Metabolic characterization and RNA profiling reveal glycolytic dependence of profibrotic phenotype of alveolar macrophages in lung fibrosis

Abstract

Metabolic reprogramming has been intrinsically linked to macrophage activation. Alveolar macrophages are known to play an important role in the pathogenesis of pulmonary fibrosis. However, systematic characterization of expression profile in these cells is still lacking. Furthermore, main metabolic programs and their regulation of cellular phenotype are completely unknown. In this study, we comprehensively analyzed the expression profile and main metabolic programs in alveolar macrophages from mice with or without experimental pulmonary fibrosis. We found that alveolar macrophages from both bleomycin and active TGF-β1-induced fibrotic mouse lungs demonstrated a primarily profibrotic M2-like profile that was distinct from the well-defined M1 or any of the M2 subtypes. More importantly, we found that fibrotic lung alveolar macrophages assumed augmented glycolysis, which was likely attributed to enhanced expression of multiple key glycolytic mediators. We also found that fatty acid oxidation was upregulated in these cells. However, the profibrotic M2-like profile of fibrotic lung alveolar macrophages was not dependent on fatty acid oxidation and synthesis or lipolysis, but instead on glycolysis, in contrast to the typical IL-4-induced macrophages M(IL-4). Additionally, glutaminolysis, a key metabolic program that has been implicated in numerous pathologies, was not required for the profibrotic M2-like phenotype of these macrophages. In summary, our study identifies a unique expression and metabolic profile in alveolar macrophages from fibrotic lungs and suggests glycolytic inhibition as an effective antifibrotic strategy in treating lung fibrosis.

macrophages play a central role in inflammation, host defense, tissue remodeling, and tumor progression (10, 27, 39). They demonstrate a great deal of phenotypic plasticity when exposed to diverse stimuli in various pathological conditions. To better understand macrophage pathobiology, macrophages are generally divided into two types based on phenotypic polarity and activator categorization, namely M1 and M2, a definition emulating the Th1/Th2 classification of T cells (10, 16, 21, 27, 39, 43).

Macrophages activated by IFN-γ plus lipopolysaccharide (LPS) are M1 macrophages (also referred to as classically activated macrophages) (16, 27, 30), whereas various alternatively activated macrophages are defined as M2 macrophages (16, 27, 30). Functionally diversified M2 macrophages have been further divided into three subtypes, namely M2a, M2b, and M2c (22). Typical M2 macrophages are those activated by the Th2 cytokines IL-4 and IL-13 and are known as M2a (23, 40). Macrophages activated by exposure to immune complex plus Toll-like receptor (TLR) agonists are called M2b, whereas cells activated by IL-10, TGF-β, or glucocorticoids are termed M2c macrophages (1, 22, 23). To better represent the diverse stimuli used to establish polarized macrophage phenotype and improve reproducibility of experimental methods and results, a set of new nomenclatures that are named by stimuli has been proposed in the field to replace the old ones, such as M(LPS + IFN-γ), M(IL-4), M(IL-10), and M(TGF-β1) (29).

Alveolar macrophages are critical elements of the pulmonary defensive immune system. They are also widely recognized to have a major role in the pathogenesis of lung fibrosis (9, 36). Although it is generally believed that alveolar macrophages in fibrotic lungs are alternatively activated M2 cells (4, 15, 37, 42), it has not been well characterized what expression profile these cells assume. It is unclear how this profile is developed and what the difference is between these cells and alveolar macrophages activated in vitro by defined growth factors and molecules.

Metabolic reprogramming has been intrinsically linked to macrophage activation (7, 25, 32). Early evidence came from O'Neill and coworker’s in studies demonstrating that glycolysis was required for LPS induction of proinflammatory cytokine IL-1β in macrophages (33, 41). Another relevant study also found that glycolysis was required for TLR-mediated activation of dendritic cells (11). Fatty acid oxidation (FAO) and intracellular lipolysis were shown to be essential to differentiation of M(IL-4) (11). Despite this progress, the status of major metabolic programs in alveolar macrophages of fibrotic lungs and their roles in regulating the phenotype associated with fibrotic pathology in these cells are unknown.

In this study, we extensively investigated the expression profile in alveolar macrophages from fibrotic mouse lungs. We delineated the similarity and disparity in the expression profile of profibrotic and M2 markers between these cells and various in vitro established alveolar macrophage subtypes [M(IL-4), M(IL-10), and M(TGF-β1)]. We interrogated main metabolic programs in fibrotic lung alveolar macrophages and identified glycolysis, but not FAO or glutaminolysis, as being essential to the profibrotic M2-like phenotype in these cells. Our study suggests that glycolytic inhibition may be an effective approach to treat lung fibrosis.

