Developing hybrid approaches to predict pKa values of ionizable groups

Authors

  • Shawn Witham,

    1. Computational Biophysics and Bioinformatics, Department of Physics, Clemson University, Clemson, South Carolina 29634
    Search for more papers by this author
    • Shawn Witham and Kemper Talley contributed equally to the manuscript.

  • Kemper Talley,

    1. Computational Biophysics and Bioinformatics, Department of Physics, Clemson University, Clemson, South Carolina 29634
    Search for more papers by this author
    • Shawn Witham and Kemper Talley contributed equally to the manuscript.

  • Lin Wang,

    1. Computational Biophysics and Bioinformatics, Department of Physics, Clemson University, Clemson, South Carolina 29634
    Search for more papers by this author
  • Zhe Zhang,

    1. Computational Biophysics and Bioinformatics, Department of Physics, Clemson University, Clemson, South Carolina 29634
    Search for more papers by this author
  • Subhra Sarkar,

    1. School of Computing, Clemson University, Clemson, South Carolina 29634
    Search for more papers by this author
  • Daquan Gao,

    1. Computational Biophysics and Bioinformatics, Department of Physics, Clemson University, Clemson, South Carolina 29634
    Search for more papers by this author
  • Wei Yang,

    1. Department of Chemistry and Biochemistry, Institute of Molecular Biophysics, Florida State University Tallahassee, Florida 32306
    Search for more papers by this author
  • Emil Alexov

    Corresponding author
    1. Computational Biophysics and Bioinformatics, Department of Physics, Clemson University, Clemson, South Carolina 29634
    • Computational Biophysics and Bioinformatics, Department of Physics, Clemson University, Clemson, SC 29634
    Search for more papers by this author

  • The authors state no conflict of interest.

Abstract

Accurate predictions of pKa values of titratable groups require taking into account all relevant processes associated with the ionization/deionization. Frequently, however, the ionization does not involve significant structural changes and the dominating effects are purely electrostatic in origin allowing accurate predictions to be made based on the electrostatic energy difference between ionized and neutral forms alone using a static structure and the subtle structural changes be accounted by using dielectric constant larger than two. On another hand, if the change of the charge state is accompanied by a large structural reorganization of the target protein, then the relevant conformational changes have to be explicitly taken into account in the pKa calculations. Here we report a hybrid approach that first predicts the titratable groups whose ionization is expected to cause large conformational changes, termed “problematic” residues, and then applies a special protocol on them, while the rest of the pKas are predicted with rigid backbone approach as implemented in multi-conformation continuum electrostatics (MCCE) method. The backbone representative conformations for “problematic” groups are generated with either molecular dynamics simulations with charged and uncharged amino acid or with ab-initio local segment modeling. The corresponding ensembles are then used to calculate the titration curves of the “problematic” residues and then the results are averaged to obtain the corresponding pKa. Proteins 2011; © 2011 Wiley-Liss, Inc.

Ancillary