Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Bronchial epithelium is the first target of a high concentration of inspired oxygen, and epithelial damage is a typical feature in human airway diseases. Antioxidant enzymes (AOEs) can be postulated to be responsible for the protection of bronchial epithelial cells against high oxygen tension, oxidants, and pollutants in vivo. Although the regulation of AOEs by hyperoxia has been studied extensively in experimental animals and various cell types, many of the results of these studies are controversial (1- 14). Only few studies are available on the direct effect of oxygen on the AOEs in human cells (15), and most of these have been conducted on vascular endothelial cells (15).

Hyperoxia causes upregulation of messenger RNA (mRNA) levels and/or increase in the specific activities of the major AOEs (1, 11, 12), most prominently manganese superoxide dismutase (MnSOD) (3, 9, 11, 12), but also copper-zinc superoxide dismutase (CuZnSOD) (3, 5, 7), catalase (CAT) (5, 9), glutathione reductase (GR) (3, 9), and glutathione peroxidase (GPx) (3, 5, 7, 9, 12) in various animal-cell models in vivo and in vitro. Animals preexposed to sublethal hyperoxia show resistance to consequent exposure to 100% O₂, and this resistance correlates with increased total SOD activity (1). However, these data are inconsistent, and unchanged or decreased specific activities of CAT and GPx after hyperoxic exposures in vivo have been reported (8, 14); this discrepancy is even more obvious in studies performed in vitro (11).

Very little is known about regulation of the major AOEs by oxygen in human cells (15). The effect of hyperoxia on CAT in human bronchial epithelium has been investigated in two studies (18, 19), which have shown that CAT is constitutively expressed in human bronchial epithelium. No corresponding studies of the expression of other main AOEs, their regulation, and/or their significance in protecting human bronchial epithelial cells have been conducted. We investigated the effect of hyperoxia on MnSOD, CuZnSOD, CAT, and GPx, using cells of a human bronchial epithelial cell line (BEAS-2B), and further assessed the resistance of hyperoxia-preexposed cells to subsequent exogenous oxidants.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture

Transformed BEAS-2B cells (20, 21) were obtained from the National Cancer Institute, Laboratory of Human Carcinogenesis (Dr. C. Harris, Bethesda, MD). Cells were cultured on uncoated petri dishes in serum-free hormone-supplemented medium according to the manufacturer's instructions (bronchial epithelial cell growth medium [BEGM]; Clonetics Inc., San Diego, CA).

Oxidant and Cytokine Exposures

Cells were exposed to hyperoxia (95% O₂, 5% CO₂) for 16 to 72 h. Subconfluent cells in BEGM medium were placed into a Plexiglas chamber, through which gas flow was adjusted so that the high oxygen concentration remained stable during the 16- to 72-h incubation. To confirm the inducibility of MnSOD, normoxic cells were also treated with human recombinant tumor necrosis factor- (TNF-) (10 ng/ml, 5 × 10⁷ U/mg protein, 16 to 48 h). TNF- was provided by Boehringer Ingelheim, Europe (Vienna, Austria). In additional experiments, cells grown under normoxic conditions (21% O₂, 5% CO₂) and cells that had been exposed to hyperoxia (48 h) were further exposed to 5 to 50 µM menadione (Sigma Chemical Co., St. Louis, MO) or to 0.1 to 5 mM H₂O₂ for 4 h. Furthermore, nontreated cells and cells treated with buthionine sulfoximine (BSO; Sigma) (50 µM and 500 µM, for the last 18 h of the 48-h exposure to hyperoxia) were exposed to 0.1 to 5 mM H₂O₂. After the exposures, the cells were collected for analysis as subsequently described.

Lactate Dehydrogenase Release

Lactate dehydrogenase (LDH) release into the medium was measured spectrophotometrically, using pyruvic acid (Sigma) as the substrate (22). Total LDH was measured in cell lysates obtained by treatment with 1% Triton X-100. The results are expressed as percent of activity released into medium.

