AMPK Activation: Respiratory Panacea?

Improvements in perinatal care have led to the increased survival of preterm infants (1). The development of the gas exchange apparatus of the lung occurs during late lung development, and these infants are at risk of developing bronchopulmonary dysplasia (BPD) (2). BPD is a condition of chronic lung disease caused by the disruption of pulmonary development with superimposed lung injury characterized by alveolar simplification, microvascular disturbances, fibrosis, and variable airway injury (3). BPD remains the most common complication of extreme preterm birth, affecting ∼35% of U.S. newborns of extremely low gestational age (1). To date, there is poor understanding of the mechanisms directing the saccular and alveolar stages of lung development, the so-called late lung development. Uncovering these developmental pathways is critical to develop strategies either to promote lung development or to protect lungs from injury and boost lung repair.

Data strongly support the role of inflammation in BPD pathogenesis, but the exact pathophysiology remains unclear (2). In this issue of the Journal, Soni and colleagues (pp. 279–287) report on their examination of the effect of 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR), a direct AMP-activated protein kinase (AMPK) activator, in a murine hyperoxic model of BPD with a focus on the role of autophagy in AMPK-induced macrophage responses (4). Previously, the authors have reported a link between autophagy, AMPK, and increased expression of proinflammatory genes in alveolar macrophages in a murine model of neonatal hyperoxia–induced lung injury (5). Autophagic activity was induced during alveolarization in macrophages and alveolar type II cells in murine lungs; hyperoxic exposure was associated with impaired autophagic activity and insufficient AMPK activation, and autophagy-deficient Becn1^+/− mice had increased injury when exposed to hyperoxia with alveolar macrophages that exhibited increased expression of proinflammatory genes (5). In the study by Soni and colleagues, AICAR treatment resulted in AMPK activation and increased autophagic markers in the lung and BAL cells from neonatal hyperoxia–exposed mice. This was associated with reduced alveolar simplification, increased density of ACTA2-positive myofibroblasts at the septal tips, and an increase in CD31 expression, a pan–endothelial cell marker, in the lung, along with decreased expression of proinflammatory (M1-like) and increased expression of proreparative (M2-like) genes in alveolar macrophages. When they examined the effects of AICAR in mice with a myeloid cell–specific deficiency of Becn1 (Becn1^F/F;LysM-cre), they did not find the same antiinflammatory phenotype in macrophages, indicating that the effects of AICAR on alveolar macrophages are autophagy dependent (4).

Human lung macrophage differentiation has been previously linked to the development of BPD in preterm infants. Sahoo et al. measured changes in lung macrophage gene expression in preterm infants at risk of BPD and found that patients who eventually developed BPD had higher inflammatory mediator expression in tracheal aspirate macrophages, even on the first day of life, and that macrophages from patients who were resilient to BPD dynamically expressed a set of genes associated with the M2 phenotype (6). Lung macrophages, however, do not always display a clear M1 or M2 functional phenotype, and the development of fetal lung macrophages into mature alveolar macrophages is influenced by activation and may include features of both proinflammatory and alternative activation paradigms (7). Autophagy has been implicated previously in aberrant lung development, and increased autophagy through regulatory-associated protein of mechanistic target of rapamycin (MTOR) complex 1 (RPTOR) antagonism has been linked to improved lung structure and cardiac remodeling in mouse pups exposed to hyperoxia (8). In this context, the beneficial effects of autophagy were mostly attributed to antiapoptotic effects in lung epithelial cells (8). Autophagy-mediated control of macrophage polarization, along with other aspects of macrophage regulation, has been studied extensively in other disease paradigms and has been associated with NFκB and mechanistic target of rapamycin activity (9). Similarly, AMPK regulation was crucial for phagocytosis-induced macrophage skewing from the M1 phenotype toward the M2 phenotype and resolution of inflammation during a skeletal muscle regeneration model (10). Moreover, AMPK’s direct input into autophagy through the protein kinases ULK1 and ULK2 is well described (11). In sum, the study by Soni and colleagues brings together observations previously made in other organs and disease paradigms to highlight, for the first time, a novel role for autophagy in lung macrophage maturation during lung development.

AMPK activation has been associated with favorable metrics in BPD models in the past (12, 13). Endothelial AMPK1 knockdown potentiated experimental BPD (14). Autophagic activity was also detected in alveolar epithelial type 2 cells in hyperoxic experimental BPD (5). AICAR and AMPK have been shown to have effects on other models of lung injury through neutrophilic activity regulation (15) or expression of genes involved in cellular senescence (16). Moreover, in a recent study, AICAR and AMPK were associated with increased heme oxygenase-1 (HO-1) in the lung and decreased lung injury (17). HO-1 breaks down free heme, released into the circulation after the turnover of red blood cells, to carbon monoxide (CO) and biliverdin. The cytoprotective effects of HO-1 are linked to autophagy (18). CO activates autophagy in lung and other tissues (19). CO also negatively regulates NLRP3 inflammasome activation in macrophages (20). All of the aforementioned mechanisms could come into interplay in the favorable phenotype of the neonatal mice exposed to hyperoxia treated with AICAR (4). Therefore, a major limitation of Soni and colleagues’ study is that it does not examine the role of AMPK activation on other AMPK-related pathways or the contribution of AMPK-driven enhanced autophagy in cell types other than macrophages in the pathogenesis of neonatal hyperoxia–induced lung injury. Along the same lines, the authors did not investigate what happens to the overall phenotype when the effects of AICAR on macrophage differentiation are abrogated in Becn1^F/F;LysM-cre mice.

Soni and colleagues, expanding on paradigms from other disease models, highlight a role for autophagy-dependent AMPK macrophage polarization in hyperoxia-induced perturbations to lung alveolarization in a rodent model of BPD. Detailed steps of this pathway, the relative contribution of macrophage polarization to lung development and repair, and the overall favorable effects of AMPK activation on lung injury require further elucidation but show promising therapeutic potential.