Cell-specific expression of manganese superoxide dismutase protein in the lungs of patients with respiratory distress syndrome, chronic lung disease, or persistent pulmonary hypertension.

The developmental profile of manganese superoxide dismutase (MnSOD) and its regulation in hyperoxia vary between species. We hypothesized that MnSOD increases in human lung in response to oxygen treatment, although this response could be restricted to certain cell types and depend on gestational age. Therefore, the cell-specific expression of pulmonary immunoreactive MnSOD protein was investigated during development, and in patients with respiratory distress syndrome (RDS), chronic lung disease (CLD), or persistent pulmonary hypertension (PPHN). Throughout ontogenesis, all cell types expressed MnSOD, but the most intense positivity was found in bronchiolar epithelium and (pre-) type-II pneumocytes. MnSOD protein did not increase during development. The MnSOD staining pattern in arterial endothelium was more intense in RDS patients than in age-matched controls, but this may be related to induction of MnSOD by increased blood flow rather than by oxygen. MnSOD expression in other cell types of RDS, CLD, or PPHN patients did not differ from that in age-matched controls. We conclude that, in terms of mitochondrial enzymatic superoxide scavenging capacity, preterm infants are not more vulnerable than term infants to oxygen-induced lung injury at physiological oxygen concentrations. However, the inability to induce MnSOD in response to oxygen treatment may result in a poor outcome.

Nuclear factor Y activates the human xanthine oxidoreductase gene promoter.

To study the regulation of the human xanthine oxidoreductase (XOR) gene, we cloned 1.9 kb of the promoter region. In reporter gene assays, a construct encompassing nucleotides between 142 to +42 conferred maximal basal activity of the XOR promoter in 293T cells, in comparison with shorter (-92 to +42) or longer (up to -1937 to +42) constructs. The promoter activity was low in NIH-3T3 cells. The most active construct contained a putative CCAAT motif at -119 to -123. Electrophoretic mobility shift assays showed that this sequence binds the ubiquitous nuclear factor Y (NF-Y). Mutation of the CCAAT motif (CTGAT) abolished the NF-Y binding and considerably reduced the promoter activity. Our data suggest an important functional role for NF-Y in the transcriptional activation of the human XOR gene.

Antioxidant defense mechanisms of human mesothelioma and lung adenocarcinoma cells.

The development of drug resistance of tumors is multifactorial and still poorly understood. Some cytotoxic drugs generate free radicals, and, therefore, antioxidant enzymes may contribute to drug resistance. We investigated the levels of manganese superoxide dismutase (Mn SOD), its inducibility, and its protective role against tumor necrosis factor-alpha and cytotoxic drugs (cisplatin, epirubicin, methotrexate, and vindesin) in human pleural mesothelioma (M14K) and pulmonary adenocarcinoma (A549) cells. We also studied other major antioxidant mechanisms in relation to oxidant and drug resistance of these cells. A549 cells were more resistant than M14K cells toward both oxidants (hydrogen peroxide and menadione) and all the cytotoxic drugs tested. M14K cells contained higher basal Mn SOD activity than A549 cells (28.3 +/- 3.4 vs. 1.8 +/- 0.3 U/mg protein), and Mn SOD activity was significantly induced by tumor necrosis factor-alpha only in A549 cells (+524%), but the induction did not offer any protection during subsequent oxidant or drug exposure. Mn SOD was not induced significantly in either of these cell lines by any of the cytotoxic drugs (0.007-2 microM, 48 h) tested when assessed by Northern blotting, Western blotting, or specific activity. A549 cells contained higher catalase activity than M14K cells (7.6 +/- 1.3 vs. 3.6 +/- 0.5 nmol O(2). min(-1). mg protein(-1)). They also contained twofold higher levels of glutathione and higher immunoreactivity of the heavy subunit of gamma-glutamylcysteine synthetase than M14K cells. Experiments with inhibitors of gamma-glutamylcysteine synthetase and catalase supported our conclusion that mechanisms associated with glutathione contribute to the drug resistance of these cells.

