Simulations show increased lipid interdigitation in transmembrane coupling of protein condensates

Transmembrane coupling of protein condensates via membrane-mediated interactions: A simulation study

Chain Collapse, Reduced Dielectric, and Water Release Drive Protein Phase Separation.

Biomolecular condensates represent unique microenvironments that organize intracellular biology and promote biochemical reactions. However, the biomolecular interactions driving condensate phase separation are often weak, transient, and heterogeneous. Investigating the structural biology and chemical properties of condensate interiors has therefore proven experimentally challenging, often requiring the use of perturbative probes. To overcome this challenge, we combine label-free optical scattering and vibrational spectroscopy approaches spanning ultraviolet, visible, mid-infrared, and terahertz wavelengths with deep-learning-based ensemble prediction of intrinsically disordered protein conformations. This suite of label-free approaches provides quantitative insights into protein-protein/protein-solvent interactions and the chemical properties of condensate interiors. Investigating the N-terminal domain of the RNA DEAD-box helicase 4 (DDX4), our experimental and computational results support a model of phase separation involving protein chain collapse, reduced dielectric, and water release. These molecular events are expected to enhance the strength of multivalent protein-protein interactions within condensates, creating a positive feedback loop important for condensate growth and phase separation.

Cholesterol-Dependent Structure and Dynamics of Curved Lipid Vesicles Revealed by Dry MARTINI Simulations

Cholesterol is a key regulator of membrane structure and dynamics, yet its effects on large curved vesicles under implicit-solvent coarse-grained conditions remain incompletely understood. Equilibrating large Dry MARTINI vesicles is challenging because transient membrane deformations can arise during the early stages of equilibration. Here, we developed a leaflet-specific restrained-equilibration protocol that preserves vesicle geometry while allowing local lipid relaxation. All restraints were removed before production simulations, and all reported results were obtained from unbiased trajectories. Using this protocol together with the Dry MARTINI force field and the TS2CG membrane builder, we simulated ∼50 nm DOPC vesicles containing 0–40 mol% cholesterol in three independent 20 µs production simulations for each membrane composition. Increasing cholesterol concentration produced a consistent structural reorganization of the membrane, characterized by increased membrane thickness and lipid-tail ordering, and decreased species-specific Voronoi area per lipid, lipid-tail interdigitation, solvent-accessible surface area, and vesicle shape anisotropy. Cholesterol flip-flop increased progressively with cholesterol concentration, whereas DOPC flip-flop exhibited a reproducible non-monotonic dependence with a maximum near 20 mol% cholesterol. Comparison with our previous explicit-solvent MARTINI simulations showed that the major cholesterol-dependent structural trends were preserved across both solvent representations, whereas species-specific lipid packing, lipid-tail interdigitation, and the absolute magnitude of lipid flip-flop remained sensitive to the solvent representation. Overall, Dry MAR-TINI combined with the restrained-equilibration protocol provides an efficient framework for studying large curved cholesterol-containing vesicles.

Interplay Between Cholesterol Concentration and Membrane Curvature in Liposomes Revealed by Molecular Dynamics Simulations

Liposomes are widely used as model membranes and nanoscale drug delivery systems, where cholesterol plays a key role in regulating bilayer structure and dynamics. However, how cholesterol concentration influences the structure and dynamics of liposome and how this influence is dependent on membrane curvature are not fully understood at the molecular level. In this work, coarse-grained molecular dynamics simulations using the MARTINI force field were employed to examine the concentration-dependent behavior of cholesterol in planar and curved membranes composed of cholesterol and unsaturated phospholipids, namely DOPC. More specifically, a planar lipid bilayer and an approximately 50-nm liposome were simulated to represent two extreme limits of small and large curvature, respectively. Increasing cholesterol concentration led to thicker membranes and reduced solvent exposure, consistent with cholesterol’s condensing effect. Membrane curvature enhanced interleaflet coupling and increased tail interdigitation relative to planar systems. Notably, DOPC flip-flop rate in spherical bilayers exhibited a non-monotonic dependence on cholesterol content, reflecting a balance between curvature-induced packing stress and cholesterol-driven ordering. These findings provide molecular-level insight into how cholesterol and curvature together shape the structure and dynamics of unsaturated lipid bilayers.

Membrane-Mediated Interaction of Biomolecular Condensates at Biomembranes.

Cellular organization is achieved through compartmentalization into membrane-bound organelles and biomolecular condensates, formed by liquid-liquid phase separation of biopolymers. These condensates behave as liquid droplets with interfacial tensions in the μN/m-mN/m range, and interact with lipid membranes by wetting and deforming them. Using the Helfrich Hamiltonian, triangulated interfaces and membranes, and energy minimization, we analyze the wetting-to-wrapping transition of single condensates at initially planar membranes. For a given membrane stiffness, when the ratio of condensate-cytosol interfacial tension to the membrane tension exceeds a critical value, with increasing adhesion strength, the condensates undergo multiple transitions between nonwrapped, shallow-wrapped, deep-wrapped, and complete-wetting states. The deep-wrapped state is characterized by a neck-stabilized morphology that suppresses complete wrapping; a transition to a complete-wetting state occurs at sufficiently high adhesion strength. At high ratios of the membrane tension over the condensate-cytosol interfacial tension, the deep-wrapped state vanishes at a triple point, and the condensates transition directly from shallow-wrapped to complete-wetting with increasing adhesion strength. Furthermore, we quantify membrane-mediated interactions between two partial-wrapped condensates. High membrane tension induces repulsion by reducing the adhered area at short separations, whereas low membrane tension promotes attraction via cooperative wrapping. Upon contact, the condensates fuse into a nearly spherical droplet, reflecting the dominance of interfacial tension over membrane bending. Overall, our results provide a quantitative framework for understanding condensate organization at biological membranes and guiding the design of biomolecular condensates for drug-delivery applications.

Quantifying the Energetics of Protein-Mediated Stabilization of Membrane Deformation.

Membrane curvature plays a central role in a wide range of biological processes, yet a quantitative and transferable description of its underlying energetics remains challenging. In this work, we develop a "reaction coordinate" within the Umbrella-Sampling framework to compute the free-energy cost associated with local membrane bending in both coarse-grained and atomistic simulations. The methodology is systematically validated across diverse lipid environments, including pure bilayers with varying tail unsaturation, tail length, and headgroup chemistry, where it reproduces established trends in membrane-mechanical rigidity. Extension to mixed lipid systems further captures composition-dependent modulation of bending energetics, including cholesterol-induced responses. Finally, we apply the framework to membrane-active proteins with distinct curvature-generation mechanisms and demonstrate that the method successfully captures the protein modulation of the free-energy cost of local membrane deformation. Furthermore, the method is transferable to atomistic membrane systems while maintaining reasonable computational efficiency, providing a unified approach for quantifying membrane bending energetics across lipids, proteins, and simulation scales.