Upsala J Med Sci 81: 193-200, 1976 The Fluid Double Polar-Nonpolar-Polar Leaflet Model for Biological Membranes ERIK CERVEN From the Institute of Medical and Physiological Chemistry, Biomedical Center, University of Uppsala, Uppsala, Sweden ABSTRACT A model for biological membranes is proposed accord- ing to which the plasma membrane consists of two functionally different polar-nonpolar-polar leaflets separated by a polar space. The binding of water- soluble proteins, integral lipoproteins and spanning proteins to a biological membrane as well as possible Conformations of interphase peptides partitioned be- tween polar and nonpolar layers are discussed. A model for the diffusion of water soluble proteins across adenylate kinase ( 5 ) , NDP kinase (2), ATPase (3, 43), and protein kinase (6, 44). Phenomena such as ghost resealing (c$: 26), secretion of proteins apparently without exocytosis (23, 32), and the transport of polynucleotides across intact plasma membranes (cf.: 36)are difficult to explain with the traditional lipid bilayer concept of a plasma membrane. This concept for biological mem- branes in general has previously been questioned nonpolar layers Of a membrane is described. Three complete biological membranes containing two leaflets and an inter-leaflet space are defined. These are: 1: The inner nuclear membrane + the perinuclear space and the endoplasmatic cisternae + the outer nuclear membrane and the endoplasmatic reticulum, 2: the inner mitochondrial membrane + the mitochondrial intermembraneous space + the outer mitochondrial membrane and 3: the cytoplasmic leaflet of the plasma membrane + an intramembraneous space in the plasma membrane + the outer leaflet of the plasma membrane. on morphological ( 4 3 , chemical (28) and bio- functional (28) grounds. Perhaps the most con- vincing objection is that lipids can be extracted from Some biological membranes without altering the usual trilaminar appearance in electronmicro- scopic pictures (18). Also the nature of the ma- terial in the sheet which does not st& with osmium remains undetermined because according to the traditional view strong hydro- phobic bonds would be expected to be present thus preventing, for example, tangential splitting of this region on freeze-etching (6: 51). INTRODUCTION The fluid mosaic model proposed by S.J. Singer and G.L. Nicholson in 1972 (46) represented con- siderable progress in our understanding of several plasma membrane phenomena but it does not satisfactorily explain all of them, in particular several findings that have been made in this labo- ratory during the past 10 years relating to the as- sociation of cytoplasmic enzymes to the plasma membrane (4 l), albumin-inhibited leakage of such enzymes into the extracellular medium (54) and the apparent existence of an intramembraneous pool of solutes (4). All cells hitherto investigated in this laboratory display extracellular adenylate metabolism including ATP synthesis (1, 5 , 42), THE MODEL As an approach to these and other questions as- sociated with the traditional lipid bilayer concept, the fluid double polar-nonpolar-polar leaflet model for biological membranes illustrated in Fig. 1 to 7 is proposed. According to this model, biological membranes consist of two functionally different polar-nonpolar-polar (PNP) leaflets each containing water soluble and integral (46) pro- teins, lipids, and extended polypeptides partitioned between polar and nonpolar phases. One leaflet is in addition enriched in carbohydrate residues shared by the polar and nonpolar layers. The structure of the plasma membrane, for example, Upsala J Med Sci 81 194 Erik Cervin P, '. I p5 r P? . I 4 -- Fig. 1. The principal structure of a biological mem- brane. Light areas indicate polar and dark areas indi- cate nonpolar residues (P = polar layer, N = nonpolar layer, I = intermembraneous space). The indices and the arrow indicate the polarity of the two