Solid-Supported Lipid Membranes: Formation, Stability and Applications

This thesis presents a comprehensive investigation of the formation of supported lipid membranes with vesicle hemifusion, their stability under detergents and organic solvents and their applications in molecular biology.

In Chapter 3, we describe how isolated patches of DOPC bilayers supported on glass surfaces are dissolved by various detergents (decyl maltoside, dodecyl maltoside, CHAPS, CTAB, SDS, TritonX-100 and Tween20) at their CMC, as investigated by fluorescence video microscopy. In general, detergents partition into distal leafets of bilayers and lead to the expansion of the bilayers through a rolling motion of the distal over the proximal leaflets, in agreement with the first stage of the established 3-stage model of lipid vesicle solubilization by detergents. However, solubilization pathways are different for different detergents. For decyl/dodecyl maltoside, CHAPS and SDS, dissolution of bilayers starts only once the bilayers reach a critical lipid:detergent ratio in the bilayers, whereas for CTAB, TritonX and Tween20, dissolution starts once detergents partition into the bilayers. From the expansion dynamics of bilayers, we conclude that the energy loss due to the partitioning of detergents is balanced by the released heat and the increment of the planar bilayer's curvature energy. Moreover, in contrast to a previous view that detergents need to be present on both leaflets of bilayers to initiate solubilization, we find that detergents in the distal leaflets of bilayers are sufficient to induce micellization. Lastly, we estimate the energy barrier for detergents to partition into lipid bilayers and find it is on the order of 3 to 7 kT.

Subsequently, we study the partitioning of organic solvents (methanol, ethanol, isopropanol, propanol, acetone and chloroform) into isolated bilayer patches on glass in Chapter 4 with fluorescence microscopy. The area expansion of bilayers due to the partitioning of organic solvents is measured. From the titration of organic solvents, we measured the rate of area expansion as a function of the volume fraction of organic solvents, which is proposed to be a measure of strength of interactions between solvents and membranes. From the same experiments, we also measure the maximum expansion of bilayers (or the maximum binding stoichiometry between organic solvents and lipids) before structural breakdown, which depends on the depth of penetration of solvents to the membranes. From the partitioning dynamics, it is observed that bilayers expand the area through the same rolling motion as described for detergent partitioning in Chapter 3. The energy loss due to the partitioning of organic solvents was balanced by the released heat and the increment of planar bilayer's curvature energy. Upon desorption of organic solvents, the bilayers shrink the area through the edges and by forming pores within the bilayers, concomitantly increase the contour length of bilayer edges, indicating self-healing is impossible for this system.

In Chapter 5, we investigate the formation of sparsely-tethered bilayer lipid membranes (stBLMs) with vesicle hemifusion. In vesicle hemifusion, lipid vesicles in contact with a hydrophobic alkyl-terminated self-assembled monolayer (SAM) deposit a lipid monolayer to the SAM surface, thus completing the bilayer. Electrical Impedance Spectroscopy and Neutron reflectivity are used to probe the integrity of stBLMs in terms of their insulating and structural properties. Preparation conditions are screened for those that are optimal for stBLM formation. Concentrations of lipid vesicles, hydrophobicity of SAMs, the presence of calcium and high concentrations of salt are identified as the key parameters. We show that stBLMs can be formed with vesicles of different compositions. Vesicle hemifusion opens up a new route in preserving the chemical compositions of stBLMs and facilitating membrane proteins incorporation.

In Chapter 6, we visualize the hemifusion pathway of giant unilamellar vesicles (GUVs) with planar hydrophobic surfaces at the single vesicle level with fluorescence video microscopy. When a GUV hemifuses to a surface, its outer leaflet breaks apart and remains connected to the surface presumably through a hemifusion diaphragm. Lipids from the outer leaflet are transferred to the surface as a lipid monolayer that expands radially outward from the hemifusion diaphragm, thereby forming the loosely packed outer hemifusion zone. The tension of the outer leaflet rises as lipids are transported to the outer hemifusion zone until it is large enough to rupture the GUV. Therefore, a pore forms near the hemifusion diaphragm through which the water encapsulated within the GUV is expelled. Additionally, lipids ipped from the inner to outer leaflet via the pore and are transferred to the surface. The hemifusion of lipids from both leaflets to the surface leads to the formation of the inner hemifusion zone densely packed with lipids. The inner and outer hemifusion zones expand radially as concentric discs with a rate of about 1000 m2=s, suggesting the expansion is driven by surface hydrophobicity. The spreading dynamics of the lipid monolayer is consistent with a model where the energy dissipated by friction between the monolayer and the surface is equal to the difference of surface energy when the surface is covered by a lipid monolayer. The mechanism revealed in this work provides insights into membrane reorganization and improves the understanding of vesicle-surface interactions.

The tethering of membranes is a common process regulating membrane fusion throughout the secretory pathway. The Prof. Linstedt laboratory at the CMU Department of Biological Sciences has been focusing on the elucidation of the mechanism of Golgi membrane tethers, GRASP (Golgi Reassembly And Stacking Proteins), which are essential for the formation of the characteristic ribbon structure of the Golgi apparatus. Typically, such work involves in vivo expression of these proteins which tracks how specific mutations affect organelle morphology. However, it remains unclear in such investigations, whether a certain active protein is sufficient to trigger the formation of the characteristic structure or if other factors, for example auxiliary proteins, are involved. Therefore, in Chapter 7, we develop an in vitro assay employing stBLMs and lipid vesicles to examine the functionality of GRASP in membrane tethering. Membrane-bound GRASP on opposing membranes dimerizes and tethers fluorescently-labeled vesicles to stBLMs. The fluorescence intensity of images taken at stBLM surfaces is used to quantify the tethering activity. Both wild type and mutant proteins were studied to shed light on the molecular mechanism of tethering. We show that the GRASP domain is sufficient and necessary for membrane tethering. In addition, the tethering capability of GRASP is impaired when the internal ligands and the binding pockets participating in dimerization are deleted and mutated. Membrane anchors, sizes of vesicles and membrane compositions are explored for their influence on the outcomes of the assay. Furthermore, preliminary analysis from neutron reflectivity measurements shows that both the internal ligands and binding pockets are exposed instead of buried toward the membrane surface. In summary, we establish a functional assay for studying GRASP activity in vitro. This assay may also be used for studies of similar supramolecular structure formation processes in molecular biology.