Molecular determinants of lipid membrane dynamic properties
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
The cell membrane is a quasi-2D viscous fluid, and a highly heterogeneous mixture based on a foundation of amphipathic lipids. Membrane material properties must be regulated by the cell to remain biologically functional. Viscosity in particular has been shown to be an important physical property: bacterial cells grown under different environmental conditions have been shown to regulate their viscosity, and viscosity has been linked to cellular respiration. Many experimental measurements rely on viscosity as a parameter of the analysis, and yet experimental measurements of membrane viscosity are often difficult and vary widely between techniques. ☐ Molecular dynamics simulations present a uniquely satisfying method to measure membrane properties. As all the positions and momenta in the system are known, hydrodynamic properties can be estimated directly from physical principles alone, without relying on molecular probes which may disrupt the system. As cells regulate their mechanical properties by varying lipid composition, we took a physics-based approach and measured membrane viscosity for a variety of single-component bilayers, using the all-atom CHARMM36 force field in GROMACS. This allowed us to evaluate how specific changes to lipid chemical structure affected membrane properties. ☐ We used a Green-Kubo method based on the fluctuation-dissipation theorem to calculate the membrane viscosity. These kinds of autocorrelation function analyses are guaranteed to have some level of often-neglected statistical bias. We present a thorough analysis of the origin and form of the systematic bias inherent to autocorrelation estimators. We also present a method to overcome this bias when performing numerical curve fitting: many of the most common fitting forms for autocorrelation estimators have bias terms that can be directly computed, and thus the bias can be accounted for in the curve fits. ☐ Molecular dynamics simulation also has the capability to emulate various experimental techniques for a more direct comparison to experimental data and analysis. In collaboration with inelastic X-ray scattering experiment, we simulated three component lipid bilayers in the liquid ordered and liquid disordered phases, directly computing the dynamic structure factor of the membrane (the observable of their scattering method). This analysis showed dynamic behavior unique to the liquid ordered phase from a new perspective and allowed us to investigate the physical origins of these dynamic modes.
Description
Keywords
Cell membranes, Molecular dynamics simulation, Viscosity, Bacterial cells, X-ray scattering experiment