MATERIALS AND METHODS

Reagents.

5-(Tetradecyloxy)-2-furoic acid and tetrahydrolipstatin (orlistat) were from Cayman Chemical. Medical-grade bleomycin that is free of LPS contamination was from Besse Medical. CB-839 was from Calbiochem. Palmitate-BSA was from Sigma-Aldrich. Oligomycin, carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP), rotenone, antimycin, etomoxir (ETO), 2-deoxy-d-glucose (2-DG), and LPS were from Sigma-Aldrich. 3-(3-Pyridinyl)-1-(4-pyridinyl)-2-propen-1-one (3PO) was from Axon Medchem. Mouse recombinant TGF-β1, IL-4, and IL-10 were from Peprotech. Mouse IFN-γ was from R&D Systems.

Experimental pulmonary fibrosis model.

C57BL/6 mice were purchased from The Jackson Laboratory. To establish bleomycin-induced lung fibrosis, 8- to 10-wk-old mice were anesthetized with isoflurane. Mouse tongue was gently pulled out with forceps, and bleomycin (1.5 U/kg body wt dissolved in 50 μl saline) was dropped in the oropharyngeal cavity. The tongue was held outside the mouth by forceps to allow all liquids to be inhaled in the lung. To establish TGF-β1-induced lung fibrosis, adenovirus expressing activated murine TGF-β1 (~6 × 10⁶ plaque-forming units/g body wt) or the control adenovirus (46) was delivered to mouse lungs by oropharyngeal instillation as described above for bleomycin delivery. The animal protocol was approved by the University of Alabama at Birmingham (UAB) Institutional Animal Care and Use Committee.

Isolation of alveolar macrophages.

After death, mouse tracheas were exposed, and a “V” excision was made. Bronchoalveolar lavage (BAL) was performed with 1 ml HBSS containing 5 mM EDTA for three times. BAL cells were collected by centrifuge. After red blood cell lysis, BAL cells were resuspended with RPMI media containing 10% FBS and seeded in 48-well plates for 30 min. The cells were then thoroughly washed with PBS to remove all unattached cells, and attached macrophages were used for various assays and treatments or harvested in TRIzol reagent for RNA purification. We have confirmed in our previous study that this method produces 100% macrophages (2).

RNA sequencing assay.

RNA sequencing (RNA-seq) and Ingenuity Pathways Analysis were performed by the UAB Heflin Center for Genomic Science. RNA-seq data were deposited to Gene Expression Omnibus and are unrestrictedly accessible with accession number GSE98468.

Real-time PCR.

mRNA levels of various mouse genes were determined by real-time PCR using the SYBR Green Master Mix Kit (Roche). Primer sequences were listed in Table 1.

Table 1.

