Attractive membrane domains control lateral diffusion

Simulations and integral-equation theories for dipolar density interacting disks

Integral equation theories (IETs) based on the Ornstein-Zernike (OZ) relation can be used as an analytical tool to predict structural and thermodynamic properties and phase behavior of fluids with low numerical cost. However, there are no studies of the IETs for the dipolar density interaction potential in two-dimensional systems, a relevant interdomain interaction in lipid monolayers with phase coexistence. This repulsive interaction arises due to the excess dipole density of the domains, which are aligned perpendicular to the interface. This work studies the performance of three closures of the OZ equation for this novel system: Rogers-Young (RY), modified hypernetted chain (MHNC), and variational modified hypernetted chain (VMHNC). For the last two closures the bridge function of a reference system is required, with the hard disk being the most convenient reference system. Given that in two dimensions there is no analytical expressions for the hard disk correlation functions, two different approximations are proposed: one based on the Percus-Yevick (PY) approximation, and the other based on an extension of the hard spheres Verlet-Weis-Henderson-Grundke (LB) parametrization. The accuracy of the five approaches is evaluated by comparison of the pair correlation function and the structure factor with Monte Carlo simulation data. The results show that RY closure is satisfactory only for low-structured regimes. MHNC and VMHNC closures perform globally well, and there are no significant differences between them. However, the reference system in some cases affects their performance; when the pair correlation function serves as the measure, the LB-based closures quantitatively outperform the PY ones. From the point of view of its applicability, LB-based closures do not have a solution for all studied interaction strength parameters, and, in general, PY-based closures are numerically preferable.

Molecular transport in systems containing binding obstacles

We studied the movement of particles in crowded environments by means of extensive Monte Carlo simulations. The dynamic lattice liquid model was employed for this purpose. It is based on the cooperative movement concept and allows the study of systems at high densities. The cooperative model of molecular transport is assumed: the motion of all moving particles is highly correlated. The model is supposed to mimic lateral motion in a membrane and therefore the system is two-dimensional with moving objects and traps placed on a triangular lattice. In our study the interaction (binding) of traps with moving particles was assumed. The conditions in which subdiffusive motion appeared in the system were analysed. The influence of the strength of binding on the dynamic percolation threshold was also shown. The differences in the dynamics compared to systems with impenetrable obstacles and with systems without correlation in motion were presented and discussed. It was shown that in the case of correlated motion the influence of deep traps is similar to that of impenetrable obstacles. If the traps are shallow a recovery to normal diffusion was observed for longer time periods.

Influence of Bilayer Size and Number in Multi-Bilayer DOPC Simulations at Full and Low Hydration

Biophysical studies of model cell membranes at full and low hydration are usually carried out using scattering measurements on multi-bilayer systems. Molecular simulations of lipid bilayers aimed at reproducing those experimental conditions are usually conducted using single bilayers with different amounts of water. These simulation conditions may lead to artifacts arising from size effects and self-interactions because of periodic boundary conditions. We have tested the influence of the size and number of bilayers on membrane properties using the Lipid14 force field for lipids in molecular dynamics simulations of 1,2-dioleoyl- sn-glycero-3-phosphocholine bilayers at full hydration (44 water molecules per lipid), low hydration (18 water molecules per lipid), and dehydration (9 water molecules per lipid). A number of additional simulations were conducted with the Slipids force field for comparison. We have found that the average area per lipid (APL), thickness, mass density profiles, and acyl tail order parameters are insensitive to the size and the number of bilayers for all hydration states. The Lipid14 force field can also successfully reproduce the experimentally observed decrease in APL and corresponding increase in bilayer thickness upon dehydration, reflecting the increase in ordering as the system becomes more gel-like. Additionally, decreasing hydration levels were associated with a trend away from normal lateral diffusion and toward more subdiffusive regimes across both force fields. In summary, at least for the Lipid14 force field, the use of a single bilayer with 128 phospholipid molecules provides an adequate representation of multi-bilayer systems at varying levels of hydration.

