Lung EC-SOD overexpression attenuates hypoxic induction of Egr-1 and chronic hypoxic pulmonary vascular remodeling

Eva Nozik-Grayck, Hagir B. Suliman, Susan Majka, Joseph Albietz, Zachary Van Rheen, Kevin Roush, Kurt R. Stenmark

Abstract

Although production of reactive oxygen species (ROS) such as superoxide (O2·−) has been implicated in chronic hypoxia-induced pulmonary hypertension (PH) and pulmonary vascular remodeling, the transcription factors and gene targets through which ROS exert their effects have not been completely identified. We used mice overexpressing the extracellular antioxidant enzyme extracellular superoxide dismutase (EC-SOD TG) to test the hypothesis that O2·− generated in the extracellular compartment under hypoxic conditions contributes to PH through the induction of the transcription factor, early growth response-1 (Egr-1), and its downstream gene target, tissue factor (TF). We found that chronic hypoxia decreased lung EC-SOD activity and protein expression in wild-type mice, but that EC-SOD activity remained five to seven times higher in EC-SOD TG mice under hypoxic conditions. EC-SOD overexpression attenuated chronic hypoxic PH, and vascular remodeling, measured by right ventricular systolic pressures, proliferation of cells in the vessel wall, muscularization of small pulmonary vessels, and collagen deposition. EC-SOD overexpression also prevented the early hypoxia-dependent upregulation of the redox-sensitive transcription factor Egr-1 and the procoagulant protein TF. These data provide the first evidence that EC-SOD activity is disrupted in chronic hypoxia, and increased EC-SOD activity can attenuate chronic hypoxic PH by limiting the hypoxic upregulation of redox-sensitive genes.

  • extracellular superoxide dismutase
  • redox-sensitive transcription factor
  • early growth response-1
  • tissue factor
  • pulmonary hypertension

adult and pediatric patients with lung diseases complicated by alveolar hypoxia are at risk for developing pulmonary hypertension (PH), a process that significantly increases morbidity and mortality (43). Production of reactive oxygen species (ROS) such as superoxide (O2·−) has been implicated in the pathogenesis of PH, including chronic hypoxia-induced PH (7, 17, 26, 28, 35, 41). A subset of these studies implicate a role for O2·− generated specifically in the extracellular compartment (28, 34). Extracellular superoxide dismutase (EC-SOD) is the only extracellular enzymatic scavenger of O2·− and is highly localized to the lung and vascular tissue (30, 36, 37). EC-SOD activity decreases in other models associated with oxidative stress, including hyperoxia and bleomycin-induced lung injury (9, 13, 29, 38, 46), although the impact of hypoxia on EC-SOD expression and activity in the lung is unknown.

Interestingly, Kamezaki et al. (23) recently reported that gene transfer of the antioxidant enzyme EC-SOD protected rats against monocrotaline-induced PH, suggesting that extracellular O2·− is important in the disease process. Mice overexpressing EC-SOD have been utilized in models of vascular and lung diseases to confirm a role for extracellular O2·− in the pathogenesis of a number of disease models associated with fibrosis and tissue remodeling (2, 5, 8, 10, 39). The effects of EC-SOD overexpression on chronic hypoxic PH and pulmonary vascular remodeling have not been previously studied. One important function of ROS is their contribution as signaling molecules to regulate the expression of redox-sensitive genes (14, 27). EC-SOD levels, and thus extracellular O2·−, modulate the classic hypoxia-inducible transcription factor, HIF-1α, as well as its downstream target, erythropoietin, in the hypoxic kidney (44, 53). The contribution of extracellular O2·− and EC-SOD in the regulation of transcription factors and downstream genes in the pulmonary circulation has not been examined. Among many redox-regulated transcription factors, early growth response-1 (Egr-1) is of interest because we and others have shown it increases in the lung and pulmonary vascular cells early in response to hypoxia, and Egr-1 is known to activate a number of downstream targets critical to the vascular remodeling process (3, 12, 22, 24, 4951). One important Egr-1-regulated downstream gene is the procoagulant molecule, tissue factor (TF), which has been implicated in vascular remodeling in response to a variety of stimuli including hypoxia in both the systemic and pulmonary circulations (4752).

The overall goals of this study were thus to evaluate the changes in EC-SOD expression and activity in the lung in response to hypoxia; to determine if lung-specific overexpression of EC-SOD would protect against PH and pulmonary vascular remodeling; and, if so, to determine if any of the protective effects might be achieved through suppression of Egr-1/TF pathways. Our approach was to utilize wild-type (WT) and transgenic mice overexpressing human EC-SOD in the lung and assess the effects of hypoxia on physiological, pathological, and biochemical parameters indicative of PH. We present new data demonstrating a loss of EC-SOD activity in the lung after exposure to chronic hypoxia, a protective effect of lung EC-SOD overexpression on chronic hypoxic PH and vascular remodeling, and a novel role for extracellular O2·− in the hypoxic induction of Egr-1 and the downstream target TF in the murine lung.

MATERIALS AND METHODS

Mouse strains.

