Hemin toxicity in a human epithelioid sarcoma cell line.

The major cytotoxic component of hemin was identified as metal free protoporphyrin IX in an epithelioid sarcoma cell line (VA-ES-BJ) and a glioblastoma cell line (U-373 MG) by exposing the cell lines to the iron chelator deferoxamine, tin-protoporphyrin IX, and protoporphyrin IX. The contribution of lipid peroxidation and free radical generation to toxicity was examined using DL-buthionine-[S,R]-sulfoximine (BSO), and 21-aminosteroid (lazaroid, U74500A). Hemin caused significantly greater toxicity in VA-ES-BJ than in U-373 MG. While exogenous PpIX was more toxic than hemin in both cell lines, this toxicity was not due to iron depletion following intracellular heme formation since ferric citrate did not reverse PpIX toxicity. Pre-treatment with BSO enhanced hemin toxicity in the VA-ES-BJ cell line but not in U-373 MG, suggesting different modes of toxicity in the two cell lines. Exposure to lazaroid protected only VA-ES-BJ from protoporphyrin-induced toxicity implicating a specific sensitivity to lipid peroxidation and/or free radical generation by this cell line. These characteristics of the VA-ES-BJ cell line distinguish it from the glioblastoma and emphasize its utility for exploring cytotoxic effects of hemin and its precursors.

Camptothecin cytotoxic effects in vitro: dependency on exposure duration and dose.

A survey of in vitro cytotoxic effects of camptothecin in human epitheliod sarcoma, colon, breast and ovarian carcinomas, glioblastoma, and neuroblastoma (PNET) cell lines, was done. We chose the MTT assay to measure survival and observed that 24 h exposures to camptothecin caused consistently greater toxicity than 1 h exposures. The LD50 for camptothecin was in the 12.5-25 ng/ml range. There was a 10-fold range of growth rates measured by OD after 5 days exposure and varied expression of MDR1 in these cell lines--none of which could be correlated with tumor sensitivity to drug. The most sensitive cell lines were colon and glioblastoma, and the most resistance were ovarian, breast and epithelioid sarcoma.

Va-es-bj - an epithelioid sarcoma cell-line.

A tumor cell line was derived and established from a bone marrow aspirate of a 41-year-old male who presented with motor symptoms of cord compression and an epidural tumor with osseous, pulmonary and hepatic metastases. The epidural tumor was an epithelioid sarcoma staining positive for cytokeratin markers of epithelial differentiation AE-1 and AE-3. This cell line was composed of lowly refractile, multinucleated giant or mononuclear round or elongated cells with monopolar or bipolar short processes. When permitted to grow beyond confluence these cells clumped up and formed mounds progressing to free floating spheroids. The cells were characterized by chromosomal triploidy with marker chromosomes, rapid growth in nude mice and secretion of immunoreactive and biologically functional granulocyte macrophage colony stimulating factor. This epithelioid sarcoma cell line stained positive for AE-1, AE-3, vimentin, epithelial membrane antigen and was designated VA-ES-BJ.

Reversion of p-glycoprotein mediated multidrug-resistance to vincristine and adriamycin by psc-833, a cyclosporine derivative in human neuroblastoma cell-lines.

We evaluated the effect of PSC-833 a cyclosporine derivative for its MDR1 reversing activity on vincristine and adriamycin resistance in human neuroblastoma cell lines. The cell lines have high, intermittent and low MDR1 expression. Resistance to vincristine and adriamycin was inverse to the degree of MDR1 expression. Resistance could be lowered with longer exposures to drug in all three cell lines. PSC-833 was able to reverse resistance in the SK-N-FI cells to a greater degree than in the two cell lines with lower expression of MDR1. These data suggest that MDR1 may play a protective role against vincristine and adriamycin in human neuroblastoma cells, and PSC-833 can reverse this protection. The duration of exposure to drug alters the response and the effect of PSC-833 on reversal of resistance. Where MDR1 is minimally expressed, PSC-833 is ineffective. Overall activity of vincristine and its modulation by PSC-833 were more consistent than that of adriamycin. Multiple drug resistance mechanisms are complex and vary among different cell lines when challenged with MDR1 drugs and reversing agents.

