Aberrant tissue repair and persistent inflammation following oxidant-mediated acute lung injury (ALI) can lead to the development and progression of various pulmonary diseases, but the mechanisms underlying these processes remain unclear. Hyperoxia is widely used in the treatment of pulmonary diseases, but the effects of this oxidant exposure in patients undergoing recovery from ALI are not clearly understood. Nrf2 has emerged as a crucial transcription factor that regulates oxidant stress through the induction of several detoxifying enzymes and other proteins. Using an experimental model of hyperoxia-induced ALI, we have examined the role of oxidant stress in resolving lung injury and inflammation. We found that when exposed to sublethal (72 h) hyperoxia, Nrf2-deficient, but not wild-type mice, succumbed to death during recovery. When both genotypes were exposed to a shorter period of hyperoxia-induced ALI (48 h), the lungs of Nrf2-deficient mice during recovery exhibited persistent cellular injury, impaired alveolar and endothelial cell regeneration, and persistent cellular infiltration by macrophages and lymphocytes. Glutathione (GSH) supplementation in Nrf2-deficient mice immediately after hyperoxia remarkably restored their ability to recover from hyperoxia-induced damage in a manner similar to that of wild-type mice. Thus, the results of the present study indicate that the Nrf2-regulated transcriptional response and, particularly GSH synthesis, is critical for lung tissue repair and the resolution of inflammation in vivo and suggests that a dysfunctional Nrf2-GSH pathway may compromise these processes in vivo.

Oxygen supplementation (also known as hyperoxia) is used to support critically ill patients with noninfectious and infectious acute lung injury (ALI)3 and its more severe form, acute respiratory distress syndrome, as well as emphysema. Acute exposure to hyperoxia (≤72 h) has been shown to induce lung inflammation and injury, leading to an impairment in respiratory function, whereas prolonged exposure (>96–120 h) causes lethality in rodents (1). Because of its similar pathologic features, hyperoxia exposure has been widely used as an experimental model of ALI/acute respiratory distress syndrome. Using this experimental system and positional cloning studies in inbred mice, we have previously identified Nrf2, a cap’n’collar basic leucine zipper transcription factor, as a candidate susceptibility gene for hyperoxia-induced ALI (HALI) (2). We have shown that Nrf2-deficient (Nrf2−/−) mice are more susceptible than wild-type (Nrf2+/+) mice to HALI (3). A deficiency in this transcription factor leads to diminished levels of both basal (3, 4) and inducible expression of several genes encoding enzymes and proteins that are critical for detoxifying the reactive oxygen and/or nitrogen species generated by hyperoxia (5).

In its constitutive state, Nrf2 is primarily localized in the cytosol; however, in response to pro-oxidant or oxidant exposure, it is translocated into the nucleus, where it up-regulates gene expression (see recent review, Ref. 6). This transcriptional induction by Nrf2 is principally mediated by the antioxidant response element (7). We have recently demonstrated an association of NRF2 promoter polymorphism with increased variation in ALI risk in a well-characterized clinical at-risk trauma group (8), suggesting that Nrf2 deficiency can enhance susceptibility to ALI. Recently, Arisawa et al. (9) have reported a significant association of NRF2 promoter polymorphisms with the development of gastric mucosal inflammation, either independently or through an interaction with Helicobacter pylori infection. These findings provide further support for the notion that these transcription factor promoter polymorphisms play a role in disease susceptibility. Consistent with this idea, the protective roles of Nrf2-regulated antioxidant enzymes (5), such as thioredoxin (10), peroxiredoxin (11), and Nqo1 (12), in the pathogenesis of HALI have been demonstrated in vivo using genetic models. Collectively, these studies suggest that an imbalance between pro-oxidant load and the antioxidant defense system could potentially enhance the lung tissue’s susceptibility to oxidant stress, thereby contributing to lung injury.

The resolution of lung injury and inflammation following pro-oxidant insult plays a prominent role in the restoration of normal lung structure and function. However, it is unclear why ALI completely resolves in some individuals, with restoration of normal lung structure and function, whereas in others, this syndrome leads to the development of progressive lung disease. Although the redox imbalance caused by hyperoxia has been implicated in the development of HALI (13), the exact roles of oxidant stress in regulating the resolution of lung injury and inflammation following hyperoxic insult remain unclear. We hypothesized that the ability of the Nrf2-regulated antioxidant transcriptional response to mitigate the redox imbalance caused by hyperoxia is critical for the effective resolution of HALI. In this study, we demonstrate impairment in the resolution of lung injury and inflammation in mice lacking Nrf2 and further report that glutathione (GSH) supplementation after hyperoxia exposure can rescue this defect in Nrf2−/− mice, suggesting a critical role for Nrf2-regulated GSH in resolving HALI.

The Nrf2-sufficient (Nrf2+/+) and Nrf2-deficient (Nrf2−/−) CD-1/ICR strains of mice (6- to 8-wk-old female mice, 25–30 g body weight) were exposed to hyperoxia or room air as previously described (2). After exposure, lung injury was assessed by alveolar permeability, whereas lung inflammation was evaluated by differential cell counts in bronchoalveolar lavage (BAL) fluid in the right lobes as previously described (2). Left lung lobes were inflated to 25 cm of water pressure and fixed with 0.8% low-melting agarose in 1.5% buffered paraformaldehyde for 24 h, and 5-μm lung sections were cut and stained with H&E. The remainder of the BAL was centrifuged and the supernatant was stored at −80°C. BAL protein concentration was measured by a Bio-Rad protein assay (catalog no. 500-0006). Differential cell counts were performed after staining the cells with a Diff-Quik stain kit (catalog no. B4132-1A; Dade Behring). All experimental protocols were approved by the animal care use committee of Johns Hopkins University.

Mice received in some experiments the antioxidant, GSH ester (5 mmol/kg body weight; Sigma-Aldrich), i.v. at every 24 h for 72 h or a similar volume of vehicle (PBS) immediately following hyperoxic insult.