Human cystathionine gamma-lyase: developmental and in vitro expression of two isoforms.

Cystathionine gamma-lyase (CGL) is the last enzyme of the trans-sulphuration pathway, which converts methionine into cysteine. To study the possible differences in enzymic activity of the two human cystathionine gamma-lyase isoforms characterized earlier, these were separately expressed in human kidney embryonic 293T cells. Furthermore, developmental changes in the expression of the two mRNA forms as well as the enzymic activity in human liver were studied, as it has been postulated that a change in the relative expression of CGL isoforms causes the postnatal increase in CGL activity. Transfection with the longer isoform increased the CGL activity 1.5-fold, while the activity of the cells transfected with the shorter form did not differ from the basal activity. In human liver samples, CGL activity was only detected in adult tissue (68+/-9 nmol of cysteine/h per mg of protein), whereas activity in fetal, premature and full-term neonatal liver tissue was undetectable. In contrast, strong mRNA expression of both mRNA isoforms was detected from the 19th gestational week onwards and the longer form of CGL appeared to be predominant. The expression of the two mRNA forms varied in parallel. In conclusion, we have shown that only cells overexpressing the longer form of CGL have increased activity, and CGL appears to be regulated at the post-transcriptional level during development.

Expression of gamma-glutamylcysteine synthetase during development.

Prematurity has been associated with low glutathione (GSH) concentrations in bronchoalveolar lavage fluid as well as in leukocytes from tracheal aspirates and peripheral blood. To elucidate whether this is caused by deficient GSH synthesis, the expression and activity of gamma-glutamylcysteine synthetase (glutamate-cysteine ligase, GCS, EC 6.3.2.2), the rate-limiting enzyme for GSH synthesis, were measured from fetal, neonatal, and adult human liver, lung, and kidney samples. The highest activity was measured in the liver, in which mRNA expression of the catalytic GCS heavy and the regulatory light subunits, as well as activity, were, on average, similar in the various stages of development. Although GCS light subunit mRNA concentrations in the lung were higher in neonates than in fetuses and adults, enzyme activities were similar. In the adult kidney, mean enzyme activity was somewhat higher than in fetal or neonatal kidney, but this may be accounted for by the variation in the small number of samples. In conclusion, GCS is expressed and active already in the second trimester and thus low GSH concentrations found in preterm neonates appear not to be explained by deficient GSH synthesis. Other factors, such as limited availability of the GSH precursor cysteine or increased GSH consumption, may account for the lower concentrations of GSH found in preterm infants.

Expression of antioxidant enzymes in human inflammatory cells.

Because antioxidant enzymes may have an important role in the oxidant resistance of inflammatory cells, we investigated the mRNA levels and specific activities of manganese and copper-zinc superoxide dismutases (Mn SOD and Cu,Zn SOD), catalase (Cat), and glutathione peroxidase, as well as the concentrations of glutathione (GSH) in human neutrophils, monocytes, monocyte-derived macrophages, and alveolar macrophages. Levels of GSH and glutathione peroxidase activity in monocytes were three times higher than in neutrophils, whereas the mRNA of Cat was 50-fold and its specific activity 4-fold higher in neutrophils. Although Mn SOD mRNA levels were higher in neutrophils, enzyme activities, as well as those of Cu,Zn SOD, were similar in all phagocytic cells. Neutrophils lost their viability, assessed by adenine nucleotide depletion, within 24 h ex vivo and more rapidly if GSH was depleted. However, neutrophils were the most resistant cell type to exogenous H(2)O(2). In conclusion, high Cat activity of neutrophils appears to explain their high resistance against exogenous H(2)O(2), whereas low GSH content and GSH-related enzymes seem to account for the poor survival of human neutrophils.

Correlation of oxidant-induced acute ATP depletion with delayed cell death in human neuroblastoma cells.

