Understanding and Manipulating Channels and Pores
The investigation of channels and pores is a delightfully varied field requiring a wide range of knowledge and experimental tools taken from a multiplicity of scientific disciplines. This wonderful diversity is well illustrated by the articles in this special thematic issue of Molecular BioSystems, entitled “Understanding and Manipulating Channels and Pores”. All the work described here involves membrane proteins, yet it has required expertise in protein expression and purification, protein chemistry, cell biology, transgenic animals, synthetic chemistry, surface chemistry, photochemistry, materials science, engineering, biophysics, and more.Like most areas of science, major advances in understanding transmembrane channels and pores have come through the invention and development of new technologies. Derek Trezise and colleagues (p. 444) present an early experimental evaluation of population patch clamp (PPC) electrophysiology, which they regard as a breakthrough technology for screening various voltage-gated and certain non voltage-gated ion channels. In PPC, which was released by Molecular Devices in 2006, ionic currents are recorded simultaneously under voltage clamp from 384 different populations of up to sixty-four tissue culture cells overexpressing a channel of interest. In a previous manifestation, recordings were made from an array of single cells, in which case many of the wells had to be discounted because of poor seals or failure of a cell to express the protein under investigation. And still today, university research groups and many pharmaceutical companies use individual patch clamp assays, which are very slow indeed. PPC by contrast is semi-automated and offers better precision by averaging over multiple cells. Although costly, it is cheaper than single-cell arrays and allows the screening of several thousand compounds a week. This is a large number, so far as membrane proteins are concerned. Therefore, as well as facilitating the discovery of new therapeutic agents, PPC and related technologies will be crucial in screening for unwanted side effect of drugs that are not intended to act on ion channels, a problem that has prevented the advance of many drugs in clinical trials and led to withdrawals from the marketplace.
In their review (p. 453), Mancia and Hendrickson describe the heterologous expression of G-protein coupled receptors (GPCRs), with a view to structural studies. This work originally appeared as a chapter in the volume ‘Structural Biology of Membrane Proteins’ in the RSC Biomolecular Sciences series www.rsc.org/publishing/books/bss/index.asp. Membrane proteins comprise ∼25% of the proteins encoded in both prokaryotic and eukaryotic genomes, yet they represent only a small fraction of proteins with known three-dimensional structures. The difficulty of structure determination by X-ray crystallography arises from the relatively low abundance of membrane proteins in the cell, their need for a membrane-like environment and, in the case of mammalian membrane proteins, their instability. Ion channels often have stable homologs in bacteria, but GPCRs do not and besides we would like to know structures of the “real thing”. The authors report the state of the art of GPCR expression in both bacteria and eukaryotic cells. For example, 3 mg of functional rat neurotensin receptor can currently be purified to homogeneity from a 20 L culture of E. coli; very little by the standards of soluble proteins, but a notable achievement. Certain GPCRs can be obtained in functional form after refolding. For example, a human leukotriene receptor was obtained from E. coli in inclusion bodies and refolded in detergent. Bovine rhodopsin has been expressed at up to 10 mg L−1 in human embryonic kidney 293 (HEK-293) cells, but this is certainly the exception rather than the rule. Further, although GPCRs capable of binding their ligands have been expressed successfully by several groups, no laboratory has yet been able to produce crystals that diffract well. Indeed MacKinnon's structure of a rat voltage-gated potassium channel obtained from the yeast Pichia pastoris is the only crystal structure of a mammalian membrane protein obtained by heterologous expression.
Membrane proteomics aspires to identify the membrane protein constituents of cells and their posttranslational modifications, as well protein–protein interactions in membranes, and changes in expression and modification in different cellular states, including disease. Macher and Yen (p. 435) articulate the very real challenges presented in studying water-insoluble proteins that are distributed between the plasma membrane and a variety of organelles, usually in low overall abundance even though they represent a large fraction of protein-encoding genes. The authors focus on “gel-free” technologies. They first describe efforts to enrich for membrane proteins from homogenates. The review focuses on plasma membrane proteins, which are the easiest targets as they can be enriched by relying on the fact that they are often glycosylated. The use of lectin-affinity techniques (the authors' area of expertise) and selective chemical modification of the oligosaccharide appendages are described. The authors also describe general approaches such as cell-surface biotinylation, followed adsorption with streptavidin reagents. The clean separation of membrane proteins from organelles presents a greater challenge that has not yet been met. After enrichment, the most powerful of modern approaches for the identification of proteins and their modifications is mass spectrometry. This in turn requires enzymatic digestion of the target proteins into fragments amenable to high-resolution analysis. But membrane protein, for example multipass proteins with small extramembraneous loops, can be hard to digest. The use of organic solvents and detergents for extraction and subsequent digestion by enzymes are described, as well as chemical cleavages that can be used in denaturing environments.
While powerful analytical tools such as mass spectrometry can be used to analyze native biological membranes, quite different techniques are needed for the functional analysis of individual membrane proteins and small assemblies of them. The use of tethered bilayer lipid membranes (tBLM) is an emerging approach of great promise, described here by Ingo Köper (p. 381). Alternative model systems have severe disadvantages. For example, in the case of lipid vesicles, there is poor access to one side of the bilayer. Planar lipid bilayers are unstable, while conventional solid-supported bilayers offer only a thin water layer at one face, prohibiting the incorporation of many membrane proteins and providing a tiny ionic reservoir. In tBLM, the inner leaflet is in part covalently attached to the solid surface and a spacer arm dictates the volume of the reservoir between the support and the bilayer. Köper describes conditions for success in this potentially treacherous field. Recent work has led to the development of stable bilayers with high electrical resistance for the investigation of ion channel function, requiring, for example, a toolbox of synthetic anchor lipids for tethering and a reduction in the surface roughness of the underlying gold or oxide support. Applications of tBLM in biosensing can be envisioned, especially if the high-resistance, large-reservoir bilayers can be further enhanced for reliable single-channel recording, which would allow stochastic detection. However, the incorporation of proteins with large internal extramembraneous domains still cannot be taken for granted and we can expect further advances in this area as indicated by the author.
