The effect of reducing agents, including N-acetylcysteine (NAC), dithiothreitol (DTT), and 2-mercaptoethanol (2-ME) on nuclear transcription factor-kappa B (NF-kappa B) activation and manganese superoxide dismutase (MnSOD) expression was investigated in a pulmonary adenocarcinoma (A549) cell line. NAC, DTT, and 2-ME each activated the transcription factor NF-kappa B and increased steady-state levels of MnSOD mRNA and enzyme activity in these cells. In addition, NAC, DTT, and 2-ME increased chloramphenicol acetyltransferase (CAT) activity in cells transfected with a construct containing the CAT gene under the control of the rat MnSOD promoter. SOD and catalase (500 U/ml) plus ethanol (1 mM) did not inhibit activation of NF-kappa B or elevation of steady-state MnSOD mRNA levels by NAC, DTT, or 2-ME. Controls in which comparable amounts of O2-. to those produced by thiols were generated by hypoxanthine and xanthine oxidase, or in which H2O2 was added directly, had neither activated NF-kappa B nor elevated MnSOD mRNA. This shows that reactive oxygen intermediates, which may be formed during autooxidation, may not contribute to activation of NF-kappa B. Because the MnSOD promoter also contains potential binding sites for other transcription factors, such as promoter-selective transcription factor-1 (SP-1), activator protein-1 (AP-1), AP-2, adenosine 3',5'-cyclic monophosphate-regulator element binding factor (CREB), and transcription factor IID complex (TFIID), the effect of thiols on their activation also were evaluated. In contrast to findings with NF-kappa B, there was only minor activation of AP-1 by thiols, and none of the other transcription factors were activated by thiols. AP-1 activation was inhibited by catalase (500 U/ml) plus SOD plus ethanol (1 mM). Addition of 700 microM H2O2 also activated AP-1, and catalase at 500 U/ml prevented this activation. This indicates that H2O2 produced as a result of autooxidation of thiols can activate AP-1 but not NF-kappa B. Thus a close association between exposure to reducing agents, activation of NF-kappa B, and elevation of MnSOD gene expression is demonstrated.

Manganese superoxide dismutase (MnSOD) is a mitochondrial enzyme that dismutates potentially toxic superoxide radical into hydrogen peroxide and dioxygen. This enzyme is critical for protection against cellular injury due to elevated partial pressures of oxygen. Thioredoxin (TRX) is a potent protein disulfide reductase found in most organisms that participates in many thiol-dependent cellular reductive processes and plays an important role in antioxidant defense, signal transduction, and regulation of cell growth and proliferation. Here we describe induction of manganese superoxide dismutase by thioredoxin. MnSOD mRNA and activity were increased dramatically by low concentrations of TRX (28 microM). Elevation of MnSOD mRNA by TRX was inhibited by actinomycin D, but not cycloheximide, occurring both in cell lines and primary human lung microvascular endothelial cells. mRNAs for other antioxidant enzymes including copper-zinc superoxide dismutase and catalase were not elevated, demonstrating specificity of induction of MnSOD by TRX. Thiol oxidation by diamide or alkylation by chlorodinitrobenzene inhibited MnSOD induction, further indicating a requirement for reduced TRX. Because both oxidized and reduced thioredoxin (28 microM) induced MnSOD mRNA, the intracellular redox status of externally added Escherichia coli oxidized TRX was determined. About 45% of internalized E. coli TRX was reduced, with 8% in fully reduced form and about 37% in partially reduced form. However, when TRX reductase and nicotinamide adenine dinucleotide (NADPH) were added to the extracellular medium with TRX, more than 80% of E. coli TRX was found to be in a fully reduced state in human adenocarcinoma (A549) cells. Although lower concentrations of oxidized TRX (7 microM) did not induce MnSOD mRNA, this concentration of TRX, when reduced by NADPH and TRX reductase, increased MnSOD mRNA six-fold. In additional studies, MCF-7 cells stably transfected with the human TRX gene had elevated expression of MnSOD mRNA relative to vector-transfected controls. Thus, both endogenously produced and exogenously added TRX elevate MnSOD gene expression. These findings suggest a novel mechanism involving reduced TRX in regulation of MnSOD.

TNF alpha and IL-1 each can activate NF-kappa B and induce gene expression of manganese superoxide dismutase (MnSOD), a mitochondrial matrix enzyme which can provide critical protection against hyperoxic lung injury. The regulation of MnSOD gene expression is not well understood. Since redox status can modulate NF-kappa B and potential kappa B site(s) exist in the MnSOD promoter, the effect of thiols (including NAC, DTT and 2-ME) on TNF alpha and IL-1 induced activation of NF-kappa B and MnSOD gene expression was investigated. Activation of NF-kB and increased MnSOD expression were potentiated by thiol reducing agents. In contrast, thiol oxidizing or alkylating agents inhibited both NF-kappa B activation and elevated MnSOD expression in response to TNF alpha or IL-1. Since protease inhibitors TPCK and TLCK can inhibit NF-kappa activation, we also investigated the effect of these compounds on MnSOD expression and NF-kappa B activation. TPCK and TLCK each inhibited MnSOD gene expression and NF-kappa B activation. Since the MnSOD promoter also contains an AP-1 binding site, the effect of thiols and thiol modifying agents on AP-1 activation was investigated. Thiols had no consistent effect on AP-1 activation. Likewise, some of the thiol modifying compounds inhibited AP-1 activation by TNF alpha or IL-1, whereas others did not. Since diverse agents had similar effects on activation of NF-kappa B and MnSOD gene expression, we have demonstrated that activation of NF-kappa B and MnSOD gene expression are closely associated and that reduced sulfhydryl groups are required for cytokine mediation of both processes.

Parenteral injection of the cytokines interleukin-1 and tumor necrosis factor, or of endotoxin (lipopolysaccharide), protects rats against lethal pulmonary oxygen toxicity. To determine the potential importance of manganese superoxide dismutase (MnSOD) in this model, we measured MnSOD mRNA and activity in lung. In addition, we confirmed that increases in activities were related to changes in MnSOD protein, which was measured using an enzyme-linked immunosorbentassay (ELISA) technique. After cytokine or endotoxin administration, increases in lung MnSOD mRNA occurred promptly (4 h), with or without hyperoxic exposure. In parallel, lung MnSOD protein and activity were increased after 24 h, and protein levels remained elevated after 52 h. MnSOD activity and protein levels were closely correlated. Neither lung copper-zinc superoxide dismutase (CuZnSOD) mRNA nor activity increased following administration of cytokines. Small increases in CuZnSOD mRNA, which did not exceed those in beta-actin mRNA, occurred early (4 h) after endotoxin, but CuZnSOD activity was unchanged. Immunohistochemistry was used to demonstrate in which cell types the increase in MnSOD protein occurred after cytokine or endotoxin administration. In agreement with ELISA findings, immunoreactive MnSOD appeared to be increased in lung parenchyma, but not in lung neutrophils, 24 h after cytokine or endotoxin treatment. MnSOD was heavily concentrated in alveolar type II cells. However, the numbers of surfactant protein D-positive (type II) cells in lung sections did not appear to be increased after treatment with cytokines or endotoxin. We conclude that early and sustained increases in endogenous MnSOD, but not CuZnSOD or other antioxidant enzymes, are associated with protection of rat lungs against hyperoxic damage by cytokines or endotoxin.