leaflets. The proportions between the different layers and the inter- membraneous space are not necessarily the same as in this illustration, but may depend on the composition and the amount of the different chemical constituents. A B C D E Fig. 2. Association of watersoluble proteins to one leaflet of a biological membrane. The capitals A-E indicate increased complexity of the interactions be- tween associated proteins and the membrane. from the cytoplasmic surface and outward is given by the following general scheme (see Fig. 1). The layers of the N2 leaflet P1. Polar bound (= associated and covalently bound to the nonpolar layer) residues. N2. Fluid (9) nonpolar residues. P3. Polar bound residues. The I4 space Intermembraneous polar space containing solutes and enzymes but susceptible to conventional electron microscopic preparatory techniques. The layers of the N6 leaflet P5. Polar bound residues. N6. Fluid nonpolar residues. P7. Polar bound residues. Upsala J Med Sci 81 A. I ll Jll I Y P Fig. 3. Development of watersoluble membrane-as- sociated proteins into integral (A-B) and spanning (B) proteins. The Roman numbers I-IV indicate increased content of nonpolar amino acid residues involved in binding the integral protein to a membrane. A B C D E F G Fig. 4. Possible modes of binding of an integral protein spanning (46) across two leaflets of a biological mem- brane e.g. the plasma membrane (A-D). Stabilization of close spatial arrangement of two leaflets of a bio- logical membrane by indirect (E) and direct (F) polar bonds and by fusion of the N2 and N6 leaflets (G). The binding between a membrane and globular proteins that do not become denatured when extracted (e.g. amino acid-binding proteins from bacteria, 10) is illustrated in Fig. 2. The substitution of polar bonds involved in this type of binding for nonpolar bonds resulting in proteins integrated (46) in the nonpolar layer (e.g. a Nai- and- K+- stimulated ATPase, 1O)is illustrated in Fig. 3 A and B. Two types of spanning (46)proteins can be visualized in this model, mainly polar-nonpolar- polar (PNP, see Fig. 4A and B) and mainly polar- nonpolar-polar-nonpolar-polar (PNPNP, see Fig. 4 C and D) tentatively illustrated by glycophorin A (35) and the band 3 protein (25, 35) from ery- throcytes respectively. The spanning proteins pre- sumably hold the two leaflets of the plasma membrane together. This function could also be accomplished by indirect or direct polar bonds between the P3 and P5 layers as illustrated in Fig. 4 E and F or by fusion (c$: 29) of the N2 and N6 layers (Fig. 4 G). It is to be expected that a PNP leaflet heavily The fluid double proteolipid leaflet model f o r biological membranes 195 Fig. 5. Conformations of oligo-peptides partitioned between polar (P) and nonpolar (N) layers of a mem- brane without reference to possible stabilizing interactions with adjacent molecules. Oligopeptides with alternating polar and nonpolar amino acids represented by glu-leu-glu-leu (A-B) and leu-glu-leu-glu (C-D). Part of a ring structure (E I) and possibilities for direct (E 11) or indirect (E 111) electrostatic stabilization illustrated by a conformation of the oligopeptide glu-glu-trp-leu-glu-glu. Upsala J Med Sci 81 196 Erik Cewkn loaded with electronlucent and possibly displace- able enzymes bound as illustrated in Fig. 2 E would yield an electron microscopic picture reminiscent of a lipid bilayer. The reason for this is that polar and nonpolar residues of a mem- brane are closer and therefore firstly more sus- ceptible to cross-linking, secondly more electron dense and thirdly, since they are linked together by both polar and nonpolar bonds, also less extractable. However, on careful examination this type of trilaminar membrane (see Fig. 2E) would be expected to reveal globular structures re- sembling those found in mitochondrial and endo- plasmatic reticulum membranes (47). The pro- found effect on the interleaflet