Session A10: Membranes I

The dynamic softening and intrinsic stiffening effect of cholesterol on membranes

Small patches of the plasma membrane of the cell must be reshaped from a nearly flat geometry into a sphere (vesicle) as part of the normal transport of proteins and lipids inside as well as out of the cell. The key parameter characterizing how the membrane resists this deformation is the bending modulus, the force constant for membrane curvature deformations. 

The plasma membrane of cells is made up of hundreds of different types of lipids. Each of these lipids has a varied effect on membrane mechanical properties, including the bending modulus. Cholesterol is both the most abundant single lipid species as well as the most unusual, having a rigid ring structure that, consistent with intuition, often does strongly perturb the mechanics of simple model membranes.

Molecular simulation evidence will be presented demonstrating that the effect of cholesterol varies with the surrounding composition of the lipid matrix. The key parameter appears to be its preference for concave membrane leaflet curvature. Depending on how strongly cholesterol couples to local curvature, it may have a dramatic softening effect at long timescales, and/or a strong stiffening effect at short timescales. This complex behavior may be at the heart of discrepancies between techniques that measure the stiffness response at short timescales or at equilibrium.   

Cholesterol Hydrogen Bonding Interactions in Lipid Membranes: Insights from Raman Fingerprint Modes

While Raman spectroscopy has long provided insights into lipid membrane structure, progress has been largely limited to the vibrations of common functional groups within lipids. The more unique "fingerprint" vibrations of the entire molecule can be identified by comparisons to reference measurements, but a lack of detailed knowledge of their motion impairs their spectral interpretation. Here we will show that time-dependent density functional theory (TDDFT) can calculate accurate Raman spectra for molecules up to 100 atoms, including the effects of conformational variation and hydrogen bonding. A comparison between the calculated spectrum of cholesterol and measurements on purified cholesterol powder will be presented, identifying 34 vibrational modes. Furthermore, Raman peaks of cholesterol that are sensitive to hydrogen bonds at the hydroxyl group will be identified by TDDFT and confirmed with comparisons to powder and lipid vesicle spectra. Similar results on anthraquinones, flavonoids, and carotenoids will be presented. Together these results illustrate that TDDFT calculations create new opportunities for spectral interpretation in Raman spectroscopy and advance its applications in molecular structure analysis for biomolecules in their natural environment.   

Scaling relations of elastic and structural properties in cholesterol-rich lipid membranes

The elastic properties of cell membranes play a key role in cell functions. Cholesterol – an essential component of eukaryotic cell membranes– is known to cause lipid condensation or a decrease in the average area per lipid (A_L). However, the interplay between cholesterol-driven lipid packing and emergent elastic properties of lipid membranes remains largely unexplored. Theories based on mean field calculations have predicted different dependences of the membrane bending modulus, κ, and the area compressibility modulus, K_A, on A_L. Here, we examine these relations in cholesterol-rich lipid membranes with various degrees of acyl chain unsaturation using neutron spin-echo spectroscopy (NSE), solid-state ²H NMR relaxometry, and molecular dynamics simulations. We find that cholesterol stiffens all studied membranes irrespective of the degree of chain unsaturation over the accessible length and time scales, with an inverse dependence of κ and K_A on A_L. To generalize the molecular mechanisms driving this dependence, we consider normalized variables that eliminate experimental factors leading to different observed values of elastic membrane parameters. These structure-property relations inform design rules for engineered membranes and artificial cells with real-world functionalities.

    

Author not Attending

The spontaneous curvature model or Helfrich's curvature model is widely used to describe the shape changes and behavior of fluid membranes. The discretization of Helfrich's continuum model on triangulated meshes requires a mesh with a uniform distribution of vertices. The dynamic trian- gulation algorithm is the state of the art method to simulate mesh shape changes while maintaining a uniform vertex distribution. The dynamic triangulation method is used in Monte Carlo simula- tions but is not compatible with Molecular Dynamics. Here we introduce a Molecular Dynamics simulation method called dynamic area redistribution that can be used to simulate the fluid behavior of biomembranes using two-dimensional meshes with fixed connectivity. The application of an MD compatible method for fluid membranes is discussed, and a simple red blood cell composite model is proposed.   

Thermosensing through membrane mechanics

Recent years have seen enormous progress in our understanding of the molecular basis for thermosensing and mechanosensing. In particular, it has been shown that thermosensing and mechanosensing both rely crucially on ion channels that are activated by thermal or mechanical stimuli. We explore here the possibility that thermal and mechanical activation of ion channels, as well as other membrane proteins, may be coupled through the physical properties of lipid bilayer membranes. In particular, experiments have shown that key mechanical properties of lipid bilayers, such as the bilayer hydrophobic thickness and bending rigidity, change with temperature over biologically relevant temperature ranges. We use this experimental data to estimate the effect of temperature changes on lipid bilayer-dependent contributions to the transition energy between different conformational states of membrane proteins. Our calculations suggest that the elastic coupling of lipid bilayer properties and membrane protein conformational state may provide a generic physical mechanism for temperature sensing through membrane mechanics.   