Adenine Nucleotide Depletion

Cells were preincubated with 0.1 mM (¹⁴C)-adenine (specific activity: 51 to 55 mCi/mmol; Amersham International, Amersham, UK) in BEGM medium for 48 h, starting at the same time as the hyperoxia exposure. The labeled medium was then removed and the cells were washed and exposed to menadione or H₂O₂ in serum-free RPMI 1640 medium (Gibco Europe, Paisley, UK) as described previously. After incubation, the medium was frozen until analysis. The cells were extracted with 0.42 N perchloric acid. Purine nucleotides adenosine triphosphate, adenosine diphosphate, and adenosine monophosphate (ATP, ADP, and AMP, respectively) in neutralized cell extract, and nucleotide catabolic products (xanthine, hypoxanthine, and uric acid) in the medium were separated by thin layer chromatography as described (23). The results are expressed as percent distribution of radioactivity among nucleotides retained in the cell, nucleotides in the medium, and their catabolic products.

Enzyme Activities

Cells were detached with trypsin, and were pelleted and immediately frozen at 80°C until analysis. Total SOD activity was measured spectrophotometrically by the method of McCord and Fridovich (24). MnSOD activity was distinguished from CuZnSOD activity by its resistance to 1 mM potassium cyanide. GR activity was analyzed by measuring the oxidation of nicotinamide adenine diphosphate (NADPH) in the presence of oxidized glutathione (25), and GPx activity was analyzed by measuring NADPH oxidation in the presence of t-butylhydroperoxide, glutathione, and GR (25). CAT activity was determined with an oxygen electrode, as described earlier (26). Enzyme activities are expressed as units per milligram of protein. Protein was measured by the method of Lowry and colleagues (27).

Northern Blot Analysis

Cells were collected in 4 M thiocyanate buffer, and the samples were frozen immediately at 80°C. Total RNA was extracted from the cells by the acid phenol-chloroform method (28). RNAs 10 µg/lane (20 µg/lane for MnSOD) were electrophoresed on a 1% agarose gel containing 0.36 M formaldehyde. Samples were capillary transferred onto Hybond-N nylon filters (Amersham International) and cross-linked to the filters by UV illumination (UV Stratalinker 1800; Stratagene, La Jolla, CA). Prehybridization was done at 58.5°C for > 1 h in buffer containing 50% deionized formamide, 5× standard saline citrate (SSC), 50 mM sodium phosphate (pH 6.5), 5× Denhardt's reagent, and 100 µg/ml herring-sperm DNA. ³²P-labeled complementary RNA (cRNA) probes were transcribed from cDNA clones representing nucleotides 596 to 987 of human MnSOD (29), 127 to 457 of human CuZnSOD (30), 537 to 2,218 of human CAT (31), and 533 to 624 of rat GPx (32), all cloned into the pSP65 vector (Promega Co., Southampton, UK). The transcripts were purified with NucTrap columns (Stratagene) and added to the prehybridization solution at 2 × 10⁶ cpm/ ml. Hybridization was then done overnight at 58.5°C with shaking. After washing with 2× SSC and 0.2× SSC at room temperature, and with 0.2× SSC and 0.1% sodium dodecyl sulfate (SDS) at 58.5°C, autoradiography was done at 80°C using Kodak BioMax MR film (Eastman Kodak Co., Rochester, NY). Following autoradiography, the filters were rehybridized with a -actin control probe transcribed from pTRI--actin plasmid (Ambion, Austin, TX). AOE mRNA expressions were quantified relative to actin expression, using an X-Rite 331 transmission densitometer (X-Rite, Grandville, MI). The MnSOD, CuZnSOD, CAT, and GPx cDNAs were kindly provided by Dr. Y.-S. Ho of Wayne State University, Detroit, MI.

Glutathione Content

Cells were collected in 2 N perchloric acid containing 2 mM ethylenediamine tetraacetic acid (EDTA). After neutralization with a solution containing 2 M KOH and 0.3 M N-morpholinopropanesulfonic acid (MOPS), total cellular glutathione content was determined spectrophotometrically by measuring the reduction of 5,5'-dithiobis(2-nitrobenzoic acid) (Sigma) by NADPH in the presence of GR (Sigma) (33). Cellular glutathione content is expressed as nmol/mg protein.

H₂O₂ Consumption

H₂O₂ consumption by normoxic and hyperoxia-exposed intact cells after 48 h was analyzed by measuring the H₂O₂-dependent oxidation of 4-hydroxy-3-methoxyphenylacetic acid (homovanillic acid [HVA]; Sigma) in the presence of horseradish peroxidase (HRP; Sigma) with the modified method (34) of Ruch and colleagues (35). Incubations were conducted at 37°C in phosphate-buffered saline (PBS). H₂O₂ (500 µM) was added to the medium, and H₂O₂ consumption was analyzed on the basis of samples drawn after 10, 20, and 30 min incubation. The exact H₂O₂ concentration was determined spectrophotometrically, using the molar extinction coefficient of 44 M¹ cm¹ at 240 nm.

