Enhanced epithelial sodium channel activity in neonatal Scnn1b mouse lung attenuates high oxygen-induced lung injury

Abstract

Prolonged oxygen therapy leads to oxidative stress, epithelial dysfunction, and acute lung injury in preterm infants and adults. Heterozygous Scnn1b mice, which overexpress lung epithelial sodium channels (ENaC), and their wild-type (WT) C57Bl6 littermates were utilized to study the pathogenesis of high fraction inspired oxygen (FI_O2)-induced lung injury. Exposure to high FI_O2 from birth to postnatal (PN) day 11 was used to model oxidative stress. Chronic exposure of newborn pups to 85% O₂ increased glutathione disulfide (GSSG) and elevated the GSH/GSSG redox potential (E_h) of bronchoalveolar lavage fluid (BALF). Longitudinal X-ray imaging and Evans blue-labeled-albumin assays showed that chronic 85% O₂ and acute GSSG (400 µM) exposures decreased alveolar fluid clearance (AFC) in the WT lung. Morphometric analysis of WT pups insufflated with GSSG (400 µM) or amiloride (1 µM) showed a reduction in alveologenesis and increased lung injury compared with age-matched control pups. The Scnn1b mouse lung phenotype was not further aggravated by chronic 85% O₂ exposure. These outcomes support the hypothesis that exposure to hyperoxia increases GSSG, resulting in reduced lung fluid reabsorption due to inhibition of amiloride-sensitive ENaC. Flavin adenine dinucleotide (FADH₂; 10 µM) was effective in recycling GSSG in vivo and promoted alveologenesis, but did not impact AFC nor attenuate fibrosis following high FI_O2 exposure. In conclusion, the data indicate that FADH₂ may be pivotal for normal lung development, and show that ENaC is a key factor in promoting alveologenesis, sustaining AFC, and attenuating fibrotic lung injury caused by prolonged oxygen therapy in WT mice.

INTRODUCTION

Hyperoxia mediates lung injury (1–3) and increases the risk of developing bronchopulmonary dysplasia (BPD) in preterm infants delivered before 37 wk gestation (4–6). Attenuated lung development, fibrosis, and airway smooth muscle hyperplasia threaten life and characterize BPD (7). Oxidative stress has been shown to interrupt normal lung development in newborns who develop BPD by altering cellular signaling mechanisms (8). Therefore, identifying the mechanisms by which oxidative stress promotes BPD will offer opportunities for targeted interventions to improve survival for preterm infants requiring higher than atmospheric fraction of inspired oxygen (FI_O2) therapy (9–12).

High FI_O2 therapy leads to oxidative stress in the preterm lungs by overwhelming an immature antioxidant system. Mature lungs are rich in the antioxidant glutathione (GSH), which scavenges reactive oxygen species in a reaction catalyzed by glutathione peroxidase (GPx) to protect cells from oxidative stress and cell injury (13). Specifically, GPx converts GSH to its oxidized form, glutathione disulfide (GSSG). Subsequently, GSSG is recycled back to its reduced form by glutathione reductase (GR) and associated cofactors such as flavin adenine dinucleotide (FAD) to maintain oxidation/reduction (redox) balance. Because GSH and GPx levels are deficient in the preterm lungs (14–18), these newborns are especially vulnerable to oxidative stress.

Epithelial sodium channels (ENaC) are amiloride-sensitive sodium channels located on the apical surface of airway epithelia. At birth, ENaC plays a critical role in pulmonary absorption of fetal lung fluid (19, 20). Low levels of ENaC expression makes the preterm lungs vulnerable to oxidative stress (21, 22) because the preterm lungs are not yet fully developed and contain 25% more water than term infant lungs (23). Moreover, the preterm lungs are made up of saccular structures lined with thick cuboidal cells and capillaries that are distant from the epithelium, which limits effective gas exchange due to increased distance for diffusion of gases (24, 25). Due to underdeveloped lungs and immature antioxidant systems, even small decrements in lung ENaC activity can lead to alveolar flooding and adverse health outcomes in premature lungs (22, 26–28).

Our research group recently showed that GSSG inhibits canonical ENaC activity using single channel and whole cell current measurements (29, 30). As an extension of our previous work, we herein hypothesize that supplemental O₂ therapy increases the risk for lung injury by increasing extracellular GSSG levels, which thereby inhibits ENaC activity and adversely affects normal lung function. To support this hypothesis, we used the term mouse model for preterm lungs and exposed pups to high FI_O2 of 85% for 11 consecutive days to model oxidative stress. Scnn1b mice that overexpress lung ENaC were used to evaluate whether increased sodium channel activity could alter the course of lung injury following chronic high oxygen exposure. FADH₂ treatment was given to wild-type (WT) pups following chronic 85% O₂ exposure to evaluate its effectiveness in reducing oxidative stress and attenuating further lung injury. The data indicate that regulatory factors that sustain GSH/GSSG redox balance and lung ENaC activity may be important therapeutic targets for preventing high oxygen-induced lung injury in preterm infants.

MATERIALS AND METHODS

Alveolar Fluid Clearance

Sodium channel activity controls the alveolar fluid clearance (AFC) rate from the airspace into the interstitial space (31, 32), which can be measured using Evans blue-labeled-albumin assays (33) and real-time X-ray imaging (34–36).