Direct quantification of ligand-induced lipid and protein microdomains with distinctive signaling properties

Lipid rafts are ordered lipid domains that are enriched in saturated lipids, such as the ganglioside GM1. While lipid rafts are believed to exist in cells and to serve as signaling platforms through their enrichment in signaling components, they have not been directly observed in the plasma membrane without treatments that artificially cluster GM1 into large lattices. Here, we report that microscopic GM1-enriched domains can form, in the plasma membrane of live mammalian cells expressing the EphA2 receptor tyrosine kinase in response to its ligand ephrinA1-Fc. The GM1-enriched microdomains form concomitantly with EphA2-enriched microdomains. To gain insight into how plasma membrane heterogeneity controls signaling, we quantify the degree of EphA2 segregation and study initial EphA2 signaling steps in both EphA2-enriched and EphA2-depleted domains. By measuring dissociation constants, we demonstrate that the propensity of EphA2 to oligomerize is similar in EphA2-enriched and -depleted domains. However, surprisingly, EphA2 interacts preferentially with its downstream effector SRC in EphA2-depleted domains. The ability to induce microscopic GM1-enriched domains in live cells using a ligand for a transmembrane receptor will give us unprecedented opportunities to study the biophysical chemistry of lipid rafts.

 

    

Force-sensing in Piezo ion channels

In 2010, Piezo proteins were discovered, which has led to stunning progress in the elucidation of the molecular basis for mechanosensation. Piezo channels are mechanosensitive ion channels that locally curve the membrane into a dome shape. Membrane elasticity theory predicts that the curved shape of the Piezo dome deforms the surrounding lipid membrane into a membrane footprint, which may amplify Piezo's sensitivity to applied forces and explain the observed modulation of Piezo gating through the cytoskeleton. We directly test the membrane elasticity theory of Piezo's membrane footprint through cryo-electron tomography of lipid bilayer vesicles deformed by Piezo channels, finding quantitative agreement between observed and predicted membrane shapes, with no free parameters. On this basis, it becomes possible to deduce elastic properties of membrane proteins solely from membrane shape measurements. We thus derive a force-distortion relationship for the Piezo dome, from which we infer the Piezo dome's intrinsic radius of curvature and bending stiffness in free-standing lipid membranes mimicking cell membranes, and predict Piezo's gating curve, with no free parameters. Our results suggest that Piezo's intrinsic curvature, membrane footprint, small stiffness, and large area are the key properties of Piezo that give rise to low-threshold, high-sensitivity mechanical gating.   

The role of protein constriction in the fission of membrane tubes

Membrane remodeling, such as fusion and fission, is involved in a variety of basic, cellular processes. When unaided, the free energy barriers for such remodeling can be prohibitively high, so biological systems employ proteins as catalysts. This work investigates the influence of proteins, such as dynamin, which constrict membrane tubes in order to lower the barrier to fission. We employ self-consistent field theory and utilize the string method to find the Minimum Free Energy Path (MFEP) in order to determine the most likely pathway for the transition. During fission, the tube first partially collapses into a worm-like micelle, which then ruptures, resulting in two capped tubes. The free energy barrier to fission depends strongly on membrane tension. Simply constricting the membrane aids the initial partial collapse, however dynamin also inserts its PH domains between head groups, distorting the membrane. Our results suggest that this distortion plays a critical role in reducing the free energy barrier to fission.   

A Dynamic and Mechanical Membrane Model to Study Membrane Remodeling Kinetics

Remodeling membranes requires an induced curvature change, which is essential in multiple cellular processes including fission and fusion, exo- and endocytosis, and matrix formation in mitochondria. Most membranes prefer to be flat and bending of the membrane is thus energetically unfavorable. Therefore, a wide range of proteins facilitates curvature inducing of cellular membranes. However, the coupled kinetics of protein assembly and membrane dynamics are too fast to resolve experimentally; current models have investigated membrane remodeling dynamics in the presence of proteins, but not coupled with explicit proteins obeying reaction and diffusion dynamics. In this research, we bridge the gap between protein and membrane dynamic models with a mechanical membrane model that integrates reaction-diffusion dynamics to investigate the mechanisms of protein driven membrane remodeling. We start off with a static mechanical membrane with the limit surface method to introduce dynamics according to brownian motion, which is validated against expected membrane fluctuations. We implement flat and spherical membrane models with different boundary conditions and couple membrane dynamics with reaction-diffusion models of single membrane binding proteins. This dynamic and mechanical membrane model will ultimately provide an open source resource to the community for detailed understanding of membrane remodeling in cell biology.   

Lipid number asymmetry: The hidden dimension of mammalian plasma membranes

The transbilayer distribution of lipids in mammalian plasma membranes (PMs) is functionally important and incompletely understood. It is generally assumed that the two leaflets of lipid bilayers must contain similar numbers of phospholipids (PLs) due to the constraint that their areas must be evenly matched. Contrary to this assumption, our recent detailed lipidomics analysis of live human erythrocytes reveals a large phospholipid imbalance between PM leaflets, with the cytoplasmic leaflet possessing almost 2-fold more PLs than the exoplasmic one. This surprising finding challenges our understanding of living membrane organization and structure. Extensive atomistic simulations guided by the lipidomics data reveals that a large PL imbalance between PM leaflets can be sustained via highly asymmetric distribution of cholesterol. We confirm that cholesterol increases membrane tolerance for PL imbalances in model membranes and cells. Driven by preferential interactions with saturated lipids and its tendency to 'fill gaps', we show that cholesterol is poised for enrichment in the PM exoplasmic leaflet and confirm this prediction in live red blood cells using a novel FRET-based assay. The resulting lipid number asymmetries give rise to unique PM biophysical properties including differential permeability of the two bilayer leaflets and substantial differential stress. Thus, lipid number asymmetry of major membrane constituents presents a largely unexplored dimension of membrane organization.