Short-duration hyperoxia causes genotoxicity in mouse lungs: protection by volatile anesthetic isoflurane

Abstract

High concentrations of oxygen (hyperoxia) are routinely used during anesthesia, and supplemental oxygen is also administered in connection with several other clinical conditions. Although prolonged hyperoxia is known to cause acute lung injury (ALI), whether short-duration hyperoxia causes lung toxicity remains unknown. We exposed mice to room air (RA or 21% O₂) or 60% oxygen alone or in combination with 2% isoflurane for 2 h and determined the expression of oxidative stress marker genes, DNA damage and DNA repair genes, and expression of cell cycle regulatory proteins using quantitative PCR and Western analyses. Furthermore, we determined cellular apoptosis using TUNEL assay and assessed the DNA damage product 8-hydroxy-2′-deoxyguanosine (8-Oxo-dG) in the urine of 60% hyperoxia-exposed mice. Our study demonstrates that short-duration hyperoxia causes mitochondrial and nuclear DNA damage and that isoflurane abrogates this DNA damage and decreases apoptosis when used in conjunction with hyperoxia. In contrast, isoflurane mixed with RA caused significant 8-Oxo-dG accumulations in the mitochondria and nucleus. We further show that whereas NADPH oxidase is a major source of superoxide anion generated by isoflurane in normoxia, isoflurane inhibits superoxide generation in hyperoxia. Additionally, isoflurane also protected the mouse lungs against ALI (95% O₂ for 36-h exposure). Our study established that short-duration hyperoxia causes genotoxicity in the lungs, which is abrogated when hyperoxia is used in conjunction with isoflurane, but isoflurane alone causes genotoxicity in the lung when delivered with ambient air.

INTRODUCTION

Oxygen therapy is a common medical necessity that has significant negative side effects (11, 54, 55). For example, in addition to direct pulmonary toxicity, hyperoxia has been implicated in the progression of several pulmonary disorders, including bronchopulmonary dysplasia (5, 58), acute respiratory distress syndrome (9, 23, 35), persistent pulmonary hypertension of the newborn (1, 50), and lung fibrosis (20). In addition to its routine use during anesthesia, oxygen is also administered in connection with several other clinical conditions (28, 43). Additionally, pure oxygen is often delivered to patients before specific procedures (e.g., intubation, extubation, and endoscopy) because of the potential for brief episodes of airway loss and transient hypoxemia (43). Breathing 100% oxygen before performance of these procedures increases the time that apnea can be tolerated should misadventure prevent immediate establishment of a patent airway. Considering its wide application, it is difficult to eliminate high oxygen concentrations from clinical use. On the contrary, it would be highly desirable to co-deliver oxygen with a gas that can prevent or eliminate the toxic effects of high oxygen exposure. Whereas prolonged breathing of oxygen at high concentrations is known to cause lung damage in animal experiments (22, 36, 42), the effects of short-duration exposure to hyperoxia on lung injury remains unknown. Almost all animal or cell culture experiments use 24 to 96 h of 90%–95% oxygen exposure to induce lung injury or to induce cell death in cultured cells (36). Thus, there is a paucity of studies that address the effect of short-duration hyperoxia exposure on lung injury. Because hyperoxia is used for short duration in several clinical situations, we investigated whether short-duration hyperoxia causes pulmonary oxygen toxicity.

Isoflurane is a halogenated volatile anesthetic that is widely used perioperatively. Although published results show both beneficial and detrimental effects of isoflurane on murine lung inflammation and sepsis, the role of isoflurane on hyperoxia-induced lung injury remains, to our knowledge, unexplored. For example, a recent publication indicates that isoflurane protects against endotoxin-mediated lung injury by preserving epithelial tight junction integrity (24) and in inflammation during sepsis (34). Isoflurane has also been shown to decrease apoptosis of endothelial cells exposed to tumor necrosis factor alpha (TNFα) (4). However, other findings indicate that isoflurane produces reactive oxygen species (ROS), which are critical for preconditioning against lipopolysaccharide (LPS)-mediated acute lung injury (ALI) (10). Although isoflurane is almost always used in combination with high oxygen concentrations for short durations (1–5 h or more) during surgery, the detrimental effects of short-duration hyperoxia on lung and the effect of isoflurane on possible lung injury caused by short-duration hyperoxia exposure remain unknown. Furthermore, the effect of isoflurane on hyperoxia-induced ALI has never been investigated, although several studies have reported the beneficial effect of isoflurane or isoflurane plus oxygen in alleviation of lung injury and inflammation caused by endotoxin or sepsis (3, 32, 34).

It has been established that elevated levels of ROS are generated in response to hyperoxia in the mitochondria and by NADPH oxidase (Nox) activation (15, 16, 18, 57). Since ROS can induce DNA lesions, as well as damage proteins and lipids, it is believed that the extensive damage to lung cells and lung architecture occurs as a consequence of hyperoxic insult. Mechanistically, it has been demonstrated that hyperoxic conditions contribute to lung injury and its progression as a consequence of enhanced DNA damage (6, 45, 49, 60). The cellular response to DNA damage is a complex process involving a number of proteins that recognize the process and initiate a signaling cascade, which results in the activation, or inhibition, of various effector molecules required for regulating DNA repair, cellular proliferation, and apoptosis (2, 45). Base excision repair (BER) is a major pathway involved in the repair of oxidative DNA damage caused by hyperoxia. Overexpression of apurinic endonuclease 1 (APE1) or Oxo guanine glycosylase (Ogg1) has been shown to protect cells against hyperoxia-mediated lung injury (59). Both of these enzymes belong to the BER class that removes nucleotide bases that have been damaged by oxidative stress (59).

In this report, we investigated the effect of short-duration (2 h) hyperoxia (60%) on cultured mouse alveolar type II cells (MLE-12) and primary type I, type II, and endothelial cells isolated from murine lung. In addition, we also utilized human primary microvascular endothelial cells from lung (HMVEC-L) and the mouse lung in vivo for these studies. We demonstrate that short-duration hyperoxia causes increased production of O₂^·− by mitochondria and NADPH oxidases that causes pulmonary genotoxicity. In contrast, isoflurane delivered with hyperoxia decreases O₂^·− generation and consequently alleviates oxygen toxicity. However, isoflurane alone produces significant O₂^·− via NADPH oxidase. Isoflurane in normoxia causes significant lung genotoxicity and apoptosis of lung cells. Furthermore, we show that lung injury due to prolonged inhalation of hyperoxia is significantly reduced when delivered with low concentrations of isoflurane.

MATERIALS AND METHODS

Cell culture and oxygen exposure.

