Regulation of spontaneous and TNF/IFN-induced IL-6 expression in two human ovarian-carcinoma cell lines.

Autocrine and paracrine production of interleukin-6 (IL-6) is considered to be involved in the ongoing proliferation of ovarian-cancer cells. In view of the variability in IL-6 expression between various ovarian-cancer cells, we questioned whether differences in IL-6-gene regulation might be observed in ovarian tumor cells with and without IL-6 expression. The CAOV-3 cell line spontaneously secreted IL-6, which was enhanced by tumor necrosis factor-alpha (877 +/- 89 vs. 8,452 +/- 1,762 pg/ml, x +/- sd, p < 0.01). The electrophoretic mobility shift assay (EMSA) demonstrated that basic IL-6 expression was associated with DNA binding of activator protein-1 (AP-1) and nuclear factor IL-6 (NF-IL6). Nuclear factor kappa-B (NF-KB), which consisted mainly of p65-NF-KB was induced in response to TNF-alpha stimulation. A2780 cells did not express IL-6, either spontaneously or after stimulation with TNF-alpha. EMSAs, showed spontaneous AP-1 but no NF-IL6 or NF-KB DNA binding. TNF-alpha stimulation enhanced AP-1 and induced NF-KB but no NF-IL6 DNA binding in these cells. NF-IL6 protein, however, was detected in nuclear extracts of these cells by Western blotting. In contrast, IL-6-promoter transfection studies showed no difference in promoter activation between CAOV-3 and A2780. This study reveals that differential IL-6-gene expression observed in ovarian-cancer cell lines is independent of NF-IL6 activation.

Differential regulation of IL-6 promoter activity in a human ovarian-tumor cell line transfected with various p53 mutants: involvement of AP-1.

In human ovarian carcinomas, the p53 tumor-suppressor gene is frequently mutated. Interleukin-6 (IL-6) in these tumors is known to stimulate tumor-cell proliferation. In order to evaluate the effect of several p53 phenotypes on the IL-6 promoter activity, the human ovarian wild-type (wt)-p53 cell line A2780 was stably transfected with an empty plasmid (CMV) or (m)-175-, m-248- or m-273-p53. Electrophoretic mobility-shift assays revealed differences in activator protein-1 (AP-1) DNA-binding activity in the various clones. The CMV and m-273 clone had comparable amounts of AP-1. The m-175 clone displayed the least and m-248 the most pronounced AP-1 binding. Supershift analysis of AP-1/DNA complexes with antibodies against the AP-1 sub-units, c-Fos, FosB, Fra-1, Fra-2, c-Jun, JunB, and JunD, revealed that the AP-1/DNA complexes in the various clones had different compositions. Fra proteins were basically present only in m-175 and m-248 AP-1. IL-6-promoter activity was evaluated in the presence and absence of the AP-1 binding site which showed that the m-175-transfected clone has a transcriptional suppressing AP-1, whereas the CMV and the m-273 clones have an activating AP-1. Exposure of the p53 clones to tumor-necrosis factor-alpha (TNF-alpha) clearly altered the AP-1/DNA complex composition. IL-6-promoter activity was enhanced by TNF-alpha irrespective of the presence of an AP-1 binding site, while the degree of activation differed in the various clones, being most pronounced in the m-175 and m-248 clones. The results demonstrate that the basic and activated IL-6-promoter activity is differently regulated in the various p53 clones, possibly due to alterations in the AP-1 composition.

Mutual interactions between p53 and growth factors in cancer.

The function of p53 tumour suppressor protein is determined by various intrinsic properties of the protein. The effect of p53 DNA-binding, and protein-protein interactions are determined by the conformation of the protein. Thus, p53 fulfils its role in cell cycle control and the onset of apoptotic cell death, which is altered when the wild-type p53 (wt-p53) conformation is changed due to mutation. This review focuses on the communal interactions of wt- and mutant p53 (m-p53) with growth factors and shows that m-p53 affects different cell biological functions that determine the malignant behaviour of cells. P53, for instance, affects the response of cells to growth factors and growth factor-withdrawal. Furthermore, p53 is involved in the expression of several growth factor- and growth factor receptor genes. These data suggest that restoration of the wt-p53 phenotype in tumour cells with m-53 might not only affect cell cycle control and apoptotic mechanisms but could also reduce autocrine growth and restore sensitivity to physiological growth inhibitors.

Enhanced sensitivity to tumor necrosis factor-alpha in doxorubicin-resistant tumor cell lines due to down-regulated c-erbB2.

