Mechanism of expression of Thomsen-Friedenreich (T) antigen at the cell surface of a mammary adenocarcinoma

The Thomsen-Friedenreich (T) antigen and its disaccharide component Gal beta 1,3GalNAc, which is recognized by the plant lectin peanut agglutinin (PNA), have been proposed as useful tumor markers because of their apparently specific occurrence in certain types of carcinomas. We have investigated the mechanism for the appearance of the disaccharide at the cell surface of ascites 13762 rat mammary adenocarcinoma cells using pulse-chase glucosamine labeling, proteolysis, and PNA precipitation of the cell-surface sialomucin ASGP-1. Glucosamine-labeled disaccharide appears at the cell surface in less than 10 min. Although the appearance of larger oligosaccharides continues to increase, the appearance of labeled disaccharide levels off within an hour. Analysis of intracellular vs. cell surface-labeled oligosaccharides showed that all disaccharide synthesized more than an hour before reaching the cell surface is converted to larger oligosaccharides. Thus, the presence of the disaccharide at the cell surface results from its synthesis late in the transit pathway of the sialomucin to the cell surface. We propose that the presence of T antigen at the surface of carcinoma cells results from an aberration of the pathway for O-linked glycosylation in these cells, probably caused by inappropriate localization of the enzymes involved in synthesis of the disaccharide.     

Biosynthesis of the cell surface sialomucin complex of ascites 13762 rat mammary adenocarcinoma cells from a high molecular weight precursor

Cell surfaces of metastatic 13762 ascites rat mammary adenocarcinoma cells are covered with a sialomucin complex composed of the high Mr sialomucin ASGP-1 (approximately 600,000) and a concanavalin A-binding, integral membrane glycoprotein ASGP-2 (120,000). Antibodies prepared against ASGP-2 and deglycosylated ASGP-1 react on immunoblots of ascites cells or their isolated microvilli with the Mr = 120,000 species and the high Mr sialomucin, respectively. No cross-reactivity was observed. Under complex dissociating conditions, anti-ASGP-2 immunoprecipitated primarily components of Mr = 120,000 and about 400,000 from lysates of cells labeled for 1 h with mannose, glucosamine, and threonine. Under similar conditions, anti-ASGP-1 immunoprecipitated the Mr = 400,000 component and a second major labeled component of about 330,000. Pulse-chase labeling with 35S-labeled amino acids followed by immunoprecipitation with anti-ASGP-2 indicated a precursor-product relationship for the Mr = 400,000 component, designated pSMC-1 (precursor, sialomucin complex), and ASGP-2. Similar pulse-chase analyses of threonine-labeled cells using anti-ASGP-1 showed equivalent amounts of immunoprecipitated pSMC-1 and pSMC-2, both of which disappeared with kinetics similar to those observed for pSMC-1 immunoprecipitated with anti-ASGP-2. A precursor-product relationship of both pSMC-1 and pSMC-2 to ASGP-1 was suggested by combined precipitations with anti-ASGP-1 and peanut agglutinin, which precipitates ASGP-1 specifically. Immunoblot and lectin blot analyses indicated that pSMC-1 and pSMC-2 from the immunoprecipitates bind anti-ASGP-2, anti-ASGP-1, and concanavalin A. Moreover, these three components can also be labeled with mannose; the mannose was removed from 30-min pulse-labeled anti-ASGP-2 immunoprecipitates by incubation with endo-beta-N-acetylglucosaminidase H, indicating the presence of only high mannose N-linked oligosaccharides in pSMC-1. One-dimensional peptide maps of 35S-labeled pSMC-1 and Mr = 120,000 ASGP-2 showed several corresponding bands. These results indicate that both ASGP-1 and ASGP-2 can be synthesized from a common high Mr precursor. We propose that complex is formed from pSMC-1 by proteolytic cleavage to yield Mr = 120,000 ASGP-2 plus the precursor to ASGP-1 early in the transit pathway from the endoplasmic reticulum to the cell surface.     