MLE-12 cells were cultured in HITES (hydrocortisone, insulin, transferrin, estradiol and selenium) medium (14) supplemented with 2% fetal bovine serum. Actively growing MLE-12 cells were exposed to normoxia [21% oxygen, room air (RA)] or hyperoxia (60% oxygen) along with or without 2% isoflurane for 2 or 16 h in a Modular Incubator Chamber (Billups-Rothenberg, San Diego, CA). HMVEC-L were cultured in endothelial basal medium (EBM) with growth supplements (Lonza).

Antibodies.

Anti-p53 (sc-6243), anti-Cdc25A (sc-97), anti-Chk1 (sc-8408), anti-cyclin B1 (sc-7393), anti-Ape1 (sc-17774), anti-SPC (sc-7706), and anti-AQP5 (sc-28628) antibodies were purchased from Santa Cruz Biotechnology (Dallas, TX). 8-Hydroxy-2′-deoxyguanosine (8-Oxo-dG) antibodies (MAB 3560) were obtained from Millipore (Burlington, MA). Anti-phospho p53 (Ser15) antibody (cat. no. 9284), anti-phospho Cdc25C (Ser216) antibody (cat. no. 9528), anti-phospho Chk1 (Ser345) antibody (cat. no. 2341), anti-phospho ERK1/2 (Thr202/Tyr204) antibody (cat. no. 9101), and anti-p21 antibody (cat. no. 2947) were procured from Cell Signaling Technology (Danvers, MA). Anti-Tom20 antibody (ab-186734), anti-Ogg1 antibody (ab-135940), and anti-NOX4 antibody (ab-133303) were purchased from Abcam (Cambridge, MA).

Exposure of mice to hyperoxia and isoflurane.

C57BL/6J male mice (12–14 wk old) were purchased from Charles River Laboratories and used in this study. All procedures were approved by the Institutional Animal Care and Use Committee at the Texas Tech University Health Sciences Center, Lubbock, and University of Texas Health Science Center at Tyler and were consistent with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. Mice (n = 5 to 6 per group) were exposed to 21% O₂, 60% O₂, 21% O₂ plus 2% isoflurane, or 60% O₂ plus 2% isoflurane for 2 h (2 l/min). The lower and higher dosages of isoflurane (0.25%) were used to demonstrate the effect of non-anesthetic versus anesthetic dose of isoflurane on genotoxic effects of hyperoxia. An isoflurane vaporizer was used to mix oxygen and isoflurane. In some cases, mice were allowed to recover under normoxic conditions for up to 24 h after isoflurane exposure. When we exposed mice to isoflurane or isoflurane plus 60% oxygen, mouse body temperature (monitored by rectal probe) dropped to 29°C by 30 min. We provided heat supplementation by warming base by a heat pad [Homeothermic Blanket Systems with Flexible Probe 507222-F (Harvard Apparatus, Holliston, MA) for monitoring the temperature] and utilizing a lamp, and with this arrangement the temperature was maintained between 35.9°C and 36.5°C. At the end of these treatments, both anesthetized and non-anesthetized control groups of mice were euthanized by cervical dislocation followed by thoracotomy. The lungs were perfused with PBS via right ventricular puncture, dissected out, and processed immediately.

Isolation of pulmonary epithelial and endothelial cells.

Endothelial cells from mouse lungs were isolated as described by Sobczak et al. (52). In brief, 6–8-week-old mice were anesthetized with ketamine (100 mg/kg) and xylazine (10 mg/kg), and after opening of the thoracic cavity, lungs were perfused from the right ventricle, excised, and placed in DMEM containing 1% penicillin/streptomycin. Lung lobes were minced into small pieces and digested in DMEM containing 1 mg/ml collagenase II from Worthington for 45 min. Then the mixture was triturated by passing through a 16-guage cannula and filtered with a 70-μm strainer. After centrifugation, the pellet was dissolved in DMEM, and endothelial cells were selected by incubating with anti-platelet endothelial cell adhesion molecule antibodies (BD Biosciences)-bound sheep anti-rat Dynabeads (Invitrogen) for 30 min, washed, and trypsinized to detach cells from Dynabeads. Finally, the cells were suspended in EBM-2 medium and plated in gelatin-coated plates. Type II alveolar epithelial cells from mouse lungs were isolated as described in Ref. 16. Briefly, lungs were teased apart after immersing in 1.5 ml dispase for 45 min, filtered through 100- and 40-µM nylon filters, and seeded onto anti-CD45 and anti-CD32 coated 100-mm dishes. Supernatant containing type II alveolar cells was suspended in DMEM with 10% FBS and seeded onto six-well plates or in cover glass for microscopy.

PCR array analysis.

Mice were exposed to hyperoxia with or without 2% isoflurane, and after appropriate length of treatment or recovery, they were euthanized and the lungs were isolated. The lungs were perfused with nuclease-free PBS and preserved in RNALater (Ambion). RNA was isolated from the lung tissue using Trizol Reagent, and equal amounts of RNA were subjected to reverse transcription using an RT² First Strand Kit (Qiagen). Gene expression analysis was performed by RT² Profiler PCR Array (Mouse Oxidative Stress and Antioxidant Defense, PAMM-065-ZC, and Mouse DNA Damage Signaling Pathway, PAMM-029-ZC) following the manufacturer’s protocol.

Western blotting.

Protein extracts were prepared from appropriately treated MLE-12 cells or lungs using radioimmunoprecipitation assay buffer. Minced lung pieces were homogenized in cold radioimmunoprecipitation assay buffer in a glass dounce homogenizer. The homogenates were centrifuged at 15,000 g for 15 min at 4°C. Protein concentrations were determined with the BCA protein assay kit (Pierce Chemical, Rockford, IL). Protein extracts were analyzed for Western blotting using their specific antibodies as previously published (18, 39).

Immunofluorescence microscopy.

MLE-12 cells, primary type II cells, or primary endothelial cells were cultured on glass coverslips and exposed to hyperoxia with or without 2% isoflurane. The cells were then washed, fixed, permeabilized, and probed with rabbit anti-8-Oxo-dG and mouse anti-Tom 20 antibodies followed by Alexa Fluor 488- and Alexa Fluor 568-conjugated donkey anti-rabbit and anti-mouse secondary antibodies. Cells were counterstained with DAPI. A stack of fluorescent images was obtained along the z-axis at 60-nm intervals via a ×63/1.40 numerical aperture objective using a Zeiss Axio Imager Z2 upright fluorescent microscope and deconvolved using AxioVision 4.9 software. Lung tissue sections were deparaffinized, hydrated, permeabilized, blocked, and immunostained with anti-8-Oxo-dG and mouse anti-Tom 20 antibodies. A stack of fluorescent images was obtained along the z-axis at 100-nm intervals via a ×40/1.40 numerical aperture objective and was deconvolved.

