Alternative (M2)-polarized macrophages possess high capacities to produce specialized proresolving mediators (SPM; i.e., resolvins, protectins, and maresins) that play key roles in resolution of inflammation and tissue regeneration. Vacuolar (H+)-ATPase (V-ATPase) is fundamental in inflammatory cytokine trafficking and secretion and was implicated in macrophage polarization toward the M2 phenotype, but its role in SPM production and lipid mediator biosynthesis in general is elusive. In this study, we show that V-ATPase activity is required for the induction of SPM-biosynthetic pathways in human M2-like monocyte-derived macrophages (MDM) and consequently for resolution of inflammation. Blockade of V-ATPase by archazolid during IL-4–induced human M2 polarization abrogated 15-lipoxygenase-1 expression and prevented the related biosynthesis of SPM in response to pathogenic Escherichia coli, assessed by targeted liquid chromatography–tandem mass spectrometry–based metabololipidomics. In classically activated proinflammatory M1-like MDM, however, the biosynthetic machinery for lipid mediator formation was independent of V-ATPase activity. Targeting V-ATPase in M2 influenced neither IL-4–triggered JAK/STAT6 nor the mTOR complex 1 signaling but strongly suppressed the ERK-1/2 pathway. Accordingly, the ERK-1/2 pathway contributes to 15-lipoxygenase-1 expression and SPM formation in M2-like MDM. Targeting V-ATPase in vivo delayed resolution of zymosan-induced murine peritonitis accompanied by decreased SPM levels without affecting proinflammatory leukotrienes or PGs. Together, our data propose that V-ATPase regulates 15-lipoxygenase-1 expression and consequent SPM biosynthesis involving ERK-1/2 during M2 polarization, implying a crucial role for V-ATPase in the resolution of inflammation.
Macrophages are innate immune cells with marked plasticity that, depending on their phenotype, promote acute inflammation and host defense but also can support the resolution of inflammation and the return to homeostasis (1–3). Polarization of macrophages toward proinflammatory M1-like and proresolving M2-like phenotypes is governed by epigenetic and cell survival pathways, tissue microenvironment, and extrinsic factors like microbial products and secreted cytokines (3). CD4+ T cell subtype-derived cytokines determine macrophage polarization, in which the M1 phenotype is obtained by exposure to TH1-related IFN-γ, particularly in the presence of the TLR agonist LPS, whereas the M2 subtype is achieved upon exposure to TH2-derived IL-4 or IL-13 (3, 4).
One hallmark characteristic of the opposing functions of M1 and M2 is their differential lipid mediator (LM) profile that they produce, due to their subtype-specific expression of LM-biosynthetic enzymes that distinguish their proinflammatory or proresolving phenotypes (1, 5, 6). Whereas M1 are characterized by marked expression of proinflammatory 5-lipoxygenase (LOX)/5-LOX–activating protein (FLAP), and cyclooxygenase (COX) pathways with substantial leukotriene (LT)B4 and PGE2 production, the LM profile of M2 is dominated by the 15-LOX-1 pathway along with strong biosynthesis of specialized proresolving mediators (SPM) (6). These SPM encompass a novel superfamily of highly potent bioactive LM, including lipoxins (LX), resolvins (Rv), maresins (MaR), and protectins (PD), that terminate inflammation and promote tissue regeneration (7, 8). Accumulating evidence suggests that the ability of M2 to generate SPM critically determines their inflammation-resolving features (9). However, the signaling pathways involved in SPM biosynthesis during M2 polarization are incompletely understood and remain to be elucidated.