Cerebral ventricular fluid distribution of subcutaneous granulocyte-macrophage colony stimulating factor.

After treating a child suffering from disseminated primitive neuroectodermal tumor and hydrocephalus with bilateral ventriculostomies, we administered intravenous high dose thiotepa followed by a single subcutaneous dose of GM-CSF 24 hours later. The appearance and clearance of GM-CSF were measured from both ventricles, one of which was surrounded by tumor. Peak levels of GM-CSF were recorded simultaneously in both ventricles 11 hours after injection. Complete clearance from injection required 15 hours and 31 hours for the tumor-free right ventricle and the tumor-involved left ventricular wall, respectively. Tumor response was ephemeral and limited to ventricular fluid WBC, protein and LDH decreases. Microglia were detected; however, there was no evidence of anti-tumor activity in biopsied tumor tissue. Tumored regions of the brain may have perturbation of GM-CSF distribution and clearance which may contribute to the lack of microglial activity.

A saturation threshold for taxol cytotoxicity in human glial and neuroblastoma cells.

The cytotoxic effects of taxol at concentrations of 0.001-1.0 microgram/ml were determined in two human glioblastoma multiforme, two neuroblastoma and two primitive neuroectodermal tumor cell lines. The neuroectodermal cell lines were established from previously treated patients, while the glioblastomas were from untreated patients. At exposure durations of 1, 4 and 24 h there was an inverse taxol concentration-survival relationship for all six cell lines as measured by the MTT method. Significant differences in sensitivity to taxol among these cell lines were observed; the most resistant cell line SK-N-FI is characterized by very high levels of MDR1 expression and the most sensitive SK-N-AS by very low levels. An additional level of complexity concerns a saturation threshold for taxol-induced cytotoxic effects which when reached precludes additional effects of prolonged or additional exposure. Tumors of the brain and peripheral nervous system appear to be sensitive to taxol. However, dosage necessary to maximize cytocidal effects in tumors requires knowledge of at least the range of each tumors constitutive sensitivity to taxol and a way to optimize drug delivery.

Sequential addition of deferoxamine and hemin inhibits glioma tumor-cell growth.

The toxicity of deferoxamine, a potent iron chelator may be antagonized by hemin a potential iron source. Using the MTT assay we explored the effects of different concentrations and schedules of deferoxamine and hemin or deferoxamine and iron free tin protoporphyrin (SnPP) in two neuroblastoma (VA-N-BR, SK-N-AS), and two glioblastoma multiforme (VA-MG-SL, and U-373 MG) cell lines. In these cell lines, survival after exposure to 10 mug/ml deferoxamine for three days ranged from 28% to 59%. Incubation with hemin alone, had variable effects on growth depending on the cell line. Concomitant exposure to equimolar concentrations of deferoxamine and hemin resulted in the reversal of deferoxamine induced toxicity. Surprisingly, in the glioblastoma multiforme cell lines sequential exposure to deferoxamine and then hemin resulted in additional toxic effects rather than abrogation of deferoxamine toxicity. Sequential exposure of all cell lines to deferoxamine, hemin, and the chemotherapeutic agent thiotepa resulted in enhanced toxicity over any drug used alone. Exposure of the cells to deferoxamine and SnPP or deferoxamine and FeCl3 did not result in increased toxicity. These results implicate iron as the toxic element but indicate that the iron is only toxic when presented to the cell bound to protoporphyrin, such as found in hemin.

Establishment of a new cell line, VA-N-BR, from a primitive neuroblastoma tumor of the abdomen.

