Proteomics of total membranes and subcellular membranes

Membrane proteins are key molecules in the cell and are important targets for drug development. Much effort has, therefore, been directed towards research of this group of proteins, but their hydrophobic nature can make working with them challenging. Here we discuss methodologies used in the study of the membrane proteome, specifically discussing approaches that circumvent technical issues specific to the membrane. In addition, we review several techniques used for visualization, qualification, quantitation and localization of membrane proteins. The combination of the techniques we describe holds great promise to allow full characterization of the membrane proteome and to map the dynamic changes within it essential for cellular function.     

Evidence for tubulin-binding sites on cellular membranes: plasma membranes, mitochondrial membranes, and secretory granule membranes

We describe the interaction of pure brain tubulin with purified membranes specialized in different cell functions, i.e., plasma membranes and mitochondrial membranes from liver and secretory granule membranes from adrenal medulla. We studied the tubulin-binding activity of cellular membranes using a radiolabeled ligand-receptor assay and an antibody retention assay. The tubulin-membrane interaction was time- and temperature-dependent, reversible, specific, and saturable. The binding of tubulin to membranes appears to be specific since acidic proteins such as serum albumin or actin did not interfere in the binding process. The apparent overall affinity constant of the tubulin- membrane interaction ranged between 1.5 and 3.0 X 10(7) M-1; similar values were obtained for the three types of membranes. Tubulin bound to membranes was not entrapped into vesicles since it reacted quantitatively with antitubulin antibodies. At saturation of the tubulin-binding sites, the amount of reversibly bound tubulin represents 5-10% by weight of membrane protein (0.4-0.9 nmol tubulin/mg membrane protein). The high tubulin-binding capacity of membranes seems to be inconsistent with a 1:1 stoichiometry between tubulin and a membrane component but could be relevant to a kind of tubulin assembly. Indeed, tubulin-membrane interaction had some properties in common with microtubule formation: (a) the association of tubulin to membranes increased with the temperature, whereas the dissociation of tubulin- membrane complexes increased by decreasing temperature; (b) the binding of tubulin to membranes was prevented by phosphate buffer. However, the tubulin-membrane interaction differed from tubulin polymerization in several aspects: (a) it occurred at concentrations far below the critical concentration for polymerization; (b) it was not inhibited at low ionic strength and (c) it was colchicine-insensitive. Plasma membranes, mitochondrial membranes, and secretory granule membranes contained tubulin as an integral component. This was demonstrated on intact membrane and on Nonidet P-40 solubilized membrane protein using antitubulin antibodies in antibody retention and radioimmune assays. Membrane tubulin content varied from 2.2 to 4.4 micrograms/mg protein. The involvement of membrane tubulin in tubulin-membrane interactions remains questionable since erythrocyte membranes devoid of membrane tubulin exhibited a low (one-tenth of that of rat liver plasma membranes) but significant tubulin-binding activity. These results show that membranes specialized in different cell functions possess high- affinity, large-capacity tubulin-binding sites...     

Proteomics of total membranes and subcellular membranes

Membrane proteins are key molecules in the cell and are important targets for drug development. Much effort has, therefore, been directed towards research of this group of proteins, but their hydrophobic nature can make working with them challenging. Here we discuss methodologies used in the study of the membrane proteome, specifically discussing approaches that circumvent technical issues specific to the membrane. In addition, we review several techniques used for visualization, qualification, quantitation and localization of membrane proteins. The combination of the techniques we describe holds great promise to allow full characterization of the membrane proteome and to map the dynamic changes within it essential for cellular function.     

Bending membranes

It is widely assumed that peripheral membrane proteins induce intracellular membrane curvature by the asymmetric insertion of a protein segment into the lipid bilayer, or by imposing shape by adhesion of a curved protein domain to the membrane surface. Two papers now provide convincing evidence challenging these views. The first shows that specific assembly of a clathrin protein scaffold, coupled to the membrane, seems to be the most prevalent mechanism for bending a lipid bilayer in a cell. The second reports that membrane crowding, driven by protein-protein interactions, can also drive membrane bending, even in the absence of any protein insertion into the bilayer.     

Fusion of synaptic vesicle membranes with planar bilayer membranes.

