OBJECTIVE:The member of the tumor necrosis factor family LIGHT (lymphotoxin-like inducible protein that competes with glycoprotein D for herpesvirus entry on T cells; TNFSF14 (tumor necrosis factor super family protein 14) is primarily expressed in lymphocytes, in which it induces the expression of pro-inflammatory cytokines and alterations of lipid homeostasis. Recently, the protein was shown to be upregulated in obesity and to induce cytokine secretion from adipocytes.RESEARCH METHODS AND PROCEDURES:Using an automated complementary DNA (cDNA) screen, LIGHT was identified to inhibit adipose differentiation. As cellular models for adipogenesis mouse 3T3-L1, human SGBS (Simpson-Golabi-Behmel syndrome) and primary human preadipocytes differentiated in vitro were used as well as primary human adipocytes to study adipocyte functions. Analysis of lipid deposition by Oil Red O staining, mRNA expression by quantitative reverse transcriptase-PCR, nuclear factor (NF)-κB activation as well as protein secretion by enzyme linked immunosorbent assay and Luminex technology was performed.RESULTS:LIGHT was found to inhibit lipid accumulation in the three models of preadipocytes in a dose-dependent manner without cytotoxic effects. This inhibition of differentiation was probably because of interference at early steps of adipogenesis, as early exposure during differentiation showed the strongest effect, as assessed by decreased peroxisome proliferator-activated receptor-γ (PPARγ) and CCAAT/enhancer-binding protein-α (C/EBPα) mRNA expression. In contrast to TNFα, basal and insulin-stimulated glucose uptake and lipolysis of terminally differentiated mature adipocytes were not altered in the presence of LIGHT. At a concentration sufficient to inhibit differentiation, secretion of proinflammatory cytokines was not significantly induced and NF-κB activity was only modestly induced compared with TNFα.CONCLUSION:LIGHT is a novel inhibitor of human adipocyte differentiation without adversely influencing central metabolic pathways in adipocytes.

We have used a combination of high throughput functional genomics, computerized database mining and expression analyses to discover novel human tumor suppressor genes (TSGs). A genome-wide high throughput cDNA phenotype screen was established to identify genes that induce apoptosis or reduce cell viability. TSGs are expressed in normal tissue and frequently act by reduction of growth of transformed cells or induce apoptosis. In agreement with that and thus serving as platform validation, our pro-apoptotic hits included genes for which tumor suppressing activities were known, such as kangai1 and CD81 antigen. Additional genes that so far have been claimed as putative TSGs or associated with tumor inhibitory activities (prostate differentiation factor, hRAS-like suppressor 3, DPH2L1-like and the metastasis inhibitor Kiss1) were confirmed in their proposed TSG-like phenotype by functionally defining their growth inhibitory or pro-apoptotic function towards cancer cells. Finally, novel genes were identified for which neither association with cell growth nor with apoptosis were previously described. A subset of these genes show characteristics of TSGs because they (i) reduce the growth or induce apoptosis in tumor cells; (ii) show reduced expression in tumor vs. normal tissue; and (iii) are located on chromosomal (LOH-) loci for which cancer-associated deletions are described. The pro-apoptotic phenotype and differential expression of these genes in normal and malignant tissue make them promising target candidates for the diagnosis and therapy of various tumors.

We describe a novel genetic screen that is performed by transfecting every individual clone of an expression clone collection into a separate population of cells in a high-throughput mode. We combined high-throughput functional genomics with experimental validation to discover human genes that ameliorate cytotoxic responses of neuronal HT-22 cells upon exposure to oxidative stress. A collection of 5,000 human cDNAs in mammalian expression vectors were individually transfected into HT-22 cells, which were then exposed to H(2)O(2). Five genes were found that are known to be involved in pathways of detoxification of peroxide (catalase, glutathione peroxidase-1, peroxiredoxin-1, peroxiredoxin-5, and nuclear factor erythroid-derived 2-like 2). The presence of those genes in our "hit list" validates our screening platform. In addition, a set of candidate genes was found that has not been previously described as involved in detoxification of peroxide. One of these genes, which was consistently found to reduce H(2)O(2) -induced toxicity in HT-22, was GFPT2. This gene is expressed at significant levels in the central nervous system (CNS) and encodes glutamine-fructose-6-phosphate transaminase (GFPT) 2, a rate-limiting enzyme in hexosamine biosynthesis. GFPT has recently also been shown to ameliorate the toxicity of methylmercury in Saccharomyces cerevisiae. Methylmercury causes neuronal cell death in part by protein modification as well as enhancing the production of reactive oxygen species (ROS). The protective effect of GFPT2 against H(2)O(2) toxicity in neuronal HT-22 cells may be similar to its protection against methylmercury in yeast. Thus, GFPT appears to be conserved among yeast and men as a critical target of methylmercury and ROS-induced cytotoxicity.