Genetic and pharmacologic evidence links oxidative stress to ventilator-induced lung injury in mice

Abstract

Rationale: Mechanical ventilation (MV) is an indispensable therapy for critically ill patients with acute lung injury and the adult respiratory distress syndrome. However, the mechanisms by which conventional MV induces lung injury remain unclear.

Objectives: We hypothesized that disruption of the gene encoding Nrf2, a transcription factor that regulates the induction of several antioxidant enzymes, enhances susceptibility to ventilator-induced lung injury (VILI) and that antioxidant supplementation attenuates this effect.

Methods: To test our hypothesis and to examine the relevance of oxidative stress in VILI, we assessed lung injury and inflammatory responses in Nrf2-deficient (Nrf2(-/-)) mice and wild-type (Nrf2(+/+)) mice after an acute (2-h) injurious model of MV with or without administration of antioxidant.

Measurements and main results: Nrf2(-/-) mice displayed greater levels of lung alveolar and vascular permeability and inflammatory responses to MV as compared with Nrf2(+/+) mice. Nrf2 deficiency enhances the levels of several proinflammatory cytokines implicated in the pathogenesis of VILI. We found diminished levels of critical antioxidant enzymes and redox imbalance by MV in the lungs of Nrf2(-/-) mice; however, antioxidant supplementation to Nrf2(-/-) mice remarkably attenuated VILI. When subjected to a clinically relevant prolong period of MV, Nrf2(-/-) mice displayed greater levels of VILI than Nrf2(+/+) mice. Expression profiling revealed lack of induction of several VILI genes, stress response and solute carrier proteins, and phosphatases in Nrf2(-/-) mice.

Conclusions: Our data demonstrate for the first time a critical role for Nrf2 in VILI, which confers protection against cellular responses induced by MV by modulating oxidative stress.

Figure 1.

Effect of mechanical ventilation (MV) on lung injury and inflammatory responses in Nrf2^−/− mice. (A) Mice (n = 3–5 per group) were subjected to MV at low Vt (LVt, 12 ml/kg) and high Vt (HVt, 30 ml/kg) for 2 hours and allowed to recover for 1 hour before bronchoalveolar lavage fluid (BALF) was obtained and protein was quantified as detailed in Methods. In parallel, control mice were sham-operated, anesthetized, and spontaneously ventilated (SpV) under identical conditions of MV exposure. (A) The effect of LVt and HVt on BAL protein leakage in Nrf2^+/+ and Nrf2^−/− mice. *P ⩽ 0.05 versus SpV control of the same genotype. (B) Albumin concentration in the BALF of various experimental groups (as in A) was determined using ELISA. *P ⩽ 0.05 versus SpV or LVt groups of the same genotype. ^#P ⩽ 0.05, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to HVt. (C) The number of neutrophils in the BALF collected from the Nrf2^−/− mice and Nrf2^+/+ mice was counted and expressed as a percentage (top) of total cells (bottom). Values represent means ± SEM (n = 3–5). *P ⩽ 0.05 versus SpV control of the same genotype; ^#P ⩽ 0.05, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to HVt.

Figure 2.

Effect of mechanical ventilation (MV) on pulmonary neutrophil infiltration and vascular protein leakage in Nrf2^−/− mice. (A) Assessment of neutrophil infiltration in lung parenchyma of Nrf2^+/+ and Nrf2^−/− mice subjected to high Vt (HVt). Mice (n = 7–8 mice/group) were subjected to MV at HVt for 2 hours, and then lungs were immediately fixed with 0.2% of low-melting agarose with 10% buffered formalin at 25 cm of water pressure. Immunostaining analysis of parenchymal neutrophils in the tissue sections was assessed using specific antineutrophil antibody. Original magnification: ×10. (B) Neutrophils in the alveolar space were quantified using digital images of tissue sections stained with antineutrophil antibody. Data represent means ± SEM of an average of at least 15 high-power fields for each exposure group. *P ⩽ 0.04, HVt versus spontaneous ventilation (SpV) of Nrf2^+/+ mice. **P ⩽ 0.01, HVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.04, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to HVt. (C) Vascular leakage was assessed by the extravasation of Evans blue into lung parenchyma. Mice (n = 6 per group) were injected with Evans blue into the jugular vein 30 minutes before termination of MV at HVt for 2 hours. Mice in the SpV group also received Evans blue. Lungs were immediately frozen, and Evans blue in the lung tissues of all experimental groups was quantified and expressed as μg/mg lung tissue. Values represent n = 6 per experimental condition. *P ⩽ 0.02, HVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.001, HVt versus SpV of Nrf2^−/− mice.