In this article, we show that the vacuolar (H+)-ATPase (V-ATPase) plays a crucial role in the induction of SPM–biosynthetic pathways during polarization of monocyte-derived macrophages (MDM) to the M2 phenotype. V-ATPases are ATP-dependent proton-translocating macromolecular complexes that acidify lysosomes, endosomes, Golgi apparatus, and certain secretory granules in eukaryotic cells and participate in cellular pH homeostasis, receptor-mediated endocytosis, virus and toxin entry, intracellular trafficking, and protein degradation and processing (10). In macrophages, V-ATPase regulates lysosomal and cytoplasmic pH homeostasis with consequences for reactive oxygen species formation, bactericidal activity, and lysosomal enzyme secretion (11–13). However, little is known about the role of V-ATPase in either LM biosynthesis or the polarization of human macrophages. V-ATPase was shown to participate at M2 polarization in mice, but SPM biosynthesis or expression of 15-LOX-1 was not addressed (14). For human polarized M1 and M2, contrasting phagosome pH regulation and maturation related to V-ATPase was demonstrated (15), and targeting of V-ATPase during polarization of human MDM elevated TNF-α release and ROS formation in M1, but not in M2 (16). Hence, V-ATPase might impact human M1/M2 polarization, but whether V-ATPase is involved in the regulation of LM-biosynthetic pathways related to inflammation and its resolution is unknown. Therefore, we here aimed to reveal the role of V-ATPase in the induction of phenotype-specific profiles of LM with proinflammatory and proresolving properties during human MDM polarization. Our results show that V-ATPase regulates 15-LOX-1 expression and consequent SPM biosynthesis in an ERK-1/2-dependent fashion during acquiring an M2 phenotype and thus imply critical roles of this signaling pathway in the resolution of inflammation.
Materials and Methods
Deuterated and nondeuterated LM standards for ultra-performance liquid chromatography–tandem mass spectrometry (UPLC–MS-MS) quantification were purchased from Cayman Chemical/Biomol (Hamburg, Germany). Archazolid A (ArchA) was isolated from Archangium gephyra as previously described (17). Bafilomycin A1 was obtained from Sigma-Aldrich (Taufkirchen, Germany), CP-690,550 and GSK-2033 were obtained from Tocris Bioscience (Bristol, U.K.), JQ-1 was from AdooQ Bioscience (Irvine, CA), LY294002 was from Cytoskeleton (Denver, CO), skepinone-L and Torin 1 were from Cayman Chemical (Ann Arbor, MI), and U0126 was from Enzo Life Sciences (Farmingdale, NY). AS1517499 and all other reagents were obtained from Sigma-Aldrich unless mentioned otherwise.
Male CD-1 mice (33–39 g, 6–8 wk; Charles River Laboratories, Calco, Italy) were housed in a controlled environment (21 ± 2°C) and provided with standard rodent chow and water. Animals were allowed to acclimate for 4 d prior to experiments and were subjected to a 12 h light/12 h dark schedule. Mice were randomly assigned for the experiments, which were conducted during the light phase. The experimental procedures were approved by the Italian Ministry according to International and National law and policies (European Union Directive 2010/63/EU and Italian DL 26/2014 for animal experiments.
Cell isolation and polarization of MDM
Leukocyte concentrates from freshly withdrawn peripheral blood of male and female healthy adult human donors (age 18–65 y) were provided by the Institute of Transfusion Medicine at the University Hospital Jena, Jena, Germany. The experimental protocol was approved by the ethical committee of the University Hospital Jena. All methods were performed in accordance with the relevant guidelines and regulations. PBMC were separated using dextran sedimentation, followed by centrifugation on lymphocyte separation medium (Histopaque-1077; Sigma-Aldrich). PBMC were seeded in RPMI 1640 (Sigma-Aldrich) containing 10% (v/v) heat-inactivated FCS, 100 U/ml penicillin, and 100 μg/ml streptomycin in cell culture flasks (Greiner Bio-One, Frickenhausen, Germany) for 1.5 h at 37°C and 5% CO2 for adherence of monocytes. For differentiation of monocytes to MDM and polarization toward M1 and M2, published criteria were used (4). M1 were generated by incubating monocytes with 20 ng/ml GM-CSF (PeproTech, Hamburg, Germany) for 6 d in RPMI 1640 supplemented with 10% FCS, 2 mmol/L glutamine (Biochrom/Merck, Berlin, Germany), and penicillin–streptomycin (Biochrom/Merck), followed by treatment with 100 ng/ml LPS and 20 ng/ml INF-γ (PeproTech). M2 were generated by incubating monocytes with 20 ng/ml M-CSF (PeproTech) for 6 d, followed by treatment with 20 ng/ml IL-4 (PeproTech). Routinely, cells were polarized for 48 h unless stated otherwise. ArchA (30 nM) and other inhibitors were added 15 min prior to addition of polarization agents (INF-γ, LPS, IL-4) unless mentioned otherwise.