For identification of DNA damage, TUNEL staining was performed using an In Situ Cell Death Detection Kit as per the manufacturer’s instructions. Briefly, the lung tissue sections were deparaffinized, washed with PBS, blocked with 3% H2O2 in methanol, permeabilized, and then incubated with 50 μl of TUNEL mixture for 1 h at 37°C. The sections were analyzed under fluorescent microscope and images were captured.

Deparaffinized lung tissue sections were permeabilized with 0.1% Triton X-100 and blocked with 5% BSA and 1% appropriate serum. The sections were permeabilized and incubated with anti-surfactin protein C (SPC) (catalog no. WRAB-SPC; Seven Hills Bioreagents), anti-CD34 (catalog no. 16-034; eBioscience), anti-Ki-67 (catalog no. RM 9106; Thermo Fisher Scientific), or anti-CD68 (catalog no. SC-9139; Santa Cruz Biotechnology) Abs overnight at 4°C. The sections were washed five times in PBS and incubated with green-fluorescent Alexa Fluor 488 donkey anti-rabbit IgG Ab (catalog no. A21206; Invitrogen) or red-fluorescent Alexa Fluor 594 rabbit anti-mouse IgG Ab (catalog no. A11029; Invitrogen) for appropriate time periods. Finally, the slides were washed in PBS and mounted using 4′,6-diamidino-2-phenylindole, immunostaining was observed using a fluorescence microscope (Nikon Eclipse TE 2000-S), and images were captured.

Total GSH levels were measured using a total GSH detection kit (catalog no. 900-169; Assay Designs) per the manufacturer’s instructions. Briefly, 5% homogenates of lung tissues were prepared in metaphosphoric acid. To 5 μl of homogenate, freshly prepared assay reaction mixture was added and the absorbance was measured immediately using a plate reader at 405 nm at 1-min intervals over a 10-min period. The total GSH levels assayed in triplicate were calculated using the standard graph.

The expression levels of various genes in the lungs of mice exposed to RA or hyperoxia and 72-h recovery were quantified in triplicate by TaqMan gene expression assays (Applied Biosystems) using Gapdh and mitochondrial ribosomal protein L32 (Mrpl32) (n = 3–4/group) as internal control genes. The absolute expression values for each gene were normalized to that of Gapdh/Mrpl32 and values from room air samples set as one unit.

All data involving animal experimentation were collected in a double-blinded fashion. Data are expressed as the mean ± SD (n = 3–5 for each condition). ANOVA was used to compare means of multiple groups. For paired data, Student’s t test was used. Significance in all cases was defined as p < 0.05.

Continuous exposure to >90% hyperoxia for 4–7 days results in lethality in various strains of mice, including CD1/ICR (1). To examine the role of oxidative stress modifier Nrf2 in the resolution of HALI, we exposed wild-type (Nrf2+/+) and Nrf2-deficient (Nrf2−/−) mice to hyperoxia for 72 h and then allowed the animals to recover under normoxia for various periods of time. A striking difference in the survival rate was observed between the two genotypes (Fig. 1,A). Nrf2−/− mice that were exposed to hyperoxia and then allowed to recover in room air died within 6 h after restoration of normoxia. In contrast, 20% of Nrf2+/+ mice died in the same period, while the remaining Nrf2+/+ mice survived for the entire 80-h recovery period after hyperoxic insult (Fig. 1,A). Lung histology revealed greater levels of hemorrhage and perivascular/peribronchiolar edema in the lung tissue of Nrf2−/− mice than those of wild-type Nrf2+/+ mice (see supplemental Fig. S14). Consistent with our earlier results, we found a significant increase in hyperoxia-induced alveolar permeability (Fig. 1,B) and in the sloughing of lung epithelial cells (Fig. 1,D) but not total inflammatory cell accumulation (Fig. 1 C) in the BAL fluid of Nrf2−/− mice, when compared with similarly exposed Nrf2+/+ mice. These results suggest that the Nrf2-regulated transcriptional response is critical for survival in recovery after HALI.

FIGURE 1.

Effects of sublethal hyperoxia on Nrf2−/− mice in recovery. A, Survival time for Nrf2+/+ and Nrf2−/− mice (n = 16/genotype) exposed to 72- hyperoxia. Mice were allowed to recover under normoxia for various time periods. B, Total protein in the BAL fluid of room air- and hyperoxia-exposed Nrf2+/+ and Nrf2−/− mice. C, Total inflammatory cells in the BAL fluid of room air- and hyperoxia-exposed Nrf2+/+ and Nrf2−/− mice. D, Total epithelial cells in the BAL fluid of room air- and hyperoxia-exposed Nrf2+/+ and Nrf2−/− mice. Data in B–D represent the mean values of five mice with SD. *, p < 0.001 vs room air control of the same genotype; #, p < 0.001, Nrf2−/− mice vs Nrf2+/+ mice subjected to hyperoxia. □, The room air controls and ▪, hyperoxia.

FIGURE 1.

Effects of sublethal hyperoxia on Nrf2−/− mice in recovery. A, Survival time for Nrf2+/+ and Nrf2−/− mice (n = 16/genotype) exposed to 72- hyperoxia. Mice were allowed to recover under normoxia for various time periods. B, Total protein in the BAL fluid of room air- and hyperoxia-exposed Nrf2+/+ and Nrf2−/− mice. C, Total inflammatory cells in the BAL fluid of room air- and hyperoxia-exposed Nrf2+/+ and Nrf2−/− mice. D, Total epithelial cells in the BAL fluid of room air- and hyperoxia-exposed Nrf2+/+ and Nrf2−/− mice. Data in B–D represent the mean values of five mice with SD. *, p < 0.001 vs room air control of the same genotype; #, p < 0.001, Nrf2−/− mice vs Nrf2+/+ mice subjected to hyperoxia. □, The room air controls and ▪, hyperoxia.

