Effects of antioxidant enzyme overexpression on the invasive phenotype of hamster cheek pouch carcinoma cells.

To examine the role of reactive oxygen species on the invasive phenotype of cancer cells, we overexpressed manganese- and copper-zinc-containing superoxide dismutases (MnSOD, CuZnSOD) and catalase (Cat) in hamster cheek pouch carcinoma (HCPC-1) cells in vitro using adenoviral vector-mediated gene transfer. Hamster cheek pouch carcinoma cells were transduced with these adenoviral vector constructs alone, or in combination, at concentrations [i.e., multiplicity of infectivity (MOI)] of 100 MOI each. The Escherichia coli beta-galactosidase reporter construct was used as a control virus. Protein expression was examined by Western blot analysis and enzymatic activities were measured using spectrophotometry. To observe the effects of transgene overexpression on in vitro tumor cell invasion, we used the membrane invasion culture system, an accurate and reliable method for examining tumor cell invasion, in vitro. This assay measures the ability of tumor cells to invade a basement membrane matrix consisting of type IV collagen, laminin, and gelatin. MnSOD overexpression resulted in a 50% increase in HCPC-1 cell invasiveness (p < .001); co-overexpression of MnSOD with Cat partially inhibited this effect (p < .05). Moreover, co-overexpression of both SODs resulted in a significant increase in invasiveness compared with the parental HCPC-1 cells (p < .05). These changes could not be correlated with the 72 kDa collagenase IV or stromolysin activities using zymography, or the downregulation of the adhesion molecules E-cadherin or the alpha4 subunit of the alpha4beta1 integrin. These results suggest that hydrogen peroxide may play a role in the process of tumor cell invasion, but that the process does not rely on changes in matrix metalloproteinase activity in the cells, or the expression of cell adhesion molecules.     

Expression of superoxide dismutases, catalase, and glutathione peroxidase in glioma cells

Four primary antioxidant enzymes were measured in both human and rat glioma cells. Both manganese-containing superoxide dismutase (MnSOD) and copper-zinc-containing superoxide dismutase (CuZnSOD) activities varied greatly among the different glioma cell lines. MnSOD was generally higher in human glioma cells than in rat glioma cells and relatively higher than in other tumor types. High levels of MnSOD in human glioma cells were due to the high levels of expression of MnSOD mRNA and protein. Heterogeneous expression of MnSOD was present in individual glioma cell lines and may be due to subpopulations or cells at different differentiation stages. Less difference in CuZnSOD, catalase, or glutathione peroxide was found between human and rat glioma cells. The human glioma cell lines showed large differences in sensitivity to the glutathione modulating drugs 1,3-bis (2-chloroethyl)-1-nitrosourea (BCNU) and buthionine sulfoximine (BSO). A good correlation was found between sensitivity to BCNU and the activities of catalase in these cell lines. Only one cell line was sensitive to BSO and this line had low CuZnSOD activity.     

Esteblishment and characterization of the line of Fagopyrum tataricum morphogenic callus tolerant to aminotriazole

It is shown that the inhibitor of catalase 3-amino-1,2,4-triazole (AT) at the concentration of 2 mM affected differently growth of tartary buckwheat (Fagopyrum tataricum (L.) Gaertn.) callus lines differing in the morphogenecity. In some cases, AT induced the death of a great fraction of non-morphogenic callus cells; in other cases, it inhibited growth and reduced viability of morphogenic callus. The death of non-morphogenic callus cells may be related to the accumulation of hydrogen peroxide and the development of oxidative stress. After morphogenic callus treatment with AT, we obtained a modified line 1?C8 AT tolerant to AT and differing from the original line in morphology, cell sizes, proliferative activity, and some biochemical characteristics. In the 1?C8 AT line, catalase was sensitive to this inhibitor action. In this case, catalase inactivation with AT did not increase the content of hydrogen peroxide in the cells, which may indicate the compensatory functioning of another/others mechanism(s) destroying hydrogen peroxide.     