We have studied the relationship between tumor necrosis factor (TNF)-sensitivity and doxorubicin-resistance in our doxorubicin-resistant cell line panel consisting of the parental cell line GLC4 plus GLC4-Adr2x and GLC4-Adr350x with respective resistance factors of 2 and 350 compared with GLC4. At the highest dose of 1000 ng/ml TNF, GLC4 was almost completely resistant to TNF, whereas 37% and 68% growth inhibition was observed in GLC4-Adr2x and GLC4-Adr350x, respectively. Sensitivity to TNF appeared to correlate inversely with the expression and gene copies of topoisomerase IIalpha in these cell lines. The gene encoding for c-erbB2 is in the proximity of the topoisomerase IIalpha gene and its product is a known cause for TNF-resistance. We found that our doxorubicin-resistant cell lines with decreased topoisomerase IIalpha gene copies have a simultaneous decrease in c-erbB2 gene copies, probably due to linkage between these 2 genes. This reduced number of c-erbB2 gene copies resulted in decreased c-erbB2 expression and subsequently in increased sensitivity to TNF. Additionally, we tried to establish some of the mechanisms associated with TNF-resistance in GLC4 related to c-erbB2 overexpression. There was no difference in TNF-receptor-1 expression between the cell lines. Compared with the TNF-sensitive GLC4-Adr350x, GLC4 appeared to have a decreased activation of NF-kappaB after exposure to TNF that might indicate a reduced TNF-receptor function. In GLC4, increased Bcl-2 expression was found, a protein described to confer TNF-resistance. Moreover, it was demonstrated that combining TNF with the poly-ADP-ribose polymerase (PARP) inhibitors 6-aminonicotinamide and 3-aminobenzamide did not affect TNF-sensitivity in GLC4 and GLC4-Adr350x, excluding a pivotal role of PARP in TNF-resistance in these cell lines. In conclusion, our data show that doxorubicin-resistant tumor cell lines with decreased topoisomerase IIalpha gene copies can become sensitive to TNF due to loss of c-erbB2 gene copies. We also found that several mechanisms associated with c-erbB2 overexpressing contribute to TNF-resistance in GLC4.

The combined effects of IL-3 and PSC 833 on daunorubicin- and mitoxantrone cytotoxicity in two growth factor-dependent leukemic cell lines.

In the present study it was investigated whether and by which mechanisms the co-administration of interleukin-3 (IL-3) and the P-glycoprotein blocker PSC 833 can augment mitoxantrone (MX) and daunorubicin (DAU) cytotoxicity in two human growth factor dependent P-glycoprotein (P-gp) positive myeloid leukemic cell lines, Mo-7 and GF-D8. Cytotoxicity was determined in MTT assay. Increased cytotoxicity occurred in Mo-7 cells preincubated with 24h IL-3 followed by 1 h MX (cell survival: 85% +/- 6 vs 68% +/- 2, at 0.05 microM MX, P < 0.005) or DAU (79% +/- 8 vs 62% +/- 9 at 0.8 microg/ml DAU, P < 0.05). Similar results were obtained for the GF-D8 cell line. In this cell line, at 0.5 microM MX the cell survival decreased from 84% +/- 13 to 61% +/- 19 (P < 0.05) and at 5.0 microg/ml DAU from 102% +/- 8 to 69% +/- 5, (P < 0.002). The IL-3 administration did not affect the P-gp and bcl-2 protein expression, cellular MX concentration or MX efflux but coincided with an increased percentage of cells in S-phase and topoisomerase II (topo II)-alpha mRNA and topo II activity especially in the Mo-7 cell line. PSC 833 enhanced DAU cytotoxicity in both cell lines. The administration of IL-3 plus PSC 833 in the Mo-7 cell line resulted in an additive effect on DAU cytotoxicity. At 0.8 microg/ml DAU and 2 microg/ml PSC 833, the percentage surviving cells decreased from 62% +/- 9 in the absence of IL-3 to 37% +/- 3 in the presence of IL-3 (P < 0.01). The additive effect of combined treatment was most pronounced in GF-D8 cells which also had the highest P-gp expression. In contrast, PSC 833 did not modulate the MX effects, irrespective of the presence of IL-3. In summary, the results demonstrate that the combined administration of IL-3 and PSC 833 can enhance the cytotoxic effects of DAU but not MX in these P-gp positive cell lines whereas the effects of MX could be modulated by factors which influence topo II activity.

Expression of macrophage colony-stimulating factor (M-CSF), interleukin-6, (IL-6), interleukin-1 beta (IL-1 beta), interleukin-11 (IL-11) and tumour necrosis factor-alpha (TNF-alpha) in p53-characterised human ovarian carcinomas.

Ovarian carcinoma is often associated with overexpression of cytokines that may exert autocrine and paracrine growth effects, as well as genetic alterations in (proto)oncogenes and tumour suppressor genes, such as p53. The p53 protein is not only involved in the regulation of cell cycle and apoptosis, it is also involved in the in vitro regulation of IL-6 gene expression. In this study, 30 tumours of patients with a primary diagnosis of human ovarian carcinoma were characterised for p53 expression with immunohistochemistry and analysed for the expression of M-CSF, IL-6, IL-1 beta, IL-11 and TNF-alpha with Northern blotting. Nuclear and cytoplasmic p53 staining was observed in 27% (8/30), cytoplasmic staining in 30% (9/30), and no p53 staining in 43% (13/30) of the tumours. In 70% (21/30) of the tumours, M-CSF mRNA was expressed, in 40% (12/30) TNF-alpha, and in 30% (9/30) IL-6. None of the tumours expressed IL-1 beta or IL-11. The expression of TNF-alpha occurred more frequently in M-CSF positive tumours compared to M-CSF negative tumours (52% (11/21) versus 11% (1/9), P < 0.05). TNF-alpha expression was also associated with better responses to chemotherapy (P < 0.02). M-CSF expression was associated with nuclear p53 staining (P < 0.05). The p53 positive tumours more frequently expressed one or more cytokines (88%) compared with p53 negative tumours (54%, P < 0.05). This study suggests that mutations in the p53 gene might be associated with cytokine overexpression, especially M-CSF.