Distribution of ferritin-conjugated lectins on sialidase-treated membranes of human erythrocytes

The labeling of sialidase-treated, human erythrocyte membranes with ferritin-conjugates of four plant lectinss, concanavalin A, Ricinus communis hemagglutinin, Bauhinia purpurea hemagglutinin and Arachis hypogoea hemagglutinin, is reported. Among these ferritin-conjugated lectins, ferritin-conjugated concanavalin A and ferritin-conjugated R. communis hemagglutinin were found in clusters on the sialidase-treated membranes, whereas ferritin-conjugated B. pupurea hemagglutinin and ferritin-conjugated A. hypogoea hemagglutinin were found in a random distribution on the membranes. Furthermore, when the membranes were labeled with a mixture of concanavalin A and ferritin-conjugated B. purpurea hemagglutinin, ferritin particles were found in clusters, indicating that the membrane receptors for B. purpurea hemagglutinin were forced to move together with those for concanavalin A. A method for thentitative estimation of the clustering of ferritin particles on the membranes was also devised and applied to the labeling of sialidase-treated, human erythrocyte membranes with the above four ferritin-conjugated lectins.     

A statistical evaluation of the binding of ferritin-conjugated lectins to the surface of rat cells. Topographical variations as a function of lectin concentration and cell type.

Statistical procedures were utilized to evaluate the concentration dependence of labeling by ferritin-conjugated lectins on four different rat cells: hepatocytes, normal thymocytes, Friend virus-induced rat tumor cells and feline sarcoma virus-induced rat sarcoma cells. Labeling by ferritin conjugates of concanavalin A, wheat germ agglutinin and Ricinus communis agglutinins I and II was quantitated by counting the number of ferritin granules on 600 Angstrom membrane segments. Relationships between the arithmetic means and variances for sample populations from each cell and ferritin-lectin combination were used to define four types of topographical distributions: uniform/ordered, uniform/random, random and clustered. It was found that the distribution and/or density of surface-bound lectin was concentration-dependent for all four ferritin-lectins. The nature of this dependency was complex and varied with both lectin and cell type.     

Localization of the binding sites for the Ricinus communis, Agaricus bisporus and wheat germ lectins on human erythrocyte membranes.

Freeze-etch electron microscopy has been utilized to localize the binding sites for the Ricinus communis, Agaricus bisporus and wheat germ lectins on human erythrocyte membranes and to determine the relation of these different glycoprotein receptors to the intramembranous particles. A. bisporus lectin, which could be visualized directly on the surface of erythrocyte membranes, and ferritin conjugates of wheat germ agglutinin showed a distribution that correlates exactly with the intramembranous particles at all lectin concentrations tested. The binding sites for both of these lectins are located on the major sialoglycoprotein of the membrane. The R. communis agglutinin-ferritin conjugate which binds to receptors on membrane glycoproteins that are distinct from the major sialoglycoprotein showed a close correlation with the intramembranous particles at low lectin concentrations and a poor correlation at high lectin concentrations. High concentrations resulted in virtually complete coating of the surface of trypsinized ghosts which displayed marked aggregation of the intramembranous particles. We conclude that the intramembranous particles of erythrocyte membranes contain at least two glycoproteins and that some membrane lectin receptors are not associated with the intramembranous particles.     

A method for the direct demonstration of the lectin-binding components of the human erythrocyte membrane.

1. A method which allows the characterization of lectin-binding components is described. This method should be useful in defining the nature and heterogeneity of these components in cell membranes. 2. The method, which we have used on erythrocyte "ghosts", involves the fixation of "ghost" components after sodium dodecyl sulphate/polyacrylamide-gel electrophoresis and incubation with purified 125I-labelled lectins. 3. Each of the four lectins used shows an individual pattern of reactivity towards "ghosts" components. Band 3, the major membrane-penetrating glycoprotein, is bound by the lectins from Ricinus communis and Phaseolus vulgaris (phytohaemagglutinin) and by concanavalin A. The major erythrocyte sialoglycoprotein is bound by the lectins from R. communis, P. vulgaris and Maclura aurantiaca. 4. Three of the lectins displays binding for other membrane components, some of which are not demonstratable by conventional protein- and carbohydrate-staining techniques.     