distance in bio- logical membranes of conventional electron microscopic preparatory techniques (48) which are known to produce leakage of cytoplasmic proteins, shrinking and swelling of specimens, is documented in the case of mitochondria (49) chloroplasts (50) and plasma membranes (1 7). In all these cases, certain preparation techniques seem to preserve a broader interleaflet distance. This favours the view that inter-leaflet spaces are polar and excludes the possibility of strong non- polar bonds in this region. Tangentially extended polypeptides with polar residues directed outward and nonpolar residues directed inward in the proteolipid leaflets are imagined to constitute some kind of a lipid ske- leton and in addition to mediate the attachment of watersoluble enzymes and carriers. These interphase polypeptides could be expected to require a secondary structure different from the a- helix, the P-pleated sheat and the random coil, perhaps making them even more susceptible to cross-linking agents used in electron microscopy and to other frequently used denaturing proce- dures. Some possible configurations of such inter- phase polypeptides are illustrated in Fig. 5 . As shown in Fig. 5 A-D, polypeptides with alternating polar and nonpolar residues are readily partitioned between the water and lipid phases of a biological membrane. Exclusively nonpolar peptides may form a ring structure with the nitrogen atoms of the peptide bond directed towards the center. Part of such a ring located in the interphase between polar and nonpolar layers Upsala J Med Sci 8I is shown in Fig. 5 E. Numerous possibilities for stabilization of an interphase conformation by electrostatic interactions exist, some of which are illustrated in Fig. 5 E. These interactions could be direct, between positively and negatively charged amino acid residues, or indirect, mediated by cations or anions that link similarly charged groups together. This model imposes functional similarities be- tween the plasma membrane and the mitochond- rial membranes because, in both cases, a pool of enzymes involved in adenylate metabolism (e.g. adenylate kinase, 5 , 16, 33, and NDP kinase, 2, 33) is located in the I4 space. This pool is acces- sible from the outside through a N6 leaflet permeable to nucleotides (33) and containing carbohydrates (8, 37, 46). On the other hand the inner N2 leaflet is assumed to be impermeable to nucleotides necessitating the presence of various carriers in both inner mitochondrial (33) and plasma membranes. Phenomena associated with intact cells such as crypticity (15, 19) of enzymes (15, 19) and receptors (14) could be explained by this similarity. Weak interactions between the P3 and P5 layers are imagined to catalyze the fluid state- dependent fusion of the N2 and N6 layers (see Fig. 4 G) resulting in nonpolar intermembraneous bridges. The fusion points are possible sites for diffusion of cytoplasmic enzymes across the plasma membrane as illustrated in Fig. 6 A. They could thus be responsible for the fact that albu- min inhibits leakage of several cytoplasmic en- zymes from some tumor cells without affecting or displacing other enzymes expressing their activity at the cell surface (54)indicating differential ac- cessability of albumin to the plasma I4 and intra- cellular spaces of macromolecules. On the other hand, preferential and specific entry of globular proteins through one PNP leaflet must be considered (see Fig. 6 B) because some soluble enzymes are enriched in the mitochondrial I4 space (33) which because of continuous great changes of shape (21) can not have any fusion points between its two leaflets. If all transport of macromolecules across the plasma membrane not mediated by exo- and endocytosis (38) occurred via the I4 space any function of enzymes em- The fluid double proteoligid leaflet model f o r biological membranes 197 A Fig. 6. Diffusion of watersoluble proteins across