List of sequences of real-time PCR primers

Genes	Sense	Antisense
Mouse tubulin α1	GGATGCTGCCAATAACTATGCTCGT	GCCAAAGCTGTGGAAAACCAAGAAG
Mouse arginase-1	TGACTGAAGTAGACAAGCTGGGGAT	CGACATCAAAGCTCAGGTGAATCGG
Mouse YM-1	ATGAAGCATTGAATGGTCTGAAAG	TGAATATCTGACGGTTCTGAGGAG
Mouse FIZZ-1	AGGTCAAGGAACTTCTTGCCAATCC	AAGCACACCCAGTAGCAGTCATCCC
Mouse PD-L1	GGAGATCACAGCCAGGGCAAAAC	ACCGTGGACACTACAATGAGGAACAAC
Mouse MSR1	TGTCAGAGTCCGTGAATCTACAGCAAA	CAGTGTCTGTGAGTGTTCCCAGTCCTT
Mouse PD-L2	GGCAGTACCGTTGCCTGGTCAT	GGGGTCCTGATGTGGCTGGTGT
Mouse MMP-8	GGTTACCCCAAAAGCATACCAAGC	CTCTGTGACTGACAAAATTAAATGCAAAA
Mouse MMP-12	CACTTCCCAGGAATCAAGCCTAAAAT	AAAACCAGCAAGCACCCTTCACTACA
Mouse TIMP-2	GCAACCCCATCAAGAGGATTCAGT	CTTCTGGGTGATGCTAAGCGTGTC
Mouse OPN	GCCGAGGTGATAGCTTGGCTTATG	CTCTCCTGGCTCTCTTTGGAATGC
Mouse PDGF-α	CTGTTGTAACACCAGCAGCGTCAAGT	CATTGGCAATGAAGCACCATACATAG
Mouse GluT1	TGCTGCTCAGTGTCATCTTCATCCC	CATCATCTGCCGACCCTCTTCTTTC
Mouse HK2	CTTACCGTCTGGCTGACCAACAC	CTCCATTTCCACCTTCATCCTTCT
Mouse PFK-L	GGCTGATTGGCTATTCATTCCTGA	CCCTCTGCGATGATGATGATGTTC
Mouse LDHB	CTGCTCGATTCCGCTACCTCATG	TGCACCTCCTTCCAGTTCTCACT
Mouse IL-6	CCCAATTTCCAATGCTCTCCTA	AGGAATGTCCACAAACTGATATGCT
Mouse IL-12 p40	CCAAATTACTCCGGACGGTTCAC	CAGACAGAGACGCCATTCCACAT
Mouse CCR-7	AAACGTGCTGGTGGTGGCTCTC	ACCGTGGTATTCTCGCCGATGT
Mouse Gls1	CAGCGGGATTATGACTCCAGAACA	TGGGGTGCAGACCATTATAGCAAC
Mouse GluD1	CTTCAAGATGGTGGAGGGCTTCTT	CACTCACGTCAGTGCTGTAACGGA

YM-1, chitinase 3-like 3; FIZZ-1, resistin-like molecule-α; PD-L1, programmed cell death 1 ligand 1; MSR1, macrophage scavenger receptor 1; PD-L2, programmed cell death 1 ligand 2; MMP-8, matrix metalloproteinase 8; MMP-12, matrix metalloproteinase 12; TIMP-2, tissue inhibitor of metalloproteinase 2; OPN, osteopontin; PDGF-α, platelet-derived growth factor-α; GLUT1, glucose transporter-1; HK2, hexokinase 2; PFK-L, phosphofructokinase, liver type; LDHB, lactate dehydrogenase B; CCR-7, C-C chemokine receptor type 7; Gls1, glutaminase 1; GluD1, glutamate dehydrogenase 1.

Western blotting.

Western blotting was performed as previously described (2). Anti-hexokinase 2 (HK2) antibody was from Cell Signaling. Anti-β-actin antibody was from Sigma-Aldrich. Anti-6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3 (PFKFB3) antibody was from Abcam.

Real-time cell metabolism assay.

XF-24 Extracellular Flux Analyzer (Seahorse Bioscience) was used for real-time recording of extracellular acidification rate (ECAR) and oxygen consumption rate (OCR). Alveolar macrophages were seeded in Seahorse XF-24 microplates. Before analysis, the cells were incubated in OCR or ECAR media for 1 h at 37°C in room air without CO₂ supplementation. Cells were sequentially treated with 1.5 μg/ml oligomycin, 4.5 μM FCCP, 1 μM rotenone, and 0.5 μM antimycin. Real-time ECAR and OCR were recorded according to the manufacturer's manual. For FAO assay, the cells were incubated in FAO assay media for 1 h at 37°C in room air without CO₂ supplementation before analysis. Macrophages were sequentially treated with palmitate-BSA, 1.5 μg/ml oligomycin, 4.5 μM FCCP, and 200 µM ETO. Real-time OCR was recorded.

Glucose consumption assay.

Cell supernatants were diluted 1:200. Glucose levels in the supernatants were determined using the Glucose Colorimetric/Fluorometric Assay Kit (Biovision) according to the manufacturer's instructions.

Statistical analysis.

One-way ANOVA followed by the Bonferroni test was used for multiple group comparisons. The Student's t-test was used for comparison between two groups. P < 0.05 was considered statistically significant.

RESULTS

RNA-seq and Ingenuity pathways analysis show that alveolar macrophages from fibrotic mouse lungs present both M1 and M2 expression profiles.

