Main

ROS have been implicated in the pathogenesis of various lung diseases including cancer, pulmonary fibrosis, adult respiratory distress syndrome, emphysema, chronic bronchitis, and pleural disease. Many chemical and environmental agents including hyperoxia, ozone, UV radiation, nitrogen and sulfur oxides, etc. are potent initiators of ROS generation in the lung (1, 2). In experimental models, it has been observed that the specific targets of hyperoxic insult to the lung seem to be the vascular endothelial cells and the epithelial cells of the alveoli (3, 4). ROS result in ultrastructural changes in the cytoplasm of pulmonary capillary endothelial cells and cause focal hypertrophy and altered metabolic activity (5). The generation of the superoxide anion and hydrogen peroxide can result in formation of lipid hydroperoxides in the cell membrane, which may severely disrupt its function and result in death of epithelial cells of the alveoli (4).

GPx are antioxidant enzymes that detoxify hydrogen peroxide and lipid hydroperoxides using reduced glutathione. E-GPx is genetically distinct from the well-studied C-GPx (6), phospholipid hydroperoxide glutathione peroxidase (7), and an additional cytoplasmic glutathione peroxidase in the gastrointestinal tract (8, 9). In adults, most of the blood plasma E-GPx activity is derived from the kidney, with additional sites of synthesis of E-GPx mRNA in lung, heart, and intestine (10, 11).

E-GPx activity has been found in the human bronchoalveolar lavage fluid: at least 60% of the cell-free bronchoalveolar lavage fluid GPx activity in normal adults is due to E-GPx. Bronchial epithelial cells and alveolar macrophages are potential sources of pulmonary airspace E-GPx because these cells synthesize both E-GPx mRNA and E-GPx protein in vitro (12).

Because the extracellular environment, especially the cell membrane, is exposed and damaged by ROS exposure, this study was designed to evaluate the effect of exposure of mice to hyperoxia on E-GPx and C-GPx in the lung. Increased activity of E-GPx could play a role in the protection of the lung and the mammalian organism from oxidant damage from exposure to hyperoxia.

METHODS

Animals.

Adult C57BL mice (approximately 20 wk old, 25 g), maintained on standard mouse chow and water ad libitum, were used in a protocol approved by the Institutional Review Board of Stanford University. Exposure of the mice to hyperoxic or normoxic atmosphere was conducted in polystyrene chambers. Mice exposed to hyperoxia received an FiO2 >99%. Mice exposed to normoxia comprised the control group and received ambient room air (FiO2 = 21%) in similar chambers. The animals were housed in their respective chambers for 72 h, with food and water available ad libitum. The hyperoxic and normoxic exposures were carried out at the same time on groups of five mice each. The entire series of exposures was performed twice. Virtually identical results were obtained in the two series of exposures. Except as noted, all presented data are pooled from both series of exposures.

Preparation of lung homogenate.

After exposure to either normoxia or hyperoxia for 72 h, the mice were killed by exposure to carbon dioxide gas. Lungs were rapidly removed en bloc without perfusion, rinsed in PBS, and placed in protein lysis buffer solution containing 10 mM Tris, 150 mM NaCl, 0.5% Triton X-100, 1% Nonidet P-40, 10 mM NaN3, 10 μg/mL leupeptin, 0.5 mM para-amino benzamide, 1.0 mM phenylmethylsulfonyl fluoride, at pH 7.5. Lungs (50 mg) were homogenized on ice in 1 mL of lysis buffer by using a Tissue Tearor (Biospecs Products, Inc., Racine, WI). Samples were prepared for SDS-PAGE, assay of protein concentration, and assay of GPx activity. All samples were saved at −70°C until further processing.

Isolation of plasma.

Immediately after sacrifice, 1 mL of blood was removed from the right ventricle of the mouse into a syringe containing heparin and was placed on ice. Plasma was separated from the cellular components by centrifugation at 1000 ×g for 10 min at 4°C and was kept at −70°C until analyzed.

GPx activity.

