Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Eukaryotic cells suffer from oxygen toxicity through the conversion of molecular oxygen to more cytotoxic-reduced species of superoxide anion, hydrogen peroxide, and hydroxyl radical (1, 2). These reactive oxygen species (ROS) cause cellular damage through covalent modifications of lipids, proteins, and nucleic acids. Although all cellular compartments are susceptible to damage by ROS, the nucleus is thought to be especially sensitive to oxygen because it contains a rich source of metals that maintain chromatin structure and act as cofactors for DNA binding of various transcription factors (1). Redox cycling of metals has the potential to catalyze the formation of ROS from molecular oxygen through the Haber-Weiss reaction. Oxidative DNA damage is manifested as formation of dimers, nucleotide loss, and DNA fragmentation (2). Changes in oxidant status resulting in DNA strand breaks are also observed following exposure to ozone, ionizing radiation, and bleomycin. Exposure to high levels of molecular oxygen is also used clinically in adults and neonates suffering from respiratory distress syndrome, who require supplemental oxygen to maintain blood PO₂ levels (3). Unfortunately, hyperoxia injures pulmonary epithelial and endothelial cells and such treatment, although obviously necessary and beneficial, is also associated with increased morbidity and mortality.

Numerous studies have shown that cells protect themselves from oxidative stress through the expression of various antioxidant enzymes that detoxify ROS to less toxic species (4). However, chronic exposure to ROS will ultimately overwhelm the antioxidant capacity of the cell and lead to oxidative damage of essential cellular molecules, such as nuclear DNA. One mechanism by which cells may protect themselves from DNA damage is to couple DNA repair processes with the cell cycle (5). In this hypothesis, cells transiently growth arrest during the cell cycle at G₁ and G₂ phases in order to determine if the preceding M or S phase, respectively, has been processed correctly and to prevent fixation of mutations by repairing damaged DNA. When DNA damage overwhelms the repair processes, it is believed that this signals cells to undergo programmed cell death or apoptosis. The cellular protooncogene p53 is one molecule that plays a central role in regulating the cellular response to DNA damage. Cells that suffer from DNA damage by ionizing radiation, bleomycin, or drugs that inhibit DNA topoisomerase activity accumulate p53 protein, which induces growth arrest in G₁ (6, 7). Higher levels of p53 can also promote apoptosis in some cells (8). Expression of mutant forms of p53 is often observed in tumors, especially from the lung, which have escaped from normal proliferative and apoptotic signals (9). Because p53 regulates cell proliferation and apoptosis, it is clearly involved in regulating whether a cell lives or dies following DNA damage.

Because the principal role of the lung is to facilitate the exchange of gases between the external environment and the blood, pulmonary epithelial cells are constantly exposed to higher levels of oxidant gases than other tissues (10). Although oxidant injury is dependent on the physical and chemical form of the pollutant, as well as duration and concentration of exposure, the distal respiratory and alveolar epithelium are often damaged by exposure to oxidant gases. Chronic exposure to molecular oxygen leads to epithelial cell death, edema, and inflammation that ultimately result in morbidity and mortality (11, 12). Relief of the oxidant stress is often associated with significant epithelial and endothelial cell proliferation and differentiation, which is essential for replacing sick and dying cells that were injured during the exposure period (13, 14). Because hyperoxia alters cell proliferation and survival, the present study was designed to determine whether DNA integrity, p53 expression, and epithelial cell gene expression were altered in lungs of adult mice exposed to > 95% oxygen.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Exposure to Hyperoxia

Normal 8-wk-old male C57Bl/6J mice were obtained from The Jackson Laboratory (Bar Harbor, ME) and housed under pathogen-free conditions for 1 wk prior to exposure. Mice were exposed to room air (controls) or to > 95% oxygen by placing the cages inside a Plexiglas chamber of approximately 32 × 14 × 24 inches. The gas was prewarmed, humidified by bubbling in sterile distilled water, and filtered through a 0.22-µm pore size filter before passage through the chamber at a flow rate of 6 liters/min resulting in 12 complete changes of air per hour. Oxygen concentrations were monitored each day with a miniOXI analyzer from Catalyst Research Corp. (Owings Mills, MD) and maintained at > 95%, and the relative humidity was approximately 50%. Animals were permitted food and water ad libitum and were killed after 72 h of exposure with pentobarbital (65 mg/kg injected intraperitoneally). All exposures were reviewed and approved by the University of Rochester University Committee on Animal Resources. The lungs were exposed, the left lobe was ligated and removed for isolation of RNA, and the right lobes were inflation-fixed for histology through the trachea with 100 mM cacodylic acid, pH 7.4, with 2% glutaraldehyde at 10 cm of pressure for 15 min. Lungs were further fixed for 12 h by immersion in the same buffer, dehydrated through graded alcohol, embedded in paraffin, and 5-µm sections were prepared. Protein homogenates were prepared from some lungs that were first perfused through the right cardiac ventricle with 10 ml of Dulbecco's phosphate-buffered saline (PBS) containing 1% glucose, 0.25 mg/ml gentamicin, and 0.2 mM EGTA followed by lavaging twice with 1 ml of the same lacking EGTA. An average of three to four mice were analyzed per exposure time for each parameter studied.