Evans Blue-Labeled-Albumin Assay

Newborn pups were anesthetized with intraperitoneal injections of ketamine (NDC 0143-9509-01; 100 mg/kg body wt; West-Ward Eatontown, NJ) and xylazine (NDC 13985-704-10; 10 mg/kg body wt; MWI Animal Health, Boise, ID). A 26-gauge blunt tip cannula was inserted caudally into the lumen of the trachea and secured using nylon suture. Saline solution containing 5% bovine serum albumin (BSA; MDL No. MFCD00130384) and Evans blue dye (0.1 mg/mL) was instilled into the airspace via the secured cannula in a final volume of 35 μL/g body weight. Saline solution consisted of 140 mM NaCl, 5 mM KCl, 1 mM CaCl₂, and 10 mM HEPES; pH = 7.4 and 315 mOsmol/L. Evans blue-labeled-albumin containing saline was recovered 5–30 min following instillation to measure change in albumin protein concentration (one determinant of AFC) using a Nanodrop 2000 (Thermo Scientific, Waltham, MA), and calculated as follows:

$$\text{AFC}=\left[\left(V_{\text{i}}-V_{\text{f}}\right)/V_{\text{i}}\right]\times \text{1}00,$$

where V_i is the volume of instilled Evans blue-labeled albumin and V_f is the final alveolar fluid calculated as follows:

$$V_{\text{f}}=\left(V_{\text{i}}\times {\text{EB}}_{\text{i}}\right)/{\text{EB}}_{\text{f}},$$

where EB is the concentration of Evans blue-labeled albumin in the instilled solution (i) and the final alveolar fluid (f).

Real Time X-Ray Imaging

Longitudinal X-ray imaging of freely breathing anesthetized pups was performed as previously described (34–36). Immediately following tracheal fluid instillations (5 μL saline/g body weight), neonatal pups were placed into an In-Vivo Multispectral Imaging System (Bruker, Billerica, CT) and were X-ray imaged at 5 min intervals up to 2 h with an acquisition period of 120 s/image. X-ray settings were as follows: 2 × 2 binning, 180-mm field of view, 149 µA X-ray current, 35 kVp, and 0.4-mm aluminum filter. Anesthesia was maintained with isoflurane in oxygen delivered using a flow rate of 1.5 L/min via a mouse nose cone. X-ray density of the flooded upper left lobe of the newborn lung was defined as the region of interest (ROI = 5 mm²), and was quantified using Carestream Health MI software. All ROIs were background corrected and normalized to the initial X-ray intensity (I_o) to quantify changes in AFC fluoroscopically expressed as I − I_o. Peak fluid volume and rate of fluid reabsorption were calculated from measured I − I_o values and modeled as previously described by our group (35). Briefly, the steady-state value of fluid balance, K, is the ratio of the rate of absorption to the rate of secretion (35). Integration of this differential equation with the addition of the rate of elimination gives the model for in vivo fluid clearance using measured I_o values in the following equation:

$$\text{F}\left(t\right)=\left(K\left(\text{1}-e^{-k_{\text{a}}t}\right)\right)e^{-k_{\text{e}}t}$$

where F(t) is the amount of surface fluid in the lung at any time, t; K is the steady state or peak amount of lung fluid; k_a is the rate of absorption of fluid; and k_e is the rate of drug elimination. The form of the function is consistent with fluid rapidly rising at rate k_a to a peak value, K, before gradually decreasing toward initial levels as drug is at rate k_e. The value, K, provides a measure of the maximal effect of any of the treatments and can be statistically compared.

GSH/GSSG Redox Potential

The ratio of GSH/GSSG is an indicator of oxidative stress and can be measured using high-performance liquid chromatography (HPLC) or commercially available assays (GSH Assay Kit, ab239709, Abcam, Cambridge, MA). HPLC was used to analyze GSH and GSSG concentrations in postnatal (PN) days 3–11 lungs that yield low volume bronchoalveolar lavage fluid (BALF). The GSH Assay Kit was utilized in tissue culture supernatant and PN day 11 lung studies, as indicated. Both assays allowed for calculating the redox potential of the GSH/GSSG redox pair in accordance with the Nernst equation where: E_h = E_o + RT/2F ln [(GSSG)/(GSH)²]; E_o is the standard potential for the redox couple, R is the gas constant, T is the absolute temperature (K), and F is Faraday’s constant. Calculated E_h serves as a quantitative expression of the redox pair’s affinity for electrons (i.e., oxidative stress).

High-Performance Liquid Chromatography

Anesthetized pups were cannulated and lavaged with 400 µL sterile phosphate buffered saline (PBS) under a dissecting microscope. The BALF was centrifuged at 6,000 rpm and the supernatant transferred to a new microfuge tube to obtain cell-free BALF. BALF samples were acidified with 5% perchloric acid (Pubchem CID 24247) containing an internal standard of 50 mmol/L γ-glutamyl-glutamate (Pubchem CID 92865). After derivatization with iodoacetic acid and dansyl chloride (Pubchem CID 5240 and 11801, respectively), the GSH, GSSG, and Cys fractions were separated by HPLC using an amino μBondaPak column (Waters, Milford, MA). Fluorescent detection was used to separate and quantitate the dansyl derivatives relative to the fluorescence of the γ-glutamyl-glutamate standard.

GSH Assay Kit

Manufacturer protocol was followed without deviation. Briefly, BALF or tissue culture supernatants were combined with 5-sulfosalicylic acid (Pubchem CID 665025) and centrifuged for removal of protein. The samples were subsequently processed with Ellman’s Reagent and glutathione reductase, which reacts to generate 2-nitro-5-thiobenzoic acid in the presence of GSH. In this way, GSH and GSSG can be measured on a standard microplate reader at an absorbance of 415 nm.

Histopathology

Neonatal lung tissue was removed en bloc, formalin-fixed, paraffin-embedded, cut into 4-µm sections and histochemically labeled. Stained lung samples were visualized and imaged using an Olympus IX71 microscope and compatible DP80 dual sensor camera attachment (Olympus, Tokyo, Japan) for subsequent morphometric analysis. Radial alveolar counts (RAC) were calculated as described by Emery and Mithal (37). Mean linear intercepts (MLI) were measured as described by Dunnill (38) using the equation MLI = N⋅L/m, where N is the number of traverses, L is the length of the traverse, and m is the total number of intercepts by the traverses.