Hematoxylin-eosin staining.

C57BL6 mice (12–14 wk) were exposed to 60% oxygen with or without 2% isoflurane to study the effect of short-duration hyperoxia on mouse lung injury. We also exposed mice to either 95% oxygen or 95% oxygen with 0.25% isoflurane for 36 h to study the effect of isoflurane on ALI in prolonged hyperoxia. After the exposure, mice were euthanized, and the lungs were flushed with warm PBS via the right ventricle. After blood was removed from the lung, the entire organ was inflated via instillation of 4% paraformaldehyde through the trachea for 12 h. The inflated lung was removed en bloc and immersed in this fixative. Fixed lung was paraffin mounted and sectioned using a microtome. Lungs sections were stained with hematoxylin and eosin. The stained sections were observed under a light microscope for assessment of lung injury indices (41). Lung injury score was implemented as suggested by an official American Thoracic Society Workshop Report for histological evidence of lung injury as published (41). In brief, the score criteria are as follows: A, neutrophils in alveolar space; B, neutrophils in interstitial space; C, hyaline membranes; E, alveolar septal thickening. Each of these indicators was scored as per Table-1 of the lung injury scoring system (41). The score was calculated as number of each alphabet/field and number of field × 100. The injury score is reported as a continuous value between 0 and 1 (41).

8-Oxo-dG ELISA.

Urine was collected from mice exposed to normoxia, normoxia plus isoflurane, hyperoxia, and hyperoxia plus isoflurane. For urine collection, individual mice were allowed to urinate on a clear polyethylene surface, and the urine was aspirated by a micropipette into microcentrifuge tubes. The level of 8-Oxo-dG in the urine of mice was determined using an HT 8-oxo-dG ELISA Kit (R & D Systems, cat. no. 4380-096-K) following supplier’s instructions.

TUNEL assay.

Type 2 epithelial cells and endothelial cells from mouse lung (12–14 wk) were isolated and cultured on glass coverslips and exposed to 21% O₂ or 60% O₂ with/without 2% isoflurane for 24 h. At the end of the exposure period, cells were fixed and permeabilized, and apoptotic type 2 and endothelial cells were labeled with TdT-mediated fluorescein-dUTP-X nick end labeling (TUNEL) kit (cat. no. 11684795910, In Situ Cell Death Detection Kit, Sigma). Cells were also counterstained with DAPI. Immunofluorescence images were obtained using Zeiss LSM880 fluorescent confocal microscope via ×20 objective. TUNEL-positive type 2 and endothelial nuclei were counted using Zeiss Zen 2.3 software. Deparaffinized and hydrated lung sections were subjected to TUNEL assay using In Situ Cell Death Detection Kit from Roche Applied Science (cat. no. 11684795910) following the supplier's protocol.

Flow cytometry and annexin V apoptosis assay.

MLE-12 cells or HMVEC-Ls were exposed to 21% or 60% O₂ with or without 2% isoflurane for 2 h or 24 h, and then the cells were trypsinized using TrypLE reagent. The collected cells were washed and incubated with 1X Annexin V binding buffer containing Annexin V-FITC (cat. no. 556547, BD Bioscience) as per manufacturer instructions. After incubation, the cells were washed with 1X binding buffer and subjected to FACS analysis using BD FACS CALIBUR, and the results were analyzed using FlowJo software.

Detection of superoxide anion by electron paramagnetic resonance spectrometry.

Superoxide formation in MLE-12 cells was detected by electron paramagnetic resonance (EPR) spectroscopy using spin probe 1-hydroxy-3-methoxycarbonyl-2, 2,5,5-tetramethylpyrrolidine hydrochloride (CMH). After appropriate treatments, MLE-12 cells were resuspended in Krebs-Henseleit Buffer (118 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO₄, 1.25 mM CaCl₂, 1.2 mM KH₂PO₄, 25 mM NaHCO₃, 11 mM glucose, pH 7.4) and incubated with 0.1 mM CMH in the presence of 25 µM desferoxamine and 2.5 µM diethyldithiocarbamate at 37°C. At the end of incubation period, the reaction mix was loaded in an EPR capillary tube, and the generated superoxide was detected as paramagnetic nitroxide radical (CM^·) using a Bruker EMX Nano spectrometer at room temperature. Mitochondrial superoxide formation in MLE-12 cells was detected by EPR spectroscopy using spin probe 1-hydroxy-4-[2-(triphenylphosphonio)-acetamido]-2,2,6,6-tetramethylpiperidine (mitoTEMPO-H). After appropriate treatments, MLE-12 cells were resuspended in Krebs-Henseleit Buffer and incubated with 50 µM mitoTEMPO-H at 37°C for 15 min. Superoxide generated by MLE-12 cells was detected as paramagnetic nitroxide radical using Bruker EMX Nano spectrometer at room temperature. EPR spectra were acquired under following scan conditions: microwave frequency, 9.63 GHz; power, 0.31 mW; attenuation 25 dB; modulation frequency, 100 kHz; modulation amplitude, 1.00 G; sweep time, 60 s; time constant, 1.28 s; receiver gain, 40 dB; magnetic field, 3400–3460 G.

Statistical analysis.

Data are expressed as the mean ± SEM. Multiple experimental groups are compared by analysis of variance with Tukey’s posttest comparison of means using GraphPad Prism software. The Student’s t-test was used when comparing means from two experimental groups. A value of P ≤ 0.05 was considered statistically significant. Unless otherwise stated, the figures represent a minimum of three mice (n = 3) that were used to determine mean values and other analyses.

RESULTS

Exposure of MLE-12 cells to 60% oxygen for 24 h causes mitochondrial and nuclear DNA damage.

We exposed MLE-12 cells to 60% oxygen for 2 h and measured the level of the oxidative DNA damage product 8-Oxo-dG (44) using immunofluorescence. However, we did not observe 8-Oxo-dG, as there was no fluorescence in these cells (Fig. 1A). Next, we increased the exposure to 24 h, and as shown in Fig. 1, B and C, this treatment caused significant mitochondrial DNA damage, as detected by co-localization of mitochondrial membrane protein Tom-20 and 8-Oxo-dG. Additionally, 8-Oxo-dG accumulations were also found in the nucleus, demonstrating nuclear DNA damage as well (Fig. 1B). In contrast, when oxygen and isoflurane were used in combination, the level of 8-Oxo-dG was significantly decreased (Fig. 1B). Interestingly, isoflurane in normoxia caused significant mitochondrial and nuclear DNA damage as visualized by accumulation of 8-Oxo-dG in the nucleus and mitochondria (Fig. 1B). However, isoflurane in combination with hyperoxia caused significant reduction in 8-Oxo-dG accumulations in both nucleus and mitochondria. Collectively, our data presented in Fig. 1 show that in cultured cells 60% O₂ exposure for 2 h does not cause detectable mitochondrial or nuclear DNA damage, but exposure to higher duration (>16 h) of 60% O₂ induces significant DNA damage. Furthermore, our data also show that isoflurane is able to abrogate DNA damage induced by prolonged exposure to hyperoxia, but isoflurane with normoxia causes significant nuclear and mitochondrial DNA damage.