To comprehensively delineate phenotypic alterations of alveolar macrophages in fibrotic lungs, we performed RNA-seq assay on these cells from mice that were treated with saline or bleomycin [raw data accessible with accession no. GSE98468 and differentially expressed genes listed in Supplemental Table 1 (Supplemental data for this article may be found on the journal website.)]. Ingenuity pathways analysis showed that the enriched top canonical pathways were all those involved in tissue fibrosis and Th1/2 activation (Fig. 1). These data are consistent in general with the overwhelming evidence in literature that macrophages play crucial roles in the pathogenesis of organ fibrosis (45). Furthermore, these results suggest that alveolar macrophages in fibrotic lungs present a mixed expression profile that shares those of both M1 and M2, a phenotypic dichotomy that has long been used to describe opposite functional polarity of macrophage activation (10, 16, 21, 27, 31, 39, 43).

Fig. 1.

RNA sequencing (RNA-seq) and Ingenuity Pathways Analysis show that alveolar macrophages from fibrotic mouse lungs present both M1 and M2 expression profile. C57BL/6 mice (10 wk old) were instilled with saline or bleomycin (BLM) it. Later (2 wk), mice were killed, and bronchoalveolar lavage (BAL) was harvested. After red blood cell lysis, BAL cells were resuspended with RPMI media containing 10% FBS and seeded in 48-well plates for 30 min. The plates were then thoroughly washed with PBS to remove all unattached cells, and attached macrophages were harvested in TRIzol and total RNA purified. RNA-seq and Ingenuity Pathways Analysis were performed. The top canonical pathways identified by comparing the expression profile between the two groups were presented. iCOS, inducible costimulator; iCOSL, inducible costimulator ligand.

Alveolar macrophages from fibrotic mouse lungs acquire a distinct expression profile that crosses multiple M2 subtypes [M(IL-4), M(IL-10), and M(TGF-β1)].

Given the revelation by the RNA-seq and pathway analyses that alveolar macrophages in bleomycin-induced fibrotic lungs possessed both the M1 and M2 expression profile, we next determined if this phenotype resembled that of polarized macrophages generated in vitro by well-defined cytokines and/or microbial molecules. We induced alveolar macrophages into M(IL-4), M(IL-10), and M(TGF-β1); we activated alveolar macrophages with classical M1 stimuli LPS plus IFN-γ [M(LPS + IFN-γ)]. We examined a set of widely accepted mouse M2 macrophage markers, arginase 1 (Arg1), resistin-like molecule-α (FIZZ-1), chitinase 3-like 3 (YM-1), macrophage scavenger receptor 1 (MSR1), and programmed cell death 1 ligand 2 (PD-L2) in in vitro activated alveolar macrophages and those from mice treated with saline or bleomycin (10, 16, 19, 21, 27, 31, 39, 43). We found that IL-4, but not IL-10 or TGF-β1, markedly upregulated Arg1, FIZZ-1, YM-1, and PD-L2 in alveolar macrophages (Fig. 2A). We also found that IL-10, but neither IL-4 nor TGF-β1, induced MSR1 (Fig. 2A). These data suggest that, despite being termed as M2 macrophages in general, these subtypes express distinct sets of M2 markers. Nevertheless, alveolar macrophages from fibrotic mouse lungs demonstrated remarkably increased Arg1, MSR1, and PD-L2 but not other M2 markers like FIZZ-1 or YM-1 (Fig. 2B and data not shown). Together, these data suggest that alveolar macrophages from fibrotic mouse lungs do not belong to any of the well-characterized M2 subtypes but rather become a unique class possessing an array of phenotypic markers that are shared by more than one subtype.

Fig. 2.

Alveolar macrophages from fibrotic mouse lungs acquire a distinct expression profile that crosses multiple M2 subtypes. A, D, and F: alveolar macrophages were prepared as in Fig. 1. The cells were treated with media (con), 5 ng/ml mouse IL-4, 5 ng/ml mouse IL-10, 10 ng/ml mouse TGF-β1 (A and D), or 10 ng/ml lipopolysaccharide (LPS) plus 5 ng/ml mouse IFN-γ (F) for 24 h. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. B, C, and E: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. A, D, and F: n = 3–4, mean ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001 compared with the “con” group. B, C, and D: n = 3–4, mean ± SE, *P < 0.05, **P < 0.01, and ***P < 0.001 compared with the “saline” group. Experiments were repeated two times with representative results presented.

