Interactions of a laminin-binding peptide from a 33-kDa protein related to the 67-kDa laminin receptor with laminin and melanoma cells are heparin-dependent

A laminin-binding peptide (peptide G), predicted from the cDNA sequence for a 33-kDa protein related to the 67-kDa laminin receptor, specifically inhibits binding of laminin to heparin and sulfatide. Since the peptide binds directly to heparin and inhibits interaction of another heparin-binding protein with the same sulfated ligands, this inhibition is due to direct competition for binding to sulfated glycoconjugates rather than an indirect effect of interaction with the binding site on laminin for the 67-kDa receptor. Direct binding of laminin to the peptide is also inhibited by heparin. This interaction may result from contamination of the laminin with heparan sulfate, as binding is enhanced by the addition of substoichiometric amounts of heparin but inhibited by excess heparin and two heparin-binding proteins. Furthermore, laminin binds more avidly to a heparin-binding peptide derived from thrombospondin than to the putative receptor peptide. Adhesion of A2058 melanoma cells on immobilized peptide G is also heparin-dependent, whereas adhesion of the cells on laminin is not. Antibodies to the beta 1-integrin chain or laminin block adhesion of the melanoma cells to laminin but not to peptide G. Thus, the reported inhibition of melanoma cell adhesion to endothelial cells by peptide G may result from inhibition of binding of laminin or other proteins to sulfated glycoconjugate receptors rather than from specific inhibition of laminin binding to the 67-kDa receptor.

Characterization of the adenine binding sites of two Dolichos biflorus lectins

The seed lectin and a stem and leaf lectin (DB58) from Dolichos biflorus have high-affinity hydrophobic sites that bind to adenine. The present study employs a centrifugal filtration assay to characterize these sites. The seed lectin contains two identical sites with Ka's of 7.31 x 10(5) L/mol whereas DB58 has a single site with a Ka of 1.07 x 10(6) L/mol. The relative affinities of these sites for a host of adenine analogs and derivatives were determined by competitive displacement assays. The most effective competitors for adenine were the cytokinins, a class of plant hormone, for which the lectins had apparent Ka's of 1.96 x 10(5)-4.90 x 10(4) L/mol. Direct binding of the cytokinin 6-(benzylamino)purine (BAP) to both lectins showed positive cooperativity for only the seed lectin, indicating the interaction of this ligand with more than one class of hydrophobic binding site. Fluorescence enhancement assays demonstrate cooperativity between hydrophobic sites of the seed lectin and also suggest that BAP binds to more than one class of site.

Heparin- and sulfatide-binding peptides from the type I repeats of human thrombospondin promote melanoma cell adhesion

Peptides from the three type I repeats of human endothelial cell thrombospondin, containing the consensus sequence-Trp-Ser-Xaa-Trp-, bind to sulfated glycoconjugates including heparin and sulfatide. The peptides are potent inhibitors for the binding of thrombospondin, laminin, or apolipoprotein E to these ligands. The thrombospondin peptides that inhibit heparin binding, but not adjacent peptides from the thrombospondin sequence containing the previously identified adhesive motif Val-Thr-Cys-Gly, promote melanoma cell adhesion when immobilized on plastic. Melanoma cell adhesion to the immobilized peptides is inhibited by soluble recombinant heparin-binding fragment of thrombospondin. The peptides also inhibit heparin-dependent binding of thrombospondin or laminin to human melanoma cells. The active peptides lack any previously identified heparin-binding consensus sequences and most do not contain any basic amino acids. Studies with homologous peptides showed that the tryptophan residues are required for binding. Adjacent basic residues in the second type I repeat enhance binding to heparin but not to sulfatide. Thus the type I peptides of thrombospondin define a distinct class of heparin-binding peptides.

