Lucie Fransen

Recombinant tumor necrosis factor: its effect and its synergism with interferon-γ on a variety of normal and transformed human cell lines

Tumor Necrosis Factor (TNF), released by induced macrophages, causes tumor necrosis in animals, and preferentially kills transformed cells in vitro. Using pure, recombinant human TNF, we report here its cytotoxic action on several human transformed and non-transformed cell lines. Furthermore, remarkable synergism between TNF and interferon-gamma (IFN-gamma) was observed in a large number of human cell lines, especially breast, cervix and colon carcinomas. Some other human cell lines, not sensitive to TNF alone, became highly sensitive when IFN-gamma was present as well. We could not demonstrate a synergism between TNF and IFN-gamma on any of the lymphoma/leukemia cell lines tested. All normal human, non-transformed diploid cell lines were insensitive to TNF even in the presence of IFN-gamma. This study also confirms the observation that inhibition of protein synthesis by metabolic drugs (e.g. actinomycin D) remarkably enhances the sensitivity of several target cell lines to cytolysis by TNF.

Recombinant tumor necrosis factor: species specificity for a variety of human and murine transformed cell lines

Tumor necrosis factor (TNF) exhibits cytotoxic or cytostatic activity on a wide range of animal and human transformed cell lines. Using pure, recombinant human and mouse TNF, we examined the degree of species specificity of the in vitro TNF activity on a variety of human and murine transformed cell lines. This species specificity was studied for the TNF activity alone or in synergism with IFN-gamma. Recombinant human and mouse TNF behave remarkably similarly regarding the in vitro cytolytic/cytostatic activity. However, a certain degree of species-specific preference could be revealed as human cell lines needed a higher concentration of recombinant mouse TNF than of recombinant human TNF to attain a similar effect, while on mouse cells the reverse was true. Also, synergism with IFN-gamma seemed more effective when the target cell was treated with homologous TNF.

An in vitro model for the study of microglia-induced neurodegeneration: involvement of nitric oxide and tumor necrosis factor-α

The precise function of activated microglia and their secretory products remains controversial. In order to assess the role of microglial secretion products, we established an in vitro model of an inflammatory reaction in the brain by co-culturing microglial and neuronal cell lines. Upon stimulation with interferon-gamma and lipopolysaccharides, the microglial cells adopted an activated phenotype and secreted tumor necrosis factor-alpha (TNF-alpha), prostaglandin E(2) and nitric oxide (NO). Neuronal degeneration was quantified by measuring the concentrations of microtubule associated protein tau and neuron specific enolase, which are also used as diagnostic tool in Alzheimers disease, in supernatants. In activated contact co-cultures, the levels of these neuronal markers were significantly raised compared to non-activated co-cultures. NO-synthase inhibitors significantly diminished the rise of tau in activated co-cultures, while indomethacin, superoxide dismutase, or a neutralizing TNF-alpha antibody did not. When a chemical NO-donor or TNF-alpha were added to pure neuronal cultures, cell viability was significantly reduced. TNF-alpha increased neuronal sensitivity towards NO. There were indications that a part of the cells died by apoptosis. This model demonstrates a neurotoxic role for NO in microglia-induced neurodegeneration and provides a valuable in vitro tool for the study of microglia-neuron interactions during inflammation in the brain.

Both TNF receptors are required for direct TNF-mediated cytotoxicity in microvascular endothelial cells

The conditions under which tumor necrosis factor-alpha (TNF) induces apoptosis in primary microvascular endothelial cells (MVEC) were investigated. In the absence of sensitizing agents, TNF induced apoptosis after 3 days of incubation in confluent MVEC. In contrast, upon addition of the transcriptional inhibitor actinomycin D (Act. D), confluence was no longer required and apoptosis occurred already after 16 h. To assess the role of either TNF receptor (TNFR) type in apoptosis, MVEC isolated from mice genetically deficient in TNFR1 (Tnfr1o mice) or TNFR2 (Tnfr2o mice) were incubated with TNF in the presence or absence of Act. D. Under sensitized conditions, Tnfr2o MVEC were lysed like controls, whereas Tnfr1o MVEC were completely resistant, indicating an exclusive role for TNFR1. In contrast, in the absence of Act. D, confluent monolayers of wild-type cells were lysed by TNF, but both Tnfr1o and Tnfr2o MVEC were resistant to TNF-mediated toxicity, indicating a requirement for both TNFR types. Overexpression of the anti-apoptotic protein bcl-xL in MVEC led to a protection against the direct, but not the sensitized cytotoxicity of TNF. In conclusion, in pathophysiologically relevant conditions, both TNFR appear to be required for TNF-induced apoptosis in MVEC.

The lectin-like domain of tumor necrosis factor-α increases membrane conductance in microvascular endothelial cells and peritoneal macrophages

Herein, we show that TNF exerts a pH‐dependent increase in membrane conductance in primary lung microvascular endothelial cells and peritoneal macrophages. This effect was TNF receptor‐independent, since it also occurred in cells isolated from mice deficient in both types of TNF receptors. A TNF mutant in which the three amino acids critical for the lectin‐like activity were replaced by an alanine did not show any significant effect on membrane conductance. Moreover, a synthetic 17‐amino acid peptide of TNF, which was previously shown to exert lectin‐like activity, also increased the ion permeability in these cells. The amiloride sensitivity of the observed activity suggests a binding of TNF to an endogenousion channel rather than channel formation by TNF itself. This may have important implications in mechanisms of TNF‐mediated vascular pathology.

