Search

The bilayer concept was established by Singer and Nicolson in 1972 with the fluid mosaic membrane model. The outline structure of biological membranes is a lipid bilayer with integral proteins inside [1]. The membrane can be thought of as a barrier to the free entry and exit of cellular molecules and as a matrix in which (or on which) biochemical reactions take place. It was found that the structures of membranes are dynamic and contain areas of heterogeneity that are vital for their formation [2, 3]. In membranes, both proteins and lipids have lateral and rotational freedom. The functional roles of glycolipids (GLs) include maintenance of membrane structure and fluidity, lipid–protein interactions, and cell surface recognition [4].

Protein-lipid interactions play an essential role in the functioning of biomembranes and in the establishing the membrane structure. The functioning of membrane proteins is accompanied by changes in their conformation. These changes could influence the structure and physical properties of surrounding lipid environment. On the other hand, lipids can also influence the protein function. This is due to certain specificity of protein-lipid interactions. As a rule, integral proteins require specific lipids for optimal functioning and can be inhibited by other lipids1. Both peripheral and integral proteins contribute to protein-lipid interactions. Peripheral proteins are associated with membrane largely electrostatically. However nascent peripheral proteins and many soluble proteins must pass through one or more membranes during biosynthesis and therefore should interact also with inner part of the membrane. Integral proteins span the membrane transversely, establishing contacts between their hydrophobic moieties and the hydrocarbon chains of lipids. Owing to the different geometry of the hydrophobic moiety of proteins and that of lipids, as well as due to the action of electrostatic and elastic forces, regions of altered structure may arise around protein molecules. The hydrophobic interactions between integral protein and lipids are considered as a dominant. In this case the influence of the protein on its lipid environment is determined by the size of so-called hydrophobic mismatch2, i.e. the difference between hydrophobic length of protein and lipids. This reflects, e.g., in changes of phase transition temperature of lipids. On the other hand, the appearance of distorted regions around integral proteins is accompanied by changes of surface tension and the mechanical parameters of the membrane3,4. If conformational changes of integral proteins and binding of the peripheral proteins actually affects its lipid environment, then we can expect that these changes could be associated with changes of the microscopic structural state of lipid molecules as well as with the macroscopic, i.e. mechanical parameters of the membrane as a whole.

Membrane electroporation (ME) defines an electrical technique to render lipid membranes porous and permeable, transiently and reversibly, by external voltage pulses. The „electro-deformation“ of lipid vesicles is possibly a mechanism to provide larger pores leading to transmembrane transport or causing membrane rupture. Recently, electrochemical relaxation data have provided evidence that ME of a vesicle suspension is rapidly coupled with vesicle elongation. Thus the electroporative changes in the membrane structure are optically enhanced by the vesicle deformation and can be recognized in the very early stages by electrooptics. In the meantime there are numerous applications of ME to manipulate cells, organelles and tissues in medicine2, cell biology and biotechnology3, yet the exact molecular mechanism of ME is not well understood, even very controversial. We propose a general chemical-thermodynamical approach to the quantitative description of cell membrane electroporation.

In view of the many roles played by cholesterol in membranes, it is not surprising that the interactions of cholesterol with phospholipids, and their consequences with respect to membrane structure and function, have been studied intensively for more than three decades. This chapter reviews recent studies on the importance of the isooctyl side chain of cholesterol on ordering of the acyl chains, as well as how variations in the alkyl side chain length affect the conformational order of fatty acyl chains of phospholipids in bilayer membranes and the molecular packing of phospholipids in monolayers. The discussion focuses on recent biophysical studies of liposomes and monolayers in which the alkyl chain length of the sterol and/or the acyl chain length of the phospholipid were varied. The results are interpreted according to the hydrophobic mismatch effect. Also reviewed in this chapter are some recent kinetic studies of intermembrane movement and intracellular transport of sterols having a side-chain structure different from that of cholesterol. For a recent review of other aspects of sterol-phospholipid interactions, including a more detailed discussion of the use of NMR spectroscopy to establish phase diagrams than are presented here, and an overview of differential interaction of cholesterol with subclasses of phospholipids, the reader is referred to Keough et al. (1996) and reference cited therein.