Binding of elastin to Staphylococcus aureus

Many pathogenic bacteria specifically bind to components of the extracellular matrix. In this study, we report the specific association of Staphylococcus aureus with elastin, a major structural component of elastic tissue. Competition assays in which the binding of radiolabeled tropoelastin was inhibited by excess unlabeled elastin peptides, but not by other proteins, established the specificity of the interaction. Kinetic studies showed that tropoelastin binding to the bacteria was rapid and saturable. Scatchard analysis of the equilibrium binding data indicated the presence of a single class of high affinity binding sites (KD approximately 4-7 nM) with approximately 1000 sites per organism. Protease susceptibility suggested that the elastin binding moiety on S. aureus was a protein, which was confirmed by the isolation of a 25-kDa elastin-binding protein from S. aureus extracts through affinity chromatography. Using a truncated form of tropoelastin, the bacterial binding domain on elastin was mapped to a 30-kDa fragment at the amino end of the molecule. Although the precise amino acid sequence recognized by the staphylococcal elastin receptor has not been characterized, it is clearly different from the region of tropoelastin that specifies binding to mammalian elastin receptors.

Platelets adhere to sulfatides by von Willebrand factor dependent and independent mechanisms

Unstimulated human platelets from normal volunteers adhere to sulfatides (galactosylceramide-I3-sulfate) as single cells but do not adhere appreciably to other lipids including gangliosides, neutral glycolipids, phospholipids or cholesterol-3-SO4. Platelet adhesion to sulfatide is saturable and dose-dependent, reaches maximal levels in 90 to 120 min, and is not divalent cation-dependent. Because sulfatides bind von Willebrand factor (vWf) with specificity and high affinity and platelet adhesion to structurally related sulfated glycolipids is approximately proportionate to their ability to bind vWf, we examined whether vWf mediates platelet adhesion to sulfatides. Platelets from a patient with severe Type I von Willebrand's disease adhere poorly to sulfatides. However, adhesion to levels seen with normal platelets is restored by the addition of vWf. Adhesion of normal platelets can be partially inhibited by a monospecific antibody to vWf. Normal platelet adhesion to sulfatides, however, is not increased following preincubation with vWf. Both vWf binding and platelet adhesion to sulfatides can be inhibited by the sulfated polysaccharide dextran sulfate at low concentration, fucoidan at high concentrations, but not by heparin, fibrinogen, fibronectin, or the synthetic peptides Gly-Arg-Gly-Asp-Ser-Pro or Gly-Arg-Gly-Glu-Ser-Pro. Thus, adhesion to sulfatides appears to be of two types; vWf dependent (50-75%) and vWf independent (25-50%).

Antigenic and calcium binding properties of a Peptide containing the essential cysteine in lima bean lectin

Polyclonal antisera were raised against a peptide containing the cysteine residue required for carbohydrate binding activity in the lima bean lectin. The antisera were tested for cross-reactivity with (a) synthetic peptide analogs to the essential cysteine containing peptide, (b) proteolytic digests of related lectins, (c) native lectins. The antisera were specifically inhibited from binding to a peptide conjugate by free synthetic peptides. The degree of inhibition by lectin digests correlated approximately along evolutionary relationships and the degree of sequence conservation. One antiserum was found to cross-react with certain lectins in the native state. In a second set of experiments, the calcium binding properties of the synthetic peptides were investigated using metal ion-chelate chromatography and UV-difference spectroscopy. The nonapeptide and undecapeptide bound to a Ca(2+) iminodiacetic acid agarose column and were eluted with EDTA. Ultraviolet difference spectral titrations with Ca(2+) performed on the synthetic undecapeptide and a related favin derived peptide resulted in dissociation constants of approximately 6 x 10(3) per molar.

Suramin inhibits laminin- and thrombospondin-mediated melanoma cell adhesion and migration and binding of these adhesive proteins to sulfatide

