Potential of Mean Force Calculations with an Efficient On-The-Fly Scrutinization Scheme for Protein-Protein Association
Dhananjay Joshi1,2*, Jung-Hsin Lin3,4
1Taiwan International Graduate Program, Academia Sinica, Taipei, Taiwan
2Institute of Biochemical Sciences, National Taiwan University, Taipei, Taiwan
3Research Centre for Applied Science and Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan
4School of Pharmacy, College of Medicine, National Taiwan University, Taipei, Taiwan
* presenting author:Dhananjay Joshi, email:joshid@gate.sinica.edu.tw
Protein-protein association (PPA) is important in almost all biological processes such as immune response, signal transduction, DNA replication, transcription and translation etc. Estimating the statistical thermodynamics-based energetics and unraveling the underlying physical pathways of biomolecular association are essential for understanding the related molecular mechanisms. Although biophysical and biochemical experimental approaches (e.g., Surface Plasmon Resonance, Isothermal Titration Calorimetry, etc.) can provide some kinetic and thermodynamic information, the atomistic details of the physical process are apparently always missing. All-atom molecular dynamics (MD) simulation is so far still the only way to estimate energetics with deep underlying structural dynamics information. In this work, we address protein-protein association/dissociation (PPA) of the barnase/barstar protein complex by performing potential of mean force (PMF) calculations with the umbrella sampling method using all atom explicit solvent MD simulations. We demonstrate that our new strategy of PMF calculation could be efficiently used to detect both obvious and non-obvious docking decoys. Next, multiple
independent PMF runs for the same dissociation confirmed the robustness and the reliability of the designed strategy. Further, we showed that variation in pulling rates along the reaction coordinate affects the PMF profile significantly. We determined two significantly different physical pathways for the protein-protein dissociation. From this finding, we anticipate our new on-the-fly structure-based scrutinization scheme for the umbrella sampling simulations or non-equilibrium approaches can significantly improve the computational efficiency for the PMF construction.

Keywords: Molecular Dynamics Simulation, Protential of mean force, protein dynamics