Impaired Autophagic Activity Contributes to the Pathogenesis of Bronchopulmonary Dysplasia. Evidence from Murine and Baboon Models

Abstract

Bronchopulmonary dysplasia (BPD) is a common and serious complication associated with preterm birth. The pathogenesis of BPD is incompletely understood, and there is an unmet clinical need for effective treatments. The role of autophagy as a potential cytoprotective mechanism in BPD remains to be fully elucidated. In the present study, we investigated the role and regulation of autophagy in experimental models of BPD. Regulation and cellular distribution of autophagic activity during postnatal lung development and in neonatal hyperoxia-induced lung injury (nHILI) were assessed in the autophagy reporter transgenic GFP-LC3 (GFP-microtubule-associated protein 1A/1B-light chain 3) mouse model. Autophagic activity and its regulation were also examined in a baboon model of BPD. The role of autophagy in nHILI was determined by assessing lung morphometry, injury, and inflammation in autophagy-deficient Beclin 1 heterozygous knockout mice (Becn1^+/-). Autophagic activity was induced during alveolarization in control murine lungs and localized primarily to alveolar type II cells and macrophages. Hyperoxia exposure of neonatal murine lungs and BPD in baboon lungs resulted in impaired autophagic activity in association with insufficient AMPK (5'-AMP-activated protein kinase) and increased mTORC1 (mTOR complex 1) activation. Becn1^+/- lungs displayed impaired alveolarization, increased alveolar septal thickness, greater neutrophil accumulation, and increased IL-1β concentrations when exposed to nHILI. Becn1^+/- alveolar macrophages isolated from nHILI-exposed mice displayed increased expression of proinflammatory genes. In conclusion, basal autophagy is induced during alveolarization and disrupted during progression of nHILI in mice and BPD in baboons. Becn1^+/- mice are more susceptible to nHILI, suggesting that preservation of autophagic activity may be an effective protective strategy in BPD.

Keywords: GFP-LC3; autophagy; beclin; bronchopulmonary dysplasia; macrophage.

Figure 1.

Autophagic activity is induced in the lung during alveolarization. (A) Representative immunoblot analysis for GFP, LC3-I and -II, and p62 using GFP-LC3 (GFP–microtubule-associated protein 1A/1B-light chain 3) whole-lung homogenates harvested at Postnatal Day (P) 0.5, P1, and P3. Relative protein concentrations normalized to ACTB (β-actin) are indicated under the GFP and p62 blots and LC3-II/LC3-I ratio is indicated under the LC3 blot. (B) Representative immunoblot analysis for GFP and p62 using GFP-LC3^+/− (GFP^+/−) and GFP-LC3^+/−/Becn1^+/− (GFP^+/−/Becn1^+/−) whole-lung homogenates at P7. ACTB was used as a loading control. Relative protein concentrations are indicated under the blots. (C) GFP-LC3 mice were exposed to normoxia (21% O₂) starting at birth and killed on P1, 3, 5, 7, 10, or 14. (D and E) Left lungs were homogenized and used for IB and densitometric analysis of GFP (D) and p62 protein (E). Data are shown as mean ± SEM; n = 3–4 per group. ****P < 0.0001 versus all groups except for P10.

Figure 2.

Autophagic activity is initially induced and then impaired during progression of neonatal hyperoxia-induced lung injury. GFP-LC3 mice were randomized to normoxia (N; 21% O₂) or hyperoxia (H; 75% O₂) exposure within 12 hours of birth and killed on P3, 5, or 7. (A) Left lung homogenates were analyzed for GFP, p62, Becn1 (Beclin 1), AMPK (5′-AMP–activated protein kinase), phospho-AMPK (p-AMPK), and phospho-S6 (p-S6) expression by IB. Relative protein concentrations of P3, P5, and P7 samples, normalized to ACTB, were determined by densitometry and are presented in B–D, respectively. Data are shown as mean ± SEM; n = 4 per group. *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 3.