Suramin is a polysulfonated drug with several biological activities including inhibition of binding of some growth factors to cells, inhibition of tumor cell growth, and of glycosaminoglycan metabolism. We report here that suramin also inhibits binding of the adhesive glycoproteins, thrombospondin and laminin, to immobilized sulfatide with 50% inhibitory doses of 220 and 470 micrograms/ml, respectively. Sulfated glycoconjugates on melanoma cells mediate spreading on thrombospondin by binding to the amino-terminal heparin- and sulfatide-binding domain. This domain is also required for chemotaxis on thrombospondin. We therefore examined the effect of suramin on human melanoma cell spreading and migration. Suramin at 50-400 micrograms/ml specifically inhibited G361 melanoma cell spreading on thrombospondin without affecting cell attachment. Suramin also inhibited spreading of A2058 melanoma cells on thrombospondin and laminin and partially inhibited cell attachment. However, suramin had no effect on G361 or A2058 cell attachment or spreading on fibronectin. Chemotaxis of A2058 and G361 melanoma cells to thrombospondin and laminin were also specifically inhibited by suramin, as was haptotaxis of A2058 melanoma cells to laminin. However, suramin only weakly inhibited haptotaxis of G361 melanoma cells to thrombospondin, which is not mediated by the amino-terminal domain, and did not inhibit haptotaxis to fibronectin. These results suggest a new mechanism for the observed antitumor activity of suramin based on its ability to inhibit interactions of tumor cells with laminin or thrombospondin in the extracellular matrix.

Interactions of respiratory pathogens with host cell surface and extracellular matrix components

Adhesion of pathogens to proteins and glycoconjugates on the host cell plasma membrane or to components of the extracellular matrix is a critical early step in the initiation of infection. For intracellular pathogens, adhesion to the cell surface is a prerequisite to gaining entry into the cell. In all cases, adhesion to host tissue prevents elimination of the pathogens by normal clearance processes and may help the organism to evade immune surveillance by the host. Many laboratories are investigating the ligand binding specificities of bacterial receptors or adhesions and have described diverse binding specificities for adhesive proteins in the host extracellular matrix including laminin and fibronectin. Many bacteria also have adhesins that bind to carbohydrates occurring on glycolipids and glycoproteins in the apical membranes of epithelia in tissues that are targets for infection. Definition of these binding specificities and identification of the receptors that mediate adhesion may lead to development of a novel class of antibiotics whose mechanism of action is to compete with the endogenous ligands for binding to the pathogen receptors or to otherwise prevent adhesion to host tissues and thereby prevent infection.

Sulfatide-binding domain of the laminin A chain

A sulfatide-binding site on the globular end region of the long arm of laminin has been identified. Following proteolytic digestion with thermolysin, an intact fragment of the laminin A chain carboxyl-terminal domain exhibiting sulfatide-binding activity was isolated using gel filtration and heparin affinity chromatography. This fragment is composed of two peptides that are covalently linked by at least one disulfide bond and encompass the carboxyl-terminal 394 amino acids of the A chain. The clusters of charged residues in the primary structure of these fragments are sufficient for heparin-binding activity but not sulfatide binding since reduction and alkylation of the fragments abolished sulfatide binding under conditions in which heparin binding was retained. Thus, sulfatide binding requires an intact three-dimensional structure. The iodinated fragment bound to A2058 melanoma and T47D breast carcinoma cells and could be displaced by the unlabeled fragment. Based on incorporation of [35S] sulfate, both cell lines synthesize sulfated glycolipids that bind to laminin. In agreement with previous data that indicate a synergistic interaction of the sulfatide-binding domain with other laminin-binding sites on melanoma cells during attachment, the isolated sulfatide-binding fragment significantly inhibited interaction of labeled intact laminin with melanoma and breast carcinoma cells in direct binding assays.

Falciparum malaria parasitized erythrocytes bind to a carboxy-terminal thrombospondin fragment and not the amino-terminal heparin-binding region

We investigated Plasmodium falciparum parasitized erythrocyte binding to proteolytic fragments of thrombospondin and the effects of anti-thrombospondin monoclonal antibodies on this binding. Purified human platelet thrombospondin was cleaved by trypsin, chymotrypsin or thrombin. Fragments were separated by heparin-agarose affinity chromatography, removing the amino-terminal heparin-binding region. Trypsin at 5.0 micrograms ml-1 of thrombospondin cleaved thrombospondin to reduced 140 and 120 kDa fragments plus a reduced 25-kDa heparin-binding fragment. Infected erythrocytes bound to intact thrombospondin (3420 +/- 460 infected erythrocytes mm-2) and the carboxy-terminal fragment, yielding 120-140-kDa fragments on sulfhydryl reduction, but not to the 25-kDa fragment (144 +/- 104 infected erythrocytes mm-2 (mean +/- s.d., N = 4). Similar results were obtained with chymotrypsin and thrombin cleavage. When the anti-thrombospondin monoclonal antibody MA-I was added to immobilized thrombospondin prior to infected erythrocytes, adherence was inhibited by 99%. At the same concentration, MA-I inhibited adherence to C32 melanoma cells by only 35%. MA-I binds to a calcium-dependent structure at the C-terminal globular region of thrombospondin. Monoclonal antibody MA-II inhibited adherence to thrombospondin by 46%, while MA-III had no effect. These antibodies bind to the N-terminal globular region which includes the heparin-binding site and the segment connecting the two globular regions, respectively. The site(s) for infected erythrocyte binding on thrombospondin reside in the large, 140- or 120-kDa, proteolytic cleavage fragments, and not in the N-terminal heparin-binding region.