Changes in expression of a major sialoglycoprotein associated with ascites forms of a mammary adenocarcinoma

Glycoproteins of a cultured form (MR) of the 13762 rat mammary adenocarcinoma and its variants have been studied by analyses for peanut agglutinin receptors, [³H]glucosamine labeling, lactoperoxidase labeling and CsCl density gradient centrifugation. The 13762 MR cells, derived from 13762 MAT-B ascites cells, do not contain detectable ASGP-1, the predominant cell surface sialoglycoprotein of the ascites forms of the 13762 tumor.Transplantation and continued passage as ascites cells of MR cells or clonal lines derived from MR results in abrupt expression of ASGP-1 at about passage 16; it is absent in early passages of the ascites tumor. When these ascites cells are transferred to culture, ASGP-1 is again lost. No ASGP-1 is found in solid tumors derived from subcutaneous transplantation of the 13762 MR cells. The results suggest modulation of ASGP-1 content of the 13762 tumor cells.     

Biosynthesis of a tumor cell surface sialomucin. Maturation and effects of monensin

The major cell surface glycoprotein (ascites sialoglycoprotein-1 (ASGP-1] of ascites 13762 rat mammary tumor cells is a large (Mr greater than 500,000), highly glycosylated sialomucin which is present in great abundance (greater than 0.5% of total cell protein). Thus, these tumors provide a useful system for investigating the biosynthesis of O-glycosylated glycoproteins. Previous studies in this system have demonstrated that initiation of O-linked oligosaccharides occurs throughout most of the transit period of ASGP-1 from the endoplasmic reticulum to the cell surface. By pulse-chase threonine labeling and precipitation with peanut agglutinin, ASGP-1 is first observed as an immature lightly glycosylated form (Mr approximately 200,000) which is converted to a more mature, more heavily glycosylated form (designated the premature or P form) with a half-time of about 30 min. The P form is then more gradually converted into the mature ASGP-1. Analysis of glucosamine-labeled oligosaccharitols obtained from the immature form showed primarily unsialylated derivatives consisting of the structures of the size of the tetrasaccharide Gal beta 1,4GlcNAc beta 1,6(Gal beta 1,3)GalNAc and smaller, whereas the mature form showed a mixture of sialylated and unsialylated structures. Desialylation of glucosamine-labeled mature form resulted in a glycoprotein intermediate in size between the immature and mature forms, indicating that the size change with maturation is not solely due to sialylation. Treatment of the cells with 10(-6) M monensin significantly reduced the conversion of immature to mature form without inhibiting initiation of O-linked oligosaccharides and without preventing sialylation. Analysis of oligosaccharitols obtained from ASGP-1 of monensin-treated cells showed that the major oligosaccharides are trisaccharide GlcNAc beta 1,6(Gal beta 1,3)GalNAc and sialylated trisaccharide GlcNAc beta 1,6(NeuAc alpha 2,3-Gal-beta 1,3) GalNAc. These results suggest that monensin specifically disrupts the compartment of the biosynthetic pathway which adds most of the beta 1,4-Gal to the oligosaccharides of ASGP-1 and that this compartment is separate from the primary site of sialylation.     

Evidence for the association of two cell surface glycoproteins of 13762 mammary ascites tumor cells : Concanavalin A-induced redistribution of peanut agglutinin-binding proteins

The behavior of the cell surface concanavalin A (conA) receptors and of peanut agglutinin (PNA) receptors on the MAT-B1 ascites subline of the 13762 rat mammary adenocarcinoma was examined using fluorescein-labeled conA and PNA. ASGP-1, the major glucosamine-containing glycoprotein of these ascites cells, is the only PNA-binding protein observed by dodecyl sulfate electrophoresis. ASGP-2, the second most prominent component after glucosamine labeling, is the most abundant conA-binding protein. These two glycoproteins were previously shown to be associated as a complex in detergent extracts of the cells [20]. ConA-binding proteins, upon incubation with fluorescein-labeled conA (FITC-conA), redistribute on the cell surface into small and large aggregates similar, but not identical, to those seen in ‘patching’ and ‘capping’ experiments with lymphocytes. PNA-binding proteins failed to redistribute during incubation with fluorescein-labeled PNA (FITC-PNA) and appeared in a diffusely stained pattern around the circumference of the cells. However, when cells were treated with unlabeled conA followed by FITC-PNA, or with FITC-PNA followed by unlabeled conA, there was marked redistribution of the FITC-PNA. These results indicate that ASGP-1 redistributes in response to the movement of conAbinding proteins and supports our hypothesis that ASGP-1 and ASGP-2 are associated on the plasma membrane at the cell surface as well as in detergent extracts.     