Autophagic activity is impaired in baboon lungs with bronchopulmonary dysplasia (BPD). Baboons were delivered by cesarean section at 125 days of gestation (∼27 wk of human gestation) and treated with exogenous surfactant, pro re nata O₂, and mechanical ventilation for 14 days. Control baboons were delivered at 140 days of gestation and killed immediately before their first breath. (A) Lung homogenates were prepared and analyzed by IB with antibodies against cleaved caspase 3 (cl. cas3), p-S6, AMPK, p-AMPK, LC3-I and -II, p62, TFEB (transcription factor EB), BECN1, Parkin, and PINK1 (phosphatase and tensin homolog–induced kinase 1). (B) Relative protein expression levels, normalized to ACTB (except for LC3-II, which was normalized to LC3-I), were determined by densitometry. Data are shown as mean ± SEM; n = 7 per group. *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 4.

GFP-LC3 expression in neonatal murine lungs. Immunofluorescence staining was performed on cryosections of GFP-LC3 lungs with primary antibodies against SPC (surfactant protein C) (top row), ACTA2 (α-SMA) (second row), FABP4 (fatty acid-binding protein, adipocyte) (third row), p-AMPK (fourth row), and p-S6 (fifth row). The secondary antibody was Alexa Fluor 594 goat antirabbit IgG (red). Representative images are shown. White arrows indicate examples of co-localization of GFP-LC3 (green) with each of the primary antibody targets. Scale bars, 25 μm and 50 μm. A = airway; V = vessel.

Figure 5.

Becn1^+/− mice are more susceptible to neonatal hyperoxia-induced lung injury. Becn1^+/− (knockout [KO]) and wild-type (WT) littermate control mice (both on C57BL/6 background) were exposed to normoxia (21% O₂) or hyperoxia (75% O₂) between P1 and P7. At P7, mice were killed, and their lungs were inflation fixed with 10% formalin to 25 cm H₂O and embedded in paraffin. (A) Sections were stained with hematoxylin and eosin. (B) Mean linear intercept as a surrogate for alveolar diameter and alveolar septal thickness were quantified using ImageJ and NIS-Elements Basic Research (Nikon Instruments) software, respectively. (C and D) In other cohorts of mice, left lungs were snap frozen, homogenized in radioimmunoprecipitation assay buffer containing protease inhibitors, and analyzed by IB (C) and densitometry (D) for the indicated proteins. (E and F) A TUNEL assay was performed on paraffin-embedded mouse lung tissues (E), and TUNEL-positive cells were quantified (F). Data were obtained from a minimum of three separate litters and are shown as mean ± SEM; n = 7–10 per group. *P < 0.05 and **P < 0.01. Scale bars, 50 μm.

Figure 6.

Becn1^+/− lungs and alveolar macrophages exhibit enhanced inflammatory signaling. (A) BAL fluid (BALF) was obtained from Becn1^+/− (KO) and WT littermates after hyperoxia exposure between P1 and P7. Macrophages (Mϕ) and polymorphonuclear neutrophils (PMN) were enumerated on cytospins stained with Diff-Quik. (B) BALF was obtained from other cohorts of hyperoxia-exposed Becn1^+/− (KO) and WT littermate mice, and alveolar macrophages were isolated by adhesion purification. Total RNA was isolated, and qRT-PCR was performed for a panel of genes indicated on the figure. (C) Caspase 1 activity was measured in right lung homogenates from hyperoxia-exposed Becn1^+/− (KO) and WT littermates using a fluorescence-based activity assay. Relative caspase 1 activity level was expressed as a percentage of activity detected in WT samples. (D) IL-1β protein was measured in whole-lung lysates from hyperoxia-exposed Becn1^+/− (KO) and WT littermates using ELISA. All data were derived from a minimum of three different experiments with n = 6–14 per group. Data are shown as mean ± SEM; *P < 0.05 and **P < 0.01.