Statistical Analysis

Data are expressed as means ± SD. Comparisons of two groups were done with the nonparametric Mann-Whitney U test. Values of P < 0.05 were considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects of Hyperoxia on AOE Activities and mRNA Levels

Preliminary experiments indicated that significant cell injury (LDH release) was evident after 72 h exposure to hyperoxia. LDH release from control cells and from hyperoxic cells after 72-h exposures was 7.0 ± 1.0% (n = 4) and 22.0 ± 1.0% (n = 4) (P < 0.05), respectively. Therefore, hyperoxia exposures for 16 to 48 h were selected to investigate AOE regulation in these cells. MnSOD mRNA expression remained unchanged after all hyperoxia exposures, when calculated relative to -actin mRNA (Figure 1). It has to be emphasized that actin expression itself did not change during these exposures. Because TNF- is a well-known inducer of MnSOD in a number of cell types (6, 36), the cells were also exposed to TNF- to confirm this induction. TNF- caused a significant increase in MnSOD mRNA after 16 h, 24 h, and 48 h (Figure 1). The mRNA level after 48 h exposure to TNF- was significantly lower than the corresponding values after 16 h and 24 h. There was no change in the levels of mRNA for CuZnSOD, CAT, or GPx after 16 h, 24 h, or 48 h hyperoxia exposures (Figures 2-4). The activities of MnSOD, CuZnSOD, CAT, GR, and GPx did not change after 48 h exposure to hyperoxia compared with those of the normoxic controls (Table 1).

View larger version (59K): [in this window] [in a new window]   Figure 1.   Representative Northern blot of MnSOD mRNA and -actin mRNA of control (normoxic) BEAS-2B cells; cells exposed to hyperoxia for 16 h, 24 h, and 48 h; and cells exposed to TNF- (10 ng/ml) for 16 h, 24 h, and 48 h. Hybridization with a human MnSOD cRNA probe shows evidence of two transcripts of 4.0 kb and 1.0 kb. The experiments were repeated three times and indicated that only TNF- caused significant upregulation after 16 h, 24 h, and 48 h.

View larger version (50K): [in this window] [in a new window]   Figure 2.   Representative Northern blot of CuZnSOD mRNA and -actin mRNA of control (normoxic) BEAS-2B cells and cells exposed to hyperoxia for 16 h, 24 h, and 48 h. Hybridization with a human CuZnSOD cRNA probe shows evidence of two transcripts of 0.9 kb and 0.7 kb. The experiments were repeated three times. Hyperoxia had no effect on the levels of mRNA for CuZnSOD.

View larger version (47K): [in this window] [in a new window]   Figure 3.   Representative Northern blot of CAT mRNA and -actin mRNA of control (normoxic) BEAS-2B cells and cells exposed to hyperoxia for 16 h, 24 h, and 48 h. The experiments were repeated three times. Hyperoxia had no effect on the levels of CAT mRNA.

View larger version (49K): [in this window] [in a new window]   Figure 4.   Representative Northern blot of GPx mRNA and -actin mRNA of control (normoxic) BEAS-2B cells and cells exposed to hyperoxia for 16 h, 24 h, and 48 h. The experiments were repeated three times. Hyperoxia had no effect on the levels of GPx mRNA.

Cell Injury

Oxidant resistance was investigated by exposing the cells after 48 h hyperoxia to H₂O₂ or menadione for 4 h. Subsequent exposure of hyperoxic or normoxic cells to 0.5 to 1 mM H₂O₂ or to 5 µM menadione did not cause significant LDH release. LDH release from the hyperoxia-preexposed cells was significantly attenuated during the subsequent exposure to 5 mM H₂O₂, but was not significantly attenuated during the exposure to 50 µM menadione (Figure 5). Although hyperoxia as such caused significant high-energy-nucleotide depletion and accumulation of the catabolic products of these nucleotides, menadione exposure potentiated nucleotide catabolism in normoxic but not in hyperoxic cells (Figure 6), again suggesting the induction of protective mechanisms in hyperoxic cells.

View larger version (10K): [in this window] [in a new window]   Figure 5.   LDH release from control (normoxic) BEAS-2B cells and cells preexposed to hyperoxia for 48 h and subsequently exposed to H2O2 (5 mM) (A) or to menadione (50 µM) (B) for 4 h. Values are means ± SD from four to eight separate experiments. *P < 0.05 versus normoxic control.

