Modeling the allostery transition of protein molecules: A double-well network model (DWNM)

PHYS 133

Jhih-Wei Chu, jwchu@berkeley.edu, Department of Chemical Engineering, University of Carlifornia, Berkeley, 101A Gilman, UC Berkeley, Berkeley, CA 94720 and Gregory A. Voth, voth@chem.utah.edu, Center for Biophysical Modeling & Simulation, University of Utah, Department of Chemistry, 315 S. 1400 E. Rm 2020, Salt Lake City, UT 84112-0850.
A coarse-grained (CG) model is proposed to study protein structural transition. This model takes the two structures of a protein before and after the allosteric transition as input, with Cα atoms selected as the CG sites in the current implementation. Between each pair of CG sites within a cutoff distance, a double-well (DW) potential is used to describe their interaction. The model thus contains a network of interconnected double-well potentials, and is reduced to the elastic network model (ENM) near the reference structures. Compared to a plastic network model (PNM) proposed earlier (J. Mol. Biol., 353, 807, 2005), DWNM introduces roughness in the conformational free energy surface and allows multiple reaction channels to connect two structures. The results of applying DWNM to the allosteric transition of G-actin and adenylate kinase (AKE) will be described in detail. We found that DWNM pathways are consistent with the biochemical data of G-actin and compare favorably with the all-atom molecular dynamics simulations of AKE. Our results indicate that in addition to the shape of a protein molecule, the topology of the conformational free energy surface also plays a determining role in the transition pathways of protein allostery and the intermediates in between.