Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Infants with respiratory distress syndrome are frequently given high concentrations of oxygen to maintain adequate tissue oxygenation. Exposure to hyperoxia in the neonatal period, however, is a major contributing factor to the development of bronchopulmonary dysplasia (BPD) (1). BPD is characterized by impaired lung growth, airway inflammation, and increased airway resistance. Histologically, exposure to hyperoxia can lead to a decreased number of alveoli and a reduction in lung surface area (2). Neonatal mice and rats exposed to high concentrations of oxygen have been shown to develop large terminal air spaces and decreased numbers of septa, similar to lesions that are present in the lungs of infants with BPD (3). Because the majority of lung growth in children occurs in the first 2 yr of life, exposure to high levels of oxygen can affect ultimate lung growth and function (6). We were interested in determining whether apoptosis was the predominant form of cell death in neonatal lung subjected to hyperoxic stress. During both normal prenatal and postnatal lung development, apoptosis has been shown to be involved in lung remodeling in both rat lung and human fetal lung explants (7, 8). Prenatally, mesenchymal cells have been shown to undergo apoptosis, presumably as a mechanism to thin the septa and establish an adequate pulmonary alveolar-capillary interface (7). Postnatally, apoptosis appears to be involved in ridding the lung of excess numbers of fibroblasts and type 2 epithelial cells to increase lung surface area (8). Apoptosis requires the activation of the caspases, a family of cytosolic proteases. During apoptosis Bax, a proapoptotic gene, releases cytochrome c from the mitochondria, which in turn induces caspase activation (9). Morphologically, the cell undergoes a series of changes before death, including cell shrinkage, chromatin condensation, active membrane blebbing, and the formation of apoptotic bodies. In contrast to necrosis or nonapoptotic cell death, apoptosis is believed to occur in the absence of inflammation (10). It is not known whether apoptosis is in part responsible for the impaired lung growth found in developing lung exposed to hyperoxia. Previous studies in adult mice have shown that apoptosis is significantly increased in the lungs of mice exposed to hyperoxia and is associated with increased mortality (11, 12). Ward and colleagues found that transgenic mice that overexpress interleukin (IL)-6 had increased survival (12). In their model, IL-6 protected lung from the effect of hyperoxia and this was associated with a decrease in apoptotic cell number in the lung. They also found that the transgene-positive mice had an increase in Bcl-2 protein. Others have found that Bcl-2 and Bcl-xL are able to block the cytochrome c mitochondrial release induced by Bax (9). Neonatal mice have previously been shown to be more resistant to hyperoxic stress (13). This was partially secondary to an increase in lung antioxidant enzyme activity and a difference in chemokine response to hyperoxic stress when compared with adult animals (13). Also, the form of cell death that neonatal murine lungs undergo in response to hyperoxic stress may influence survival in the neonatal mouse; for example, apoptosis versus a nonapoptotic cell death. In the present study we were interested in determining whether apoptosis was a primary mechanism of cell death in the lungs of newborn mice exposed to hyperoxic stress. Therefore, we studied the effect of hyperoxia on the lungs of newborn mice in an effort to understand the role of apoptosis during hyperoxic stress in the neonate.

We previously found that 3.5-d-old CD-1 mice raised in 92% oxygen had impaired lung growth, as demonstrated by a 3-fold decrease in bromodeoxyuridine uptake (14). Using this model, we isolated lung sections from mice raised in 92% oxygen for 3.5, 4.5, and 5.5 d, and compared these with the lungs of mice raised in room air. Apoptotic cells in the lung were identified in situ using terminal deoxyribonucleotidyl transferase-mediated deoxyuridine triphosphate-biotin nick-end labeling (TUNEL).

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

RNA Extraction and Northern Blot Analysis

Timed pregnant CD-1 mice (Charles River, Wilmington, MA) were placed in a hyperoxic chamber at 18.5 d of gestation. Mice were born into hyperoxic conditions (92% FiO₂) the following morning. Mothers were rotated every 12 to 24 h to prevent death from acute oxygen toxicity. These experiments were done according to the animal protocol approved by the Animal Care Use Committee of the Johns Hopkins University School of Medicine. Lung tissue was harvested at specific time points from both hyperoxic-exposed and age-matched control neonatal mice raised in room air. RNA was isolated by the RNAzol method. Total RNA was used for Northern blot analysis.