Biosynthetic maturation of an ascites tumor cell surface sialomucin. Evidence for O-glycosylation of cell surface glycoprotein by the addition of new oligosaccharides during recycling

Previous biosynthetic studies of the ascites 13762 rat mammary adenocarcinoma cell surface sialomucin ASGP-1 (ascites sialoglycoprotein-1) showed that it is synthesized initially as a poorly glycosylated immature form, which is converted to a larger premature form (t1/2 30 min) and more slowly to the mature glycoprotein (t1/2 greater than 4 h). In the present study O-glycosylation of ASGP-1 polypeptide is shown to occur in two phases: an early phase complete in less than 30 min, which corresponds to the synthesis of the premature form, and a later phase that continues for hours and corresponds to the synthesis of the mature form. Pulse-chase labeling studies indicate that 95% of the ASGP-1 has moved to the cell surface in 2 h. Since transit to the cell surface is faster than the slow phase of addition of new oligosaccharides, some new oligosaccharides must be added after ASGP-1 has reached the cell surface. Initiation of new oligosaccharides on cell surface ASGP-1 was demonstrated directly using a biotinylation procedure to identify cell surface molecules. Glucosamine labeling of biotinylated ASGP-1 was shown to occur on galactosamine residues, which are linked to the polypeptide, establishing the addition of new oligosaccharides to the cell surface molecules. Finally, resialylation studies indicate that ASGP-1 rapidly recycles through a sialylating compartment. From these results we propose that ASGP-1 reaches the cell surface in an incompletely glycosylated state and that additional oligosaccharides are added to the glycoprotein in a second process involving recycling.     

Cell surface properties of ascites sublines of the 13762 rat mammary adenocarcinoma : Relationship of the major sialoglycoprotein to xenotransplantability

The relationship between cell surface sialoglycoprotein and xenotransplantation has been investigated in ascites sublines of the 13762 rat mammary adenocarcinoma. Two of the five sublines (MAT-C and MAT-C1) can be transplanted into mice. These two sublines also have the greatest amounts of total, trypsin-releasable and neuraminidase-releasable sialic acid. Chemical labeling using periodate treatment followed by [³H]borohydride reduction indicates that most of the protein-bound sialic acid is associated with a single major sialoglycoprotein (or family of glycoproteins) with a low mobility on polyacrylamide gels in dodecyl sulfate (SDS). This glycoprotein, denoted ASGP-1, is also labeled by lactoperoxidase and ¹²⁵I, indicating its presence at the cell surface. Metabolic labeling with [³H]glucosamine shows that ASGP-1 is the major glycosylated protein in both xenotransplantable (MAT-C1) and non-xenotransplantable (MAT-B1) sublines, representing >70% of the protein-bound label in each. The labeling studies indicate that the non-xenotransplantable subline does not have a substantially greater amount of ASGP-1 on its cell surface. Likewise cationized ferritin labeling and transmission electron microscopy (TEM) do not show substantially greater amounts of negatively charged groups distributed along the cell surfaces of MAT-C1 than of MAT-B1 cells. The results indicate that the transplantation differences between these sublines cannot be explained solely by the presence of a major sialoglycoprotein at the cell surface.     

Isolation and partial characterization of ascites sialoglycoprotein-2 of the cell surface sialomucin complex of 13762 rat mammary adenocarcinoma cells.

Sialomucins are the dominant components of the cell surfaces of some carcinoma ascites cells and have been postulated to inhibit recognition of tumours by the immune system. The sialomucin ASGP-1 (ascites sialoglycoprotein-1) of the 13762 rat mammary adenocarcinoma is associated with the cell surface as a complex with a concanavalin-A-binding glycoprotein called ASGP-2. This sialomucin complex has been purified from ascites cell microvilli by extraction with Triton X-100 and CsCl density-gradient centrifugation. ASGP-1 (which has been purified previously) and ASGP-2 were dissociated in 6 M-guanidine hydrochloride and separated by gel filtration. The molecular mass of the undenatured detergent complex of ASGP-2, estimated by gel filtration and velocity sedimentation in Triton X-100, was 148 kDa. Since the apparent molecular mass by SDS/polyacrylamide-gel electrophoresis was about 120 kDa, ASGP-2 must be a monomer as extracted from the membrane. Studies of its chemical composition indicate that it contains about 45% carbohydrate by weight, including both mannose and galactosamine. Alkaline borohydride treatment of ASGP-2 converted approx. half of the N-acetylgalactosamine to N-acetylgalactosaminitol, demonstrating the presence of O-linked oligosaccharides. Analyses of mannose-labelled Pronase glycopeptides from ASGP-2 by lectin-affinity chromatography on concanavalin A and leucocyte-agglutinating phytohaemagglutinin suggested that 40% of the label was present in high-mannose/hybrid oligosaccharides, 20% in triantennary oligosaccharides substituted on the C-2 and C-4 mannose positions and 40% in tri- or tetra-antennary oligosaccharides substituted on C-2 and C-6. The presence of polylactosamine sequences on these oligosaccharides was suggested by lectin blots and by precipitation from detergent extracts with tomato lectin. From chemical analyses and lectin-affinity studies, we estimate that ASGP-2 contains four high-mannose and 13 complex N-glycosylated oligosaccharides, plus small amounts of polylactosamine and O-linked oligosaccharides. The presence of four different classes of oligosaccharides on this glycoprotein suggests that it will be an interesting model system for biosynthetic comparisons of the different glycosylation pathways.     

