Search

Background: The importance of matrix metalloproteinases in angiogenesis, tumor growth, and metastasis is well known. However, little is known about the role of matrix metalloproteinases in the formation of hemangiomas and about the possible therapeutic use of matrix metalloproteinase inhibitors in aggressive vascular tumors., Purpose: To study the role of matrix metalloproteinase in vascular tumors, we tested the antineoplastic activity of a synthetic inhibitor of matrix metalloproteinases, batimastat, on an experimental model of hemangioma, formed by murine endothelioma cells transformed by polyoma middle-T oncogene (eEnd.1)., Methods: The effect of batimastat was studied in vivo on the formation of hemorrhaging, cavernous hemangiomas by eEnd.1 endothelioma cells injected subcutaneously in nude mice and on the angiogenic response induced by an endothelioma cell supernatant embedded in a pellet of reconstituted basement membrane (Matrigel). The effect of batimastat was investigated in vitro on endothelial cell proliferation, motility, and invasion of a layer of Matrigel., Results: Daily treatment with batimastat (30, 3, and 0.3 mg/kg at the site of eEnd.1 cell injection) inhibited tumor growth, with increased doubling time. The carboxamide derivative of batimastat, BB-374, a poor inhibitor of matrix metalloproteinase activity, was less active in reducing hemangioma growth. Histologic analysis of treated tumors indicated a reduction in the size of blood-filled spaces and in hemorrhage. Batimastat also inhibited the angiogenic response induced by cultured eEnd.1 endothelioma cell supernatant embedded in a pellet of Matrigel. Batimastat significantly inhibited endothelial cell invasion in vitro through a layer of Matrigel, but it showed no direct cytotoxic activity., Conclusions: Batimastat reduces in vivo growth of experimental hemangiomas, most probably by blocking endothelial cell recruitment by the transformed cells or by interfering with cell organization in vascular structures., Implications: These results confirm the importance of matrix metalloproteinase in endothelial cell recruitment that occurs in angiogenesis and in the formation of vascular tumors and suggest a therapeutic potential for synthetic matrix metalloproteinase inhibitors.

Polyoma virus middle T-transformed murine endothelioma cell lines provide a useful model for studying vascular lesions such as hemangiomas, hemangiosarcomas, and Kaposi's sarcoma and tumor-associated angiogenesis. In vivo they produce fast-growing, hemorrhaging, cavernous blood-filled hemangiomas, mainly formed by recruited host endothelial cells, suggesting an angiogenesis-like process underlying the lesion. The molecular mechanism(s) responsible for the recruitment of host endothelial cells by endothelioma cells has not yet been identified. We found that five different cultured endothelioma cell lines produced a soluble factor, named endothelioma-derived motility factor (EDMF) that stimulates chemotaxis (motility induced by a gradient of soluble attractant), haptotaxis (motility in response to substrate-bound attractant), and chemoinvasion (migration through a layer of reconstituted basement membrane, Matrigel) of normal human, bovine, and murine endothelial cells. The inhibitory effect of actinomycin D and of enzymatic treatment on its activity proved that EDMF is a protein. EDMF binds to heparin, since its activity was inhibited by heparin, and it was retained on a heparin-Sepharose column. Its molecular weight, as assessed by Sephacryl S-200 gel filtration, ranges from 40,000-65,000. Although in many aspects EDMF is similar to vascular permeability factor-vascular endothelial growth factor, this was not detected in endothelioma cell supernatants, as assessed by enzyme-linked immunosorbent assay, thus indicating that EDMF might be related to, but is not identical with, vascular permeability factor. Our findings support the notion that recruitment of host endothelial cells by endothelioma cells in vivo might be mediated by a still unidentified, soluble factor that stimulates and directs endothelial cell migration.