Animal studies were approved by the Institutional Animal Care and Use Committees. We tested WT C57/BL6 mice and heterozygous transgenic mice bred in the C57/BL6 background that overexpress human EC-SOD (TG). The human EC-SOD transgene is driven by the SP-C promoter to generate overexpression specifically in the lung, and, similar to native EC-SOD, contains the sequences that enable extracellular secretion of the protein (10).

Chronic hypoxic PH.

To produce chronic hypoxia, 8-wk-old mice were placed in a hypobaric hypoxia chamber at a simulated altitude of 18,000 ft (395 mmHg) for up to 35 days. In chronically hypoxic mice exposed to 35 days of hypoxia, along with their age-matched normoxic controls, the development of PH was assessed by right ventricular (RV) pressure measurements. Mice were lightly sedated with intramuscular ketamine (50 mg/kg) and xylazine (10 mg/kg), and RV pressure was measured in a closed chest via a direct RV puncture. A 25-gauge needle attached to a pressure transducer was introduced into the RV, and placement was confirmed by a live pressure tracing. Pressures were measured using the Cardiomax III Cardiac Output Computer (Columbus Instruments) connected to a Dell laptop running the Cardiomax III program v2.10. Systolic RV pressures (RVSP) were monitored for 30 s and averaged every 10 s to account for beat to beat variability. The RVSP measurements for each mouse were compared between different experimental conditions. At the end of the experimental protocol, mice were euthanized with intraperitoneal pentobarbital (150 mg/kg).

Lung tissue was harvested at several time points from normoxic and chronically hypoxic mice. Before harvesting lung tissue, lungs were flushed with 10 ml of cold saline. The right lung was flash-frozen for mRNA, protein, and activity assays. The left lung was inflation-fixed at 20 cmH2O pressure in 4% paraformaldehyde for 20 min and then dissected from the chest cavity and placed in 4% paraformaldehyde at 4°C overnight. Lungs were then transferred to 70% ethanol, paraffin embedded, and sectioned for immunohistochemistry.

SOD activity measurements.

In lung homogenates, EC-SOD was separated from intracellular SODs (Cu,Zn SOD and MnSOD) using a concavalin A sepharose column as previously described (33). EC-SOD and intracellular SOD activity levels were measured using a commercially available SOD assay kit (Dojindo Molecular Technologies). Briefly, this kit utilizes a water-soluble tetrazolium salt, WST-1 [2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfo-phenyl)-2H-tetrazolium, monosodium salt], to produce a water-soluble formazan dye upon reduction with a superoxide anion, detectable by a colorimetric assay. The rate of WST-1 reduction by superoxide anion is linearly related to SOD concentrations between 0.1 and 10 U/ml of SOD; thus, activity is calculated based on a standard curve generated with xanthine oxidase and SOD. SOD activity data was expressed as units of SOD activity per gram of tissue.

Protein expression.

Lung protein homogenates (cytosolic and membrane proteins) and nuclear protein extracts were isolated from lung tissue using a commercially available kit (NE-PER, Pierce Biotech). Proteins (10 μg nuclear protein extracts or 30 μg lung protein homogenates) were separated by gel electrophoresis and transferred to PVDF membranes as previously described. Blots were blocked with 5% milk in TBS and probed with primary antibody for 1 h followed by an appropriate HRP-conjugated secondary antibody. Blots from the lung protein homogenates were probed with the following primary antibodies: polyclonal rabbit anti-mouse EC-SOD antibody (1:1,000) to detect endogenous EC-SOD, polyclonal rabbit anti-human EC-SOD antibody (1:5,000) to detect the human EC-SOD transgene (kind gift of Dr. James Crapo, National Jewish Medical and Research Center, Denver, CO), and polyclonal rabbit TF antibody (1:500; American Diagnostics, Stamford, CT). Each blot was stripped (Restore Plus Western Blot Stripping Buffer, Pierce) and reprobed with β-actin antibody (1:6,000) for protein normalization. Blots from nuclear protein extracts were probed with a rabbit polyclonal Egr-1 antibody (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA) and rabbit polyclonal NAB2 antibody (1:5,000, Active Motif). These blots were stripped and reprobed with the nuclear loading control TATA Binding Protein (Abcam, Cambridge, MA) for normalization. Blots were developed with ECL plus (GE Healthcare) and bands quantified by densitometry. Normalized values were expressed relative to mean normoxic WT protein expression.

Immunohistochemistry.

