Solvent Effect on Protein Folding and Network Presentation of Protein Kinetics

BBL is a small independently folding domain with two main parallel helices. Combining with Replica-Exchange Molecular Dynamics and constant pH Molecular Dynamics, the solvent effect on the free energy landscape of BBL has been studied. The condition of the acidic pH reshaped the free energy surface, leading to a broadly populated denatured state basin and a low free energy barrier between the denatured state and the native state.;Rd-apocytochrome b562 is known to have hidden intermediates which exist after the rate-limiting step and are not detected in kinetic experiments. The kinetics between the hidden intermediates of Rd apocytochrome b562 was investigated using a specialized supercomputer, Anton. Markov state model was used to analyze the simulation results and a network model was proposed. Transition-path theory was employed to calculate the net effective flux from the most unfolded state towards the most folded state in the network. Experimentally undetected PUFs were identified in the network model. The network model better describes the relationship and the role of hidden intermediates in folding process.;Salt induced refolding process of BBL has been studied by Anton at 2 M, 4 M and 6 M with LiCl. The AMBER ff99SB-ILDN force field with a new set of ion parameters were tested in the refolding experiment. Generally, higher salt concentrations facilitate refolding. However, force field flaws have been found in secondary structure propensity in low pH and high salt concentration environments. Comparison of RMSD with topological measurement, Scaled Gauss Metric (SGM) pinpoint that the force field deficiency may be due to the dihedral angles of glycine in the turn region. The salt effect on solvation energy from hydrophobic interactions and electrostatic interactions has been studied. In general, at high salt concentration hydrophobic interactions contribute more to solvation energy than electrostatic interactions and such contribution increases with salt concentrations.;Inherent structure (IS) approach and geometry-based clustering methods are commonly used clustering method to analyze long-time molecular dynamics simulations. Both methods can provide important information about the thermodynamics and kinetics of the systems. A comparison of these two clustering methods in the application of Markov state model has been attempted in the system of (Ala)4. At the micro cluster level, two methods gave totally different results of space decomposition and this discrepancy leads to different macro state discretization. The large IS macro cluster merges some of the largest RMSD macro clusters, which can be explained from the structural and kinetics point of view.