To further characterize the profibrotic activity of alveolar macrophages from mice with bleomycin-induced lung fibrosis uncovered by the RNA-seq and pathway analyses, we validated by real-time PCR the markedly increased levels of a set of five well-established profibrotic mediators, including osteopontin (20, 44), platelet-derived growth factor-α (PDGF-α) (17, 28), matrix metallopeptidase 8 (MMP-8) (4), MMP-12 (24), and tissue inhibitor of metalloproteinase 2 (TIMP-2) (6), in alveolar macrophages from fibrotic mouse lungs (Fig. 2C). However, neither IL-4, IL-10, nor TGF-β1 alone induced all of these molecules in alveolar macrophages in vitro (Fig. 2D). Additionally, TIMP-2 or PDGF-α expression was not affected by any of these pro-M2 growth factors (Fig. 2D and data not shown). Altogether, these data suggest that the profibrotic activity of alveolar macrophages from fibrotic mouse lungs may a result of collective actions of various pro-M2 mediators, including IL-4, IL-10, TGF-β1, and others.

We also compared the M1 expression profile of alveolar macrophages from fibrotic mouse lungs and M(LPS + IFN-γ). We found that, although alveolar macrophages from fibrotic lungs also showed elevated expression of some classical M1 markers, such as IL-6, IL-12, and PD-L1 (Fig. 2E), they were often at levels that were barely above detection by real-time PCR, in contrast to those in M(LPS + IFN-γ) (Fig. 2F). These data suggest that, although the RNA-seq and pathway analyses indicate a M1 phenotype as well for fibrotic lung alveolar macrophages, it is a rather weak profile.

In all, based on the similarity, dissimilarity, and robustness of phenotypic marker expression between alveolar macrophages in fibrotic mouse lungs and those activated in vitro, we believe that it is more precise to define these pathogenic cells as profibrotic M2-like macrophages.

Alveolar macrophages from TGF-β1- and bleomycin-induced fibrotic lungs share a similar profibrotic M2-like profile.

To determine if the profibrotic M2-like profile present in alveolar macrophages from bleomycin-treated lungs was developed specifically in the setting of fibrotic pathogenesis, but not a response only to bleomycin stimulation itself, we examined the same set of M2 phenotypic markers and profibrotic mediators in alveolar macrophages from mice that received intratracheally administered adenovirus expressing constitutively active TGF-β1, another lung fibrosis model widely used by us and others (18, 46). We found that alveolar macrophages from TGF-β1-induced fibrotic mouse lungs demonstrated an expression pattern of these genes that was similar to that in the cells from the bleomycin lung fibrosis model (Fig. 3, A and C). These data confirmed the intrinsic correlation between the profibrotic M2-like profile of alveolar macrophages and lung fibrotic pathology. Of note, although alveolar macrophages from TGF-β1-induced fibrotic mouse lungs expressed proinflammatory M1 markers IL-12, PD-L1, and C-C chemokine receptor type 7, they did not demonstrate IL-6 upregulation like those from bleomycin-treated mouse lungs (Fig. 3B and data not shown).

Fig. 3.

Alveolar macrophages from TGF-β1- and bleomycin-induced fibrotic lungs share a similar profibrotic M2-like profile. A–C: 10-wk-old C57BL/6 control mice were it instilled control or adenovirus (Ad-con) that expressed active TGF-β1 (Ad-TGF-β1). Later (2 wk), mice were killed, and BAL were prepared. Alveolar macrophages were harvested as in Fig. 1. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR; n = 5, mean ± SE. *P < 0.05, **P < 0.01, and ***P < 0.001 compared with the “Ad-con” group.

Alveolar macrophages from fibrotic lungs assume augmented glycolysis and fatty acid oxidation.

There has been accumulating evidence that macrophage activation is intrinsically associated with metabolic reprogramming (7, 25, 32). Recent studies demonstrated that both glucose and fatty acid metabolism played critical roles in regulating macrophage polarization (7, 25, 32). To characterize the main metabolic programs in alveolar macrophages from fibrotic lungs, we used a Seahorse XF24 Extracellular Flux Analyzer to measure ECAR as a level of lactate production (a surrogate for glycolytic rate) (13) and OCR, a representation of mitochondrial respiratory activity. As shown in Fig. 4, A and B, alveolar macrophages from mouse lungs treated with bleomycin for 2 or 4 wk demonstrated remarkably augmented maximal glycolysis compared with those from saline-treated animals. Additionally, mitochondrial oxidative phosphorylation, as reflected by OCR, including that at the basal and maximal levels, was also markedly increased in these cells (Fig. 4, C and D). These data and other findings of increased ATP production-associated OCR and unchanged proton leaks (Fig. 4, C and D) suggest that alveolar macrophages from fibrotic mouse lungs are more robust in mitochondrial activity. Consistent with the elevated glycolysis in alveolar macrophages from mouse lungs treated with bleomycin, these cells also presented enhanced glucose consumption (Fig. 4E). We also found that, similar to glycolysis, FAO was upregulated in these pathological macrophages compared with that in normal cells (Fig. 4F). This finding suggests that these two metabolic programs may be critical to the profibrotic M2-like profile in fibrotic lung alveolar macrophages.