The quantity of 15 to 20 µg of total RNA was run on a 1% agarose/formaldehyde gel (equal amounts were loaded per blot), transferred to a filter (genescreen plus; Dupont, Boston, MA), and baked for 2 h at 80°C in a vacuum oven. Murine Bax (ATTC clone) and -actin (gift of Dr. S.-J. Lee, Johns Hopkins University) complementary DNA clones were labeled with ³²P and random primed as previously described (15). Blots were hybridized with the solution of 5× saline sodium phosphate ethylenediamenetetraacetic acid; 10% dextran sulfate; 50% formamide; 1% sodium dodecyl sulfate (SDS); 200 mg/ml salmon DNA; and 0.1% each of bovine serum albumin, ficoll, and polyvinylpyrrolidone, at 42°C overnight; then washed in 0.1% SDS and 0.2% saline sodium citrate five times for 20 min each time at 55°C. RNA signals were quantified using a phosphorimager program (ImageQuant v3.2).

Western Blot Analysis

Whole-cell lysates were solubilized in 2% SDS. Protein concentrations were determined using the Biorad D_C protein assay kit (Bio-Rad Laboratories, Hercules, CA). Lysates were loaded and run on a 12% SDS-polyacrylamide gel. Each protein gel contained equal amounts of protein per lane (20 µg/lane). Gels were then transferred to nitrocellulose and Western blot analysis was performed using an anti-Bax monoclonal antibody (sc-7480; Santa Cruz Biotechnologies, Santa Cruz, CA) at 1:100 dilution in 5% blotto, phosphate-buffered saline (PBS) and .05% Tween overnight at 4°C. The blots were then washed three times in PBS-Tween, incubated with a mouse immunoglobin, horseradish peroxidase-linked whole antibody (RPN 2108; Amersham, Arlington Heights, IL) for 1 h at a 1:500 dilution, then washed and developed using enhanced chemiluminescence (RPN 2106; Amersham).

Cell Lines

Murine lung bronchus cells were obtained from American Type Culture Collection (#CRL-6382; Rockville, MD) and grown to 50% confluence. The cell line is from normal murine lung and bronchus. The cells were grown in Dulbecco's modified Eagle's medium (GIBCO BRL Life Technologies, Rockville, MD) with 10% fetal bovine serum, penicillin /streptomycin, and 4 mM L-glutamine. The cells were seeded at 110,000 cells per 25-mm flask. Cells were either placed in a modular incubator chamber (Billups-Rothenberg, Del Mar, CA) and exposed to 95% O₂ and 5% CO₂, or placed in room air and 5% CO₂. Cells were gassed every day and media were changed daily. Cells were then harvested at specific time periods for counting, protein, and RNA extraction for Western and Northern blot analysis.

Assays for Apoptosis

The lungs of neonatal mice were perfused with 4% paraformaldehyde through the trachea until the lungs were fully distended, then placed in 4% paraformaldehyde overnight, dehydrated in ethanol, paraffin-embedded, and cut into 10-µm sections. Apoptosis was determined by in situ labeling using terminal deoxynucleotidyl transferase (In situ cell death detection kit, fluorescein; Boehringer Mannheim, Indianapolis, IN; catalog #1-684-795). Cells were then counterstained with 4,6-diamidino-2-phenylindole (DAPI) to identify the nuclei. Fluorescein isothiocyanate (FITC)-labeled apoptotic cells were detected under fluorescent microscopy. Under ×40 magnification, three areas of lung parenchyma were systematically and randomly photographed from each animal. Murine lung bronchus cells were grown on glass coverslips and kept in room air and 5% CO₂ or exposed to 95% O₂ and 5% CO₂ for 72 h. The cells were then stained for apoptosis using an Oncor apoptag peroxidase kit (S7100-kit; Oncor, Gaithersburg, MD). Murine lung bronchus cells were grown in T25 flasks and left in room air and 5% CO₂ or 95% O₂ and 5% CO₂ for 72 h. Cells were then counted and annexin V binding was determined using TACS Annexin V-FITC (catalog #TA4638; R&D Systems, Minneapolis, MN), following the manufacturer's protocol. Cells were stained with propidium iodide and annexin V-FITC, which allowed determination of viable, early apoptotic, or late apoptotic/ necrotic cells.