Phospholipids fulfill a number of important functions of basic cell membranes. Synthetic phospholipids, in particular phosphatidylcholines, have received considerable attention because of their ability to produce a variety of morphologies including liposomes or vesicles. These vesicular structures have been used as models for biological membranes1—3 to understand in vitro membrane structure and functions with an intention to enhance their applicability. In vesicles the attention is focussed on the both the central cavity as well as the outer surface.4 During the past 40 years a large number of papers have been published on synthetic and natural phospholipids. It is clearly reflected from the reviews1,5,6that a large number of those papers have focused on the potential of these microstructures for applications in the areas of encapsulation, controlled release, biosensors, enzyme immobilization, and functional protein reincorporation.

Fast and considerable redistribution of plasma membrane proteins of isolated cardiac myocytes, excerted by anoxia, has been found using the technique of fluorescence recovery after photobleaching (FRAP) (1 ). The question is whether there is an increase of intracellular calcium which may activate the subcortical microfilaments during anoxia under our conditions and what triggers this calcium increase and protein redistribution?

Over the past few years, my laboratory has been engaged in studies on phospholipid metabolism using the ciliate protozoan Tetrahymena thermophila as a model system for eukaryotic phospholipid metabolism (Smith, 1983; Smith, 1984; Smith, 1986b; Smith, et al, 1992b; Smith and O’Malley, 1978). Tetrahymena is characterized by having high concentrations of phosphonolipids — 2-aminoethylphosphonoglyceride and 2-aminoethylphos-phonoceramide — as well as phosphatidylcholine and phosphatidylethanolamine as the major phospholipids (Smith, 1985; Smith and Giegel, 1981; Smith and Giegel, 1982; Smith, et al., 1992; Smith and O’Malley, 1978; Thompson, 1972; Thompson, et al, 1971).

To this point we have been concerned with the molecular and macromolecular building blocks of membranes. This chapter will focus on how these building blocks are assembled to become functional membranes.

Figure 1 shows a schematic view of an ensemble of membranes in a living cell. Each of these structures consists of a large quantity of various constituents: proteins, glycolipids, lipids and others. Each membrane has a specific composition, and plays a specific role in the functioning of the cell. For instance, the Golgi apparatus, shown in Figure 1, is mainly used by the cell to sort out proteins produced by ribosomes in the endoplasmic reticulum (another membrane structure) and send them towards different parts of the cell.

Carbohydrates attached to lipid and proteins are a major element of the cell surface, where, anchored in the membrane they modulate interactions with the outside world. Glycolipids may be divided into two major classes distinguished by the nature of the hydrophobic anchor. The glycolipids of bacteria and plants generally consist of a mono-or oligo-saccharide glycosidically linked to the glycerol 3-position of 1,2-diacylglycerol (Curatolo, 1987a). The major glycolipids of animals are in the second class which consist of carbohydrate glycosidically linked to ceramide glycosphingolipids (Curatolo, 1987a). Glycolipids are intimately involved with membrane structure, recognition, immune function, interaction with toxins and biological pathogens, and growth control in both normalcy and disease (Curatolo, 1987a; Thompson and Tillack, 1985; Critchley, 1979). While the important roles that these lipids play is well established, how these roles are fulfilled and influenced by the environment in which these molecules exist is far from being understood. For these reasons, there has been increasing interest in attempting to systematically delineate various physio-chemical properties of these lipids as pure systems and in mixtures (Curatolo, 1987b). The ultimate goal of such studies is to gain insight into how these parameters may influence the various biological roles that the glycolipids can assume.

The liquid condensed and liquid expanded physical states in an insoluble monomolecular film at the air-water interface and the gel and liquid crystelline states in fully hydrated bilayers are discussed and compared. It is postulated that, under reasonable restraints, the monolayer may be regarded as an adequate model membrane system. The interfacial behavior of a number of fatty acid spin-label and fluorescent probes are then outlined both for pure and mixed monomolecular films. As in bilayers and membranes these probes exhibit perturbation and immiscibility phenomena which should be taken into account when investigating membrane structure. While the perturbations of any individual probe may be small, they appear to be significant where bilayer profiles are concerned.