GPx activity was measured by an adaptation (10) of the method of Beutler (13), using t-BuOOH and glutathione (GSH) as substrates. Activity was assayed in a coupled enzymatic reaction by following the oxidation of NADPH at 340 nm at 37°C in a double-beam spectrophotometer (Cary 1E, Varian, Palo Alto, CA) in the presence of glutathione reductase. The standard reaction mixture contained 0.1 M Tris-HCl, pH 8.0, 2.0 mM NADPH, 0.5 mM EDTA, 2 mM GSH, 1 U glutathione reductase, and either plasma or lung homogenate in a total volume of 0.99 mL. The reaction was started by the addition of 10 μL t-BuOOH to the sample cuvette only (final concentration, 70 μM). One unit of GPx activity is defined as the amount of enzyme that oxidizes 1 μmol NADPH/min. GPx activity was expressed as U/mg protein and U/mL in plasma and as U/mg protein and U/μg DNA in lung. Protein concentration was determined using a modified Lowry assay kit (Bio-Rad), with bovine IgG as the standard. DNA was determined using the diphenylamine method as described (14, 15).

Isolation of total RNA and analyses of mRNA.

Total RNA from lung tissue was isolated using the Ultraspec RNA isolation system (Biotecx, Houston, TX) based on the method of Chomczynski and Sacchi (16). RNA (10 μg) was transferred to positively charged nylon membranes (Nytran Plus, Schleicher & Schuell, Keene, NH) by both Northern and slot blot methods. Murine E-GPx cDNA (clone 47-A1, a gift from Dr. R. Maser, University of Kansas) was labeled with [α-32P]-dCTP (specific activity, 3000 Ci/mmol; NEN, Boston, MA) to specific activities > 5 × 108 cpm/μg by using a random-primer polymerase kit (RadPrime Labeling Kit, Life Technologies, Gaithersberg, MD). The blots were prehybridized and hybridized in 5X Denhardts, 50% deionized formamide, 10% dextran sulfate, 250 μg/mL salmon sperm DNA, 1% SDS, 2X SSC at 42°C overnight. The blots were then washed at 55°C with the most stringent washing in 0.1X SSC (sodium chloride/sodium citrate, where 1X = 0.15 M NaCl/0.015 M Na3 citrate. 2H2O, pH 7.0)/1% SDS before exposure to a phosphor imaging screen. The screens were scanned and quantified using a Bio-Rad GS-505 molecular imager (Bio-Rad, Hercules, CA). The blots were stripped by boiling in 0.1% SDS for 1 min and checked for background before the next probe was used. The blots were then sequentially probed with murine C-GPx (a gift from Dr. N. Imura, Kitasato University, Japan), GAPD, and α-tubulin cDNA probes prepared and used in the same manner.

Western blot analysis.

For analysis of the amount of E-GPx and C-GPx protein, 50 μg of mouse lung protein (from the mouse lung protein homogenate described above) or mouse plasma was boiled in SDS sample buffer and electrophoresed on 12.5% SDS-polyacrylamide gels under reducing conditions (10 mM DTT) (17). After electrophoresis, the proteins were transferred to a membrane (Hybond-PVDF, Amersham Life Science) in a semidry electroblotter (Owl Scientific, Portsmouth, NH). The membranes were blocked with PBS containing 6% (wt/vol) nonfat dry milk at room temperature for 1 h and were then incubated for 1 h with either rabbit anti-mouse E-GPx antibody or rabbit anti-mouse C-GPx antibody at a 1:20 000 dilution. After washing the membranes in PBS containing 0.05% polyethylene-sorbitan monolaurate (PBS/Tween 20), the immunoblots were incubated with the secondary antibody, goat anti-rabbit IgG-horseradish peroxidase at a 1:10 000 dilution (Pierce, Rockford, IL) for 1 h at room temperature. The blots were again washed vigorously with PBS/Tween 20. The blots were then developed with chemiluminescent substrate (Super Signal, Pierce, Rockford, IL). Blots were exposed to a chemiluminescence phosphor screen and quantified using a Bio-Rad GS-505 molecular imager. Lanes with mouse plasma and kidney were used as controls on each gel. Similar methods were used for quantification of albumin in lung and plasma samples, except that 10% SDS-PAGE was used. In addition, lower amounts of mouse plasma (1 μg) and mouse lung (12.5 μg) were electrophoresed per lane.