Statistical Analysis

Statistical calculations were performed using the SPSS 8.0 statistical package for Windows (SPSS, Chicago, IL) and Microsoft Excel 97 statistical package for Windows. Differences in measured variables between experimental and control groups were determined using comparison of the means using one-way analysis of variance (ANOVA) statistical calculations and Student's t test (two-tailed). Statistical difference was accepted at P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Survival of Neonatal Mice in Hyperoxia

At 3.5 d of life, neonatal mice born and raised in 92% oxygen had a minimal mortality rate (range, 0 to 11%). Mortality within the litters, however, increased with each subsequent day of exposure to hyperoxia. At 5.5 d of life mice exposed to hyperoxia had mortality rates between 14 and 34%, and at 6.5 d of life mice exposed to hyperoxia had mortality rates ranging between 18 and 47%. Survival was significantly higher in mice at 3.5 d of hyperoxia compared with mice at 6.5 d of hyperoxia (P < 0.03) (Figure 1).

View larger version (15K): [in this window] [in a new window]   Figure 1.   Percentage of neonatal mice surviving in 92% hyperoxia. Mice were born into 92% hyperoxia. Each group consisted of two litters that were fed by the same group of mothers. Average size of each group was 29 animals ± 2.6. The mortality rate at 6.5 d of hyperoxia was significantly greater than the mortality rates at 2.5 and 3.5 d (P < .02, P < .03, respectively). Comparisons of the means were done by one-way ANOVA; n = 3.

We next weighed the mice and found that the weights of mice born and raised in 92% hyperoxia were significantly lower at 4.5 d (O₂, 2.35 g ± 0.5; room-air mice, 2.92 g ± 0.51; P < 0.00002) and 5.5 d (O₂, 2.67 g ± 0.64; room air, 3.14 g ± 0.47; P < 0.00003) when compared with mice raised in room air (Figure 2). These findings are similar to those of Northway and colleagues, who found a significant difference in body weights between O₂-treated and room-air mice after 120 h of 100% hyperoxia in C57BL/Ka neonatal mice (16).

View larger version (9K): [in this window] [in a new window]   Figure 2.   Comparison of body weights between neonatal mice in hyperoxia and room air. Mice in both room air and hyperoxia were weighed daily. At 4.5 d the average weight of the oxygen-treated mice (2.35 g ± 0.5) was significantly less than that of the room-air mice (2.92 g ± 0.51, *P < .00002). At 5.5 d the average weight of the oxygen-treated mice (2.67 g ± 0.64) was significantly less than that of the room-air mice (3.14 g ± 0.47; **P < .00003). The 9.5-d time point represents mice exposed to 92% oxygen for 6.5 d and allowed to recover in room air for 3 d. There was no significant difference in body weight between the oxygen-treated/recovery group and the room-air group. Error bars represent standard error of the mean. At each time point from .5 to 5.5 d the weights of each litter were averaged. Each time point had between two and eight litters. For the 9.5-d mice: O2/recovery group n = 6, room air n = 5.

We then allowed mice exposed to 92% hyperoxia for 6.5 d to recover in room air for 3 d. Their weights were compared with the weights of mice raised in room air for 9.5 d. We found no significant difference in body weights between the two groups (6.5 d O₂/3 d room air, 3.93 g ± .61; room air, 4.0 g ± 0.75; P < .87).

Apoptosis in Neonatal Murine Lung Exposed to Hyperoxia

A TUNEL assay was performed to determine the number of cells undergoing apoptosis in the lungs of neonatal mice exposed to hyperoxia for 3.5, 4.5, and 5.5 d or room air. From the lungs of each animal, FITC-labeled cells, representing apoptotic cells, were counted from three randomly selected lung fields (Figure 3). An increase in the number of cells undergoing apoptosis (apoptotic cells/lung field) was found in the hyperoxic-treated group at 3.5 d (O₂-treated, 18.3 cells ± 11.7; room air, 4.9 cells ± 1.8; P < .0002), 4.5 d (O₂-treated, 18.2 cells ± 8.0; room air, 4.1 cells ± 2.8; P < .00001), and 5.5 d of age (O₂-treated, 30.3 cells ± 13.8; room air, 1.9 cells ± 1.3; P < .00001) which was significantly greater than the number of apoptotic cells found in the lungs of mice of the same age raised in room air (Figure 4). The number of cells undergoing apoptosis increased the longer the exposure time to oxygen, with the number of apoptotic cells in 3.5-d oxygen-treated lung being significantly less than the number of apoptotic cells in 5.5-d oxygen-treated lung (P < .03).