Quantification of Fibrosis Using Image Analysis

Trichrome-labeled images were obtained using light microscopy and ×20 objective. Digital images were imported into Fiji (an image processing program distributed by ImageJ with open source plug-ins) to threshold visual fields occupied by collagen as described by Hadi et al. (39). Subsequent binary images were created to calculate the proportion of collagen to the total area analyzed.

Hydroxyproline Assay

Lung hydroxyproline content is indicative of collagen deposition and was determined using a kit following the manufacturer’s protocol (BioVision, Milpitas, CA).

Animal Models

Preclinical animal models were used in accordance with the Guide for the Care and Use of Laboratory Animals and reported following the Animal Research: Reporting In Vivo Experiments Guideline (40, 41). Animal protocols were reviewed and approved by The University of Utah Institutional Animal Care Committee. All animals were given ad libitum access to food and water, and housed under a normal light/dark cycle. Ambient temperature was kept between 22.2°C and 23.3°C in the vivarium.

Term Mouse Model of Preterm Lung Injury

Mice delivered at term are in the saccular stage of lung organogenesis with thick septa and capillaries that are still distant from the airway epithelium. The term mouse lung closely resembles the immature lungs of infants delivered before 37 wk of pregnancy and is therefore employed to study preterm lung injury. WT C57Bl6 mice (Taconic Bioscience, Rensselaer, NY) were mated with Scnn1b mice that overexpress ENaC under the CCSP promoter. The Scnn1b mice were obtained from Dr. Alessandra Livraghi-Butrico (Marsico Lung Institute, Chapel Hill, NC) who has extensively characterized the phenotype of Scnn1b mouse lung (42, 43). We confined our evaluation of Scnn1b lung to pups that are >21 days old because parenchymal abnormalities are modest in neonates compared with adult mice. All pups were genotyped on PN day 7 by submitting tail snips to TransnetYX, Cordova, TN. Newborn mice were randomly assigned to the treatment groups described in results.

Lung Injury Model

In rodents, chronic exposure to high FI_O2 leads to disruption of lung development and fibrotic lung injury that parallels changes seen in BPD. As such, we chose a high oxygen-induced lung injury model to investigate the role of lung ENaC in preterm lung injury. Newborn litters were paired so that lactating dams could be swapped daily between hyperoxic (85% O₂) and normoxic (21% O₂) chambers in order to avoid maternal oxygen toxicity. Newborns were continuously exposed to 85% O₂ or 21% O₂ beginning on the day of birth through PN days 3–18, as indicated. Pups assigned to 85% O₂ were housed in a Plexiglass chamber with continuous oxygen monitoring (Proox model 110; Biospherix, Redfield, NY). Newborn litters were not redistributed to separate males from females. In rodent studies, sex is not a relevant biological variable in high oxygen-induced lung injury (44, 45). A subset of pups maintained on 85% O₂ through PN day 11 were randomized to receive either saline or FADH₂ (10 µM) intranasally for seven consecutive days at room air.

Blood Gas Analysis and Respiratory Rate

On PN day 11, arterial blood samples were drawn from the left ventricle using a 22-gauge needle attached to a 1-mL luer slip syringe containing 23.5 IU/mL of dry lithium heparin (Fisher Scientific). The needle was disconnected and replaced with a filtered cap immediately following collection. Arterial blood samples were transported on ice to the ABL800 Flex blood gas analyzer (Radiometer America, Brea, CA) housed in adjoining clinical facilities in the University of Utah medical campus. Oximetry and blood gas measurements were measured from 35 µL blood samples using the FLEXMODE and MICROMODE of the ABL800 Flex blood gas analyzer. Respiratory rates were measured by counting the number of times the abdomen rises over the course of one minute.

Nasal Insufflation

Newborn mice were nasally insufflated with PBS, GSSG (400 µM), or amiloride (1 µM). Unanesthetized pups were held in a supine position and a 20-µL pipette was used to place the solution to be insufflated directly over the nostrils. Insufflation volumes increased from 5 µL volumes on PN days 1, 3, and 5 to 20 µL volumes on PN days 7, 9, and 11. FADH₂ (10 µM) insufflations in 20 µL volumes were delivered daily on PN days 11–18.

Tissue Culture

A549 cells were purchased from ATCC (Manassas, VA) and cultured in a 37°C humidified 5% CO₂ incubator in complete media consisting of Dulbecco’s MEM supplemented with 10% FBS and 1 µM dexamethasone (Thermo Fisher Scientific, Pittsburg, PA). A549 cells were subcultured onto 60-mm tissue culture dishes and treated with 1–100 µM FADH₂ for 72 h before measuring GSH and GSSG concentrations.

Statistical Analysis

Sigma Plot software was used for statistical tests and graphical analysis. All data are summarized as means ± SE and plotted as scatter or box plots. Box plots show the minimum, first quartile, third quartile, and maximum values with a horizontal line going through the box at the median. Single data comparisons were performed using paired Student’s t tests. Multiple comparisons were performed using one-way analysis of variance followed by post hoc testing in Sigma Plot. P values of <0.05 were considered statistically significant.