Overexpression of manganese or copper-zinc superoxide dismutase inhibits breast cancer growth

We have studied the effects of overexpression of superoxide dismutase (SOD), a tumor suppressor protein that dismutes superoxide radical to H2O2, on breast cancer cell growth in vitro and xenograft growth in vivo. No previous work has directly compared the growth-suppressive effects of manganese SOD (MnSOD) and copper-zinc SOD (CuZnSOD). We hypothesized that either adenoviral MnSOD (AdMnSOD) or adenoviral CuZnSOD (AdCuZnSOD) gene therapy would suppress the growth of human breast cancer cells. After determining the antioxidant profiles of three human breast cell lines, MCF 10A, MDA-MB231, and MCF-7, we measured the effects of MnSOD or CuZnSOD overexpression on cell growth and survival in vitro and in vivo. Results demonstrated that infection with AdMnSOD or AdCuZnSOD increased the activity of the respective enzyme in all three cell lines. In vitro, overexpression of MnSOD or CuZnSOD decreased not only cell growth but also clonogenic survival in a dose- and transgene-dependent manner. In vivo, treatment of tumors with AdMnSOD or AdCuZnSOD decreased xenograft growth compared to controls. The first direct comparison of MnSOD to CuZnSOD overexpression indicated that CuZnSOD and MnSOD were similarly effective at suppressing cancer cell growth.     

Induction of superoxide dismutase and cytotoxicity by manganese in human breast cancer cells.

MnCl2 induced manganese-containing superoxide dismutase (MnSOD) expression (mRNA, immunoreactive protein, and enzyme activity) in human breast cancer Hs578T cells. The induction of MnSOD immunoreactive protein in Hs578T cells was inhibited by tiron (a metal chelator and superoxide scavenger), pyruvate (a hydrogen peroxide scavenger), or 2-deoxy-d-glucose (DG, an inhibitor of glycolysis and the hexose monophosphate shunt), but not by 5,5-dimethyl-1-pyrroline-1-oxide (a superoxide scavenger), N-acetyl cysteine (a scavenger for reactive oxygen species and precursor of glutathione), diphenylene iodonium (an inhibitor of flavoproteins such as NADPH oxidase and nitric oxide synthase), or SOD (a superoxide scavenger). Northern blotting demonstrated that tiron or DG affected at the mRNA level, while pyruvate affected Mn-induced MnSOD expression at both the mRNA and protein levels. These results demonstrate that Mn can induce MnSOD expression in cultured human breast cancer cells. Mn also induced apoptosis and necrosis in these cells. Since inhibitors of Mn-induced MnSOD induction did not affect cell viability, MnSOD induction is probably not the cause of the Mn-induced cell killing.     

Oxygen free radicals mediate the induction of manganese superoxide dismutase gene expression by TNF-alpha

In this study, the hypothesis that oxygen free radicals act as intracellular messengers is examined. Treatment of human oral carcinoma SCC-25 cells with 200 ng/ml human TNF-alpha for 6 h greatly increased manganese superoxide dismutase (MnSOD) gene expression as detected by western blotting, RT-PCR, and nuclear run-on experiments. In the presence of the oxygen free radical spin trapping reagent, 5,5-dimethyl pyrroline-N-oxide (DMPO), the induction of MnSOD gene expression by TNF-alpha was significantly reduced. Electron paramagnetic resonance experiments showed that the production of oxygen free radicals was enhanced in TNF-alpha treated cells. Taken together, these observations suggest that the induction of MnSOD expression by TNF-alpha is at least partially mediated by intracellular formation of oxygen free radicals, and that superoxide is most likely the initiating species involved in the mediation of MnSOD gene expression by TNF-alpha.     

Superoxide Mediates Direct Current Electric Field-Induced Directional Migration of Glioma Cells through the Activation of AKT and ERK

Direct current electric fields (DCEFs) can induce directional migration for many cell types through activation of intracellular signaling pathways. However, the mechanisms that bridge extracellular electrical stimulation with intracellular signaling remain largely unknown. In the current study, we found that a DCEF can induce the directional migration of U87, C6 and U251 glioma cells to the cathode and stimulate the production of hydrogen peroxide and superoxide. Subsequent studies demonstrated that the electrotaxis of glioma cells were abolished by the superoxide inhibitor N-acetyl-l-cysteine (NAC) or overexpression of mitochondrial superoxide dismutase (MnSOD), but was not affected by inhibition of hydrogen peroxide through the overexpression of catalase. Furthermore, we found that the presence of NAC, as well as the overexpression of MnSOD, could almost completely abolish the activation of Akt, extracellular-signal-regulated kinase (Erk)1/2, c-Jun N-terminal kinase (JNK), and p38, although only JNK and p38 were affected by overexpression of catalase. The presenting of specific inhibitors can decrease the activation of Erk1/2 or Akt as well as the directional migration of glioma cells. Collectively, our data demonstrate that superoxide may play a critical role in DCEF-induced directional migration of glioma cells through the regulation of Akt and Erk1/2 activation. This study provides novel evidence that the superoxide is at least one of the “bridges” coupling the extracellular electric stimulation to the intracellular signals during DCEF-mediated cell directional migration.     