Structural and functional aspects of tumor cell sialomucins

Sialomucins are abundant on the surfaces of certain ascites tumor cells and have been implicated in the escape of tumors from immune destruction and metastasis. They are large, highly glycosylated glycoproteins which are rich in serine and threonine and have a variety of 0-linked oligosaccharides. The sialomucin (ASGP-1) or 13762 rat mammary adenocarcinoma ascites cells represents more than 0.5% of the total cell protein and can be isolated from cell membranes by centrifugation in 4 M guanidine hydrochloride-cesium chloride. ASGP-1 can also be isolated from membranes or cells by nonionic detergent extraction as a 1:1 complex with a second glycoprotein ASGP-2. Studies with the fluorescent lectins peanut agglutinin, which binds ASGP-1, and Concanavalin A, which binds ASGP-2, indicate that the glycoproteins are present at the cell surface as a complex. ASGP-1 is shed into cell culture medium or ascites fluid, apparently by a proteolytic cleavage mechanism. 13762 ascites cells grown in culture or as solid tumors lose their ASGP-1. The sialomucin reappears with extensive passage of the tumor cells in ascites form. Studies on the biosynthesis of ASGP-1 indicate that carbohydrate is being added over nearly the entire period of transit of ASGP-1 from the site of polypeptide synthesis to the plasma membrane. The negatively charged, rod-like structure of the sialomucins suggests that they may play a role in inhibiting recognition or binding processes necessary for the immune destruction of these tumor cells.     

Sialoglycoprotein differences between xenotransplantable and nonxenotransplantable ascites sublines of the 13762 rat mammary adenocarcinoma

MAT-B1 and MAT-CI rat ascites mammary adenocarcinoma cells differ in morphology, lectin receptor mobility, and xenotransplantability. Since these properties may be related to cell surface organization, the predominant sialoglycoproteins of these sublines have been investigated by chemical labeling, proteolysis, and alkaline borohydride elimination. Treatment of both sublines with periodate and tritiated borohydride labels one major sialoglycoprotein (ASGP-1) with a low electrophoretic mobility on polyacrylamide gels in dodecyl sulfate. Treatment of labeled or unlabeled cells with trypsin releases about 30% of the total cell sialic acid without significant decrease in cell viability. Gel filtration in pyridine-acetate buffer or in dodecyl sulfate indicates that the released materials are very heterogeneous, and that most of the MAT-C1 sialoglycopeptides are larger than sialoglycopeptides of MAT-B1. Amino acid compositions are quite similar for the released material from the two sublines, but they differ substantially in sialic acid. Further degradation of trypsin-released material with Pronase gives products which are included in a column of mixed Bio-Gel P-10 and P-30 and which also indicate a larger average size for MAT-C1 sialoglyco-peptides. Oligosaccharides from the sialoglycopeptides were obtained by alkaline borohydride treatment of trypsin-released, labeled material and fractionated by chromatography on Bio-Gel P-2. The oligosaccharide(s) comprising the major peak from MAT-C1 cells was larger in size than most of the material from MAT-B1 cells and contained galactosaminitol, galactose, glucosamine, sialic acid, and fucose. These results suggest that MAT-C1 ASGP-1 has more complex oligosaccharides than MAT-B1 ASGP-1, a difference which may play an important role in the differences in cell behavior between the sublines, including transplantability. Regardless of whether the ASGP-1 plays a role in transplantation, investigations of the sialoglycoproteins of these sublines provide a potentially valuable tool for understanding some of the mechanisms by which tumor cells control their cell surface properties.     