Close modal

To define the role of Nrf2 in the resolution of HALI, we exposed mice to hyperoxia for a shorter (48-h) period of time and then assessed lung injury and inflammation during the recovery period. Unlike the lethality that was observed after a 72-h exposure, we found that Nrf2−/− as well as Nrf2+/+ mice that had been exposed to hyperoxia for only 48 h were able to survive for 21 days when allowed to recover in room air (Fig. 2,A). No striking differences in lung histopathology were seen between the Nrf2−/− and Nrf2+/+ mice immediately after 48 h of hyperoxia (see supplemental Fig. S2). BAL fluid analysis revealed that the degree of hyperoxia-induced damage in terms of lung alveolar permeability (Fig. 2,B), total inflammatory cellular infiltration (Fig. 2,C), and epithelial cell sloughing (Fig. 2 D) in the Nrf2−/− mice exposed to hyperoxia for 48 h was not significantly different from that of the Nrf2+/+ mice. Consistent with these results, we found no significant difference between the wet:dry ratios for the lungs of Nrf2−/− and Nrf2+/+ mice after 48 h of hyperoxia (data not shown).

FIGURE 2.

Effects of shortened sublethal hyperoxia on Nrf2−/− mice in recovery. A, Survival time for Nrf2+/+ and Nrf2−/− mice (n = 16/genotype) exposed to 48-h hyperoxia. Mice were allowed to recover under normoxia for various time periods. All of the Nrf2+/+ and Nrf2−/− mice were survived through 21 days of recovery. Total protein (B), inflammatory cells (C), and epithelial cells (D) in the BAL fluid obtained room air- and hyperoxia-exposed Nrf2+/+ and Nrf2−/− mice. Graphs represent the mean values of five mice with SD. *, p < 0.001 vs room air control of the same genotype. □, The room air controls and ▪, hyperoxia.

FIGURE 2.

Effects of shortened sublethal hyperoxia on Nrf2−/− mice in recovery. A, Survival time for Nrf2+/+ and Nrf2−/− mice (n = 16/genotype) exposed to 48-h hyperoxia. Mice were allowed to recover under normoxia for various time periods. All of the Nrf2+/+ and Nrf2−/− mice were survived through 21 days of recovery. Total protein (B), inflammatory cells (C), and epithelial cells (D) in the BAL fluid obtained room air- and hyperoxia-exposed Nrf2+/+ and Nrf2−/− mice. Graphs represent the mean values of five mice with SD. *, p < 0.001 vs room air control of the same genotype. □, The room air controls and ▪, hyperoxia.

Close modal

Although 48 h of hyperoxia produced a comparable degree of HALI in the Nrf2−/− and Nrf2+/+ mice, we saw striking differences in lung histology and alveolar permeability and also in the inflammatory responses in the BAL fluid between these two mouse strains during the recovery period (Fig. 3). Histopathologic analysis revealed that hemorrhage and cellular infiltration into the alveolar space persisted in the Nrf2−/− mice during the first 7 days of recovery, with a peak at 72 h. In contrast, the lung hemorrhage subsided in the Nrf2+/+ mice as early as 24 h after the beginning of the recovery period (Fig. 3,A). BAL fluid analysis revealed elevated levels of protein and a cellular infiltration of lymphocytes and macrophages, along with an increase in the level of epithelial cell sloughing in Nrf2−/− mice after 72 h of recovery, when compared with the corresponding Nrf2+/+ mice. The protein concentrations and epithelial cell sloughing were noticeably decreased in the Nrf2−/− mice at 7 and 21 days, respectively (Fig. 3,B). However, consistent with the histologic observations in the lungs, we found a persistent cellular infiltrate composed of lymphocytes and macrophages in the Nrf2−/− mice through 21 days of recovery (Fig. 3,B). The levels of inflammatory cytokines, IL-6 and CXCL2, in the lung tissues of mice exposed to hyperoxia were elevated compared with room air-exposed mice. However, we found significantly greater levels of IL-6 and CXCL2 in Nrf2−/− mice compared with Nrf2+/+ mice exposed to hyperoxia. The increased levels of IL-6 and CXCL2 were sustained in Nrf2−/− mice during the 3-day recovery, whereas the levels of these cytokines were subsided in Nrf2+/+ mice (Fig. 3 C).These results suggest that Nrf2 is critical for the effective resolution of lung injury and inflammation following hyperoxia.

FIGURE 3.

Effect of hyperoxia on lung histology and inflammatory responses in Nrf2−/− mice after injury. Mice were exposed to 100% oxygen for 48 h and then allowed to recover up to 21 days. Nrf2+/+ and Nrf2−/− mice lungs were harvested immediately after hyperoxia and at the 24-h, 72-h, and 7-day, and 21-day recovery and the right lobe was used for BAL collection while the left lobe was inflated and fixed as described above. A, Lung histology at 48-h exposure and at the 24-h, 72-h, 7-day, and 21-day during recovery. B, Graphs represent the total protein, total cells, neutrophils, macrophages, lymphocytes, and epithelial cells in the BAL fluid of Nrf2+/+ and Nrf2−/− mice. C, Expression levels of IL-6 and CXCL2 as assayed by using quantitative RT-PCR. Data are mean with SD, n = 3–4/each experimental group. *, p < 0.001 vs room air control of the same genotype; #, p ≤ 0.001, Nrf2−/− mice vs Nrf2+/+ mice subjected to hyperoxia and recovery.

FIGURE 3.

Effect of hyperoxia on lung histology and inflammatory responses in Nrf2−/− mice after injury. Mice were exposed to 100% oxygen for 48 h and then allowed to recover up to 21 days. Nrf2+/+ and Nrf2−/− mice lungs were harvested immediately after hyperoxia and at the 24-h, 72-h, and 7-day, and 21-day recovery and the right lobe was used for BAL collection while the left lobe was inflated and fixed as described above. A, Lung histology at 48-h exposure and at the 24-h, 72-h, 7-day, and 21-day during recovery. B, Graphs represent the total protein, total cells, neutrophils, macrophages, lymphocytes, and epithelial cells in the BAL fluid of Nrf2+/+ and Nrf2−/− mice. C, Expression levels of IL-6 and CXCL2 as assayed by using quantitative RT-PCR. Data are mean with SD, n = 3–4/each experimental group. *, p < 0.001 vs room air control of the same genotype; #, p ≤ 0.001, Nrf2−/− mice vs Nrf2+/+ mice subjected to hyperoxia and recovery.