Figure 2.

Effect of mechanical ventilation (MV) on pulmonary neutrophil infiltration and vascular protein leakage in Nrf2^−/− mice. (A) Assessment of neutrophil infiltration in lung parenchyma of Nrf2^+/+ and Nrf2^−/− mice subjected to high Vt (HVt). Mice (n = 7–8 mice/group) were subjected to MV at HVt for 2 hours, and then lungs were immediately fixed with 0.2% of low-melting agarose with 10% buffered formalin at 25 cm of water pressure. Immunostaining analysis of parenchymal neutrophils in the tissue sections was assessed using specific antineutrophil antibody. Original magnification: ×10. (B) Neutrophils in the alveolar space were quantified using digital images of tissue sections stained with antineutrophil antibody. Data represent means ± SEM of an average of at least 15 high-power fields for each exposure group. *P ⩽ 0.04, HVt versus spontaneous ventilation (SpV) of Nrf2^+/+ mice. **P ⩽ 0.01, HVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.04, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to HVt. (C) Vascular leakage was assessed by the extravasation of Evans blue into lung parenchyma. Mice (n = 6 per group) were injected with Evans blue into the jugular vein 30 minutes before termination of MV at HVt for 2 hours. Mice in the SpV group also received Evans blue. Lungs were immediately frozen, and Evans blue in the lung tissues of all experimental groups was quantified and expressed as μg/mg lung tissue. Values represent n = 6 per experimental condition. *P ⩽ 0.02, HVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.001, HVt versus SpV of Nrf2^−/− mice.

Figure 2.

Effect of mechanical ventilation (MV) on pulmonary neutrophil infiltration and vascular protein leakage in Nrf2^−/− mice. (A) Assessment of neutrophil infiltration in lung parenchyma of Nrf2^+/+ and Nrf2^−/− mice subjected to high Vt (HVt). Mice (n = 7–8 mice/group) were subjected to MV at HVt for 2 hours, and then lungs were immediately fixed with 0.2% of low-melting agarose with 10% buffered formalin at 25 cm of water pressure. Immunostaining analysis of parenchymal neutrophils in the tissue sections was assessed using specific antineutrophil antibody. Original magnification: ×10. (B) Neutrophils in the alveolar space were quantified using digital images of tissue sections stained with antineutrophil antibody. Data represent means ± SEM of an average of at least 15 high-power fields for each exposure group. *P ⩽ 0.04, HVt versus spontaneous ventilation (SpV) of Nrf2^+/+ mice. **P ⩽ 0.01, HVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.04, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to HVt. (C) Vascular leakage was assessed by the extravasation of Evans blue into lung parenchyma. Mice (n = 6 per group) were injected with Evans blue into the jugular vein 30 minutes before termination of MV at HVt for 2 hours. Mice in the SpV group also received Evans blue. Lungs were immediately frozen, and Evans blue in the lung tissues of all experimental groups was quantified and expressed as μg/mg lung tissue. Values represent n = 6 per experimental condition. *P ⩽ 0.02, HVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.001, HVt versus SpV of Nrf2^−/− mice.

Figure 3.

Effect of mechanical ventilation (MV) on cytokine expression in the lungs of Nrf2^+/+ and Nrf2^−/− mice. Lung tissues of Nrf2^+/+ and Nrf2^−/− mice subjected to spontaneous ventilation (SpV) or high Vt (HVt), as detailed in Figure 1, were homogenized and centrifuged. Supernatants were collected, and protein was estimated by the BCA method. Equal amounts of protein were used for the cytokine multiplex bead-based immunoassay kit as detailed in Methods. The concentration of each cytokine in the lung tissues was quantified using a standard curve, and the levels of each cytokine were expressed as pg/mg lung tissue. (A) Cytokine profiles altered by MV that differed between Nrf2^+/+ and Nrf2^−/− mice. (B) Cytokine profiles altered by MV that did not differ between two genotypes. Data are means ± SEM (n = 3–4 mice per group). *P ⩽ 0.05, HVt versus SpV groups of each genotype. **P ⩽ 0.001, HVt versus SpV groups of the respective genotype. ^#P ⩽ 0.05, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to HVt.