Microvascular murine endothelial cells lines transformed by middle T oncogene of polyoma virus maintain the biological characteristics of nontransformed microvascular endothelial cells (EC). By using cell lines originated from different anatomical districts (thymus, brain, heart, and skin), we demonstrated that murine granulocyte-colony-stimulating factor (G-CSF) induces proliferation of murine microvascular endothelial cells at nanomolar concentrations without any cooperation with fetal calf serum. The proliferative effect on murine cells is less than that elicited by epidermal growth factor (EGF), used as standard for this function. G-CSF also promotes the migration of tEnd.1 endothelial cell line assayed by Boyden chamber technique. The analysis of transcript for G-CSF receptor (G-CSFR) by Northern blot hybridization and by reverse-transcriptase polymerase chain reaction (RT-PCR) shows that these cell lines have specific mRNA, with the size of that present in myeloid cells. These results indicate that G-CSF operates in the microvascular endothelial cells by a mechanism related to the presence of a specific receptor.

Background: Stable anchorage of circulating cancer cells to the vasculature is a critical step in the formation of hematogenous metastases. Although the basement membrane glycoprotein laminin clearly plays a crucial role in this event, the exact interactive pathways among cancer cells, laminin, and the vessel wall have not been elucidated. In a previous study, we identified synthetic peptide G, which contains the laminin-binding domain of the 67-kd laminin receptor and which inhibits tumor cell adhesion to endothelial cells., Purpose: To assess the role of the interaction between laminin and the 67-kd laminin receptor in hematogenous metastasis formation, we studied the effect of peptide G on melanoma cell behavior in vivo and in vitro., Methods: The effect of peptide G and control peptides was studied in vivo on lung retention and colonizing potential of murine (B16BL6) and human (A2058) melanoma cells injected intravenously in C57BL/6 and nude mice, respectively. In addition, their effect on cell adhesion and chemotaxis to laminin and on binding of iodine 125-labeled laminin to cells was studied in vitro., Results: In vivo, pretreatment of cells with peptide G resulted in a two- to 10-fold significant increase in the number of experimental lung metastases. A significant relative increase in lung retention of peptide G-treated tumor cells was observed 48 hours after injection, although after 4 hours a partial reduction was observed. In vitro, peptide G significantly increased laminin binding and cancer cell adhesion to laminin and subendothelial matrix, whereas chemotaxis to laminin was significantly inhibited., Conclusions: Peptide G differentially affected the biological response of cancer cells to laminin. In vitro, it increased laminin binding and cell adhesion to laminin and subendothelial matrix, whereas it inhibited cell chemotaxis to laminin. In vivo, the overall effect of peptide G was an augmentation of lung metastasis., Implications: Our findings suggest that direct adhesion of tumor cells to the subendothelial matrix is a main pathway for hematogenous metastases and that tumor cell-matrix interaction may be more relevant than tumor cell-endothelial cell attachment in this process.

Background: Extracellular matrix components are known to modulate mesangial cell functions as adhesion, motility and proliferation. Among other extracellular matrix components, mesangial cells have been recently described to secrete thrombospondin (TSP), a high molecular weight glycoprotein, produced by several cell types, and known to play a role in embryogenesis, wound healing, angiogenesis, and tumorigenesis. The aim of this work was the functional and molecular characterization of TSP interactions with mesangial cells., Experimental Design: Adhesion of mesangial cells to TSP-coated plastic, and chemotaxis in the Boyden chamber assay were tested. In order to identify TSP active domains, heparin, known to bind to the amino-terminal region of TSP, four monoclonal antibodies directed against specific domains of the molecule, and TSP fragments obtained by enzymatic digestion were used., Results: We found that TSP induces mesangial cell adhesion and chemotaxis, in a dose dependent manner. Adhesion was inhibited by antiserum against TSP, and by an anti-CD36 monoclonal antibody tested in the presence of heparin, but not by the peptide Gly-Arg-Gly-Asp-Ser. We have also found that only the carboxy-terminal end of TSP retains the adhesive properties of the molecule, while all the fragments tested showed some degree of chemotactic activity., Conclusions: We conclude that TSP modulates mesangial cell adhesion and motility, thus acting as a potential autocrine and paracrine regulator of mesangial cell functions in normal and pathologic conditions.