View larger version (80K): [in this window] [in a new window]   Figure 3.   Apoptotic cells in the peripheral airways of hyperoxic-exposed neonatal mice. On the left, 10-µm sections of paraffin-embedded peripheral lung from newborn mice born and raised in 92% hyperoxia for either 3.5, 4.5, or 5.5 d; on the right, peripheral lung sections from mice born and raised in room air for 3.5, 4.5, or 5.5 d. Apoptotic cells are labeled with FITC (green) and nuclei are labeled with DAPI (blue). The white arrows point to apoptotic cells in the peripheral lung. An increase in number of apoptotic cells is found in the hyperoxic-exposed lung.

View larger version (74K): [in this window] [in a new window]   Figure 4.   Resolution of apoptosis in the lungs of neonatal mice previously exposed to hyperoxia. Top panel: peripheral lung from 9.5-d-old mouse raised in room air. Bottom panel: peripheral lung from 9.5-d-old mouse exposed to 92% hyperoxia for 4.5 d and allowed to recover for 3 d in room air. No increase in apoptotic-labeled cells was found in the lung previously exposed to hyperoxia compared with the room-air control lung. The alveolar air spaces of the previously O2-treated lung appear larger and less complexed than the room-air lung.

In mice exposed to 92% oxygen for 6.5 d and allowed to recover for 3 d in room air, no difference in the number of apoptotic cells in the lungs was found compared with mice raised in room air for 9.5 d (6.5 d O₂/3 d room air, 2.3 cells ± 0.9; room air, 1.8 cells ± 0.8; P < .2). We did, however, notice that the size of the alveolar air spaces in the mice treated with oxygen and recovered in room air appeared larger and less complex compared with mice raised in room air, suggesting an inhibition of lung growth in the oxygen-treated animals (Figure 5). This is similar to findings reported by Randell and associates, who showed, using morphometric measurements, that rats receiving 7 d of 95% hyperoxia at birth had abnormally enlarged alveolar ducts and decreased alveolar surface area at 40 d of age (17).

View larger version (11K): [in this window] [in a new window]   Figure 5.   Apoptosis in neonatal murine lung exposed to 92% hyperoxia and room air. Apoptotic cells, in neonatal murine lung exposed to 92% hyperoxia and room air, were identified by in situ labeling using TUNEL. TUNEL cells were counted from three randomly selected areas of lung under fluorescent microscopy at ×40 magnification. Significant increases in the number of apoptotic-labeled cells were found at 3.5, 4.5, and 5.5 d in the oxygen-treated lung compared with room-air murine lung (*P < .0002, **P < .00001, and ***P < .00001). *9.5: 9.5-d oxygen-treated mice were exposed to 92% O2 for 6.5 d and allowed to recover in room air for 3 d. No significant difference in the number of apoptotic-labeled cells was found in the lungs of these mice compared with room-air controls. Error bars represent standard deviations. n = 6 for 3.5-d O2-treated mice and 4.5- and 5.5-d room-air mice; n = 5 for 4.5- and 5.5-d O2-treated mice; n = 4 for 3.5-d room-air mice; and n = 3 for 9.5-d-old mice.

Bax Expression in Neonatal Murine Lung and Murine Lung Bronchus Cells

Lung from mice treated with 92% oxygen was analyzed for Bax messenger RNA (mRNA) expression and compared with room-air controls. The proapoptotic gene Bax has previously been shown to be induced during hyperoxic injury in adult murine lung (11). We found that Bax mRNA levels were induced in neonatal murine lung at as early as 1.5 d of hyperoxic exposure, with levels increasing 2.5-fold above baseline by 4.5 d of hyperoxic exposure (Figure 6a). After 2 d of room-air recovery, however, Bax mRNA levels returned to baseline (Figure 6b). This suggests a correlation between increasing numbers of apoptotic cells and increasing mRNA levels of Bax during exposure to hyperoxia in neonatal lung.