Distribution and nature of saccharides of plasma membranes of rat ascites hepatoma cells as detected by ferritin-conjugated lectins and dialysed iron

Electron microscopic cytochemical studies were made on saccharides involved in the plasma membranes of rat ascites hepatoma cells (AH7974F) using ferritin-conjugated lectins and dialysed iron (DI). In the rat hepatoma cells, saccharide receptors for each of the three lectins used (concanavalin A (ConA), wheat germ agglutinin (WGA) and Ricinus communis agglutinin (RCA)) were shown to be distributed homogeneously throughout the plasma membranes. When the cells were agglutinated, however, the saccharide receptors for each lectin appeared to form clusters on the plasma membranes. The cluster formation induced by one lectin was found to lead to a changed distribution of saccharide receptors for another lectin. None of the cluster formation types induced by lectins yield any noticeable effects upon the distribution of DI reactive acidic saccharides on the plasma membranes.     

SOME SURFACE CHARACTERISTICS OF THE CHLOROPLAST ENVELOPE

Pronase, cationic ferritin, and ferritin-conjugated plant lectins were used to study the chloroplast envelope. Negative charges (binding cationic ferritin) are fairly uniformly distributed over the envelope surfaces in contact with the hyaloplasm and are not appreciably altered by mild pronase treatment of isolated plastids. All surfaces of stroma-free thylakoids previously exposed to the stroma uniformly bind cationic ferritin. Ricin_II-ferritin binding to the membranes of the chloroplast envelope indicates that galactolipids are distributed in the outer membrane in such a way that their galactose moieties are exposed on the envelope surface. In addition, the outer surface of the inner membrane (the intermembrane face) contains uniformly distributed galactose which binds ricin_II when this membrane is exposed to the reaction medium. Isolated vesicles of the chloroplast envelope bind ricin_II, while isolated envelope vesicles as well as the envelopes of intact chloroplasts failed to bind concanavalin A. Thylakoid surfaces showed minor binding of ricin_II as well as concanavalin A.     

Localization of globoside and Forssman glycolipids on erythrocyte membranes

Using the freeze-etch technique, the membrane localization of globoside, a principal glycolipid in human erythrocytes, and Forssman antigen, the chief glycolipid in sheep erythrocytes was evaluated using ferritin and colloidal gold as morphological markers for rabbit antibodies prepared against these glycolipids. Brief trypsinization of human red cell ghosts markedly aggregated intramembranous particles and permitted labeling of globoside, which appeared in a clustered arrangement. The aggregates of ferritin-anti-globoside differed from those of ferritin-wheat germ agglutinin, a label for glycophorin, which corresponded with the aggregates of intramembranous particles. Double-labeling of human trypsinized ghosts with anti-globoside/ Staphylococcal protein A-colloidal gold and ferritin-wheat germ agglutinin indicated that the patterns of labeling were different and that the aggregates of globoside did not bear a direct relationship to the intramembranous particles, which represent transmembrane proteins. Resealed sheep erythrocyte ghosts labeled with ferritin-conjugated rabbit anti-Forssman showed small clusters of Forssman glycolipid on the erythrocyte surface, which could be markedly aggregated with a second goat anti-rabbit antibody, indicating relative mobility of the small glycolipid domains. The distribution of ferritin-anti-Forssman label in sheep ghosts treated at pH 5.5 to aggregate intramembranous particles also did not show definite correspondence between intramembranous particles and the clusters of ferritin-anti-Forssman.     

Immunoelectron microscope localization of cytochrome P-450 on microsomes and other membrane structures of rat hepatocytes

Localization of cytochrome P-450 on various membrane fractions of rat liver cells was studied by direct immunoelectron microscopy using ferritin-conjugated antibody to the cytochrome. The outer surfaces of almost all the microsomal vesicles were labeled with ferritin particles. The distribution of the particles on each microsomal vesicle was usually heterogeneous, indicating clustering of the cytochrome, and phenobarbital treatment markedly increased the labeled regions of the microsomal membranes. The outer nuclear envelopes were also labeled with ferritin particles, while on the surface of other membrane structures such as Golgi complexes, outer mitochondrial membranes and plasma membranes the labeling was scanty and at the control level. The present observation indicates that cytochrome P-450 molecules are localized exclusively on endoplasmic reticulum membranes and outer nuclear envelopes where they are probably distributed not uniformly but heterogeneously, forming clusters or patches. The physiological significance of such microheterogeneity in the distribution of the cytochrome on endoplasmic reticulum membranes is discussed.     