Susceptibility of Human Head and Neck Cancer Cells to Combined Inhibition of Glutathione and Thioredoxin Metabolism

Increased glutathione (GSH) and thioredoxin (Trx) metabolism are mechanisms that are widely implicated in resistance of cancer cells to chemotherapy. The current study determined if simultaneous inhibition of GSH and Trx metabolism enhanced cell killing of human head and neck squamous cell carcinoma (HNSCC) cells by a mechanism involving oxidative stress. Inhibition of GSH and Trx metabolism with buthionine sulfoximine (BSO) and auranofin (AUR), respectively, induced significant decreases in clonogenic survival compared to either drug alone in FaDu, Cal-27 and SCC-25 HNSCC cells in vitro and in vivo in Cal-27 xenografts. BSO+AUR significantly increased glutathione and thioredoxin oxidation and suppressed peroxiredoxin activity in vitro. Pre-treatment with N-acetylcysteine completely reversed BSO+AUR-induced cell killing in FaDu and Cal-27 cells, while catalase and selenium supplementation only inhibited BSO+AUR-induced cell killing in FaDu cells. BSO+AUR decreased caspase 3/7 activity in HNSCC cells and significantly reduced the viability of both Bax/Bak double knockout (DKO) and DKO-Bax reconstituted hematopoietic cells suggesting that necrosis was involved. BSO+AUR also significantly sensitized FaDu, Cal-27, SCC-25 and SQ20B cells to cell killing induced by the EGFR inhibitor Erlotinib in vitro. These results support the conclusion that simultaneous inhibition of GSH and Trx metabolism pathways induces oxidative stress and clonogenic killing in HNSCCs and this strategy may be useful in sensitizing HNSCCs to EGFR inhibitors.     

Increased oxidative stress created by adenoviral MnSOD or CuZnSOD plus BCNU (1,3-bis(2-chloroethyl)-1-nitrosourea) inhibits breast cancer cell growth

Superoxide dismutases (SODs) have been found to decrease tumor formation and angiogenesis. SOD gene therapy, as with many other gene transfer strategies, may not completely inhibit tumor growth on its own. Thus, concomitant therapies are necessary to completely control the spread of this disease. We hypothesized that intratumoral injection of AdSOD in combination with 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) chemotherapy would synergistically inhibit breast cancer growth. Our data indicate that BCNU when combined with SOD overexpression increased oxidative stress as suggested by elevated glutathione disulfide (GSSG) production in one of three breast cancer cell lines tested, at least in part due to glutathione reductase (GR) inactivation. The increased oxidative stress caused by BCNU combined with adenovirally expressed SODs, manganese or copper zinc SOD, decreased growth and survival in the three cell lines tested in vitro, but had the largest effect in the MDA-MB231 cell line, which showed the largest amount of oxidative stress. Delivery of MnSOD and BCNU intratumorally completely inhibited MDA-MB231 xenograft growth and increased nude mouse survival in vivo. Intravenous (iv) BCNU, recapitulating clinical usage, and intratumoral AdMnSOD delivery, to provide tumor specificity, provided similar decreased growth and survival in our nude mouse model. This cancer therapy produced impressive results, suggesting the potential use of oxidative stress-induced growth inhibitory treatments for breast cancer patients.     

Squamous cell carcinoma cells differentially stimulate NK cell effector functions: the role of IL-18

Tumor cells stimulate natural killer (NK) cell effector functions, but the regulation of cytokine secretion and cytolysis is incompletely understood. We tested whether oral and pharyngeal squamous cell carcinoma cell lines differentially stimulated NK cell interferon-gamma (IFN-gamma) secretion and cytolysis using a clone of the NK-92-transformed human NK cell line, NK92.35. SCC-4 and SCC-25 cells, but not FaDu or Cal 27 cells, stimulated robust NK92.35 IFN-gamma secretion. All four carcinoma cell lines were lysed by NK92.35 cells. These findings indicate that carcinoma cells differentially stimulate NK cell IFN-gamma secretion and cytolysis. In Transwell experiments, a combination of SCC-4 or SCC-25 cell soluble factors and contact with FaDu cells synergistically stimulated NK92.35 cell IFN-gamma secretion. Stimulatory SCC-4 cells constitutively secreted IL-18, a cytokine that potently augments IFN-gamma secretion by T cells and NK cells. In contrast, poorly stimulatory FaDu cells produced little or no IL-18, but synergized with recombinant IL-18 to stimulate NK92.35 IFN-gamma secretion. mAb to IL-18 or IL-18 receptor diminished SCC-4-stimulated IFN-gamma secretion by NK92.35 cells and by nontransformed NK cells. Thus, IL-18 was necessary for optimal carcinoma stimulation of NK cell IFN-gamma secretion. In vivo, oral and upper aerodigestive tract epithelia and carcinomas produced IL-18, but one squamous cell carcinoma had heterogeneous IL-18 expression. Thus IL-18 production can account for squamous cell carcinoma differential stimulation of NK cell effector functions in vitro and may be important for stimulation of NK cells in vivo.     