Incubation of MDM for LM formation and LM metabololipidomics
MDM (2 × 106/ml) were incubated in PBS containing 1 mM CaCl2. To evoke LM biosynthesis pathogenic Escherichia coli (serotype O6:K2:H1, ratio 1:50) was added for 180 min at 37°C. The supernatants were then transferred to 2 ml of ice-cold methanol containing 10 μl of deuterium-labeled internal standards (200 nM d8-5S-HETE, d4-LTB4, d5-LXA4, d5-RvD2, d4-PGE2 and 10 μM d8–arachidonic acid [AA]) to facilitate quantification and sample recovery. Sample preparation was conducted by adapting published criteria (6). In brief, samples were kept at −20°C for 60 min to allow protein precipitation. After centrifugation (1200 × g, 4°C, 10 min), 8 ml acidified H2O was added (final pH = 3.5), and samples were subjected to solid phase extraction. Solid phase cartridges (Sep-Pak Vac 6cc 500 mg/6 ml C18; Waters, Milford, MA) were equilibrated with 6 ml methanol and 2 ml H2O before samples were loaded onto columns. After washing with 6 ml H2O and additional 6 ml n-hexane, LM were eluted with 6 ml methyl formate. Finally, the samples were brought to dryness using an evaporation system (TurboVap LV; Biotage, Uppsala, Sweden) and resuspended in 100 μl methanol–water (50/50, v/v) for UPLC–MS-MS automated injections. LM profiling was analyzed with an Acquity UPLC system (Waters) and a QTRAP 5500 Mass Spectrometer (AB Sciex, Darmstadt, Germany) equipped with a Turbo V Source and electrospray ionization. LM were eluted using an Acquity UPLC BEH C18 column (1.7 μm, 2.1 × 100 mm; Waters, Eschborn, Germany) at 50°C with a flow rate of 0.3 ml/min and a mobile phase consisting of methanol–water–acetic acid of 42:58:0.01 (v/v/v) that was ramped to 86:14:0.01 (v/v/v) over 12.5 min and then to 98:2:0.01 (v/v/v) for 3 min. The QTrap 5500 was operated in negative ionization mode using scheduled multiple reaction monitoring coupled with information-dependent acquisition. The scheduled multiple reaction monitoring window was 60 s, optimized LM parameters (collision energy, entrance potential, declustering potential, collision cell exit potential) were adopted (18), and the curtain gas pressure was set to 35 ψ. The retention time and at least six diagnostic ions for each LM were confirmed by means of an external standard (Cayman Chemical/Biomol). Quantification was achieved by calibration curves for each LM.