View larger version (24K): [in this window] [in a new window]   Figure 6.   Induction of Bax mRNA levels in murine lung exposed to 92% hyperoxia. (a) Total RNA from neonatal murine lung was isolated from mice born and raised in either 92% hyperoxia or room air, at different time points. Northern blot analysis was performed using Bax and -actin probes. Levels of Bax mRNA were normalized to -actin levels. By 4.5 d, mRNA levels of Bax were increased 2.5-fold above room-air controls. (b) mRNA levels of Bax returned to baseline in the lungs of mice exposed to 4.5 d of hyperoxia and allowed to recover in room air for 2 d.

We then exposed murine lung bronchus cells to 95% O₂ and 5% CO₂ and found that cells underwent growth arrest by 48 h of hyperoxia (Figure 7a). We then evaluated Bax protein levels in room air and hyperoxia. At 24 and 48 h, Bax protein levels were increased in cells exposed to hyperoxia; however, at later time points Bax protein levels appeared to decrease. In room-air cells Bax protein was increased at later time points, possibly representing an increase in apoptosis from cell overgrowth (Figure 7b). Using a TUNEL assay, we found at 72 h a population of cells from both room-air and hyperoxic-treated cells that were apoptotic (Figure 8a). The majority of cells, however, did not stain for apoptosis (data not shown). We next used annexin V binding to determine whether necrotic cell death was the primary mode of cell death in oxygen-treated murine lung bronchus cells. Annexin V binding was done to determine the percentage of viable, apoptotic, or necrotic/ late apoptotic cells in room-air and hyperoxic-treated cells. The viability of room-air cells by annexin V binding at 72 h was 60.4% ± 7.0%, and for O₂-treated cells 29.1% ± 1.4% (P < .002). In the oxygen-treated cells, 43.3% ± 18.5% of the cells were undergoing necrotic cell death by 72 h, in contrast to only 11.1% ± 6.1% of the room-air cells (P < .05) (Figure 8b).

View larger version (28K): [in this window] [in a new window]   Figure 7.   Growth of murine lung bronchus cells in 95% hyperoxia and room air. (a) Growth of murine bronchus cells in room air ( filled diamonds) and 95% O2 and 5% CO2 (gray squares). Cells in hyperoxia underwent growth arrest by 48 h of 95% oxygen. (b) Protein expression of Bax and p21 in murine lung bronchus cells exposed to hyperoxia and room air.

View larger version (64K): [in this window] [in a new window]   View larger version (13K): [in this window] [in a new window]   Figure 8.   Annexin V binding and apoptosis in murine lung bronchus cells. (a) Apoptotic cells in both hyperoxic and room-air murine lung bronchus cells. Arrows point to apoptotic cells. Left panel: cells grown in 95% O2 and 5% CO2 for 72 h. Right panel: cells grown in room air and 5% CO2 for 72 h. (b) Annexin V binding in murine lung bronchus cells grown in 95% hyperoxia and room air for 72 h. *P < .002, **P < .05, n = 3.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this study we found that high concentrations of oxygen led to an increase in apoptosis in the peripheral lung of neonatal mice. The number of cells undergoing apoptosis increased the longer the exposure to oxygen, and the number of apoptotic cells correlated with the degree of lung injury. The increase in apoptosis during hyperoxia was associated with an induction of Bax mRNA expression. Although it was not possible to identify specific cell types undergoing apoptosis, the apoptotic cells were primarily found along the alveolar surface of the peripheral airways. We also studied the effect of hyperoxia on murine lung bronchus cells. These cells underwent both growth arrest and cell death when exposed to high concentrations of oxygen. The murine lung bronchus cells, however, exhibited primarily necrotic cell death when exposed to 72 h of hyperoxia.

