Biogenesis of plasma membrane glycoproteins. Purification and properties of two rat liver plasma membrane glycoproteins

As a preliminary to a study of the biogenesis of individual plasma membrane glycoproteins, the marker enzyme nucleotide pyrophosphatase (NPPase) and a major rat liver plasma membrane sialoprotein, subsequently found to be identical with the enzyme dipeptidyl peptidase IV (DPP IV), were purified 10,000- and 2,000-fold, respectively, from rat liver. Both were amphipathic proteins which formed defined micellar complexes with detergents and aggregated in their absence. Gel filtration, sucrose density gradient centrifugation, and polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate showed the Triton X-100 complex of NPPase to contain a single 150,000-dalton peptide, while that of DPP IV was composed of two 120,000-dalton subunits; each complex also contained about 150,000-dalton Triton X-100. Trypsin cleaved the detergent complexes with release of major hydrophilic fragments which no longer bound detergent micelles; the accompanying change in peptide size was small for NPPase and undetectable for DPP IV, which also retained the dimer structure of its native form. DPP IV was the only major glycoprotein in rat liver plasma membrane which bound strongly to wheat germ agglutinin. Monospecific rabbit antibodies against NPPase and DPP IV precipitated the antigens without affecting their enzymatic activities.     

Identification of two lysosomal membrane glycoproteins

Two murine lysosome-associated membrane proteins, LAMP-1 of 105,000-115,000 D and LAMP-2 of 100,000-110,000 D, have been identified by monoclonal antibodies that bind specifically to lysosomal membranes. Both glycoproteins were distinguished as integral membrane components solubilized by detergent solutions but not by various chaotropic agents. The lysosome localization was demonstrated by indirect immunofluorescent staining, co-localization of the antigen to sites of acridine orange uptake, and immunoelectron microscopy. Antibody binding was predominantly located at the limiting lysosomal membrane, distinctly separated from colloidal gold-labeled alpha-2-macroglobulin accumulated in the lumen during prolonged incubation. LAMP-1 and LAMP-2 also appeared to be present in low concentrations on Golgi trans-elements but were not detected in receptosomes marked by the presence of newly endocytosed alpha-2-macroglobulin, or in other cellular structures. LAMP-1 and LAMP-2 were distinguished as different molecules by two-dimensional gel analysis, 125I-tryptic peptide mapping, and sequential immunoprecipitations of 125I-labeled cell extracts. Both glycoproteins were synthesized as a precursor protein of approximately 90,000 D, and showed a marked heterogeneity of apparent molecular weight expression in different cell lines. LAMP-2 was closely related or identical to the macrophage antigen, MAC-3, as indicated by antibody adsorption and tryptic peptide mapping. It is postulated that these glycoproteins, as major protein constituents of the lysosomal membrane, have important roles in lysosomal structure and function.     

Platelet plasma membrane lectin activity

The lectin activity of human platelet and erythrocyte membranes was evaluated using trypsinized, formalinized erythrocytes from eight species. Platelet membranes had the greatest lectin activity against cow erythrocytes, but also had significant activity against human, sheep, electric eel, and rabbit erythrocytes. In contrast, erythrocyte membranes only had low lectin activity against electric eel erythrocytes with no activity against the other types of erythrocytes tested. The platelet membrane lectin activity was found to reside in protein molecules on the external surface of the platelet plasma membrane. The lectin activity of platelet membranes was inhibited by amino sugars and some basic amino acids: N-acetylated amino sugars and other neutral sugars were without effect. These results demonstrate that the external surface of the platelet plasma membrane has a specific lectin activity.     