Immunohistochemistry

Sections were deparaffinized and hydrated prior to blocking of endogenous peroxidase with hydrogen peroxide- methanol. Sections were blocked with 1.5% normal goat serum-0.5% bovine serum albumin and then incubated in primary antibody in humidified chambers. Nonspecific binding was removed by washing extensively in PBS before incubating with biotinylated goat anti-rabbit IgG and avidin-enzyme complex. Sections were reacted with 3,3'-diaminobenzidine (Sigma Chemical Co., St. Louis, MO) and counterstained with hematoxylin or methyl green.

Some modifications to this protocol were used for specific assays. Terminal transferase dUTP nick end-labeling (TUNEL) staining was performed using an ApopTag kit obtained from Oncor (Gaithersburg, MD), which recommended that sections be treated with proteinase K prior to addition of terminal transferase. Immunohistochemical detection of p53 was performed using goat anti-rabbit CM5 antibody (Novacastra, Newcastle, UK), which recommended boiling sections in 0.01 M citrate for 15 min prior to reacting with the primary antibody (15). The specificity of this antisera for p53 has been described previously (16).

RNA Extraction and Analysis

Lungs were homogenized in 4 M guanidine isothiocyanate, 0.5% N-laurylsarcosine, 20 mM sodium citrate, and 0.1 M 2-mercaptoethanol using a Techmar homogenizer (Techmar Company, Cincinnati, OH). RNA was extracted using acid phenol and phase lock columns (5 Prime-3 Prime, Boulder, CO) and resuspended in diethylpyrocarbonate-treated water. Poly(A)⁺ RNA was prepared using oligo(dT) spin columns (Pharmacia Biotech, Piscataway, NJ). The amount of RNA in an aqueous solution was determined by absorbance at 260 nm. RNA was electrophoretically separated on a 1.2% agarose-formaldehyde gel and transferred to Nytran. Blots were prehybridized and hybridized at 65°C in 1% bovine serum albumin, 7% sodium dodecyl sulfate, 0.5 M sodium phosphate, and 1 mM EDTA. Radioactive cDNA probes were prepared by random primer labeling (GIBCO-BRL, Grand Island, NY) and [³²P]dCTP (3,000 Ci/mmol; New England Nuclear, Boston, MA). Hybridized blots were washed briefly in 1% bovine serum albumin, 40 mM sodium phosphate, and 2 mM EDTA twice at room temperature prior to stringent washing in the same buffer at 65°C. Washed blots were exposed to Kodak (Rochester, NY) X-Omat AR film and differences in loading samples normalized to the expression of L32. RNA blots were probed with a 1.7-kb human p53 cDNA (17) and normalized to expression of mRNA for the ribosomal subunit L32, which is not altered by hyperoxia (18). Because L32 protein is a subunit of the ribosome, its mRNA expression is directly proportional to the expression of total ribosomal RNA.

S1 Nuclease Protection Analysis

Details of the probes and methods for S1 nuclease protection analysis have been described previously (19). Briefly, mouse-specific cDNA probes were end-labeled with ³²P by T4 polynucleotide kinase and hybridized to 2 µg of total RNA from mouse lung. Annealed products were digested with S1 nuclease and the protected fragments resolved by electrophoresis through denaturing sequencing gels and visualized by autoradiography of dried gels. Quantitative measurements of undigested probes were made using a PhosphorImager (Du Pont, Boston, MA) and statistical analysis was performed using StatView software (Abacus Concepts, Berkeley, CA).