Sulfation of the tumor cell surface sialomucin of the 13762 rat mammary adenocarcinoma

ASGP-1, the major cell surface sialomucin of the 13762 ascites rat mammary adenocarcinoma, is at least 0.5% of the total ascites cell protein and has sulfate on 20% of its O-linked oligosaccharide chains. We have used this system to investigate the O-glycosylation pathway in these cells and to determine the temporal relationship between sulfation and sialylation. The two major sulfated oligosaccharides (S-1 and S-2) were isolated as their oligosaccharitols by alkaline borohydride elimination, anion exchange HPLC, and ion-suppression HPLC. From structural analyses S-1 is proposed to be a branched, sulfated trisaccharide -O4S-GlcNAc beta 1,6-(Gal beta 1,3)-GalNAc and S-2 its sialylated derivative -O4S-GlcNAc beta 1,6-(NeuAc alpha 2,3-Gal beta 1,3)-GalNac. Pulse labeling with sulfate indicated that sulfation occurred primarily on a form of ASGP-1 intermediate in size between immature and mature sialomucin. Pulse-chase analyses showed that the intermediate could be chased into mature ASGP-1. The concomitant conversion of S-1 into S-2 had a half-time of less than 5 min. Monensin treatment of the tumor cells led to a 95% inhibition of sulfation with the accumulation of unsulfated trisaccharide GlcNAc beta 1,6-(Gal beta 1,3)-GalNAc and sialylated derivative GlcNAc beta 1,6-(NeuAc alpha 2,3-Gal beta 1,3)-GalNAc. These data suggest that sulfation of ASGP-1 is an intermediate synthetic step, which competes with beta-1,4-galactosylation for the trisaccharide intermediate and thus occurs in the same compartment as beta-1,4-galactosylation. Moreover, sulfation precedes sialylation, but the two are rapidly successive kinetic events in the oligosaccharide assembly of ASGP-1.     

Shedding of the major plasma membrane sialoglycoprotein from the surface of 13762 rat ascites mammary adenocarcinoma cells

ASGP-1 (ascites Sialoglycoprotein 1) the major sialoglycoprotein of 13762 rat ascites mammary adenocarcinoma cells, is shed from MAT-B1 (nonxenotransplantable) and MAT-C1 (xenotransplantable) sublines when incubated in vitro after labeling in vivo with [³H]glucosamine. The rates of shedding of label in both particulate and soluble form are similar for the two sublines, but the turnover of label in the cells is 80% greater for MAT-C1 cells ($\text{t}{}_{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}$ 2.4 days) than for MAT-B1 cells ($\text{t}{}_{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}$ 4.1 days). Shed soluble ASGP-1 was smaller than ASGP-1 in the particulate fraction by gel filtration in dodecyl sulfate. By CsCl density gradient centrifugation, gel filtration, and sucrose density gradient centrifugation, all in 4 m guanidine hydrochloride, the shed soluble ASGP-1 was found to be slightly more dense and smaller than ASGP-1 purified from membranes. No differences in sialic acid or oligosaccharides released by alkaline borohydride treatment were found between the shed soluble ASGP-1 and purified ASGP-1. These results suggest that the shed soluble ASGP-1 is released from the membrane by a proteolytic cleavage. This mechanism is supported by the inhibition of the release of soluble shed ASGP-1 by aprotinin, a protease inhibitor. Soluble ASGP-1 in ascites fluid is also smaller by gel filtration, but is more heterogeneous, suggesting a similar release mechanism in vivo followed by more extensive degradation in the ascites fluid.     

Ehrlich ascites tumor cell surface labeling and kinetics of glycocalyx release

Ehrlich ascites tumor cells spontaneously release cell surface material (glycocalyx) into isotonic saline medium. Exposure of these cells to tritium-labeled 4,4′-diisothiocyano-1,2′-dihenylethane-2,2′-disulfonic acid (³H₂DIDS) at 4°C leads to preferential labeling of the cell surface coat. We have combined studies of the kinetics of ³H₂DIDS-label release, the effects of enzymatic treatment, and cell electrophoretic mobility to characterize the ³H₂DIDS-labeled components of the cell surface. Approximately 73% of the cell-associated radioactivity is spontaneously released from the cells after 5 h at 23°C. The kinetics of release is consistent with the first-order loss of two fractions; a slow (τ_½ = 360 min) component representing 33% of the radioactivity, and a fast (τ_½ = 20 min) component representing 26%. The remaining 14% of the labile binding may reflect mechanically induced surface release. Trypsin (1 μ/ml) also removes approximately 73% of the labeled material within 30 min and converts the kinectics of release to that of a single component (τ_½ = 5.5 min). The specific activity (SA) of material released by trypsin immediately after labeling is 83% of the SA of the material spontaneously los in 1 h. However, trypsinization following a 2-h period of spontaneous release yields material of reduced (43%) SA. Neither ³H₂DIDS labeling nor the initial spontaneous loss of labeled material alters cell electrophoretic mobility. However, extended spontaneous release is accompanied by a significant decrease in surface charge density. Trypsinization immediately following labeling or after spontaneous release (2 h) reduces mobility by 32%. We have tentatively identified the slowly released compartment as contributing to cell surface negativity.     