non- polar layers of a membrane. The diffusion of a protein across a double leaflet membrane is shown in A. The Roman numbers 1-111 represent successive events while I V and V represent possible final events. Succes- sive events in the diffusion of a globular protein across a single leaflet membrane are shown in B. Fig. 7 . The principal outline and polarity of three complete biological membranes, the inner nuclear membrane and the outer nuclear membrane combined with the endoplasmatic membranes, the mitochondrial membranes and the plasma membrane(s). The cor- responding interleaflet spaces are the perinuclear space with the cisternae of the endoplasmatic reticulum, the intermembraneous space of the mitochondria and the intramembraneous space of the plasma membrane. N=nucleus, M=mitochondrion, P=plasma mem- brane. The arrows indicate the polarity from the N2 to the N6 leaflet. bedded in or between the two leaflets would be disturbed. This simple mechanism for transport of macromolecules across biological membranes implies that only adhesion sites could be required since low entropy adhesion would favour dis- placement of the underlying nonpolar fluid lipid residues. On the basis of ultrastructural evidence and biochemical characterization the polarity of at least three complete biological membranes can be determined (see Fig. 7). By definition the associa- tion of energy transducers such as coupling fac- tors (33) and actin (30, 3 1) and covalently bound DNA, characterizes a N2 leaflet as exemplified by the inner nuclear membrane (12, 13, cJ: 34) the inner mitochondrial membrane (33, cJ: 53) and the cytoplasmic leaflet of the plasma mem- brane (24 cJ: 22). Also the presence of adenylyl- cyclase (40) is a characteristic feature of some N2 leaflets. A comparatively high content of carbo- hydrates characterizes a N6 leaflet as for example the outer nuclear membrane (7, 37), the endo- plasmatic reticulum ( 4 3 , the outer mitochondrial membrane (8, 37)and the exterior leaflet of the plasma membrane (46). This characterization is supported by ultrastructural evidence indicating a globular appearance of the mitochondrial and endoplasmatic membranes (47) which resemble each other and are thinner than the plasma membrane (47). The facts that the nuclear mem- branes are thin and rarely display a trilaminar structure (52) indicate that these membranes will be possible to characterize as stacked with glo- bular enzymes in analogy with the structure of mitochondrial and endoplasmatic membranes (47). The associated I4 spaces in these cases are the perinuclear space combined (52) with the cisternae of the endoplastmatic reticulum, the intermembraneous space of mitochondria and a corresponding intramembraneous space in the plasma membrane (4). These three complete biological membranes containing N2 and N6 leaflets as well as an I4 space together with their polarity are illustrated in Fig. 7. The nature and functions of the inter- membraneous space with its associated enzymes and solutes is interesting in view of the fact that this space containing adenylate kinase and NDP Upsala J Med Sci 81 198 Erik Cervkn kinase enlarges during oxidative phosphorylation in rat liver mitochondria (21). Similar fluctuations of the I4 spaces of other biological membranes could be expected to be associated with their characteristic functions. The demonstration that the endoplasmatic cisternae also contain adeny- late kinase and NDP kinase in vivo will support generalizations facilitating the interpretation of phenomena associated with any of these three complete biological membranes by their mutual analogy. DISCUSSION This model for biological membranes offers the possibility of a new approach to several plasma membrane phenomena. The involvement of GTP in adenylyl cyclase activity (39) is one example. The enhanced transport of amino acids across the nuclear N2 leaflet brought about by exogenous ATP (27) may