Fig. 4.

Alveolar macrophages from fibrotic lungs assume augmented glycolysis and fatty acid oxidation. A and B: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 and 4 wk and seeded in Seahorse XF-24 cell culture microplates. The cells were treated sequentially with oligomycin (Oligo), carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP), and rotenone (Rot) plus antimycin (AA). A: real-time extracellular acidification rate (ECAR) was recorded. B: basal and maximal ECAR were plotted. n = 3 mice for saline and 4 mice for 2 and 4 wk postbleomycin. *P < 0.05 and **P < 0.01 compared with the saline group. C: oxygen consumption rate (OCR) of the experiment in A was recorded. D: basal, maximal, proton-leak associated, ATP production-associated, and reserved OCR were plotted. *P < 0.05, **P < 0.01, and ***P < 0.001 compared with the saline group. E: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. The cells were seeded in a 96-well plate and cultured for 24 h. Levels of supernatant glucose were determined. Glucose consumptions were the difference between glucose levels in the conditioned media and that in the original media. n = 5 mice for saline and bleomycin. Mean ± SE. *P < 0.05 compared with the saline-treated group. F: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. The cells were seeded in Seahorse XF-24 cell culture microplates. The cells were treated by sequential treatments with palmitate-BSA, Oligo, FCCP and etomoxir (ETO). Real-time OCR was recorded. n = 5 mice for saline and bleomycin.

Key glycolytic enzymes are upregulated in alveolar macrophages from fibrotic lungs.

We have shown that alveolar macrophages from fibrotic lungs adopt augmented glycolysis. To delineate the underlying mechanism responsible for this enhanced program, we determined levels of key glycolytic enzymes and cell surface glucose transporter and found that there was significantly increased expression of HK2, phosphofructokinase, liver type (PFK-L), PFKFB3, lactate dehydrogenase B (LDHB) and glucose transporter 1 (GluT1) in alveolar macrophages from bleomycin-induced fibrotic lungs (Figs. 5A–5B). Consistently, the expression of these glycolysis-associated molecules was also upregulated in cells from mice with TGF-β1-induced lung fibrosis (Fig. 5C). These findings suggest that the augmented glycolysis in alveolar macrophages from fibrotic lungs is caused by elevated glycolytic mediators in these cells.

Fig. 5.

Key glycolytic enzymes are upregulated in alveolar macrophages from fibrotic lungs. A and B: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. n = 4, Mean ± SE. *P < 0.05 compared with the saline group. Experiments were repeated two times with representative results presented. B: Western blotting was performed with the alveolar macrophages from mice treated with saline or bleomycin for 2 wk. Densitometric analysis of protein band intensities was performed, and fold change was calculated. Mean ± SD. ***P < 0.001 compared with the saline group. C: alveolar macrophages harvested from Ad-con and Ad-TGF-β1 mice. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. n = 5, Mean ± SE. *P < 0.05, **P < 0.01, and ***P < 0.001 compared with the Ad-con group.

Glycolysis is required for the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs.

We next investigated if augmented glycolysis was required for the presentation of profibrotic M2-like profile in fibrotic lung alveolar macrophages. As shown in Fig. 6, A and B, we found that both one and two days' treatment with 2-DG, a potent inhibitor of the key glycolytic enzyme HK2, effectively reversed the profibrotic M2-like profile. We also treated the cells with 3PO, a specific inhibitor of another critical glycolytic enzyme, PFKFB3 (38, 48), and found that this compound attenuated the expression of the profibrotic and M2 markers Arg1, MMP-8, MMP-12, and osteopontin (Fig. 6C). Taken together, these data suggest that glycolysis is essential to the maintenance of the profibrotic M2 profile in alveolar macrophages from fibrotic lungs.

Fig. 6.