Platelet membrane glycoproteins and their function: an overview

The membrane glycoproteins (GP) of human platelets act as receptors that mediate two important functions, adhesion to the subendothelial matrix and platelet-platelet cohesion, or aggregation. Many of these glycoprotein receptors exist as noncovalently linked heterodimers, including those that belong to the supergene family of adhesion receptors called the integrins. Human platelets contain at least five members of this integrin family, including a collagen receptor (GP Ia-IIa; alpha 2, beta 1), a fibronectin receptor (GP Ic-IIa; alpha 5, beta 1), a laminin receptor (GP Ic'-IIa; alpha 6, beta 1), a vitronectin receptor (VnR; alpha v, beta 3), and a promiscuous, activation-dependent receptor that is thought to be the receptor most responsible for fibrinogen-dependent, platelet-platelet cohesion (GP IIb-IIIa; alpha IIb, beta 3). Some, but not all, of the integrins bind to a tripeptide sequence, arginine-glycine-aspartic acid (RGD), on the adhesive proteins. In addition to the integrins, platelets contain other membrane glyco-proteins: GP Ib-IX, a receptor for von Willebrand factor, which is thought to be the receptor most responsible for platelet adhesion to the subendothelial matrix in a flowing system; GP V, which may be associated with GP Ib-IX and whose function remains unknown; and GP IV (GP IIIb), which functions as a receptor for thrombospondin and collagen.     

Biogenesis of plasma membrane glycoproteins. Tracer kinetic study of two rat liver plasma membrane glycoproteins in vivo

Antibodies to purified nucleotide pyrophosphatase (NPPase) and dipeptidyl peptidase IV (DPP IV) were used to study the biogenesis of these rat liver plasma membrane glycoproteins in vivo. Following injection of tritiated leucine, the radioactivity in NPPase and DPP IV decayed at markedly different rates in the plasma membrane, with apparent half-lives of about 1 and 5 days, respectively. In short term experiments, labeling of total plasma membrane proteins was rapid and insensitive to colchicine, while labeling of both NPPase and DPP IV showed a lag of about 15 min, followed by colchcine-sensitive/cycloheximide-insensitive increases to half-maximal and maximal values at about 1 and 2 h, respectively. A peak of labeled DPP IV in rough microsomes at 15 min showed increased mobility on polyacrylamide gels and was largely inaccessible to antibodies in intact microsomes, consistent with its being an underglycosylated precursor, exposed on the cisternal side of the rough endoplasmic reticulum. In contrast, the behavior of unlabeled DPP IV in preparations of rough microsomes and Golgi was consistent with its being contributed by contaminating right-side-out plasma membrane vesicles. This conclusion was also necessary to fit the tracer kinetic data to a simple membrane-flow model, which gave precursor pools (1 microgram/g of liver) and fluxes (1 microgram/h/g of liver) for both DPP IV and NPPase which were about 3 orders of magnitude less than those for the synthesis of rat serum albumin. Thus, unlike hepatoma tissue culture cells (Doyle, D., Baumann, H., England, B., Friedman, E., Hou, E., and Tweto, J. (1978) J. Biol. Chem. 253, 967-973), normal rat liver does not contain large amounts of preformed intracellular plasma membrane precursors.     

Detection of glycoproteins in the Acanthamoeba plasma membrane

In the present study we have shown that glycoproteins are present in the plasma membrane of Acanthamoeba castellanii by utilizing different radioactive labeling techniques. Plasma membrane proteins in the amoeba were iodinated by 125I-lactoperoxidase labeling and the solubilized radiolabeled glycoproteins were separated by lectin-Sepharose affinity chromatography followed by polyacrylamide gel electrophoresis. The periodate/NaB3H4 and galactose oxidase/NaB3H4 labeling techniques were used for labeling of surface carbohydrates in the amoeba. Several surface-labeled glycoproteins were observed in addition to a diffusely labeled region with Mr of 55,000-75,000 seen on electrophoresis, which could represent glycolipids. The presence of glycoproteins in the plasma membrane of Acanthamoeba castellanii was confirmed by metabolic labeling with [35S]methionine followed by lectin-Sepharose affinity chromatography and polyacrylamide gel electrophoresis.     