RESULTS

Injury model of hyperoxia exposure.

In this injury model of hyperoxia exposure, the adult mice tolerated 48 h of FiO2 >99% without apparent distress. By 72 h, the mice exposed to hyperoxia had become somewhat lethargic with more labored and shallow respirations. Protein concentration of the plasma of normoxia- and hyperoxia-exposed mice was not significantly different (30.38 ± 0.62 and 31.08 ± 0.74 mg/mL for air- and oxygen-exposed animals, respectively;p= 0.47), indicating that the plasma volume of the hyperoxia-exposed animals had not changed.

Plasma GPx activity and E-GPx protein are increased after hyperoxiaexposure.

GPx activity was determined in the plasma from mice exposed to either hyperoxia or air. Hyperoxia exposure was associated with an increase of 55% in plasma GPx activity: the plasma GPx activity was 1.595 ± 0.046 and 2.477 ± 0.120 U/mL for normoxia- and hyperoxia-exposed mice, respectively (Fig. 1A). The increase in plasma GPx activity was similar and statistically significant whether analyzed on the basis of U/mL plasma (shown in Fig. 2A, p< 0.0001) or U/mg plasma protein (Fig. 1B, p< 0.0001), indicating that potential changes in plasma volume of the mice did not account for the increase in plasma GPx activity seen in hyperoxia-exposed mice. We quantified E-GPx protein in the plasma by Western blotting to verify that the increase in plasma GPx activity was due to an increase in E-GPx protein in the plasma. E-GPx protein (a subunit size of 22 kD) in plasma from hyperoxia-exposed mice was also increased relative to control mice (results from exposure series 1 are shown in Fig. 1C;p< 0.05 in each of the two series of exposures). By this technique, the amount of E-GPx protein increased by 50%.

Figure 1
figure 1

Plasma GPx activity and E-GPx protein. Plasma GPx activity was measured as described in “Methods.” Shown are the results for both series of exposures. A, The results are presented as U/mL plasma (p< 0.0001, unpaired t test). B, Plasma GPx activity is presented as U/mg plasma protein (p< 0.0001, unpaired t test). C, The amount of E-GPx protein in the plasma in exposure series 1 was quantified by Western blot as described in “Methods” (p< 0.05, unpaired t test).

Full size image
Figure 2
figure 2

Lung GPx activity. Lung GPx activity was measured as described in “Methods.” Shown are the results for the mice in both series of exposures, normalized to lung protein (A) (p= 0.4, unpaired t test) and lung DNA (B) (p= 0.4, unpaired t test) concentrations.

Full size image

Lung GPx activity is unchanged after hyperoxia exposure.

GPx activity in lung was determined after exposure to either normoxia or hyperoxia. GPx activity of the homogenates prepared from the lungs of mice exposed to either hyperoxic or normoxic conditions for 72 h and the lung GPx activities from these mice are shown in Figure 2. The lung GPx activities were 0.146 ± 0.017 and 0.126 ± 0.016 U/mg protein for normoxia- and hyperoxia-exposed mice, respectively. Although the lung GPx activity seemed to trend to lower activity in the hyperoxia-exposed mice, the lung GPx activities for the two groups of mice were not statistically different (unpaired t test, p= 0.4068). The measured GPx activity in lung homogenates was also normalized to DNA content of the lung homogenates of the two groups of mice. The DNA concentration of the lungs did not differ in the two groups: 62.48 ± 10.98 and 61.08 ± 7.79 μg DNA/mg protein for air- and oxygen-exposed animals, respectively;p= 0.92, unpaired t test). Consequently, GPx activity normalized to DNA concentration also was not significantly different, although the trend was in the same direction as for GPx activity normalized to protein.

Lung E-GPx protein is increased with hyperoxia exposure.

Because plasma GPx activity and the amount of E-GPx protein in plasma were increased after exposure to hyperoxia, the amounts of E-GPx and C-GPx proteins in the lungs were also determined. E-GPx and C-GPx protein in lung homogenate were quantified by Western blots by using excess antibodies specific for the individual enzymes. E-GPx protein content was >6 times higher in the lung homogenates from mice exposed to hyperoxia relative to normoxia-exposed mice (see Fig. 3A for the mice in series 1). The level of C-GPx protein content (subunit size of 23 kD) in the lung homogenates was decreased by approximately 25% (Fig. 3B), in agreement with the downward trend of total GPx activity in the lungs of mice exposed to hyperoxia (Fig. 2). The ratio of E-GPx:C-GPx protein in the lung was increased approximately 7-fold in the two series of exposures.