Glycolysis is required for the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs. A and B: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. The cells were then treated with 1 mM 2-deoxy-d-glucose (2-DG) for 24 (A) or 48 (B) h. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. n = 3–4, Mean ± SD. **P < 0.01 and ***P < 0.001 compared with the “−BLM-2-DG” group. ^#P < 0.05, ^##P < 0.01, and ^###P < 0.001 compared with the ““+BLM-2-DG” group. Experiments were repeated two times with representative results presented. C: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. The cells were then treated with 30 μM 3-(3-pyridinyl)-1-(4-pyridinyl)-2-propen-1-one (3PO) for 24 h. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. n = 3, Mean ± SD. *P < 0.05 compared with the “+BLM-3PO” group.

Fatty acid oxidation and synthesis are dispensable for the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs.

We also examined the role of mitochondrial FAO in preservation of the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs. We treated the cells with ETO, a commonly used inhibitor of carnitine palmitoyltransferase-1A that is a key enzyme mediating fatty acid entry in mitochondria (35). As shown in Fig. 7A, we found that ETO had hardly any effect on the profibrotic M2-like profile in fibrotic lung alveolar macrophages. Also given that previous studies found that intracellular lipolysis-fueled FAO was necessary for IL-4 induction of M2a macrophages (11), we determined if this pathway was critical to the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs. To test this hypothesis, we treated the cells with orlistat, a clinically used lipase inhibitor (11), and found, consistent with the failed effect of FAO inhibition on presentation of the profibrotic M2-like profile, blocking lipolysis had no impact on the phenotype of alveolar macrophages from fibrotic lungs (Fig. 7B). Additionally, fatty acid synthesis did not show any regulation of the profibrotic M2-like profile either (Fig. 7C). Altogether, these data suggest that, although fatty acid metabolism is crucial to M2a macrophage activation, it is dispensable to the acquirement of profibrotic M2-like phenotype of alveolar macrophages in lung fibrosis.

Fig. 7.

Fatty acid oxidation and synthesis are dispensable for the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs. A–C: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. The cells were then treated with 100 μM ETO (A), 40 μM orlistat (B), or 5 μM 5-(tetradecyloxy)-2-furoic acid (TOFA, C) for 48 h. #P < 0.05 compared with the “+BLM-ETO” group. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. n = 3–4, Mean ± SD. **P < 0.01 and ***P < 0.001 compared with the “−BLM” groups that were treated with control media. Experiments were repeated two times with representative results presented.

Glutaminolysis is not required for the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs.

Although there is currently no evidence that glutaminolysis regulates macrophage activation, we still investigated the role of this metabolic program in establishment of the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs. We treated fibrotic lung alveolar macrophages with CB-839, a specific inhibitor of glutaminase 1 (Gls1) (12), and found that this compound had no effect on the expression of the profibrotic and M2 markers (Fig. 8A). Consistent with this finding, expression of two primary glutaminolytic mediators, Gls1 and glutamate dehydrogenase 1 (3, 5), was unchanged in these cells (Fig. 8B). These data suggest that, although it is a key metabolic program that has been implicated in numerous pathologies (47), glutaminolysis seems irrelevant to the profibrotic M2-like phenotype of alveolar macrophages in lung fibrosis.

Fig. 8.

Glutaminolysis is not required for the profibrotic M2-like profile in alveolar macrophages from fibrotic lungs. A: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. The cells were then treated with 1 μM CB-839 for 48 h. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. n = 3–4, Mean ± SD. **P < 0.01 and ***P < 0.001 compared with the “−BLM-CB-839” group. Experiments were repeated two times with representative results presented. B: alveolar macrophages were harvested from mice treated with saline or bleomycin for 2 wk. Total RNAs were purified, and levels of indicated genes were determined by real-time PCR. n = 4, Mean ± SE.

DISCUSSION

In our previous studies, we found that pharmacological inhibition of glycolysis diminished both bleomycin- and TGF-β1-induced lung fibrosis in mice (46). Although the mitigation of lung fibrosis was then attributed to alleviated profibrotic activity of lung myofibroblasts led by glycolytic inhibition, other types of pulmonary cells were likely also involved. In the current study, we provided convincing evidence that inhibition of glycolysis in alveolar macrophages from fibrotic lungs reversed the profibrotic profile in these cells. Taken together, our studies suggest that glycolytic dysregulation in both lung myofibroblasts and alveolar macrophages participates in the pathogenesis of pulmonary fibrosis. In addition, these findings further boost our confidence that targeting aberrant glycolysis in the lung is effective in treating pulmonary fibrosis and is a novel therapeutic deserving further investigation.

