This project focuses on the study of membranes, proteins and carbohydrates by molecular dynamics computer simulation. Progress is reported under each Aim listed above Aim 1. Understand Model Membranes. Analytical expressions for the diffusion constants of tethered lipids were developed within the context of the Saffman Delbruck (S-D) hydrodynamic model for diffusion of membrane components. It is shown that the S-D parameters (membrane thickness and ratio of membrane to water viscosity) required to fit simulation results are borderline physical. This implies that the S-D model is significantly less useful that previously believed. This work will be submitted for publication shortly. The simulations used to demonstrate these discrepancies with the S-D model were also used to compare with experiment (4), and are described in detail in Aim 3. The lateral pressure profiles of cardiolipin (a 4-chained lipid) and diacyl lipids have been compared in efforts to understand the spontaneous curvature of lipid bilayer directly from computer simulation.
Aim 2. Develop Simulation Methodology. Force field (FF) development and testing continued for lipids (3), carbohydrates (2,6,7), and ethers (5). A new lipid FF, C36, for use in the CHARMM (1) program was submitted for publication and published (3). It successfully reproduces structural (surface areas, chain order parameters), mechanical (area compressibility modulus), and dynamic (frequency dependent NMR relaxation times) quantities for six different lipids. This is by far the most comprehensive and parameterization of lipids in the field, and many research groups are already using it. Work is presently underway to test C36 on cardiolipin, phosphoinositols (by combining the lipid and carbohydrate FF), and sphingomyelin (which requires development of new terms). These advances will enable the simulation of increasingly more realistic models of the cell membrane. A comprehensive reworking of the lipid-ion parameters is also underway. FF terms the glycosidic linkages for the new carbohydrate FF were published (2), and translational diffusion constants and NMR spin-lattice relaxation times were evaluated from simulations of glucose, trehalose, and maltose. Results agree well with experiment, and further validate the FF. The comparison required the determination of the valid range for viscosity scaling based on the concentration dependent viscosity of the solute in question. Hydrodynamic calculations yielded insight to hydration patterns for these sugars. The work in press in the Journal of Physical Chemisty B (7). Simulations on a set of branched tri-glucose were also carried out, and the anti/syn ratios were shown to agree well with experiment. The results of this study will be submitted for publication shortly. A coarse-grained (CG) model for polyethylene oxide (PEG) was published (5). FF terms for the interaction of PEG with membrane lipids were developed based on comparisons with all-atom MD simulations.
Aim 3. Simulate Complex Membranes CG simulations of different ratios of PEGylated and normal lipids were carried out using the new FF described in Aim 2. The phase behavior (vesicles, bicelles and micelles as concentration increased) agree well with experiment. Analysis of the lipid distributions on the bicelle surfaces revealed some segregation of PEGylated lipids to the edges, but not nearly so much as had been assumed in interpretation of the experimental data. This work has been submitted to the Journal of the American Chemical Society. Simulations of tethered dimers and trimers of lipids in a bilayer were carried out to explore an experimental observation that the diffusion constants of pleckstrin homology (PH) domains complexed to PIP3 and tethered to each other were inversely proportional to the number of proteins in the complex. This is reminiscent of the free-draining limit in solution hydrodynamics, and is inconsistent with the Saffman-Delbruck model with the usual parameters. This observation not only has practical applications (e.g., determination of the stoichiometry of membrane-protein complexes), but also has significant implications for theoretical description of bilayer lipid dynamics. The combined experimental and simulation paper is in press the Biophysical Journal (4). In an effort to help understand experimental observations of pooling of PIP2 on the surface of nascent phagosomes, Langevin dynamics simulations of PIP2 in a corral of actin filaments were carried out. The simulations revealed that translational diffusion of PIP2 could be slowed down substantially by the electrostatic fence made by the actin, but not enough to explain the experimentally observed reduction. This implies that the compartmentalization of PIP2 on this cell surface is being effected by other components of the membrane, not actin alone. A paper describing these results has been submitted to the Biophysical Journal.
Showing the most recent 10 out of 52 publications