Fig. 1.

Isoflurane protects against mitochondrial and nuclear DNA damage in MLE-12 cells exposed to 60% oxygen. A: accumulation of 8-Oxo-dG in the mitochondria of MLE-12 cells following 24-h exposure to 60% O₂. Top, cells exposed to room air; second panel from top, cells exposed to 60% O₂; third panel from top, cells exposed to 21% O₂ plus 2% isoflurane; bottom, cells exposed to 60% O₂ plus 2% isoflurane. B: panels are similar to A, but cells are exposed to normoxia, normoxia plus isoflurane, hyperoxia, or hyperoxia plus isoflurane for 24 h. C: 8-Oxo-dG levels (mean gray value). *Significantly higher, P < 0.05 compared with 21% exposure; **significantly lower compared with 60% O₂ exposure, ANOVA P < 0.05; n = 3 for each. Iso, isoflurane; 8-Oxo-dG, 8-hydroxy-2′-deoxyguanosine.

Mice exposed to 60% oxygen for 2 h show significant mitochondrial 8-Oxo-dG accumulations in the lung, but not in the presence of isoflurane.

Our cell culture data show that mouse alveolar epithelial cells (MLE-12) are resistant to induction of DNA damage after 2 h of 60% O₂ exposure. Therefore, we sought to determine whether in in vivo the lung would sustain any DNA damage due to 2-h exposure to 60% O₂, as most lung injury studies using hyperoxia use >60% oxygen exposure for 48–96 h (16, 17, 46). We determined whether short-term 60% O₂ exposure damages lung DNA in vivo by analyzing the levels of 8-Oxo-dG in the lung following 2 h of exposure. As shown in Fig. 2A, high levels of 8-Oxo-dG were observed in the lungs of mice exposed to 60% O₂ for 2 h (Fig. 2A, top left, and Fig. 2B, left), indicating significant mitochondrial DNA damage in the lungs of these animals. We also exposed mice to 2 h of 60% oxygen and recovered them for 24 h in RA. As shown in Fig. 2C, the level of 8-Oxo-dG was decreased in 60%-oxygen-exposed animals after 24-h recovery in RA. In contrast, mice breathing 60% oxygen along with 2% isoflurane had significantly lower levels of 8-Oxo-dG (Fig. 2, B and C), indicating that this anesthetic protects against pulmonary DNA damage caused by short-duration hyperoxic exposure.

Fig. 2.

Isoflurane protects against accumulation of 8-Oxo-dG released from mitochondria in the lungs of mice exposed to 60% oxygen. A: lung sections from mice exposed to 2 h of 60% oxygen in the presence or absence of 2% isoflurane were labeled with Tom-20 and 8-Oxo-dG antibodies for immunofluorescence. Top, 8-Oxo-dG; middle, Tom20; bottom, 8-Oxo-dG plus Tom20. B: higher magnification image from A. Lung alveolar cells with 8-Oxo-dG/Tom20 co-staining in 60% O₂. C: effect of isoflurane on 8-Oxo-dG/Tom20 staining after 2-h exposure, following 24-h recovery. D: isoflurane prevents the release of 8-Oxo-dG in the urine of mice exposed to 60% oxygen. Mice (n = 3) were exposed to 60% O₂ in the presence or absence of 2% isoflurane for 2 h. After 4 or 24 h, the urine was collected and 8-Oxo-dG levels were determined by ELISA (Novus Biologicals). *Significantly increased compared with normoxic controls; **significantly decreased compared with hyperoxia. One way analysis of variance followed by Tukey’s multiple comparison (n = 3). 8-Oxo-dG, 8-hydroxy-2′-deoxyguanosine.

We next examined whether 8-Oxo-dG detected in the lungs in hyperoxia appears in the urine of mice exposed to high levels of oxygen, as 8-Oxo-dG is known to be excreted in the urine in response to DNA damage by other agents (12, 25, 31). As shown in Fig. 2D, urinary levels of 8-Oxo-dG (as determined by ELISA) were significantly increased 4 h posttreatment. When isoflurane was used in conjunction with oxygen, the level of 8-Oxo-dG was decreased to baseline levels, indicating significant protection by the anesthetic. Additionally, after 24 h of recovery in normoxia, the level of 8-Oxo-dG was decreased in animals exposed to 60% oxygen. However, there was no change in the levels of 8-Oxo-dG in animals exposed to hyperoxia plus isoflurane.

Isoflurane prevents apoptosis of pulmonary cells exposed to hyperoxia.

One of the consequences of DNA damage is the induction of DNA repair. If the damage is extensive, then the process of energy-dependent apoptosis ensures that the damaged cells are eliminated to prevent deleterious mutations. Accordingly, we determined whether short-duration hyperoxia-induced DNA damage signal culminates in the apoptosis of lung cells in vivo. As shown in Fig. 3, A and B, significant pulmonary apoptosis occurred in mice lung exposed to 60% O₂ for 2 h. However, apoptosis was significantly decreased when isoflurane was co-delivered along with the high concentrations of oxygen. Next, we allowed mice to recover for 24 h in RA following 2-h exposure to 60% oxygen alone or oxygen plus isoflurane. As shown in Fig. 3, A and B, the oxygen only group showed a high percentage of cells with TUNEL-positive staining after 24 h of recovery. However, the isoflurane plus oxygen group demonstrated a significantly lower percentage of TUNEL-positive staining that was not significantly different than the lungs of mice in normoxia. We further determined specific cell types of lung that are associated with enhanced apoptosis due to exposure to 24-h hyperoxia. Isolated type II cells from lungs of mice exposed to 60% O₂ showed TUNEL-positive nuclei demonstrating significant apoptosis, but when isoflurane was mixed with hyperoxia, the level of apoptosis was significantly decreased (Fig. 3, C and D). Furthermore, isoflurane in RA also caused significant apoptosis, which is in agreement with our MLE-12 cells exposed to 16 h with isoflurane in RA (Fig. 1B). In contrast to type II cells, the endothelial cells from lung did not show significant apoptosis in response to hyperoxia. Additionally, these cells also did not show enhanced apoptosis when exposed to isoflurane in normoxia (Fig. 3, E and F). Next, we determined the effect of hyperoxia and isoflurane on lung type II and I cells in situ in lung sections. As shown in Fig. 3, G and H, both type I and type II cells underwent apoptosis in hyperoxia and also in normoxia with isoflurane, but the magnitude of apoptosis was higher in type I cells compared with type II cells. Isoflurane in combination with 60% O₂ significantly decreased apoptosis caused by hyperoxia alone in both cell types (Fig. 3, G and H). We also performed annexin V labeling and performed apoptosis assay of MLE-12 and HMVEC-L using flow cytometry. As shown in Fig. 4A, isoflurane in normoxia caused significant apoptosis of MLE-12 cells similar to that obtained with 60% O₂ hyperoxia. Isoflurane significantly decreased apoptosis of MLE-12 cells in hyperoxia. We also evaluated the apoptosis of endothelial cells using this method, and as shown in Fig. 4B, endothelial cells did not show significant apoptosis. Isoflurane alone in normoxia also failed to show significant apoptosis in contrast to MLE-12 cells.