Cells derived from retinoblastomas grow slowly in vitro and only very rarely form tumors in nude mice. Matrigel, a mixture of components normally found in basement membranes, promotes the growth of Y-79 and WERI-Rb1 retinoblastoma (Rb) cells when added to suspension cultures of the 2 Rb cell lines. It also substantially increases cell adhesion in vitro. Y-79 cells, seeded into a Matrigel matrix, form round colonies over a 3-week period similar to those of control, weakly metastatic murine melanoma cells. In vivo, s.c. co-injection of Matrigel with either Y-79 or WERI-Rb 1 cells into nude mice promotes retinoblastoma tumor formation. Transplantation of as few as 1,000 cells allows for xenografting under these conditions, while no tumors were observed in the absence of Matrigel, even at 10 x 10(6) cells/inoculum. The tumors produced have the expected morphology and express an mRNA for a highly specific retina/retinoblastoma marker protein, the interphotoreceptor retinoid-binding protein. Thus, the xenografts obtained maintain the original morphological and molecular characteristics of the injected cells and represent a useful model for in vivo studies of retinoblastoma growth and treatment.

We previously reported that thrombospondin (TSP) induces endothelial cell (EC) adhesion, spreading and motility, suggesting that it can play a role in angiogenesis. We then studied whether TSP might modulate EC response to known angiogenic stimuli in vitro. Here we describe that TSP inhibits EC chemotactic response to basic fibroblast growth factor (bFGF). Furthermore, TSP and its 140 kD fragment reduce EC proliferative response to serum and bFGF. These data support the indicated role of TSP and its 140 kD fragment in angiogenesis and in related pathologies including tumor malignancy.

Stable attachment of cancer cells to the endothelium is a key step in the formation of metastasis. In this study, we have investigated the possibility that interaction between laminin and its Mr 67,000 high-affinity receptor (67 LR) could play a major role in this process. Scatchard analysis of laminin-binding studies showed that bovine aortic endothelial cells exhibit 46,000 high-affinity receptors that mediate, at least in part, the attachment of highly invasive melanoma cells. This endothelial cell-melanoma cell interaction was significantly inhibited by soluble laminin and by anti-laminin antibodies. Peptide G, an eicosapeptide derived from the complementary DNA sequence of the 67 LR precursor (IPCNNKGAHSVGLMWWMLAR) that specifically binds to laminin and presumably contains the active ligand-binding site of the receptor, specifically prevented attachment of the melanoma cells to both the bovine aortic endothelial cell monolayer and human umbilical vein endothelium. Thus, peptide G may selectively interfere with the metastatic cascade by inhibiting tumor cell attachment to endothelium via the laminin-67 LR pathway and is a potential new antimetastatic agent.

Natural or recombinant neutrophil activating cytokine (IL-8) induced migration across polycarbonate filters of human A 2058 melanoma cells. Anti-IL-8 antibodies blocked IL-8 induced melanoma cell migration. Checkerboard experiments revealed a gradient-dependent response of A2058 melanoma cells to IL-8. Filters exposed to IL-8 and washed supported melanoma cell migration, thus implying a haptotactic component in the response. The homologous polypeptide platelet factor 4 was inactive. The observation that IL-8 affects melanoma cells emphasizes the need for a comprehensive analysis of the spectrum of action of platelet factor 4-related peptides. The effect of the inflammatory cytokine IL-8 on melanoma cells may be relevant to augmented secondary localization of tumors at sites of inflammation.

Three malignant tumors (3LL carcinoma, mFS6 sarcoma, and B16 F1 melanoma) were transplanted in mice with congenital (beige) or acquired (anti-asialo GM1-treated) defects of natural killer cell (NK) activity. The macrophage content of the neoplastic tissues was not influenced by host NK activity levels. These data suggest that NK cell-mediated resistance does not play an appreciable role in the regulation of the levels of tumor-associated macrophages in established malignancy.