Phospholipids are believed to play at least two distinct roles in biological membranes. As integral parts of membrane structure, phospholipids are believed to impart much of the uniqueness to membranes and to provide a matrix in which membrane proteins are imbedded. A second role is their requirement for the function of certain membrane-associated enzymes.

The significance of the study of function, structure, and biogenesis of biological membranes and the relevance of the ensuing results to the progress of our understanding of a variety of phenomena related to the structure, differentiation, and normal functioning of the cell does not need to be emphasized here. The exponentially growing literature has been reviewed during the last years, and the main aspects of the field have been covered. The “noninitiated” reader is referred to reviews dealing with the membrane structure,1,2 modulation of composition and structure,3 methods of isolation and analysis of membrane proteins and lipoprotein components,4 and re-constitution from isolated lipoprotein complexes.5 Also, pertinent information can be found in reviews dealing with artificial and natural membrane properties,6–8 membranes of mycoplasma,9 photosynthetic bacteria,10 and biogenesis of mitochondria11 and chloroplasts.12,13

The development and application of the freeze-fracture technique to membrane research (Moor and Muhlethaler, 1963; Muhlethaler et al., 1965) led to a better understanding of the complex structural arrangement of lipids and proteins. It is now used worldwide for topographical studies, being the only available method for producing a direct high-resolution image of large areas of membrane surfaces and internal fracture faces. At first, however, it was thought that freezing would cause phase transformation and structural rearrangement of the original membrane structure. This has been disproved with the help of indirect methods such as X-ray diffraction. As shown by Gulik-Krzywicki and Costello (1977), a perturbation of the molecular organization of hydrocarbon chains does not occur provided freeze quenching is carried out rapidly enough to avoid the formation of large ice crystals between lipid lamellae. Based on these results, much work has been carried out to improve the main preparative steps such as freezing, fracturing, and replication. In addition to these procedures, new methods for labeling membrane components, special devices for rapid quenching, handling procedures for split membranes, and elaborate image-processing procedures became available. Some of these achievements will be discussed in the following chapters.

With respect to the general background of membrane modification we refer to our recent review article in “Membranes and Transport”, Vol. 1, edited by A.N. Martonosi, Plenum Press, New York and London, 1982, pages 173–184. From this article we include a list of reagents which have successively been used by various authors for the exploration of membrane structure and function (Tables 1 and 2 at the end of this article).

The increased knowledge of the properties of membrane lipids (Ansell et al., 1973) and of lipid-protein interactions (Singer, 1971; Lenaz, 1973, 1977; Vanderkooi, G., 1974) allows a better understanding of the role of lipids in membrane structure and functions. Nevertheless, a unifying picture of such a role is lacking, and it is often tacitly assumed that lipids have different roles; this is indeed the main conclusion emerging from analysis of the literature. In fact, lipids in membranes have different functions, affecting enzymic activity positively or negatively, being determinants of permeability properties and transport and being involved in the action of membrane binding sites and receptors. Moreover, they are determinants of membrane phenomena involving fusion processes (e.g., cell movement, pinocytosis, cell division, cell adhesion, secretion). In such functions, lipids may be specific or not. The physical state of a lipid, besides the specific chemical nature of certain groups, appears to be very important in its functions. It seems therefore appropriate to assign to lipids many different roles.

Many of the basic hypotheses concerning membrane structure were developed in the period 1900–1945. From 1945 to the present, intensive experimental study of these hypotheses has left our concepts of the role of lipids in the membrane largely unchanged. It has also greatly extended the evidence for a variety of roles for proteins in the membrane. Information about the way in which proteins enter into membrane structure, and how they carry out their roles, has been remarkably extended since 1960. In this review, I shall analyze the development of the basic hypotheses in terms of the interplay between hypothesis, experiment, and theory, as it occurred during the period 1930–1970.