The ability to manipulate the structures of ion channels and pores is permitting a better understanding of how these proteins work and providing tools for many applications in biotechnology. One approach is to take a biological channel or pore of known three-dimensional structure and “engineer” it by mutagenesis or by targeted chemical modification. Another approach is to synthesize these structures from scratch as described here by Sakai, Mareda and Matile (p. 388). The authors are sufficiently modest and scholarly enough to begin with highlights from other research groups, which they follow with striking examples from their own laboratory in Geneva. Synthetic pores, formed in lipid bilayers from an astonishing variety of structurally diverse constituent molecules (their Fig. 1), exhibit properties associated with biological channels and pores, including a defined unitary conductance, ion selectivity, voltage gating and susceptibility to open-channel blockers. By using known chemistry, it is also possible to build-in features such as light sensitivity. The Geneva lab has focused on rigid-rod scaffolds as a basis for synthetic pores. They make a characteristically provocative statement “Somehow the antithesis to foldamers, these simple rods bypass all folding problems because they do not fold. Interestingly, rigid, i.e. unbendable, incompressible, unfoldable and ununfoldable rods do not exist in biology. They are, however, common and useful in the materials sciences.” With their original oligophenyl rods, the authors have prepared pores that are ion selective, and pores that gate in response to the applied potential or chemical stimulation by pH change, ions, small molecules, and macromolecules, as well as pores with catalytic activity. They also show that the use of pores as optical transducers of chemical reactions (for example, the hydrolysis of a blocker can lead to dye release from a phospholipid vesicle) has potential applications in drug discovery and molecular sensing. New oligophenyl structures include the formation of a transmembrane structure with an internal π stack, rather than a central channel, which can be used to generate photosynthetic activity. The authors have recently introduced new rod structures and the possibilities for this large family of synthetic channels and pores appears to be inexhaustible. All that is lacking is hard structural evidence for the proposed transmembrane structures, which will perhaps be forthcoming from new NMR techniques.
Another, equally remarkable, but quite different development in synthetic pores are the conical nanopores produced by Martin and his collaborators (p 397). These structures are formed in variety of robust polymer sheets by using a track-etch process. The polymer is first bombarded with high energy particles to produce ion tracks. In a typical example, poly(ethylene terephthalate), the tracks can be converted into pores by treatment with strong base, which etches the polymer by hydrolyzing ester bonds. But the polymer sheets are 5 to 10 μm thick, and therefore an anisotropic etching procedure is crucial for producing conical nanopores in which the properties of the high-resistance narrow end mimic those of a biological pore. Martin and colleagues describe how the procedures for etching have been painstakingly optimized so that single nanopores with reproducible dimensions and geometry can be prepared. Considerations included the composition of the etching solution and the stop solution, manipulation of the temperature and applied potential, and the use of cosolvents. The spectacular results are documented in beautiful electron micrographs of replicas of the nanopores. The authors go on to describe how functions characteristic of biological pores, such as voltage-gating, can be introduced into the nanopores by chemical modifications of the internal surface. Prototype sensors derived from single conical nanopores are also described. Small molecules or macromolecules, such as DNA, can be detected in a “resistive pulse” mode: that is the conductance of a pore changes as an analyte molecule moves through it. Proteins could be detected after modification of the internal surface of a conical nanopore with small ligands or binding proteins such as antibodies. The nanopore surface is first plated with a thin layer of gold to permit thiol-attachment chemistry. In this case, analyte binding leads to permanent blockage of the pore. Just as in the case of synthetic pores, there remains a vast unexplored territory.
The final paper in our special issue illustrates the extraordinary power of targeted protein modification; that is, modification a specific site dictated by site-directed mutagenesis. Gorostiza and Isacoff describe ion channels and pores with optical switches and triggers (p 416). By using molecular modeling, mutagenesis, chemical synthesis and a refined knowledge of ion channel biophysics, the authors and their collaborators at Berkeley have produced both voltage- and ligand-gated ion channels that can be switched on with light and off with light of a different wavelength. The on–off response is rapid and can be spatially controlled, potentially at the submicron level. The authors explore the history of the approach and outline all its variants (dubbed nanokeys, nanotoggles and nanotweezers) in their delightful Fig. 1. They show, that in combination with fluorescence detection of cell signaling, their approach allows non-invasive, all-optical experiments on cells and tissues. They go on to describe an experimental tour de force in which a behavior, the escape response to touch of chemically-modified transgenic zebrafish was controlled by light in a reversible manner. You can see videos of these stunning experiments in the supplementary data for Neuron 54, 535 (2007), and then find out here about how it was done. It is possible that photoswitchable ion channels can be further developed for controllable therapies of human neurological disease in combination with gene therapy, and employed in other areas such as the delivery of encapsulated drugs with remotely controlled release.
I do hope you will flip through all the papers in this issue. I am certain that you will quickly be drawn in and want to read in full about these stimulating new developments in membrane biology.
| This journal is © The Royal Society of Chemistry 2007 |