The effects of estradiol and two synthetic progestins (ORG2058 and R5020) on the expression of the high-affinity, metastasis-associated laminin receptor in two human breast carcinoma cell lines were examined. The T47D cell line contains estrogen and progesterone receptors, but the MDA-MB 231 cell line lacks both receptors. Treatment of T47D cells with 10(-9) M estradiol alone results in a three-fold increase (P less than or equal to .05) in the steady-state level of laminin receptor mRNA determined by RNA blot analysis as well as in cell-surface, laminin receptor expression that is evaluated by immunofluorescence. No effects of estradiol on the receptor-negative MDA-MB 231 cells were observed. Untreated and steroid-treated MDA-MB 231 cells had higher levels of laminin receptor mRNA than did untreated or estradiol-treated T47D cells. A more dramatic increase (five-fold; P less than or equal to .005) of mRNA and cell-surface expression in T47D cells was observed after treatment with estradiol plus 10(-8) M progestin or with progestin alone. Estradiol treatment also increased chemotaxis and haptotaxis of T47D cells but not of MDA-MB 231 cells to laminin; it had no effect on the attachment of these latter cells to laminin. Interestingly, treatment with estradiol plus progestin or progestin alone significantly increased the attachment of T47D cells to laminin but did not have an effect on either haptotaxis or chemotaxis to laminin. These results suggest that the various cell-laminin interactions are mediated by different mechanisms. The augmentation of laminin receptor mRNA by estrogen and progesterone treatment in hormone receptor-positive cells, but not in cells that lack these receptors, may relate functionally to the difference in the clinical aggressiveness between classes of breast cancers.

Polyclonal antisera were made against biochemically purified laminin receptor protein as well as against synthetic peptides deduced from a complementary DNA clone corresponding to the COOH-terminal end of the laminin receptor (U.M. Wewer et al., Proc. Natl. Acad. Sci. USA, 83: 7137-7141, 1986). These antisera were used to study the potential role of laminin receptor in laminin-mediated attachment and haptotactic migration of human A2058 melanoma cells. The anti-laminin receptor antisera reacted with the surface of suspended, nonpermeabilized melanoma and carcinoma cells. The anti-laminin receptor antisera blocked the surface interaction of A2058 cells with endogenous laminin, resulting in the inhibition of laminin-mediated cell attachment. The A2058 melanoma cells migrated toward a gradient of solid phase laminin or fibronectin (haptotaxis). Anti-laminin antiserum abolished haptotaxis on laminin but not on fibronectin. Synthetic peptide GRGDS corresponding to the fibronectin cell-binding domain inhibited haptotaxis on fibronectin but not on laminin. Both types of anti-laminin receptor antisera inhibited haptotaxis on laminin but not on fibronectin. Using immunohistochemistry, invading human carcinoma cells in vivo exhibited a marked cytoplasmic immunoreactivity for the receptor antigen. Together these findings indicate a specific role for the laminin receptor in laminin-mediated migration and that the ligand binding of the laminin receptor is encompassed in the COOH-terminal end of the protein.

Pseudopodia protrusion is a prominent feature of actively motile cells in vitro and invading tumour cells in vivo; however, the function and regulation of pseudopodia are poorly understood. Tumour autocrine motility factor (AMF) represents a new class of cytokines which are secreted by tumour cells and embryonic cells and induce random motility in the producer cells or in heterologous cells with appropriate receptors. Here we report that a major effect of this factor is to induce the extension of cell pseudopodia before cell translocation. Using a new method to quantify and isolate pseudopodia, we find that human breast carcinoma cell AMF (at concentrations of 1 nM or below) stimulates random pseudopodia formation in a dose-dependent and time-dependent manner. Anti-AMF antibodies inhibit pseudopodia protrusion and cell motility, showing the importance of pseudopodia formation during locomotion. AMF-stimulated motility and pseudopodia formation occur on a wide variety of adhesive substrata which suggests that certain intrinsic motility events are independent of the attachment mechanism. Induced pseudopodia show a prominent axial actin network in the electron microscope. The number of laminin receptor and fibronectin RGD recognition sites is increased by a factor of 20 in the induced pseudopodia when compared to the average distribution in unstimulated cells. Exploratory pseudopodia regulated by cell-derived motility factors contain receptors for matrix proteins and could serve as 'senseorgans' essential to the process of cell locomotion.