We correlated the adenine nucleotide (AN) levels and energy charge (EC) at the end of a transient oxidative exposure to the delayed death of neuronal cells. When wild-type (WT) or Bcl-2-overexpressing (BCL-2) human neuroblastoma cells (Paju) were exposed to 250 microM H(2)O(2) for 60 min, the EC of WT cells was unchanged, but that of BCL-2 cells decreased from 0.91 +/- 0.03 to 0.67 +/- 0.02. Depletion of ANs was significantly greater in BCL-2 (66.7 +/- 2%) than in WT (38.8 +/- 2%) cells. Proliferation of both lines decreased, averaging 63 +/- 17% of control by 48 h. Exposure to 5 mM H(2)O(2) caused no further change in ANs in BCL-2 cells but in WT cells decreased the EC to 0.45 +/- 0.08 and depleted ANs to 41 +/- 9% of control; after 24 h, WT cells became pyknotic and showed DNA fragmentation but no chromatin condensation, whereas BCL-2 cells died by delayed necrosis. After 10 mM H(2)O(2), EC dropped to 0.15 +/- 0.1, and both lines were acutely killed. The EC after an oxidative insult correlated well with further growth of both cell lines. A significant decline in EC led to delayed death. Bcl-2 did not protect against the fall in EC or AN depletion, but, although survival was not improved, the mechanism of death appeared to be different.

Cellular expression of xanthine oxidoreductase protein in normal human tissues.

Xanthine oxidoreductase is an important cytoplasmic source of reactive oxygen species, and has been implicated in the pathogenesis of ischemia-reperfusion damage. Because the cellular localization of this protein remains unclear, our aim was to study its distribution in fresh normal human tissue obtained at surgery. For immunohistochemical studies we purified the protein from human milk and raised a polyclonal antibody in rabbits. In the liver the protein was preferentially localized to the periportal hepatocytes and it was absent from the perivenous region. In the proximal intestine, the protein was expressed in epithelial cells and goblet cells. Lactating mammary gland acinar cells showed intense staining. Small vessel vascular endothelial cells of the intestine, mammary gland, and skeletal muscle showed immunoreactivity, but in the kidney, glomerular endothelial cells were negative. No cells in the heart, brain, or lung expressed the enzyme protein. The observed localization of the xanthine oxidoreductase protein is consistent with previously observed enzyme activities in the organs studied. The widely assumed exclusive localization to capillary endothelium obviously does not apply to humans.

Delayed impairment of cerebral oxygenation after deep hypothermic circulatory arrest in children.

BACKGROUND: Clinical studies of deep hypothermic circulatory arrest (DHCA) have focused only on the immediate postoperative period. However, experimental findings suggest impairment of cerebral oxygenation at 2 to 8 hours after reperfusion. METHODS: In 10 children who had DHCA for heart operations, transcerebral differences of hemoglobin oxygen saturation and plasma hypoxanthine, xanthine, and lactoferrin concentrations were measured in concurrently obtained cerebral venous, arterial, and mixed venous samples up to 10 hours postoperatively. RESULTS: Compared with preoperative levels (57% +/- 7%), cerebral venous oxygen saturation was not significantly reduced until 2 hours (44% +/- 6%) and 6 hours (42% +/- 5%) after DHCA (p < 0.05). A statistically significant transcerebral (ie, cerebral vein versus artery) concentration difference of hypoxanthine was observed at 30 minutes (3.6 +/- 0.9 micromol/L), 1 hour (3.4 +/- 1.1 micromol/L), and 2 hours (3.1 +/- 0.8 micromol/L) after DHCA but not preoperatively (0.4 +/- 0.2 micromol/L). A transcerebral concentration difference of lactoferrin occurred 30 minutes after DHCA (196 +/- 70 microg/mL) but not preoperatively (16 +/- 20 microg/mL). CONCLUSIONS: Cerebral venous oxygen saturation of hemoglobin decreased as late as 2 to 6 hours after DHCA, in association with impaired cerebral energy status. Neutrophil activation in the cerebral circulation occurred 30 minutes after reperfusion.

Irreversible conversion of xanthine dehydrogenase into xanthine oxidase by a mitochondrial protease.

