New insights into tumor necrosis factor-alpha in cancer: distinct isoforms exert opposing effects on tumor associated myeloid cells and tumorigenesis

Abstract

TNF-α, produced by most malignant cells, orchestrates the interplay between malignant cells and myeloid cells, which have been linked to tumor growth and metastasis. Although TNF-α can exist as one of two isoforms, a 26-kDa membrane tethered form (mTNF-α) or a soluble 17-kDa cytokine (sTNF-α), the vast majority of published studies have only investigated the biological effects of the soluble form. We demonstrate for the first time that membrane and soluble isoforms have diametrically opposing effects on both tumor growth and myeloid content. Mouse lung and melanoma tumor lines expressing mTNF-α, generated small tumors devoid of monocytes versus respective control lines or lines expressing sTNF-α. The lack of myeloid cells was due to a direct effect of mTNF-α on myeloid survival via induction of cell necrosis by increasing reactive oxygen species. Using cultured RAW 264.7 monocytic cell line and L929 fibroblasts we showed that mTNF-α increased reactive oxygen species (ROS)-mediated cytotoxicity, independent of caspase-3 activity. Although TNF-α receptors (TNFR) on target cells were required for this effect, we observed that mTNF-α-induced cell death could be mediated through both TNFR-1 and the death domain-lacking TNFR-2. ROS generation and cytotoxicity were inhibited by a mitochondrial respiratory chain inhibitor but not by an inhibitor of NADPH oxidase. Furthermore, mTNF-α mediated cytotoxicity was independent of RIP-1, a serine/threonine kinase which serves as a main adaptor protein of sTNF-α induced programmed necrosis, but rather depended on ceramide signaling pathways. Furthermore, we found that human none-small-cells lung carcinomas (NSCLCs) expressed varying levels of both soluble and membrane TNF-α. Analysis of publicly accessible NSCLC microarray database showed that gene expression patterns favoring mTNF-α were predictive of improved lung cancer survival.