Close modal

The alveolar and endothelial cells of the lung have been reported to undergo death during hyperoxia exposure, but during recovery, they regenerate to restore normal lung structure and function. To determine whether this process has been compromised in mice lacking Nrf2, we performed immunohistochemical analyses of the lung tissues obtained from mice after exposure to hyperoxia and during recovery (Fig. 4). Immunostaining of lung sections with an anti-SPC Ab revealed low levels of type II cells in the lungs of hyperoxia-exposed Nrf2−/− and Nrf2+/+ mice (Fig. 4,A). However, we found that the levels of anti-SPC Ab staining in wild-type mice after 72 h and 7 days of recovery were nearly comparable to those of room air-exposed control mice. In contrast, only weak staining was observed in the Nrf2−/− mice after 72 h and 7 days of recovery (Fig. 4,A, bottom panel). We also assessed the status of endothelial cells using an anti-CD34 Ab. As was observed for SPC, only weak CD34 staining was seen in the lungs of Nrf2−/− and Nrf2+/+ hyperoxia-exposed mice at 0 h of recovery, but endothelial cell staining for CD34 was noticeable in Nrf2+/+ mice after 72 h of recovery and was comparable to that of the room air-control group (Fig. 4,B, top panel). In Nrf2−/− mice, the level of CD34 staining was noticeably lower than that of the corresponding Nrf2+/+ mice after 72 h of recovery. However, after 7 days of recovery, the CD34 staining in the Nrf2−/− mice was stronger than at 72 h and was comparable to that of the room air-exposed control group (Fig. 4 B, bottom panel).

FIGURE 4.

Effect of Nrf2 deficiency on type II epithelial and endothelial cell regeneration and cell proliferation after injury. Lung tissues from the Nrf2+/+ and Nrf2−/− mice exposed to room air and 48-h hyperoxia, 72-h and 7-day recovered, were inflated and fixed. Lung sections were immunostained using anti-SPC (A) and anti-CD34 (B), anti-CD68 (C), and anti-Ki-67 (D) Abs to stain type II epithelium, endothelium, macrophages, and proliferating cells, respectively. The positively stained cells are indicated with arrows. Images shown are representative lung sections from three different mice in each group.

FIGURE 4.

Effect of Nrf2 deficiency on type II epithelial and endothelial cell regeneration and cell proliferation after injury. Lung tissues from the Nrf2+/+ and Nrf2−/− mice exposed to room air and 48-h hyperoxia, 72-h and 7-day recovered, were inflated and fixed. Lung sections were immunostained using anti-SPC (A) and anti-CD34 (B), anti-CD68 (C), and anti-Ki-67 (D) Abs to stain type II epithelium, endothelium, macrophages, and proliferating cells, respectively. The positively stained cells are indicated with arrows. Images shown are representative lung sections from three different mice in each group.

Close modal

In contrast to the diminished levels of type II and endothelial cell staining, we observed elevated levels of macrophage infiltration, as indicated by immunostaining analysis with anti-CD68 Ab in the lungs of Nrf2-deficient mice during recovery (Fig. 4,C). The macrophage staining was barely detectable in the Nrf2+/+ mice. These results are consistent with the BAL fluid analysis, which revealed elevated levels of macrophages and lymphocytes in the lungs of Nrf2−/− mice during recovery (Fig. 3,B). We next examined the effect of Nrf2 deficiency on lung cell proliferation. Lung tissue sections were immunostained with Abs specific for Ki-67, a marker of cell proliferation. As shown in Fig. 4,D, the Ki-67 staining was very weak or undetectable in the lung tissues of Nrf2−/− and Nrf2+/+ mice exposed to room air and hyperoxia. However, we observed increased levels of cell proliferation, as indicated by immunostaining analysis with anti- Ki-67 Ab, in the lungs of wild-type mice compared with Nrf2-deficient mice during recovery (Fig. 4 D).These results suggest that the Nrf2-dependent transcriptional response is critical for regeneration of alveolar epithelium and endothelium during recovery from hyperoxia.

Because the generation of reactive oxygen/nitrogen species has been identified as a mechanism that contributes to the alveolar cell damage and impairment of endothelial and epithelial cell regeneration after hyperoxia (14, 15), we assessed the extent of the cellular injury and the expression levels of DNA damage and repair signaling molecules, p53 and p21, in the lung tissues of hyperoxia-exposed and room air-recovered mice by TUNEL staining and by real-time PCR analysis, respectively (Fig. 5). Although the number of TUNEL-positive cells was essentially equivalent in hyperoxia-exposed Nrf2−/− and Nrf2+/+ mice immediately after exposure to hyperoxia, the numbers gradually declined in the Nrf2+/+ mice after 24 and 72 h of recovery and had decreased to undetectable levels after 7 days of recovery (Fig. 5,A and see supplemental Fig. S3 for fluorescent images). In contrast, we saw persistent cellular damage in the Nrf2−/− mice through day 7 of recovery, although the level of damage was lower after both 72 h and 7 days of recovery than immediately after hyperoxia (Fig. 5,B). The expression levels of p53 and p21 were elevated by 5- and 25-fold, respectively, in Nrf2+/+ mice after hyperoxia exposure and remained elevated at 72 h of recovery compared with room air-exposed lung tissues. In contrast, the induction of p53 and p21 expression was not detectable in Nrf2−/− mice after hyperoxia and recovery through 7 days (Fig. 5 C).

FIGURE 5.