Figure 3
figure 3

E-GPx and C-GPx protein in lungs. E-GPx and C-GPx protein levels were quantified by Western blot as described in “Methods.” Shown are the results from exposure series 1, with comparable results obtained in exposure series 2. In (A), the amount of E-GPx protein is increased in hyperoxia-exposed lungs (p< 0.005, unpaired t test). In (B), the amount of C-GPx protein in hyperoxia-exposed lungs is not significantly different from normoxia-exposed lungs (p= 0.075, unpaired t test). In (C), a ratio of E-GPx:C-GPx is presented to relate the changes in amounts of these proteins when comparing the lungs of hyperoxia- and normoxia-exposed mice (p< 0.01, unpaired t test). These ratios provide no information about the relative amounts of E-GPx or C-GPx.

Full size image

Plasma contamination does not account for lung E-GPx.

To determine whether the increase in lung E-GPx protein was due to increased plasma in hyperoxia-treated lungs, lung homogenates were also probed with a rabbit anti-mouse albumin antibody in a Western analysis to determine how much albumin was present in the lung homogenate. The albumin values were then used to calculate the volume of plasma that was present in the lung homogenates, based on the value obtained for a plasma control performed on the same blot. By use of this method, the lungs of normoxia- and hyperoxia-exposed mice contained 0.66 ± 0.05 and 1.00 ± 0.09 μL plasma/mg lung homogenate protein, respectively, indicating an increased amount of blood or plasma leak in the lungs of hyperoxia-exposed mice. On the basis of the measured plasma GPx activity of the mice in the different groups, the calculated amount of E-GPx (in units) in the lung due to plasma volume in the lung was determined. In addition, the amount of E-GPx protein in lung (from the data in Fig. 3A) was converted to the total amount of E-GPx (in units) on the basis of a plasma standard of known activity that was quantified at the same time. The results of these calculations of total lung E-GPx are presented in Figure 4, with the assumption that all E-GPx protein in lungs is enzymatically active. By these calculations, the amount of E-GPx protein associated with plasma contamination in the lung accounted for approximately 10% of the total lung E-GPx in the normoxia-exposed control mice. In hyperoxia-exposed mice, the amount of E-GPx protein calculated to be due to the amount of plasma in the lungs accounted for only 6% of total lung E-GPx protein. Therefore, the majority of E-GPx protein present in the lung in both normoxic and hyperoxic conditions is probably synthesized and secreted by lung cells. By comparing the data from Figures 2 and 4, it can be seen that lung E-GPx would account for approximately 10% of lung GPx activity in normoxic lungs. However, in the lungs of mice exposed to hyperoxia, lung E-GPx would account for approximately 45% of total lung GPx activity.

Figure 4
figure 4

Contribution to lung E-GPx by plasma E-GPx. The total amount of E-GPx protein in the lungs of normoxia- and hyperoxia-exposed mice was calculated on the basis of the quantification of E-GPx protein from lung and plasma on the same Western blot and the plasma GPx activity in the different mice. The amount of E-GPx protein in the lungs that is attributed to the presence of plasma in the lung is calculated on the basis of the quantification of mouse albumin in the lung samples and plasma standards probed on the same Western blot. The data are presented as units of E-GPx, with the assumption that all E-GPx protein in lungs is enzymatically active. Shown are the data from exposure series 1, with comparable results obtained in exposure series 2.

Full size image

Effect of hyperoxia on lung E-GPx mRNA.

