Building Model Membranes with Lipids and Proteins: Dangers and Challenges

Abstract

First, it is necessary to understand why biochemists wish to reconstitute membrane proteins into lipid vesicles and nonnative systems, and indeed there are several different reasons why this path could be taken. It is important to realize that there can be multiple objectives, as the shortcomings of a particular method for one objective might be an advantage for another. Historically, reconstitution has been aimed at both structural and functional studies. Structural studies can include not only the obvious determination of structure but also more subtle aspects of structure and the interplay of environment and structure, or the dynamics of the structure. Similarly, there are different aspects of function that can be targeted, many of which are accessible only in reconstituted systems.Structural studies, such as high-resolution three-dimensional (3D) structural determination, can be conducted using 2D crystals formed in lipid membranes by using electron diffraction (e.g., Wang and Kühlbrandt 1992); however, these structures rarely compete with the structures obtained from X-ray diffraction (XRD) of 3D crystals. However, this approach can be particularly rewarding when the organization of lipids around a membrane protein is of interest. For example, initial studies of aquaporins by X-ray crystallography did not find any associated lipids (Murata et al. 2000); however, 2D crystals of aquaporin can be formed in several different lipids, and indeed in the structure obtained by electron diffraction (Gonen et al. 2005), many lipids were resolved in the crystal structure, including a few that were not in direct contact with the protein. Reconstitution for such structural studies, which can provide high quality 3D structures and information on the immediate environment, this is one possible reason to make artificial membranes. The objective is high protein density, and so reconstitutions are typically at lipid-to-protein ratio of less than 0.5 mg/mg.For functional studies, reconstitution is often necessary because membrane proteins have transport activities. So ion channels are typically reconstituted into lipid bilayers for electrophysiological or transport measurements; for example, the KcsA I. Mus-Veteau (ed.), Membrane Proteins Production for Structural Analysis,