Paraformaldehyde-fixed and paraffin-embedded lung sections were immunostained as previously described with the following antibodies: polyclonal rabbit anti-mouse EC-SOD antibody (1:1,000), polyclonal rabbit anti-human EC-SOD (1:500), mouse monoclonal α-smooth muscle actin antibody (α-SMA; 1:100, Clone 1A4), and rabbit monoclonal Ki67 antibody (1:200 Lab Vision NeoMarkers, Ki67 immunoperformed by IHC Services, Aurora, CO). In brief, lung sections were deparaffinized with citrisolve and rehydrated in graded ethanol. Slides were treated with 3% hydrogen peroxide in methanol for 5 min to block endogenous peroxidase activity and treated with 0.01% pronase for 20 min for antigen retrieval. The block, secondary antibody, and ABC reagent were provided in the RTU Universal Elite ABC Kit (Vector Laboratories) for polyclonal rabbit antibodies and the Mouse-on-Mouse kit (Vector Laboratories) for the mouse monoclonal α-SMA antibody. The slides were developed with ImmPact DAB diluent (Vector Laboratories) and counterstained with hematoxylin. Slides were dehydrated with ethanol followed by citrisolve before placement of coverslips. Tissue sections were examined by light microscopy and photographed using a final magnification of ×200. The number of small vessels (<50 μm) with positive α-SMA staining was counted in 10 peripheral high-powered (×20) lung fields to evaluate for muscularized small pulmonary vessels. The number of proliferating Ki67-positive cells within the vessel wall was measured by counting Ki67-positive cells within the intimal and medial layer of muscularized pulmonary arteries between 50 and 150 μm. At least seven vessels were counted for each mouse lung. The analysis of α-SMA and Ki67 staining was evaluated by an investigator blinded to treatment group and genotype.

GSH/GSSG measurements.

Oxidative stress in lung tissue was evaluated using a commercially available glutathione assay kit (Cayman Chemical, Ann Arbor, MI). The kit measures GSH by an enzymatic recycling method using glutathione reductase for the quantification of total tissue glutathione. Fifty milligrams of lung tissue was homogenized in MES buffer (MES, 0.1 M phosphate, and 2 mM EDTA, pH 6.0) and centrifuged at 10,000 g for 15 min at 4°C. The supernatant was removed and deproteinated for analysis (Centrifugal Filter Units, 10,000 MW Cut-Off, R&D Systems, MN). We tested both total glutathione in lung homogenates after reducing GSSG to GSH as well as GSSG alone according to the manual instructions and calculated the ratio of reduced GSH to oxidized GSSG.

Picrosirius red staining of collagen.

The impact of EC-SOD overexpression on chronic hypoxia-induced pulmonary vascular remodeling was further assessed by staining collagen in lung sections with picrosirius red (6). Lung sections were deparaffinized with citrisolve and rehydrated in graded ethanol followed by deionized H2O. Sections were treated with 400 μl of 0.2% phosphomolybdic acid (Electron Microscopy Services, Hatfield, PA) for 5 min followed by 400 μl of 0.1% Sirius Red in saturated picric acid (Electron Microscopy Services) for 90 min at room temperature. The tissue sections were washed with 0.01 N HCl and rinsed with 70% ethanol. The sections were counterstained with hematoxylin. The slides were then dehydrated in graded alcohols followed by citrisolve, and after drying, mounting media and coverslips. Pulmonary arteries adjacent to the junction between the terminal bronchioles and alveolar ducts were photographed by light microscopy to allow visual comparisons between similar sized mid-sized blood vessels.

mRNA expression.

Mouse EC-SOD, Egr-1, and TF mRNA expression was determined by semiquantitative RT-PCR. Briefly, total RNA was isolated from the lung using TRIzol reagent (Invitrogen, Carlsbad, CA). One microgram of RNA was reverse-transcribed using a 2:1 mixture of oligo dT to random hexamer primers and Superscript III enzyme (Invitrogen) according to Invitrogen instructions. mRNA expression was amplified for 30 cycles by the PCR using mouse specific primers for Egr-1 and TF. Primers were EC-SOD forward: 5′-GCACTTCCACAGACC CAGAT-3′, reverse: 5′-GGTGAGGGTGTCAGAGTGGT-3′; Egr-1 forward: 5′-GCA GAT CTC TGA CCC GTT CGG -3′, Egr-1 reverse: 5′-CCG AGT CGT TTG GCT GGG ATA-3′; TF forward: 5′-GCC GGT ACC CAT CAC TCG CTC CCT CCG-3′, TF reverse: 5′-TTC CTC CGT GGG ACA GAG AGG ACC TTT G-3′. Products were run on an agarose gel, and band intensity was measured by densitometry. Egr-1 and TF amplicons were normalized to β-actin amplicons, and normalized values were expressed relative to normoxic WT mRNA expression.

Reagents.

Reagents, unless specified, were obtained from Sigma/Aldrich Chemical (St. Louis, MO).

Data analysis.

Data were expressed as means ± SE. Comparisons were made between normoxic and hypoxic mice and between genotypes by one-way ANOVA followed by Bonferroni post hoc analysis using Prism GraphPad software. Statistical significance was defined as P < 0.05.

RESULTS

Lung EC-SOD activity and expression in chronic hypoxia.

We measured the effect of chronic hypoxia on lung EC-SOD protein activity, expression, and localization in WT mice. The impact of hypoxia on endogenous EC-SOD activity was determined in WT mice at early (1-day) and late (35-day) time points. Lung EC-SOD activity was increased following the 1-day exposure of WT mice. However, by 35 days of hypoxia, lung EC-SOD activity was decreased to 60% of baseline activity (Fig. 1A).

Fig. 1.