The amounts of E-GPx, C-GPx, GAPD, and tubulin mRNA in lung were quantified by both Northern and slot blot assays of total RNA in this injury model of the effects of hyperoxia on lung. Similar results were obtained with both techniques. On the basis of slot blot analysis of identical amounts of total RNA, the absolute amounts of E-GPx and C-GPx mRNA in lung increased 87% (p< 0.01) and 215% (p< 0.001), respectively, after 3 d of hyperoxia. The absolute amount (similarly on the basis of total RNA) of GAPD and tubulin mRNA in lung also increased after hyperoxia by 55% (p< 0.01) and 83% (p< 0.0001), respectively, in a manner similar to that found by Ho et al. (18). The level of 28S ribosomal RNA in the preparations of total lung RNA from air- and hyperoxia-exposed animals was not different as quantified by ethidium bromide staining of gels prepared for Northern blot analysis (data not shown). There were no significant differences when the amounts of lung E-GPx and C-GPx mRNA were normalized to either GAPD or tubulin levels (Fig. 5). In addition, the ratio of E-GPx:C-GPx mRNA levels was the same for the lungs of mice exposed to normoxia or hyperoxia (data not shown). By Northern blot analysis, the size of the transcripts for E-GPx (1.9 kb) and C-GPx (1.2 kb) was not altered at 72 h of exposure to hyperoxia. Thus, although there were changes in the amounts of E-GPx, C-GPx, GAPD, and tubulin mRNA levels relative to total RNA, there were no changes in the amount of E-GPx or C-GPx mRNA relative to either GAPD or tubulin or in the size of the transcripts for E-GPx or C-GPx in lungs measured after 72 h of exposure to hyperoxia.

Figure 5
figure 5

E-GPx and C-GPx mRNA content in lung. Total RNA isolated from the lungs of hyperoxia- and normoxia-exposed mice was isolated and probed by slot blot. Shown are the data for the mice from both series of exposures. In (A), the amount of E-GPx mRNA is normalized to the amount of GAPD (p= 0.21, unpaired t test) and tubulin mRNA (p= 0.51). In (B), C-GPx mRNA was normalized to GAPD (p= 0.06, unpaired t test) and tubulin mRNA (p= 0.41, unpaired t test).

Full size image

DISCUSSION

The pulmonary system is susceptible to oxidant-induced injury in a variety of therapeutic and toxic situations. The family of GPx is an important enzymatic component of the mechanisms for detoxifying ROS in the lung (19). Prior studies have implicated a significant role for GPx in preventing pulmonary oxidant damage. Selenium deficiency was associated with decreased rat lung GPx activity and worsened pulmonary injury after hyperoxic exposure (20). Although these studies illustrate the importance of availability of selenium in prevention of oxidant-mediated pulmonary injury, the relative importance of the different selenium-containing enzymes required further study. Besides GPx, other selenium-dependent enzymes include types I and III iodothyronine-5′-deiodinase, selenoprotein P, and thioredoxin reductase (21–27).

Recently, mice made deficient in C-GPx by targeted inactivation of the gene for C-GPx were developed and evaluated for susceptibility to the effects of hyperoxia (28). Mice without C-GPx showed no increased sensitivity to hyperoxia in survival studies at 10 wk of age. The tissues of the C-GPx–deficient mice exhibited no deficit in the rate of consuming extracellular hydrogen peroxide, and no increased content of protein carbonyl groups and lipid peroxidation, compared with the wild-type mice. These results suggest that the tissues in C-GPx null mice can still effectively decompose hydrogen peroxide, possibly by other antioxidant enzymes such as catalase, selenium-independent glutathione-peroxidizing enzymes such as glutathione S-transferases, nonenzymatic mechanisms, or other members of the GPx family such as E-GPx. A possible role of E-GPx in preventing injury due to hyperoxia had not been evaluated.

In the present study, the amount of E-GPx in plasma and lung was increased after exposure of mice to hyperoxia in an injury model. Plasma GPx activity was increased by approximately 50% whether normalized to either plasma protein concentration or volume of plasma. E-GPx protein was significantly increased in both the plasma and lungs of mice exposed to hyperoxia. Although total GPx activity (using the substrates t-BuOOH and GSH) in the lungs of mice exposed to hyperoxia was not significantly changed, there was a large increase in the amount of E-GPx protein in hyperoxia-exposed lungs. By analyzing the content of albumin in the lung, it was determined that this increase in E-GPx protein in the lungs of hyperoxia-exposed mice was not due to the amount of plasma in the lungs. In these studies, we could not verify that the lung E-GPx protein was enzymatically active, a problem inherent in Western analysis. Unfortunately, the antibodies used in these studies in mice are not able to immunoprecipitate the differing GPx activities. In addition, the substrate specificity of E-GPx and C-GPx, although somewhat different, cannot currently be used as a basis for differentiation of these activities in crude preparations. Thus, the only available means of quantifying the different amounts of GPx rely on antibodies and studies such as these.