Isolation and characterization of mucin from the serum of cystic fibrosis patients

Using monoclonal antibody 19-9, elevated levels of the sialyl Lea antigen (NeuAc alpha 2-3Gal beta 1-3[Fuc alpha 1-4]GlcNac beta 1-3R) are detected in serum from most cystic fibrosis patients. We now report further characterization of the serum antigen and evidence that it is on a mucin glycoprotein and not on glycolipids. The antigen has an apparent molecular weight greater than 2 X 10(6) by gel filtration on Sephacryl S-400. On density gradient centrifugation, the antigen has a density of 1.54 g/ml in cesium chloride and 1.42 g/ml in cesium chloride/4 M guanidine HCl. Immunostaining with monoclonal antibody 19-9 of lipid extracts from cystic fibrosis patient serum and erythrocytes does not detect any antigen on glycolipids. The antigen was purified by gel filtration and density gradient centrifugation. After tritium labeling of the sialic acid residues, sodium dodecyl sulfate gel electrophoresis separates two subunits with apparent molecular weights of 200,000 and 400,000. All of the labeled sialic acid is released as low molecular weight oligosaccharides after mild alkaline borohydride degradation. The purified antigen contains fucose, galactosamine, glucosamine, and galactose but no mannose and is enriched in the amino acids threonine, serine, glycine, proline, and alanine. The purified antigen binds several antibodies recognizing epitopes common to many mucins. Thus, the physical, biochemical, and immunochemical properties of the purified antigen indicate that the sialyl Lea antigen is present on mucins in the serum of cystic fibrosis patients.

Mono-sulfated globotetraosylceramide from human kidney

A novel sulfated glycosphingolipid that belongs to "globo-series" was isolated from human kidney. This lipid was purified from a pooled kidney preparation by chloroform-methanol extraction, mild alkaline treatment, DEAE-Sephadex and silicic acid column chromatographies, and preparative TLC. The structure and the properties were studied by IR spectroscopy, proton NMR spectroscopy, negative secondary ion-mass spectrometry, solvolysis, periodate oxidation, compositional and methylation analyses, monoclonal antibodies, and a sulfatide-binding protein. From the results of the above analyses, the structure of this glycolipid was proposed to be HSO3-3GalNAc beta 1-3Gal alpha 1-4Gal beta 1-4Glc beta 1-1ceramide. This sulfated lipid reacted with a monoclonal anti-SSEA-3 (stage-specific embryonic antigen-3) (MC-631) (Kannagi, R., Cochran, N.A., Ishigami, F., Hakomori, S., Andrews, P.W., Knowles, B.B., & Solter, D. (1983) EMBO J. 2, 2355-2361), whose epitope is R-3GalNAc beta 1-3Gal alpha 1-4Gal beta 1-R', on TLC and solid-phase radioimmunoassay. This lipid also bound to the 125I-labeled sulfatide-binding protein, thrombospondin. The yield of this sulfated glycolipid was 34 pmol/g of tissue, which was about 0.028, 0.16, and 18 mol% of galactosyl- and lactosylceramide sulfates, and globopentosylceramide sulfate (Nagai, K.-i., Roberts, D.D., Toida, T., Matsumoto, H., Kushi, Y., Handa, S., & Ishizuka, I. (1989) J. Biol. Chem. 264, in press), respectively, in human kidney.