Lung extracellular superoxide dismutase (EC-SOD) activity in chronic hypoxia. Lung EC-SOD was separated from intracellular SODs using a concavalin A column, and EC-SOD activity levels were measured using a SOD assay kit (Dojindo Molecular Technologies) after 0, 1, or 35 days of hypobaric (Hypo) hypoxia in wild-type (WT) mice (A) and mice overexpressing lung EC-SOD (B). EC-SOD activity was expressed as U/g tissue. *P < 0.05 vs. WT normoxic lungs compared by 1-way ANOVA; n = 4–5.

In contrast, the normoxic EC-SOD TG mice showed an approximately five- to sixfold increase in lung EC-SOD activity levels compared with WT mice, similar to our previously published report (10). The EC-SOD activity levels in the EC-SOD TG mice remained elevated and did not significantly change in response to either a 1-day or 35-day exposure to hypoxia. After 35 days of hypoxia, the EC-SOD activity in the EC-SOD TG mice was sevenfold greater than WT mice due to the decreased activity levels measured in WT mice at this late time point (Fig. 1B). Intracellular SOD (SOD1 and SOD2) activity remained similar in both strains of mice under normoxia (521 ± 40 U/g tissue in WT vs. 464 ± 195 U/g tissue in EC-SOD TG, n = 5 each, P > 0.05) with no significant change following exposure to 35 days of hypoxia.

Although we detected a small increase in EC-SOD activity at 1 day of hypoxia in the lungs of WT mice, we measured a decrease in GSH/GSSG ratio after 1 day of hypoxia, indicative of increased oxidative stress in response to hypoxia (Fig. 2). In contrast, in mice overexpressing EC-SOD, there was an increased GSH/GSSG ratio compared with WT mice at baseline that did not significantly decrease in response to hypoxia (Fig. 2).

Fig. 2.

Overexpression of EC-SOD protects against oxidative stress. The ratio of reduced to oxidized glutathione (GSH/GSSG) was measured in lung tissue using a glutathione assay kit (Cayman Chemical) in WT and EC-SOD TG mice exposed to 1 day of hypoxia. This time point was selected to test whether there was evidence for oxidative stress despite the small increase in EC-SOD activity levels observed at 1 day of hypoxia and test the impact of EC-SOD on the redox state of the lung. *Decreased GSH/GSSG vs. WT normoxic lungs; #increased GSH/GSSG vs. WT normoxic lungs. P < 0.05 by 1-way ANOVA; n = 4.

We evaluated changes in EC-SOD protein expression and localization using antibodies specific either for endogenous mouse EC-SOD or the human EC-SOD transgene. Endogenous EC-SOD protein expression in WT mice did not significantly change after a 1-day exposure to hypobaric hypoxia (Fig. 3A, left). After 35 days of hypoxia, in concordance with the decrease in EC-SOD activity, there was a decrease in EC-SOD protein expression (Fig. 3B, left). We tested EC-SOD mRNA expression in the WT lungs after 1 and 35 days of hypoxia to evaluate the level of regulation. mRNA levels, evaluated by the densitometry of the EC-SOD signal relative to β-actin, did not change after 1 or 35 days of hypoxia (P > 0.05 by 1-way ANOVA, n = 4, data not shown). In the EC-SOD TG mice, baseline protein expression of the mouse EC-SOD was similar to normoxic WT mice. The expression of endogenous mouse EC-SOD in the EC-SOD TG mice did not change after 1 day (Fig. 3A, right) or 35 days of hypoxia (Fig. 3B, right). Protein expression of the human EC-SOD protein in EC-SOD TG mice did not change in response to hypoxic exposure (Fig. 3C).

Fig. 3.

Lung EC-SOD protein expression in chronic hypoxia. Endogenous mouse EC-SOD (mEC-SOD) protein expression in lungs of WT and EC-SOD TG mice exposed to 1 day (A) and 35 days (B) of hypoxia was evaluated by Western blot analysis using a polyclonal rabbit anti-mouse EC-SOD antibody (1:1,000). C: Western blot analysis of human EC-SOD (hEC-SOD) transgene expression in mice exposed to hypoxia using a polyclonal rabbit anti-human EC-SOD antibody (1:5,000). Lanes from a representative blot are shown with the densitometric measurements of EC-SOD standardized to β-actin and analyzed by 1-way ANOVA. *P < 0.05 vs. normoxic WT lung; n = 4–7.

EC-SOD was localized predominantly within and around pulmonary arteries of hypoxia-exposed WT mice, similar to the localization described previously in lungs of normoxic mice, although the intensity of staining decreased in the 35-day hypoxic lung. In the EC-SOD TG mice, localization of endogenous mouse EC-SOD was similar after both 1 and 35 days of hypoxia (Fig. 4A). As the human EC-SOD antibody does not recognize mouse EC-SOD, the WT mouse lung did not stain for human EC-SOD. Human EC-SOD protein was distributed through the lung parenchyma as well as around pulmonary arteries in 1-day (not shown) and 35-day hypoxia-exposed EC-SOD TG mice (Fig. 4B).

Fig. 4.