Lung cellular injury of Nrf2+/+ and Nrf2−/− mice exposed to hyperoxia and during recovery. A, TUNEL staining was performed using an In Situ Cell Death Detection Kit (Roche) on lung tissues of Nrf2+/+ and Nrf2−/− mice exposed to hyperoxia and allowed to recover. B, Graph represents the percentage of TUNEL-positive cells of five fields per mice for each group (n = 5). C, Graphs represent the relative expression levels of p53 and p21 (n = 3). *, p ≤ 0.001, room air vs experimental groups of same genotype; #, p ≤ 0.01, Nrf2−/− mice vs Nrf2+/+ mice in recovery.

FIGURE 5.

Lung cellular injury of Nrf2+/+ and Nrf2−/− mice exposed to hyperoxia and during recovery. A, TUNEL staining was performed using an In Situ Cell Death Detection Kit (Roche) on lung tissues of Nrf2+/+ and Nrf2−/− mice exposed to hyperoxia and allowed to recover. B, Graph represents the percentage of TUNEL-positive cells of five fields per mice for each group (n = 5). C, Graphs represent the relative expression levels of p53 and p21 (n = 3). *, p ≤ 0.001, room air vs experimental groups of same genotype; #, p ≤ 0.01, Nrf2−/− mice vs Nrf2+/+ mice in recovery.

Close modal

Disruption of Nrf2 leads to a redox imbalance as a result of a decrease in, or lack of, antioxidant enzyme expression. To determine whether diminished levels of antioxidant gene induction could contribute to dysfunctional lung repair following exposure to hyperoxia, we assessed Gclc and Gpx2 gene induction as well as the levels of GSH in the lungs of Nrf2−/− and Nrf2+/+ mice immediately after hyperoxia exposure and during recovery. These two enzymes were chosen because they are prototypical targets of Nrf2 and regulate GSH (5). As shown in Fig. 6, we found that the level of hyperoxia-induced Gclc and Gpx2 expression was severalfold greater in the Nrf2+/+ than in the Nrf2−/− mice. Furthermore, although the induction of Gclc decreased during recovery, it remained elevated above basal levels throughout the 72-h recovery period in these mice. On the other hand, we saw no apparent induction of Gclc or Gpx2 expression in the lungs of the Nrf2−/− mice. These differences in the level of Gclc mRNA expression in the lungs of Nrf2+/+ and Nrf2−/− mice were also confirmed by Western blot analysis (Fig. 6 B). Consistent with Gclc expression, the GSH levels were induced in Nrf2+/+ mice after hyperoxia and remained at the levels of room air-exposed control mice after the 3-day recovery. On the contrary, the GSH levels were decreased in Nrf2−/− mice after hyperoxia and reached to room air control levels after the 3-day recovery.

FIGURE 6.

The expression levels of Nrf2 target genes in Nrf2+/+ and Nrf2−/− mice. cDNA was prepared from total RNA isolated from the lungs of Nrf2+/+ and Nrf2−/− mice exposed to 48-h hyperoxia and after recovery at day 3. A, The expression levels of Gclc and Nqo1 were determined using TaqMan real-time probes. The values represented are mean with SD. B, Western blot analysis of Gclc in the protein lysates of lungs of Nrf2+/+ and Nrf2−/− mice exposed to 48-h hyperoxia and after injury at day 3. C, Total GSH levels in the lungs of Nrf2+/+ and Nrf2−/− mice exposed to room air, hyperoxia, and after recovery. The values represent the mean and SD of triplicates of three mice per group. #, p < 0.01, Nrf2−/− mice vs Nrf2+/+ mice in recovery.

FIGURE 6.

The expression levels of Nrf2 target genes in Nrf2+/+ and Nrf2−/− mice. cDNA was prepared from total RNA isolated from the lungs of Nrf2+/+ and Nrf2−/− mice exposed to 48-h hyperoxia and after recovery at day 3. A, The expression levels of Gclc and Nqo1 were determined using TaqMan real-time probes. The values represented are mean with SD. B, Western blot analysis of Gclc in the protein lysates of lungs of Nrf2+/+ and Nrf2−/− mice exposed to 48-h hyperoxia and after injury at day 3. C, Total GSH levels in the lungs of Nrf2+/+ and Nrf2−/− mice exposed to room air, hyperoxia, and after recovery. The values represent the mean and SD of triplicates of three mice per group. #, p < 0.01, Nrf2−/− mice vs Nrf2+/+ mice in recovery.

Close modal

We have previously shown that type II cells lacking the Nrf2 gene undergo cellular stress and proliferate poorly in vitro (16, 17); however, GSH supplementation can rescue the Nrf2-deficient cells and overcome this phenotypic defect in vitro. We have now observed a similar phenotype in freshly cultured endothelial cells isolated from the lungs of Nrf2−/− mice (data not shown). To confirm that Nrf2-regulated GSH signaling is critical for the resolution of lung injury in vivo, we administered exogenous antioxidant (GSH ester, 5 mmol/kg body weight every 24 h) to Nrf2−/− mice immediately after 48 h of hyperoxia and then allowed the mice to recover in room air (Fig. 7). The GSH ester was used for these experiments instead of N-acetyl-l-cysteine because Nrf2 deficiency is associated with decreased levels of Gclc, a key enzyme required for the conversion of N-acetyl-l-cysteine to GSH (see Fig. 6). PBS was used as vehicle control. Lung inflammation and injury were analyzed after 72 h of recovery as described above. As expected, GSH supplementation attenuated the lung injury and inflammatory cell accumulation in the lungs of Nrf2−/− mice (Fig. 7,A). Histopathologic examination of the lungs revealed diminished levels of cellular infiltration in the GSH-supplemented mice as compared with the vehicle-treated mice, and immunostaining with the anti-SPC Abs and anti-CD34 Abs (data not shown) revealed a restoration of both the alveolar epithelium and the endothelium in the GSH-treated mice but not the vehicle-treated mice (Fig. 7,B). Differential cell counts in the BAL fluid of the GSH-treated Nrf2−/− mice revealed that the levels of total inflammatory cells, macrophages, and lymphocytes had decreased to levels similar to those of room air-exposed controls (Fig. 7,C). We have measured GSH levels in the lungs of Nrf2−/− mice during 24 and 72 h recovery following GSH ester administration. Indeed, GSH supplementation significantly increased GSH levels in the lungs of Nrf2−/− mice after the 24- and 72-h recovery (Fig. 7 D).