The majority of todays membranes used in microfiltration, ultra-filtration or reverse osmosis consists of a symmetric or asymmetric microporous structure, which may in case of the asymmetric structure carry a more or less dense skin on the surface. The membranes may differ considerably in their structure, their function and in the way they are produced[1]. Preparation procedures for the different membrane types are described in patents and publications in detailed recipes, which are deeply rooted in empiricism. Superficially the preparation technique of a microporous polyethylene tube made by extrusion seems to have very little in common with the preparation of an asymmetric skin-type reverse osmosis membrane made by the precipitation technique described by Loeb and Sourirajan[2]. A more detailed analysis, however, reveals that the formation of both membrane types is determined by the same basic mechanism and governed by a process referred to by Kesting[3] as phase inversion. In fact, basically all polymeric ultrafiltration and reverse osmosis membranes and the majority of the microfiltration membranes, no matter how different their structures and their mass transport properties may be, are made by the phase inversion process. This process involves the conversion of lipid homogeneous polymer solutions of two or more components into a two-phase system with a solid, polymer-rich phase forming the rigid membrane structure and a liquid, polymer-poor phase forming the membrane pores. Phase separation can be achieved with any polymer mixtures, which forms, under certain conditions of temperature and composition, a homogeneous solution and separates at a different temperature or composition into two phases. To obtain a microporous medium such as a membrane requires both phases to be coherent. If only the solid phase is coherent and the liquid phase incoherent, a closed-cell foam structure will be obtained. If the solid phase is incoherent, a polymer powder will be obtained instead of a rigid structure.

Methodologic as well as conceptual progress during the past decade has made it possible for renal physiologists to “peek” inside the epithelial “black box” of renal tubular function with a greater degree of confidence than ever before. Among the more important developments have been: (1) evolution of the fluid mosaic model of plasma membrane structure (Singer and Nicolson, 1972), (2) emergence of a role for the cytoskeleton as a regulator of membrane function (Nicolson and Poste, 1976), (3) maturation of the concept that electrochemical potential gradients can provide the driving force for transport systems, i.e., Na+ gradient, chemiosmotic hypotheses (Crane, 1977; Mitchell, 1976) without being coupled directly to metabolic intermediates, (4) application of sophisticated electrophysiologic measurements to the elucidation of renal transport processes at opposing membrane surfaces (Gottschalk and Lassiter, 1973), (5) in vitro perfusion of isolated tubule segments (Burg and Orloff, 1973), (6) biochemical separation of epithelial cellular organelles, especially isolation and separation of clean membrane fractions from luminal (brush border) and antiluminal (basolateral) membrane surfaces (Heidrich et al., 1972), and (7) in vivo characterization of substrate interactions with luminal as distinct from antiluminal nephron surfaces (Silverman et al., 1970a,b).

A general model of active transport is proposed, based on the flexoelectric membrane properties, with regard to liquid crystalline membrane structure. The physical principles of the flexoelectric effect are briefly reviewed and it is admitted that the spherical deformation of planar membrane produces a flexoelectric polarization which results in depolarizing electric field. The orientational state of integral protein globuli in this field is considered and it is found to be similar to that of a trigger. It is shown that a change of the globula dipole moment by adhering to the negative end of a proton may cause its reorientation accompanied with translocation of the proton to the outer membrane side. The theoretical formulae describing such a mechanism are derived and both the value and the sign of transmembrane potential are calculated as a function of flexoelectric membrane coefficients. Two chanal feedback control leading to stable self-regulation both of the membrane curvature and transmembrane potential, is involved in this model. In view of such mechanism a natural explanation may be given for a number of active transport features: curved membrane sectors are found to be metabolitically active; transport energy is secured by membrane itself; assymmetry and vector character of the transport are clearly demonstrated. There exist at least two experiments which are consistent with the notion of flexoelectricity and permit in principle to measure the value of flexoelectric membrane coefficient.

Our understanding about the chemical composition and functional properties of platelet membrane systems has been advancing very rapidly at both the morphological and molecular levels over the last ten years or so. This is well emphasized by the fact that in certain areas of membrane biochemistry the use of the platelet as a model cell for probing the more general aspects of membrane structure and behavior, as also for studying drug transport and disease-related membrane defects, has resulted in a popularity rating for the platelet almost equal to that enjoyed earlier by the red cell. Although both the red cell and the platelet are anucleate cells, in some respects the platelet with its highly interactive surface membrane, its mitochondria, internal membranes, and lysosomes and storage organelles more reflects the behavioral and metabolic activities of other cells in the body than does the red cell with its high level of functional specialization. Since the platelet circulates as a poised and potentially highly reactive cell, responding rapidly to external stimuli, and is also most sensitive to even minor biochemical changes in the surrounding milieu, considerably more attention has been focused on the plasma membrane and its constituents and properties than on the various intracellular membrane systems that are equally well developed for functional needs. These latter membrane elements include not only endoplasmic-reticulumlike (ER) structures often referred to as dense tubular membrane system (DTS), but also include the boundary membranes of the many different granular organelles residing within the cytoplasmic matrix (mitochondria, lysosomes, α-granules, dense body granules, etc.).