Flu virion as a substrate for proteolytic enzymes

The proteolysis of flu virions of the strain A/Puerto Rico/8/34 (subtype H1N1) by enzymes of various classes was studied to develop an approach to the study of the structural organization and interaction of the major protein components of the virion: hemagglutinin (HA), transmembrane homotrimeric glycoprotein, and matrix protein M1 forming a layer under the lipid membrane. Among the tested proteolytic enzymes and enzymic preparations (thermolysin, trypsin, chymotrypsin, subtilisin Carlsberg, pronase, papain, and bromelain), the cysteine proteases bromelain and papain and the enzymic preparation pronase efficiently removed HA ectodomains, while chymotrypsin, trypsin, and subtilisin Carlsberg deleted only a part of them. An analysis by MALDI TOF mass spectrometry allowed us to locate the sites of HA hydrolysis by various enzymic preparations. Bromelain, papain, trypsin, and pronase split the polypeptide chain after the K177 residue located before the transmembrane domain (HA₂ 185–211). Subtilisin Carlsberg hydrolyzed the peptide bond at other neighboring points: after L178 (a major site) or V176. The hydrolytic activity of bromelain measured by a highly specific chromogenic substrate of cysteine proteases Glp-Phe-Ala-pNA was almost three times higher in the presence of 5 mM β-mercaptoethanol than in the presence of 50 mM. However, the complete removal of ectodomains of HA by the high-and low-activity enzyme required identical time intervals. In the absence of the reducing reagent, the removal of HA by bromelain proceeded a little more slowly and was accompanied by significant fragmentation of protein M1. The action of trans-epoxysuccinyl-L-leucylamido)(4-guanidino)butane (E-64), a specific inhibitor of cysteine proteases, and HgCl₂ On the hydrolysis of proteins HA and M1 by bromelain was investigated.     

Flu virion as a substrate for proteolytic enzymes

The proteolysis of flu virions of the strain A/Puerto Rico/8/34 (subtype H1N1) by enzymes of various classes was studied to develop an approach to the study of the structural organization and interaction of the basic protein components of the virion environment: hemagglutinin (HA), transmembrane homotrimeric glycoprotein, and matrix protein M1 forming a layer under the lipid membrane. Among the tested proteolytic enzymes and enzymic preparations (thermolysin, trypsin, chymotrypsin, subtilisin Carlsberg, pronase, papain, and bromelain), the cysteine proteases bromelain and papain and the enzymic preparation pronase efficiently deleted HA ectodomains, while chymotrypsin, trypsin, and subtilisin Carlsberg deleted only a part of them. An analysis by MALDI TOF mass spectrometry allowed us to locate the sites of HA hydrolysis by various enzymic preparations. Bromelain, papain, trypsin, and pronase split the polypeptide chain after the K177 residue located before the transmembrane domain (HA2 185-211). Subtilisin Carlsberg hydrolyzed the peptide bond at other neighboring points: after L178 (a basic site) or V176. The hydrolytic activity of bromelain measured by a highly specific chromogenic substrate of cysteine proteases Glp-Phe-Ala-pNA was almost three times higher in the presence of 5 mM beta-mercaptoethanol than in the presence of 50 mM. However, the complete removal of exodomains of HA, HA, and low-activity enzyme by the HA high- and low-activity enzyme required identical time intervals. In the absence of the reducing reagent, the removal of HA by bromelain proceeded a little more slowly and was accompanied by significant fragmentation of protein Ml1. The action of trans-epoxysuccinyl-L-leucylamido)butane (E-64), a specific inhibitor of cysteine proteases, and HgCl2 on the hydrolysis of proteins HA and M1 by bromelain was investigated.     

