Structure and Elasticity of Liquid Ordered / Liquid Disordered Domains

Biological membranes regulate diverse processes on the cellular level, including material transport, signal transduction and various chemical reactions. An about 4 nm thin layer of phospholipids and proteins is the central element of these membranes and there is ample evidence that lipids and proteins mutually regulate their function. Of particular interest is the organization of lipids and proteins into functional platforms, also known as rafts, to perform specific signaling or transport tasks. Sphingomyelin and cholesterol are considered to be the major lipid components of these rafts. Clearly, basic physicochemical understanding of raft properties is of need for ground- breaking developments in human health care. Biophysical studies on complex mixtures of "raft lipids" can contribute significantly to such insights. These mixtures show phase separation into coexisting liquid ordered (L o ) and liquid disordered (L a) domains. Lo domains show several features of membrane rafts, including enrichment in cholesterol and capability of glycophosphatidylinositol (GPI)-anchored proteins to insert into these lipid environments. However, the recently discovered exclusion of raft-associating transmembrane proteins (TMP) from Lo domains questions their applicability as models for membrane rafts. The aim of the project is to derive lipid-only models for membrane rafts that recover the partitioning of raft-associating TMP, thus creating improved and experimentally stringently controlled lipid-only models of membrane rafts. The goal will be achieved by deriving those structural and elastic properties of Lo and La domains that couple to partitioning of TMP using a broad selection of biophysical techniques. In particular, we will determine lateral packing density (area per lipid), membrane thickness, spontaneous curvature, bending rigidity and Gaussian curvature modulus for several lipid mixtures, from which we will calculate the preferential insertion of protein models into a given lipid environment. This way we will understand why TMP proteins do not partition into current lipid-only raft models and how to tune Lo phase properties such that TMP partition into these environments. We will test the optimized Lo phases by measuring the partitioning of a well-established peptide model for TMP in comparison to its behavior in raft-like structures of cell-derived vesicles. Work will be performed in collaboration with several key experts in the fields of membrane biophysics and cell biology. Our results will the basis for several follow-up applications and will thus be an important contribution to close the gap between membrane biophysics and cell biology.

Cell membranes are complex composites of lipids, proteins and carbohydrates and play a pivotal role in cell physiology (e.g. compartmentalization of diverse biochemical processes, cell-cell communication, etc.). The hierarchical and dynamical structure of cell membranes is of particular importance to fulfill these functions. A current hypothesis of membrane structure claims that nanoscopically small and dynamic domains (termed membrane rafts) facilitate the assembly of certain proteins needed for e.g. signaling events or material transport through the membrane. However, direct experimental evidence for the existence of such rafts in live cells has been, due to the complexity of the system, elusive so far. Artificial membranes, in turn, composed of lipids only and in particular mixtures thereof are well known to exhibit under certain conditions domains. The aim of the project was to study physical properties of such domains in detail in order to gain insight into their role in determining preferred protein partitioning into a given lipid environment, deemed to be of high physiological importance for cellular function. To this end we developed x-ray and neutron scattering techniques, which enabled us to determine domain physical parameters under in-situ conditions, i.e. simultaneously, when they are coexisting. This way we were able to explore the nanoscopic structural properties of the domains which gave us several interesting insights. For example, we found that rigid domains lose their thickness upon heating, while coexisting soft domains become increasingly thick, up to the point where both domains mix into a homogenous bilayer. This temperature behavior can be understood in terms of a lipid exchange between the domains by diffusion. Another interesting feature of such systems is that domains align in multibilayer systems, such that like domains are located on top of each other. This is surprising, because membranes are separated by a layer of water. Our measurements showed that there is a delicate balance of repulsive and attractive interactions, which is responsible for this behavior. In particular we found fluctuation-driven repulsion to be significantly different between soft and rigid domains. Finally, the obtained physical parameters of both domains allowed us to derive the preferred partitioning of transmembrane proteins in dependence of their overall shape. Convex-shaped proteins were found to be preferentially located in soft domains, while concave proteins would diffuse to rigid domains, respectively. Intriguingly calculations suggested that proteins with an asymmetric shape would prefer asymmetric lipid bilayers. This correlates with membrane asymmetry found in natural systems, a feature which is needed for physiological function, and therefore might be a significant lead as to the understanding the origin membrane asymmetry.