View larger version (21K): [in this window] [in a new window]   Figure 6.   Adenine nucleotide depletion (A) and purine catabolite accumulation (B) in control (normoxic) BEAS-2B cells and cells preexposed to hyperoxia for 48 h and subsequently exposed to H2O2 (100 µM) or menadione (50 µM). Values are means ± SD from eight separate experiments. *P < 0.05 versus normoxic control. #P < 0.05 versus corresponding value in unexposed cells.

H₂O₂ Consumption

Given that the cells showed a tendency toward increased oxidant resistance after hyperoxia exposure, and that the levels of MnSOD, CuZnSOD, CAT, GR, and GPx were unchanged, H₂O₂ consumption by normoxic and hyperoxia-preexposed cells was measured to see whether other H₂O₂ consumption mechanisms are responsible for this effect. These experiments, standardized relative to total cell protein, indicated a similar H₂O₂-scavenging capacity (Figure 7).

View larger version (14K): [in this window] [in a new window]   Figure 7.   Consumption of exogenously added H2O2 (500 µM) by control (normoxic) BEAS-2B cells and cells exposed to hyperoxia for 48 h. Same amounts of cells were incubated in PBS for 10 to 30 min in the presence of HVA and HRP. H2O2 was measured from aliquots drawn during the incubation of 30 min. Values are means ± SD from eight separate experiments.

Effect of Hyperoxia Exposure on Cellular Glutathione Content

Given that hyperoxia can cause an increase of intracellular glutathione content, at least in cultured endothelial cells (12, 40), and because this effect has not been confirmed in human bronchial epithelial cells, glutathione levels were measured in control (normoxic) and hyperoxia-exposed cells. The results, which are shown in Figure 8, indicate that total glutathione was significantly increased after 24-h and 48-h hyperoxia exposures as compared with that of normoxic controls.

View larger version (12K): [in this window] [in a new window]   Figure 8.   Cellular glutathione content of control (normoxic) BEAS-2B cells and cells exposed to hyperoxia for 16 h, 24 h, and 48 h. Values are means ± SD from nine separate experiments. *P < 0.05 versus normoxic control.

Effect of BSO Treatment on the Cell Injury of Hyperoxic Cells

Further experiments, in which nontreated and BSO-treated hyperoxia-exposed cells were subsequently exposed to H₂O₂, were conducted. BSO (50 µM and 500 µM) caused a significant depletion of glutathione (to 13.5% and to 9.0% of the original glutathione content in 18 h, n = 5 or 6 separate experiments in duplicate), but was nontoxic to these cells (data not shown). When nontreated and BSO-treated hyperoxia-exposed cells were exposed to 1 mM or 5 mM H₂O₂, very similar LDH release was observed in the nontreated and BSO-treated groups (n = 4 to 6 separate experiments at both BSO concentrations, data not shown). In additional experiments, subsequent H₂O₂ exposure (100 µM) caused similar nucleotide depletion in both groups, the high-energy-nucleotide levels being 42.5 ± 1.7% in the nontreated cells and 36.5 ± 3.9% in the BSO (500 µM)-treated cells (n = 4 separate experiments in duplicate).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our findings indicate that in transformed human bronchial epithelial (BEAS-2B) cells, the levels of the AOEs MnSOD, CuZnSOD, CAT, and GPx remain unchanged after hyperoxia exposure, when followed to the point at which no significant cell injury can be detected. MnSOD upregulation by TNF- was noted in these cells, as shown earlier in many other cell types (6, 36). Hyperoxia-preexposed cells showed increased oxidant resistance during subsequent oxidant exposure, despite unchanged AOE levels. Glutathione was significantly increased by hyperoxia, but this also did not explain the increased oxidant resistance.

The effect of hyperoxia on the regulation of AOEs in rat lung has been extensively investigated (1, 39). Most but not all studies have shown upregulation of MnSOD mRNA and/or an increase in the specific activity of MnSOD in lung homogenates or alveolar type II pneumocytes with hyperoxia (3). Also, the results with H₂O₂-scavenging AOEs are contradictory. The induction of CAT, GPx, and GR has been minor or insignificant, or the activities of these enzymes have declined in hyperoxia (4, 7, 8). Immunoreactive enzyme protein, enzyme activity, and mRNA levels for the AOEs after hyperoxia do not necessarily correlate with each other (7, 15, 41, 42), which indicates the complexity of the regulatory mechanisms of these enzymes and may explain some of the discrepancies in various studies.

