Identification of oxidative stress and Toll-like receptor 4 signaling as a key pathway of acute lung injury

Abstract

Multiple lung pathogens such as chemical agents, H5N1 avian flu, or SARS cause high lethality due to acute respiratory distress syndrome. Here we report that Toll-like receptor 4 (TLR4) mutant mice display natural resistance to acid-induced acute lung injury (ALI). We show that TLR4-TRIF-TRAF6 signaling is a key disease pathway that controls the severity of ALI. The oxidized phospholipid (OxPL) OxPAPC was identified to induce lung injury and cytokine production by lung macrophages via TLR4-TRIF. We observed OxPL production in the lungs of humans and animals infected with SARS, Anthrax, or H5N1. Pulmonary challenge with an inactivated H5N1 avian influenza virus rapidly induces ALI and OxPL formation in mice. Loss of TLR4 or TRIF expression protects mice from H5N1-induced ALI. Moreover, deletion of ncf1, which controls ROS production, improves the severity of H5N1-mediated ALI. Our data identify oxidative stress and innate immunity as key lung injury pathways that control the severity of ALI.

Figures

Figure 1

TLR4 Is a Susceptibility Gene for Acute Lung Injury (A) Changes in lung elastance after acid or saline treatment in WT and tlr4−/− mice. n = 8–10 for acid-treated groups, n = 6 for saline-treated groups. ∗∗p < 0.01 for the whole time course. (B) Lung edema formation after acid or saline treatment. ∗p < 0.05. (C) Lung histopathology. H&E staining. Original magnifications × 200. (D) Lung elastance after acid treatment of WT mice transplanted with WT bone marrow (BM) (WT BM → WT), WT mice receiving tlr4−/− BM (tlr4−/− BM → WT), tlr4−/− mice transplanted with WT BM (WT BM → tlr4−/−), and tlr4−/− mice receiving tlr4−/− BM (tlr4−/− BM → tlr4−/−). n = 6 for each group. (E) Lung edema formation and (F) lung histopathology 3 hr after acid injury in WT BM → WT, tlr4−/− BM → WT, WT BM → tlr4−/−, and tlr4−/− BM → tlr4−/− chimeras. In (D) and (E), ∗p < 0.05 comparing WT BM → WT or WT BM → tlr4−/− mice. n = 6 for each group. H&E staining. Original magnifications × 200. Data in (A), (B), (D), and (E) are mean values ± SEM.

Figure 2

TRIF-TRAF6-NFκB-Cytokine Signaling Mediates the Severity of Acid Aspiration-Induced Acute Lung Injury (A) Lung elastance after acid or saline treatment of WT and trif−/− mice. n = 6–10 for acid-treated groups, n = 4 for saline-treated groups. ∗∗p < 0.01 for the whole time course. (B) Reduced inflammation and hyaline membrane formation in acid-treated trif−/− mice. H&E staining. Original magnifications × 200. (C) TRAF6 protein expression in bone marrow macrophages isolated from Lys-Cre(−) TRAF6flox/flox (TRAF6MC-WT) and Lys-Cre(+) TRAF6flox/flox (TRAF6MC-KO) mice. β-actin protein is shown as control. Data from two different mice (1, 2) are shown for each genotype. (D) Changes in lung elastance in TRAF6MC-WT (n = 6) and TRAF6MC-KO mice (n = 4) following acid aspiration. ∗p < 0.05 for the whole time course. (E) Reduced inflammatory cell infiltration, bleeding, and hyaline membrane formation in acid-treated TRAF6MC-KO mice. H&E staining. Original magnifications × 200. (F) Immunolocalization of Ser276-phosphorylated NFκBp65 in lung tissue of saline-treated control WT mice and acid-treated WT, myd88−/−tlr4−/−, trif−/−, and TRAF6MC-KO mice. Note the accumulation of phosphorylated NFκBp65 in nuclei of macrophages (red arrows) in acid-treated WT and myd88−/− mice, which is absent in lung macrophages (black arrows) from acid-treated tlr4−/−, trif−/−, and TRAF6MC-KO mice. Original magnifications × 400. (G) IL-6 levels in lung tissue after acid treatment in WT Balb/c, WT BL6, control TRAF6MC-WT (wild-type), tlr4−/− (Balb/c background), tlr9−/− (BL6 background), tlr3−/− (BL6), myd88−/− (Balb/c) trif−/− (BL6), irf3−/− (BL6), and TRAF6MC-KO mice. IL-6 levels were determined by whole lung tissue ELISA. n = 5–8 animals for each group. ∗p < 0.05. (H) Lung elastance after acid treatment in interleukin-6 mutant (il-6−/−) and control WT mice. n = 5–6 each group. ∗p < 0.05. (I) Improved lung histopathology in acid-treated il-6−/− mice. H&E staining. Original magnifications × 200. Lungs were analyzed 3 hr after treatment. (J) Schematic diagram depicting the role of TLR4 in ALI. Data in (A), (D), (G), and (H) are mean values ± SEM.