Membrane interaction of TNF is not sufficient to trigger increase in membrane conductance in mammalian cells

Tumor necrosis factor (TNF) can trigger increases in membrane conductance of mammalian cells in a receptor‐independent manner via its lectin‐like domain. A lectin‐deficient TNF mutant, lacking this activity, was able to bind to artificial liposomes in a pH‐dependent manner, but not to insert into the bilayer, just like wild type TNF. A peptide mimicking the lectin‐like domain, which can still trigger increases in membrane currents in cells, failed to interact with liposomes. Thus, the capacity of TNF to trigger increases in membrane conductance in mammalian cells does not correlate with its ability to interact with membranes, suggesting that the cytokine does not form channels itself, but rather interacts with endogenous ion channels or with plasma membrane proteins that are coupled to ion channels.

Soluble Interleukin-5 Receptor Alpha Is Increased in Acute Exacerbation of Chronic Obstructive Pulmonary Disease

Background: During chronic obstructive pulmonary disease (COPD) exacerbations (AE-COPD), an influx of eosinophils into the bronchial mucosa has been described. Eosinophilic cationic protein (ECP) and soluble interleukin-5 receptor α (sIL5Rα) are secreted by eosinophils and increased in eosinophilic airway diseases. Methods: We studied ECP and sIL5Rα expression in patients with COPD compared to healthy controls and smokers and investigated a possible association to viral exacerbations of COPD. Expression of sIL5Rα in serum was analyzed by ELISA and ECP by the Uni-Cap system. Induced sputum from patients with COPD was analyzed for six different respiratory viruses by nested PCR. Results: ECP and sIL5Rα were significantly elevated in AE-COPD subjects (n = 54) compared to healthy controls (n = 11, p = 0.018). Furthermore, there was a significant increase in sIL5Rα, but not in ECP, in 30 patients with virus-associated AE-COPD compared to smokers without COPD (n = 16) and healthy controls. The increase in FEV1 after resolution of the AE-COPD correlated with the decrease in sIL5Rα (r = 0.269, p = 0.034). Conclusions: sIL5Rα is increased in AE-COPD and not affected by smoking like ECP. sIL5Rα is increased in patients with virus-associated AE-COPD compared to smokers and controls. Concentrations of sIL5Rα mirror changes in the clinical status and lung function. These data support the involvement of eosinophils in acute exacerbations of COPD.

Lectin-deficient TNF mutants display comparable anti-tumour but reduced pro-metastatic potential as compared to the wild-type molecule.

In this study, we characterised the anti‐tumour as well as the pro‐metastatic activities of TNF mutants deficient in their lectin‐like activity. 16 , 19 We report that, despite reduced systemic toxicity as compared to wild‐type (wt) mTNF, a (T104A) and a (T104A‐E106A‐E109A) mTNF mutant (triple mTNF) retained most of their necrotic and tumouristatic activities, as measured in a CFS‐1 fibrosarcoma and a B16BL6 melanoma tumour model, respectively. These mutants also conserved their anti‐angiogenic activity, as measured in an in vitro endothelial morphogenesis assay. 26 In contrast, the pro‐metastatic activity of the T104A and the triple mTNF mutants in the CFS‐1 fibrosarcoma and the 3LL‐R Lewis lung carcinoma tumour model was significantly lower than that of the wt molecule. These results thus indicate that the lectin‐like domain of TNF is not implicated in its necrotic, tumouristatic and anti‐angiogenic activities, but that it can contribute to the pro‐metastatic effect of the cytokine. In conclusion, in view of their reduced systemic toxicity and pro‐metastatic capacity, but their retained anti‐tumour activities, lectin‐deficient TNF mutants might prove to be therapeutically interesting alternatives to wt TNF.

Protein kinase activities in immune complexes of simian virus 40 large T-antigen and transformation-associated cellular p53 protein.

Immune complex kinase assays in the simian virus 40 system were performed by incubation of immunoprecipitates containing tumor antigens with [gamma-32P]ATP, followed by analysis of any phosphoacceptor proteins. These assays yielded mainly the viral large T-antigen and, in particular, the associated cellular p53 as endogenous substrates. The nature of these substrates was confirmed by proteolysis techniques. Under specific conditions, casein could be used as an exogenous substrate as well. The kinase reactions showed preference for ATP and MgCl2 instead of GTP or MnCl2. Both phosphoserine and phosphothreonine, but in no case phosphotyrosine, were detected after an immune complex kinase reaction. Apparently, several in vivo phosphorylation sites were recognized in vitro in both large T-antigen and p53, but the presence of some artifactual sites could not be completely excluded. Although contaminating kinases were detectable in the immune complexes, at least the p53 molecules were phosphorylated in vitro in a more specific way. This followed from several characteristics of the immune complex kinase reactions and especially from the strong inhibition of p53 phosphorylation by two anti-large-T monoclonal antibodies. It was shown that large T-antigen showed associated kinase activity, although none of our results could unambiguously demonstrate an intrinsic kinase activity of this protein. Finally, anti-p53 monoclonal antibodies only slightly affected in vitro phosphorylation reactions, whereas a p53 molecule from a simian virus 40-free, chemically transformed human cell line was not phosphorylated in vitro under any condition tested. Thus, it is highly unlikely that the p53 molecule per se carries intrinsic or even associated kinase activities.