RESULTS

Chronic Exposure to 85% O₂ Attenuates Lung Alveolar Fluid Clearance in Neonates

The rate of neonatal AFC was determined by measuring the percent change in Evans blue-labeled albumin and using real time X-ray imaging techniques. Figure 1A shows that chronic 85% O₂ exposure significantly decreased the rate of AFC compared with age-matched control pups housed at 21% O₂; data shown for three separate litters using the Evans blue-labeled albumin assay for determining AFC. Figure 1B shows representative chest X-ray images of PN day 11 pups housed under 21% O₂ (Fig. 1B, top) and 85% O₂ (Fig. 1B, bottom) 120 min following saline challenges delivered into the lungs. Hyperoxic lung chest X-rays were more opaque compared with X-rays of neonatal lungs maintained at room air for 11 days, thus indicating that there is unresolved fluid in neonatal lungs maintained on high FI_O2. Figure 1C quantifies the rate of AFC in neonatal lung based on change in X-ray opacity (I − I_o) following saline challenges (up to 120 min observation). The rate of fluid absorption in neonatal lungs housed at 21% O₂ (white bars; K_a = 0.99 ± 0.97) and peak fluid volume (K = 190.91 ± 64.30) was significantly higher than neonatal lungs maintained on 85% O₂ (light gray bars; K_a = 0.08 ± 0.05; K = −52.98 ± 22.60). The K_a and K values were calculated from n = 8 pups from three litters, P < 0.05.

Figure 1.

Effect of high oxygen on lung fluid clearance [postnatal (PN) day 11]. A: rate of alveolar fluid clearance (AFC) calculated using Evans blue-labeled-albumin as a measure of lung fluid clearance. Chronic 85% O₂ exposure of newborn pups (gray bars) significantly decreases AFC compared with control pups on 21% O₂ (white bars). Data represent n = 8 from 3 litters and triplicate measures of AFC; *=P < 0.05. B: representative X-ray image of 21% O₂ (top) and chronic 85% O₂ (bottom) treated mice on PN day 11 and 120 min following lung saline challenges. C: quantification of changes in X-ray density from radiographs obtained longitudinally (2–120 min) following saline challenges in 21% O₂ (control; white bars), chronic 85% O₂ (light gray bars), and 400 µM glutathione disulfide (GSSG) (dark gray bars) treatments on PN day 11. Data shown represent full time course for n = 8 from 3 litters for each treatment group. I − I_o (y-axis) represents change in X-ray intensities, where an increase in I − I_o values indicates lung fluid clearance and negative values represent greater X-ray opacities (i.e., alveolar flooding). Peak fluid volumes and rate of fluid reabsorption differ significantly from the control group (†=P < 0.05; ††=P < 0.001 based on the model described in materials and methods and post hoc tests with ANOVA). R² values represent curve fitting using nonlinear regression.

Chronic Exposure to 85% O₂ Increases GSH/GSSG E_h in Neonatal BALF

To better understand how neonatal exposure to high FI_O2 decreases AFC (shown in Fig. 1), we measured GSH and GSSG in bronchoalveolar fluid (BALF) obtained from pups housed under 21% O₂ or 85% O₂ using HPLC analysis. Figure 2A shows that normalized GSH levels are depleted in hyperoxic lung BALF compared with age-matched lungs housed under normoxia (P < 0.05 at PN day 7). GSH levels continue to increase in normoxic lungs, as expected. Corresponding (total) GSSG levels in BALF obtained from hyperoxic lungs were significantly higher on PN day 7 versus normoxic BALF (Fig. 2B), with many measurements remaining high on PN day 11. Since the ratio of GSH to GSSG determines the redox coupling, we calculated the GSH/GSSG E_h of BALF on PN day 11. The GSH/GSSG E_h of hyperoxic BALF was −111.03 ± 21.25 mV, which is significantly higher than the calculated E_h value of BALF obtained from age-matched control pups maintained on 21% O₂ (−147.52 ± 5.39 mV) shown in Fig. 2C.

Figure 2.

Chronic 85% O₂ exposure elevates glutathione (GSH)/glutathione disulfide (GSSG) redox potential (E_h) in neonatal lung bronchoalveolar lavage fluid (BALF). A and B: newborn pups were chronically exposed to 21% or 85% O₂ immediately after birth. On postnatal (PN) days (D) 3, 7, and 11, BALF was collected and high-performance liquid chromatography (HPLC) analyzed for GSH (normalized to total Cys measured by HPLC) and GSSG (normalized to total GSH) levels. Data shown represent n = 3 from three independent experiments (litters) and nonlinear curve fit; *=P < 0.05. C: GSH/GSSG redox potential of BALF measured from PN day 11 pups housed under 21% O₂ or 85% O₂ (n = 6 from 3 litters); *=P < 0.05.

The data indicate that chronic 85% oxygen exposure results in a shift in the GSH/GSSG pool toward GSSG and implicates GSSG as an important determinant of newborn lung injury. Figure 1C shows that GSSG decreased AFC; dark gray bars show that tracheal insufflation of GSSG (400 µM) on PN day 11 mouse lung (previously housed at 21% O₂) significantly decreased fluid absorption (K_a = 0.03 ± 0.01) and peak fluid volume (K = −200.75 ± 32.43) compared with age-matched control neonatal lungs insufflated with saline (P < 0.001) and compared with pups on chronic high FI_O2 of 85% (P < 0.05). The finding that GSSG inhibited AFC extends our previous studies showing that GSSG significantly decreases ENaC activity (29, 30).

GSSG and Amiloride Impairs Alveolar Development

We insufflated WT C57Bl6 pups with GSSG (400 µM) or amiloride (1 µM) every other day between PN days 1 and 11 to better understand the roles of GSSG and ENaC in neonatal lung injury. Figure 3A shows representative hematoxylin and eosin (H&E) stained lungs insufflated with saline (vehicle; Fig. 3A, left), GSSG (Fig. 3A, middle), or amiloride (Fig. 3A, right). H&E labeled panels show rarefication and simplification of alveoli in GSSG and amiloride-treated lungs that are typical of high oxygen-induced lung damage in preterm infants (3, 46). Morphometric analysis of all H&E labeled slides (Fig. 3, B and C) show that GSSG and amiloride significantly decreases radial alveolar counts (RAC; an indicator of lung development) and increases mean linear intercepts (MLI; a stereological assessment of airspace size).

Figure 3.

