Mitochondrial DNA damage and subsequent activation of Z-DNA binding protein 1 links oxidative stress to inflammation in epithelial cells

Abstract

This report identifies mitochondrial DNA (mtDNA) as a target and active mediator that links low-level oxidative stress to inflammatory response in pulmonary epithelial cells. Extrusion of mtDNA into the bronchoalveolar lavage fluid occurs as an early event in mice subjected to cigarette smoke injury, concomitantly with the depletion of mtDNA in the lung tissue. In cultured lung epithelial cells, prolonged, low-level oxidative stress damages the mtDNA, without any detectable damage to the nuclear DNA. In turn, cellular depletion of the mtDNA occurs, together with a transient remodeling of cellular bioenergetics and morphology - all without any detectable impairment in overall cell viability. Damaged mtDNA first enters the cytoplasm, where it binds to Z-DNA binding protein 1 (ZBP1) and triggers inflammation via the TANK-binding kinase 1 /interferon regulatory factor 3 signaling pathway. Fragments of the mtDNA are subsequently released into the extracellular space via exosomes. MtDNA-containing exosomes are capable of inducing an inflammatory response in naïve (non-oxidatively stressed) epithelial cells. In vivo, administration of isolated mtDNA into the in lungs of naïve mice induces the production of pro-inflammatory mediators, without histopathologic evidence of tissue injury. We propose that mtDNA-specific damage, and subsequent activation of the ZBP1 pathway, is a mechanism that links prolonged, low-level oxidative stress to autocrine and paracrine inflammation during the early stages of inflammatory lung disease.

Figure 1

Mitochondrial DNA is released into the bronchoalveolar lavage fluid as an early event in a murine model of cigarette smoke induced lung injury. Early presence of mtDNA (A) but not of nuclear DNA (B) in BALF of mice exposed to cigarette smoke induced lung injury. MtDNA content is depleted (C) and mtDNA integrity is impaired (D) in the lung tissue of smoke injured mice. 6–8 animals were used for each experimental end-point. Data represent average ± SEM. **p < 0.01 vs. control.

Figure 2

Non-cytotoxic levels of oxidative stress induce mitochondrial DNA depletion, impair cellular bioenergetics and stimulate inflammation. BEAS2B cells were treated with several concentrations of GOx for 1 h and immediately post-challenge as well as 24 h later were measured: (A) mitochondrial DNA integrity, (B) mitochondrial DNA content, (C) apoptotic/necrotic cell death; (D) mitochondrial respiration; (E) glycolysis. (F) 3-D reconstruction of cellular morphology of BEAS2B at 24 h after GOx-treatment was visualized using ATPA-specific antibody (green), β-tubulin (red) and nucleus (DAPI, blue). (G) Amount of pro-inflammatory mediators in medium of BEAS2B cells measured at 24 h post GOx-treatment. Data represent average ± SEM of n = 5 biological replicates. (H) Lack of enhanced production of IL-6 in BEAS 2B cells stimulated with inactive GOx at 24 h. Representative data of n = 3 independent experiments are shown. *p < 0.05, **p < 0.01 vs. control.

Figure 3

Damaged mitochondrial DNA activates ZBP1/TBK1/IRF3 signaling pathway. Interaction between BrDU-labelled mtDNA and TFAM (A) and BrDU-labelled mtDNA and ZBP1 (B) in control and GOx-treated cells at 1 h. (C) Interaction between ZBP1/TBK1, IRF3/TBK1, ZBP1/P-Tyr and ZBP1/P-Ser in control and GOx-treated cells at 1 h. (D) Interaction between IRF3/TBK1 in pre-treated with 3 μM of CsA in GOx-treated BEAS 2B cells. (E) Level of ZBP1 depletion, shown by Western blotting. (F) Expression of IL-6 and IL-8 in unstressed and GOx-treated (0.006 U/ml for 1 h) in control and ZBP1-depleted BEAS2B cells. Representative images of n = 3 independent experiments are shown. Data represent average ± SEM of n = 5 biological replicates. *p < 0.05 vs. control.

Figure 4

Oxidatively stressed cells induce release of damaged mtDNA via exosomes. (A) Comparison of mitochondrial and nuclear DNA in exosomal fraction isolated at 24 h from BEAS 2B cells treated with 0.006 U/ml GOx. (B) Time-dependent release of the mtDNA from control and BEAS 2B cells treated with 0.006 U/ml of GOx. (C) Protein level of CD63, marker of exosomes, in BEAS 2B cell extract and in exosomal fraction. MtDNA release from cells pretreated for 1 h with (D) 300 nM of Mito-TEMPO; (E) 3 μM cyclosporine-A or (F) 3 μM colchicine in control and 0.006 U/ml of GOx-treated BEAS 2B cells. Data represent average ± SEM of n = 5 biological replicates. Representative images of n = 3 independent experiments are shown. **p < 0.01 vs. control.

Figure 5

Inflammation respond to mitochondrial DNA in vivo. (A) Inflammatory response triggered in BALF of naïve mice at 24 h post intratracheal administration of 1 μg of isolated mtDNA. (B) Time-dependent expression of IL-1α in BEAS 2B cells incubated 100 ng of isolated mtDNA. (C) Model of inflammation induced by mtDNA/ZBP1 pathway. Data represent average ± SEM using n = 6–8 animals or n = 3 for biological replicates per experimental end-point. *p < 0.05, **p < 0.01 vs. control.