Western Blot Analysis

Perfused and lavaged lungs were homogenized at 4°C in 50 mM Tris (pH 7.4), 150 mM sodium chloride, 2 mM EDTA, 25 mM sodium fluoride, 25 mM -glycerol phosphate, 0.1 mM sodium vanadate, 1 mM phenylmethylsulfonyl fluoride, 0.2% Triton X-100, 0.3% Nonidet P-40 (NP-40), 10 µg/ml leupeptin, 10 µg/ml pepstatin, and 10 µg/ml aprotinin. Cell lysates were centrifuged at 13,000 rpm at 4°C and the supernatants aliquoted. Approximately 20% of the lysate was stored at 80°C for determination of protein concentration and the remainder was immediately boiled in 2× Laemmli buffer (1× is 62.5 mM Tris [pH 6.8], 2% sodium dodecyl sulfate [SDS], 10% glycerol, 0.025% bromophenol blue, and 5% 2-mercaptoethanol). Protein concentration was determined using a modified Lowry assay (Bio-Rad, Hercules, CA) and bovine serum albumin as a standard. Proteins (50 µg) were electrophoretically separated by size on 10% polyacrylamide-SDS gels and transferred to nitrocellulose. Membranes were blocked in PBS containing 5% nonfat dry milk overnight at 4°C before incubating with the anti-p53 CM5 antibody at a 1:1,000 dilution at room temperature for 1 h. Nonspecific interactions were removed by washing the membranes in PBS containing 0.05% Tween 20 before incubating the blots in goat anti-rabbit peroxidase-conjugated secondary antibody at 1:2,000 (Jackson ImmunoResearch Labs, West Grove, PA). Blots were extensively washed again and conjugates were visualized by chemiluminescence (Amersham, Arlington Heights, IL) and exposure to Kodak Bio-Max film.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Previous studies have shown that adult mice exposed to hyperoxia for 3-4 days suffer from extensive alveolar epithelial and endothelial cell death (12). To determine whether exposure to hyperoxia resulted in DNA damage in mouse lung, adult C57Bl/6J male mice were exposed to room air or > 95% oxygen for 72 h, killed, and their lungs were examined for changes in DNA integrity by TUNEL staining. This assay uses the enzyme terminal transferase to add digoxigenin-conjugated nucleotides to free 3'-hydroxyl groups on DNA, which can then be visualized as a brown stain. Examination of control lungs revealed lightly stained TUNEL-positive nuclei throughout the lung surrounded by numerous TUNEL-negative cells. This low level of TUNEL staining was equivalently observed in terminal bronchioles (Figures 1a and 1c) and throughout the parenchyma (Figures 2a and 2c). Although modest TUNEL staining was observed in epithelium, fibroblasts, smooth muscle, and endothelium, it was not observed in all cells of these lineages or in macrophages. In contrast, lungs from mice exposed to hyperoxia for 72 h exhibited intensely stained TUNEL-positive nuclei surrounded by TUNEL-negative cells. Increased TUNEL staining was observed in the epithelium of the terminal bronchioles (Figures 1b and 1d) and throughout the parenchyma (Figures 2b and 2d). TUNEL-positive nuclei in the parenchyma were observed in cells that contained large intracellular inclusion bodies, indicative of alveolar type II cells, endothelial cells surrounding capillaries and in thin alveolar septae, suggestive of either type I or endothelial cells. Changes in TUNEL staining were restricted to the parenchyma and distal conducting airways and were not observed in histologic sections reacted in the absence of terminal transferase (Figures 1a and 2b, inset). Furthermore, TUNEL staining was not observed in esophagus or the nonapoptotic crypt cells of the intestine of room air- or oxygen-exposed animals (data not shown).

View larger version (112K): [in this window] [in a new window]   Figure 1.   Identification of DNA damage in adult mouse lung exposed to hyperoxia. Adult mice were exposed to room air (a, c) or > 95% oxygen for 72 h (b, d), killed, and their lungs analyzed for the presence of free 3'-hydroxyl ends that were detected immunohistochemically as a brown stain within the nucleus. Filled arrows depict strongly reactive nuclei and thin arrow depicts negative nuclei. Open arrows mark the airway-alveolar junction, which is magnified in (c) and (d). Insets in (a) and (c) are sections from room air- or oxygen-exposed lungs reacted in the absence of terminal transferase. Sections were counterstained with methyl green. Bar in (a) and (c) = 40 µm. Inset magnification is reduced 50%.

View larger version (134K): [in this window] [in a new window]   Figure 2.   Identification of DNA damage in the parenchyma. Adult mice were exposed to room air (a, c) or > 95% oxygen for 72 h (b, d) and their lungs stained for the presence of free 3'-hydroxyl ends that were detected immunohistochemically as a brown stain. Heavy arrow depicts reactive nuclei, thin arrow depicts negative nuclei. Sections were counterstained with methyl green. Bar in (a) and (c) = 40 µm.