Lung EC-SOD immunolocalization in chronic hypoxia. A: immunolocalization of endogenous mouse EC-SOD in WT and EC-SOD TG mice exposed to 1 and 35 days of hypoxia using a polyclonal rabbit anti-mouse EC-SOD antibody (1:200) followed by a goat anti-rabbit secondary IgG antibody conjugated to HRP, developed with DAB (brown stain) and counterstained with hematoxylin (blue stain). B: immunolocalization of the human EC-SOD transgene in WT (absent staining) and EC-SOD TG mice exposed to 35 days of hypoxia using a polyclonal rabbit anti-human EC-SOD antibody (1:500). Bar represents 50 μm.

Overexpression of EC-SOD protects against chronic hypoxic PH.

To establish whether EC-SOD and extracellular O2 contribute to the development of chronic hypoxic PH, we tested whether lung-specific overexpression of EC-SOD imparted protection against increases in RVSP, an indirect measure of pulmonary artery pressures. WT and EC-SOD TG mice had similar baseline RVSP under normoxic conditions. WT mice exposed to 35 days of hypobaric hypoxia developed PH shown by an increase in RVSP. In contrast, although RVSP increased from baseline in EC-SOD TG mice, the degree of PH was significantly less in EC-SOD TG mice exposed to 35 days of hypoxia (Fig. 5).

Fig. 5.

Overexpression of EC-SOD limited the development of chronic hypoxic pulmonary hypertension. Chronic hypoxic pulmonary hypertension was assessed by right ventricular systolic pressure (RVSP, mmHg) measurements via direct RV puncture in WT mice (white bars) and EC-SOD TG mice (black bars) exposed to 35 days of hypobaric hypoxia. *P < 0.05 vs. normoxic WT and #P < 0.05 vs. normoxic EC-SOD TG by 1-way ANOVA; n = 9–10.

Overexpression of EC-SOD attenuated chronic hypoxic pulmonary vascular remodeling.

A pathognomonic feature of chronic hypoxic PH is the remodeling of small pulmonary vessels. This process is characterized by several changes including cell proliferation within the pulmonary artery wall, muscularization of small, previously non-muscularized vessels, and deposition of extracellular matrix protein, including collagen in the pulmonary arteries. We found that overexpression of EC-SOD attenuated each of these markers of pulmonary vascular remodeling. We used Ki67 nuclear staining as a marker for cell proliferation. We detected an increase in proliferating cells within the medial and intimal layers of similar sized pulmonary arteries (50–200 μm) of WT mice in response to a 3-day exposure to hypoxia (Fig. 6, A and B). In the EC-SOD TG mice, there was a similar number of proliferating cells in the pulmonary arteries under normoxic conditions and at 1 day of hypoxia compared with WT mice. Following 3 days of hypoxic exposure, an increase in the number of proliferating cells was observed, but the number of proliferating cells remained significantly less than observed in the WT mice (Fig. 6, A and B). Interestingly, we also observed a marked increase in Ki67 immunoreactivity throughout the lung parenchyma in response to 1 and 3 days of hypoxia in WT mice. This early increase in Ki67 staining was significantly decreased in the EC-SOD TG mice. In addition, proliferation in the lung was dramatically decreased at 35 days of hypoxia in both strains of mice (Fig. 6A).

Fig. 6.

Overexpression of EC-SOD limits pulmonary vascular remodeling, as shown by cell proliferation, muscularization of small vessels, and collagen deposition. Cell proliferation was evaluated by Ki67 immunostaining with a rabbit monoclonal Ki67 antibody (1:200 Lab Vision, NeoMarkers) in lung sections of mice exposed to 0, 1, 3, or 35 days of hypoxia. A: the number of Ki67 positive nuclei in cells within the medial and intimal layer of similar sized pulmonary arteries (50–200 μm) was counted in 7–10 vessels for each lung by an investigator blinded to the experimental conditions. The average number of Ki67 positive cells in the pulmonary artery per animal was analyzed. *P < 0.05 by 1-way ANOVA vs. normoxic WT (n = 3–4). B: Ki67 staining in representative pulmonary arteries of WT and EC-SOD TG mice exposed to normoxia or 3 days of hypoxia (brown color with DAB detection system, arrow indicates Ki67 positive cell within vessel wall). Bar represents 50 μm. C: muscularization was evaluated by α-smooth muscle actin (α-SMA) staining (mouse monoclonal α-SMA antibody, 1:100, Clone 1A4, Sigma) and in small pulmonary vessels after 0, 1, 3, and 35 days of hypoxia. Small vessels (<50 μm) with α-SMA staining within >75% of the circumference of the vessel were counted in 10 peripheral high-powered fields (HPF; ×20 objective) by an investigator blinded to the experimental conditions. The average number of stained vessels in each HPF per animal was analyzed by 1-way ANOVA. *P < 0.05 vs. normoxic WT (n = 4–6). D: representative images of α-SMA immunostaining (brown color with DAB detection system) in small pulmonary vessels (arrow). Bar represents 50 μm. E: chronic hypoxic-induced pulmonary vascular remodeling was also evaluated by visualizing collagen deposition around similar sized pulmonary arteries (arrow) located at the junction of the terminal bronchiole and alveolar duct after staining with picrosirius red (red signal). Sections were counterstained with hematoxylin. A representative image was selected for each mouse strain at 35 days of hypoxia. The bar represents 50 μm.