Figure 4.

Effects of mechanical ventilation (MV) on antioxidant gene expression and reduced glutathione/oxidized glutathione (GSH/GSSG) levels. (A) Mice were subjected to spontaneous breathing (sham-treated) or MV at HVt for 2 hours, and total RNA and protein were isolated as detailed in Methods. cDNA was produced by reverse-transcription of RNA (100 ng) extracted from the left lungs of mice (n = 3 per group) followed by polymerase chain reaction amplification using primers specific for mouse Gclc, Gclm, and GPx2. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. (B) Western blot analysis of glutathione-S-transferase (GST) isoenzymes in the lungs of Nrf2^+/+ and Nrf2^−/− mice. (C) Bands were quantified using a Bio-Rad Gel Doc 2000 System. Data are means ± SEM from two independent experiments and are expressed as percentage increases relative to the vehicle-treated Nrf2^+/+ group. Effect of MV on the specific activities of GST (D) and GPx (E) enzymes in the lungs of Nrf2^+/+ and Nrf2^−/− mice. Values represent means ± SEM (n = 3 per group). *P ⩽ 0.05, HVt versus control (SpV) of the same genotype. (F) GSH/GSSG ratios of lung tissue samples of Nrf2^+/+ and Nrf2^−/− mice subjected to SpV and MV at HVt were determined as detailed in Methods. Data are presented as means ± SEM (n = 3–4). *P ⩽ 0.05, HVt versus SpV control of Nrf2^−/− mice.

Figure 4.

Effects of mechanical ventilation (MV) on antioxidant gene expression and reduced glutathione/oxidized glutathione (GSH/GSSG) levels. (A) Mice were subjected to spontaneous breathing (sham-treated) or MV at HVt for 2 hours, and total RNA and protein were isolated as detailed in Methods. cDNA was produced by reverse-transcription of RNA (100 ng) extracted from the left lungs of mice (n = 3 per group) followed by polymerase chain reaction amplification using primers specific for mouse Gclc, Gclm, and GPx2. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. (B) Western blot analysis of glutathione-S-transferase (GST) isoenzymes in the lungs of Nrf2^+/+ and Nrf2^−/− mice. (C) Bands were quantified using a Bio-Rad Gel Doc 2000 System. Data are means ± SEM from two independent experiments and are expressed as percentage increases relative to the vehicle-treated Nrf2^+/+ group. Effect of MV on the specific activities of GST (D) and GPx (E) enzymes in the lungs of Nrf2^+/+ and Nrf2^−/− mice. Values represent means ± SEM (n = 3 per group). *P ⩽ 0.05, HVt versus control (SpV) of the same genotype. (F) GSH/GSSG ratios of lung tissue samples of Nrf2^+/+ and Nrf2^−/− mice subjected to SpV and MV at HVt were determined as detailed in Methods. Data are presented as means ± SEM (n = 3–4). *P ⩽ 0.05, HVt versus SpV control of Nrf2^−/− mice.

Figure 4.

Effects of mechanical ventilation (MV) on antioxidant gene expression and reduced glutathione/oxidized glutathione (GSH/GSSG) levels. (A) Mice were subjected to spontaneous breathing (sham-treated) or MV at HVt for 2 hours, and total RNA and protein were isolated as detailed in Methods. cDNA was produced by reverse-transcription of RNA (100 ng) extracted from the left lungs of mice (n = 3 per group) followed by polymerase chain reaction amplification using primers specific for mouse Gclc, Gclm, and GPx2. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. (B) Western blot analysis of glutathione-S-transferase (GST) isoenzymes in the lungs of Nrf2^+/+ and Nrf2^−/− mice. (C) Bands were quantified using a Bio-Rad Gel Doc 2000 System. Data are means ± SEM from two independent experiments and are expressed as percentage increases relative to the vehicle-treated Nrf2^+/+ group. Effect of MV on the specific activities of GST (D) and GPx (E) enzymes in the lungs of Nrf2^+/+ and Nrf2^−/− mice. Values represent means ± SEM (n = 3 per group). *P ⩽ 0.05, HVt versus control (SpV) of the same genotype. (F) GSH/GSSG ratios of lung tissue samples of Nrf2^+/+ and Nrf2^−/− mice subjected to SpV and MV at HVt were determined as detailed in Methods. Data are presented as means ± SEM (n = 3–4). *P ⩽ 0.05, HVt versus SpV control of Nrf2^−/− mice.