Cells that suffer from DNA damage accumulate p53 protein through an unknown post-translational stabilization mechanism (6, 7). To determine whether changes in DNA integrity were also associated with changes in p53 expression, sections of lung from normoxic and hyperoxic-exposed mice were immunostained for expression of p53 protein. Examination of the distal conducting airway epithelium from control lungs revealed faint nuclear and cytoplasmic staining (Figure 3a). In contrast, alveolar cells were essentially nonreactive at the dilution of antiserum used, suggesting that the airway epithelium expresses higher levels of p53 protein (Figure 3c). Lungs from mice exposed to hyperoxia displayed markedly increased cytoplasmic and nuclear staining in terminal bronchiolar epithelium with significantly less, but still detectable, increases in alveolar epithelium (Figures 3b and 3d). In general, p53 immunostaining was more intense within airway epithelial cells than alveolar cells, suggesting that normal airway epithelial cells express and accumulate more p53 in response to oxidant stress. Expression of p53 was not observed in fibroblasts underlying the airway epithelium (Figure 3b) or in large conducting airways. Sections reacted in the absence of the primary antibody were devoid of staining, similar to the TUNEL controls in the insets of Figures 1a and 1b (data not shown).

View larger version (116K): [in this window] [in a new window]   Figure 3.   Immunohistochemical localization of p53 in hyperoxic mouse lung. Adult mice were exposed to room air (a, c) or > 95% oxygen for 72 h (b, d), killed, and their lungs immunostained for p53, which was identified as a dark brown stain. Heavy arrows depict immunopositive cells, arrowheads depict weakly stained cells, and thin arrows depict nonreactive cells. Open arrows mark airway-alveolar junctions. Sections were counterstained with hematoxylin. Bar in (a) = 40 µm.

To determine whether changes in p53 immunoreactivity reflected an increase in total p53 protein, Western blot analysis was performed on homogenates of lavaged and perfused lungs. Expression of p53 protein was readily detected in lungs of room air-exposed mice and was found to increase in homogenates from hyperoxia-treated animals (Figure 4a). Scanning densitometry of blots obtained from four exposed mice revealed that p53 expression increased 3.0 (± 0.9)-fold. The ability to detect p53 protein in lung homogenates that were perfused and lavaged supports the immunostaining results, which demonstrated that the normal adult lung expresses low levels of p53 in the conducting airway epithelium. Coomassie staining revealed that homogenates of hyperoxic lungs contained a large increase in serum albumin, which leaks into the pulmonary interstitium as alveolar cells die (data not shown). Thus, although the increase in p53 protein by Western blot analysis was observed in equally loaded homogenates, the precise quantitation of this change is difficult to determine owing to the increase in serum albumin and is most likely an underestimation of the actual increase. To determine whether changes in p53 expression were reflected in changes in p53 mRNA levels, poly(A)⁺ RNA was prepared from control and hyperoxia-exposed mice and hybridized with a p53 cDNA. Northern blot analysis identified a single 2.0-kb mRNA transcript as previously reported (20) whose abundance was not markedly altered in response to hyperoxia (Figure 4b). Hybridization for the ribosomal mRNA L32, whose expression is unaltered by oxygen exposure, confirmed that the blot was equally loaded (data not shown) (18).

View larger version (28K): [in this window] [in a new window]   Figure 4.   p53 expression in hyperoxic mouse lung. Adult mice were exposed to room air or > 95% oxygen for 72 h, killed, and RNA and protein extracts prepared from perfused and lavaged lungs. (a) Equal amounts of protein (50 µg) were separated by SDS-PAGE, transferred to nitrocellulose, and immunoblotted for p53. Lane 1, protein from a mouse exposed to room air; lane 2, protein from an oxygen-exposed mouse. Similar results were observed in four separately exposed mice. Immunoreactive protein was not observed in the absence of primary antibody (not shown). (b) Poly(A)+ RNA (5 µg) was separated by size, transferred to Nytran, and hybridized with 32P-labeled p53. A single RNA transcript of 2.0 kb was identified in control mice (lanes 1 and 2) and from oxygen-exposed mice (lanes 3 and 4).

