Reliable Redox-Potential Simulations of Proteins

Area of Science:

• Computational chemistry
• Biophysical chemistry
• Quantum mechanics

Background:

• Protein redox potentials are vital for biological processes but difficult to compute due to large system sizes.
• Hybrid quantum mechanical/molecular mechanics (QM/MM) methods often fail due to nonpolarizable force fields.
• Pure quantum mechanical (QM) calculations on cluster models are preferred but face truncation errors.

Purpose of the Study:

• To determine the optimal cluster model size for accurate density functional theory (DFT) based redox potential calculations in proteins.
• To develop and validate a snapshot-based approach for computing standard redox potentials using DFT on truncated QM/MM geometries.
• To compare the accuracy of DFT cluster models against traditional QM/MM methods for redox potential prediction.

Main Methods:

• DFT calculations were performed on cluster models of varying sizes (200-1500 atoms) for ten proteins.
• A snapshot-based approach combined QM/MM molecular dynamics with DFT on ~500-atom cluster models.
• Redox potentials were calculated from vertical reduction energies derived from DFT.

Main Results:

• Cluster models of approximately 500 atoms were found to be sufficient, with truncation errors smaller than typical DFT errors.
• The snapshot-based DFT approach achieved a mean absolute error (MAE) of 0.12 V for standard redox potentials, comparable to intrinsic DFT error.
• Traditional QM/MM energies resulted in a significantly higher MAE of 0.36 V.

Conclusions:

• ~500-atom DFT cluster models provide accurate protein redox potential predictions, overcoming limitations of larger systems and QM/MM methods.
• The proposed snapshot approach offers a reliable and computationally efficient method for calculating redox potentials in complex biological systems.
• These findings advance the prediction of electronic structure changes in biological processes like electron transfer and catalysis.