Platelet membrane glycoproteins: role in primary hemostasis and component antigens

The biochemical details of the platelet surface as they relate to normal platelet function have been elucidated through study of labeled membranes from both normal platelets and those with congenitially defective function. Several cytoadhesive glycoprotein complexes which are integral components of the platelet membrane have been demonstrated to act as important receptors for extracellular matrix macromolecules. Glycoproteins Ia/IIa (collagen receptor), Ic/IIa (fibronectin receptor), and IIb/IIIa (fibrinogen receptor) belong to a family of cytoadhesive complexes called the integrins, while glycoprotein Ib/IX, the major von Willebrand receptor, has different features. These same major glycoproteins comprise all of the alloantigens and most of the autoantigens that have been characterized. Glycoprotein IIb/IIIa contains the alloantigens, PlA (Zw), Bak (Lek), and Pen (Yuk), as well as the most frequent target antigenic sites for anti-platelet autoantibodies. Because a number of platelet alloantigens were discovered independently by more than one group, nomenclature is confusing at present, although a system analogous to that used for histocompatibility antigens has been proposed. Precise identification of the antigenic epitopes has not yet been accomplished for all of the platelet antigens. Current research efforts include characterization of antigenic epitopes, elucidation of mechanisms by which platelet immunization occurs, and determination of the clinical implications of the presence of various platelet antibodies.     

Identification of plasma membrane glycoproteins involved in the contact-dependent inhibition of growth of diploid human fibroblasts

Growth of normal, nontransformed cells is regulated by the interplay between growth stimulating compounds and growth inhibiting cell-cell contacts. We have previously shown that the growth of normal diploid human fibroblasts is mainly regulated by a specific class of plasma membrane glycoproteins (R. J. Wieser and F. Oesch (1986) J. Cell Biol. 103, 361-367). Because it was found that immobilization of the glycoproteins involved in contact-dependent inhibition of growth is an essential step in the recovery of the biological activity of the glycoproteins, we developed a technique for a first characterization of the active compounds. After SDS-PAGE separation of plasma membrane glycoproteins, they were transferred onto nitrocellulose. The nitrocellulose was cut along the separation track into circles which fit into wells of a 96-well microtiter plate. Culturing human diploid fibroblasts on the nitrocellulose circles resulted in characteristic growth patterns, which were dependent upon the source and the treatment of the plasma membrane proteins which had been separated. Five major inhibitory fractions with apparent molecular masses of 300, 170, 90, 50, and 25 kDa have been identified in plasma membranes from confluent fibroblast cultures.     

Topography of glycoproteins in the chick synaptosomal plasma membrane

Chick brain synaptosomes or synaptic subfractions were treated with neuraminidase (EC 3.2.1.18) and/or galactose oxidase (EC 1.1.3.9) preparations in which proteolytic activity was inhibited with phenylmethanesulfonyl fluoride followed, after washing, by reductive incorporation of sodium boro[³H]hydride to identify galactose residues exposed on the synaptosomal external surface. Control experiments to demonstrate restriction of labeling to the external surface involved comparing the radioactivity in synaptoplasmic, soluble polypeptides isolated after labeling with labeled, isolated synaptoplasm and examining incorporation into fractions incubated without enzymes. Intactness of the synaptic plasma membrane after labeling was shown by trypsin digestion studies. Polypeptides were separated on sodium dodecyl sulfate polyacrylamide gels and were detected by a liquid scintillation counting procedure. Eleven major radioactive peaks were found after galactose oxidase treatment and reduction of isolated synaptic membranes. When intact synaptosomes were labeled, the same components were detected. When isolated synaptic membranes or intact synaptosomes were treated with neuraminidase before galactose oxidase treatment, three additional components were labeled. These results suggest that (a) chick synaptic membranes have a complex mixture of glycoproteins, (b) all major chick synaptic membrane glycoproteins labeled by galactose oxidase have most or all carbohydrate groups exposed at the exterior surface of the synaptosome, (c) all major, externally-disposed polypeptides of these synaptic membranes are glycoproteins.     