according to this hypothesis cor- respond to similar phenomena in the I4 space and N2 leaflet of the plasma membrane. Furthermore, various effects of treating intact cells with phospholipase C and other exogenous enzymes may be interpreted in terms of increased exposure of the I4 space and the N2 leaflet of the plasma membranes. The finding that certain amino acids apparent- ly are deprotonated during transport across the plasma membrane (1 1) may indicate that under certain conditions the I4 space is capable of main- taining a concentration of univalent cations, dif- ferent from that on either side of the enclosing PNP-leaflets and may have important implica- tions for the function of the mitochondria and the endoplasmatic reticulum. In the former case, swelling of the I4 spaceiscorrelated with oxidative phosphorylation and in the latter case inhibition of protein synthesis by puromycin is known to cause swelling of the endoplasmatic cisternae (20). The possibility that carriers generate a sodium gradient primarily across the N2 leaflet is interesting in view of the knowledge that inward transport of many metabolites is correlated with comigration of sodium (10). Due to the fact that sodium participates in an electrochemical gradient Upsala J Med Sci 81 (lo), such an arrangement would be possible to detect as a transient positive potential when in- serting an electrode into a cell. If sodium-depend- ent transport of solutes occurs primarily across the N2 leaflet, this transient positive potential would be detected primarily in cells that are characterized by a high rate of sodium-dependent transport and by an N6 leaflet that is semiperme- able to sodium ions. The compartment idea for the subcellular loca- tion of macromolecules and the idea of vesicles as the unique mediators of transmembraneous transport of macromolecules are concerned by this interpretation of the structure of biological membranes. Several apparently unrelated obser- vations dealing with, for example, the location of serum proteins in the nucleus (55) and the leakage of cytoplasmic proteins from intact cells (23, 32, 54)may be accounted for by assuming that the nuclear and plasma membranes and probably other intracellular membranes as well are perme- able to certain macromolecules by diffusion as implied in this membrane model. This permeabi- lity, is an intersting weakness of tumor cells be- cause of the possibility of introducing macro- molecular agents that are harmful for their proli- ferative metabolism. ACKNOWLEDGEMENT This work was supported by grants from the Swedish Medical Research Council (project B77- 13X-228 -13B), The University of Uppsala and a personal grant from Uddeholms AB. I would like to thank Drs. Gunnar Agren and Gunnar Ronquist for valuable criti- cizm. REFERENCES 1. Agren, G., Ponten, J., Ronquist, G. & Wester- mark, B.: Formation of extracellular adenosine triphosphate by normal and neoplastic human cells in culture. J Cell Physiol 77: 331-335, 1971. 2. - Nucleoside Diphosphate Kinase at the Cell Surface of Neoplastic Human Cells in Culture. J Cell Physiol 83: 91-101, 1974. 3. - Demonstration of an ATPase at the cell surface of intact normal and neoplastic human cells in culture. J Cell Physiol 78: 171-176, 1971. The fluid double PJ 4. Agren, G. & Ronquist, G.: On the Availability of Certain Metabolites at the Outer Surface of Normal and Malignant Cells for the Membranous de Novo synthesis of ATP and Other Nucleotides. Upsala J'Med Sci 8 0 : 1-4, 1975. 5 . - Formation of extracellular adenosine triphos- phate by tumour cells. Acta Physiol Scand 7 5 : 6. - Isolation of "P-Labelled Phosphorylserine and Phosphorylthreonine from Ehrlich Mouse-ascites Tumour Cells, Suspended in an isotonic Medium, Containing '2P-Labelled Nucleoside Triphos- phates or Inorganic Pyrophosphates. Acta Chem Scand 2 5 : 2931-2934, 1971. 7. Bosmann, H.B.: Molecules at the External Nuclear Surface. J Cell Biol 59: 601-614, 1973. 