Irreversible conversion of xanthine dehydrogenase (XDH) to its oxygen free radical producing oxidase (XO) form occurs through an uncharacterized proteolytic process, which was studied in human liver. Upon incubation of fresh unfrozen liver cytosol, XDH remained intact. When recombinant human XDH was coincubated with subcellular fractions of human liver, the mitochondrial intermembrane space was shown to contain a heat-labile activity that converted XDH irreversibly to XO. This activity is resistant to inhibitors of all major groups of proteases. We postulate that this novel type of proteolytic enzyme is released into the cytosol upon mitochondrial damage.

Oral iron is sufficient for erythropoietin treatment of very low birth-weight infants.

UNLABELLED: The aim of this study was to compare two different doses and means of administration of iron in recombinant human erythropoietin (rHuEPO)-treated very low birth-weight (VLBW) infants. VLBW infants (n = 41) were randomized to one of three groups. Fourteen infants were treated with rHuEPO (300 IU/kg three times a week s.c.) and oral iron (ferrofumarate, 6 mg of iron/kg per day). Another 14 infants received the same erythropoietin dose and intramuscular iron (ferroxypolymaltose, once 12 mg of iron/kg weekly). Thirteen infants were treated with the same dose of intramuscular iron but did not receive rHuEPO. After the 3-week study period, haemoglobin concentrations and reticulocyte counts were similar in the rHuEPO-treated groups and both were higher than in the group not receiving rHuEPO (P < 0.001). In both rHuEPO-treated groups the transferrin receptor concentration increased from 6.8-7.2 mg/l to 10.5-11.3 mg/l. CONCLUSION: In erythropoietin-treated very low birth weight infants the iron need for erythropoiesis can be met by oral administration of iron.

Pharmacokinetics of intravenous N-acetylcysteine in pre-term new-born infants.

BACKGROUND: Reactive oxygen species have been considered to play a role in several clinical complications in pre-term infants. The aim of this study was to determine the pharmacokinetics of intravenous N-acetylcysteine in pre-term neonates. This information is needed to evaluate the use of N-acetylcysteine as an antioxidant in this patient group. METHODS: N-acetylcysteine was infused intravenously in ten patients (gestational age 24.9-31.0 weeks, weight 500-1384 g) for 24 h (3.4-4.6 mg/kg/h), starting 2.0-11.2 h from birth (study I) and in six patients (gestational age 25.9-29.7 weeks, weight 520-1335 g) for 6 days (0.3-1.3 mg/kg/h), starting at the age of 24 h (study II). Arterial plasma N-acetylcysteine and cyst(e)ine concentrations were determined from timed samples taken during (study I and II) and after (study I) the N-acetylcysteine infusion. RESULTS: In study I, the mean elimination half-life of N-acetylcysteine was 11 h (range 7.8-15.2 h). The mean plasma clearance of N-acetylcysteine was 37 ml/kg/h (range 13-62 ml/kg/h) and the mean volume of distribution was 573 ml/kg (range 167-1010 ml/kg). The plasma clearance and volume of distribution correlated with weight (r = 0.81, P < 0.01, and r = 0.78, P < 0.01, respectively) and with gestational age (r = 0.71, P < 0.05, and r = 0.64, P < 0.05, respectively). In study II, the steady-state concentration of N-acetylcysteine was reached in 2-3 days in five of six patients during a constant infusion. CONCLUSIONS: The pharmacokinetics of N-acetylcysteine in pre-term infants depend markedly on weight and gestational age. The elimination of N-acetylcysteine is much slower in pre-term new-borns than in adults.

Expression and developmental profile of antioxidant enzymes in human lung and liver.