Energy transduction in biological membranes is supported by functionally distinct protein complexes embedded in the lipid bilayer matrix. Various biophysical methods introduced to the field of biomembranes yielded indispensible information about structure and function of the lipid bilayers and their embedded protein complexes. The demonstration of the lateral and rotational mobility of membrane proteins (1,2) undoubtedly influenced the current concepts of membrane structure and function to a large extent. In principle there are three main methods which can yield quantitative information on lateral and rotational mobility of membrane proteins. These are: fluorescence, phosphorescence, and absorption polarization techniques; fluorescence photobleaching recovery; and saturation transfer in electron parametric resonance (for reviews see 3,4). Here the use of phosphorescence (triplet) probes for measuring the rotational diffusion of intrinsic and extrinsic membrane proteins will be described. Besides yielding information on internal fluctuations within the protein, this technique enables the measurement of comparatively slow rotational diffusion (> 0.1 Pa s = 1 cp), i.e., in biological membranes. Thus it extends the application range to time domains which are more relevant to the biological action of enzymes.

The correlation between chain order and chain mobility is notwell understood as yet, and it is thus important to keep the two concepts apart.....it should also be obvious that the distinction between time-averaged structural parameters (such as order parameters) and dynamic parameters (such as relaxation times, correlation times and microviscosity) refers to membranes in general and is independent of the specific technique employed. This is also illustrated by fluorescence spectroscopy, where it has been realized only recently that the interpretation of fluorescence measurements exclusively in terms of microviscosity is incorrect, but that the fluorescence anisotropy is equally dependent on the ordering of the fluorescent probe in the membrane...

By “metabolism of myelin,” we refer to the molecular events involved in the synthesis of myelin components, and the subsequent assembly, maintenance, and turnover of the myelin sheath. The deposition of myelin involves coordination of the synthesis of its various lipid and protein components, and the interaction of these components to give a stable membrane. Degradation of myelin components occurs as a reaction of the myelin sheath to injury, but may also be required for normal maintenance and remodeling of the membrane. The topographic features of the myelin sheath can be expected to limit partially the metabolism of myelin. The roles of membrane fluidity and cytoplasmic inclusions in the turnover of compact myelin lamellae are not well understood, but obviously metabolism in this system is integrally linked to both anatomical and membrane structure.

The structure and function of cellular membranes in seeds are dynamic. Substantive changes occur as water is imbibed, as the axis initiates elongation, and, at least in dicot seeds, as the reserves are mobilized from the cotyledons. This brief summary seeks to review some of the more recent experiments concerning the structure of the cellular membranes in seeds during each of these processes.

In the past decade fluorescence probe molecules have become increasingly useful in unraveling the asymmetric distribution of proteins, lipids, and sterols in biological membranes. The intrinsic sensitivity and the ability to discriminate between extremely short-lived membrane events provide fluorescence methodology with distinct advantages over NMR, ESR, X-ray scattering, infrared spectroscopy, calorimetry, and optical rotary dispersion-circular dichroism (Badley, 1976). Advances in fluorescence microscopy such as fluorescence photobleaching recovery (Peters et al., 1974; Edidin and Fambrough, 1973; Axelrod et al., 1976), fluorescence lifetime determination by phase and modulation or by photon counting (Ware, 1971; Weber, 1981; Isenberg, 1975), and computer-centered spectrofluorimetry (Holland et al., 1973, 1977; Christman et al., 1980, 1981; Wampler, 1976) have greatly enhanced the utility of fluorescence probe molecules in determination of biological membrane asymmetry. An evaluation of the asymmetric structure of membranes requires recognition of two possible asymmetric distributions of components either in the plane of the membrane (lateral, horizontal) or normal to the plane (transbilayer, vertical). Both of these asymmetric aspects of membrane structure may have important physiological consequences as detailed herein.