SDS-PAGE and Western blot
Cell lysates of MDM, corresponding to 2 × 106 cells, were separated on 8% (for cPLA2-α), 10% (for 5-LOX, 15-LOX-1, 15-LOX-2, COX-1, COX-2, STAT6, phospho-STAT6, Akt, phospho-Akt, p70S6K, phospho-p70S6K, cRaf, phospho-cRaf, MEK, phospho-MEK, ERK-1/2, phospho-ERK-1/2, p38 MAPK, phospho-p38 MAPK, c-Myc, β-actin, and GAPDH), and 16% (for FLAP and late endosomal and lysosomal adaptor and MAPK and mTOR activator 1 [Lamtor1]) polyacrylamide gels and blotted onto nitrocellulose membranes (Amersham Protran supported 0.45 μm nitrocellulose; GE Healthcare, Freiburg, Germany). The membranes were incubated with the following primary Abs: rabbit polyclonal anti-cPLA2-α, 1:1000 (2832S; Cell Signaling Technology); rabbit polyclonal anti-5-LOX, 1:1000 (to a peptide with the C-terminal 12 aa of 5-LOX: CSPDRIPNSVA; kindly provided by Dr. M. E. Newcomer, Louisiana State University, Baton Rouge, LA); mouse monoclonal anti-15-LOX-1, 1:500 (ab119774; Abcam, Cambridge, U.K.); rabbit polyclonal anti-15-LOX-2, 1:500 (ab23691; Abcam); rabbit polyclonal anti-COX-1, 1:1000 (4841S; Cell Signaling Technology); rabbit polyclonal anti-COX-2, 1:1000 (4842S; Cell Signaling Technology); mouse polyclonal anti-STAT6, 1:1000 (ab88540; Abcam); rabbit polyclonal anti–phospho-STAT6 (Tyr641), 1:1000 (ab28829; Abcam); mouse monoclonal anti-Akt, 1:1000 (2920; Cell Signaling Technology); rabbit polyclonal anti-phospho-Akt, 1:1000 (9271S; Cell Signaling Technology); rabbit monoclonal anti-p70S6K, 1:1000 (2708T; Cell Signaling Technology); mouse monoclonal anti–phospho-p70S6K (Thr389) 1:1000 (9206S; Cell Signaling Technology); rabbit polyclonal anti-MEK1/2, 1:1000 (9122; Cell Signaling Technology); rabbit polyclonal anti–phospho-MEK1/2 (Ser217/Ser221), 1:1000 (9121S; Cell Signaling Technology); rabbit monoclonal anti-ERK1/2, 1:1000 (4695S; Cell Signaling Technology); mouse monoclonal anti–phospho-ERK1/2 (Thr202/Tyr204), 1:1000 (9106; Cell Signaling Technology); rabbit monoclonal anti-p38 MAPK, 1:1000 (8690S; Cell Signaling Technology); rabbit polyclonal anti–phospho-p38 MAPK (Thr180/Tyr182), 1:1000 (9211S; Cell Signaling Technology); rabbit monoclonal anti-cMyc, 1:1000 (5605S; Cell Signaling Technology); rabbit polyclonal anti-FLAP, 1:1000 (ab85227; Abcam); rabbit monoclonal anti-Lamtor1 1:1000 (8975T; Cell Signaling Technology); mouse monoclonal anti-β-actin, 1:1000 (3700S; Cell Signaling); rabbit polyclonal anti–β-actin, 1:1000 (4967S; Cell Signaling Technology) and rabbit monoclonal anti-GAPDH, 1:1000 (5174S; Cell Signaling). Immunoreactive bands were stained with IRDye 800CW Goat anti-Mouse IgG (H+L), 1:10,000 (926-32210; LI-COR Biosciences, Lincoln, NE), IRDye 800CW Goat anti-Rabbit IgG (H+L), 1:15,000 (926 32211; LI-COR Biosciences) and/or IRDye 680LT Goat anti-Mouse IgG (H+L), 1:40,000 (926-68020; LI-COR Biosciences), and visualized by an Odyssey infrared imager (LI-COR Biosciences). Data from densitometric analysis were background corrected.
Determination of cytokine levels
Unpolarized MDM were treated with ArchA (30 nM) or vehicle (0.1% DMSO) 30 min before cells were polarized to M1 or M2 for 48 h. For measurement of extracellular cytokine levels, supernatants were collected by centrifugation (2000 × g, 4°C, 10 min). The cytokines IL-6, IL-10, and TNF-α were analyzed by in-house–made ELISA kits (R&D Systems, Bio-Techne).