Domain size polydispersity effects on the structural and dynamical properties in lipid monolayers with phase coexistence

In lipid monolayers with phase coexistence, domains of the liquid-condensed phase always present size polydispersity. However, very few theoretical works consider size distribution effects on the monolayer properties. Because of the difference in surface densities, domains have excess dipolar density with respect to the surrounding liquid expanded phase, originating a dipolar inter-domain interaction. This interaction depends on the domain area, and hence the presence of a domain size distribution is associated with interaction polydispersity. Inter-domain interactions are fundamental to understanding the structure and dynamics of the monolayer. For this reason, it is expected that polydispersity significantly alters monolayer properties. By means of Brownian dynamics simulations, we study the radial distribution function (RDF), the average mean square displacement and the average time-dependent self-diffusion coefficient, D(t), of lipid monolayers with normally distributed size domains. For this purpose, we vary the relevant system parameters, polydispersity and interaction strength, within a range of experimental interest. We also analyze the consequences of using a monodisperse model to determine the interaction strength from an experimental RDF. We find that polydispersity strongly affects the value of the interaction strength, which is greatly underestimated if polydispersity is not considered. However, within a certain range of parameters, the RDF obtained from a polydisperse model can be well approximated by that of a monodisperse model, by suitably fitting the interaction strength, even for 40% polydispersities. For small interaction strengths or small polydispersities, the polydisperse systems obtained from fitting the experimental RDF have an average mean square displacement and D(t) in good agreement with that of the monodisperse system.

Scaling relationships for translational diffusion constants applied to membrane domain dissolution and growth

We compare the way that relationships for diffusion constants scale with the size of diffusing membrane domains and the geometry of their environments. Then, we review our experimental work on the dynamics of dissolution/growth of membrane domains in crowding induced mixing, phase separation, and Ostwald ripening in a highly confined environment. Overall, the scaling relationships applied to diffusion constants obtained by fits to our dynamic data indicate that dissolution and growth is influenced by the diffusion of clusters or small domains of lipids, in addition to kinetic processes and geometrical constraints.

Regulation of phase boundaries and phase-segregated patterns in model membranes

Demixing of components has long been described in model membranes. It is a consequence of non-ideal lateral interactions between membrane components, and it causes the presence of segregated phases, forming patches (domains) of different properties, thus introducing heterogeneity into the membrane. In the present review we first describe the processes through which domains are generated, how they grow, and why they are rounded, striped or fractal-like, as well as why they get distributed forming defined patterns. Next, we focus on the effect of an additive on a lipid mixture, which usually induces shifts in demixing points, thus stabilizing or destabilizing the phase-segregated state. Results found for different model membranes are summarized, detailing the ways in which phase segregation and the generated patterns may be modulated. We focus on which are, from our viewpoint, the most relevant regulating factors affecting the surface texture observed in model membranes. This article is part of a Special Issue entitled: Emergence of Complex Behavior in Biomembranes edited by Marjorie Longo.

Fractal analysis of lateral movement in biomembranes

Lateral movement of a molecule in a biomembrane containing small compartments (0.23-μm diameter) and large ones (0.75 μm) is analyzed using a fractal description of its walk. The early time dependence of the mean square displacement varies from linear due to the contribution of ballistic motion. In small compartments, walking molecules do not have sufficient time or space to develop an asymptotic relation and the diffusion coefficient deduced from the experimental records is lower than that measured without restrictions. The model makes it possible to deduce the molecule step parameters, namely the step length and time, from data concerning confined and unrestricted diffusion coefficients. This is also possible using experimental results for sub-diffusive transport. The transition from normal to anomalous diffusion does not affect the molecule step parameters. The experimental literature data on molecular trajectories recorded at a high time resolution appear to confirm the modeled value of the mean free path length of DOPE for Brownian and anomalous diffusion. Although the step length and time give the proper values of diffusion coefficient, the DOPE speed calculated as their quotient is several orders of magnitude lower than the thermal speed. This is interpreted as a result of intermolecular interactions, as confirmed by lateral diffusion of other molecules in different membranes. The molecule step parameters are then utilized to analyze the problem of multiple visits in small compartments. The modeling of the diffusion exponent results in a smooth transition to normal diffusion on entering a large compartment, as observed in experiments.