Figure 5.

Effects of exogenous antioxidant on mechanical ventilation (MV)-induced lung inflammation and injury. (A and B) Wild-type and Nrf2^−/− mice were pretreated for 30 minutes with N-acetyl-l-cysteine (NAC) (100 mg/kg body weight) before MV. Mice were randomly assigned to spontaneous ventilation (SpV) or high Vt (HVt) groups. Mice were subjected to MV for 2 hours and then allowed to recover for 1 hour. Bronchoalveolar lavage fluid (BALF) was collected to measure the total cells (A) and neutrophils (B) as detailed previously. Data are presented as means ± SEM (n = 4–6). The albumin concentration in the BALF was analyzed by ELISA. (C) The values of the static control group were considered to be 100%. MV-induced albumin leakage is presented as the percentage increase over static control (n = 4–6 animals/group).*P < 0.05 versus vehicle control. (D) The effect of NAC intervention on MV-induced vascular leakage was assessed by extravasation of Evans blue into the lung parenchyma. Nrf2^+/+ and Nrf2^−/− mice were pretreated with NAC as described previously and subjected to MV at HVt or SpV for 2 hours. Evans blue was injected into the jugular vein 30 minutes before termination of ventilation, and extravasation of Evans blue in the lung tissue was quantified and expressed as μg/mg lung tissue. Values represent n = 5 to 6 mice per experimental condition. The values for the static control group were considered to be 100%. MV-induced albumin leakage in the lung is expressed as the percentage increase over the static control. *P ⩽ 0.05. EBD = Evans blue dye.

Figure 6.

Lung injury and inflammatory response of Nrf2^−/− and Nrf2^+/+ mice subjected to mechanical ventilation (MV) at low Vt (LVt) for 4 hours. (A) Mice (n = 5–6 mice per group) were subjected to MV at LVt or spontaneous ventilation (SpV) for 4 hours, and lung injury and neutrophil infiltration were assessed as detailed in Figure 2. Total cells (A), neutrophils (B), macrophages (C), and protein (D) in the bronchoalveolar lavage fluid (BALF) were analyzed as in Figure 1. *P ⩽ 0.05, LVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.05, LVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.05, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to LVt. (E) Neutrophils in the alveolar space were quantified by immunohistochemistry using antineutrophil antibody as in Figure 2B. Data represent means ± SEM (n = 5–6). *P ⩽ 0.001, LVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.004, HVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.002, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to LVt. (F) Vascular leakage was assessed by the extravasation of Evans blue into lung parenchyma. Mice (n = 6–7 per group) subjected to SpV or MV at LVt for 4 hours were injected with Evans blue into the jugular vein 30 minutes before termination. *P ⩾ 0.02, LVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.001, HVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.05, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to LVt. EBD = Evans blue dye.

Figure 6.

Lung injury and inflammatory response of Nrf2^−/− and Nrf2^+/+ mice subjected to mechanical ventilation (MV) at low Vt (LVt) for 4 hours. (A) Mice (n = 5–6 mice per group) were subjected to MV at LVt or spontaneous ventilation (SpV) for 4 hours, and lung injury and neutrophil infiltration were assessed as detailed in Figure 2. Total cells (A), neutrophils (B), macrophages (C), and protein (D) in the bronchoalveolar lavage fluid (BALF) were analyzed as in Figure 1. *P ⩽ 0.05, LVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.05, LVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.05, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to LVt. (E) Neutrophils in the alveolar space were quantified by immunohistochemistry using antineutrophil antibody as in Figure 2B. Data represent means ± SEM (n = 5–6). *P ⩽ 0.001, LVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.004, HVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.002, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to LVt. (F) Vascular leakage was assessed by the extravasation of Evans blue into lung parenchyma. Mice (n = 6–7 per group) subjected to SpV or MV at LVt for 4 hours were injected with Evans blue into the jugular vein 30 minutes before termination. *P ⩾ 0.02, LVt versus SpV of Nrf2^+/+ mice. **P ⩽ 0.001, HVt versus SpV of Nrf2^−/− mice. ^#P ⩽ 0.05, Nrf2^−/− mice versus Nrf2^+/+ mice subjected to LVt. EBD = Evans blue dye.