Fluorescent staining for flow cytometric analysis of M1 or M2 was performed in FACS buffer (PBS with 0.5% BSA, 2 mM EDTA, and 0.1% sodium azide). MDM were treated either with vehicle or 30 nM ArchA during 48 h polarization. Nonspecific Ab binding was blocked by using mouse serum for 10 min at 4°C prior Ab staining. Subsequently, MDM were stained with fluorochrome-labeled Ab mixtures at 4°C for 20 min. The following Abs were used: FITC anti-human CD14 (2 μg/test, clone M5E2; BioLegend, San Diego, CA), PE anti-human CD54 (1 μg/test, clone HA58; eBioscience, San Diego, CA), allochycocyanin-H7 anti-human CD80 (0.25 μg/test, clone L307.4; BD Bioscience), PE-Cy7 anti-human CD163 (2 μg/test, clone RM3/1; BioLegend), PerCP-eFluor710 anti-human CD206 (0.06 μg/test, clone 19.2; eBioscience). Upon staining, MDM (M1 or M2) were analyzed using a FACS Canto Plus flow cytometer (BD Bioscience), and data analyzed using FlowJo X Software (BD Bioscience).
Determination of cell viability
The viability of MDM was assessed by MTT assay as described (16). Briefly, MDM after 6 d of differentiation of monocytes were preincubated with test compounds for 15 min at 37°C (5% CO2), polarization agents were added, and cells were incubated for 48 h. Staurosporine (1 μM), a pan-kinase inhibitor and inducer of apoptosis, was used as positive control. MTT solution was added, and cells were further incubated for 4 h and lysed in a buffer containing 10% (w/v) SDS.
MDM after 6 d of differentiation of monocytes were pretreated with ArchA (30 nM) for 30 min before polarization for 48 h to M1 and M2, and 106 cells were seeded onto glass coverslips. Lysotracker (1:100 dilution; Invitrogen, Thermo Fisher Scientific) was added to the cells for 1 h at 37°C. The fluorescence was visualized with a Zeiss Axio Observer.Z1 inverted microscope (Carl Zeiss, Jena, Germany) and an LCI Plan-Neofluar 63×/1.3 Imm Corr DIC M27 objective. Images were taken with an AxioCam MR3 camera and were acquired, cut, linearly adjusted in the overall brightness and contrast, and exported by the AxioVision 4.8 software (Carl Zeiss).
Zymosan-induced peritonitis in mice
For the zymosan-induced peritonitis, mice (n = 6 per experimental group) received ArchA (1 mg/kg) or vehicle (0.5 ml of 0.9% saline solution containing 2% DMSO) i.p., 60 min prior to peritonitis induction by zymosan, as previously described (19). Zymosan (Sigma-Aldrich) was prepared as a final suspension (2 mg/ml) in 0.9% (w/v) saline and injected i.p. (0.5 ml), followed by a peritoneal lavage with 3 ml of cold PBS. Cells in the inflammatory exudates were immediately counted (dilution 1:10) by using a Bürker chamber and vital trypan blue staining, and then samples were centrifuged (18,000 × g, 5 min, 4°C). The blood (∼0.7–0.9 ml) was collected by intracardiac puncture using citrate as anticoagulant, and plasma was obtained by centrifugation at 800 × g at 4°C for 10 min. LM levels in the supernatants of peritoneal exudates as well as in the plasma were assessed by UPLC–MS-MS analyses as described above (LM metabololipidomics).
Results are expressed as mean ± SEM of n observations, where n represents the number of experiments with separate donors performed on different days, or the number of animals per group (n = 6), as indicated. Analyses of data were conducted using GraphPad Prism 7 software (San Diego, CA). LM data were log transformed for statistical analysis and analyzed by paired t test. For multiple comparisons, ANOVA with Bonferroni or Dunnett post hoc tests were applied as indicated. Two-tailed t test was used for comparison of two groups. The criterion for statistically significant is p < 0.05.