We presented solid data that the augmented glycolysis in alveolar macrophages from fibrotic lungs was caused by elevation of important glycolytic mediators, such as GluT1, HK2, PFK-L, and LDHB (8). However, there was little evidence from the RNA-seq analysis that expression of mitochondrial regulators of oxidative phosphorylation was significantly altered in these cells. Therefore, the enhanced FAO found in the profibrotic alveolar macrophages may be simply the result of increased expression of fatty acid transporters, such as long-chain fatty acid transport protein 1 (data not shown). Regardless, given that blocking FAO failed to have an impact on the profibrotic M2-like profile in fibrotic lung alveolar macrophages, the role of this metabolic program in these cells in the pathogenesis of lung fibrosis remains to be elucidated. It should be noted that lack of effect of FAO on fibrotic lung alveolar macrophages is contradictory to its well-defined regulation of M2 polarization by IL-4 in vitro (11, 25). Additionally, the generally M1-controlling glycolysis (33) is shown to be essential to the profibrotic M2-like profile in fibrotic lung alveolar macrophages. Altogether, these findings highlight the complexity of phenotypic development of macrophages in in vivo pathological settings, such as lung fibrosis.

Although the presentation of M2 phenotype appeared more prominent than M1 in alveolar macrophages from fibrotic lungs, we found that glycolytic mediators GluT1, HK2, PFK-L, and LDHB were only induced by M1 activator IFN-γ plus LPS, but not by M2 inducers IL-4, IL-10, or TGF-β1 in alveolar macrophages in vitro (data not shown). This invites an intriguing question of how these glycolytic mediators are regulated during pathological fibrogenesis in the lung. Elucidation of the underlying mechanism will certainly help to identify novel therapeutic targets for treating this disease.

We found that glycolytic inhibition mitigated the increased expression of M2 phenotypic markers in alveolar macrophages from fibrotic lungs. We also found that blocking glycolysis reversed the levels of profibrotic mediators in these cells. However, one cannot conclude from these findings that there is a correlation between the M2-like and the profibrotic phenotype. It has been previously shown that ablation of the M2 phenotypic marker Arg1 leads to aggravated tissue fibrosis (34). Nevertheless, given the profile of incomplete array of M2 markers in fibrotic lung alveolar macrophages, it remains to be elucidated if the profibrotic activity of alveolar macrophage relies on the presentation of M2 phenotype in lung fibrosis.

It has been shown that alveolar macrophages from mice with bleomycin-induced lung fibrosis are remarkably heterogeneous, consisting of resident macrophages and/or macrophages differentiated from infiltrated monocytes (14, 26). Although the current study is incapable of differentiating the contribution of individual macrophage subtypes to the profibrotic expression profile and metabolic programs that we observed in these cells as a group, the revelation of the dependence of the pathogenic phenotype on glycolysis shall probably lead to new avenues for treating lung fibrosis. In addition, the limitation of the methodology in the current study to the static two-time-point analysis did not allow an opportunity to investigate these pathologies at the time preceding full-blown fibrosis or at resolution phase. Therefore, a continuous tracking of metabolic reprogramming and phenotypic migrating of single cells would be an ideal and ultimate approach.

We showed that alveolar macrophages from both bleomycin- and TGF-β1-induced fibrotic lungs demonstrated a highly similar profibrotic M2-like profile. These data suggest that such characteristic is acquired in the specific setting of pathological fibrogenesis but is not the result of direct activation by bleomycin or TGF-β1 in the lungs. This notion is further reinforced by the finding that TGF-β1-treated alveolar macrophages did not mirror the expression profile possessed by alveolar macrophages from TGF-β1-induced fibrotic lungs. Despite the identification of the close profile of profibrotic and metabolic gene expression in alveolar macrophages from the two models, we caution against an immediate translational implication before similar phenotypic alterations are validated in these cells from human patients. Therefore, the more significant take in our future study may be to delineate the main metabolic programs and their roles in regulating the profibrotic phenotype of alveolar macrophages in idiopathic pulmonary fibrosis.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants HL-135830 and HL-114470.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

N.X., H.C., J.G., S.B., S.G., and S.D. performed experiments; N.X., H.C., and G.L. analyzed data; N.X., H.C., R.-M.L., and G.L. interpreted results of experiments; N.X. and G.L. edited and revised manuscript; N.X., H.C., J.G., S.B., S.G., E.A., and G.L. approved final version of manuscript; G.L. conceived and designed research; G.L. prepared figures; G.L. drafted manuscript.