Characterization and dynamics of O-linked glycosylation of human cytokeratin 8 and 18.

The glycosylation of human cytokeratin (CK) 8 and 18 was studied after metabolic labeling of HT29 colonic cells with [3H]glucosamine. Labeling of CK8/18 was not inhibited by tunicamycin, suggesting that glycosylation was not N-linked. Acid hydrolysis of CK8 and CK18, purified from [3H]glucosamine-labeled cells, generated free glucosamine. In the presence of UDP-[3H]galactose, galactosyltransferase catalyzed the labeling of cytokeratin 8 and 18. beta-Elimination of the [3H]galactose- labeled CK8/18 generated the disaccharide N-acetyllactosaminitol, indicating that cytokeratin 8 and 18 contain single O-linked N-acetylglucosamine residues. Using chemical analysis, the stoichiometry of glycosylation was found to be 1.5 and 2 molecules of N-acetylglucosamine/protein molecule of CK8 and CK18, respectively. Peptide maps of [3H]glucosamine-labeled CK8/18 showed that multiple peptides were labeled with the amino sugar. The biosynthetic and degradation rates of the carbohydrate moiety were faster than the protein core as determined by metabolic radiolabeling or pulse-chase experiments, respectively. Our results show that CK8 and 18 are glycosylated at multiple sites with a single O-linked N-acetylglucosamine. Furthermore, CK8/18 glycosylation is a dynamic process which is likely to have functional relevance.     

Active cell aggregation by immature rat Sertoli cells in primary culture: a role for cell surface glycoproteins.

Follicle stimulating hormone (FSH) stimulates “colony formation” by immature rat Sertoli cells in primary culture. “Colony formation” involves cell aggregation. Consequently, the involvement of cell surface glycoproteins in cell aggregation was investigated by treatment of dissociated 10-day rat testis cells with sodium metaperiodate, glucosamine, various lectins, tunicamycin, and puromycin. Treatment of control cultures with 5 μM glucosamine stimulated cell aggregation; however, glucosamine did not affect FSH-stimulated cultures. Treatment of dissociated testis cells with 5 μM sodium metaperiodate, 10 μg/ml castor bean agglutinin (ricin), or 2.5 μg/ml horseshoe crab agglutinin inhibited FSH stimulation of cell aggregation. A similar inhibition of cell aggregation was observed following addition of 10 μg/ml puromycin or tunicamycin to culture media from 0- to 18-hours incubation. Treatment with soybean agglutinin, concanavalin A, or wheat germ agglutinin had no effect. The galactose-specific lectins, Ricin, Ricinus communis agglutinin I, and Bendeirea simplicifolia agglutinin, inhibit the FSH stimulation of ³H-aminoacid incorporation as well as cell aggregation in 24-hour cultres. The inhibition of cell aggregation by sodium metaperiodate treatment was reversed with 5 μM sodium borohydride reduction. Sodium metaperiodate treatment did not alter cell viability (as assayed with trypan blue dye exclusion), did not alter cell attachment, nor significantly decrease ¹²⁵I-FSH binding by cultured testis cells. The results suggest that FSH stimulation of cell aggregation by immature rat Sertoli cells requires cell surface glycoprotein interactions. Furthermore, the specificity of lectin inhibition suggests that glycoproteins with terminal galactose and sialic acid residues are required for the FSH induction of cell aggregation.     