Chemical agents with different capacities to penetrate into the membrane and with different chemical reactivities can be used to gain information concerning the location of transport sites in the membrane structure and the particular functional ligands. If the agents are highly specific in their interactions and if their inhibitory effects are irreversible, they can also be used as probes to identify the transport components. Several examples are cited using the human red blood cell as a model. The anion transport system in particular has been studied by the use of nonpenetrating irreversible inhibitors, and more recently with a photoaffinity probe, NAP-taurine. In the dark the latter is transported in competition with the normal inorganic anions but after exposure to light, it becomes fixed in an irreversible bond that allows identification of the sites of its transport. It is proposed that anion transport involves a transmembrane protein of about 90,000 daltons that forms a channel through the lipid bilayer. The exchange of anions occurs via a gating mechanism containing a specific anion-binding site. Transport of water, cations and sugars may also involve similar transmembrane protein channels.

The analysis of the conductance properties of the skeletal muscle membrane has proven to be considerably more difficult than earlier investigations of axon excitability. The principal reason for this difficulty is the unique internal membrane structure of muscle, consisting of an infolding of the surface membrane to form a transverse tubular membrane structure, as well as the sarcoplasmic reticulum, which appears to be electrically coupled to the tubular system. A voltage-clamp measurement of this complex system is always beset with the difficulty of charging the tubular membrane in a uniform manner so as to allow measurement of the associated membrane currents. The voltage-clamp analyses that have been done probably have had varying amounts of uncontrolled tubular membrane current. As a result, the quantitative analyses of these experiments may not be entirely valid.

Myelin was, for many years, the focal point for studies of membrane structure, largely because of its abundance and accessibility for chemical analyses, and its ordered laminar structure, which made it amenable to physical measurements. It is widely recognized that myelin is a highly specialized membrane structure, both in structure and in function, and that it is largely a “living relic” of the plasma membrane of glial or Schwann cells, arising as a result of cellular differentiation processes which occur during development of the nervous system. In order to gain some insight into and perspective of the complexities of this membranous structure, it is important to relate the growing body of structural information on myelin to the general principles of membrane structure that have emerged from the numerous studies of other membrane systems. For the purpose of discussing the molecular organization of myelin, it will be sufficient to summarize those concepts of membrane structure which have gained a significant measure of acceptance. The literature in this field has become immense during the past 5 years, and all aspects of membrane structural studies summarized here have been extensively discussed and reviewed (Gitler, 1972; Levine, 1972; Singer and Nicolson, 1972; Bretscher, 1973; Singer, 1974; Vanderkooi, 1974; Wallach and Winzler, 1974).

Early electron micrographs suggested that all membranes have the same basic structure. In the late 1950s and early 1960s, J. D. Robertson proposed his unit membrane theory of membrane structure. He suggested that all membranes, whether from plants, animals or micro-organisms have a common structural pattern based on a bimolecular lipid layer with protein associated with the two faces; the pattern, in fact proposed by Danielli and Dayson (see chapter 6). Robertson’s theory was based on electron-microscope studies of myelin, though in later work he examined a wide variety of membranes and found that they all appeared to have the same structure as myelin.

Conformational transitions in membrane-bound macromolecules probably provide the most general and powerful mechanisms for chemical and physical signalling processes in biology. Changes in the conformations of membrane proteins and glycoproteins are exploited in cell biology to modulate enzymatic activity, to change membrane permeability, and to initiate the membrane reorganization events involved in endocytosis, in secretion and in the processing of ligands and receptors. In recent years, we have explored the consequences of conformational transitions in synthetic macromolecules bound to lipid bilayer membranes by adsorption or via hydrophobic anchoring groups.1-4 The present chapter examines the roles of polymer conformation and solvation in controlling membrane structure in such systems.