Glutathione disulfide (GSSG) and amiloride induces neonatal lung injury. A: representative hematoxylin and eosin staining of postnatal (PN) day 11 lung sections. Scale bar = 100 µm. B and C: nasal insufflation of GSSG (400 µM) and amiloride (1 µM) significantly decreases radial alveolar counts (RAC) and increases mean linear intercepts (MLI), respectively, in PN day 11 mouse lungs vs. vehicle instilled control lungs. *=P < 0.05 in n = 3 from three separate litters; 21 regions of interest (ROIs) analyzed per group.

Lung Injury Does Not Worsen in Scnn1b Mouse Lung under Chronic 85% O₂ Exposure

We housed Scnn1b mice (that overexpress lung ENaC) in FI_O2 of 85% or 21% O₂ from birth to PN day 11 in order to study ENaC function under pro-oxidizing conditions. Figure 4 shows alveologenesis and lung architecture are preserved to a greater degree in Scnn1b pups compared with WT littermates co-housed at 85% O₂ through PN day 11. Representative trichrome staining (Fig. 4A) shows pronounced alveolar simplification, thickened alveolar septa, exudate, and (blue) collagen fiber labeling in WT pups compared with Scnn1b littermates co-housed under hyperoxic conditions, as well as age-matched WT and Scnn1b (littermates) maintained at 21% O₂. Figure 4B shows representative binary images of trichrome-labeled lung sections that have threshold values (equal across all experimental groups) set to show areas occupied by collagen.

Figure 4.

Transgenic Scnn1b neonatal lungs are protected from high oxygen-induced lung injury. A: representative Masson’s trichrome staining of postnatal (PN) day 11 wild-type (WT) and Scnn1b mouse lungs housed at 21% O₂ and chronic 85% O₂. Scale bar = 50 µm. B: representative binary images used for calculation of the percentage fibrosis per area. Scale bar = 40 µm. C and D: morphometric analysis of WT and Scnn1b neonatal lungs under 21% O₂ and chronic 85% O₂. Significant difference in radial alveolar counts (RAC) and mean linear intercepts (MLI) measurements were observed between WT vs. Scnn1b littermates at room air; WT mice housed under chronic 85% O₂ vs. WT 21% O₂; and between Scnn1b-85% O₂ vs. WT-85% O₂. Data represent n= 3 pups from three separate litters; 21 regions of interest (ROIs) analyzed per group; *=P < 0.05; ns = not significant. E: Fiji analysis detected more collagen (fibrosis) in WT mice chronically housed at 85% O₂. Data represent n = 3 pups from three separate litters; 9 ROIs analyzed per group; Multiple ANOVA between all groups; *=P < 0.05. F: hydroxyproline quantitation of collagen was significantly reduced in transgenic Scnn1b lung (n = 3 from 3 litters) vs. WT lung (n = 3 from 3 litters) housed under chronic 85% O₂ on PN day 11; *=P < 0.05. The % collagen calculated and hydroxyproline concentrations did not differ between WT and Scnn1b littermates at room air. G: alveolar fluid clearance (AFC) rates were significantly higher in Scnn1b vs. WT lungs measured on PN day 11 following chronic 85% O₂ exposure. *=P < 0.05, Data represent n= 12 from three separate litters.

Morphometric analysis indicates that ENaC overexpression supports normal lung development in pups exposed to chronic 85% O₂. Radial alveolar count (a measure of alveologenesis; Fig. 4C) and mean linear intercept (an indicator of alveolar size; Fig. 4D) measurements of Scnn1b neonatal lung on high oxygen did not differ significantly from control lungs maintained at 21% O₂. Quantification of fibrosis using image analysis (Fig. 4E) and hydroxyproline assays (Fig. 4F) indicate that ENaC overexpression (Scnn1b) attenuates further hyperoxic lung injury. Furthermore, Scnn1b pups sustained their ability to clear flooded airspaces following chronic 85% O₂ exposure. AFC rates (Fig. 4G) measured on PN day 11 were significantly higher in Scnn1b pups versus WT pups that were co-housed under chronic 85% O₂. When left unstressed at room air, neonatal Scnn1b mice did not exhibit higher rates of AFC compared with WT (data not shown; n = 5 litters evaluated at 21% O₂).

Arterial blood gas (ABG) measurements, respiration, and additional H&E studies of WT and Scnn1b newborn lungs (Fig. 5) were evaluated to add additional support and evidence for the beneficial effects of sustaining lung ENaC activity under pro-oxidizing conditions (as opposed to Scnn1b mice potentially benefitting from chronic 85% O₂ exposure). Figure 5, A and B shows that WT Pa_O2 and Pa_CO2 were 91.02 ± 9.92 mmHg and 29.13 ± 3.87 mmHg, respectively; ABG measurements from Scnn1b pups did not differ significantly: Scnn1b Pa_O2 = 102.53 ± 8.90 mmHg and Scnn1b Pa_CO2 = 31.41 ± 4.35 mmHg. ABG data indicate that both WT and Scnn1b mice have adequate ventilation under baseline room air conditions. Chronic 85% O₂ exposure in WT pups reduced respiration per minute (RPM) compared with baseline control and Scnn1b pups maintained on 85% O₂ (Fig. 5C). A decrease in RPM has also been reported in mice exposed to hyperoxia for 28 days (47). H&E labeling of PN day 11 Scnn1b pups (from 6 separate litters spanning 4 generations) shows very modest airway mucus obstruction in ENaC overexpressing lungs. Together, Figs. 4 and 5 show that enhanced AFC in congenic Scnn1b mice does not result in asphyxiating mucus plug formation on PN day 11. Our phenotypic characterization and >85% survival rates of Scnn1b mice (data not shown) are consistent with the study by Livraghi-Butrico et al. (42).

Figure 5.