8. Bosmann, H.B., Myers, M.W., Dehond, D., Ball, R. & Case, K.R.: Mitochondria] Autonomy. Sialic Acid Residues on the Surface of Isolated Rat Cerebral Cortex and Liver Mitochondria. J Cell Biol 55: 147-160. 1972. Characteristics of Lipids and Cell Membranes. Quarterly Reviews of Biophysics 8, 2 : 185-235, 1975. 10. Christensen, H.N.: Biological Transport, second edition, W.A. Benjamin, Inc., London, Amster- dam, Don Mills, Ontario, Sydney, Tokyo, 1975. 11. Christensen, H.N. & Handlogten, M.E.: A Cycle of Deprotonation and Reprotonation Energizing Amino-Acid Transport? Proc Nat Acad Sci USA 12. Clay, W.F., Katterman, F.R.H., & Bartels, P.G.: Chromatin and DNA Synthesis Associated with Nuclear Membrane in Germinating Cotton. Proc Nat Acad Sci USA 72: 3134-3138, 1975. 13. Comings, D.E. & Kakefuda, T.: Initiation of Deoxyribonucleic Acid Replication at the Nuclear Membrane in Human Cells. J Mol Biol 33: 14. Cuatrecasas, P.: Insulin Receptor of Liver and Fat Cell Membranes. Fed Proc, vol 32, 8: 1838- 1846, 1973. 15. Duchon, G. & Collier, H.B.: Enzyme Acitivities of Human Erythrocyte Ghosts. J Membrane Biol 16. Dunkley, C.R., Manery, J.F. & Dryden, E.F.: The conversion of ATP to IMP by Muscle Surface Enzymes. J Cell Physiol 68: 241-247, 1966. 17. Elbers, P.F.: The Cell Membrane: Image and In- terpretation. In Recent Progress in Surface Science, Vol 2 (eds. J.F. Danielli, K.G.A. Pank- hurst, and A.C. Riddiford) pp. 443-503, 1964. 18. Fleischer, S., Fleischer, B. & Stoeckenius, W.: Fine Structure of Whole and Fragmented Mito- chondria after Lipid Depletion. Fed Proc 2 4 : 296, 1965. 124-128, 1969. 9. Chapman, D.: Phase Transitions and Fluidity 7 2 : 23-31, 1975. 225-229, 1968. 6: 138-157, 1971. poteolipid leaflet model for biological membranes 199 19. Gezelius-Kjallman, C., Ronquist, G. & Agren, G.: Membrane-bound Inorganic Pyrophosphatase of Human Erythrocytes. Acta Chem Scand 25: 20. Goldblatt, P.J.: The Endoplasmic Reticulum. I n Handbook of Molecular Cytology (ed. A. Lima- de-Faria) pp. 1101-1 129. North-Holland Publish- ing Company, Amsterdam & London 1972. 21. Hackenbrock, C.R.: Ultrastructural Bases for Metabolically Linked Mechanical Activity in Mitochondria, I, 11. J Cell Biol 3 0 : 269-297, 1967, J Cell Biol 37: 345-369, 1968. 22. Hall, M.R., Meinke, W., Goldstein, D.A. & Ler- ner, R.: Stynthesis of Cytoplasmic Membrane- associated DNA in Lymphocyte Nucleus. Nature New Biology 234: 227-229. 23. Hallak, G.J. & Wilkinson, J.H.: Membrane Permeability in Normal Human Lymphocytes and Lymphocytes from Patients with Chronic Lym- phatic Leukaemia. Clin Chim Acta 69: 341-349, 1976. 24. Hynes, R.O.: Cell Surface Proteins and Malignant Transformation. Biochim Biophys Acta 458: 25. Jenkins, R.E. & Tanner, J.A.: The Major Human Erythrocyte Membrane Protein. Biochem J 147: 26. Johnson, R.M.: The Kinetics of Resealing of Washed Erythrocyte Ghosts. J Membrane Biol 27. Karjalainen, E.: Amino Acid Transport into Cell Nuclei, Doctoral Dissertation, The University of Helsinki, 1969. 28. Korn, E.D.: Structure of Biological Membranes. Science 153: 1491-1498, 1966. 29. Korn, E.D., Bowers, B., Batzri, S., Simmons, S.R. & Victoria, E.J.: Endocytosis and Exocytosis: Role of Microfilaments and Involvement of Phos- pholipids in Membrane Fusion. J Supramol Struct 30. Laki, K.: Actin as an Energy Transducer. J Theor Biol 4 4 : 117-130, 1974. 31. Laki, K. & Ladik, J.: Protein Energy Converters. NATO Adv Study Inst Ser B: 681-98, 1975. 32. Lampen, J.O.: Movement of Extracellular En- zymes across Cell Membranes. In Symposia of the Society for Experimental Biology, Number XXVIII: Transport at the Cellular Level: pp. 35 1-374 Cambridge: At the University Press, London 1974. 33. Lehninger, A.L.: Mitochondria] Compartmenta- tion and the Integration of Energy-Yielding Reac- tions. I n Biochemistry pp. 395-416 Worth Pub- lishers, Inc., N.Y. 1970. 