The appearance of newly muscularized small pulmonary vessels was evaluated by immunohistochemical staining for α-SMA. We found, as expected, a significant increase in the number of small (<50 μm) vessels positive for α-SMA in the lungs of 35-day hypoxic WT mice. The number of α-SMA-expressing small (<50 μm) pulmonary vessels measured in WT mice at 35 days of hypoxia significantly decreased in the lungs of hypoxia-exposed EC-SOD TG mice (Fig. 6, C and D).

Another aspect of vascular remodeling is the deposition of collagen around pulmonary arteries, which is most obvious in the mid-sized pulmonary arteries, which normally contain small amounts of collagen. Prominent increased collagen deposition, shown by picrosirius red staining, was noted in the lungs of chronically hypoxia-exposed WT mice. A decrease in staining was observed in similar sized vessels at the terminal bronchiolar/alveolar duct junction in the EC-SOD TG littermates (Fig. 6E).

EC-SOD regulates hypoxia-induced pulmonary Egr-1 mRNA and protein expression.

Once we established that EC-SOD overexpression protected against chronic hypoxic vascular remodeling, we tested whether EC-SOD overexpression could prevent the hypoxic upregulation of a transcription factor (Egr-1) and a downstream target, TF, thought to be critical to the remodeling process. In WT mice, exposure to 1 day of hypoxia increased lung Egr-1 gene expression by 4.4-fold, along with a concomitant increase in nuclear Egr-1 protein by 2-fold (P < 0.05) (Fig. 7, A and B, left). In striking contrast to the WT mice, no upregulation of Egr-1 mRNA or nuclear protein was observed in EC-SOD TG mice (Fig. 7, A and B, right).

Fig. 7.

Overexpression of EC-SOD prevented the hypoxic induction of lung Egr-1 mRNA and nuclear protein expression. A: Egr-1 mRNA was measured by semiquantitative RT-PCR in the lungs of mice exposed to hypobaric hypoxia (18,000 ft) for 1 day. Bands from a single representative blot are shown along with a bar graph that displays the relative densitometry of Egr-1/β-actin standardized to WT normoxia. *P < 0.05 vs. normoxic WT by 1-way ANOVA (n = 3–4). B: Egr-1 protein in nuclear extracts from the lungs of mice exposed to hypobaric hypoxia for 1 day was compared by Western blot analysis. Bands from a single representative blot are shown along with a bar graph that displays the relative densitometry of Egr-1 normalized to the TATA nuclear loading control. *P < 0.05 vs. normoxic by 1-way ANOVA (n = 4–7). C: the protein expression of the major corepressor of Egr-1, NAB2, was compared by Western blot analysis in nuclear protein extracts from the lungs of mice exposed to hypobaric hypoxia for 1 day. Bands from a single representative blot are shown.

We provided further supporting evidence that EC-SOD overexpression blocks Egr-1 expression by testing for the coinduction of the NGF1-A binding protein, NAB2. It has been shown in other injury models that Egr-1 can induce the expression of its own corepressor molecule, NAB2. The binding of NAB2 to Egr-1 in turn represses transcription of Egr-1 target genes, including Egr-1 itself, to establish a negative feedback loop for Egr-1 activity (18, 25, 42, 45). After 1 day of hypoxia, when Egr-1 expression increased in WT mice, we also detected a corresponding increase in nuclear NAB2 protein. In contrast, in the mice overexpressing EC-SOD that did not exhibit hypoxic induction of Egr-1, the corepressor protein NAB2 also remained low (Fig. 7C).

EC-SOD regulates the hypoxic induction of TF, a downstream gene target of Egr-1.

TF is an important downstream gene target of Egr-1 that is upregulated in the lung in response to hypoxia and initiates a procoagulant state that contributes to pulmonary vascular remodeling (51). Published studies in transgenic mice lacking Egr-1 have already demonstrated the critical link between Egr-1 and TF; therefore, the current studies were designed specifically to examine the impact of EC-SOD overexpression on TF expression. Lung TF mRNA and protein both increased in WT mice exposed to 24 h of hypoxia (Fig. 8, A and B, left). Notably, hypoxia did not induce TF mRNA and protein in the lungs of EC-SOD TG mice (Fig. 8, A and B, right).

Fig. 8.

Overexpression of EC-SOD prevented the hypoxic induction of the Egr-1-responsive gene target, tissue factor (TF), in the lung. A: TF mRNA was measured by semiquantitative RT-PCR in the lungs of mice exposed to hypobaric hypoxia (18,000 ft) for 1 day. Bands from a single representative blot are shown along with a bar graph that displays the relative densitometry of TF/β-actin standardized to WT normoxia. *P < 0.05 vs. normoxic WT by 1-way ANOVA (n = 4). B: TF protein from the lungs of mice exposed to hypobaric hypoxia for 1 day was compared by Western blot analysis. Bands from a single representative blot are shown along with a bar graph that displays the relative densitometry of TF/β-actin. *P < 0.05 vs. normoxic WT by 1-way ANOVA (n = 4–7).