Effect of lung epithelial sodium channels (ENaC) overexpression and hyperoxia on arterial blood gas (ABG), respiration, and airways obstruction. A and B: arterial blood sampling was conducted in anesthetized wild-type (WT) and Scnn1b pups on postnatal (PN) day 11 at room air. Pa_O2 symbols in A correspond with Pa_CO2 symbols in B. Average ABG values did not significantly differ between WT (n = 9) and Scnn1b (n = 7) pups from four separate litters. C: breathing frequency (respiration per minute; RPM) was significantly reduced in WT mice housed under 85% O₂ compared with Scnn1b littermates similarly housed, and compared with WT/Scnn1b littermates maintained at 21% O₂. Multiple ANOVA between all groups; *=P < 0.05; n = 3 from two litters. D: representative H&E images of WT and Scnn1b at room air. H&E images of Scnn1b lungs are from six different litters spanning four transgenerations. H&E, hematoxylin and eosin.

FADH₂ Enhances the Rate of GSSG Recycling and Improves Fetal Lung Development

GSSG recycling to GSH is dependent on glutathione reductase and adequate availability of the electron carrier FADH₂. Figure 6 shows that 1–100 µM FADH₂ alters total levels of GSH and GSSG (Fig. 6A), which leads to significant lowering of the GSH/GSSG E_h in A549 cell supernatant (Fig. 6B). Likewise, Fig. 6C shows that 10 µM FADH₂ is also effective in lowering GSH/GSSG E_h in vivo in WT mice. FADH₂ significantly lowered oxidative stress following chronic 85% O₂ treatment in BALF versus control (saline treated) BALF. Together, Fig. 6, D–F indicates that FADH₂ treatment following high oxygen exposure did not significantly alter neonatal AFC (Fig. 6D) nor reverse fibrotic lung injury (Fig. 6, E and F). Lowering oxidative stress by insufflating FADH₂, however, did result in improved fetal lung development following high oxygen exposure (Fig. 6, E, G, and H). FADH₂ significantly increased RAC (Fig. 6G) and lowered MLI (Fig. 6H) measurements in oxidatively stressed lungs compared with vehicle (saline) treated lungs. Collectively, the data suggest that extracellular application of FADH₂ (e.g., nasal insufflation) can restore redox balance and improve fetal lung development following supplemental oxygen therapy.

Figure 6.

Reduced FAD (FADH₂) lowers glutathione (GSH)/glutathione disulfide (GSSG) redox potential (E_h) and improves fetal lung development. A: in vitro measurements of GSH (left) and GSSG (right) concentrations following 0, 1, 10, and 100 µM FADH₂ treatments in A549 cells. B: calculated GSH/GSSG E_h following 1–100 µM FADH₂ treatment in A549 cells. *=P < 0.05 vs. control (CTR); n = 3 from three independent experiments performed in triplicate wells. C: 10 µM FADH₂ significantly decreases GSH/GSSG E_h in vivo; *=P < 0.05 in bronchoalveolar lavage fluid (BALF) obtained from three separate litters maintained on 85% O₂ ± FADH₂ treatments; individual E_h measurements shown. D: FADH₂ did not significantly alter alveolar fluid clearance (AFC) rates in neonatal lungs following chronic O₂ exposure evaluated in BALF obtained from five separate litters (individual measurements from pups shown). E: Masson’s trichrome staining of neonatal lung. Scale bar = 50 µm. F: hydroxyproline measurements of collagen significantly increased in hyperoxic + FADH₂ treated mouse lung (n = 7 from 7 litters) vs. vehicle control lungs (n = 4 from 4 litters) assayed on postnatal (PN) day 18; *=P < 0.05. G and H: morphometric analysis of Masson’s trichrome stained lungs shows that FADH₂ significantly increases radial alveolar counts (RAC) and decreases mean linear intercepts (MLI) measurements in FADH₂-treated lungs following chronic 85% O₂ exposure; *=P < 0.05 in three litters (n = 3) with a minimum of 12 regions of interest (ROIs) analyzed for each group.

DISCUSSION

Our study shows that exposing newborn C57Bl6 mice to chronic 85% O₂ from birth to PN day 11 depletes lung GSH, increases lung GSSG, and attenuates the lung’s ability to clear edematous fluid. The dysmorphic phenotype observed in hyperoxic newborn lungs (see Fig. 4) resembled neonatal lungs insufflated with GSSG and amiloride (see Fig. 3). In view of previous work that has established inhibitory effects of GSSG (29) and amiloride (for review, see Ref. 48) on lung ENaC activity, we accordingly hypothesize that increasing ENaC activity under pro-oxidizing conditions could attenuate high oxygen-induced lung injury in newborn lungs. In support of this hypothesis, our results show that Scnn1b mice (that overexpress lung ENaC) exhibited less edematous and fibrotic lung injury (compared with their WT littermates) following 11 days of chronic 85% O₂ exposure. Because FADH₂ is an important cofactor in GSSG recycling, we also hypothesized that FADH₂ treatments could also protect the lungs from high oxygen-induced injury (presumably by limiting GSSG bioavailability). In support of this hypothesis, our results show that FADH₂ does indeed lower the GSH/GSSG redox potential (E_h) and attenuate lung injury caused by chronically high FI_O2 exposure. Additional studies are needed to delineate the signal transduction pathway between FADH₂ and ENaC activity to establish a direct causal connection between GSH/GSSG E_h and transepithelial transport in the lungs. In summary, the data indicate that regulating GSH/GSSG redox balance and lung ENaC activity may be important therapeutic strategies for preventing high oxygen-induced lung injury in preterm infants. The following discussion provides additional context for our current work.