The prevailing view of membrane structure presented in many undergraduate textbooks is that of proteins floating in the plane of a fluid phospholipid bilayer (Singer and Nicolson, 1972). Many observations in the last 10 years indicate, however, that biological membranes do not behave as simple two-dimensional solutions of proteins. Measurements of the rates of lateral diffusion of membrane proteins in a variety of systems have revealed populations of proteins that are not mobile. The proteins that are mobile move at rates 10- to 100-fold slower than predicted on the basis of membrane viscosity (reviewed by Cherry, 1979). The diffusion constants of these slowly diffusing proteins are increased 200-fold in areas of the cell where the membrane is separated from the cytoplasm (Tank et al., 1982). Furthermore, diffusion of proteins may be nonrandom in some cells (Smith et al., 1979) and can require metabolic energy as in formation of caps of surface-labeled proteins in lymphocytes (Unanue and Karnovsky, 1973). These examples indicate possibilities for long-range interactions and organization in cell membranes and have led to proposals that at least some membrane proteins have direct interactions with underlying cytoplasmic proteins (Singer, 1974; Nicolson, 1974; Edelman, 1976; Bourguignon and Singer, 1977).

The electrical potential across the platelet plasma membrane changes in response to specific stimuli (e.g., thrombin, ADP) and, conversely, alteration of the trans-membrane potential affects the platelet sensitivity to these activating agents. These electrical changes are mediated by a redistribution of cations across the plasma membrane, and therefore it may be assumed that conformational changes of membrane proteins are involved in this membrane transport process. Therefore, even though a discussion of platelet membrane potential cannot yet be related to the membrane glycoproteins, an understanding of this phenomenon is essential for a complete description of platelet membrane structure and function.

Plasma membranes of both gram-positive and gram-negative bacteria perform a variety of functions including transport of ions and metabolites, secretion of proteins, energization, biosynthetic stages in cell wall peptidoglycan, lipopolysaccharide, and capsular polysaccharide formation, and the biosynthesis of membrane lipids. All of these functions are organized in a single membrane system, in contrast to the functional compartmentalization characteristic of the membranous organelles (e.g., mitochondria, endoplasmic reticulum, Golgi, and lysosomes) of eukaryotic cells (Stanier, 1970; Carlile, 1980). In addition to plasma membranes, gram-negative bacteria possess a distinctive outer membrane (Inouye, 1979), and certain groups of bacteria possess specialized intracellular membranes such as chromatophores of photosynthetic bacteria, cytomembranes of nitrifying organisms, and spore membranes of sporulating bacteria. The only other membrane structure seen in many bacteria, especially in gram-positive organisms, is the mesosome. The origins and functions of mesosomes have been widely reviewed and there is still much speculation and some evidence as to their role and significance in the bacterial cell and cell cycle (Ghosh, 1974; Salton and Owen, 1976; Higgins et al?,1981). Central to an understanding of the structure-function relationships of these various bacterial membrane systems are (1) resolution of the complexity of the components in plasma membranes, outer membranes of gram-negative bacteria, and specialized membranes such as chromatophores, and (2) establishment of the molecular architecture and sidedness of the membranes and derived vesicles. Much progress has been made in recent years by resolving the variety of polypeptides in membrane structures by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (SDS-PAGE) and by the application of two-dimensional (“crossed”) immunoelectrophoresis (CIE) in the identification of membrane antigens and/or enzymes under essentially nondenaturing solubilization conditions with nonionic detergents such as Triton X-100(Owen and Salton, 1975a; Owen and Smyth, 1977; Smyth et al., 1978).

The functional complexity of the eukaryotic membrane requires a similar structural complexity. One of the major components of the membrane is the sterol molecule, such as cholesterol (in animals), sitosterol (in plants), and ergosterol (in fungi). These neutral lipid molecules act not only as a bulk lipid, but also act to stabilize the fluidity of the membrane structure, thus allowing for protein function under conditions such as high or low temperature under which the phospholipid hydrocarbon chains might either gel or become too fluid for proper physiological functioning.

Intact membranes as well as membrane components have been studied extensively by various physical and chemical methods. Important data have accummulated but more specific structural information on the atomic level is still necessary in order to obtain a detailed understanding of lipid-lipid and lipid-protein interactions and of variations in structure and composition of lipids observed in different types of membranes. Only then will it be possible to explain the function of the different constituents and their significance for various membrane properties.