A fluid mosaic model is presented for the gross organization and structure of the proteins and lipids of biological membranes. The model is consistent with the restrictions imposed by thermodynamics. In this model, the proteins that are integral to the membrane are a heterogeneous set of globular molecules, each arranged in an amphipathic structure, that is, with the ionic and highly polar groups protruding from the membrane into the aqueous phase, and the nonpolar groups largely buried in the hydrophobic interior of the membrane. These globular molecules are partially embedded in a matrix of phospholipid. The bulk of the phospholipid is organized as a discontinuous, fluid bilayer, although a small fraction of the lipid may interact specifically with the membrane proteins. The fluid mosaic structure is therefore formally analogous to a two-dimensional oriented solution of integral proteins (or lipoproteins) in the viscous phospholipid bilayer solvent. Recent experiments with a wide variety of techniqes and several different membrane systems are described, all of which abet consistent with, and add much detail to, the fluid mosaic model. It therefore seems appropriate to suggest possible mechanisms for various membrane functions and membrane-mediated phenomena in the light of the model. As examples, experimentally testable mechanisms are suggested for cell surface changes in malignant transformation, and for cooperative effects exhibited in the interactions of membranes with some specific ligands. Note added in proof: Since this article was written, we have obtained electron microscopic evidence (69) that the concanavalin A binding sites on the membranes of SV40 virus-transformed mouse fibroblasts (3T3 cells) are more clustered than the sites on the membranes of normal cells, as predicted by the hypothesis represented in Fig. 7B. T-here has also appeared a study by Taylor et al. (70) showing the remarkable effects produced on lymphocytes by the addition of antibodies directed to their surface immunoglobulin molecules. The antibodies induce a redistribution and pinocytosis of these surface immunoglobulins, so that within about 30 minutes at 37 degrees C the surface immunoglobulins are completely swept out of the membrane. These effects do not occur, however, if the bivalent antibodies are replaced by their univalent Fab fragments or if the antibody experiments are carried out at 0 degrees C instead of 37 degrees C. These and related results strongly indicate that the bivalent antibodies produce an aggregation of the surface immunoglobulin molecules in the plane of the membrane, which can occur only if the immunoglobulin molecules are free to diffuse in the membrane. This aggregation then appears to trigger off the pinocytosis of the membrane components by some unknown mechanism. Such membrane transformations may be of crucial importance in the induction of an antibody response to an antigen, as well as iv other processes of cell differentiation.

In 1972 the Fluid-Mosaic Membrane Model of membrane structure was proposed based on thermodynamic principals of organization of membrane lipids and proteins and available evidence of asymmetry and lateral mobility within the membrane matrix [S. J. Singer and G. L. Nicolson, Science 175 (1972) 720-731]. After over 40years, this basic model of the cell membrane remains relevant for describing the basic nano-structures of a variety of intracellular and cellular membranes of plant and animal cells and lower forms of life. In the intervening years, however, new information has documented the importance and roles of specialized membrane domains, such as lipid rafts and protein/glycoprotein complexes, in describing the macrostructure, dynamics and functions of cellular membranes as well as the roles of membrane-associated cytoskeletal fences and extracellular matrix structures in limiting the lateral diffusion and range of motion of membrane components. These newer data build on the foundation of the original model and add new layers of complexity and hierarchy, but the concepts described in the original model are still applicable today. In updated versions of the model more emphasis has been placed on the mosaic nature of the macrostructure of cellular membranes where many protein and lipid components are limited in their rotational and lateral motilities in the membrane plane, especially in their natural states where lipid-lipid, protein-protein and lipid-protein interactions as well as cell-matrix, cell-cell and intracellular membrane-associated protein and cytoskeletal interactions are important in restraining the lateral motility and range of motion of particular membrane components. The formation of specialized membrane domains and the presence of tightly packed integral membrane protein complexes due to membrane-associated fences, fenceposts and other structures are considered very important in describing membrane dynamics and architecture. These structures along with membrane-associated cytoskeletal and extracellular structures maintain the long-range, non-random mosaic macro-organization of membranes, while smaller membrane nano- and submicro-sized domains, such as lipid rafts and protein complexes, are important in maintaining specialized membrane structures that are in cooperative dynamic flux in a crowded membrane plane. This Article is Part of a Special Issue Entitled: Membrane Structure and Function: Relevance in the Cell's Physiology, Pathology and Therapy.

The successful penetration of endothelial basement membranes is an important process in the formation of hematogenous tumor metastases. Heparan sulfate (HS) proteoglycan is a major constituent of endothelial basement membranes, and we have found that HS-degradative activities of metastatic B16 melanoma sublines correlate with their lung-colonizing potentials. The melanoma HS-degrading enzyme is a unique endo-beta-D-glucuronidase (heparanase) that cleaves HS at specific intrachain sites and is detectable in a variety of cultured human malignant melanomas. The treatment of B16 melanoma cells with heparanase inhibitors that have few other biological activities, such as N-acetylated N-desulfated heparin, results in significant reductions in the numbers of experimental lung metastases in syngeneic mice, indicating that heparanase plays an important role in melanoma metastasis. HS-degrading endoglycosidases are not tumor-specific and have been found in several normal tissues and cells. There are at least three types of endo-beta-D-glucuronidases based on their substrate specificities. Melanoma heparanase, an Mr approximately 96,000 enzyme with specificity for beta-D-glucuronosyl-N-acetylglucosaminyl linkages in HS, is different from platelet and mastocytoma endoglucuronidases. Elevated levels of heparanase have been detected in sera from metastatic tumor-bearing animals and malignant melanoma patients, and a correlation exists between serum heparanase activity and extent of metastases. The results suggest that heparanase is potentially a useful marker for tumor metastasis.

The locations of distant secondary tumors in many clinical cancers and animal tumors are nonrandom, and their distributions cannot be explained by simple anatomical or mechanical hypotheses based on the simple lodgment or trapping of tumor cell cmboli in the first capillary bed encountered. Evidence from certain experimental tumor systems supports Paget's 'seed and soil' hypothesis on the nonrandom distributions of metastases, in which the unique properties of particular tumor ceils ('seeds') and the different characteristics of each organ microenvironment ('soil') collectively determine the organ preference of metastasis. Experimentally, differential tumor cell adhesion to organ-derived microvessel endothelial cells and organ parenchymal cells, differential invasion of basement membranes and organ tissues, and differential responses to organ-derived growth-stimulatory and -inhibitory factors all appear to be important determinants in explaining the organ preference of metastasis. Each tumor system may achieve organ specificity because of its own unique set of multiple metastasis-associated properties and responses to host microenvironments. As neoplasms progress to more highly malignant states multisite metastases are more likely and organ-specific metastases may be masked or circumvented owing to stochastic events, tumor cell diversification, host selection processes, and increased production of tumor autocrine molecules that may modulate adhesion, invasion, growth, and other properties important in metastasis. The importance of each of these properties, however, appears to vary considerably among different metastatic tumor systems. These and other tumor cell and host properties may eventually be used to predict and explain the unique metastastic distributions of certain human malignancies.

No abstract