An established cell line was derived from a neuroblastoma originating in the abdomen of a six year old male. This patient had increased urinary homovanillic acid on two occasions, and the tumor had a unique pattern of local peritoneal and hepatic dissemination but not distant spread. The cells were initially highly refractile round cells forming firmly adherent cell clumps which later formed irregularly shaped bodies with mono- or bipolar short processes attached to the plastic surface of the culture flask. Chromosomal tetraploidy and marker chromosomes distinguished these cells from Ewings sarcoma, neuroepithelioma and the common disseminating neuroblastoma. Immunohistochemical staining positive for neurofilament, chromogranin, desmin and vimentin suggested it to be relatively undifferentiated. When heterotransplanted to nude mice, these cells expressed only desmin and vimentin reactivity. This neuroblastoma cell line was established and designated as VA-N-BR. It appears to be different from the more common partially differentiated neuroblastoma of childhood.

Deferoxamine and human neuroblastoma and primitive neuroectodermal tumor cell lines.

Deferoxamine at concentrations of 3.28 microM to 32.8 microM for five days causes in vitro growth inhibitory and cytolytic activities in human neuroblastoma and neuroectodermal cell lines. These effects are most likely due to intracellular iron depletion and vary with each cell line tested. A 3.28 microM threshold for cytolytic effects was observed in the most sensitive cell lines SK-N-DZ and SK-PN-LI, while proportionate responses ranging from lysis to relative growth inhibition was observed in the more refractory VA-N-BR, SK-N-LO and SK-N-AS cell lines. Cytolytic effects may represent an artifact of the in vitro setting where maximum exposure of cells to the drug can be achieved. Different sensitivities to deferoxamine in controlled in vitro conditions suggest variable anti-tumor effects can be expected in the clinical setting. Deferoxamine in patients may require a maximum tolerated dosage as a constant infusion for greater than 72 hours.

Human meningeal sarcoma heterotransplants in Nu/Nu mice treated with L-phenylalanine mustard (L-PAM) and dianhydrogalactitol (DAG).

The clinical characteristics and response to therapy of a patient with meningeal sarcoma, one of four patients over a twenty-five year period at Memorial Sloan-Kettering Cancer Center, are described. The light and electron microscopic characteristics of the primary tumor and as a heterotransplant in nude mice showed minimal differences. The tumor was resistant to conventional chemotherapeutic agents, both in the patient, in vitro as a cell culture and ex vivo as a heterotransplant. Only a combination of L-phenylalanine mustard and dianhydrogalactitol produced a limited response in both the patient and the mice. This combination may have some utility in future cases.

Effects of vitamin E on fibroblast fibronectin.

Fibronectin is a glycoprotein found mainly on fibroblast cell surfaces. Cultured human skin fibroblasts grow as monolayers in regular linear and whorl-shaped patterns. Exposure to a certain vitamin E analogue, D-alpha-tocopheryl-polyethylene-glycol-1000-succinate, induces concentration-dependent morphological and pattern changes. Using indirect immunofluorescence techniques, fibronectin, a cell protein associated with cell structure and intercellular patterns, was identified within the cells and the intercellular spaces. At a low concentration, the vitamin E analogue caused inhibition of the intercellular distribution of fibronectin without changes in intracellular fibronectin or significant disruption of the morphological patterns. At higher concentrations morphological changes ensued. These data suggest that the mechanism of vitamin E-induced morphological and intercellular pattern changes may not be manifested through inhibition of fibronectin deposition or secretion.

Human neuroblastoma cells and 13-cis-retinoic acid.

To a limited extent, we corroborated a previous report of human neuroblastoma sensitivity to 13-cis-retinoic acid. Seven cultured human neuroblastoma, two primitive neuroectodermal tumor, and one melanoma cell line were exposed to 0.001 to 10.0 microM 13-cis-retinoic acid for six to fourteen days. The neuroblastoma cell line, SK-N-DZ, was the only cell line lysed by all concentrations of 13-cis-retinoic acid. The other cell lines were refractory to concentrations as high as 10 microM. Increased cell process formation was observed in three neuroblastoma, SK-N-SH, SK-N-BE, SK-N-LE, and one melanoma cell line. We conclude that sensitivity to 13-cis-retinoic acid is unevenly distributed among histogenetically similar tumors from different patients.