Effects of soft interactions and bound mobility on diffusion in crowded environments: a model of sticky and slippery obstacles

Crowded environments modify the diffusion of macromolecules, generally slowing their movement and inducing transient anomalous subdiffusion. The presence of obstacles also modifies the kinetics and equilibrium behavior of tracers. While previous theoretical studies of particle diffusion have typically assumed either impenetrable obstacles or binding interactions that immobilize the particle, in many cellular contexts bound particles remain mobile. Examples include membrane proteins or lipids with some entry and diffusion within lipid domains and proteins that can enter into membraneless organelles or compartments such as the nucleolus. Using a lattice model, we studied the diffusive movement of tracer particles which bind to soft obstacles, allowing tracers and obstacles to occupy the same lattice site. For sticky obstacles, bound tracer particles are immobile, while for slippery obstacles, bound tracers can hop without penalty to adjacent obstacles. In both models, binding significantly alters tracer motion. The type and degree of motion while bound is a key determinant of the tracer mobility: slippery obstacles can allow nearly unhindered diffusion, even at high obstacle filling fraction. To mimic compartmentalization in a cell, we examined how obstacle size and a range of bound diffusion coefficients affect tracer dynamics. The behavior of the model is similar in two and three spatial dimensions. Our work has implications for protein movement and interactions within cells.

Dipolar interactions between domains in lipid monolayers at the air-water interface

A great variety of biologically relevant monolayers present phase coexistence characterized by domains formed by lipids in an ordered phase state dispersed in a continuous, disordered phase. From the difference in surface densities between these phases, inter-domain dipolar interactions arise. These interactions are relevant for the determination of the spacial distribution of domains as well as their dynamics. In this work, we propose a novel way of estimating the dipolar repulsion using a passive method that involves the analysis of images of the monolayer with phase coexistence. This method is based on the comparison of the pair correlation function obtained from experiments with that obtained from Brownian dynamics simulations of a model system. As an example, we determined the difference in dipolar density of a binary monolayer of DSPC/DMPC at the air-water interface from the analysis of the radial distribution of domains, and the results are compared with those obtained by surface potential determinations. A systematic analysis for the experimentally relevant parameter range is given, which may be used as a working curve for obtaining the dipolar repulsion in different systems.

Non-Brownian diffusion in lipid membranes: Experiments and simulations

The dynamics of constituents and the surface response of cellular membranes-also in connection to the binding of various particles and macromolecules to the membrane-are still a matter of controversy in the membrane biophysics community, particularly with respect to crowded membranes of living biological cells. We here put into perspective recent single particle tracking experiments in the plasma membranes of living cells and supercomputing studies of lipid bilayer model membranes with and without protein crowding. Special emphasis is put on the observation of anomalous, non-Brownian diffusion of both lipid molecules and proteins embedded in the lipid bilayer. While single component, pure lipid bilayers in simulations exhibit only transient anomalous diffusion of lipid molecules on nanosecond time scales, the persistence of anomalous diffusion becomes significantly longer ranged on the addition of disorder-through the addition of cholesterol or proteins-and on passing of the membrane lipids to the gel phase. Concurrently, experiments demonstrate the anomalous diffusion of membrane embedded proteins up to macroscopic time scales in the minute time range. Particular emphasis will be put on the physical character of the anomalous diffusion, in particular, the occurrence of ageing observed in the experiments-the effective diffusivity of the measured particles is a decreasing function of time. Moreover, we present results for the time dependent local scaling exponent of the mean squared displacement of the monitored particles. Recent results finding deviations from the commonly assumed Gaussian diffusion patterns in protein crowded membranes are reported. The properties of the displacement autocorrelation function of the lipid molecules are discussed in the light of their appropriate physical anomalous diffusion models, both for non-crowded and crowded membranes. In the last part of this review we address the upcoming field of membrane distortion by elongated membrane-binding particles. We discuss how membrane compartmentalisation and the particle-membrane binding energy may impact the dynamics and response of lipid membranes. This article is part of a Special Issue entitled: Biosimulations edited by Ilpo Vattulainen and Tomasz Róg.

