“KEAP”ing Alveolar Macrophage Mitochondria Content in Chronic Obstructive Pulmonary Disease

Chronic obstructive pulmonary disease (COPD) is associated with persistent macrophage infiltration in the lung parenchyma and alveolar space even after smoking cessation (1), which correlates with disease severity and areas of lung destruction (2). Macrophages isolated from BAL in COPD are derived from both tissue-resident macrophages with the capacity for self-renewal and circulating peripheral blood monocytes that migrate to the lung under a variety of stimuli (3, 4). Despite the increased abundance of airway and alveolar macrophages (hereafter termed alveolar macrophages [AMs]) in COPD, recurrent respiratory infections are abundant, accelerating disease progression and mortality (5). This may be attributed to the inability of these cells to carry out efficient phagocytosis and efferocytosis (6, 7), which may be closely linked to mitochondrial dysfunction in these cells (8–10).

Cellular energy metabolism governs macrophage adaptation and responses to a variety of environmental signals (11). At steady state, macrophages can generate ATP via several metabolic pathways, including aerobic oxidative phosphorylation (OXPHOS), glycolysis, and fatty acid oxidation. Macrophages display robust metabolic plasticity to reprogram ATP production from the most efficient metabolic pathways available. For example, stimulated bone marrow–derived macrophages predominantly use glycolysis (11) to rapidly generate ATP for proinflammatory effector functions, whereas AMs that inhabit a low-glucose, high-lipid environment in the healthy lung rely on OXPHOS to meet energy demands (12, 13). AMs from individuals with COPD display a defect in mitochondrial respiration (10, 13), with lower compensatory glycolysis and alterations in metabolite abundance (4, 12). Yet how this metabolic rewiring relates to the functional capacity of AMs and whether this phenomenon can be therapeutically reversed with beneficial effects had yet to be defined.

In this issue of the Journal, Ryan and colleagues (pp. 998–1011) build on their own earlier findings (14) to show that both alveolar and circulating peripheral blood monocyte-derived macrophages (MDMs) from individuals with COPD have impaired phagocytosis and efferocytosis and show for the first time that defective efferocytosis correlates with clinical features of COPD (15). Both phagocytosis and efferocytosis, two fundamental functions of macrophages, are processes with high energy demand that rely on the generation of ATP (16). In this study, the authors find that both AMs and MDMs are metabolically impaired in COPD. Specifically, the authors first show using total RNA sequencing of AMs isolated from BAL that metabolic processes are among the most significantly suppressed pathways in COPD AMs. Using liquid chromatography–mass spectroscopy and Seahorse (Agilent) analysis, the authors extensively characterize this metabolic defect to show that COPD AMs and MDMs have lower ATP concentrations, lower OXPHOS, lower spare respiratory capacity (mitochondrial reserve to generate ATP via OXPHOS), and lower glycolytic reserve (the maximum capacity of glycolysis to generate ATP). Next, the authors show that COPD AMs display impaired metabolic plasticity in response to stress or incubation with apoptotic neutrophils. Although baseline glycolysis is low in COPD AMs compared with AMs from healthy control subjects, the comparable ratio of how much energy the cell generates through OXPHOS compared with glycolysis (the ratio of oxygen consumption rate to extracellular acidification rate) was lower. Consistent with these findings, the authors show that the abundance of glycolytic metabolites is higher in COPD AMs compared with those from control subjects in the resting state. This may imply that COPD AMs may have more available glucose intermediates to harness glycolysis in the absence of OXPHOS to generate the ATP required for efferocytosis. Consistently, oligomycin, an inhibitor of OXPHOS, suppressed efferocytosis in MDMs, whereas inhibiting glycolysis suppressed efferocytosis only when OXPHOS was blocked. These results show for the first time that in AMs, efferocytosis is dependent on efficient mitochondrial respiration, and in conditions in which mitochondrial respiration is impaired, glycolysis fills the energy gap, consistent with findings in other phagocytes (16, 17).

Next, to interrogate why AMs in COPD need to maximize glycolysis, the authors revisit their RNA sequencing data and find that a gene called ME-1 (malic enzyme transcript 1) is one of the most significantly downregulated metabolic genes in COPD AMs compared with those from healthy control subjects. ME-1, a cytosolic, NADP-dependent enzyme, facilitates the production of pyruvate for use in mitochondria in the tricarboxylic acid (TCA) cycle. The authors show that ME-1 protein expression is lower in COPD AMs, and a loss of ME-1 in the human leukemia monocytic cell line THP-1 via CRISPR resulted in a similar metabolic phenotype to COPD AMs, namely, lower ratios of oxygen consumption rate to extracellular acidification rate, reduction in TCA intermediates, and increased glycolytic flux. Loss of ME-1 in THP-1 cells was also associated with reduced efferocytosis. ME-1 is a target gene of NRF2 (nuclear factor erythroid 2–related factor 2), and the authors have used NRF2 agonists in the past to restore macrophage phagocytosis in COPD (14). In this study, activation of NRF2 via the KEAP-1 (Kelch-like ECH-associated protein 1) inhibitor KI-696 significantly restored ME-1 expression, increased the abundance of TCA cycle intermediaries, and partially restored cellular energetics in COPD AMs. Finally, KI-696 functionally rescued COPD AM and MDM efferocytosis. Importantly, KI-696 could not completely rescue efferocytosis and was unable to restore efferocytosis in ME-1–deficient cells, implying that KI-696 efficacy is mediated through ME-1 but may not be solely driving the observed metabolic defects.

This is the first study to directly link a metabolic defect in COPD macrophages to functional impairment, in this case, impaired efferocytosis, and there are several intriguing observations. The first is that despite the severe loss in mitochondrial respiration in COPD AMs, mitochondrial DNA and protein abundance were similar between COPD AMs and those from control subjects. This suggests that the physical pool of mitochondria may not differ between healthy and COPD AMs, in contrast to other studies showing decreased mitochondrial DNA concentrations in alveolar cells and muscles of patients with severe COPD (18, 19). The authors also show that lactate concentrations are higher in the BAL fluid of patients with COPD, confirming an altered metabolic extracellular environment in COPD BAL (20, 21), which may in turn alter AM immunometabolism. Although AM efferocytosis appears to depend on mitochondrial OXPHOS, the process of engulfing apoptotic cells itself has been shown to increase glycolysis and suppress OXPHOS in macrophages, which requires further clarification (16, 22, 23).

Some outstanding questions remain. ME-1 also regulates fatty acid biosynthesis, and COPD AMs are lipid rich with lipid activation signatures (4), suggesting that fatty acid metabolism may play a yet undefined role. Given that COPD AMs are a heterogeneous population, it will be interesting to delineate using emerging single-cell metabolomic approaches if resident or MDM-derived AMs in the COPD lung have distinct immunometabolic features. In summary, this study shows that restoring AM immunometabolism using the NRF2 agonist KI-696 restores efferocytosis function in COPD AMs warranting future further extensive examination.