Mono-sulfated globopentaosylceramide from human kidney

A novel sulfated glycosphingolipid that belongs to the "globo-series" was isolated from human kidney. This lipid was purified from a pooled kidney preparation by chloroform/methanol extraction, mild alkaline treatment, DEAE-Sephadex and silicic acid column chromatographies, and preparative thin layer chromatography. The structure and the properties were studied by infrared spectroscopy, two-dimensional proton magnetic resonance spectroscopy, negative secondary ion mass spectrometry, solvolysis, compositional and methylation analyses, monoclonal antibodies, and sulfatide-binding proteins. From the results of the above analyses, the structure of this glycolipid was proposed to be HSO3-3Gal beta 1-3GalNAc beta 1-3Gal alpha 1-4Gal beta 1-4Glc beta 1-1ceramide. The proton resonance at 3.93 ppm of the H-3 of the sulfated nonreducing terminal galactose of this lipid was downfield-shifted (delta 0.48 ppm), as compared with H-3 of the internal beta-galactose because of the electronegativity of the sulfate ester. This sulfated lipid reacted with a monoclonal anti-SSEA-3 (MC-631) (Kannagi, R., Cochran, N. A., Ishigami, F., Hakomori, S., Andrews, P. W., Knowles, B. B., and Solter, D. (1983) EMBO J. 2, 2355-2361), whose epitope is R-3GalNAc beta 1-3Gal alpha 1-4Gal beta 1-R', on thin layer chromatograms and solid-phase radioimmunoassay. This lipid also bound to the 125I-labeled sulfatide-binding protein, thrombospondin. The yield of this sulfated glycolipid was 0.19 nmol/g of tissue, which was about 0.09 and 0.5 mol % of galactosyl and lactosyl sulfatides in human kidney.

Sialic acid-dependent adhesion of Mycoplasma pneumoniae to purified glycoproteins

Several purified glycoproteins including laminin, fetuin, and human chorionic gonadotropin promote dose-dependent and saturable adhesion of Mycoplasma pneumoniae when adsorbed on plastic. Adhesion to the proteins is energy dependent as no attachment occurs in media without glucose. Adhesion to all of the proteins requires sialic acid, and only those proteins with alpha 2-3-linked sialic acid are active. The alpha-subunit of human chorionic gonadotropin also promotes attachment, suggesting that a simple biantennary asparagine-linked oligosaccharide is sufficient for binding. Soluble laminin, asparagine-linked sialyloligosaccharides from fetuin, and 3'-sialyllactose but not 6'-sialyllactose inhibit attachment of M. pneumoniae to laminin. M. pneumoniae also bind to sulfatide adsorbed on plastic. Dextran sulfate, which inhibits M. pneumoniae binding to sulfatide, does not inhibit attachment on laminin, and 3'-sialyllactose does not inhibit binding to sulfatide, suggesting that two distinct receptor specificities mediate binding to these two carbohydrate receptors. Both 3'-sialyllactose and dextran sulfate partially inhibit M. pneumoniae adhesion to a human colon adenocarcinoma cell line (WiDr) at concentrations that completely inhibit binding to laminin or sulfatide, respectively, and in combination they inhibit binding of M. pneumoniae to these cells by 90%. Thus, both receptor specificities contribute to M. pneumoniae adhesion to cultured human cells.

Adhesion of Mycoplasma pneumoniae to sulfated glycolipids and inhibition by dextran sulfate