Stiffness of lipid monolayers with phase coexistence

The surface dilational modulus--or compressibility modulus--has been previously studied for monolayers composed of pure materials, where a jump in this modulus was related with the onset of percolation as a result of the establishment of a connected structure at the molecular level. In this work, we focused on monolayers composed of two components of low lateral miscibility. Our aim was to investigate the compressibility of mixed monolayers at pressures and compositions in the two-phase region of the phase diagram, in order to analyze the effect of the mechanical properties of each phase on the stiffness of the composite. In nine different systems with distinct molecular dipoles and charges, the stiffness of each phase and the texture at the plane of the monolayer were studied. In this way, we were able to analyze the general compressibility of two-phase lipid monolayers, regardless of the properties of their constituent parts. The results are discussed in the light of the following two hypotheses: first, the stiffness of the composite could be dominated by the stiffness of each phase as a weighted sum according to the percentage of each phase area, regardless of the distribution of the phases in the plane of the monolayer. Alternatively, the stiffness of the composite could be dominated by the mechanical properties of the continuous phase. Our results were better explained by this latter proposal, as in all the analyzed mixtures it was found that the mechanical properties of the percolating phase were the determining factors. The value of the compression modulus was closer to the value of the connected phase than to that of the dispersed phase, indicating that the bidimensional composites displayed mechanical properties that were related to the properties of each phases in a rather complex manner.

Electrostatic field effects on membrane domain segregation and on lateral diffusion

Natural membranes are organized structures of neutral and charged molecules bearing dipole moments which generate local non-homogeneous electric fields. When subjected to such fields, the molecules experience net forces that can modify the lipid and protein organization, thus modulating cell activities and influencing (or even dominating) the biological functions. The energetics of electrostatic interactions in membranes is a long-range effect which can vary over distance within r^-1 to r^-3. In the case of a dipole interacting with a plane of dipoles, e.g. a protein interacting with a lipid domain, the interaction is stronger than two punctual dipoles and depends on the size of the domain. In this article, we review several contributions on how electrostatic interactions in the membrane plane can modulate the phase behavior, surface topography and mechanical properties in monolayers and bilayers.

Rheological properties of a two phase lipid monolayer at the air/water interface: effect of the composition of the mixture

Many biologically relevant monolayers show coexistence of discrete domains of a long-range ordered condensed phase dispersed in a continuous, disordered, liquid-expanded phase. In this work, we determined the viscous and elastic components of the compressibility modulus and the shear viscosity of monolayers exhibiting phase coexistence with the aim at elucidating the contribution of each phase to the observed monolayer mechanical properties. To this purpose, mixed monolayers with different proportions of distearoylphosphatidylcholine (DSPC) and dimyristoylphosphatidylcholine (DMPC) were prepared and their rheological properties were analyzed. The relationship between the phase diagram of the mixture at 10 mN m(-1) and the rheological properties was studied. We found that the monolayer shear viscosity is highly dependent on the presence of domains and on the domain density. In turn, the monolayer compressibility is only influenced by the presence of domains for high domain densities. For monolayers that look homogeneous on the micrometer scale (DSPC amount lower that 23 mol %), all the analyzed rheological properties remain similar to those observed for pure DMPC monolayers, indicating that in this proportion range the DSPC molecules contribute as DMPC to the surface rheology in spite of having hydrocarbon chains four carbons longer.

Diffusion in a fluid membrane with a flexible cortical cytoskeleton

We calculate the influence of a flexible network of long-chain proteins, which is anchored to a fluid membrane, on protein diffusion in this membrane. This is a model for the cortical cytoskeleton and the lipid bilayer of the red blood cell, which we apply to predict the influence of the cytoskeleton on the diffusion coefficient of a mobile band 3 protein. Using the pressure field that the cytoskeleton exerts on the membrane, from the steric repulsion between the diffusing protein and the cytoskeletal filaments, we define a potential landscape for the diffusion within the bilayer. We study the changes to the diffusion coefficient on removal of one type of anchor proteins, e.g., in several hemolytic anemias, as well as for isotropic and anisotropic stretching of the cytoskeleton. We predict an overall increase of the diffusion for a smaller number of anchor proteins and increased diffusion for anisotropic stretching in the direction of the stretch, because of the decrease in the spatial frequency as well as in the height of the potential barriers.