DISCUSSION

The purpose of this study was to evaluate whether overexpression of EC-SOD, the only known extracellular enzymatic defense against O2·−, was protective in chronic hypoxic PH. We report the novel findings that: 1) chronic hypoxia alters the expression and activity of EC-SOD in the lung, 2) overexpression of EC-SOD limited the development of chronic hypoxic PH and pulmonary artery remodeling, and 3) overexpression of EC-SOD prevented the early hypoxic induction of lung Egr-1 and its corepressor NAB2 as well as the Egr-1 downstream gene target, the procoagulant regulator TF. To date, this is the first study to report the changes of EC-SOD in the chronically hypoxic lung and to show that overexpression of EC-SOD blunts the development of chronic hypoxic PH and attenuates redox-sensitive gene expression in the lung.

An increasing number of studies support the concept that O2·− generation increases in the lung in response to low oxygen tension and contributes to chronic hypoxic PH and remodeling (11, 20, 21, 28). This study provides additional evidence of hypoxia-induced oxidative stress by demonstrating a decrease in the ratio of reduced glutathione to oxidized glutathione within the lungs of WT mice exposed to hypoxia. Several enzymatic sources of ROS have been identified in hypoxic lung including NOX2 and NOX4 containing NADPH oxidases and xanthine oxidase (17, 21, 28, 31). Studies in models of PH other than chronic hypoxia also demonstrate a role for ROS via NADPH oxidase and uncoupled endothelial nitric oxide synthase in disease pathogenesis (15). Furthermore, in a number of rat studies, treatment with different antioxidant strategies, including dimethylthiourea, allopurinol, N-acetylcysteine, or tempol, attenuated the development of chronic hypoxic PH (7, 21, 26). This study extends these studies by focusing specifically on the impact of hypoxia on EC-SOD and the protective effects of EC-SOD overexpression in chronic hypoxic PH.

In the present study, we measured changes in EC-SOD activity, expression, and localization in response to hypoxia and defined how the EC-SOD overexpressing transgenic mice differ from their WT counterparts under hypoxic conditions. We observed a modest increase in lung EC-SOD activity in WT mice exposed to 1 day of hypoxia without a concomitant increase in protein expression. This posttranslational change in EC-SOD activity is likely an early adaptive response to increased extracellular O2·− production as evidenced again by the change in the redox state of the lung at this early time point. Exposure to hypoxia for 35 days resulted in a significant decrease in EC-SOD activity and protein expression. This finding is consistent with hyperoxic lung injury in adult mice, which also demonstrated impaired EC-SOD activity and protein expression in response to the altered oxygen tension (38). In this study, we detected a corresponding decrease in EC-SOD protein expression in WT mice. EC-SOD mRNA did not decrease following 35 days of hypoxia, suggesting that the decrease in EC-SOD activity is regulated at the translational and posttranslational level. EC-SOD activity may be regulated at the posttranslational level by cleavage of the heparin binding domain, loss of copper incorporation, or protein modification such as nitration (28, 38). Further studies are indicated to better understand the mechanisms responsible for EC-SOD expression and activity. The EC-SOD TG mice exhibited high EC-SOD activity levels that did not change in response to hypoxia, with the human EC-SOD transgene accounting for 75–87% of the lung EC-SOD activity. The activity levels in EC-SOD TG mice were substantially higher than in WT mice at both 1 and 35 days of hypoxia. The SP-C promoter that drives human EC-SOD expression in the transgenic mice is not known to be regulated by hypoxia. These mice had a higher portion of reduced to oxidized glutathione at baseline compared with WT mice with no significant change in the redox state of the lung in response to hypoxia, further demonstrating the protective effects of EC-SOD overexpression. The marked differences in EC-SOD activity without a difference in intracellular SOD activity make the use of these two strains, the WT mice and EC-SOD TG mice, useful to test the contribution of extracellular O2·− to the development of chronic hypoxic PH.