A virulent strain of Mycoplasma pneumoniae was metabolically labeled with [3H]palmitate and studied for binding to glycolipids and to WiDr human colon adenocarcinoma cells. The organism binds strongly to sulfatide and other sulfated glycolipids, such as seminolipid and lactosylsulfatide which all contain terminal Gal(3SO4) beta 1-residues and weakly to some neolactoseries neutral glycolipids. M. pneumoniae do not bind gangliosides including the sialylneolacto-series and other neutral glycolipids that were tested. Only metabolically active M. pneumoniae cells bind to sulfatide, as binding is maximal in RPMI medium at 37 degrees C and almost completely abolished in nutrient-deficient medium or by keeping the cells at 4 degrees C. Dextran sulfate but not other sulfated or anionic polysaccharides at 10 micrograms/ml completely inhibits binding of M. pneumoniae to purified sulfatide. Dextran sulfate does not inhibit binding to the neolacto-series neutral glycolipids. Dextran sulfate partially inhibits adhesion of M. pneumoniae to cultured human colon adenocarcinoma cells (WiDr). The biological relevance of these data is suggested by our finding that sulfatide occurs in large amounts in human trachea, lung, and WiDr cells. Thus, there are at least two distinct receptors that mediate binding of M. pneumoniae to cells: glycolipids containing terminal Gal(3SO4) beta 1-residues as reported here, and glycoproteins containing terminal NeuAc alpha 2-3Gal beta 1-4GlcNAc sequences (Roberts, D. D., Olson, L. D., Barile, M. F., Ginsburg, V., and Krivan, H. C. (1989) J. Biol. Chem. 264, 9289-9293).

Structural study of the sugar chains of human platelet thrombospondin

The asparagine-linked sugar chains of human platelet thrombospondin were released as oligosaccharides by hydrazinolysis. About 12 mol of sugar chains was released from one thrombospondin molecule. This was converted to radioactive oligosaccharides by sodium borotritide reduction after N-acetylation, and separated into one neutral and four acidic fractions by paper electrophoresis. More than 90% of the oligosaccharides were recovered in the acidic fraction. The acidic oligosaccharides were mostly converted to neutral oligosaccharides by sialidase treatment, indicating that they are sialyl derivatives. The neutral and sialidase-treated acidic oligosaccharides were further fractionated by Bio-Gel P-4 column chromatography. Structural study of each oligosaccharide by sequential exoglycosidase digestion and methylation analysis revealed that the thrombospondin contains mono-, bi-, tri-, and tetraantennary complex-type sugar chains in addition to a small amount of high-mannose type. Approximately 70% of the complex-type sugar chains was fucosylated at asparagine-linked N-acetylglucosamine residue and 19% of the biantennary complex-type sugar chains was bisected.

Studies of the receptors on melanoma cells for Plasmodium falciparum infected erythrocytes

We investigated whether thrombospondin plays a role in the binding of Plasmodium falciparum parasitized erythrocytes to C32 melanoma cells. Twelve patient isolates bound variably to melanoma cells, with good correlation between the degree of binding to cells and binding to thrombospondin. With a synchronous preparation of asexual parasites, acquisition of the capacity to bind to thrombospondin occurred at the same parasite stage as binding to melanoma cells. Development of parasites to trophozoites and schizonts correlated with binding of parasitized erythrocytes to thrombospondin and melanoma cells. The infected erythrocyte receptor for thrombospondin was destroyed by mild trypsinization, as was the receptor for melanoma cells. Although these results suggest similarity in the melanoma cell receptor and thrombospondin receptor for infected cells, other results showed that thrombospondin cannot alone be the melanoma cell receptor. Binding to other melanoma cell lines did not correlate with thrombospondin secretion: the RPMI 8252 and G361 cell lines bound few or no infected cells, yet secreted 50-100% as much thrombospondin as C32 cells. Iodinated thrombospondin bound in similar amounts to C32 cells and to noncytoadherent C361 melanoma cells. Binding and nonbinding melanoma cells did not differ in quantity of surface thrombospondin by radioimmunoassay. Thus, although purified, immobilized, thrombospondin binds parasitized erythrocytes, expression of thrombospondin alone on melanoma cells is not sufficient to mediate adherence.

Serum mucin-associated antigen levels of cystic fibrosis patients are related to their ages and clinical statuses

Mucin levels are generally elevated in sera from many cystic fibrosis (CF) patients as measured by radioimmunoassay using monoclonal antibody 19-9, which is directed against the mucin-associated sialyl Lea antigen. Antibody 19-9 can only be used to measure mucin-associated antigen levels in those patients who are genetically able to make detectable levels of mucin-associated sialyl Lea epitope. Serial studies of 20 patients followed over 3-5 y showed that their serum mucin-associated antigen levels varied directly with respect to the severity of their disease and inversely with their Shwachman-Kulczycki clinical scores (p less than 0.001) and Brasfield chest roentgenographic scores (p less than 0.02). Serum mucin-associated antigen levels in samples from 89 CF patients were generally higher in the older patients (p less than 0.025). Serum mucin-associated antigen levels of CF patients who were colonized with Pseudomonas aeruginosa did not significantly differ from those of uninfected CF patients. The mean serum mucin-associated antigen level of CF patients colonized with Pseudomonas was higher than the mean mucin level of six non-CF bronchiectatic patients whose lungs were colonized with Pseudomonas (p = 0.053). Serum mucin-associated antigen levels are thus related to CF patients' ages and clinical statuses.