Structures of the O-linked oligosaccharides of the major cell surface sialoglycoprotein of MAT-B1 and MAT-C1 ascites sublines of the 13762 rat mammary adenocarcinoma

Structures of the principal O-glycosides from the major cell surface sialoglycoprotein (ASGP-1) of the MAT-B1 and MAT-C1 ascites sublines of the 13762 rat mammary adenocarcinoma have been determined. Oligosaccharitols were released by alkaline borohydride treatments of ASGP-1 and purified by gel filtration, DEAE-Sephadex ion exchange chromatography, and high performance liquid chromatography. On the basis of carbohydrate composition, methylation analysis, periodate oxidation, and exoglycosidase digestion, the five major oligosaccharides released by mild alkaline borohydride were assigned the following structures: Component II-3: (NeuAc alpha 2----3Gal beta 1----4GlcNAc beta 1----6)Ga 1 NAcOH(3----1 betaGa 1 3----2 alpha NeuAc) III-2a: (Ga 1 beta 1----4G1cNAc beta 1----6)Ga 1 NAcOH(3----1 beta Ga 1 3----2 alpha NeuAc) III-2c: (Ga 1 alpha 1----3Ga 1 beta 1----4G1cNAc beta 1----6) Ga 1 NAcOH(3----1 beta Ga 1 3----2 alpha NeuAc) IV-1a: (Ga 1 beta 1----4G 1 cNAc beta 1----6)Ga 1 NAcOH(3----1 beta Ga 1) IV-1c: (Ga 1 alpha 1----3Ga 1 beta 1----4G 1 cNAc beta 1----6) Ga 1 NAcOH(3----1 beta Ga 1) Fucosylated derivatives of III-2a, IV-1a, and IV-1c were found in smaller amounts with the fucose tentatively assigned to the 2-position of the lactosamine galactose. Components II-3, III-2a, and the fucosylated derivative of III-2A were found in both MAT-B1 and MAT-C1 sublines. The alpha-galactosides were found in detectable quantities only in subline MAT-B1. Oligosaccharides from MAT-C1 cells were enriched in sialic acid when compared to those from MAT-B1 cells. These results suggest that the 13762 ascites sublines, which bear different oligosaccharides, will provide models useful for the investigation of mechanisms regulating the expression of structures of the larger O-linked oligosaccharides.     

Glycoproteins from human colonic adenocarcinoma. Isolation and characterization of cell surface carcinoembryonic antigen from a cultured tumor cell line.

Alterations in cell surface glycoproteins have been implicated in malignancy. We examined surface membrane proteins of a cultured cell line, SKCO-1, which had been derived from a human colonic adenocarcinoma. Cell surface labeling of SKCO-1 cells with galactose oxidase, followed by reduction with sodium borotritide, revealed five major labeled glycoproteins upon sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis. At least three additional labeled glycoproteins could be detected if galactose oxidase treatment was preceded by neuraminidase treatment. Some, but not all, of the glycoproteins could be iodinated by lactoperoxidase. The predominantly labeled glycoprotein (GPI) had a molecular weight of 200,000 and co-migrated in SDS gel with carcinoembryonic antigen (CEA). GPI was not removed from the cell surface by EDTA, hypertonic saline, or sonication but was released from the membrane by detergents. This glycoprotein was subsequently purified using lectin-agarose columns and gel filtration. GPI was judged homogenous by protein- and carbohydrate-stained SDS-polyacrylamide gels and had an amino acid composition similar to that of CEA. The carbohydrate composition of GPI was qualitatively similar to CEA but quantitatively distinct. GPI had a greater proportion of sialic acid and galactosamine and less fucose and glucosamine than CEA. Immunological studies, however, demonstrated identity between GPI and CEA. A study of the turnover rate of GPI showed it to have a half-life of 5 days.     

Topography and microfilament core association of a cell surface glycoprotein of ascites tumor cell microvilli

Membrane-microfilament interactions are being investigated in microvilli isolated from 13762 rat mammary ascites tumor cells. These microvilli are covered by a sialomucin complex, composed of the sialomucin ascites sialoglycoprotein-1 (ASGP-1) and the associated concanavalin A (Con A)-binding glycoprotein ASGP-2. Limited proteolysis of the microvilli releases large, highly glycosylated fragments of ASGP-1 from the microvilli and increases the association of ASGP-2 with the Triton-insoluble microvillar microfilament core (Vanderpuye OA, Carraway CAC, Carraway, KL: Exp Cell Res 178:211, 1988). To analyze the topography of ASGP-2 in the membrane and its association with the microfilament core, microvilli were treated with proteinase K for timed intervals and centrifuged. The pelleted microvilli were extracted with Triton X-100 for the preparation of microfilament cores and Triton-soluble proteins or with 0.1 M carbonate, pH 11, for the preparation of microvillar membranes depleted of peripheral membrane proteins. These microvilli fractions were analyzed by dodecyl sulfate gel electrophoresis, lectin blotting with Con A and L-phytohemagglutinin, and immunoblotting with anti-ASGP-2. The earliest major proteolysis product from this procedure was a 70 kDa membrane-bound fragment. At longer times a 60 kDa released fragment, 30-40 kDa Triton-soluble fragments, and 25-30 kDa membrane- and microfilament-associated fragments were observed. Phalloidin shift analysis of microfilament-associated proteins on velocity sedimentation gradients indicated that the 25-30 kDa fragments were strongly associated with the microfilament core. From these studies we propose that ASGP-2 has a site for indirect association with the microfilament core near the membrane on a 15-20 kDa segment.     