The preparation of rat liver lysosome membranes. Several membrane proteins contain complex oligosaccharides

Lysosome membranes were isolated, and membrane proteins and glycoproteins were characterized by electrophoresis and lectin probes of nitrocellulose blots. Rat liver lysosomes were isolated on a discontinuous metrizamide gradient and characterized by subcellular marker enzymes. Lysosomes were lysed by hypotonic freeze-thaw shock and membranes were isolated. The release of beta-N-acetylhexosaminidase was used to monitor the disruption of the lysosomes. Proteins of lysosome membranes were analyzed by sodium dodecyl sulfate - polyacrylamide gel electrophoresis. There were at least 30 proteins present and several were glycoproteins. Nitrocellulose blots of lysosome membrane proteins were probed with a panel of lectins, including concanavalin A, Ulex europaeus agglutinin I, Lotus tetragonolobus agglutinin, soybean agglutinin, peanut agglutinin, and Ricinus communis agglutinin I. Peanut agglutinin and Ricinus communis agglutinin I binding were also examined after neuramidase treatment of lysosome membranes. Ten proteins bound concanavalin A, and neuraminidase pretreatment revealed six proteins that bound Ricinus communis agglutinin I and three proteins that bound peanut agglutinin. The other lectins tested did not bind to any lysosome membrane proteins. These results indicate that lysosome membranes contain several glycoproteins, some of which contain sialic acid terminating complex oligosaccharides.     

Localization of the binding sites for the Ricinus communis, Agaricus bisporus and wheat germ lectins on human erythrocyte membranes

Freeze-etch electron microscopy has been utilized to localize the binding sites for the Ricinus communis, Agaricus bisporus and wheat germ lectins on human erythrocyte membranes and to determine the relation of these different glycoprotein receptors to the intramembranous particles. A. bisporus lectin, which could be visualized directly on the surface of erythrocyte membranes, and ferritin conjugates of wheat germ agglutinin showed a distribution that correlates exactly with the intramembranous particles at all lectin concentrations tested. The binding sites for both of these lectins are located on the major sialoglycoprotein of the membrane. The R. communis agglutinin-ferritin conjugate which binds to receptors on membrane glycoproteins that are distinct from the major sialoglycoprotein showed a close correlation with the intramembranous particles at low lectin concentrations and a poor correlation at high lectin concentrations. High concentrations resulted in virtually complete coating of the surface of trypsinized ghosts which displayed marked aggregation of the intramembranous particles. We conclude that the intramembranous particles of erythrocyte membranes contain at least two glycoproteins and that some membrane lectin receptors are not associated with the intramembranous particles.     

Characterization of the isolation membranes and the limiting membranes of autophagosomes in rat hepatocytes by lectin cytochemistry

The isolation membranes and the limiting membranes of autophagosomes in rat hepatocytes were characterized by lectin cytochemistry using concanavalin A (ConA), Ricinus communis agglutinin 120 (RCA-120), and wheat germ agglutinin (WGA). We found that RCA-120, ConA, and WGA bind to these membranes. The distribution of the lectins on the isolation membranes was heterogeneous, mainly found on the rims, which we referred to as the peripheral dilated portion. When the rims fused and thus formed autophagosomes the apparent sites of fusion were strongly labeled by the lectins. After autophagosomes were transformed to autolysosomes by fusion with lysosomes, the limiting membranes became more densely and homogeneously labeled with the lectins. We previously reported that cytochrome P-450 does not exist on the limiting membranes of the autophagosomes. Taken together, these results suggest that the isolation membranes may originate not from endoplasmic reticulum membranes but from some post-Golgi membranes that contain complex type N and/or O-linked oligosaccharide chains.     