Interactions of thrombospondin with sulfated glycolipids and proteoglycans of human melanoma cells

Human melanoma cell spreading on thrombospondin substrates and chemotaxis in a gradient of soluble thrombospondin requires the amino-terminal heparin/sulfatide-binding domain of thrombospondin. Some melanoma cell lines attach but do not spread or respond in chemotaxis assays. Sulfated glycoconjugates produced by melanoma cells that could mediate these activities were identified by metabolic labeling with [35S] sulfate and tested for their ability to bind thrombospondin. Heparan sulfate proteoglycans that bind thrombospondin are made by both spreading and non-spreading melanoma cell lines. Thrombospondin binds with high affinity to a high molecular weight heparan sulfate proteoglycan, but not to the major chondroitin sulfate. The active heparan sulfate proteoglycan can be partially purified by affinity chromatography on thrombospondin-agarose or hydrophobic interaction with octyl-Sepharose. Thrombospondin binding requires the amino-terminal domain and is inhibited by monoclonal antibody A2.5 or fucoidan. Binding activity is lost following degradation of the proteoglycan with heparatinase or nitrous acid. [35S]Sulfate labels several melanoma cell glycolipids including galactosylceramide-I3-sulfate, lactosyl ceramide-II3-sulfate, and sulfated glucuronosylparagloboside. The latter glycolipid was detected in three cell lines that spread on thrombospondin but not in the nonspreading C32 melanoma cells. Thrombospondin binds to the isolated glycolipid, and the glycolipid and an antibody to this structure inhibit cell spreading on thrombospondin substrates. Thus, the presence of glycoconjugates with terminal nonreducing glucuronosyl 3-sulfate correlates with melanoma cell spreading on thrombospondin, whereas expression of heparan sulfate proteoglycans that bind thrombospondin does not.

Sulfated glycolipids and cell adhesion

The adhesive glycoproteins laminin, thrombospondin, and von Willebrand factor bind specifically and with high affinity to sulfatides, and it is this binding that probably accounts for their ability to agglutinate glutaraldehyde-fixed erythrocytes. The three proteins differ, however, in the inhibition of their binding to sulfatides by sulfated polysaccharides. Fucoidan strongly inhibits binding of both laminin and thrombospondin, but not of von Willebrand factor, suggesting the involvement of laminin or thrombospondin, or other unknown sulfatide-binding proteins in specific cell interactions that are also inhibited by fucoidan. Thrombospondin adsorbed on plastic promotes the attachment and spreading of some melanoma cells. Interestingly, fucoidan and an antibody against the sulfatide-binding domain of thrombospondin selectively inhibit spreading but not attachment to thrombospondin-coated surfaces. Sulfatides, but not neutral glycolipids or gangliosides, when adsorbed on plastic also promote attachment and spreading of some cultured cell lines. Direct adhesion of melanoma cells requires high densities of adsorbed sulfatide. In the presence of laminin, however, specific adhesion of some cell types to sulfatide is strongly stimulated and requires only low densities of adsorbed lipid, suggesting that laminin is mediating adhesion by crosslinking receptors on the cell surface to sulfatide adsorbed on the plastic. Although thrombospondin also binds to sulfatides and to melanoma cells, it does not enhance but rather inhibits direct and laminin-dependent melanoma cell adhesion to sulfatide, presumably because it is unable to bind simultaneously to ligands on opposing surfaces. Thus, sulfated glycolipids can participate in both laminin- and thrombospondin-mediated cell adhesion, but their mechanisms of interaction are different.

