heparitin-sulfate and Fibrosarcoma

Extracellular membrane-bound and secreted heat shock protein 90 (Hsp90) is known to be involved in cell motility and invasion. The mechanism of Hsp90 anchoring to the plasma membrane remains obscure. We showed that treatment of human glioblastoma A-172 and fibrosarcoma HT1080 cells with sodium chlorate, heparinase, and heparin causes a prominent loss of 2 Hsp90 cytosolic isoforms, Hsp90α and Hsp90β, from the cell surface and strongly inhibits the binding of exogenous Hsp90 to cells. We revealed that Hsp90α and Hsp90β are partly colocalized with heparan sulfate proteoglycans (HSPGs) on the cell surface and that this colocalization was sensitive to heparin. The results demonstrate that cell surface HSPGs are involved in the binding/anchoring of Hsp90α and Hsp90β to the plasma membrane.

Topics: Cell Line, Tumor; Cell Membrane; Cell Movement; Fibrosarcoma; Heparin; Heparitin Sulfate; HSP90 Heat-Shock Proteins; Humans; Membrane Glycoproteins; Protein Binding

Basal keratinocytes of the epidermis adhere to their underlying basement membrane through a specific interaction with laminin-5, which is composed by the association of alpha3, beta3, and gamma2 chains. Laminin-5 has the ability to induce either stable cell adhesion or migration depending on specific processing of different parts of the molecule. One event results in the cleavage of the carboxyl-terminal globular domains 4 and 5 (LG4/5) of the alpha3 chain. In this study, we recombinantly expressed the human alpha3LG4/5 fragment in mammalian cells, and we show that this fragment induces adhesion of normal human keratinocytes and fibrosarcoma-derived HT1080 cells in a heparan- and chondroitin sulfate-dependent manner. Immunoprecipitation experiments with Na2 35SO4-labeled keratinocyte and HT1080 cell lysates as well as immunoblotting experiments revealed that the major proteoglycan receptor for the alpha3LG4/5 fragment is syndecan-1. Syndecan-4 from keratinocytes also bound to alpha3LG4/5. Furthermore we could show for the first time that unprocessed laminin-5 specifically binds syndecan-1, while processed laminin-5 does not. These results demonstrate that the LG4/5 modules within unprocessed laminin-5 permit its cell binding activity through heparan and chondroitin sulfate chains of syndecan-1 and reinforce previous data suggesting specific properties for the precursor molecule.

Topics: Animals; Binding Sites; Cell Adhesion; Cell Adhesion Molecules; Cell Line; CHO Cells; Chondroitin ABC Lyase; Chondroitin Sulfates; Cricetinae; Embryo, Mammalian; Fibrosarcoma; Gene Expression; Heparitin Sulfate; Humans; Immunoblotting; Immunosorbent Techniques; Kalinin; Keratinocytes; Kidney; Laminin; Membrane Glycoproteins; Polysaccharide-Lyases; Proteoglycans; Recombinant Proteins; Sulfates; Sulfur Radioisotopes; Syndecan-1; Syndecans; Transfection; Tumor Cells, Cultured

Perlecan is a major heparan sulfate proteoglycan of basement membranes and cell surfaces. Because of its strategic location and ability to store and protect growth factors, perlecan has been implicated in the control of tumor cell growth and metastatic behavior. To test the role of perlecan in malignancy, we generated several stably transfected clones of HT-1080, a human fibrosarcoma cell line, harboring a perlecan cDNA in the antisense orientation. Surprisingly, clones with a reduced synthesis of perlecan mRNA and protein core grew faster, formed larger colonies in semisolid agar, and induced faster formation of s.c. tumors in nude mice than the wild-type cells. Their growth properties in vitro were independent of exogenous basic fibroblast growth factor. Reduction of perlecan expression was associated with three distinct properties typical of tumor cells with a more aggressive phenotype: enhanced migration through 8-microm-pore filter, increased invasion in Matrigel-coated filters, and heightened adhesiveness to type IV collagen substrata. These results thus provide the first evidence that perlecan may inhibit the growth and invasiveness of fibrosarcoma cells in a basic fibroblast growth factor-independent pathway and raise the possibility that perlecan may prevent the infiltration of host tissues in mesenchymal neoplasms.