Nature of the lectin-induced activation of plasma membrane Mg2+ATPase.

The Mg2+ATPase activity of liver plasma membranes decreases markedly with increasing temperature above 30 degrees. This negative temperature dependency is counteracted by the binding of wheat germ agglutinin, concanavalin A, or Ricinus communis agglutinin (at concentrations greater than or equal 0.5 mg/ml) to membranes prior to assay of the enzyme. With one of these lectins bound, the enzyme has a single energy of activation between 20 degrees and 45 degrees. The binding of dimeric succinyl concanavalin A, soybean agglutinin, fucose-binding lectin from Lotus tetragonolobus, or the leucoagglutinin from Phaseolus vulgaris does not alter the temperature dependency of the enzyme. The latter two lectins, however, do prevent the concanavalin A-induced activation of the enzyme at 37 degrees. At saturating substrate concentrations, the enzyme is not inhibited by any of the lectins tested over a wide range of concentrations. Cytochalasin B and colchicine separately or in combination have little influence on the lectin-induced enhancement of enzyme activity. Chlorpromazine and vinblastine sulfate each partially prevent the activation and in combination do so completely. Treatment of the membranes with the detergent Lubrol-PX or phospholipase A prevents activation of the enzyme by concanavalin A. The results are consistent with a restriction by the lectin of an environment which is normally too disordered for maximal enzyme activity above 30 degrees.     

Modulation of WGA binding sites on marrow sinus endothelium in state of stimulated erythropoiesis: a possible mechanism regulating the rate of cell egress

To study the regulation of cellular and molecular traffic across the marrow-blood barrier, rat marrow endothelial surface was incubated with ferritin-conjugated concanavalin A, wheat germ agglutinin (WGA), recinus communis agglutinin I, and phytohemagglutinin. Normal animals were compared with those after erythropoietic stimulation (phenylhydrazine-induced hemolysis, phlebotomy). A selective and significant reduction in the density of WGA receptors, but not other lectins was noted congruent to the degree of reticulocytosis. Neuraminidase treatment also reduced WGA binding sites and the surface negative charge as detected by polycationic ferritin (PCF). Thus, the reduction in WGA binding sites, may reflect a decrease in the density of membrane sialic acid, rendering the endothelial surface charge less negative and providing an electrostatic attraction for the negatively charged surface of reticulocytes. The findings may also be explained by an increase in the frequency of WGA-excluding fenestrae in the endothelium. These areas, lacking sialic acid, may provide unstable areas in the membrane suitable for the passage of cells and molecules in both directions. It is concluded that, by modulating the density of sialic acid residues, the endothelium may regulate the traffic of cells and molecules across the marrow-blood barrier.     

Cell surface of mouse blastocysts at the trophectoderm-uterine interface during the adhesive stage of implantation

The molecular basis for the acquisition of adhesiveness between blastocysts and uterine luminal epithelium is an interesting problem in reproductive biology. It is rather difficult to study implantation-stage blastocysts of mice because during the implantation period each blastocyst becomes lodged within a crypt formed by decidualizing stroma. After trophectoderm adheres to uterine luminal epithelium, it is not possible to flush intact blastocysts from the uterus by standard recovery procedures. By identifying implantation sites with the Evans blue technique and splitting or gently separating the apposed epithelium of finely trimmed sites, it was possible to expose nonadhesive and adhesive trophectoderm to polycationized ferritin (PCF) and a series of ferritin-conjugated lectins. Examination by transmission electron microscopy revealed that both adhesive and nonadhesive trophectoderm bound PCF, concanavalin A, wheat-germ agglutinin, Ricinus communis agglutinin I, and Limulus polyhemus agglutinin, but not Dolicos biflorus agglutinin or peanut agglutinin. Nonadhesive trophectoderm bound succinylated wheat germ agglutinin but adhesive trophectoderm did not. There was no apparent difference in the relative amounts of each lectin bound to adhering and nonadhering cells.     