Fig. 3.

Hyperoxia-induced apoptosis of lung cells is abrogated in the presence of 2% isoflurane. A: mice were exposed to 2 h in 60% oxygen in the presence and absence of 2% isoflurane, and lung sections were analyzed for TdT-mediated fluorescein-dUTP-X nick end labeling (TUNEL)-positive nuclei by immunofluorescence. B: percent TUNEL-positive nuclei. *Significantly increased compared with normoxic controls; **significantly decreased compared with hyperoxia. One way analysis of variance followed by Tukey’s multiple comparisons (n = 3). C–F: hyperoxia and isoflurane in normoxia induce apoptosis in mouse lung type II cells but not in isolated lung endothelial cells. Type 2 epithelial cells (C and D) and endothelial cells (E and F) were isolated from lung and cultured on glass coverslips and exposed to 21% O₂ or 60% O₂ with/without 2% isoflurane for 24 h. At the end of exposure period, cells were labeled with TUNEL kit and were counterstained with DAPI. Immunofluorescence images were obtained using Zeiss LSM880 fluorescent confocal microscope via ×20 objective. TUNEL-positive type 2 and endothelial nuclei were counted using Zeiss Zen 2.3 software and plotted as bar graph (n = 3). *P < 0.01 vs. 21% O₂; **P < 0.01 vs. 21% O₂ + 2% isoflurane. G: type 1 and type 2 epithelial cells in lung undergo apoptosis in hyperoxia as well as in isoflurane plus normoxia, and isoflurane prevents hyperoxia-induced apoptosis. Apoptotic cells in lungs sections were labeled with TUNEL kit. Type 1 and type 2 cells were marked using anti-AQP5 and anti-SPC antibodies, respectively. Fluorescence images were obtained through ×63 objective using Zeiss LSM880 fluorescent confocal microscope equipped with Airyscan detector. H: TUNEL-positive type 1 and type 2 cells were counted using Zeiss Zen 2.3 software, and mean ± SE was plotted as bar graph (n = 3) *P < 0.01 vs. 21% O₂; **P < 0.01 vs. 21% O₂ + 2% isoflurane. Iso, isoflurane.

Fig. 4.

Hyperoxia induced apoptosis in MLE-12 cells (A) and not in HMVEC-Ls (B), and isoflurane blocks hyperoxia-induced apoptosis. MLE-12 cells or HMVEC-Ls were exposed to 21% or 60% O₂ with or without 2% isoflurane for 2 h or 24 h. Apoptosis assay was performed as described in materials and methods. Percentage of the cell population undergoing apoptosis (annexin V-FITC positive or annexin V-FITC plus PI positive) was obtained and the data plotted as a bar graph (n = 3). *P < 0.01 vs. normoxia (21% O₂); **P < 0.01 vs. hyperoxia (60% O₂). HMVEC-L, human primary microvascular endothelial cells from lung; Iso, isoflurane; PI, propidium iodide.

Isoflurane induces O₂^·− production with ambient air but inhibits O₂^·− generation when mixed with hyperoxia.

Several studies indicate that isoflurane causes ROS generation (10, 24, 48); however, other studies demonstrate that isoflurane provides protection against sepsis or LPS-dependent cytotoxicity. Thus, the role of isoflurane as an inducer or inhibitor of ROS such as O₂^·− has not yet been established. We determined the effect of isoflurane on O₂^·− generation in the presence of air or hyperoxia using EPR spectrometry. As shown in Fig. 5, A and B, isoflurane mixed with air (or 21% oxygen) produced significant levels of O₂^·− in MLE-12 cells. In contrast, when it was mixed with 60% oxygen, the level of O₂^·− was significantly decreased (Fig. 5, A and B), suggesting that isoflurane can be an O₂^·− generating agent when mixed with air but scavenges hyperoxia-dependent O₂^·− production. Furthermore, our data in Fig. 5, A and B show that diphenyliodonium (DPI), a nonselective NADPH oxidase (Nox) inhibitor significantly inhibited O₂^·− generation by hyperoxia; however, this inhibition was not complete, suggesting that mitochondria could account for DPI-resistant O₂^·− generation in hyperoxia. However, when isoflurane was mixed with 60% oxygen in the presence of DPI, complete inhibition of O₂^·− generation was achieved (Fig. 5, A and B). Next, we determined the effect of DPI on isoflurane-mediated O₂^·− release in normoxia. As shown in Fig. 5, A and B, DPI almost completely inhibited the CMH signal, indicating NADPH oxidase is a major source of O₂^·− by isoflurane. Taken together, these experiments show that Nox is a major source of O₂^·− in normoxia by isoflurane.

Fig. 5.

Hyperoxia induces superoxide generation via mitochondria and Nox, but isoflurane in normoxia induces superoxide generation predominantly via Nox. MLE-12 cells (n = 3) were exposed to 21% O₂ or 60% O₂ with or without 2% isoflurane for 24 h, and superoxide formation in the presence of Nox inhibitor, DPI, or vehicle was determined by EPR using CMH as described in materials and methods (A). Absolute spin counts were obtained using Xenon nano 1.2 software and plotted as bar graph (B). *P < 0.01 vs. normoxia (21% O₂); **P < 0.01 vs. hyperoxia (60% O₂); †P < 0.01 vs. 60% O₂ + DPI. MLE-12 cells (n = 3) were exposed to 21% O₂ or 60% O₂ with or without 2% isoflurane for 24 h, and mitochondrial superoxide formation was determined by EPR using mitoTEMPO-H as described in materials and methods (C). Absolute spin counts were obtained as mentioned for C (D). *P < 0.01 vs. normoxia (21% O₂); **P < 0.01 vs. hyperoxia (60% O₂). CMH, 1-hydroxy-3-methoxycarbonyl-2, 2,5,5-tetramethylpyrrolidine hydrochloride; DPI, diphenyliodonium; EPR, electron paramagnetic resonance; Iso, isoflurane; mitoTEMPO-H, 1-hydroxy-4-[2-(triphenylphosphonio)-acetamido]-2,2,6,6-tetramethylpiperidine; Nox, NADPH oxidase.