A major question is the mechanism of the increase in E-GPx protein in the lungs of adult mice exposed to injurious amounts of oxygen. It has previously been shown that E-GPx mRNA is present in the lung of humans and mice (12, 29). In this study, the amounts of E-GPx, C-GPx, GAPD, and tubulin mRNA were determined only at the end of the exposure period, 72 h. Although there were hypoxia-associated increases in the amounts of E-GPx and C-GPx mRNA in the total RNA preparations at this time point, there was no significant change in the amount of E-GPx mRNA relative to tubulin or GAPD mRNA (two genes commonly used to normalize the amounts of genes under study) because the amounts of these genes had also increased. These results also must be cautiously interpreted because the relative amount E-GPx mRNA could have changed at an earlier time point. For example, an increase in E-GPx mRNA content at an earlier time could result in an increase in E-GPx protein translation to account for the increase in E-GPx protein observed at the 72-h time point. Alternatively, there could be a hyperoxia-associated increase in the stability of the E-GPx protein in lung in the absence of a change in the translation of E-GPx mRNA at earlier time points. Last, one needs to also include the possibility that the plasma E-GPx is preferentially sequestered in the lungs of hyperoxia-exposed mice to a greater extent than the increase in albumin (an index of the amount of plasma) in the lungs. Resolution of these issues will require further study.

Although most plasma E-GPx is derived from the kidney, the plasma E-GPx levels in anephric patients is 5–25% of the normal level (30). This indicates that E-GPx in plasma from anephric individuals can be derived from sources other than kidneys. On the basis of the prevalence of E-GPx mRNA in other tissues, one candidate for an additional source of E-GPx in plasma is the lung (10–12). Further investigations are needed to determine whether the increase in E-GPx in the plasma of hyperoxia-exposed mice is derived from kidney, lung, or other sources.

Induction of antioxidant enzymes by exposure to oxidant stimuli, including hyperoxia, has been shown by several authors. Van Golde et al. (31) demonstrated induction of superoxide dismutase (SOD), catalase, and C-GPx in chick eggs with 48 h of exposure to an FiO2 of 60%. In mammals, induction of antioxidant enzymes required initial hyperoxia exposure for 48 h, then 24 h at room air, followed by restimulation in hyperoxic conditions. Such treatment resulted in a 2-fold increase in the mRNA levels of Cu-Zn SOD, catalase, and C-GPx (32). The cell lineage that results in the increased production of antioxidant enzymes is unknown, but guinea pig alveolar macrophage exposed to hyperoxia (FiO2 > 95%) for 3 d resulted in a significant increase in GPx activity (33).

As yet, there is no evidence that the increase in the amount of E-GPx in the lungs of hyperoxia-exposed mice has a protective function, as this model was an injury model and not a tolerance model. Further studies will be required to determine whether the amount of E-GPx in the lung is also increased in animals exposed to lower FiO2. Experiments with transgenic or knockout mice might also illuminate whether E-GPx in the lung has a protective role against exposure to hyperoxia. In previous work, it has been shown that primary human bronchial epithelial cells and alveolar macrophages synthesize E-GPx mRNA and synthesize and secrete E-GPx protein in vitro (12). Further investigations will also be needed to identify the cells that are potentially involved in producing more E-GPx after hyperoxia exposure.

These studies were performed in adult mice aged 26 wk. A comprehensive study of the possible role of E-GPx in prevention or amelioration of hyperoxic injury in mice will necessarily include a comparison of neonatal, young, and old animals. Neonatal mice (34) and rats (35) are more resistant to hyperoxia-induced lung injury. Evaluation of the kinetics of E-GPx in these animals (transgenic and wild-type) may illuminate whether E-GPx in lung is related to the age-associated variability in hyperoxic lung injury.