UV-irradiation induces oxidative damage to mitochondrial DNA primarily through hydrogen peroxide: analysis of 8-oxodGuo by HPLC

Roles of reactive oxygen species (ROS) in damage to mitochondrial DNA (mtDNA) following ultraviolet (UV)-irradiation were investigated in the human hepatoma cell line SK-HEP-1. We altered the intracellular status of ROS by the overexpression of manganese superoxide dismutase (MnSOD) and/or catalase. Using HPLC, we analyzed 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodGuo), known as a marker of damage to DNA molecules. UV-irradiation resulted in the accumulation of 8-oxodGuo in these cells. The overexpression of MnSOD enhanced the accumulation of 8-oxodGuo by UV. The co-overexpression of catalase inhibited the accumulation of 8-oxodGuo by UV in MnSOD-transfectants. The overexpression of MnSOD reduced the colony forming capacity in SK-HEP-1 cells and the co-overexpression of catalase with MnSOD stimulated the capacity compared to control. UV-irradiation inhibited the colony forming capacity in these cells; no difference was observed among the capacities of control, MnSOD- and catalase-transfectants. However, the overexpression of MnSOD/catalase significantly rescued the reduction of colony forming capacity by UV-irradiation. Our results suggest that the accumulation of hydrogen peroxide plays a key role in the oxidative damage to mtDNA of UV-irradiated cells, and also that the overexpression of both MnSOD and catalase reduces the mtDNA damage and blocks the growth inhibition by UV. Our results also indicate that the increased activity of MnSOD may lead to a toxic effect on mtDNA by UV-irradiation.     

UV-irradiation induces oxidative damage to mitochondrial DNA primarily through hydrogen peroxide: Analysis of 8-oxodGuo by HPLC

Roles of reactive oxygen species (ROS) in damage to mitochondrial DNA (mtDNA) following ultraviolet (UV)-irradiation were investigated in the human hepatoma cell line SK-HEP-1. We altered the intracellular status of ROS by the overexpression of manganese superoxide dismutase (MnSOD) and/or catalase. Using HPLC, we analyzed 8-oxo-7,8-dihydro-2′-deoxyguanosine (8-oxodGuo), known as a marker of damage to DNA molecules. UV-irradiation resulted in the accumulation of 8-oxodGuo in these cells. The overexpression of MnSOD enhanced the accumulation of 8-oxodGuo by UV. The co-overexpression of catalase inhibited the accumulation of 8-oxodGuo by UV in MnSOD-transfectants. The overexpression of MnSOD reduced the colony forming capacity in SK-HEP-1 cells and the co-overexpression of catalase with MnSOD stimulated the capacity compared to control. UV-irradiation inhibited the colony forming capacity in these cells; no difference was observed among the capacities of control, MnSOD- and catalase-transfectants. However, the overexpression of MnSOD/catalase significantly rescued the reduction of colony forming capacity by UV-irradiation. Our results suggest that the accumulation of hydrogen peroxide plays a key role in the oxidative damage to mtDNA of UV-irradiated cells, and also that the overexpression of both MnSOD and catalase reduces the mtDNA damage and blocks the growth inhibition by UV. Our results also indicate that the increased activity of MnSOD may lead to a toxic effect on mtDNA by UV-irradiation.     

Selenium supplementation sensitizes renca cells to tert-butylhydroperoxide induced loss of viability

Mouse renal carcinoma (renca) cells growing exponentially in foetal bovine serum (1%) supplemented with selenium (1 microM, sodium selenite) were exposed to oxidative insult. It was found that glutathione peroxidase activity increased (44%), while the activities of catalase, glutathione disulfide reductase, and level of total glutathione did not change due to selenium supplementation. Selenium supplementation made renca cells susceptible to tert-butylhydroperoxide induced cell death, while it did not affect the viability when the cells were exposed to hydrogen peroxide. It suggested that the contribution of glutathione peroxidase in antioxidant defense mechanism of renca cells was possibly not crucial and the function of catalase might be important especially against hydrogen peroxide.     