Information about the effects of oxygen on the molecular regulation of human AOEs is complex and even less understood than that for experimental animals (2, 9, 10, 13). In one recent study, no change in CAT was observed in the bronchial epithelium of healthy individuals breathing 100% oxygen for 12 h (18). Instead of using primary cells, we used a transformed human bronchial epithelial cell line (BEAS-2B), which is easy to maintain and is therefore suitable for induction and cytotoxicity studies in vitro. Transformed cells do not represent bronchial epithelial cells in vivo, but this is also true for primary bronchial epithelial cells, which lose their typical morphology and a significant part of their antioxidant capacity when cultured in vitro (43). BEAS-2B cells contain lower AOE levels than do bronchial epithelial cells in situ (43), but in our study they expressed all AOEs. In these cells MnSOD is induced by TNF-, as shown in the present study and in a recent study that we conducted (37). In the present study, the mRNA levels of the AOEs were not upregulated by hyperoxia, and the AOE activities remained unchanged under these experimental conditions. Because hyperoxia had no effect on AOE upregulation, other mechanisms are involved in the protection of these cells after prolonged oxygen exposure.

The induction of MnSOD in rat lungs in vivo is maximal at days 3 to 5 of hyperoxia exposure (7), which is also the time required for the recruitment of activated inflammatory cells into the lungs (34, 44). Because MnSOD, but not the other AOEs, is upregulated by many cytokines (TNF-, IL-1, IL-6, IFN-) (36, 38, 41), the in vivo data can be at least partly explained by inflammation, nuclear factor-B (NF-B) activation, and a subsequent increase in MnSOD transcription (45). In the present study, MnSOD was upregulated after 16 to 24 h of TNF- treatment, and the mRNA level and specific activity of MnSOD in TNF--treated cells was higher than in normoxic and hyperoxic cells. Although inflammatory cytokines, such as TNF-, are important in the upregulation of MnSOD, the significance of MnSOD induction on oxidant resistance is still unclear. Our previous study, for instance, showed that TNF--treated BEAS-2B cells with increased MnSOD activity were more sensitive to oxidants than were untreated control cells (37).

In the present study, hyperoxic cells with unchanged MnSOD activity were more resistant to subsequent oxidant exposure than were normoxic cells. Therefore, other mechanisms can be postulated to protect these cells against oxidants, or the complex balance of many AOE pathways is more important than any individual enzyme. In addition to MnSOD, another potent antioxidative factor in human bronchial epithelial cells has been suggested to be glutathione (46). This possibility was not the main focus of our study. However, total glutathione was found to be increased in hyperoxia. This finding is in agreement with previous studies, which have shown increases in glutathione content, cystine uptake, and -glutamyl transpeptidase activity in other cell types exposed to hyperoxia (47, 48).

Increased glutathione and enzymes related to glutathione synthesis in hyperoxic cells might also explain the increased oxidant resistance of hyperoxia-exposed cells in the present study. To test this hypothesis further, the cells were pretreated with BSO, which causes glutathione depletion. However, BSO did not enhance cell injury of hyperoxic cells during subsequent exposure to H₂O₂, indicating that glutathione was not responsible for the cell protection against oxidants. H₂O₂ consumption by the hyperoxic and normoxic cells was also identical, suggesting that other mechanisms than the major AOEs or cellular glutathione explain the observed resistance of BEAS-2B cells in these experimental conditions. The involvement of glutathione in protection against menadione, however, remains unclear, because this quinone can conjugate with glutathione and thus be rendered less toxic. Furthermore, our findings do not exclude the importance of glutathione in the oxidant resistance of bronchial epithelial cells in vivo, since theoretically nondifferentiated cells may respond to hyperoxia differently than primary cells or cells in vivo. The resistance to oxidant stress (and possibly also resistance to hyperoxia) may be partly determined by inducible defense mechanisms other than AOEs or glutathione content, as the expression of several yet unidentified transcripts is known to be modulated after nonlethal oxidant exposure (49).

In conclusion, oxygen appears not to be a direct regulator of AOE gene expression in human BEAS-2B bronchial epithelial cells, and neither increases in AOE-specific activities nor in glutathione concentration are necessary for the development of increased oxidant resistance in these cells.