A comparison of direct and indirect techniques using ferritin-conjugated ligands for the localization of concanavalin A binding sites on isolated hepatocyte plasma membranes

Concanavalin A binding sites have been localized on isolated plasma membranes both by a direct technique involving ferritin-concanavalin A and by an indirect technique in which membranes were treated successively with concanavalin A, rabbit anti-concanavalin A, and ferritin-conjugated sheep anti-rabbit F(ab')2. Binding studies showed that, at saturation, less than 25% of the concanavalin A binding sites were accessible to ferritin-concanavalin A. The decreased binding was apparently related to steric factors, since membranes saturated with the conjugated ligand were able to bind additional concanavalin A, and since the conjugated ligand, once bound to the membrane, caused the same inhibition of the membrane-bound enzyme 5'-nucleotidase as concanavalin A. Nonspecific binding sites accounted for 10% of the total binding of ferritin-concanavalin A and were localized mainly on the cytoplasmic side of the membrane, whereas specific sites were on the external side. The indirect technique, which was expected to increase the binding of ferritin-conjugate to the membrane, resulted in the binding of ferritin to less than 15% of the concanavalin A binding sites, and did not decrease the nonspecific binding.     

Receptor distribution and the mechanism of enhanced erythrocyte agglutination by soybean agglutinin

We have examined the role of receptor clustering in intact erythrocyte membranes exhibiting enhanced lectin-mediated cell agglutination by analyzing freeze-fracture and freeze-etch images of human erythrocytes labeled with ferritin-conjugated soybean agglutinin. We find that trypsinization and fixation of intact erythrocytes, in either order, causes no alteration of the random distribution of ferritin-conjugated soybean agglutinin on the surfaces of these cells as compared to their distribution on the surfaces of fixed erythrocytes and untreated erythrocyte ghosts. Furthermore, clustering of the intramembranous particles in the membrane of intact erythrocytes was not found with any of the cells described above.We conclude that clustering of the soybean agglutinin receptors is not a major factor involved in the enhanced agglutination of intact trypsinized erythrocytes. Caution is necessary in transferring information obtained with erythrocyte ghosts, where clustering can be induced, to intact erythrocytes.     

CHARACTERIZATION AND TOPOGRAPHY OF THE GLYCOPROTEINS OF ADRENAL CHROMAFFIN GRANULES

The glycoproteins of the membranes of bovine chromaffin granules were characterized by two polyacrylamide gel electrophoresis systems. Five components (I-V) were demonstrated with apparent molecular weights ranging in the unreduced form from 45,000 to 150,000. Glycoprotein I was identified as the enzyme dopamine β-hydroxylase. Four of these glycoproteins (with the exception of component IV) were apparently also present in the membranes of pig and horse chromaffin granules. The soluble proteins of chromaffin granules contained at least three glycoproteins. Only glycoprotein I (dopamine β-hydroxylase) was present both in the soluble content and in the membranes of chromaffin granules. Affinity chromatography with lectins demonstrated that from the soluble proteins only dopamine β-hydroxylase was adsorbed by concanavalin A, whereas none of these proteins reacted with wheat germ lectin and Ricinus communis agglutinin. Three membrane proteins including dopamine β-hydroxylase and glycoprotein II as major components were adsorbed by concanavalin A, whereas wheat germ lectin bound only component II and a small amount of component III. By electron microscopy it was demonstrated that concanavalin A did not bind to intact chromaffin granules whereas ruthenium red and cationized ferritin did. Isotope labelling after galactose oxidase treatment revealed that at least the carbohydrate portion of the major glycoproteins is present on the inner side of the granule membranes facing the content.     

Changes of lectin-binding sites on the embryonic muscle cell surface in the developing ascidian, Halocynthia aurantium