Figure 3

Formation of Oxidized Phospholipids in Acute Lung Injury (A) Increase in IL-6 production from baseline in WT, tlr4−/−, myd88−/−, and trif−/− alveolar macrophages treated with LPS, BAL fluid from normal control, or BAL fluid from acid-treated WT mice. ∗∗p < 0.01. Data are from four separate experiments. (B) Increases in ROS expression in alveolar macrophages obtained from WT mice 60 min after treatment with saline (background control, blue) or acid (white). A representative histogram is shown among five separate experiments. (C) Increased TLR4 surface expression in alveolar macrophages from WT mice treated with saline (control) or acid. Data are from five separate experiments. ∗p < 0.05. (D) Immunohistochemistry for OxPLs detected by the mAb EO6 in lungs of saline-treated control (upper panel) and acid-treated WT mice (lower panel). OxPLs were localized to inflammatory exudates lining the injured alveoli (arrows) in acid-treated lungs. Original magnifications × 400. Lungs were analyzed 3 hr after acid treatment. (E) BAL fluid from acid-treated mice (BAL acid) induces large amounts of IL-6 in WT but not tlr4−/− alveolar macrophages. BAL fluid plus an isotype-matched control Ab was compared to BAL fluid plus the mAb EO6. ∗p < 0.05. Data are from four separate experiments. (F) Increase in IL-6 from baseline in peritoneal macrophages isolated from WT and tlr4−/− mice in response to nonoxidized PLs or oxidized PLs. ∗∗p < 0.01. (G) Lung elastance in WT mice following intratracheal administration of saline, nonoxidized PLs, and oxidized PLs. n = 4–6 for each group. ∗∗p < 0.01 for the whole time course. (H) Lung elastance in WT and tlr4−/− mice following treatment with nonoxidized PLs or oxidized PLs. n = 4 for each group. ∗∗p < 0.01 for the whole time course. Data in (A), (C), and (E)–(H) are mean values ± SEM.

Figure 4

Oxidized PAPC Induces ALI and IL-6 Production via TLR4 (A) Increases in IL-6 production from baseline (unstimulated control) in lung tissue macrophages from WT mice in response to PAPC (10 μg/ml) or OxPAPC (10 μg/ml) in the presence of an isotype-matched control mAb or the mAb EO6. ∗p < 0.05 comparing mAb EO6(−). ∗∗p < 0.01 comparing OxPAPC EO6(−) and PAPC EO6(−). Data are from four separate experiments. (B) Increase in IL-6 from baseline (control) in lung tissue macrophages isolated from WT, tlr4−/−, myd88−/−, and trif−/− mice in response to LPS (1 ng/ml) or OxPAPC (10 μg/ml). ∗p < 0.05 comparing OxPAPC-treated WT or myd88−/− macrophages ± EO6. ∗∗p < 0.01 compared to OxPAPC EO6(−) treated WT macrophages. Data are from three separate experiments. (C) Lung elastance in WT mice following intratracheal administration of PAPC or OxPAPC (20 μg/g body weight). n = 4 for each group. ∗∗p < 0.01 for the whole time course. (D) Western blots of IκBα and β-actin in lungs isolated from mice 90 min after saline treatment (control), OxPAPC, or PAPC (20 μg/g body weight). Blots are representative of three separate experiments. (E) Lung elastance at 1.5 hr in WT, tlr4−/−, myd88−/−, and trif−/− mice following OxPAPC challenge (20 μg/g body weight). The correct genetic background controls are shown. n = 4 for each group. ∗p < 0.05. Data are mean values ± SEM.