Glycoconjugates and cell adhesion: the adhesive proteins laminin, thrombospondin and von Willebrand's factor bind specifically to sulfated glycolipids

The adhesive glycoproteins laminin, thrombospondin and von Willebrand's factor bind specifically and with high affinity to sulfated glycolipids, and it is this binding that probably accounts for their ability to agglutinate glutaraldehyde-fixed erythrocytes. The 3 proteins differ, however, in the effect of sulfated polysaccharides on their binding to sulfatides. Fucoidan strongly inhibits binding of both laminin and thrombospondin, but not of von Willebrand's factor, suggesting the involvement of laminin or thrombospondin or other unknown sulfatide-binding proteins in specific cell interactions that are also inhibited by fucoidan. Thrombospondin adsorbed onto plastic promotes the attachment and spreading of G361 melanoma cells. Interestingly, fucoidan and an antibody directed against the sulfatide-binding domain of thrombospondin selectively inhibit spreading but not attachment. Sulfatides, but not neutral glycolipids or gangliosides, when adsorbed onto plastic also promote attachment and spreading of G361 melanoma cells. Direct adhesion of G361 cells requires high densities of sulfatide. In the presence of laminin, however, specific adhesion of G361 cells to sulfatide is strongly stimulated and requires only low densities of adsorbed lipid, suggesting that laminin mediates adhesion by cross-linking receptors on the melanoma cell surface to sulfatide adsorbed onto the plastic. Although thrombospondin binds to sulfatide and to G361 cells, it does not enhance but rather inhibits direct and laminin-dependent G361 cell adhesion to sulfatide, presumably because it is unable to bind simultaneously to ligands on opposing surfaces. Thus, sulfated glycoconjugates participate in both laminin- and thrombospondin-mediated cell adhesion, but their mechanisms of interaction are different.

Many pulmonary pathogenic bacteria bind specifically to the carbohydrate sequence GalNAc beta 1-4Gal found in some glycolipids

Pneumonia is one of the most common causes of death from infectious disease in the United States. To examine the possible role of carbohydrates as adhesion receptors for infection, several pulmonary pathogenic bacteria were studied for binding to glycosphingolipids. Radiolabeled bacteria were layered on thin-layer chromatograms of separated glycosphingolipids, and bound bacteria were detected by autoradiography. The classic triad of infectious bacteria found in cystic fibrosis, Pseudomonas aeruginosa, Haemophilus influenzae, and Staphylococcus aureus, along with other bacteria commonly implicated in typical pneumonia, such as Streptococcus pneumoniae, Klebsiella pneumoniae, and certain Escherichia coli, bind specifically to fucosylasialo-GM1 (Fuc alpha 1-2Gal beta 1-3GalNAc beta 1-4Gal beta 1-4Cer), asialo-GM1 (Gal beta 1-3GalNAc beta 1-4Gal beta-1-4Galc beta 1-1Cer), and asialo-GM2 (GalNAc beta 1-4Gal beta 1-4Glc beta 1-1Cer). Bacteria maintained in nutrient medium bind better than the same cells suspended in buffer. They do not bind to galactosylceramide, glucosylceramide, lactosylceramide, trihexosylceramide, globoside, paragloboside, Forssman glycosphingolipid, or several other glycosphingolipids tested, including the gangliosides GM1, GM2, GM3, GD1a, GD1b, GT1b, and Cad. The finding that these pathogens do not bind to lactosylceramide suggests that beta 1-4-linked GalNAc, which is positioned internally in fucosylasialo-GM1 and asialo-GM1 and terminally in asialo-GM2, is required for binding. beta-N-Acetylgalactosamine itself, however, is not sufficient for binding, as the bacteria did not bind to globoside, which contains the terminal sequence GalNAc beta 1-3Gal. These data suggest that these bacteria require at least terminal or internal GalNAc beta 1-4Gal sequences unsubstituted with sialyl residues for binding. Other bacteria, including Mycoplasma pneumoniae, Streptococcus pyogenes, Salmonella species, and some E. coli, do not bind to the GalNAc beta 1-4Gal sequence. The biological relevance of these data is suggested by our finding that substantial amounts of asialo-GM1 occur in human lung tissue.