Topics: Animals; Cell Adhesion; Cell Division; Cell Movement; Collagen; DNA, Antisense; Fibroblast Growth Factor 2; Fibrosarcoma; Gene Expression Regulation, Neoplastic; Heparan Sulfate Proteoglycans; Heparitin Sulfate; Humans; Male; Mice; Mice, Nude; Neoplasm Invasiveness; Proteoglycans; RNA, Messenger; Transfection; Tumor Cells, Cultured

Fibronectin (FN) has a complex pattern of alternative splicing at the mRNA level. One of the alternatively spliced segments, EDA, is prominently expressed during biological processes involving substantial cell migration and proliferation, such as embryonic development, malignant transformation, and wound healing. To examine the function of the EDA segment, we overexpressed recombinant FN isoforms with or without EDA in CHO cells and compared their cell-adhesive activities using purified proteins. EDA+ FN was significantly more potent than EDA- FN in promoting cell spreading and cell migration, irrespective of the presence or absence of a second alternatively spliced segment, EDB. The cell spreading activity of EDA+ FN was not affected by antibodies recognizing the EDA segment but was abolished by antibodies against integrin alpha5 and beta1 subunits and by Gly-Arg-Gly-Asp-Ser-Pro peptide, indicating that the EDA segment enhanced the cell-adhesive activity of FN by potentiating the interaction of FN with integrin alpha5beta1. In support of this conclusion, purified integrin alpha5beta1 bound more avidly to EDA+ FN than to EDA- FN. Augmentation of integrin binding by the EDA segment was, however, observed only in the context of the intact FN molecule, since the difference in integrin-binding activity between EDA+ FN and EDA- FN was abolished after limited proteolysis with thermolysin. Consistent with this observation, binding of integrin alpha5beta1 to a recombinant FN fragment, consisting of the central cell-binding domain and the adjacent heparin-binding domain Hep2, was not affected by insertion of the EDA segment. Since the insertion of an extra type III module such as EDA into an array of repeated type III modules is expected to rotate the polypeptide up to 180 degrees at the position of the insertion, the conformation of the FN molecule may be globally altered upon insertion of the EDA segment, resulting in an increased exposure of the RGD motif in III10 module and/or local unfolding of the module. Our results suggest that alternative splicing at the EDA exon is a novel mechanism for up-regulating integrin-binding affinity of FN operating when enhanced migration and proliferation of cells are required.

Topics: Alternative Splicing; Animals; Base Sequence; Binding Sites; Cell Adhesion; Cell Membrane; Cell Movement; CHO Cells; Cricetinae; Fibroblasts; Fibronectins; Fibrosarcoma; Genetic Vectors; Heparitin Sulfate; Humans; Mice; Molecular Sequence Data; Peptide Fragments; Rats; Recombinant Fusion Proteins; Tumor Cells, Cultured

The effects of various glycosaminoglycans (GAGs) on the growth rate of normal fibroblasts and a fibrosarcoma cell line (HT 1080) were examined. Cells were grown in 96-well microplates in the absence or presence of serum mitogens, epidermal (EGF), platelet-derived (PDGF), acidic fibroblast (aFGF), or basic fibroblast growth factor (bFGF). Cell number was measured by using crystal violet to stain cell nuclei (Westergren-Thorsson, G., Onnervik, P.-O., Fransson, L.-A., and Malmström, A. J. Cell. Phys. 147, 523-530, 1991) and also by using a Coulter counter. In the presence of serum mitogens, L-iduronate (IdoA)-rich GAGs, such as dermatan sulfate, heparin, and highly sulfated heparan sulfate, inhibited proliferation of normal cells (25-35%), whereas HT 1080 cells were unaffected or slightly stimulated. Ham's F-12 supplemented with insulin and transferrin but without growth factors was able to support growth of both cell types. Under these conditions, the IdoA-rich GAGs still suppressed growth of normal cells (40-55%), whereas HT 1080 cells again responded poorly. When growth factors were added proliferation of normal fibroblasts was further stimulated, EGF being the most effective. In the presence of either EGF, PDGF, or bFGF, IdoA-rich GAGs had a sustained inhibitory effect on normal fibroblasts (30-50% at concentrations at or above 10 micrograms/ml). However, in the presence of aFGF, both IdoA-rich and IdoA-poor heparan sulfates enhanced growth (nearly twofold after prolonged exposure) suggesting a stabilization of this growth factor. In general, IdoA-rich GAGs appear to inhibit proliferation of normal cells irrespective of the type of growth factor used. Therefore, GAGs are likely to act directly on cell-derived regulatory components, either before or after internalization. As fibrosarcoma cells were much less sensitive to growth inhibition, they may contain altered receptors for GAGs.