Isoflurane decreases mitochondrial O₂^·− production in hyperoxia but does not increase mitochondrial O₂^·− generation in normoxia.

We further determined whether mitochondria are a significant source of O₂^·− for 2-h 60% O₂ exposure and whether isoflurane would decrease this O₂^·−. As shown in Fig. 5, C and D, mitoTEMPO spectra (a mitochondrial superoxide anion spin probe) were observed in 60% O₂ exposure in MLE-12 cells (Fig. 5, C and D). This spectrum was decreased to a nonsignificant extent by DPI, indicating that Nox is not the source of O₂^·− in hyperoxia as expected. However, isoflurane significantly decreased the mitoTEMPO spectra, indicating significant decrease in mitochondrial O₂^·− in the presence of isoflurane in hyperoxia (Fig. 5, C and D). There was no increase in mitoTEMPO signal in normoxia with isoflurane, indicating that isoflurane predominantly generates O₂^·− from Nox in normoxia.

Isoflurane diminishes DNA damage signaling caused by short-term hyperoxia.

We previously established that prolonged hyperoxia activates checkpoint kinase 1 and the tumor-suppressor protein p53 as a consequence of DNA damage (17, 38). Here, we determined whether short duration of hyperoxia would also initiate gene expression associated with DNA damage signaling in the lungs of mice. As shown in Fig. 6A, several DNA damage response genes associated with cell cycle arrest and the DNA damage response were significantly increased in response to short-duration hyperoxia exposure in mice lungs. These genes include RAD9, a homolog of (S. pombe) Rad9A, Ligase I (LigA), RAD51 homolog (S. cerevisiae) (Rad51), Fanconi anemia complementation group G (Fancg), X-ray repair complementing defective repair in Chinese hamster cells 2 (Xrcc2), Replication protein A1 (Apa1), Fanconi anemia complementation group D2 (Fancd2), Hus1 homolog (S. pombe) (Hus1), MutL homolog 3 (E. coli) (Mlh3), Cell division cycle 25 homolog A (S. pombe) (Cdc25A), RAD1 homolog (S. pombe) (Rad1), Fanconi anemia complementation group A (Fanca), Ataxia telangiectasia and rad3 related (ATR), and Checkpoint kinase 1 homolog (S. pombe) (Chek1). Expression of all of these genes remained at basal level when isoflurane was used in combination with hyperoxia, indicating that isoflurane alleviates the activation of DNA damage signaling by protecting DNA damage in the lung due to short-duration hyperoxia.

Fig. 6.

Isoflurane protects against DNA damage signaling by 2-h 60% oxygen exposure in mice lungs. A: PCR microarray of DNA damage response genes in the lungs of mice exposed to 2 h in 60% O₂ in the presence or absence of 2% isoflurane. B: Western analysis of p53, phosphor-p53 (Ser15), p21, Chk1, pChk1, cdc25C, cyclin B1, and GAPDH in the lung tissues of mice exposed to 60% O₂ for 2 h in the presence or absence of 2% isoflurane, and compared with ambient air in the presence of isoflurane. C: densitometry of Western analysis presented in B; (n = 3) *P < 0.01 vs. control; **P < 0.01 vs. hyperoxia (60% O₂). Iso, isoflurane.

As shown in Fig. 6, B and C, the expression of the p53 protein, as well as the occurrence of activating phosphorylation on Ser15, was dramatically increased during short exposures to 60% O₂, which points to a DNA damage and repair response (18). We also found that the expression of p21, a target gene of the p53 transcription factor, is increased (Fig. 6B) as a consequence of 60% O₂ exposure, indicating cell cycle arrest and also transcriptional activity of p53 (37). Furthermore, other G2/M (growth phase 2/metaphase) proteins, such as Cdc25C and cyclin B1 levels (Fig. 6, B and C), were increased in animals experiencing short-term hyperoxia. However, inclusion of 2% isoflurane with high oxygen treatment prevented the increases in all of these DNA damage response proteins, confirming the protective effect of isoflurane against DNA damage that occurs in response to hyperoxia.

Short-duration hyperoxia provokes the expression of Nox components, p47phox, and p22phox but not when isoflurane is co-administered.

We next sought to determine the activation of oxidative stress response genes in response to short-duration hyperoxia and the effect of isoflurane on the activation of these genes. We utilized a PCR microarray analysis of oxidative stress response genes following 2-h exposures to oxygen or oxygen plus isoflurane. As shown in Fig. 7A, the expression of p22phox (a component of NADPH oxidase) was increased ~14-fold, and p47phox (another component of NADPH oxidase) was increased more than 25-fold in response to 60% O₂. Thus, the expression of either of these genes may account for increased superoxide anion (O₂^·−) generation in response to high levels of oxygen. Interestingly, isoflurane markedly decreased the expression of all of these genes activated in hyperoxia, any of which could significantly contribute to the lung damage via generation of high levels of O₂^·−. Collectively, the results of this experiment indicate that p47phox and p22phox could be sources of O₂^·− production that occur in response to short duration of hyperoxia. Our findings also indicate that isoflurane has a significant impact in diminishing the expression of these genes. Since p22phox and p47phox are components of Nox proteins, we determined the expression of Nox4 in the lung tissue of mice exposed to normoxia or hyperoxia in the presence or absence of isoflurane. Our data presented in Fig. 7, B and C show that isoflurane in combination with hyperoxia caused significant increase in Nox4 expression, suggesting isoflurane may induce the production of H₂O₂ via Nox4 in hyperoxia, as activated Nox4 is known to produce H₂O₂ (27). However, Nox4 expression was increased in isoflurane plus 60% oxygen exposed animals. We also tested two other isoforms of Nox mRNA by RT-PCR, such as Duox1 and Duox2, and hyperoxia-induced Duox2 mRNA, but in the presence of isoflurane, this increase was diminished (Fig. 7, D and E).

Fig. 7.