V-ATPase activity is essential for SPM biosynthesis and 15-LOX-1 expression in human M2-like MDM
To study the requirement of V-ATPase activity during polarization of MDM toward proinflammatory M1- and proresolving M2-like phenotypes with focus on their differential LM profiles they produce, we took advantage of the selective V-ATPase inhibitor ArchA (17). Human monocytes were differentiated for 6 d with GM-CSF toward MGM-CSF MDM and with M-CSF toward MM-CSF MDM. Although these MDM are not primary (alveolar or peritoneal) macrophages, for which others have reported LM formation before (20, 21), they are suitable for studying the induction of LM pathways during the acquirement of different human macrophage phenotypes (5, 6). ArchA (30 nM) was added to MDM 30 min prior to 48 h polarization of MGM-CSF with LPS plus INFγ to M1 or of MM-CSF with IL-4 to M2, according to our previous report (6). To study the LM profiles by targeted LM metabololipidomics using UPLC–MS-MS, the polarized MDM were incubated with and without pathogenic E. coli (serotype O6:K2:H1, ratio 1:50) for 180 min (6). Upon exposure to E. coli, M1 and M2 produced opposing LM profiles that clearly distinguish their proinflammatory and proresolving phenotypes: M1 mainly formed COX-derived PGE2 and 5-LOX/FLAP–derived LTs/5-hydroxyeicosatatraenoic acid (5-HETE), whereas M2 generated SPM (e.g., RvD5, MaR1, PD1, and RvE3), their precursors (i.e., 17-hydroxydocosahexaenoic acid [HDHA], 14-HDHA, 7-HDHA), and 15-LOX-1–derived products (e.g., 15-HETE, 15-HEPE) with low PG and LT/5-HETE levels (Fig. 1A–C, Supplemental Table I). During M1 polarization, ArchA had little impact on LM biosynthesis and caused no significant reduction in LOX products but slightly elevated PGD2 and TxB2 levels. By contrast in M2, inhibition of V-ATPase completely prevented the formation of all SPM, their precursors, and 15-LOX-1–derived mediators (Fig. 1B, 1C, Supplemental Table I). Although some monohydroxy 5-LOX products were lowered, formation of LTB4 was not affected in M2. Moreover, solely in M2, the release of AA, eicosapentaenoic acid, and DHA was significantly impaired because of ArchA treatment (Fig. 1B, Supplemental Table I). In agreement with our previous work (16) ArchA increased the lysosomal pH during M1 and M2 polarization (Fig. 1D), assuring that V-ATPase activity was blocked in both macrophage phenotypes. To confirm on-target effects of ArchA, we used the structurally unrelated V-ATPase inhibitor bafilomycin that caused the same pattern of LM modulation as ArchA, that is, strongly impaired formation of SPM and their precursors with minor effects on 5-LOX and COX products in M2, but rather increased PG in M1 (Supplemental Fig. 1). MTT assays revealed no detrimental effects of ArchA and bafilomycin (up to 100 nM, each) on the viability of MDM during 48 h polarization (not shown).
In line with lower SPM, the protein levels of 15-LOX-1 were strongly impaired in M2 by ArchA (Fig. 2A). In contrast, 15-LOX-2 as well as 5-LOX and FLAP were not markedly affected, neither in M1 nor in M2. COX-1 protein levels were moderately decreased in M2, whereas COX-2 was elevated in M1 by ArchA. Finally, in M2, ArchA downregulated cPLA2 protein expression, but not so in M1. Again, comparable effects were obtained with bafilomycin instead of ArchA, that is, impaired 15-LOX-1 levels in M2 and elevated COX-2 in M1 without change of 5-LOX in either phenotype (Supplemental Fig. 1). Note that in control experiments, addition of ArchA to polarized M1 or M2 and subsequent exposure to E. coli for 180 min did not affect biosynthesis of LM formation or the expression of respective biosynthetic enzymes (not shown). Together, V-ATPase activity is critically required during M2 polarization for 15-LOX-1 protein induction, accompanied by increased levels of cPLA2 and COX-1 protein, but with minor importance for expression of LM pathways during M1 polarization.