Tumour-associated macrophages (TAM) isolated from five murine sarcomas had a relatively high frequency of I-A+ cells, with mean values of 27% (mFS6), 52% (MN/MCA1), 68% (N3), 62% (N4) and 98% (J3) for TAM compared to 12% for resident peritoneal macrophages. Expression of I-E in TAM was also high (29%) in the only sarcoma (N4) examined in this respect. Expression of I-A by TAM declined in culture but exposure to lymphokine supernatants maintained and increased the frequency of I-A+ cells in TAM. Transplantation of tumours into nude mice caused a marked decrease in the percentage of I-A+ TAM in the case of the N4 sarcoma (8% compared to 48%), whereas for the MN/MCA1 sarcoma the diminution was only marginal (from 53 to 41%), TAM from murine sarcomas did not constitutively release appreciable levels of interleukin-1 (IL-1) activity. Upon stimulation with bacterial lipopolysaccharides or silica, TAM showed a limited capacity to produce and release IL-1 activity compared to peritoneal macrophages. Thus the expression of I-A antigens and the IL-1-producing capacity are uncoupled in TAM from murine sarcomas. These properties of TAM could play an important role in the generation of anti-tumour immunity and/or of suppressive T-cell circuits.

In mice bearing autochthonous 3-methylcholanthrene-induced tumors metastasis was rare, with only 2 out of 47 (4%) animals showing lung secondaries and 1 showing kidney lesions. Surgical excision of autochthonous growing tumors brought only a slight increase in incidence of metastasis (5 out of 42 mice, 12%). Cell lines were established by in vivo and/or in vitro passage from two kidney metastasis found in the same host (0.13-K1 and 0.13-K3) and from a spontaneous lung metastasis found in 2 mice (mR80/43 and mR80/17) and compared to lines from the respective primary tumors (0.13; R80/17; R80/43). Cell lines from metastases and primary tumors were heterogeneous in tumorigenicity, growth rate, metastatic potential (spontaneous), and colonizing capacity (i.v. inoculation). In particular, the mR80-43 line was more metastatic to lungs upon intravenous injection than the parent R80-43 primary tumor. Similarly the 013-K1 line from a kidney secondary caused more lung nodules when inoculated intravenously than the parent 0.13 line, but this was not the case with the 013-K3 line derived from another kidney secondary in the same host. The R80-17 and mR80-17 lines had similar lung-colonizing capacity. Lung colonizing ability was not strictly correlated to the capacity to form spontaneous metastases. Changes in lung-colonizing capacity occurred in part of the lines (013, 013-K1, R80-17, mR80-17) upon in vitro or in vivo passage. These findings with lines from spontaneous metastases from three autochthonous sarcomas extend previous observations on the heterogenous behavior of transplanted metastatic neoplasms.

The antiproliferative activity of flavone acetic acid (LM 975) was investigated on human adenocarcinoma cell lines (HCC-P2998, HCC-M1544, HCC-M1410, HT 29, LoVo), on a murine colon adenocarcinoma cell line (Colon 26), on murine pancreatic adenocarcinoma cells growing in primary culture (Pan 03) and on human normal fibroblasts (N1). No cytotoxic effects were found against human normal fibroblasts. LM 975 was active against murine adenocarcinoma Pan 03 and Colon 26, known to be sensitive in vivo too and, to variable extents, on human adenocarcinoma cell lines. LM 975 in vitro cytotoxic potency was relatively low. The high concentrations (1.0-1.4 mM) required to obtain a cytotoxic effect are, however, pharmacologically reasonable since they are comparable with drug plasma levels in mice or in patients treated with tolerable doses. After a relatively short LM 975 treatment (2 h) DNA, RNA and protein synthesis were inhibited in different proportions. In more sensitive cells LM 975 appeared to inhibit RNA synthesis more than DNA and protein synthesis. Inhibition of macromolecule synthesis after 2 h exposure was completely reversed in 24 h recovery. After 2 h treatment no detectable DNA breakage was found by the alkaline elution method, thus corroborating the idea that this compound does not act by causing DNA damage.

Membrane fluidity, determined by steady-state fluorescence polarization measurements, was correlated with metastatic capacity of murine tumor-cell lines. A correlation was observed in cell lines with different metastatic potential, and was confirmed when their lung-colonizing ability was modulated by alteration of either the membrane lipid composition or the culture conditions. Two cellular functions, motility and basement membrane invasion, were affected by the membrane lipid composition, and might explain the role of membrane fluidity observed in cancer metastasis.