Air breathing, especially oxygen therapy, exposes the lung to reactive oxygen species (ROS). Antioxidant enzymes (AOEs) may protect the lung from ROS-mediated injury. Because expression of the key AOEs increases in several animal species during gestation, we investigated (1) the messenger RNA (mRNA) and activity levels of the key AOEs manganese and copper-zinc superoxide dismutases (MnSOD and CuZnSOD, respectively), catalase (CAT), and glutathione peroxidase (GPx) in adult lung samples and during ontogenesis; and (2) the difference in AOE expression between lung and liver. In the lung, the mRNA expression of MnSOD, CuZnSOD, and CAT increased toward adulthood, and GPx was unchanged. Pulmonary activities of MnSOD and CuZnSOD were unchanged, whereas CAT increased 3-fold from fetuses to adults. In the liver, the mRNA expression of MnSOD, CuZnSOD, and GPx increased, whereas that of CAT decreased toward adulthood. Hepatic activities of MnSOD and CuZnSOD increased 2-fold and 4-fold, respectively, whereas CAT was similar in fetuses and adults. Neonatal GPx activity was 2-fold higher in the lung and 6-fold higher in the liver compared with adults. The mRNA levels of MnSOD correlated positively with those of CuZnSOD and CAT in the lung, and GPx with those of MnSOD and CuZnSOD in the liver. Activities of MnSOD and CuZnSOD correlated positively in the liver. We conclude that the regulation of AOEs differs between human lung and liver, and is not tightly coordinated in either tissue.

Assignment of the locus for a new lethal neonatal metabolic syndrome to 2q33-37.

A new neonatal syndrome characterized by intrauterine growth retardation, lactic acidosis, aminoaciduria, liver hemosiderosis, and early death was recently described. The pathogenesis of this disease is unknown. The mode of inheritance is autosomal recessive, and so far only 17 cases have been reported in 12 Finnish families. Here we report the assignment of the locus for this new disease to a restricted region on chromosome 2q33-37. We mapped the disease locus in a family material insufficient for traditional linkage analysis by using linkage disequilibrium, a possibility available in genetic isolates such as Finland. The primary screening of the genome was performed with samples from nine affected individuals in five families. In the next step, conventional linkage analysis was performed in eight families, with a total of 12 affected infants, and finally the locus assignment was proved by demonstrating linkage disequilibrium to the regional markers in 20 disease chromosomes. Linkage analysis restricted the disease locus to a 3-cM region between markers D2S164 and D2S2359, and linkage disequilibrium with the ancestral haplotype restricted the disease locus further to the immediate vicinity of marker D2S2250.

Xanthine oxidoreductase gene expression and enzyme activity in developing human tissues.

Xanthine oxidoreductase (XOR) has been implicated in tissue injury following ischemia-reperfusion because of its ability to generate reactive oxygen species under these conditions. In order to elucidate its role in various organs, we quantified the levels of XOR mRNA expression and activity in developing human tissues. XOR gene expression was highest in the intestine and in the liver, which also showed the highest enzyme activities. By a sensitive RNA-PCR assay, low levels of the transcript were detected in the kidney, lung, cardiac muscle, and brain of all subjects studied. XOR activities followed the mRNA distribution, being low or undetectable in tissues other than the liver and the intestine.

Antioxidant enzyme regulation and resistance to oxidants of human bronchial epithelial cells cultured under hyperoxic conditions.

Bronchial epithelial cells are the first cells to encounter high concentrations of inspired oxygen, and their damage is a typical feature in many airway diseases. The direct effect of oxygen on the expression of the main antioxidant enzymes (AOEs) in human bronchial epithelial cells is unknown. We investigated the messenger RNA (mRNA) levels of manganese superoxide dismutase (MnSOD), copper-zinc superoxide dismutase (CuZnSOD), catalase (CAT), and glutathione peroxidase (GPx), as well as the specific activities of MnSOD, CuZnSOD, CAT, GPx, and glutathione reductase, in BEAS-2B bronchial epithelial cells exposed to hyperoxia (95% O2, 5% CO2) for 16 to 48 h. We also assessed the resistance of cells preexposed to hyperoxia to subsequent oxidant stress. Significant cell injury was observed after 72 h exposure to hyperoxia; release of lactate dehydrogenase (LDH) from control cells and cells exposed to hyperoxia for 72 h was 7.0 +/- 1.0% and 22.0 +/- 1.0%, respectively. Hyperoxia for 16 h, 24 h, or 48 h had no effect on the mRNA levels or specific activities of any of these enzymes. Despite their unchanged AOE levels, cells exposed to hyperoxia for 48 h showed increased resistance to H2O2 and menadione. Total glutathione content of the cells increased by 55% and 58% after 24 h and 48 h, respectively, compared with normoxic controls. However, glutathione depletion with buthionine sulfoximine (BSO) did not diminish the oxidant resistance of hyperoxia-exposed cells. We conclude that AOEs in human bronchial epithelial cells are not directly upregulated by high oxygen tension, and that increases in AOE-specific activities or glutathione are not necessary for the development of increased oxidant resistance in these cells.

