Improved Structure-Based Histidine pK_a Prediction for pH-Responsive Protein Design

Area of Science:

• Biochemistry and Molecular Biology
• Computational Chemistry
• Structural Biology

Background:

• Histidine's near-neutral pKa is crucial for engineering pH-sensitive biomolecules, particularly antibodies targeting acidic tumor microenvironments or cellular endosomes.
• Histidine pKa values exhibit significant variability (up to 4 pH units), influenced by protein interactions, complicating targeted mutations.
• Accurate pKa prediction is essential for designing novel therapeutic proteins, but current methods lack sufficient precision.

Purpose of the Study:

• To develop and validate an improved computational method for predicting histidine pKa values in proteins.
• To enhance the accuracy of pKa predictions beyond the current standard of 1.0 pH unit.
• To provide a tool that facilitates the rational design of pH-responsive mutations in biomolecules.

Main Methods:

• Utilized Amber force field electrostatics with a continuum solvent model.
• Incorporated limited rotameric sampling and proton optimization.
• Applied an empirical correction for buried side-chains and validated against the PKAD database.

Main Results:

• Achieved a mean unsigned error of 0.4 pH units for histidine pKa predictions.
• Evaluated 91 histidines across 38 diverse protein structures.
• Demonstrated improved accuracy compared to existing prediction standards.

Conclusions:

• The developed method significantly enhances the accuracy of histidine pKa prediction.
• This advancement will improve in-silico design of pH-responsive mutations for therapeutic applications.
• The JustHISpKa software is now available to assist researchers in protein engineering.