FIGURE 7.

Effects of exogenous GSH on hyperoxia-induced lung inflammation and injury: Nrf2−/− mice immediately after 48-h hyperoxia were given GSH ester (5 mmol/kg body weight) or vehicle PBS at every 24 h for 3 days and lungs were harvested. Lungs sections and BAL analyses were done as described above. A, Histology of lungs isolated from mice supplemented with PBS or GSH ester at 24 and 72 h (top panel) in recovery. B, Lung sections were immunostained using anti-SPC Abs to visualize the presence of type II epithelial cells. The positively stained cells are indicated with arrows. Images shown are representative of lung section from three different mice in each group. C, Graphs represent the total protein, total cells, neutrophils, macrophages, lymphocytes and epithelial cells in the BAL fluid. D, Graph represents the GSH levels in the lungs of Nrf2−/− mice at 24- and 72-h recovery following GSH supplementation *, p < 0.001 vs Nrf2−/− GSH-supplemented mice (n = 3).

FIGURE 7.

Effects of exogenous GSH on hyperoxia-induced lung inflammation and injury: Nrf2−/− mice immediately after 48-h hyperoxia were given GSH ester (5 mmol/kg body weight) or vehicle PBS at every 24 h for 3 days and lungs were harvested. Lungs sections and BAL analyses were done as described above. A, Histology of lungs isolated from mice supplemented with PBS or GSH ester at 24 and 72 h (top panel) in recovery. B, Lung sections were immunostained using anti-SPC Abs to visualize the presence of type II epithelial cells. The positively stained cells are indicated with arrows. Images shown are representative of lung section from three different mice in each group. C, Graphs represent the total protein, total cells, neutrophils, macrophages, lymphocytes and epithelial cells in the BAL fluid. D, Graph represents the GSH levels in the lungs of Nrf2−/− mice at 24- and 72-h recovery following GSH supplementation *, p < 0.001 vs Nrf2−/− GSH-supplemented mice (n = 3).

Close modal

Abnormal tissue repair and inflammation following oxidant- or toxicant-mediated injury can contribute to the development and progression of various lung diseases (18, 19). ALI produced by a relatively shortened (48–72 h) period of hyperoxia exposure was used as a model to investigate the mechanisms controlling lung injury, repair, and inflammation (20, 21, 22). In the present study, we have used a similar strategy (a sublethal 48-h exposure) to examine the role of oxidant stress in the resolution of ALI using mice deficient in the Nrf2 transcription factor which regulates cellular stress. We found that Nrf2 deficiency was associated with impaired alveolar epithelium and endothelium regeneration during recovery, and this phenotype was associated as well with persistent cellular damage and increased cellularity comprised of macrophage and lymphocyte infiltration. However, administration of GSH immediately after hyperoxia rescued these phenotypes and was able to prevent this hyperoxia-associated damage in these Nrf2−/− mice, demonstrating that Nrf2-regulated GSH synthesis can counteract the hyperoxia-induced oxidative stress, which would otherwise impair the resolution of the repair process and inflammation during recovery.

Various studies have shown that exposure to hyperoxia for 60–72 h causes the death of lung alveolar epithelial and endothelial cells, with death accompanied by vascular and alveolar permeability and inflammatory cellular infiltration (see recent reviews, Refs. 23 and 24). Regeneration of these cell types after lung injury is critical for the restoration of normal lung structure and function (25, 26). However, we have observed impairment of regeneration of the alveolar epithelium and endothelium in Nrf2−/− mice, but not in wild-type Nrf2+/+ mice during recovery from HALI. TUNEL staining revealed that hyperoxia-induced DNA damage in wild-type mice was repaired quickly, while Nrf2 deficiency led to persistent DNA injury during the recovery period (Fig. 3). It has been reported that hyperoxia- induced DNA damage promotes growth arrest in lung epithelial cells both in vitro and in vivo (27, 28, 29). We have previously shown that in response to hyperoxic insult, Nrf2 up-regulates the expression of genes encoding several antioxidant enzymes and proteins, including the Gclc and Gclm, which are required for GSH biosynthesis (5). Consistent with these results, we found increased levels of Gclc expression and GSH levels in wild-type (Nrf2+/+) mice exposed to hyperoxia and after recovery, but not in the lungs of Nrf2−/− mice. Thus, it is likely that the persistent DNA injury observed in the absence of Nrf2-regulated GSH biosynthesis may dampen interference with the repair of lung alveolar and endothelial cells in Nrf2-deficient mice.

We have recently demonstrated that freshly cultured Nrf2-deficient type II lung epithelial cells exhibit cellular stress and proliferate poorly due to the deregulation of cell cycle progression (16, 17). Although we found that Nrf2 regulates the induction of several cellular detoxifying enzymes and proteins, an Nrf2 deficiency was associated with an increase in the expression levels of genes involved in cell cycle check point regulation, cytokinesis, repair of DNA damage, and repair and centrosome duplication, as well as several growth factors and their receptors that are involved in cell proliferation (30). GSH supplementation alone was able to rescue the proliferative defect in these Nrf2−/− cells (16, 17) and also restored the expression of several genes that control cell cycle progression (30). Although these cell culture studies suggest that Nrf2-regulated, GSH-induced redox signaling plays an essential role in regulating cell cycle progression, the exact mechanisms by which Nrf2 deficiency dampens interference with the repair of alveolar epithelium and endothelium following hyperoxic insult, as well as the means by which GSH restores this defect in vivo, remains unclear and warrants further study.