Because hyperoxia has been shown to promote epithelial and endothelial cell injury and death, S1 nuclease protection analysis was performed to measure the level of expression for differentiated markers of airway and alveolar epithelial cells. The expression of cytochrome P-450 2F2 and Clara cell secretory protein (CCSP) RNAs, expressed exclusively by Clara cells, was found to decrease to approximately 25% of unexposed animals (Figure 5). In contrast, expression of surfactant protein B (SP-B), which is expressed by alveolar and airway epithelial cells, was found to increase modestly. These changes confirm and extend previous studies in which hyperoxia was shown to decrease CCSP expression and increase SP-B expression owing to its elevated expression in airway epithelium (21). This finding demonstrates that as hyperoxia injures pulmonary epithelial cells, as shown by TUNEL and p53 staining, they also respond by altering the expression of epithelial cell-specific gene products.

View larger version (24K): [in this window] [in a new window]   Figure 5.   Changes in epithelial cell gene expression in hyperoxic mouse lung. Total RNA (2 µg) was isolated from lungs of mice exposed to room air or > 95% oxygen for 72 h and hybridized with SP-B, CCSP, cytochrome P-450 2F2, and L32 cDNA probes. Annealed complexes were digested with S1 nuclease and separated by gel electrophoresis. The intensities of the protected fragments were determined by phosphorimage analysis and normalized to the signal obtained with L32. Data are represented as mean ± SEM for each gene and are expressed relative to the average expression obtained from four mice exposed to room air. Asterisks indicate significance according to ANOVA relative to control (*P < 0.07; **P < 0.01).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

This study demonstrates that hyperoxia damages nuclear DNA of pulmonary epithelial cells in vivo and is associated with accumulation of p53 and changes in epithelial cell gene expression. Areas of the lung that displayed increased TUNEL staining correlated with previous histologic studies that revealed the greatest level of cell injury in response to oxygen (12). Although normal airway epithelial cells were found to express low levels of p53, cellular regions that displayed increased TUNEL staining also accumulated p53 protein. Associated with TUNEL and p53 changes were alterations in epithelial cell-specific gene products that regulate both normal pulmonary homeostasis and protection from oxidant stress. These observations support the concept that hyperoxia injures pulmonary epithelial cells, which respond by altering cell-specific gene expression and accumulating p53.

Free radical-mediated cytotoxicity is associated with peroxidation of lipids, proteins, and nucleic acids, resulting in a modification and/or inactivation of their biologic properties (1). One method for detecting DNA strand breaks is TUNEL staining, which uses terminal transferase to add digoxigenin-conjugated nucleotides to free 3'-hydroxyl groups of DNA that are visualized with a peroxidase-conjugated anti-digoxigenin antibody. Although this method is often used to detect apoptotic cells that have nicked DNA, this method also recognizes injured cells in other pathologies (22). In the present study, lightly stained TUNEL-positive nuclei were observed in control lungs with intensely stained nuclei in hyperoxic lungs. Although the mice used in this study were housed in a pathogen-free environment, it is presently unknown why their lungs contained low levels of TUNEL staining before exposure to hyperoxia. It is unlikely that the TUNEL staining was overreacted or that the tissue was overfixed because there were numerous TUNEL-negative cells surrounding faint TUNEL-positive cells. Furthermore, although the intensity of the stain increased in oxygen-exposed animals, it was restricted to areas of the lung that have been shown to be preferentially injured by molecular oxygen. Sections of intestine harvested from the same mice were found to exhibit TUNEL-positive nuclei only in cells at the tips of the villus, but not within the crypts. Finally, sections obtained from the esophagus were completely TUNEL negative (data not shown). Because the lung has the largest mucosal surface exposed to molecular oxygen, future studies will test the hypothesis that pulmonary cells suffer from chronic low-level DNA damage that is constantly being repaired.

Expression of the cellular protooncogene p53 was found to increase in the same areas of the lung that had changes in TUNEL staining. The p53 protein was initially identified through its interactions with simian virus 40 (SV40) large T antigen and later with adenovirus and papillomavirus oncoproteins (23). p53 regulates the cell cycle because it accumulates in the cytoplasm, migrates to the nucleus at the beginning of S phase, and reaches a maximal level during mitosis. Cells transformed with p53 growth arrest in the G₁ phase of the cell cycle, in part through the p53-dependent transcriptional activation of the cyclin-dependent kinase inhibitor p21 (24). Expression of p53 is markedly elevated posttranscriptionally in response to cellular insults including radiation-induced DNA damage, hypoxia, oncogene activation, and viral infection (25). Although the mechanism by which cells recognize DNA damage and provoke a p53 response is unknown, studies have shown that transfection of cells with DNase or one molecule per cell of nicked plasmid DNA is sufficient to provoke p53-mediated cell growth arrest (26, 27). When there is irreparable cellular damage, p53 also acts to promote an apoptotic signal and the cell undergoes programmed cell death (8).