Modeling Preterm Lung Injury

This study corroborates and supports the use of animal models as a surrogate for preterm lung injury. A FI_O2 of 85% for the first 11 days of life in C57Bl6 mice caused fewer alveoli of larger size to develop in the lungs. This phenotypic outcome is in line with the body of literature showing that 40%–85% O₂ over the first 14 days of neonatal lung development is effective in recapitulating key pathological hallmarks of BPD (for review, see Ref. 49). Postnatal day 11 was also favored as an end-point for chronic 85% O₂ exposure in our studies because secondary septation is prominent during this developmental stage (50) and lung ENaC subunits have been shown to be expressed at sufficient levels to drive net fluid clearance (51) at this time point. These developmental milestones occurring near PN day 11 allow us to assess lung development as well as lung alveolar fluid clearance in our studies. Suspending oxygen exposure on PN day 11 allowed for pharmacological intervention of oxidative stress (using FADH₂) and evaluation of its effects that occur throughout alveologenesis [up to PN day 18 (50)]. Episodic delivery of 400 µM GSSG or 1 µM amiloride (intranasally administered in separate groups between PN days 1 and 11) also resulted in significant arrest of alveolar development in neonatal C57Bl6 mice. These results are consistent with our previous studies that showed GSSG and amiloride each significantly decreases lung ENaC activity (29, 52). Together, our data support the hypothesis that high oxygen-induced increases in GSSG leads to ENaC dysfunction and contributes to the pathogenesis of preterm lung injury. To further support this hypothesis, we evaluated (and then altered) the GSH/GSSG redox potential (E_h) in the neonatal mouse lung.

The Role of the GSH/GSSG Axis in Neonatal Hyperoxic Lung Injury

In vivo, the GSH/GSSG E_h ranges from −260 mV to −150 mV (53). Under pro-oxidizing conditions, the GSH concentration decreases; although GSSG is increased, it can be recycled to maintain redox balance (under favorable conditions). When E_h values are low, the GSH/GSSG redox pair serves as a better reductant, whereas more positive E_h values are associated with excessive oxidative stress, resulting in apoptosis, and several pathological disorders (53–55). In our studies, chronic exposure to 85% FI_O2 significantly increased the GSH/GSSG E_h by decreasing GSH availability and increasing the pool of GSSG. Specifically, E_h increased from −147.52 ± 5.39 mV to −111.03 ± 21.25 mV by PN day 11 of chronic 85% O₂ exposure. Our calculated GSH/GSSG E_h values are indicative of an immature antioxidant system, and are consistent with E_h values reported in the literature that similarly evaluates oxidative stress and epithelial cell proliferation in neonates (56, 57). Of particular interest, the increase in E_h correlated with a significant decrease in lung fluid clearance and is consistent with our previous study that shows an inverse relationship between (increased) E_h values and (decreased) ENaC activity (29). These outcomes also support the hypothesis that high oxygen-induced increases in GSSG leads to ENaC dysfunction, albeit we recognize that in vivo responses to oxidative stress are complex and could impact many signaling mechanisms that may also contribute to the pathogenesis of newborn lung injury.

Our studies also evaluated whether administering daily dosages of FADH₂ (after removal from chronic O₂ exposure) would benefit lung health. FAD is a redox sensitive coenzyme involved in the reversible oxidation of GSSG. We discovered that 10 µM FADH₂ improved alveolarization, but had no effect on neonatal lung fluid clearance and increased collagen deposition in hyperoxia-exposed newborn mice. This outcome raises the distinct possibility that FADH₂ effects on lung development occurs separately from AFC, and that FADH₂ change in E_h does not impact lung ENaC activity. Although the precise mechanism of effect requires further investigations, it is possible that AFC may have reacclimatized during the FADH₂ treatment period on PN days 11–18 at room air. This is a plausible explanation for the lack of FADH₂ effect on AFC, since FADH₂ alone did not alter AFC in normoxic WT pups (data not shown). We have observed FADH₂’s effect on increasing lung fluid clearance in adult C57Bl6 mice (6-wk old) that had been housed under chronic 85% O₂ for three consecutive days (data not shown). Moreover, intraperitoneal injection of FADH₂ into H1N1-infected adult C67Bl6 mice has been shown to ameliorate lung edema [measured by wet-to-dry lung weight ratios (58)]. The amount of FADH₂ administered in our study was established based on reports of in vivo efficacy and FAD pharmacokinetcs; low amounts of FADH₂ (1 µM) has been shown to maximally activate glutathione reductase (GR) within 10 min (59) and 100 mg/kg/day has been shown to lower lung injury scores in H5N1-infected mice (58). At this juncture of our research, we surmise that FADH₂ activation of GR (and its consequential increase in GSH bioavailability) may improve neonatal lung development but is not sufficient to reverse edematous and fibrotic lung injury caused by chronic high oxygen exposure.

Understanding ENaC Dysfunction in Newborn Lung Injury

To better understand the role of lung ENaC in newborn lung injury, we housed congenic Scnn1b mice under chronic 85% FI_O2 from birth to PN day 11. Scnn1b mice have airway-targeted overexpression of β-ENaC subunit and were generated by crossing the original Scnn1b-Tg mice generated on a mixed background (60) to inbred C57Bl/6N mice. Parent strains used to establish the Scnn1b colony at the University of Utah were provided by Dr. Livraghi-Butrico (UNC Chapel Hill). Approximately four transgenerations of Scnn1b mice were needed to support this study. The phenotypic evaluation of neonatal lung from each generation did not differ significantly from the parent strains; we observed similarly high levels of % survival, β-ENaC subunit overexpression, protection from airspace enlargement, and absence of hemorrhage (42). Mucus plug formation is profound in adult Scnn1b mice (42, 43), but is not prominent in PN day 11 pups (as shown in ABG, RPM, and H&E measurements; see Fig. 5 and Refs. 42, 43).