Importance of clinical exposure on verapamil enhancement of adriamycin-vincristine cytotoxicity in human neuroblastoma.

A review of the effects of Ca2+ channel blockers on vincristine and adriamycin cytotoxic activity revealed the majority of the studies to be centered on murine leukemia or human tumors of adults. We investigated the effects of verapamil, a Ca2+ channel blocker, on human neuroblastoma cells as a representative tumor of childhood. By choosing a cell line derived from a patient previously treated with vincristine but not adriamycin, we attempted to evaluate whether previous clinical exposure to a specific drug could influence the verapamil enhancement of the cytotoxic effects of that drug.

Neuroectodermal tumor and monoclonal antibodies.

Antibody-producing clones were obtained by hybridization of spleen cells from mice immunized with whole cultured human tumor cells and mouse myeloma cells, P3UX59AG8. The tumor cells were derived from a peripheral primitive neuroectodermal tumor. One of the antibodies produced by these clones reacts with the original cell line, SK-PN-DW, and other more differentiated neuroectodermal tumors such as neuroblastomas and melanomas. The cell line SK-PN-DW contains antigenic determinants recognized by monoclonal antibodies raised against melanoma, neuroblastoma, and human fetal brain. These data indicate that this primitive neuroectodermal tumor is derived from the neuroectoderm.

Ocular tumors react with anti-neuroblastoma monoclonal antibodies.

The reactivities of mouse monoclonal antibodies directed against human neuroblastoma and peripheral melanoma associated antigens, with human retinoblastoma and choroidal melanoma cell lines were tested. Segregation of antigenic determinants according to each tumor class and cell line were observed. Three retinoblastoma cell lines and one fresh tumor explant showed determinants detected by two human neuroblastoma antisera, while the choroidal melanoma showed one determinant present on a peripheral melanoma but not neuroblastomas nor retinoblastomas, suggesting certain potential distinctive tumor related determinants.

Effect of phase I and II chemotherapeutic agents against human lymphomas heterotransplanted in nude mice.

Ten chemotherapeutic agents, mostly phase I and II drugs, were tested for activity against two human lymphomas heterotransplanted in nude mice. Three of these agents have been tested in phase II trials in patients with lymphoma and found to lack activity; a corresponding lack of activity was found in lymphoma-bearing nude mice. Apart from cyclophosphamide, which is known to have activity against lymphoma and was used as a positive control, only dianhydrogalactitol (DAG) had antitumor activity in the lymphoma-bearing nude mice. Tumor regressions induced by DAG in a heterotransplanted diffuse histiocytic lymphoma were determined to be significant using a statistical method designed for such studies. These data suggest there should be further study of DAG in patients with lymphoma.

Effects of murine tumor necrosis factor on heterotransplanted human tumors.

A partially purified glycoprotein fraction (the G-200 II fraction) obtained from sera of CD-1 mice sensitized with Corynebacterium parvum and treated with endotoxin was designated as tumor necrosis factor (TNF). Human melanoma cells exposed to this factor in vitro had decreased tumorigenicity when injected into nude mice. Human melanoma, embryonal adenocarcinoma of the testis and colon carcinoma heterotransplanted in nude mice exhibited regressions in size following intraperitoneal injections of TNF. The responses were related to dose and duration of exposure.

A rationale for the treatment of metastatic neuroblastoma.

In developing a chemotherapeutic program for children with disseminated neuroblastoma, we established three human neuroblastoma lines in cell culture to study the effects of dibutyryl cyclic AMP, papaverine, 5-trifluoromethyl-2'-deoxyuridine, and cyclophosphamide on cell growth, biochemical behavior, and morphology. Based upon our studies, a clinical treatment program was designed. We have treated 15 patients with disseminated neuroblastoma and have established the optimum dose range and sequence of these drugs. Early results were promising; plans for continuation of clinical and experimental studies were discussed.