Implicit in the growth inhibitory activities of p53 is the observation that mutated forms of p53 generally provide cells with a growth advantage. Several mechanisms account for the selective growth advantage of mutant p53, including the observation that it often accumulates in the cytoplasm instead of the nucleus, has a longer half-life, and is able to inhibit wild-type p53 function (28, 29). Cells expressing mutated p53 fail to growth arrest in G₁ in response to radiation-induced DNA damage. Additional studies led to the discovery that oxidization disrupted wild-type p53 conformation with loss of DNA binding in vitro (30). Although these studies suggest that free radicals would disrupt p53 function, it has been shown that p53-mediated apoptosis was dependent on oxidative stress (31). It is clear from these studies that redox levels may alter the subcellular distribution and function of p53. In the present study, mice exposed to hyperoxia accumulated p53 in the cytoplasm and nuclei of airway and alveolar epithelial cells without significant changes in p53 mRNA levels. Weak cytoplasmic and nuclear staining for p53 was also observed in airway epithelium in control lungs and was readily detected by Western blot analysis. The observation that p53 is expressed in the normal lung epithelium and was increased during hyperoxic injury suggests that it plays a role in regulating normal lung homeostasis and the response to oxidant injury.

Exposure to oxygen resulted in changes in epithelial-specific gene products. CCSP and cytochrome P-450 2F2 are expressed exclusively by bronchiolar Clara cells and SP-B is expressed by bronchiolar and alveolar epithelial cells. Previous studies have shown that mice exposed to 100% oxygen lose expression of CCSP mRNA without death of the Clara cell because the airway was not denuded (21). Moreover, in situ hybridization revealed that whereas SP-B is normally expressed in alveolar and bronchiolar epithelium of mice, exposure to oxygen resulted in loss of expression in the alveolus with markedly increased (8- to 10-fold) uniform expression in the bronchioles (21). Increased expression of SP-B has also been documented in adult rats exposed to 85% oxygen (32). Similarly, rabbits exposed to oxygen have increased expression of another surfactant-associated protein, SP-A, which is also expressed by both airway and alveolar epithelial cells (33). The present study found that oxygen decreased expression of mRNAs for CCSP and cytochrome P-450 2F2 with a modest increase in SP-B expression. One explanation for the disparity in magnitude of the SP-B change between these studies may be due to the exposure parameters. In the previous studies, SP-B expression was found to increase markedly in adult B6C3F₁ hybrid and FVB/N inbred mice, which are relatively oxygen resistant and survive for nearly 1 wk in > 95% oxygen. Moreover, rats were exposed to 85% oxygen, which is less damaging than 95% oxygen. The present study demonstrated only a modest increase in SP-B expression using the oxygen-sensitive C57Bl/6J inbred strain of mice, which do not survive beyond 72 h in > 95% oxygen (our unpublished observations). Regardless of these differences, it is clear that hyperoxia alters cell-specific gene expression in the lung and the magnitude of the response may be accounted for, in part, by strain differences in oxidant sensitivity (34).

The cellular and tissue response to oxidative stress is complex and involves autocrine and paracrine interactions between distinct cell types that must coordinate both individual cell survival as well as survival of the organ. The cellular response to DNA damage includes accumulation of p53, which acts to inhibit cell proliferation, mediate DNA repair processes, and coordinate signals to undergo apoptosis. One study using TUNEL staining has suggested that hyperoxia induces apoptosis in the lung (35). It is unlikely that TUNEL staining in the lung is an absolute marker of apoptosis because normal lungs express weak TUNEL staining and express p53 protein. Moreover, intense TUNEL-positive cells of the airway, which would be predicted to be apoptotic, have increased expression of p53, SP-B, and transforming growth factor- (17, 21). Increased levels of TUNEL staining and p53 protein have also been observed in sections of human lung obtained from patients suffering from idiopathic pulmonary fibrosis (36). Because the airway Clara cell and the alveolar type II cell, which are both TUNEL- and p53-positive, play a significant role in repairing the lung following oxidant injury, additional studies will clarify the extent of DNA damage and the role that p53 plays in regulating normal lung homeostasis and repair processes in response to oxidant stress.