Macrophages regulate both the propagation and resolution of inflammation (31, 32). The accumulation of macrophages in the lung tissue can lead to enhanced levels of inflammatory cytokines, which play fundamental roles in the development of lung pathogenesis, including the development of ALI (31, 32, 33, 34). For example, activated macrophages upon activation release various cytokines such as TNF-α and matrix metalloproteinases, which are known to regulate lung inflammation and tissue remodeling (18, 34, 35). We found no change in the number of macrophages in either the BAL fluid and/or lung tissues of Nrf2+/+ mice immediately after hyperoxia or during recovery, as compared with room air-exposed control. The number of macrophages in the BAL fluid obtained from Nrf2-deficient mice immediately after hyperoxic insult was comparable to that of wild- type mice. However, we found a notable increase in the number of macrophages present in the BAL fluid (Fig. 3,B) and in the interstitium (Fig. 4,C) in Nrf2-deficient mice during recovery. These results suggest that the Nrf2-dependent transcriptional response may limit the inflammatory responses induced by macrophages (Fig. 1,B). Macrophages are critical for the clearance of granulocytes and the damaged cellular organelles from dead cells (32). Thus, it is possible that the elevated levels of macrophages in Nrf2−/− mice may play a role in scavenging the oxidized or degraded cellular material resulting from persistent cellular injury present in these mice. We also found elevated levels of lymphocytes, which play key roles in regulating immune function, in the lungs of Nrf2-deficient mice but not in wild-type mice during recovery (Fig. 3 B). The exact nature and specific characteristics of these macrophages and lymphocytes present in the lungs of Nrf2−/− mice, such as their activation status and function, and whether they play a role in enhancing susceptibility to bacterial or viral infection remain to be investigated.

In summary, our studies have for the first time demonstrated that Nrf2 regulates the resolution of HALI by modulating GSH levels. Since promoter polymorphisms in this transcription factor are associated with an increased risk of ALI susceptibility (8), our current studies invoke the possibility that deregulation or variation in Nrf2- induced GSH synthesis might contribute to abnormal repair and inflammation following oxidant-mediated lung injury in susceptible populations. Analyzing the Nrf2/GSH-regulated pathways modulated by hyperoxia in vivo may provide additional insights into the factors that either promote or perpetuate the resolution of lung injury and inflammation and may help contribute to the development of novel therapies targeted at progressive lung diseases associated with abnormal remodeling and repair.

We thank the Pathology Core of the ALI Specialized Centers of Clinically Oriented Research for assisting in immunohistochemical and histopathological analysis. We thank Terrance Kavanagh (University of Washington) for providing us with Gclc Abs used in the study.

The authors have no financial conflict of interest.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1

This work was funded by National Institutes of Health Grants HL66109 and ES11863 (to SPR) and Specialized Centers of Clinically Oriented Research Grants P50 HL073994 (to S.P.R. and P.H.), HL049441 (to P.H.), and CA 94076 (to T.W.K).

3

Abbreviations used in this paper: ALI, acute lung injury; HALI, hyperoxia-induced ALI; GSH, glutathione; BAL, broncho alveolar lavage; SPC, surfactin protein C.

4

The online version of this article contains supplemental material.