Topics: Blood Proteins; Cell Division; Cells, Cultured; Dermatan Sulfate; Epidermal Growth Factor; Fibroblast Growth Factor 1; Fibroblast Growth Factor 2; Fibroblasts; Fibrosarcoma; Glycosaminoglycans; Growth Inhibitors; Growth Substances; Heparitin Sulfate; Humans; Iduronic Acid; Lung; Mitogens; Platelet-Derived Growth Factor; Tumor Cells, Cultured

Recombinant nidogen fragments comprising the globular domains G1 plus G2, the rod-like domain, and the rod connected to the globe G3 were prepared from the culture media of transfected human cell clones. In addition, domains G1 and G2 were separated from each other after cleavage with chymotrypsin. The purified fragments were characterized by N-terminal sequences, electrophoresis, electron microscopy, and radioimmunoassays and the cell clones by Northern hybridization. Transfection with a construct comprising a large part of domain G3 showed high mRNA levels but no secreted protein, indicating a protein folding problem. All these fragments were used as soluble and/or immobilized ligands in binding assays. This demonstrated major binding sites on domain G2 for collagen IV and heparan sulfate proteoglycan. Affinity chromatography on zinc- and cobalt-loaded columns showed binding of domains G2 and G3 and the rod. Protein binding, but not metal binding, was abolished by reduction and alkylation of nidogen. This allowed for the isolation of several zinc-binding tryptic peptides, four from G2, two from the rod, and one from the G3 domain. Most of these short peptides contained several histidines that are likely to mediate binding. Zinc inhibited efficiently G3-mediated nidogen binding to laminin at 4 degrees C (IC50 approximately 5 microM) but less at higher temperatures. Similarly, zinc inhibited binding to collagen IV and proteoglycan at low temperatures but not at high (37 degrees C) temperatures. This indicates a complex modulation of nidogen binding to other basement membrane proteins by some, but not all, transition metals. Whether the particularly striking effects shown for zinc are of biological relevance remains to be established.

Topics: Amino Acid Sequence; Animals; Base Sequence; Basement Membrane; Binding Sites; Binding, Competitive; Cations, Divalent; Clone Cells; Cloning, Molecular; Collagen; Fibrosarcoma; Heparan Sulfate Proteoglycans; Heparitin Sulfate; Humans; Kinetics; Laminin; Membrane Glycoproteins; Mice; Molecular Sequence Data; Oligodeoxyribonucleotides; Peptide Fragments; Proteoglycans; Recombinant Proteins; Restriction Mapping; Transfection; Tumor Cells, Cultured; Zinc

The heparan sulfates synthesized in vitro by three cell lines were isolated by proteolysis and preparative anion exchange chromatography and purified free of other glycosaminoglycans by selective enzymatic degradation. The isolates from the medium of BALB/c 3T3 fibroblasts, B16.F10 melanoma cells, and a cutaneous fibrosarcoma line, along with that from the detergent-extracted cell layer of the fibroblasts, were affinity-fractionated on columns of matrix-immobilized human antithrombin III. Each heparan sulfate contained subfractions with high affinity for the proteinase inhibitor, ranging from 3-34% of the starting material. The high affinity species possessed measurable anticoagulant activities by a clotting assay (6 to 30 units/mg). Since none of the lines were derived from cell types having any known biological role in vascular homeostasis, we suggest that anticoagulant activity of the glycosaminoglycan is a random property of its primary structure.

Topics: Animals; Antithrombin III; Blood Coagulation; Cell Line; Chromatography, Affinity; Chromatography, Ion Exchange; Fibroblasts; Fibrosarcoma; Glycosaminoglycans; Heparitin Sulfate; Melanoma; Mice; Mice, Inbred BALB C; Skin Neoplasms; Tumor Cells, Cultured

We have isolated a syngeneic monoclonal antibody (HepSS-1) reactive to a murine methylcholanthrene-induced fibrosarcoma, Meth-A. HepSS-1 also bound to a wide variety of established and fresh normal cells derived from not only mice but also other species such as human, monkey, rat, hamster, and chicken. Immunoprecipitation of surface iodinated Meth-A cell extract with HepSS-1, as well as Sepharose 4B gel chromatography of Meth-A cell extract and detection of antigens recognized by HepSS-1 by a sandwich-type radioimmunoassay revealed that the HepSS-1 antigens were composed of several molecular species, with one as large as approximately 10(6) daltons. The following evidence indicates that HepSS-1 specifically recognizes an epitope present in heparan sulfate glycosaminoglycan (HS-GAG). First, treatment of Meth-A cells with heparitinase or heparinase, but not with chondroitinase ABC or hyaluronidase, resulted in the loss of HepSS-1 binding. Second, HS-GAG but not seven other types of GAG (hyaluronic acid, heparin, chondroitin, chondroitin 4-sulfate, chondroitin 6-sulfate, dermatan sulfate, and keratan sulfate) inhibited HepSS-1 binding to Meth-A cells. Third, HepSS-1 bound with HS-GAG but not with the seven other types of GAG. From the binding analysis of HepSS-1 to various modified HS-GAG and whale omega-heparin, it is additionally suggested that HepSS-1 recognizes an epitope closely related to O-sulfated and N-acetylated glucosamine. We found that NIH 3T3 cells expressed more HepSS-1 epitopes at a low cell density than at confluency and in G2 + M than in G1, whereas NIH 3T3 cells transformed with Kirsten-ras oncogene or SV-40 expressed high levels of HepSS-1 epitopes and ceased to show the density-dependent change in the amount of HepSS-1 epitopes. These observations were also reproduced by using NIH 3T3 cells transformed with a temperature sensitive Kirsten murine sarcoma virus maintained at permissive and non-permissive temperatures. Thus HepSS-1 is a first monoclonal antibody to HS-GAG and seems to be useful to elucidate changes in cell surface HS-GAG in normal cell growth and cell transformation.