34. Lestourgeon, W.M., Foser, A., Yang, Y., Bertram, J.S. & Rusch, H.P.: Contractile Proteins. Major Components of Nuclear and Chromosome Non- Upsala J Med Sci 81 1457-1460, 1971. 73-107, 1976. 393-399, 1975. 22: 231-253, 1975. 2 : 517-528, 1974. 200 Erik Cervkn Histone Proteins. Biochim Biophys Acta 379: 35. Marchesi, V.T. & Furthmayr, T.: The Red Cell Membrane. In Annual Reveiw of Biochemistry, Vol 45 (ed. E.E. Sne1l)pp. 667-698 Annual Rev Inc., Palo Alto, California 1976. 36. Mayhew, E.: Interaction of Polynucleotides with Cultured Mammalian Cells. Exptl Cell Res 86: 37. Nicolson, G., Lacorbiere, M. & Delmonte, P.: Outer Membrane Terminal Saccharides of Bovine Liver Nuclei and Mitochondria. Exp Cell Res 71: 38. Palade, G.: Intracellular Aspects of the Process of Protein Synthesis. Science 189: 347-358, 1975. 39. Pfeuffer, T. & Helmreich, E.J.M.: Activation of Pigeon Erythrocyte Membrane Adenylate Cyclase by Guanylnucleotide Analogues and Separation of a Nucletodie Binding Protein. J Biol Chem 250: 40. Robinson, G.A.R., Butcher, R.W. & Sutherland, E.W.: Cyclic AMP. Ann Rev Biochem 37: 149- 174, 1968. 41. Ronquist, G.: Studies on the formation of adeno- sine triphosphate by the cell membrane of human erythrocytes and of tumor cells. Acta Universitatis Upsaliensis, Abstr Upps Dissert Medicine, No. 5 4 : 1 , 1968. 42. - Formation of extracellular adenosine triphos- phate by human erythrocytes. Acta Physiol Scand 43. Ronquist, G . & Agren, G.K.: A Mg2'- and Caz+- stimulated Adenosine Triphosphatase at the Outer Surface of Ehrlich Ascites Tumor Cells. Cancer Res 35: 1402-1406, 1975. 44. - Isolation of '*P-Labelled Phosphorylserine and Phosphorylthreonine from Ehrlich Mouse Ascites Tumor Cells Suspended in Different Isotonic Media Containing 32P-Labelled Adenosine Tri- phosphate. Acta Chem Scand B 28: 1169-1 174, 1974. 45. Schachter, H.: The Subcellular Sites of Glycosy- lation. Biochem SOC Symp 4 0 : 57-71, 1974. 46. Singer, S.J. & Nicolson, G.L.: The Fluid Mosaic Model of the Structure of Cell Membranes. Science 175: 720-731, 1972. 47. Sjostrand, F.S.: A Comparison of Plasma Mem- brane, Cytomembranes, and Mitochondrial Mem- brane Elements with Respect to Ultrastructural 529-552, 1975. 81-94, 1974. 468-473, 1972. 867-876, 1975. 7 4 : 594-605, 1968. Features. J Ultrastruct Res 9: 561-580, 1963. 48. - The Problems of Preserving Molecular Struc- ture of Cellular Components in Connection with Electron Microscopic Analysis. J Ultrastruct Res 49. Sjostrand, F.S. & Barajas, L.: Effect of Modifica- tions in Conformation of Protein Molecules on Structure of Mitochondria1 Membranes. J Ultra- struct Res 25: 121-155, 1968. 50. Sjostrand, F.S. & Kretzer, F.: A New Freeze- Drying Technique Applied to the Analysis of the Molecular Structure of Mitochondria1 and Chloro- plast Membranes. J Ultrastruct Res 53: 1-28, 1975. 5 1. Smith, V., Ryan, J.W. & Smith, D.S.: Freeze-etch Studies of the Plasma Membrane of Pulmonary Endothelial Cells. J Cell Biol 56: 492-499, 1973. 52. Stevens, B.J. & Andrk, J.: The Nuclear Envelope. In Handbock of Molecuiar Cytology (ed. A. Lima-de-Faria) pp. 837-87 1 North-Holland Publishing Company, Amsterdam & London 1972. 53. Swift, H. & Wolstenholme, D.R.: Mitochondria and Chloroplasts: Nucleic Acids and the Problem of Biogenesis (Genetics and Biology). In Hand- book of Molecular Cytology fed. A. Lima-de- Faria) pp. 972-1046 North-Holland Publishing Company, Amsterdam & London 1972. 54. Wernstedt, C., Agren, G. & Ronquist, G.: En- zyme Activities at the Surface of Intact Ehrlich Mouse-ascites Tumor Cells with Albumin in the Isotonic Medium. Cancer Res: 35: 1536-1541, 1975. 5 5 . Zardi, L., Ziri, A. & Santi, L.: A Serum Protein Associated with Chromatin of Cultured Fibro- blasts. Science 191: 869-870, 1976. 55: 271-280, 1976. Received November 15, 1976 Address for reprints: E. Cerven Institute of Medical and Physiological Chemistry Biomedical Center University of Uppsala Box 575 S-751 23 Uppsala Sweden Upsala J Med Sci 81