1
Crapo, J. D..
1986
. Morphologic changes in pulmonary oxygen toxicity.
Annu. Rev. Physiol.
48
:
721
-731.
2
Cho, H. Y., A. E. Jedlicka, S. P. Reddy, L. Y. Zhang, T. W. Kensler, S. R. Kleeberger.
2002
. Linkage analysis of susceptibility to hyperoxia. Nrf2 is a candidate gene.
Am. J. Respir. Cell Mol. Biol.
26
:
42
-51.
3
Cho, H.-Y., A. E. Jedlicka, S. P. M. Reddy, T. W. Kensler, M. Yamamoto, L.-Y. Zhang, S. R. Kleeberger.
2002
. Role of NRF2 in protection against hyperoxic lung injury in mice.
Am. J. Respir. Cell Mol. Biol.
26
:
175
-182.
4
Cho, H. Y., S. P. Reddy, M. Yamamoto, S. R. Kleeberger.
2004
. The transcription factor NRF2 protects against pulmonary fibrosis.
FASEB J.
18
:
1258
-1260.
5
Cho, H. Y., S. P. Reddy, S. R. Kleeberger.
2006
. Nrf2 defends the lung from oxidative stress.
Antioxid. Redox Signal.
8
:
76
-87.
6
Kensler, T. W., N. Wakabayashi, S. Biswal.
2006
. Cell survival responses to environmental stresses via the Keap1-Nrf2-ARE pathway.
Annu. Rev. Pharmacol. Toxicol.
47
:
89
-116.
7
Nguyen, T., P. J. Sherratt, C. B. Pickett.
2003
. Regulatory mechanisms controlling gene expression mediated by the antioxidant response element.
Annu. Rev. Pharmacol. Toxicol.
43
:
233
-260.
8
Marzec, J. M., J. D. Christie, S. P. Reddy, A. E. Jedlicka, H. Vuong, P. N. Lanken, R. Aplenc, T. Yamamoto, M. Yamamoto, H. Y. Cho, S. R. Kleeberger.
2007
. Functional polymorphisms in the transcription factor NRF2 in humans increase the risk of acute lung injury.
FASEB J.
21
:
2237
-2246.
9
Arisawa, T., T. Tahara, T. Shibata, M. Nagasaka, M. Nakamura, Y. Kamiya, H. Fujita, S. Hasegawa, T. Takagi, F. Y. Wang, et al
2007
. The relationship between Helicobacter pylori infection and promoter polymorphism of the Nrf2 gene in chronic gastritis.
Int. J. Mol. Med.
19
:
143
-148.
10
Yamada, T., Y. Iwasaki, K. Nagata, S. Fushiki, H. Nakamura, Y. Marunaka, J. Yodoi.
2007
. Thioredoxin-1 protects against hyperoxia-induced apoptosis in cells of the alveolar walls.
Pulm. Pharmacol. Ther.
20
:
650
-659.
11
Wang, Y., S. I. Feinstein, Y. Manevich, Y. S. Ho, A. B. Fisher.
2006
. Peroxiredoxin 6 gene-targeted mice show increased lung injury with paraquat-induced oxidative stress.
Antioxid. Redox Signal.
8
:
229
-237.
12
Das, A., L. Kole, L. Wang, R. Barrios, B. Moorthy, A. K. Jaiswal.
2006
. BALT development and augmentation of hyperoxic lung injury in mice deficient in NQO1 and NQO2.
Free Radical Biol. Med.
40
:
1843
-1856.
13
MacNee, W..
2001
. Oxidants/antioxidants and chronic obstructive pulmonary disease: pathogenesis to therapy.
Novartis Found. Symp.
234
:
169
-185.
14
O'Reilly, M. A..
2001
. DNA damage and cell cycle checkpoints in hyperoxic lung injury: braking to facilitate repair.
Am. J. Physiol.
281
:
L291
-L305.
15
Comhair, S. A., S. C. Erzurum.
2002
. Antioxidant responses to oxidant-mediated lung diseases.
Am. J. Physiol.
283
:
L246
-L255.
16
Reddy, N. M., S. R. Kleeberger, H. Y. Cho, M. Yamamoto, T. W. Kensler, S. Biswal, S. P. Reddy.
2007
. Deficiency in Nrf2-GSH signaling impairs type II cell growth and enhances sensitivity to oxidants.
Am. J. Respir. Cell Mol. Biol.
37
:
3
-8.
17
Reddy, N. M., S. R. Kleeberger, J. H. Bream, P. G. Fallon, T. W. Kensler, M. Yamamoto, S. P. Reddy.
2008
. Genetic disruption of the Nrf2 compromises cell-cycle progression by impairing GSH-induced redox signaling.
Oncogene
27
:
5821
-5832.
18
Noble, P. W., D. Jiang.
2006
. Matrix regulation of lung injury, inflammation, and repair: the role of innate immunity.
Proc. Am. Thorac. Soc.
3
:
401
-404.
19
Puchelle, E., J. M. Zahm, J. M. Tournier, C. Coraux.
2006
. Airway epithelial repair, regeneration, and remodeling after injury in chronic obstructive pulmonary disease.
Proc. Am. Thorac. Soc.
3
:
726
-733.
20
Staversky, R. J., R. H. Watkins, T. W. Wright, E. Hernady, M. B. LoMonaco, C. T. D'Angio, J. P. Williams, W. M. Maniscalco, M. A. O'Reilly.
2002
. Normal remodeling of the oxygen-injured lung requires the cyclin-dependent kinase inhibitor p21Cip1/WAF1/Sdi1.
Am. J. Pathol.
161
:
1383
-1393.
21
Baleeiro, C. E., S. E. Wilcoxen, S. B. Morris, T. J. Standiford, R. Paine, III.
2003
. Sublethal hyperoxia impairs pulmonary innate immunity.
J. Immunol.
171
:
955
-963.
22
Lee, J., R. Reddy, L. Barsky, K. Weinberg, B. Driscoll.
2006
. Contribution of proliferation and DNA damage repair to alveolar epithelial type 2 cell recovery from hyperoxia.
Am. J. Physiol.
290
:
L685
-L694.
23
Bhandari, V..
2008
. Molecular mechanisms of hyperoxia-induced acute lung injury.
Front. Biosci.
13
:
6653
-6661.
24
Tang, P. S., M. Mura, R. Seth, M. Liu.
2008
. Acute lung injury and cell death: how many ways can cells die?.
Am. J. Physiol.
294
:
L632
-L641.
25
Minamino, T., I. Komuro.
2006
. Regeneration of the endothelium as a novel therapeutic strategy for acute lung injury.
J. Clin. Invest.
116
:
2316
-2319.
26
Ware, L. B..
2006
. Pathophysiology of acute lung injury and the acute respiratory distress syndrome.
Semin. Respir. Crit. Care Med.
27
:
337
-349.
27
Roper, J. M., D. J. Mazzatti, R. H. Watkins, W. M. Maniscalco, P. C. Keng, M. A. O'Reilly.
2004
. In vivo exposure to hyperoxia induces DNA damage in a population of alveolar type II epithelial cells.
Am. J. Physiol.
286
:
L1045
-L1054.
28
O'Reilly, M. A., R. J. Staversky, J. N. Finkelstein, P. C. Keng.
2003
. Activation of the G2 cell cycle checkpoint enhances survival of epithelial cells exposed to hyperoxia.
Am. J. Physiol.
284
:
L368
-L375.
29
Nyunoya, T., L. S. Powers, T. O. Yarovinsky, N. S. Butler, M. M. Monick, G. W. Hunninghake.
2003
. Hyperoxia induces macrophage cell cycle arrest by adhesion-dependent induction of p21Cip1 and activation of the retinoblastoma protein.
J. Biol. Chem.
278
:
36099
-36106.
30
Reddy, N. M., S. R. Kleeberger, M. Yamamoto, T. W. Kensler, C. Scollick, S. Biswal, S. P. Reddy.
2007
. Genetic dissection of the Nrf2-dependent redox signaling regulated transcriptional programs of cell proliferation and cytoprotection.
Physiol. Genomics
32
:
74
-81.
31
Goodman, R. B., J. Pugin, J. S. Lee, M. A. Matthay.
2003
. Cytokine-mediated inflammation in acute lung injury.
Cytokine Growth Factor Rev.
14
:
523
-535.
32
Serhan, C. N., J. Savill.
2005
. Resolution of inflammation: the beginning programs the end.
Nat. Immunol.
6
:
1191
-1197.
33
Zhang, X., D. M. Mosser.
2008
. Macrophage activation by endogenous danger signals.
J. Pathol.
214
:
161
-178.
34
Nathan, C. F..
1987
. Secretory products of macrophages.
J. Clin. Invest.
79
:
319
-326.
35
Takemura, R., Z. Werb.
1984
. Secretory products of macrophages and their physiological functions.
Am. J. Physiol.
246
:
C1
-C9.