Although the primary objective of our study was to determine whether redox regulation of ENaC plays a central role in high oxygen-induced newborn lung injury, many important observations (made in the Scnn1b mice and their WT littermate controls) warrant additional discussion, interpretation, and clarification: 1) A difference in the rate of AFC between Scnn1b mice and their WT littermates was only observed when the pups were stressed under chronic 85% O₂ and not at room air (data not shown). The explanation for this observed effect is not completely clear, especially since O₂ appears to be associated with an increase in both ENaC expression (61), as well as with NF-κB expression (61) which inhibits ENaC activity (62, 63). 2) In Fig. 3A, we show that 11 days of intermittent insufflation of amiloride into WT neonatal lungs lead to an increase in MLI measurements when compared to WT mice insufflated with the vehicle. Astute readers may recognize that likewise, Scnn1b also exhibited greater MLI (compared with WT littermate controls) under room air (Fig. 4D) and may therefore, find it difficult to reconcile these seemingly similar outcomes (due to a decrease and increase in ENaC function in Figs. 3A and 4D, respectively). We purport that amiloride inhibition of ENaC activity in Fig. 3A contributes to defective gas exchange, and as a consequence, leads to disturbances in alveolarization of neonatal lungs (64). From this viewpoint, the associated increase in MLI observed in Fig. 3A supports the notion that ENaC plays a critically important role in lung development. Indeed, early death due to defective lung fluid clearance has been reported in EnaC-deficient mice (65). MLI measurements can also be used as an index for characterizing enlarged airspace in emphysema and the associated severity of structural destruction in the lung. As such the larger MLI values in Scnn1b mice versus WT littermates at room air (shown in Fig. 4D) are likely due to the loss of mature alveoli due to hyperactive ENaC and increased inflammation; both are well characterized phenotypes of the Scnn1b mouse model of lung injury (66–68). 3) Even closer inspection of MLI measurements obtained from Scnn1b mice housed under 21% O₂ versus 85% O₂ (shown in Fig. 4D) could also be interpreted to suggest at least a trend toward improved lung morphology in Scnn1b mice following 11 days of chronic oxygen exposure. Although this interpretation represents an intriguing possibility that Scnn1b mice may benefit from O₂ exposure, it is not entirely supported by arterial blood gas measurements in Fig. 5, A and B which shows that Scnn1b mice have Pa_O2 and Pa_CO2 within normal limits (i.e., are not hypoxic). Evaluation of MLI values measured from Scnn1b and WT mice housed under 21% versus 85% O₂, however, does lead us to conclude that ENaC overexpressing mice have improved lung architecture when compared with their WT littermates under pro-oxidizing conditions (modeled by chronic 85% O₂ exposure). Moreover, an increase in hydroxyproline (a major component of collagen) was observed when comparing WT mice housed under 21% O₂ to WT mice housed in 85% O₂. Similar comparison of hydroxyproline in Scnn1b mice in normoxic versus hyperoxic chambers showed no significant difference. 4) We recognize that many pathways can lead to the pathogenesis of pulmonary fibrosis, and that biological crosstalk confounds interpretation of ENaC’s role in attenuating lung injury. Although results from our studies indicate that enhancing net Na⁺ and fluid clearance would have beneficial effects on oxidatively stressed newborn lungs, previous studies have shown that knocking out neural precursor cell expressed developmentally downregulated protein (Nedd4-2; a ubiquitin ligase that regulates endocytosis and lysosomal degradation of ENaC) increased ENaC activity and led to increased spontaneous fibrosis even in the absence of triggers such as high oxygen (69, 70). Although Nedd4-2 also leads to hyperactive sodium transport, the fibrotic effects of the Nedd4-2 knock out mice may be due to altered TGFβ signaling. In our current study, we focus on the disease progression that arises from glutathione depletion and GSSG inhibition of ENaC activity under oxidative stress (distinct from TGFβ signaling). Evaluation of total changes in the lung proteome under 21% O₂ and 85% O₂ will provide more insight into direct and indirect regulators involved in the signal transduction pathway(s) that controls lung ENaC activity and, hence, impacting the pathogenesis of lung injury. Indeed, this important area of investigation is currently underway in our laboratories.

GRANTS

This work is supported by R01HL137033 awarded to M. N. Helms. R. Paine is supported by VA Merit Grant 5I01BX001777. P. N. Mimche received funding from R01AR076489, R21DK115991, and the Mark Flapan Award from the Scleroderma Foundation. T. G. Liou received funding from R01 HL125520, the CF Foundation, Bethesda, Maryland (Grants LIOU13A0, LIOU14P0, LIOU14Y4, and LIOU15Y4), the Ben B. and Iris M. Margolis Family Foundation of Utah, and the Claudia Ruth Goodrich Stevens Endowment Fund at the University of Utah, Salt Lake City, Utah.

DISCLOSURES

During the course of the study, T. G. Liou received other support for performing clinical trials from Abbvie, Inc; CFF Therapeutics, Inc.; Corbus Pharmaceuticals Holdings, Inc; Genentech, Inc; Gilead Sciences, Inc.; Laurent Pharmaceuticals, Inc; Nivalis Therapeutics, Inc; Novartis Pharmaceuticals; Proteostasis Therapeutics, Inc; Savara, Inc; and Vertex Pharmaceuticals, Inc. None of these sponsors were involved in any way with the performance of the current study. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

G.J.G., P.N.M., R.P., W.-J.Q., and M.N.H. conceived and designed research; G.J.G., P.N.M., and M.N.H. performed experiments; G.J.G., P.N.M., R.P., T.G.L., W.-J.Q., and M.N.H. analyzed data; G.J.G., P.N.M., R.P., T.G.L., W.-J.Q., and M.N.H. interpreted results of experiments; G.J.G., R.P., and M.N.H. prepared figures; G.J.G., R.P., T.G.L., and M.N.H. drafted manuscript; G.J.G., P.N.M., R.P., T.G.L., W.-J.Q., and M.N.H. edited and revised manuscript; G.J.G., P.N.M., R.P., T.G.L., W-J.Q., and M.N.H. approved final version of manuscript.