Department of Biochemistry University of Oxford Department of Biochemistry
University of Oxford
South Parks Road
Oxford OX1 3QU

Tel: +44 (0)1865 613200
Fax: +44 (0)1865 613201
Header picture

Membrane Protein Structure

Through the use of wide-line, solid state NMR, the perturbation of bilayer surfaces induced by integral and peripheral proteins has been studied. In this context, the "molten globule" state of a membrane associated protein, cytochrome c, has been identified. These studies are being complemented with stopped-flow fluorescence and CD studies of the insertion, folding and translocation of apocytochrome c and cytochrome c, into anionic lipid bilayers. For this work, cardiolipin, synthesized using a new route resolved by us, is being employed.

This structural work on integral membrane proteins requires instrumental development (coils, probes, computational analysis) as well as bio-organic chemistry (synthesis of isotopically enriched molecules), membrane biochemistry (isolation, purification and functional description of proteins) and molecular biology (expression of proteins). The potential for the method is significant and, in the absence of other approaches, may be the only method by which the pertinent details of membrane proteins, and their regions involved in defining molecular specificity, may be described in detail for some time to come.

Using both static and magic angle solid state NMR methods, we have been resolving structural details of large, fully functional integral membrane proteins in natural and reconstituted membranes. These methods permit atomic details to be obtained of well-defined regions of integral membrane proteins, namely those regions associated with ligand or prosthetic group binding, which may be intractable by crystallographic or other spectroscopic methods. Indeed, in the absence of routine crystallography, it is likely that structure-function relationships of membrane proteins can be described at the molecular level only through the use of such alternative methods. Proteins currently under study are bacteriorhodopsin, the gastric H+/K+-ATPase, the kidney Na+/K+-ATPase, the over-expressed GalP sugar transporter,nicotinic acetylcholine receptor, phospholamban and vertebrate and invertebrate rhodopsin.

In bacteriorhodopsin, we have been able to define the conformation and orientation of retinal whilst within the protein binding site and in the membrane using NMR observation of deuterated retinal using an ab initio approach. Curvature of the retinal polyene chain has been defined for the first time, as well as the known orientation of the ß-ionone ring of the retinal with respect to the protein. Much, but not all, experimental information available for this protein is confirmed by this non-perturbing NMR approach. Also the lack of any large change in retinal orientation within the protein during the photocycle observed using this non-perturbing approach, is already helping to redefine descriptions of the proton gradient generation by the protein on photon incidence.

Membrane protein function is controlled and regulated by, amongst other factors, the local lipid environment. Lipid-protein interactions have been studied and classes of lipids described which either have non-specific, solubilizing function, a specific functional role or combination of the two rôles. Spin-label ESR and deuterium NMR methods have both been used to prove, since it was very contentious at one time, that both spectroscopic methods can give consistent views of lipid-protein interaction and the molecular specificity of such interaction. Some fifteen proteins have been studied by various groups and we have been involved with about one third of this number. The wealth of information obtained has required much bio-organic chemistry (spin-label and deuterated lipid synthesis) and membrane protein functional reconstitution technology.

More recently, lipid-protein interactions have been found to be important in producing and stabilizing 2D-arrays of integral proteins in membranes. One particular lipid, phosphatidylglycerol phosphate is essential (and the sulphate derivative less so) in the formation of 2D arrays of bacteriorhodopsin. This has been shown by electron microscopic methods and by controlled reconstitution methods and the enigma of why and how this protein forms arrays is now resolved. Other proteins are now being examined with a view to producing routine 2D arrays for structural work.

Multidisciplinary approaches to the study of membrane structure and function are therefore used in all aspects of the current work. Most biophysical methods have been used including NMR, ESR, ultracentrifugation, diffraction (X-ray, optical and neutron), differential scanning calorimetry, electron microscopy, CD and computational approaches. The general approach is thus to address a system and a problem with a range of appropriate methods, rather than specialize in one method. All this work has been, and is being funded through significant BBSRC/HSFP, EC and MRC programme grants as well as project grant support from BBSRC, EPSRC and MRC.



Share This