DNA single strand breaks and adenine nucleotide depletion as indices of oxidant effects on human lung cells.

The comet assay (single cell gel electrophoresis) is a novel method to assess DNA strand breaks in single cells. We studied the oxidant sensitivity of cultured primary and transformed (MeT-5A) human pleural mesothelial cells, as well as primary and transformed (BEAS 2B) human bronchial epithelial cells, and compared the results obtained with the Comet assay to other markers of oxidant effects on cells, such as depletion of intracellular high-energy nucleotides (ATP, ADP, AMP), accumulation of products of nucleotide catabolism (xanthine, hypoxanthine, uric acid), and release of lactate dehydrogenase (LDH). The cells were exposed for 5 min to 4 h to 50-500 microM H2O2 or to 5-50 microM menadione. Significant tail moment increase, which is a marker of DNA strand breaks in the Comet assay, and intracellular nucleotide depletion occurred simultaneously in MeT-5A and BEAS 2B cells during the first 30-60 min of exposure to H2O2 and menadione. In the Comet assay variation between the individual cells could be detected. LDH release, a marker of cell injury, showed that mesothelial cells were far more sensitive than epithelial cells to oxidant-induced lytic cell injury. MeT-5A and BEAS 2B cells contained similar intracellular antioxidant enzyme activities, which may explain their similar oxidant sensitivity in the Comet assay. A significant increase (164%) in the tail moment was detectable in MeT-5A cells exposed to 50 microM H2O2 for 30 min. This returned to control level during the 4 h of continuing exposure. A 30 min exposure of 25 microM menadione caused a 61% increase in the mean tail moment but, unlike with H2O2, the change was irreversible during the following 4 h incubation. We conclude that human pleural mesothelial cells and bronchial epithelial cells show similar oxidant sensitivity when assessed by the Comet assay, but various oxidants differ in their potency in causing DNA breaks in these cells.

Circulating xanthine oxidase and neutrophil activation during human liver transplantation.

BACKGROUND & AIMS: Oxygen free radicals, generated by xanthine oxidase (XO) and activated leukocytes, are involved in reperfusion injury in experimental liver transplantation. The roles of XO and neutrophil activation during reperfusion in clinical liver transplantation were studied. METHODS: In 10 patients undergoing liver transplantation, we assessed plasma concentrations of circulating XO by enzyme-linked immunosorbent assay (ELISA), the purine metabolites hypoxanthine, xanthine, and urate by high-performance liquid chromatography, lactoferrin by ELISA, and malondialdehyde fluorometrically up to 48 hours postoperatively. RESULTS: During reperfusion after portal vein declamping, elevated plasma concentrations of XO (52.1 ng/mL [range, 8.0-440.1]), hypoxanthine (81.62 micromol/L [48.2-108.7]), xanthine (21.01 micromol/L [8.7-22.3]), and lactoferrin (532.6 ng/mL [370.4-1326.6]) were observed compared with the preoperative levels (0 ng/mL [0-12], 1.88 micromol/L [0.62-3.15], 0.95 micromol/L [0-0.41], and 164.3 ng/mL [73.7-334.1], respectively; all P < 0.05). No changes occurred in urate or malondialdehyde. After portal vein declamping, XO, hypoxanthine, and xanthine levels were substantially greater in the hepatic than portal vein (all P < 0.05). Marginal transhepatic differences occurred in lactoferrin. CONCLUSIONS: Reperfusion during liver transplantation is associated with liberation of xanthine oxidase, hypoxanthine, and xanthine from the liver into the circulation. During reperfusion, intravascular neutrophil activation takes place in the hepatic circulation.