To investigate if V-ATPase inhibition by ArchA would compromise M2 polarization per se, flow cytometry analysis of the phenotype-characteristic surface markers CD54 and CD80 for M1, and CD163 and CD206 for M2, was performed. In M2, ArchA had no effects on CD206 and CD163 expression, excluding a general influence on M2 polarization. However, ArchA suppressed the release of the M2-related protein IL-10 in M2 but not in M1 (Fig. 2C). In M1, considerable modulation of any macrophage phenotype markers by ArchA was not evident either (Fig. 2B), and M1-like cytokines such as TNF-α and IL-6 were not impaired (Fig. 2C). In conclusion, whereas classical human macrophage phenotype markers are hardly affected by V-ATPase inhibition during polarization, the LM profile in M2 with high SPM levels that distinguishes this proresolving phenotype from the proinflammatory M1 subset is strikingly compromised, suggesting a critical role of V-ATPase in the induction of the SPM-biosynthetic pathway.
Temporal modulation of LM-biosynthetic pathways depending on V-ATPase activity
Previous data show that the expression of the LM-biosynthetic enzymes COX-2 and 15-LOX-1 are temporally modulated during macrophage polarization (6). Thus, we studied manipulation of LM-biosynthetic enzyme expression by ArchA during M1/M2 polarization in more detail. In M2, the protein levels of 15-LOX-1 and COX-1 continuously increased during polarization up to 72 h, whereas 5-LOX slightly decreased, and cPLA2 peaked at 24 h and then declined (Fig. 3A). ArchA did not affect the expression of these enzymes at 6 h but prevented the subsequent polarization-induced increase in the expression of 15-LOX-1, COX-1, and cPLA2 starting at 24 up to 72 h (Fig. 3A,). In M1, COX-2 was strongly induced after 6 h with slightly increased cPLA2 levels followed by a continuous decline of both proteins up to 72 h; COX-1 levels were unchanged and 15-LOX-1 protein was not detectable in M1 [Fig. 3B, see also (6)]. In contrast to M2, ArchA failed to decrease COX-1 and cPLA2 levels in M1, and COX-2 levels were rather increased at 24–48 h (Fig. 3B). Like in M2, 5-LOX expression in M1 was unaffected by ArchA. Control experiments assessing V-ATPase protein expression (using Abs against V0 or V1 subunit) revealed continuous presence of the protein during M1 and M2 polarization, without significant influences of ArchA (Fig. 3C, 3D).
Abrogation of 15-LOX-1 expression and SPM formation by targeting V-ATPase is independent of the JAK/STAT6 and the PI3K/Akt–mTOR complex 1–liver X receptor pathways
Induction of M2 polarization and 15-LOX-1 expression is mediated by the JAK/STAT6 pathway that activates transcription of the human ALOX15A gene (Fig. 4A) (22–24). We hypothesized that V-ATPase activity could be required for JAK/STAT6–mediated 15-LOX-1 expression in IL-4–stimulated MDM. However, V-ATPase blockage failed to abolish phosphorylation of STAT6 at Tyr641, and thus its activation, upon 48 h of IL-4 treatment (Fig. 4B) or at other time points (6, 24, or 72 h) during polarization (Fig. 4C). In contrast, the JAK-3 inhibitor CP-690,550 (25), which in analogy to ArchA blocked 15-LOX-1 (but not 5-LOX) expression (Fig. 4D) and thus reduced SPM formation (Fig. 4E), abolished STAT6 phosphorylation in M2 (Fig. 4D), as expected. This confirms that the JAK/STAT6 pathway regulates 15-LOX-1 expression and reveals its involvement in SPM formation but excludes a role for V-ATPase in JAK/STAT6-mediated 15-LOX-1 expression during M2 polarization. Although 5-LOX protein levels were unaffected by CP-690,550 (Fig. 4D), 5-LOX product formation was slightly reduced owing to JAK-3 inhibition (Fig. 4E).