Previous studies have reported varying results when they have examined the role of apoptosis during hyperoxic exposure. In mature lung and in alveolar type 2 epithelial cells grown in culture, exposure to hyperoxia has been shown to induce apoptosis (11, 12, 18). Mantell and coworkers, using adult murine lung, found as we did that the number of apoptotic cells in lung exposed to hyperoxia correlated well with the extent of lung injury (19). Hyperoxic injury, however, may cause both apoptotic and necrotic cell death (19). We found that Bax mRNA levels in neonatal lung increased and correlated with the number of apoptotic cells in neonatal lung exposed to hyperoxia. These observations, however, do not necessarily mean that apoptosis is the primary cause of cell death in neonatal lung exposed to hyperoxia. The majority of cells in the peripheral lung exposed to hyperoxia were not apoptotic by TUNEL. Indeed, death by necrosis may be the major mode of cell death caused by hyperoxia. This is supported by our findings in murine lung bronchus cells. We found that although some cells exposed to hyperoxia underwent apoptosis by TUNEL staining, the majority of cells were not apoptotic. Bax protein levels were initially increased in murine lung bronchus cells exposed to hyperoxia, however protein levels appeared to decrease the longer the exposure to oxygen. Further, as cells in room air grew to confluency, they had increases in Bax protein levels. The increase in Bax protein in the room-air cells may correlate with apoptosis from cell overgrowth.

Using annexin V binding in the murine lung bronchus cells, we found that the majority of cells were undergoing necrotic cell death when exposed to hyperoxia for 72 h. This suggests that, at least with longer exposure to high concentrations of oxygen, necrosis is a major cause of cell death in these cells. Kazzaz and colleagues also found that A549 cells, a distal lung epithelial cell line, underwent necrotic cell death when exposed to hyperoxia (20). Therefore, it is likely that a combination of apoptosis and necrosis is occurring in neonatal lung exposed to high concentrations of oxygen. Factors that may influence whether a cell undergoes apoptosis include adenosine triphosphate (ATP) levels in the cell and the type of insult to the cell. It has been shown that the level of cellular ATP can determine the balance between apoptotic and necrotic cell death (21). Allen and White found that cellular ATP levels were decreased in A549 cells exposed to 95% hyperoxia and depleted of glucose (22).

Although we have shown that apoptosis is increased in neonatal lung exposed to hyperoxia, it is not clear how important this is to lung development. An increase in apoptotic cells in the lung during development may affect ultimate growth of the lung, whereas an increase in apoptosis in adult lung may not. Under normal circumstances, apoptosis is involved in remodeling of the lung prenatally and postnatally (7). The timing of apoptosis in the development of many organs has been shown to be important. In murine lung, postnatal lung growth is delayed until after Day 2 of life (16, 23). Maximal cell proliferation takes place at Day 4, then cell proliferation declines until Day 10 (23). Scanning electron microscopy studies of neonatal murine lung have shown that an increase in alveolar number occurs until Day 14, at which time the lung begins to histologically resemble the adult lung (24). It is possible that an increase in apoptosis from hyperoxia, however minimal, during this critical phase of development may interfere with normal lung growth. Hyperoxic exposure is a major factor in the development of BPD, and infants with BPD have been shown to have markedly decreased numbers of alveoli compared with age-matched controls (4). We observed in neonatal mice initially exposed to hyperoxia and then allowed to recover in room air that their lungs had larger and less complex alveolar air spaces compared with mice raised in room air, suggesting that in our model hyperoxia is having an adverse effect on alveolar growth.

Nutritional status may also affect lung growth during the neonatal period. We did find a significant difference in survival between infant mice exposed to hyperoxia for 3.5 d and those exposed for 5.5 d. This difference may represent the effect of hyperoxia alone or the combination of hyperoxia and poor nutrition. Poorly nourished mice have been reported to have poorer outcomes when exposed to hyperoxia (25). Adequate nutrition depends on the ability of the mother to supply milk, the size of the litter, and the ability of the pup to nurse adequately. Frank and colleagues found that normally nourished rat pups had a 73% survival after 7 d of 95% hyperoxia, but in undernourished pups survival dropped to 44% (13). Therefore, both hyperoxia and poor nutrition can potentially affect lung growth.

In summary, we found that the number of apoptotic cells as measured by TUNEL is increased in the peripheral airways of neonatal murine lung exposed to high concentrations of hyperoxia. We found that the degree of apoptosis correlates with the duration of hyperoxic exposure and the degree of lung injury. Increases in apoptotic cells in the neonatal lung during a crucial period of lung development may adversely affect future lung growth. Additional studies will be needed to evaluate this further.