Figure 5

Inactivated H5N1 Avian Influenza Virus Can Induce OxPLs and ALI in Mice (A) Changes in lung elastance in WT mice following intratracheal administration of vehicle (n = 3), inactivated H1N1 (n = 6), or inactivated H5N1 (n = 8) viruses. ∗p < 0.05 for the whole time course. (B) Lung edema formation 5 hr after vehicle, inactivated H1N1, or inactivated H5N1 treatment. ∗∗p < 0.01. (C) Immunohistochemistry for influenza A nucleoprotein in lung of WT mice challenged with vehicle, inactivated H1N1, or inactivated H5N1 viruses. Influenza A protein-positive cells were present in the lower respiratory tract including alveolar macrophage (arrowhead) and pneumocytes (arrows). Original magnifications × 400. (D) ROS expression in alveolar macrophages from WT mice treated with control vehicle (blue), inactivated H1N1 virus (white, upper panel), or inactivated H5N1 virus (white, lower panel). (E) TLR4 expression in alveolar macrophages from WT mice treated with vehicle (blue), inactivated H1N1 virus (white), or inactivated H5N1 virus (white). Representative histograms are shown for five separate experiments. Data in (D) and (E) are at 1 hr after viral challenge. (F) Immunohistochemistry for influenza A nucleoprotein and EO6-detectable OxPLs in lung of WT mice infected with live H5N1 avian influenza virus. H5N1-infected mice contain large numbers of influenza A protein-positive cells (brown) in the lower respiratory tract (left panels, arrowheads). OxPLs were localized to inflammatory cells (arrowheads) as well as inflammatory exudates (arrow). Lungs were analyzed 4 days after infection. Original magnifications × 100 (upper panels) and × 200 (lower panels). Data are mean values ± SEM.

Figure 6

TLR4- or TRIF-Deficient Mice Show Less Severe ALI to Challenge with Inactivated H5N1 Avian Influenza Virus (A) Lung elastance in WT and tlr4−/− mice following administration of vehicle or inactivated H5N1 viruses. n = 5 for each group. ∗∗p < 0.01 for the whole time course. (B) Lung edema formation in WT versus tlr4−/− mice 5 hr after H5N1 challenge. ∗∗p < 0.01. (C) Lung pathology (top; H&E staining) and immunolocalization of NFκBp65 (bottom) in lung tissue of H5N1-challenged WT and tlr4−/− mice. Note nuclear accumulation of Ser276-phosphorylated NFκBp65 in macrophages from WT lung (red arrows). In H5N1-treated tlr4−/− lungs, nuclear NF-κB accumulation was mostly absent from macrophages (green arrows) albeit present in a few cells (red arrow). Original magnifications × 400. (D) Lung elastance in WT and trif−/− mice after administration of vehicle or inactivated H5N1 viruses. n = 5 for each group. ∗∗p < 0.01 for the whole time course. (E) Lung pathology (top; H&E staining) and immunolocalization of NFκBp65 (bottom) in lung tissue of H5N1-challenged WT and trif−/− mice. Original magnifications × 400. (F) BAL fluid from WT mice 5 hr after a H5N1 challenge (BAL H5N1) triggers IL-6 in WT but not tlr4−/− alveolar macrophages. BAL fluid from H5N1-treated mice plus an isotype-matched control mAb (−) was compared to BAL fluid from H5N1-treated mice plus the mAb EO6 (+). ∗p < 0.05. Data are mean values ± SEM.

Figure 7

The Oxidative Stress Machinery Controls the Severity of ALI in Response to Inactivated H5N1 Avian Influenza Viruses (A) ROS expression in control PBMCs and PBMCs from the same donor treated with inactivated H1N1 or inactivated H5N1 viruses. Data are from CD14+ gated PBMCs. Representative histograms are shown for six separate donors. (B) TLR4 expression in control human PBMCs and H5N1- or H1N1-treated PBMCs. Data are from CD14+ gated PBMCs. Representative histograms are shown for six different donors. (C) Immunofluorescence staining of TLR4 in PBMCs treated with formulation control or inactivated H5N1. TLR4 was distributed throughout the cytoplasm in control cells (top). PBMCs treated with inactivated H5N1 showed peripheralization of TLR4 (bottom). (D) Lung elastance in WT and ncf1−/− mice following intratracheal administration of inactivated H5N1 viruses. n = 4 for each group. ∗∗p < 0.01 for the whole time course. (E) Lung pathology (top; H&E staining) and immunolocalization of oxidized PLs (bottom) in lung tissue of H5N1-challenged WT and ncf1−/− mice. Original magnifications × 200. (F) Immunohistochemistry of OxPLs in lungs from two patients infected with H5N1 avian influenza virus (top) and two patients infected with SARS-coronavirus (bottom). (G) Immunohistochemistry of OxPLs in rhesus monkeys and rabbits infected with Bacillus anthracis (left panels) and Cynomolgus monkeys infected with Monkey Pox or Yersinia pestis (right). In (F) and (G), OxPLs detected by the mAb EO6 were seen in inflammatory exudates lining the injured alveoli (arrows) and macrophages (triangles). Original magnifications × 400. Data are mean values ± SEM.