Persistent activation of platelet membrane phospholipase C by proteolytic action of trypsin and thrombin

Trypsin causes rapid activation of intact platelets that mimics many actions of thrombin, including the stimulation of phospholipase C (PLC). We have examined the effects of thrombin and trypsin on PLC in a platelet membrane preparation using exogenous [3H]-phosphatidylinositol 4,5-bisphosphate (PIP2) as substrate. Trypsin induced PIP2 breakdown, which was maximal at 20 micrograms/ml, but was reduced at higher concentrations. alpha- and gamma-Thrombins also stimulated PLC-induced hydrolysis of PIP2 in membranes. This effect was inhibited by leupeptin. Exogenous [3H]phosphatidylinositol 4-monophosphate (PIP) was hydrolyzed in response to both thrombin and trypsin in the same ratio as PIP2. Activation of membrane-bound PLC persisted after removal of thrombin and trypsin. The hydrolysis of [3H]phosphatidylinositol was not activated by alpha-thrombin and trypsin. We examined the question of whether calpain was involved in the observed PLC activation by thrombin and trypsin. Although dibucaine activated a Ca2(+)-dependent protease as judged by the hydrolysis of actin-binding protein and by the activation of phosphoprotein phosphatases, it failed to stimulate the generation of phosphatidic acid in 32P-prelabeled platelets. Moreover, when PLC was assayed in the membranes, the addition of Ca2(+)-activated neutral proteinases did not increase the rate of hydrolysis of either PIP or PIP2. Our results show that proteases such as trypsin and thrombin are able to stimulate membrane-bound PLC, but this activation does not seem to be related to calpain.     

Identification of actin-binding protein as the protein linking the membrane skeleton to glycoproteins on platelet plasma membranes

Platelets have previously been shown to contain a membrane skeleton that is composed of actin filaments, actin-binding protein, and three membrane glycoproteins (GP), GP Ib, GP Ia, and a minor glycoprotein of Mr = 250,000. The present study was designed to determine how the membrane glycoproteins were linked to actin filaments. Unstimulated platelets were lysed with Triton X-100, and the membrane skeleton was isolated on sucrose density gradients or by high-speed centrifugation. The association of the membrane glycoproteins with the actin filaments was disrupted when actin-binding protein was hydrolyzed by activity of the Ca2+-dependent protease, which was active in platelet lysates upon addition of Ca2+ in the absence of leupeptin. Similarly, activation of the Ca2+-dependent protease in intact platelets by the addition of a platelet agonist also caused the membrane glycoproteins to dissociate from the membrane skeleton. Affinity-purified actin-binding protein antibodies immunoprecipitated the membrane glycoproteins from platelet lysates in which actin filaments had been removed by DNase I-induced depolymerization and high-speed centrifugation. These results demonstrate that actin-binding protein links actin filaments of the platelet membrane skeleton to three plasma membrane glycoproteins and that filaments are released from their attachment site when actin-binding protein is hydrolyzed by the Ca2+-dependent protease within intact platelets during platelet activation.     

Liposome-mediated transfer of integral membrane glycoproteins into the plasma membrane of cultured cells

Cultured mouse 3T3 cells treated with phosphatidylserine or phosphatidylserine/phosphatidylcholine (3: 7 mole ratio) liposomes containing ortho- and paramyxovirus envelope glycoproteins become susceptible to killing by virus-specific cytotoxic T lymphocytes indicating that the liposome-derived glycoproteins have been inserted into the cellular plasma membrane. Cells incubated with liposomes of similar lipid composition containing viral antigens plus a dinitrophenylated lipid hapten were killed by both virus- and hapten-specific T lymphocytes indicating that both protein and lipid components are inserted into the plasma membrane. We consider that assimilation of liposome-derived antigens into the plasma membrane results from fusion of liposomes with the plasma membrane. Cells incubated with phosphatidylcholine liposomes containing lipid haptens and viral glycoproteins were not killed by cytotoxic lymphocytes indicating that liposomes of this composition do not fuse with the plasma membrane. Liposome-derived paramyxovirus glycoproteins inserted into the plasma membrane retain their functional activity as shown by their ability to induce cell fusion. These experiments demonstrate the feasibility of using liposomes as carriers for introducing integral membrane (glyco)proteins into the plasma membrane of cultured cells and establish a new approach for studying the role of individual (glyco)proteins in the expression of specific cell surface properties.     