Characterization of surface sugars on algal cells with fluorescein isothiocyanate-conjugated lectins

Summary. We used qualitative and quantitative fluorescence microscopy of the fluorescein isothiocyanate-conjugated lectins Concanavalin A, phytohaemagglutinin-erythroagglutinin, pokeweed mitogen, and peanut agglutinin to examine sugar composition on the cell surface and cell-associated mucilage (where present) in a number of cultured and environmental algae. Lectin-binding activity was markedly different between laboratory-cultured and environmental samples of the same species. Sugar composition of the cyanobacterium Anabaena cylindrica varied with growth cycle, although no clear pattern of change was observed. Akinetes typically showed lectin-binding activity higher than that of the vegetative cells or heterocysts throughout the growth cycle. Algae with mucilage showed greater lectin binding, indicating that mucilage contained more surface sugars accessible to the lectin probe compared with the cell wall surface. A low level of galactose and N-acetyl galactosamine (detected by peanut agglutinin) was associated with the surface mucilage of most algal species. Relatively high amounts of mannose, glucose, and N-acetyl glucosamine (detected by Concanavalin A, phytohaemagglutinin, and pokeweed mitogen) were also present. Lectin binding was shown to be a highly specific and sensitive approach to the examination of cell surface chemistry of both cultured and environmental algae and to the study of biodiversity in phytoplankton. Correspondence: School of Biological Sciences, University of Manchester, 3.614 Stopford Building, Oxford Road, Manchester M13 9PT, U.K.     

Cell surface glycoproteins of 13762NF mammary adenocarcinoma clones of differing metastatic potentials

Rat 13762NF mammary adenocarcinoma cell surface glycoproteins from s.c. tumor- or lung metastases-derived cell clones of differing spontaneous metastatic potentials were examined for their relationship to metastasis. After treatment with neuraminidase, lectin-binding assays showed that highly metastatic clone MTLn3 cells express approximately twice the quantity of peanut agglutinin (PNA) binding sites (approximately 2.3 X 10(8) sites/cell) than clones of lower metastatic potential. However, the number of wheat germ agglutinin (WGA)-binding sites on the various cell clones decreased slightly as the metastatic potential of the clones increased. The quantities of concanavalin A (conA)-binding sites were similar (approximately 1.7 X 10(8) sites/cell) in all cell clones and growth conditions. Glycoprotein analysis was performed by electrophoresis on polyacrylamide gels in the presence of sodium dodecyl sulfate (SDS-PAGE) and subsequent staining with 125I-labeled lectins. SDS-PAGE gels stained with 125I-labeled conA revealed mainly one glycoprotein (Mr approximately 150 kD), and the amounts of this glycoprotein did not correlate with metastasis. Differences in WGA-binding glycoproteins were detected between s.c. tumor- and lung metastases-derived cell clones. Several desialylated glycoproteins were detected with 125I-labeled PNA after SDS-PAGE, and the labeling intensity of one (Mr approximately 580 kD) correlated with the metastatic potentials of the various cell clones. This high Mr galactoprotein was further analyzed by [3H]glucosamine metabolic labeling, solubilization, sequential gel filtration, and chondroitinase ABC treatment prior to SDS-PAGE. The 580 kD galactoprotein was expressed in increased amounts on the more highly metastatic clones. Chemical labeling of cell surface sialic acid residues using periodate treatment followed by [3H]borohydride reduction showed an additional change in a major sialoglycoprotein (Mr approximately 80 kD), which decreased in labeling intensity on clones of increasing metastatic potential. The results suggest quantitative changes in cell surface glycoproteins rather than major qualitative alterations are associated with differences in the metastatic behavior of 13762NF tumor cell clones.