Isoflurane inhibits induction of p47phox, p22phox, Fth1, and Ptgs2 gene expression induced in response to oxygen in mouse lung. A: mice (n = 3) were exposed to 2% isoflurane mixed with room air (21% O₂) or 60% O₂ for 2 h. RNA was isolated and subjected to gene expression analysis using RT² profile PCR array [Mouse Oxidative Stress and Antioxidant Defense Array, PAMM-0652Z; Real-Time quantitative PCR array was performed and analyzed using manufacturer’s protocol (Qiagen, Germantown, MD)]. n = 3 for each mouse group. *P < 0.01 vs. 60% O₂. B: expression of Nox4 in the lungs of mice following 2 h normoxia, normoxia plus isoflurane, hyperoxia, and hyperoxia plus isoflurane. C: densitometry of Western blot presented in B. *P < 0.01 vs. 21% O₂. D: RT-PCR of Duox1 and Duox2. E: densitometry analysis of D. *P < 0.01 vs. 21% O₂; **P < 0.01 vs. 60% O₂. Iso, isoflurane.

Mice breathing isoflurane in combination with 60% O₂ show increased expression of pulmonary BER genes.

We analyzed the expression levels of DNA damage response genes during hyperoxia and changes in them when the treatment included isoflurane. When we measured the fold increase of these genes, we found that several of them are upregulated in response to the inclusion of isoflurane in a 2-h exposure to 60% O₂ (Fig. 8A) Furthermore, we found that most of these DNA repair genes fell into the BER pathway (44). For example, Gadd45α, Fen1, and XPC are associated with the BER pathway (19, 21, 51). In addition, Ppp1r15α is also known to participate in DNA repair pathways (44). Taken together, our microarray study shows that isoflurane may induce DNA repair pathways, in particular those involved in BER. We determined the expression of Ogg1 and APE1 in lungs of mice following 2 h of exposure to 60% O₂ in the presence or absence of isoflurane. As shown in Fig. 8, B and C, the expression of Ogg1 and APE1 was decreased in lungs exposed to 60% oxygen. When isoflurane was mixed with the oxygen, the expression of Ogg1 was maintained, demonstrating that this major BER enzyme is rescued by isoflurane. Collectively, our data show that isoflurane may be involved in improving DNA repair, specifically via BER, when used in combination with hyperoxia.

Fig. 8.

Isoflurane induces Ogg1 expression in the lungs of mice exposed to 60% oxygen. A: mice were exposed to 2% isoflurane mixed with room air (21% O₂) or 60% O₂ for 2 h. RNA was isolated and subjected to gene expression analysis using RT² profile PCR array (Mouse Oxidative Stress and Antioxidant Defense Array, PAMM-0652Z). B: mice were exposed to normoxia (21% oxygen) or hyperoxia (60% oxygen) with or without 2% isoflurane for 2 h, then lung lysates were analyzed for Oxo guanine glycosylase (Ogg1) and APE1 expression levels by Western blotting. Blots were re-probed with anti-β-actin antibodies for normalization. C: densitometry of Western blot presented in B (n = 3), *significantly lower than 21%O₂, **significantly higher than 60%O₂. n = 3, ANOVA, Turkey’s posttest. APE1, apurinic endonuclease 1; Iso, isoflurane.

Isoflurane inhibits the phosphorylation of ERK.

Elevated concentrations of oxygen are known to cause ERK activation (7, 8, 56). Because ROS are known to cause ERK phosphorylation (i.e., activation), we next determined the effect of isoflurane on ERK phosphorylation. As shown in Fig. 9, A and B, we found ERK was activated when mice were exposed to high oxygen for short periods. However, isoflurane, inhibited hyperoxia-induced ERK phosphorylation but increased the expression of ERK protein.

Fig. 9.

Isoflurane prevents 60% O₂-induced ERK phosphorylation in mouse lungs. A: mice (n = 3) were exposed to 21% O₂, 60% O₂, and 60% O₂ plus 2% isoflurane for 2 h; lungs were isolated and homogenized in radioimmunoprecipitation assay buffer and analyzed for pERK1/2 and β-actin by Western blotting. Blot was re-probed to analyze total protein levels. B: densitometry of A; *significantly higher than 21% O₂, **significantly lower than 60% O₂. N = 3, ANOVA, Tukey’s posttest. Iso, isoflurane.

Isoflurane inhibits lung injury in mice exposed to hyperoxia.

Prolonged breathing of hyperoxia is known to cause lung damage in mice and humans. Therefore, we tested whether the protective role of isoflurane in lung injury could extend to ALI induced by longer duration and higher concentration of oxygen. As shown in Fig. 10, A and B, markers associated with lung injury were significantly higher in mice exposed to 36-h hyperoxia (41). However, in the presence of isoflurane, the expression of these markers was significantly lower (Fig. 10, A and B). These data suggest that isoflurane could also provide significant protection against deleterious effects of prolonged oxygen exposure in the lung.

Fig. 10.

Hyperoxia or isoflurane in normoxia induces lung injury in mice, but isoflurane mixed with hyperoxia protects lung from hyperoxia-induced injury. A: C57BL6 mice (12–14 wk) were exposed to either 21% or 60% oxygen with or without 2% isoflurane for 2 h. Lung sections were stained with hematoxylin and eosin. Stained lung sections were imaged using Zeiss AxioVert A1 equipped with AxioCam ICc5 color camera. B: lung injury was scored based on neutrophil infiltration, hyaline membrane deposition, proteinaceous debris in alveoli, and alveolar septal thickening (41), and mean ± SE of injury score was plotted as bar graph. *P < 0.01 vs. 21% O₂; **P < 0.01 vs. 21% O₂ + 2% isoflurane. C: C57BL6 mice (12–14 wk) were exposed to either 95% oxygen or 95% oxygen with 0.25% isoflurane for 36 h. After the exposure, mice were euthanized, and the lungs were sectioned and processed. Lung sections were stained with hematoxylin and eosin. D: lung injury was scored as mentioned for B. *P < 0.01 vs. normoxia; **P < 0.01 vs. 95% O₂. Iso, isoflurane; RBC, red blood cell.

