Salt influences protein stability through electrostatic mechanisms as well as through nonpolar Hofmeister effects. In the present work, a continuum solvation based model is developed to explore the impact of salt on protein stability. This model relies on a traditional Poisson-Boltzmann (PB) term to describe the polar or electrostatic effects of salt, and a surface area dependent term containing a salt concentration dependent microscopic surface tension function to capture the non-polar Hofmeister effects. The model is first validated against a series of cold-shock protein variants whose salt-dependent protein fold stability profiles have been previously determined experimentally. The approach is then applied to HIV-1 protease in order to explain an experimentally observed enhancement in stability and activity at high (1M) NaCl concentration. The inclusion of the salt-dependent non-polar term brings the model into quantitative agreement with experiment, and provides the basis for further studies into the impact of ionic strength on protein structure, function, and evolution.

Recently, it was demonstrated that implicit solvent models were capable of generating stable B-form DNA structures. Specifically, generalized Born (GB) implicit solvent models have improved regarding the solvation of conformational sampling of DNA [1,2]. Here, we examine the performance of the GBSW and GBMV models in CHARMM for characterizing base flipping free energy profiles of undamaged and damaged DNA bases. Umbrella sampling of the base flipping process was performed for the bases cytosine, uracil and xanthine. The umbrella sampling simulations were carried-out with both explicit (TIP3P) and implicit (GB) solvent in order to establish the impact of the solvent model on base flipping. Overall, base flipping potential of mean force (PMF) profiles generated with GB solvent resulted in a greater free energy difference of flipping than profiles generated with TIP3P. One of the significant differences between implicit and explicit solvent models is the approximation of solute-solvent interactions in implicit solvent models. We calculated electrostatic interaction energies between explicit water molecules and the base targeted for flipping. These interaction energies were calculated over the base flipping reaction coordinate to illustrate the stabilizing effect of the explicit water molecules on the flipped-out state. It is known that nucleic base pair hydrogen bonds also influenced the free energy of flipping since these favorable interactions must be broken in order for a base to flip-out of the helix. The Watson-Crick base pair hydrogen bond fractions were calculated over the umbrella sampling simulation windows in order to determine the effect of base pair interactions on the base flipping free energy. It is shown that interaction energies between the flipping base and explicit water molecules are responsible for the lower base flipping free energy difference in the explicit solvent PMF profiles.

In the present article, coarse grained Dissipative Particle Dynamics simulation with implementation of electrostatic interactions is developed in constant pressure and surface tension ensemble to elucidate how the antimicrobial peptide molecules affect bilayer cell membrane structure and kill bacteria. We find that peptides with different chemical-physical properties exhibit different membrane obstructing mechanisms. Peptide molecules can destroy vital functions of the affected bacteria by translocating across their membranes via worm-holes, or by associating with membrane lipids to form hydrophilic cores trapped inside the hydrophobic domain of the membranes. In the latter model, the affected membranes are strongly buckled, in accord with very recent experimental observations [G. E. Fantner et al., Nat. Nanotech., 5 (2010), pp. 280-285].