NERSC Initiative for Scientific Exploration (NISE) 2011 Awards

Molecular Simulations of the Release of Phosphate and ADP from F1-ATPase to Aid in the Understanding of the Function of F1-ATPase

Victor Ovchinnikov, Massachusetts Institute of Technology

Associated NERSC Project: Understanding the function of protein and nucleic acid components of cellular machines necessary for optimizing biofuel production (gc8), Principal Investigator: Martin Karplus

The focus of this proposal is to increase our understanding of the mechanism of phosphate and ADP release and the steps where they are released during the function of F1-ATPase. The understanding of this key step will ultimately allow us to understand the entire mechanism of chemomechanical coupling of F0F1-ATPase. In addition, the results will aid in modifying the systems we study to make them more useful in the biological production of energy.

The objective of the proposed study is to provide a complete model of F1-ATPase (the catalytic moiety of F0F1-ATPase) action, which is essential to understand the motor mechanism of F0F1-ATPase. Recent experiments suggest that the release of phosphate (Pi), one of the products of ATP hydrolysis, from one of the active sites generates a significant portion of the torque to drive the γ-stalk rotation of F1-ATPase. However, it is not known which of the three nonequivalent subunits of F1-ATPase is the source of ADP and Pi.

We propose to investigate the mechanisms and pathways of the release of Pi and ADP from the β-subunit in different conformations of F1-ATPase. To overcome the time scale of the actual Pi (and ADP) release, multicopy enhanced sampling (MCES) (1) simulations will be used to accelerate the release process. The simulations will be repeated multiple times to obtain sufficient statistics, where each of the simulations will start from different distribution of initial velocities. The multiple simulations will be carried out by using multiple MPI communicators that we have implemented recently in the CHARMM program. Since the multiple simulations do not require extensive parallel communications between each simulation, we expect near perfect parallel efficiency. After finding the Pi (and ADP) release pathways, the string method (2) will be applied to compute the free energy along the pathway.