DISCUSSION

In this report, we present evidence that mice exposed to 60% oxygen (relevant as a carrier gas in anesthesia) for 2 h experience mitochondrial and nuclear DNA damage, and when isoflurane is used in conjunction with hyperoxia, the DNA damage is abrogated. Our findings suggest that cell culture studies are not responsive to short-duration of hyperoxia-mediated DNA damage, in contrast to in vivo studies with exposure of mice to 2-h 60% oxygen that causes significant DNA damage. Furthermore, these studies show, for the first time, that significant DNA damage occurs in the lungs in response to short-duration hyperoxia. Additionally, we have shown that isoflurane in normoxia generates O₂^·− via activation of NADPH oxidases; however, isoflurane does not produce O₂^·− via a mitochondrial pathway. Interestingly, isoflurane inhibits hyperoxia-mediated O₂^·− production. We also demonstrated that isoflurane inhibits the expression of p47phox and p22phox (components of NADPH oxidase), which are otherwise induced in mice inhaling 60% oxygen for 2 h. Furthermore, the expression of Ogg1, Gadd34, and Gadd45 (all of which are involved in DNA repair) are increased in the presence of isoflurane plus oxygen, but not in only oxygen, demonstrating that the anesthetic induces DNA repair mechanisms. Furthermore, using a prolonged ALI model, we show that isoflurane significantly protects against ALI. Collectively, these novel data point to a previously unrecognized role for short-duration hyperoxia as a genotoxic stress and the beneficial role of isoflurane in protecting against the DNA damage and ALI caused by hyperoxia.

Several studies indicate that isoflurane causes ROS generation (48), and other studies demonstrate that lung injury caused by LPS or sepsis is prevented by isoflurane. Thus, a clear mechanism of action for this anesthetic has not yet been established. We believe that these contradictory findings are due to the way isoflurane was administered to mice. There are instances in which isoflurane was co-delivered with high oxygen concentrations and others in which it was applied with ambient air. For example, in several animal investigations in which isoflurane was used as a preconditioning agent, we noted that isoflurane was used with ambient air (33, 40, 48, 53). These findings demonstrate that isoflurane generates low levels of O₂^·−, which is crucial for its function as a preconditioning agent. On the other hand, when isoflurane was applied in combination with high concentrations of oxygen, it protected against lung injury and inflammation induced by either LPS or TNFα. Our current finding provides a compelling explanation of this property of isoflurane. Isoflurane in combination with normoxia produces O₂^·− by activating NADPH oxidase but inhibits O₂^·− generation when it is delivered with hyperoxia.

Tissue hypoxia during anesthesia administration is the primary reason that high concentrations of oxygen are routinely used perioperatively (43). However, prolonged breathing of oxygen at high concentrations is known to cause lung damage in animal experiments. By contrast, the effect of short-duration hyperoxia on lung biology has not been previously investigated. Our data further establish that short-term exposure to 60% oxygen causes significant DNA damage in mitochondria and the nucleus, to the point that damage products appear in the urine of mice. This establishes that short-duration hyperoxia is a substantial genotoxic stress in the lung. Surprisingly, when we delivered isoflurane in a hyperoxic gas mixture, we did not detect markers of DNA damage either in the lung or in cultured cells. Consequently, this establishes a strong protective role for this anesthetic during periods of hyperoxia.

Our findings also indicate that hyperoxia induces significant release of 8-Oxo-dG in the urine of mice after 4 h of exposure, indicating oxidative DNA damage and BER. 8-Oxo-dG is primarily repaired by Ogg1, which recognizes the lesion and excises it as free 8-Oxo-dG (13). Accumulation of 8-Oxo-dG is linked to various pathologies, including cancer and aging. Furthermore, free 8-Oxo-dG could pair with adenosine, causing a GC→TA mutation that is frequently encountered as a deleterious consequence of oxidized DNA (47). It has also been shown that accumulated 8-Oxo-dG interacts with Ogg1 and activates Ras and Rac1 GTPases, resulting in an intracellular signaling cascade. Addition of 8-Oxo-dG to cells has been shown to cause oxidative stress (29). Therefore, we speculated that 8-Oxo-dG accumulation in response to hyperoxia generates oxidative stress via expression and activation of p22phox and p47phox. Indeed, our data establish that p22phox and p47phox mRNA levels are significantly decreased when isoflurane is administered in combination with high oxygen tension. These results demonstrate that isoflurane inhibits 8-Oxo-dG-dependent signaling, leading to decreased transcription of these NADPH oxidase components. Thus, since hyperoxia is known to induce NADPH oxidase, it is likely that hyperoxia-stimulated O₂^·− generation in the mitochondria activates p22phox and p47phox (30), resulting in further increases in O₂^·− levels by NADPH oxidase. It is also already known that 8-Oxo-dG can directly activate NADPH oxidase when it is artificially introduced into the lungs of mice (29). This suggests that 8-Oxo-dG accumulation caused by short-duration high oxygen exposures would comprise a major mechanism of increased oxidative stress in the lung. Interestingly, we detected a significant level of 8-Oxo-dG in the urine of mice, which suggests that this marker could potentially be used to monitor the extent of DNA damage in humans exposed to high oxygen concentrations for short durations.

We observed a significant increase in DNA damage response genes and proteins due to short-duration of 60% oxygen exposure in mice lung but not when isoflurane was included with oxygen. These data suggest, for the first time, that short-duration hyperoxia causes significant DNA damage. By extrapolation, these exposures could potentially be deleterious to human lungs and be ameliorated by supplemental isoflurane. Since these genes are involved in DNA repair, cell cycle arrest or induction of apoptosis administration of hyperoxia without isoflurane could induce significant mutagenesis of lung DNA with potential for development of cancer or other diseases associated with genotoxic stress. Exposure of cells to isoflurane combined with ambient air caused increase in p53 (and pSer15), Cdc25C, and cyclin B1 expression, demonstrating that the anesthetic alone could cause activation of DNA damage response proteins when administered with air. These data are in agreement with our current finding that isoflurane in ambient air generates O₂^·− via NADPH oxidase and with other studies demonstrating increased ROS generation by isoflurane when administered to mice or cells with ambient air (10, 33, 40).

Thus, the administration of isoflurane in conjunction with oxygen could potentially alter clinical outcomes, in which the anesthetic could protect against lung injury. Hyperoxia provides a significant source of O₂^·−, as mitochondria generate high levels of O₂^·− in high concentrations of oxygen (26). Isoflurane could potentially act as a scavenger of O₂^·− in a hyperoxic condition. Collectively, our findings suggest that isoflurane inhibits DNA damage signaling in response to short periods of hyperoxia and could thereby protect the lung against arrested cell growth or apoptosis and consequently lung injury in ALI.

GRANTS

The work presented here is supported by National Heart, Lung, and Blood Institute Grant HL 1 R01 130061 to K. Das.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

K.D. conceived and designed research; V.K.-S., J.S., S.R., G.P.M., C.O., T.W., and K.D. performed experiments; V.K.-S., J.S., C.O., and K.D. analyzed data; V.K.-S., J.S., and K.D. interpreted results of experiments; V.K.-S., J.S., and K.D. prepared figures; V.K.-S. and K.D. drafted manuscript; J.W., S.I., and K.D. edited and revised manuscript; V.K.-S. and K.D. approved final version of manuscript.