(482d) Molecular Dynamics Simulations of Complex Mixed Lipid Bilayers to Model Yeast Membranes

There has been growing interest in using molecular dynamics (MD) simulations to investigate membrane-protein systems. However, these simulations have heretofore consisted of no more than three types of lipids, greatly simplifying the compositions of actual membranes. CHARMM-GUI, a web-based graphical user interface for performing various CHARMM functions, has recently been used to generate more complex lipid bilayers composed of six different lipids ? cholesterol, DOPC, DPPC, POPA, POPE, and POPS ? in order to accurately and realistically model yeast membranes. Four bilayers containing varying amounts of chain saturation, cholesterol concentration, and surface tension were created. These yeast membranes were simulated for a total of 170 ns using the atomistic CHARMM27r force field at 303.15 K. MD simulations with a high concentration of unsaturated chains (73%) showed an increase in surface area per lipid and an decrease in the SCD order parameters for DPPC, as opposed to the membrane created with a high concentration of saturated chains (60-63%). Simulations of the more saturated membranes were in a liquid-ordered state and were in agreement with experimental cholesterol-containing membranes. The unsaturated membrane simulation exhibited a larger average tilt angle of cholesterol with respect to the bilayer normal. Moreover, cholesterol in the unsaturated membrane actually oriented parallel to the bilayer surface for periods of less than a nanosecond. This result supports previous observations of parallel cholesterol existing at the center of polyunsaturated fatty acid membranes, and makes us the first to observe parallel cholesterol in a fully atomistic simulation. Currently, we are using the newly developed atomistic CHARMM35 force field to simulate the yeast bilayers with 25-hydroxysterol and cholesterol. In unsaturated bilayers, oxysterols are believed to be more thermodynamically stable in the parallel orientation compared to cholesterol, so it is our aim to elucidate the dynamics of this phenomenon with our new simulations.