Limited and selective transfer of plasma membrane glycoproteins to membrane of secondary lysosomes

Radioactive galactose, covalently bound to cell surface glycoconjugates on mouse macrophage cells, P388D1, was used as a membrane marker to study the composition, and the kinetics of exchange, of plasma membrane-derived constituents in the membrane of secondary lysosomes. Secondary lysosomes were separated from endosomes and plasma membrane on self-forming Percoll density gradients. Horseradish peroxidase, taken up by fluid-phase pinocytosis, served as a vesicle contents marker to monitor transfer of endosomal contents into secondary lysosomes. Concurrently, the fraction of plasma membrane-derived label in secondary lysosomes increased by first order kinetics (k = [56 min]-1) from less than 0.1% (background level) to a steady-state level of approximately 2.5% of the total label. As analyzed by NaDodSO4 PAGE, labeled molecules of Mr 160-190 kD were depleted and of Mr 100-120 kD were enriched in lysosome membrane compared with the relative composition of label on the cell surface. No corresponding selectivity was observed for the degradation of label, with all Mr classes being affected to the same relative extent. The results indicate that endocytosis-derived transfer of plasma membrane constituents to secondary lysosomes is a limited and selective process, and that only approximately 1% of internalized membrane is recycled via a membrane pool of secondary lysosomes.     

3-O-Methylfluorescein phosphate as a fluorescent substrate for plasma membrane Ca-ATPase

3-O-Methylfluorescein phosphate hydrolysis, catalyzed by purified erythrocyte Ca²⁺-ATPase in the absence of Ca²⁺, was slow in the basal state, activated by phosphatidylserine and controlled proteolysis, but not by calmodulin. p-Nitrophenyl phosphate competitively inhibits hydrolysis in the absence of Ca²⁺, while ATP inhibits it with a complex kinetics showing a high and a low affinity site for ATP. Labeling with fluorescein isothiocyanate impairs the high affinity binding of ATP, but does not appreciably modify the binding of any of the pseudosubstrates. In the presence of calmodulin, an increase in the Ca²⁺ concentration produces a bell-shaped curve with a maximum at 50 μM Ca²⁺. At optimal Ca²⁺ concentration, hydrolysis of 3-O-methylfluorescein phosphate proceeds in the presence of fluorescein isothiocyanate, is competitively inhibited by p-nitrophenyl phosphate and, in contrast to the result observed in the absence of Ca²⁺, it is activated by calmodulin. In marked contrast with other pseudosubstrates, hydrolysis of 3-O-methylfluorescein phosphate supports Ca²⁺ transport. This highly specific activity can be used as a continuous fluorescent marker or as a tool to evaluate partial steps from the reaction cycle of plasma membrane Ca²⁺-ATPases.     

Evaluation of periodate/lysine/paraformaldehyde fixation as a method for cross-linking plasma membrane glycoproteins

Fixation by periodate/lysine/paraformaldehyde, a method purported to cross-link specifically plasma membrane glycoproteins, was evaluated using Novikoff rat ascites hepatocellular carcinoma cells. Cells were treated with periodate/lysine, periodate/glycine, and periodate/lysine/paraformaldehyde and subsequently reduced with NaB3H4. The glycoproteins labeled with 3H were resolved by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate and visualized by fluorography. The effects of reactant concentrations on 3H-labeling of cellular components, cell viability, and cross-linkage of 3H-labeled proteins were examined. The effect of increasing the localized density of plasma membrane glycoproteins on the extent of cross-linkage by periodate and lysine was investigated using cells in which patching of the plasma membrane glycoproteins had been induced by ferritin-conjugated concanavalin A/rabbit antiferritin antiserum. Also investigated was the periodate-independent to mixtures of periodate and lysine or glycine. Results of these studies did not support a mechanism of cross-linking involving reaction between the free base lysin and aldehyde groups on periodate oxidized carbohydrate residues but suggested a complex interaction between periodate oxidized plasma membrane glycoproteins and polymeric complexes of lysine and formaldehyde.U