Endogenous defenses against the cytotoxicity of hydrogen peroxide in cultured rat hepatocytes

The catalase activity of cultured rat hepatocytes was inhibited by 90% pretreatment with 20 mM aminotriazole without effect on the activities of glutathione peroxidase or glutathione reductase, or on the viability of the cells over the subsequent 24 h. Glutathione reductase was inhibited by 85% by pretreatment with 300 microM 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU) without effect on glutathione peroxidase, catalase, or on viability. Both pretreatments sensitized the hepatocytes to the cytotoxicity of H2O2 generated either by glucose oxidase (0.05-0.5 units/ml) or by the autoxidation of the one-electron-reduced state of menadione (50-250 microM). Aminotriazole pretreatment had no effect on the GSH content of the hepatocytes. BCNU reduced GSH levels by 50%. Depletion of GSH levels to less than 20% of control by treatment with diethyl maleate, however, did not sensitize the cells to either glucose oxidase or menadione, indicating that the effect of BCNU is related to inhibition of the GSH-GSSG redox cycle rather than to the depletion of GSH. With glucose oxidase, most of the cell killing in hepatocytes pretreated with either aminotriazole or BCNU occurred between 1 and 3 h. The antioxidant diphenylphenylenediamine (DPPD) had no effect on viability at 3 h. Catalase added to the culture medium 1 h after the addition of glucose oxidase prevented the cell killing measured at 3 h. The sulfhydryl reagents dithiothreitol (200 microM), N-acetyl-L-cysteine (4 mM), and alpha-mercaptopropionyl-L-glycine (2.5 mM) prevented the cell killing with exogenous H2O2 in hepatocytes sensitized by the inhibition of catalase or glutathione reductase. With menadione, there was no killing of nonpretreated hepatocytes at 1 h, and DPPD did not prevent the cell death after 3 h. Aminotriazole pretreatment enhanced the cell killing at 3 h but not at 1 h, and DPPD was not protective. Catalase added to the medium at 1 h inhibited the cell death measured at 3 h. In contrast, menadione killed hepatocytes pretreated with BCNU within 1 h. DPPD prevented cell death at 1 h, and there was evidence of lipid peroxidation in the accumulation of malondialdehyde in the culture medium. Catalase added with menadione did not prevent the cell killing at 1 h but did prevent it at 3 h. These data indicate that catalase and the GSH-GSSG cycle are active in the defense of hepatocytes against the toxicity of H2O2.(ABSTRACT TRUNCATED AT 400 WORDS)     

A study of the relative importance of the peroxiredoxin-, catalase-, and glutathione-dependent systems in neural peroxide metabolism

Cells are endowed with several overlapping peroxide-degrading systems whose relative importance is a matter of debate. In this study, three different sources of neural cells (rat hippocampal slices, rat C6 glioma cells, and mouse N2a neuroblastoma cells) were used as models to understand the relative contributions of individual peroxide-degrading systems. After a pretreatment (30 min) with specific inhibitors, each system was challenged with either H₂O₂ or cumene hydroperoxide (CuOOH), both at 100 μM. Hippocampal slices, C6 cells, and N2a cells showed a decrease in the H₂O₂ decomposition rate (23-28%) by a pretreatment with the catalase inhibitor aminotriazole. The inhibition of glutathione reductase (GR) by BCNU (1,3-bis(2-chloroethyl)-1-nitrosourea) significantly decreased H₂O₂ and CuOOH decomposition rates (31-77%). Inhibition of catalase was not as effective as BCNU at decreasing cell viability (MTT assay) and cell permeability or at increasing DNA damage (comet test). Impairing the thioredoxin (Trx)-dependent peroxiredoxin (Prx) recycling by thioredoxin reductase (TrxR) inhibition with auranofin neither potentiated peroxide toxicity nor decreased the peroxide-decomposition rate. The results indicate that neural peroxidatic systems depending on Trx/TrxR for recycling are not as important as those depending on GSH/GR. Dimer formation, which leads to Prx2 inactivation, was observed in hippocampal slices and N2a cells treated with H₂O₂, but not in C6 cells. However, Prx-SO₃ formation, another form of Prx inactivation, was observed in all neural cell types tested, indicating that redox-mediated signaling pathways can be modulated in neural cells. These differences in Prx2 dimerization suggest specific redox regulation mechanisms in glia-derived (C6) compared to neuron-derived (N2a) cells and hippocampal slices.