Analyzing protein dynamics from fluorescence intensity traces using unsupervised deep learning network

We propose an unsupervised deep learning network to analyze the dynamics of membrane proteins from the fluorescence intensity traces. This system was trained in an unsupervised manner with the raw experimental time traces and synthesized ones, so neither predefined state number nor pre-labelling were required. With the bidirectional Long Short-Term Memory (biLSTM) networks as the hidden layers, both the past and future context can be used fully to improve the prediction results and can even extract information from the noise distribution. The method was validated with the synthetic dataset and the experimental dataset of monomeric fluorophore Cy5, and then applied to extract the membrane protein interaction dynamics from experimental data successfully.

Modeling membrane nanotube morphology: the role of heterogeneity in composition and material properties

Membrane nanotubes have been identified as dynamic structures for cells to connect over long distances. Nanotubes typically appear as thin and cylindrical tubes, but they may also have a beaded architecture along the tube. In this paper, we study the role of membrane mechanics in governing the architecture of these tubes and show that the formation of beadlike structures along the nanotubes can result from local heterogeneities in the membrane either due to protein aggregation or due to membrane composition. We present numerical results that predict how membrane properties, protein density, and local tension compete to create a phase space that governs the morphology of a nanotube. We also find that there is an energy barrier that prevents two beads from fusing. These results suggest that the membrane-protein interaction, membrane composition, and membrane tension closely govern the tube radius, number of beads, and the bead morphology.

Effect of the membrane/protein interaction nature on adsorptive properties of ultrafiltration membranes based on aromatic poly- and copolyamides

Membranes are a science intensive product of interindustry use, without which a breakthrough development of basic and high-technology sectors of economy, development of science as well as effective solution of important goals of the social sphere and problems of environment protections are impossible. Active development of medicine, pharmaceutical industry and biotechnology in the recent years contributed to the growth of scientific and commercial interest in ultrafiltration. The most important characteristic of ultrafiltration membranes is selectivity. According to existing views the basis of the separation mechanism implemented in ultrafiltration is size selectivity. However, such phenomena as concentration polarization and adsorption play a very important role in separation. Investigation of these phenomena is of significant practical interest, because its results in many ways determine the choice of the membrane media, conditions of its recovery, the mode and conditions of filtration. Adsorption of proteins on the surface of porous membranes is a rather complicated process due to the individual mechanism implemented in each specific case. Therefore, the issue of the contribution of various types of interactions into the process of adsorption still remains controversial. To determine the nature of membrane/protein interaction and clarify the factors of its control, the present work involved investigation of sorption of several proteins on a number of synthesized porous membranes based on aromatic polyamides noted for the presence, nature and concentration of ionogenic groups. The investigations were carried out using bovine serum albumin, lysozyme of hen's eggs, myoglobin and bacitracin. Adsorption of proteins by the membranes was investigated in the static mode. The protein concentration was determined using the SF-2000 spectrophotometer (experimental-design bureau Spektr) by optical density at the wave length λ=278 nm. To mathematically process the experimental data, the two-parameter models by Langmuir, Freundlich, Temkin and the three-parameter model by Langmuir-Freundlich were used. To assess the compliance degree of the experimental data to the selected mathematical models, the values of the coefficient of determination (R2) and sum squares errors (SSE) were used. It has been shown that in the case of the presence of charge in protein macromolecules and the membrane surface the role of electrostatic forces is dominant in the protein adsorption mechanism, however, the contribution of non-electrostatic interactions in the investigated membrane/protein systems is significant.

Gaussian curvature directs the distribution of spontaneous curvature on bilayer membrane necks

ABSTRACTFormation of membrane necks is crucial for fission and fusion in lipid bilayers. In this work, we seek to answer the following fundamental question: what is the relationship between protein-induced spontaneous mean curvature and the Gaussian curvature at a membrane neck? Using an augmented Helfrich model for lipid bilayers to include membrane-protein interaction, we solve the shape equation on catenoids to find the field of spontaneous curvature that satisfies mechanical equilibrium of membrane necks. In this case, the shape equation reduces to a variable coefficient Helmholtz equation for spontaneous curvature, where the source term is proportional to the Gaussian curvature. We show how this latter quantity is responsible for non-uniform distribution of spontaneous curvature in minimal surfaces. We then explore the energetics of catenoids with different spontaneous curvature boundary conditions and geometric asymmetries to show how heterogeneities in spontaneous curvature distribution can couple with Gaussian curvature to result in membrane necks of different geometries.

Design principles for robust vesiculation in clathrin-mediated endocytosis

A critical step in cellular-trafficking pathways is the budding of membranes by protein coats, which recent experiments have demonstrated can be inhibited by elevated membrane tension. The robustness of processes like clathrin-mediated endocytosis (CME) across a diverse range of organisms and mechanical environments suggests that the protein machinery in this process has evolved to take advantage of some set of physical design principles to ensure robust vesiculation against opposing forces like membrane tension. Using a theoretical model for membrane mechanics and membrane protein interaction, we have systematically investigated the influence of membrane rigidity, curvature induced by the protein coat, area covered by the protein coat, membrane tension, and force from actin polymerization on bud formation. Under low tension, the membrane smoothly evolves from a flat to budded morphology as the coat area or spontaneous curvature increases, whereas the membrane remains essentially flat at high tensions. At intermediate, physiologically relevant, tensions, the membrane undergoes a “snap-through instability” in which small changes in the coat area, spontaneous curvature or membrane tension cause the membrane to “snap” from an open, U-shape to a closed bud. This instability can be smoothed out by increasing the bending rigidity of the coat, allowing for successful budding at higher membrane tensions. Additionally, applied force from actin polymerization can bypass the instability by inducing a smooth transition from an open to a closed bud. Finally, a combination of increased coat rigidity and force from actin polymerization enables robust vesiculation even at high membrane tensions.