Similarities between TNF and exogenous oxidants on the cytotoxic response of c-Myc-expressing fibroblasts in vitro.

Normal fibroblasts are resistant to the cytotoxic activity of tumor necrosis factor-alpha, but are rendered TNF-sensitive upon oncogenic expression of c-Myc. Free radical generation has been implicated in non-cytotoxic TNF-signaling but also as a mediator of TNF-induced cell death. In this study we used Rat1 fibroblasts containing a conditionally active form of oncogenic c-Myc (MycER) to investigate single cell line TNF-induced free oxygen radical formation during the non-cytotoxic TNF-response (inactive c-Myc) and cytotoxic response (active c-Myc). The generation of reactive oxygen species (ROS) was assayed using a fluorescent probe, dichlorodihydrofluorescein (DCFH-DA), and the following cellular injury by measuring the high energy nucleotide (ATP, ADP and AMP) depletion. We found that TNF treatment of Rat1 cells containing c-Myc in an inactive form did not induce a detectable level of ROS generation. In contrast, TNF treatment of Rat1 cells containing activated c-Myc caused fluorescence reaction indicative of ROS generation within 80 min after DCFH-DA exposure of the cells. The nucleotide depletion likely reflected the action of ROS, since the nucleotide depletion caused by TNF or oxidants such as menadione or H2O2 in cells with active c-Myc was partly inhibited by the anti-oxidant N-acetylcysteine.

Endogenous antioxidant enzymes and glutathione S-transferase in protection of mesothelioma cells against hydrogen peroxide and epirubicin toxicity.

We have previously shown that cultured malignant mesothelioma cells contain elevated manganese superoxide dismutase (MnSOD) mRNA levels and activities compared with non-malignant mesothelial cells. As many cytotoxic drugs generate both superoxide and hydrogen peroxide, we assessed the relative significance of catalase and the glutathione redox cycle, as well as glutathione S-transferase (GST), in protecting these cells against hydrogen peroxide and epirubicin toxicity. Mesothelioma cell lines containing high (M38K cells) and low (M14K cells) MnSOD, and non-malignant MeT-5A mesothelial cells were selected for the study. M38K cells were the most resistant of these three cell types to hydrogen peroxide (0.1-0.5 mM, 4 h) and epirubicin (0.1-0.5 microg ml(-1), 48 h) as judged by lactate dehydrogenase (LDH) release and by high-energy nucleotide (ATP, ADP, AMP) depletion. Total glutathione was higher in M38K cells (63.8 +/- 20.3 nnmol mg(-1) protein) than in M14K (25.2 +/- 8.2 nmol mg[-1]) or MeT-5A cells (23.5 +/- 4.5 nmol mg[-1]). Furthermore, GST specific activity was higher in M38K cells (111.3 +/- 15.8 U mg[-1]) than in M14K cells (77.4 +/- 6.6 U mg[-1]) or in MeT-5A cells (68.8 +/- 7.6 U mg[-1]). Western blotting indicated the presence of GST-pi in all these cells, the reactivity again being highest in M38K cells. Depletion of glutathione by buthionine sulphoximine and inhibition of catalase by aminotriazole enhanced hydrogen peroxide toxicity in all cell types, while only the depletion of glutathione increased epirubicin toxicity. We conclude that simultaneous induction of multiple antioxidant enzymes can occur in human mesothelioma cells. In addition to the high MnSOD activity, hydrogen peroxide scavenging antioxidant enzymes, glutathione and GST can partly explain the high hydrogen peroxide and epirubicin resistance of these cells in vitro.