The interaction of synaptic vesicles with horizontal bilayer lipid membranes (BLMs) was investigated as a model system for neurotransmitter release. High concentrations (200 mM) of the fluorescent dye, calcein, were trapped within synaptic vesicles by freezing and thawing. In the presence of divalent ions (usually 15 mM CaCl2), these frozen and thawed synaptic vesicles (FTSVs) adhere to squalene-based phosphatidylserine-phosphatidylethanolamine BLMs whereupon they spontaneously release their contents which is visible by fluorescence microscopy as bright flashes. The highest rate of release was obtained in KCl solutions. Release was virtually eliminated in isotonic glucose, but could be elicited by perfusion with KCl or by addition of urea. The fusion and lysis of adhering FTSVs appears to be the consequence of stress resulting from entry of permeable external solute (KCl, urea) and accompanying water. An analysis of flash diameters in experiments where Co+2, which quenches calcein fluorescence, was present on one or both sides of the BLM, indicates that more than half of the flashes represent fusion events, i.e., release of vesicle contents on the trans side of the BLM. A population of small, barely visible FTSVs bind to BLMs at calcium ion concentrations of 100 microM. Although fusion of these small FTSVs to BLMs could not be demonstrated, fusion with giant lipid vesicles was obvious and dramatic, albeit infrequent. Addition of FTSVs or synaptic vesicles to BLMs in the presence of 100 microM-15 mM Ca2+ produced large increases in BLM conductance. The results presented demonstrate that synaptic vesicles are capable of fusing with model lipid membranes in the presence of Ca+2 ion which, at the lower limit, may begin to approach physiological concentrations.     

Taming membranes: functional immobilization of biological membranes in hydrogels

Single molecule studies on membrane proteins embedded in their native environment are hampered by the intrinsic difficulty of immobilizing elastic and sensitive biological membranes without interfering with protein activity. Here, we present hydrogels composed of nano-scaled fibers as a generally applicable tool to immobilize biological membrane vesicles of various size and lipid composition. Importantly, membrane proteins immobilized in the hydrogel as well as soluble proteins are fully active. The triggered opening of the mechanosensitive channel of large conductance (MscL) reconstituted in giant unilamellar vesicles (GUVs) was followed in time on single GUVs. Thus, kinetic studies of vectorial transport processes across biological membranes can be assessed on single, hydrogel immobilized, GUVs. Furthermore, protein translocation activity by the membrane embedded protein conducting channel of bacteria, SecYEG, in association with the soluble motor protein SecA was quantitatively assessed in bulk and at the single vesicle level in the hydrogel. This technique provides a new way to investigate membrane proteins in their native environment at the single molecule level by means of fluorescence microscopy.     

SIMS for membranes

New research demonstrates that terminally differentiated cells are more efficient than adult stem cells for cloning.     

Interaction of smooth muscle plasma membranes with flat lipid membranes

The method of implantation of smooth muscle cells from plasma membranes (PM) of the rabbit intestine into flat lipid membranes (FLM) is described. The method is based on the pretreatment of PM vesicles with asolectin liposomes in the ratio that provides the activation of membrane ATPases. Thus modified FLM possesses channel conductivity.     

Purification of cytoplasmic membranes and outer membranes from Proteus mirabilis

Summary A crude cell envelope suspension has been prepared from Proteus mirabilis after osmotic shock of penicillin-induced spheroplasts. Employing discontinuous sucrose gradients this cell envelope suspension can be fractionated into four fractions. Besides a pellet of remaining spheroplasts and an intermediate fraction with mixed composition a highly purified cytoplasmic membrane fraction and an outer membrane fraction have been obtained. The cytoplasmic membrane fraction is not contaminated with mucopeptide or outer membrane material. It has a buoyant density of 1.13 g/ml and a protein content of 38%. The specific activities of formate dehydrogenase and nitrate reductase and the content of cytochrome b₁ have increased sixfold in comparison with the crude cell envelope suspension. The outer membrane fraction contains only few contaminations with cytoplasmic membrane components and with mucopeptide.The gradient fractions have been characterized by electron microscopy and by polyacrylamide gel electrophoresis.     

Composition ofNeurospora crassa vacuolar membranes and comparison to endoplasmic reticulum,plasma membranes,and mitochondrial membranes

We have determined the carbohydrate and lipid contents of vacuolar membranes fromNeurospora crassa, and have compared them to mitochondrial membranes, endoplasmic reticulum, and plasma membranes. These four membrane fractions were clearly distinct from each other in polypeptide composition, as judged by polyacrylamide gel electrophoresis. The vacuolar membranes proved unusual in two respects: the contained very high amounts of carbohydrate and were the only membranes with significant levels of phosphatidylserine. As in other eucaryotic cells, the mitochondrial membranes were unique in having high amounts of cardiolipin but virtually no sterol. Although the endoplasmic reticulum and plasma membranes were qualitatively similar to each other, the plasma membranes could be distinguished by a higher carbohydrate content, whereas the endoplasmic reticulum had a characteristically high ratio of phosphatidylcholine to phosphatidylethanolamine. The fatty acid compositions of all four membranes were similar, except that mitochondrial membranes contained about half as much saturated fatty acids as the other three fractions.