Mice overexpressing or lacking EC-SOD have been utilized in models of vascular and lung diseases associated with fibrosis including bleomycin-induced lung fibrosis, viral pneumonia, and hyperoxic lung injury to demonstrate a role for extracellular O2·− in disease pathogenesis (1, 2, 5, 8, 10). The protective effects of EC-SOD have been attributed both to its ability to limit extracellular reactions of O2·− and to prevent the inactivation of NO by O2·−, thus preserving NO bioactivity. Chronic hypoxia leads to both an increase in O2·− and impaired NO production despite increased eNOS expression, providing a strong rationale to test the impact of EC-SOD on chronic hypoxic PH. We now demonstrate that lung-specific EC-SOD overexpression significantly attenuates the development of PH and pulmonary vascular remodeling in response to chronic hypoxia. Interestingly, in this mouse strain, the human EC-SOD transgene is selectively expressed by type II alveolar epithelial cells, although we still observed protection within the pulmonary circulation. There are several possible explanations for this observation. First, although the human EC-SOD transgene is synthesized within type II cells, we found that it is secreted into the extracellular compartment in the hypoxia-exposed lung and is detectable throughout the lung, including the vessel wall. This enables EC-SOD to directly scavenge extracellular O2·− generated by the pulmonary artery. Second, since other lung cells can influence pulmonary vascular function, it is also possible that EC-SOD protects the pulmonary circulation through its effects on other cells within the lung including both epithelial cells and inflammatory cells. It is known that bronchial tissue can influence vascular tone, and factors such as VEGF produced by the epithelial cells influence adjacent vascular cell proliferation and differentiation (4). Interestingly, we observed that changes in the lungs of hypoxia-exposed mouse, including cell proliferation and collagen deposition, were not exclusively localized to the pulmonary artery, as there were also visible changes within the lung parenchyma. It is possible that specific vascular smooth muscle cell overexpression of EC-SOD would have a different effect on chronic hypoxic PH. While this study focused on the pulmonary circulation, further investigations are needed to more fully understand the entire pulmonary response to chronic hypoxia and the contribution of extracellular O2·− to these processes.

There is a large body of literature that demonstrates the role of ROS in the regulation of gene expression, although few studies have examined the contribution of O2·− generated in the extracellular compartment. Therefore, as the first step to examine how overexpression of EC-SOD might protect the hypoxia-exposed lungs from the development of PH, we tested the impact of EC-SOD overexpression on the hypoxic induction of the redox-sensitive transcription factor Egr-1, along with its downstream gene target, TF. Egr-1 is increasingly recognized as a critically important master regulator of genes that contribute to the diverse cardiovascular and pulmonary diseases, including hypoxic PH (24). Our observation that Egr-1 and TF are increased in the hypoxia-exposed WT mouse is consistent with a series of published studies from our laboratory and others that also report an increase in Egr-1 within days of hypoxic exposures and furthermore confirm a role for Egr-1 and TF in fibrin deposition in the hypoxic pulmonary vasculature (3, 12, 22, 4951). A key observation in this study was that the hypoxic induction of Egr-1 and TF at 1 day in WT mouse lung did not occur in mice overexpressing EC-SOD. We also reported an increase in NAB2 in the WT lungs, but not EC-SOD TG lungs, in response to hypoxia. The upregulation of NAB2 in response to hypoxia has not been previously reported but is consistent with other models in which NAB2 increases along with Egr-1 (18, 25, 42, 45). As NAB2 is also a downstream gene target of Egr-1, our data indicate that the pathway to limit Egr-1 activity, by upregulating its own corepressor molecule to suppress transcriptional activity, occurs simultaneously with the upregulation of downstream targets that promote vascular remodeling (e.g., TF). The protective effects of EC-SOD on Egr-1 expression could be mediated by either decreased O2·− levels or increased NO bioavailability and are the focus of new studies.

We limited this study to one transcription factor and these two downstream targets, TF and NAB2, to provide proof of principle, although we recognize that other transcription factors as well as other Egr-1-dependent genes are probably also important. Our laboratory previously reported in pulmonary artery adventitial fibroblasts isolated from chronically hypoxic neonatal calves that Egr-1 regulated growth-related genes, cyclin D and EGFR, and that inhibition of Egr-1 in this cell culture model prevented hypoxia-induced fibroblast growth (3). Although we did not measure genes associated with proliferation, we did observe changes in cell proliferation early in response to hypoxia in the WT, but not EC-SOD TG, mice, which could be due in part to the changes in Egr-1. Recently, in other models of lung disease, several notable Egr-1 responsive gene targets have been implicated in lung injury including proinflammatory cytokines, bone morphogenetic protein receptor, and TGF-β (16, 19, 32, 40). These studies are highly interesting due to the link between these particular genes and human PH and provide the rationale for future studies of other downstream gene targets of Egr-1 in hypoxic PH.

In summary, the development of chronic hypoxic pulmonary vascular remodeling and PH was attenuated in mice overexpressing EC-SOD. The early hypoxic induction of a transcription factor, Egr-1, and its downstream target, TF, which contribute to hypoxic pulmonary vascular remodeling, were significantly blocked in the lungs of EC-SOD TG mice. The corepressor protein NAB2 was also upregulated in hypoxia-exposed WT mice but not EC-SOD TG mice. Our new data thus support our hypothesis that the generation of extracellular O2·− in hypoxia contributes to the development of PH and upregulates redox-sensitive genes relevant to pulmonary vascular remodeling. These data provide new insight into the role of ROS on chronic hypoxic PH. These data also provide a foundation for future investigations to test new therapeutic interventions with novel antioxidant strategies that target extracellular O2·− in the lung.

GRANTS

This work was funded by National Heart, Lung, and Blood Institute (NHLBI) Grant HL-086680 and March of Dimes #6-FY06-316 (both to E. Nozik-Grayck) and NHLBI Grant HL-084923-01 (to K. R. Stenmark).

Acknowledgments

We thank Dr. Sonia Flores for constructive feedback and Sebastian Albu and Erin Bolivar for excellent technical assistance.

Footnotes

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

REFERENCES

View Abstract