Lectin-binding sites on the muscle cell surface of an ascidian embryo were studied using the ferritin labeling technique. The embryos at 4-cell, gastrula, late tail-bud, and larval stages were dissociated in the Ca²⁺- and Mg²⁺-free solution with or without collagenase. Dissociated cells and fragments were prefixed, reacted with ferritin-lectin conjugates and processed for electron microscopy. Lectins used were concanavalin A (ConA) and Ricinus communis agglutinin. Ferritin particles showing lectin-binding sites were found singly or in the form of clusters on the cell surface exposed directly to the conjugates. Most of the particles of both conjugates were distributed singly and sparsely on entire surface areas of the 4-cell stage cells, whereas they were rich in population and tended to form clusters when embryos reached the gastrula stage. At the succeeding stages tested, tagged ferritin, which was single or clustering particles, was less in number as compared with those at the former stage; on the surface facing neighboring muscle cells, in particular, the ferritin particles were much fewer than those in areas of notochordal and epithelial sides. It is suggested that the embryonic muscle cells of the ascidian show stage-specific changes of cell surface carbohydrates. They have high reactivity to both lectins around the gastrula stage and bring out the regional difference of both lectin-binding sites in the tail-bud stage, namely during the period of histogenesis.     

Interaction of human erythrocyte Band 3 with Ricinus communis agglutinin and other lectins

Agglutination and competition studies suggest that human erythrocyte Band 3 can interact with both mannose/glucose- and galactose-specific lectins. Purified Band 3 reconstituted into lipid vesicles binds concanavalin A, but the nonspecific binding component, measured in the presence of alpha-methylmannoside, is very high. This glycoprotein also carries binding sites for the galactose-specific lectin Ricinus communis agglutinin. Binding was inhibited poorly by lactose, but much more effectively by desialylated fetuin glycopeptides, suggesting that the lectin recognizes a complex oligosaccharide sequence on Band 3. The glycoprotein bears two separate classes of binding sites for R. communis agglutinin. High-affinity binding sites exist which show strong positive cooperativity and correspond in number to the outward-facing Band 3 molecules. A low-affinity binding mode is abolished by 40% ethyleneglycol, suggesting the involvement of hydrophobic lectin-glycoprotein interactions. Studies on binding of R. communis agglutinin to human erythrocytes indicate positively cooperative binding to 7 X 10(5) very-high-affinity sites per cell, and lectin binding is completely inhibitable by lactose. Based on its binding characteristics in vesicles, it seems likely that Band 3 forms the major receptor for this lectin in human erythrocytes. Properties such as positive cooperativity thus appear to be a common feature of the interaction of Band 3 with a variety of lectins of different specificity, both in erythrocytes and lipid bilayers.     

A novel approach to the identification of surface receptors. The use of photosensitive hetero-bifunctional cross-linking reagent.

Methyl 4-azidobenzoimidate, a photosensitive hetero-bifunctional cross-linking reagent, was synthesized and characterized. This reagent has an imidoester at one end, which reacts spontaneously with primary amines, and an arylazide at the other end, which reacts with a variety of chemical groups upon photolysis by ultraviolet radiation. The reagent molecules were attached to concanavalin A by reactions between imidoester groups of the reagents and free amino groups of the lectin. These activated lectins were purified on a Sephadex G-25 column and showed the binding affinity to an affinity column, glucosylated Sepharose, and to the human erythrocyte ghost membrane. The activated lectins were incubated with the membranes and then unbound lectins were removed by washing. The lectins bound to receptors in the membranes were irradiated with a shortwave ultraviolet lamp to photolyze arylazides attached to the lectins, thus cross-linking the lectins and receptors together. Then the membranes were solubilized and electrophoresed. On gels, the intensity of the lectin receptor band diminished slightly and concomitantly a new band of a higher molecular weight appeared. When 125I-labeled concanavalin A was used, the new band contained the radioactivity. The extent of the appearance of the new band and the decrease of the receptor band were reduced significantly when the ultraviolet irradiation was omitted or the activated lectins were incubated with the membranes in the presence of the lectin inhibitor, alpha-methylmannoside. The irradiation of nonactivated, receptor-bound concanavalin A did not cause those changes. When the activated lectins alone were irradiated with ultraviolet, the band of the lectin dimer appeared whereas nonirradiated lectins appeared mostly as monomers. It is concluded that a small fraction of the activated lectins were cross-linked to receptors in the membrane upon photolysis. In this study, only 8 reagent molecules were attached to a tetramer of the lectin, compared with the presence of approximately 40 available free amino groups. The efficiency of such cross-links of ligands to receptors may be increased by employing longer versions of the hetero-bifunctional cross-linking reagents and also by attaching more of the reagent molecule to ligands.