajectories and correlating them with experimental retention times.
• Applying the protocol to ubiquitin mutants and subsequently to six proteins with known Capto MMC elution times, analyzing residue-level physicochemical descriptors.

Main Results:

• The simulation protocol successfully predicted interaction sites and retention times for proteins on Capto MMC resin.
• A linear model based on aromaphilicity and hydrophobicity accurately predicted protein elution order based on simulation-derived descriptors.
• The approach demonstrated robustness to outliers and high correlation with experimental data.

Conclusions:

• The developed physics-based modeling approach provides a reliable and efficient method for predicting protein-resin interactions and optimizing chromatography-based purification.
• This method reduces the need for large experimental datasets and is adaptable to various resins and biomolecules.