| Biological membranes, mainly composed of lipid bilayers and proteins, are important in all the cells. Proteins in the membrane carry out most of the membrane functions. Knowledge about the protein-lipid interactions is essential to understanding the function of both protein and lipid. Some of the proteins that are hydrophobic can effectively partition into membranes, however, for those with charged residues, there have been intense debates on the partitioning energetics and charged protein-lipid interactions. Traditional views of the hydrophobic membrane interior as a hostile environment for charged species have been challenged by recent membrane protein experiments. Furthermore, charged proteins play important roles in many protein functions, such as antimicrobial peptides, ion channels, and cell penetrating peptides. To resolve such controversies, this thesis focuses on charged protein-lipid interactions to study the peptide partitioning energetics and ion translocation mechanisms in membranes.;All-atom Molecular Dynamics (MD) simulations have been applied in the simulations of poly-Alanine peptide with one arginine (Arg) sidechain partitioning into Dipalmitoylphosphatidylcholine (DPPC) membranes. This peptide, which is a simplified model from the GWALP peptide widely used in protein partitioning experiments, is both a good comparison to experiments and an easy model for method developing. This study generated new methods to improve free energy calculations in residue mutation and movements. With these methods, we demonstrated that the partitioning of a transmembrane helix containing one Arg sidechain involved a free energy penalty of ∼ 20kcal/mol, which is affected by various factors such as peptide tilting, peptide rotation, and anchor strength. Furthermore, this study was a new attempt in simulating the partitioning of realistic peptides using all-atom models. From this study, we showed the advantages and difficulties of simulating real peptide, which provided direct connections between MD simulations and biological experiments.;In the study of mechanisms of ion translocation in membranes, membrane thickness was suggested as an important factor in previous research, while membrane polarizability has not been well investigated in the past, thus we are especially interested in the role of membrane thickness and polarizability in this thesis. We predicted two mechanisms, including the ion-induced defect mechanism which involves membrane deformations and energy cost growing with membrane thickness, and the solubility diffusion mechanism involving ion partitioning, for which we predicted a cost of 25∼30 kcal/mol according to the previous research. All-atom MD simulations of an Arg side chain analog, MguanH +, moving across bilayers of mono-unsaturated phosphatidylcholine (PC) lipids with and without cholesterol of a wide range of thicknesses have been performed, in order to study the effect of membrane thickness on the charged protein-lipid interactions. Moreover, to understand the role of polarizability on the ion translocation mechanism, both polarizable and non-polarizable models have been applied to PC bilayers of interest. For all non-polarizable membranes, the ion translocation caused membrane deformations, leading to sharp free energy barriers ranging from 14 kcal/mol to 40 kcal/mol with similar shapes and slopes, which indicated an ion-induced defect mechanism in non-polarizable models. However, in polarizable models, ion translocation was found to start with an ion-induced defect mechanism, and then transfer to a solubility-diffusion mechanism when the free energy cost reached 26 kcal/mol, from which an upper limit of ion translocation energy barrier of 26 kcal/mol has been demonstrated for the first time. Furthermore, membrane polarizability has been proved essential in sampling the changing membrane charge transport mechanisms.;With MD simulations we were able to achieve deeper understanding of the charge-lipid interactions and mechanisms governing peptide partitioning and ion translocation at the atomic level. This research will help understand a broad range of biological phenomena involving protein partitioning and translocations, such as the mechanisms of viral peptides and cell penetrating peptides, the invention of new functionalized bionanodevices or drug delivery, voltage gated ion channel function to treat various disorders, and for basic knowledge of proteins that control our nervous systems. |
