A fast and accurate computational approach to protein ionization

Area of Science:

• Computational Biology
• Biophysics
• Biochemistry

Background:

• Accurate prediction of protein electrostatic properties and titration site pK values is crucial for understanding protein function.
• Existing methods often face limitations in speed, accuracy, or applicability to both soluble and membrane proteins.

Purpose of the Study:

• To develop a fast and accurate physics-based computational method for calculating pH-dependent electrostatic effects in proteins.
• To predict the pK values of individual titration sites within protein molecules.
• To refine hydrogen atom coordinates at specific pH values.

Main Methods:

• Combines Generalized Born approximation for electrostatic energy calculations with an iterative mobile clustering approach for protonation equilibria.
• Utilizes the GBIM (Generalized Born with Implicit Membrane) CHARMm module for modeling both soluble and membrane proteins.
• Incorporates a novel algorithm for preliminary refinement of hydrogen coordinates and uses pentapeptide structures as model compounds.

Main Results:

• Achieved high accuracy in pK value prediction, with an average RMSD of approximately 0.5 pK units across 24 test proteins.
• Demonstrated very low computational cost for achieving high accuracy.
• Showed excellent agreement between predicted pH-dependent hydrogen atom assignments and experimentally observed protonation states and hydrogen-bond networks.

Conclusions:

• The developed method offers a fast, accurate, and computationally efficient way to study pH-dependent phenomena in proteins.
• Applicable to a wide range of proteins, including membrane proteins, facilitating research in enzyme catalysis, ligand binding, and protein stability.
• Provides a valuable computational protocol for investigating pH-mediated biological mechanisms.