Topics: Animals; Antibodies, Heterophile; Antibodies, Monoclonal; Antibodies, Neoplasm; Cell Cycle; Cell Transformation, Viral; Chickens; Cricetinae; Epitopes; Fibrosarcoma; Glycosaminoglycans; Heparitin Sulfate; Humans; Kirsten murine sarcoma virus; Macaca mulatta; Mice; Rats; Sarcoma, Experimental; Species Specificity

Several fibroblast and melanoma cell lines were studied with respect to their ability to degrade heparan sulfate (HS). The optimum pH for HS degradation by HS endoglycosidase (heparanase) for all cell lines is about 5.6, but the activity of the enzyme is still present at physiological pH. The gel permeation analysis of degradation products revealed that heparanase cuts HS in fragments about one-seventh of their original size. Since the optimum pH of HS endoglycosidase activity and the terminal molecular weight of degraded HS are the same in both cell lines, it is likely that fibrosarcoma and melanoma heparanases are identical enzymes. Cell extracts and intact cells of metastatic sublines degrade HS faster than do their nonmetastatic counterparts. The degradative activity of intact cells parallels those of cell extracts, but at a much lower level; moreover, conditioned media do not appreciably degrade HS, suggesting that heparanase is scarcely released into the medium; thus, considering the differences in degradative activity between cell extracts and intact cells or conditioned medium and the occurrence of cell lysis in a tumor in vivo, we suggest that the measure of degradative activity of intact cells in vitro is not indicative of a relationship to metastasis. The total cellular content of lytic enzymes could represent the real metastatic potential of proliferating cells, but it is also necessary to find an in vitro model better representing the behavior of neoplastic cells in vivo.

Topics: Animals; Cell Line; Chromatography, Gel; Fibrosarcoma; Glucuronidase; Glycoside Hydrolases; Heparitin Sulfate; Hydrogen-Ion Concentration; Male; Melanoma; Mice; Molecular Weight; Neoplasm Metastasis

Glycosaminoglycan synthesis was studied in cell populations of ultraviolet light-induced murine cutaneous fibrosarcoma cells under conditions of varying growth rates in vitro. After labeling with the precursors, 3H-glucosamine and 35SO4, sulfated glycosaminoglycans recoverable by direct proteolysis of the culture monolayers increased approximately 5-fold on a per cell basis from sparsely populated, exponential cell cultures (greater than 85% of cells in S, G2, or M phases) to stationary cultures inhibited by high cell density (greater than 50% of cells in G1). Within this cell surface-associated material, the relative ratio of heparan sulfate to the chondroitin sulfates was approximately 60/40% under conditions of exponential growth; in the growth-arrested cultures, the reverse ratio was found. The substratum attached material, obtained from the flask surface after ethyl glycol bis(beta-aminoethyl ether)-N,N'-tetraacetic acid (EGTA)-mediated detachment of the monolayers, contained relatively more hyaluronic acid, heparan sulfate, and chondroitin sulfates in the most actively proliferating cultures compared with the growth-inhibited cell populations. Furthermore, heparan sulfate and the chondroitin sulfates, which were enriched in the substratum material and in the cell pellet of exponential cultures, showed a relative shift to the cell surface-associated compartment (releasable by mild trypsinization after EGTA-mediated cell detachment) and to the compartment loosely associated with the pericellular matrix (i.e., released into the supernatant during detachment of the monolayers in the presence of EGTA). These results demonstrate that a variety of differences in the quantities, relative compositional ratios, and cell compartment distributions of hyaluronic acid and sulfated glycosaminoglycans occur in fibrosarcoma cell populations which vary in their rate of cell growth consequent to cell density in culture.

Topics: Animals; Cell Compartmentation; Cell Count; Cell Division; Cell Line; Chondroitin Sulfates; Fibrosarcoma; Glycosaminoglycans; Heparitin Sulfate; Hyaluronic Acid; Kinetics; Mice; Neoplasms, Radiation-Induced; Skin Neoplasms; Ultraviolet Rays