Expression of the monocyte chemoattractant protein-1 (MCP-1) was examined in human central nervous system tumours (glioblastomas and astrocytomas) and normal human brain. Northern blot analysis demonstrated constitutive expression of MCP-1 mRNA in 6 of 12 glioblastoma cell lines. Expression could be stimulated by interleukin (IL)-1 beta and tumour necrosis factor (TNF)-alpha in all cell lines tested. Immunoprecipitation demonstrated secretion of both isoforms, MCP-1 alpha and -beta, of the MCP-1 protein. Reverse-transcription polymerase chain reaction and Northern blot analysis on tissues demonstrated MCP-1 mRNA expression in 17 of 17 glioblastomas, 3 of 6 anaplastic astrocytomas and 6 of 6 low-grade astrocytomas, as well as in fetal brain but not in normal adult brain. In situ hybridization on 2 glioblastomas and 1 low-grade astrocytoma indicates that neoplastic astrocytes and endothelial cells express MCP-1 mRNA in vivo. Moreover, tumour cyst fluids of glioblastomas and astrocytomas were able to induce monocyte chemoattraction in an in vitro assay. This chemotactic activity was specifically neutralized by anti-MCP-1 antibodies in 9 of 10 samples, further demonstrating the production of bioactive MCP-1 in vivo and supporting an important role for this factor in the infiltration of monocytes/macrophages into tumour tissue.

A human malignant glioma cell line, LN-18, has been established in monolayer culture and subcultured for more than 115 passages. LN-18 cells grow in vitro as bipolar or stellate cells with pleomorphic nuclei, have a doubling time of about 72 h and a plating efficiency of 3%. The glial nature of these cells has been assessed by ultrastructural examination. The synthesis of glial fibrillary acidic and S-100 proteins could not be demonstrated, although the initial biopsy tissue and the early cultures were positive for the former. The presence of Ia-like antigens on the surface of these cells was demonstrated using allo and xeno antisera. LN-18 cells were also shown to synthesize large quantities of fibronectin. The injection of LN-18 cells into nude mice induced the formation of solid tumor masses that could be retransplanted every 3 weeks and showed a morphology comparable to that of the initial biopsy. Karyotype analysis revealed the presence of three marker chromosomes, constantly present before and after hetero-transplantation.

Glioblastoma cells constitutively produce various amounts of PGE2 (prostaglandin E2) in vitro. The amounts of PGE2 found in the conditioned medium of the glioblastoma cultures (less than 5 ng/ml) were not enough to inhibit the IL-2 (interleukin-2) activation of peripheral blood lymphocytes. However the amount of PGE2 produced by approximately 1 x 10(7) of the glioblastoma cells can be assumed to suppress the generation of IL-2-induced killing activity against glioblastoma cells.

In this paper, the characterization of four human malignant glioma cell lines is described. The four lines are positive for glial fibrillary acidic protein (GFAP) in variable amounts. One of them, LN 992, is positive for S-100 protein. Myelin basic protein could not be detected in any of the four lines. The four lines had high levels of CNPase activity. The karyotype shows polyploidy for all lines, with modal numbers ranging from 80 to 120 and various numbers of marker chromosomes. Particular attention has been paid to the surface phenotype and a panel of three antiglioma monoclonal antibodies (Mabs), five antimelanoma Mabs, one anti-CALLA Mab, and two anti-HLA-DR Mabs has been used in an antibody-binding radioimmunoassay for the four cell lines. Lines LN 215 and LN 235 are positive with two antiglioma Mabs, LN 992 is negative. The four lines are positive with all five antimelanoma Mabs, except for LN 992 which ist negative with Mab D5. LN 992 and LN 215 are positive with the anti-CALLA Mab N2A12. LN 308 and LN 992 are positive with anti-HLA-DR Mab D4-22. There was no correlation between the in vitro morphology of the lines and the expression of the various biochemical or surface markers. These results stress the heterogeneity of the phenotype of human malignant glioma lines. These lines will be useful tools for further immunologic studies.

Human malignant glioma cell lines and clones were incubated with various concentrations of recombinant human TNF-alpha, either alone or in combination with recombinant human IFN-gamma. The surface expression of HLA-ABC (class I) antigens and beta 2-microglobulin, was significantly enhanced by TNF-alpha alone on every cell line and clone tested. After incubation with both TNF-alpha and IFN-gamma, the surface expression of HLA-ABC antigens was only slightly higher than that observed with each cytokine alone. In contrast to IFN-gamma, TNF-alpha had no effect on the surface expression of HLA-DR (class II) antigens. Moreover, the surface expression of HLA-DR induced by IFN-gamma was unaffected by TNF-alpha. The increased expression of HLA-ABC antigens after treatment with TNF-alpha or IFN-gamma correlated with increased levels of HLA-ABC-specific mRNA. In addition, TNF-alpha, like IFN-gamma, selectively enhanced the surface expression of a tumor-associated antigen, Me14-D12, while it had no effect on the expression of various other surface antigens. In the absence of actinomycin D, TNF-alpha exhibited no direct cytotoxic/cytostatic effect on the glioma cell lines tested. These results indicate that TNF-alpha can enhance the surface expression of HLA-ABC antigens on human glioma cells in the absence of a direct cytotoxic/cytostatic effect.