IL-4 was also shown to induce M2 polarization via signaling molecules including PI3K, Akt, mTOR complex 1 (mTORC1), p70S6K, Lamtor1/V-ATPase complex, and liver X receptor (LXR) stimulation (14, 26) (Fig. 5A). Therefore, V-ATPase might be integrated in these downstream signaling pathways of IL-4 to accomplish 15-LOX-1 expression and SPM formation. ArchA impaired phosphorylation of Akt (PI3K substrate) and p70S6K (mTORC1 substrate) in M2 (Fig. 5B, 5C), with minor or no effects in M1, whereas expression of Lamtor1, the activator of mTORC1, was not affected (Fig. 5B). In control experiments, inhibitors of PI3K (i.e., LY294002) and of mTORC1 [i.e., Torin 1 (27)] reduced the phosphorylation of their substrates Akt and p70S6K, respectively (Fig. 5D), without affecting the viability of M2 within 48 h (data not shown). However, in contrast to ArchA, neither the PI3K inhibitor LY294002 nor the mTORC1 inhibitor Torin 1 prevented 15-LOX-1 expression (Fig. 5E) or SPM formation (Fig. 5F) in M2; also, the LXR antagonist GSK-2033 failed in this respect. These data suggest that V-ATPase might be of importance for PI3K/Akt and mTORC1 signaling, but these molecules as well as LXR are dispensable for IL-4–induced 15-LOX-1 protein expression and SPM biosynthesis during M2 polarization.
15-LOX-1 expression and SPM formation in M2 involves MEK/ERK-1/2 signaling
IL-4 activates the MEK/ERK-1/2 pathway (28), and a regulatory role for ERK-1/2 in mediating 15-LOX-1 expression in monocytes was shown (29). Thus, we speculated that the MEK/ERK-1/2 pathway is involved in 15-LOX-1 expression and SPM formation in M2 and that targeting of V-ATPase may suppress the MEK/ERK-1/2 pathway (Fig. 6A). Preincubation with ArchA prior to M2 polarization strongly impaired the phosphorylation of ERK-1/2 and of its upstream kinase MEK at 24, 48, and 72 h (Fig. 6B, 6C), whereas phosphorylation of p38 MAPK was rather increased (Supplemental Fig. 2). Use of NH4Cl, another tool to elevate the pH in organelles, mimicked the effects of ArchA and suppressed ERK-1/2 phosphorylation in M2 (Supplemental Fig. 3), as expected. Noteworthy, in M1, phosphorylation of ERK-1/2 and MEK was not affected by ArchA.
We next exploited the ERK-1/2 activation inhibitor U0126 (30) to test if ERK-1/2 is required for IL-4–induced 15-LOX-1 expression and SPM formation. Pretreatment with U0126 prior to M2 polarization reduced ERK-1/2 phosphorylation as expected and significantly prevented the expression of 15-LOX-1 without affecting 5-LOX protein levels (Fig. 6D). In contrast, the p38 MAPK inhibitor skepinone-L (31) failed to influence 15-LOX-1 levels or LM biosynthesis (Supplemental Fig. 2). The transcription factor c-Myc downstream of the MEK/ERK-1/2 pathway has been shown to regulate 15-LOX-1 expression in macrophages (32), and ArchA downregulated c-Myc in leukemic cells (33) (Fig. 6B). Thus, we tested if the selective c-Myc inhibitor JQ-1 (34) would suppress 15-LOX-1 protein levels (Fig. 6E). As expected, SPM formation and 15-LOX-1-related LM were markedly impaired by U0126 as well as by JQ-1, whereas PG and fatty acid release were unaffected (Fig. 6F). Also, 5-LOX product formation was reduced by U0126 (Fig. 6F) but not so 5-LOX protein expression (Fig. 6D). Finally, the STAT6 inhibitor AS1517499, used to reinforce the link between STAT6 and 15-LOX-1, blocked phosphorylation of STAT6 and expression of 15-LOX-1 (but not 5-LOX) and also ERK-1/2 phosphorylation was moderately inhibited (Supplemental Fig. 3). Together, our data show that V-ATPase is required for functional MEK/ERK-1/2 pathway to accomplish 15-LOX-1 expression and SPM formation during M2 polarization.
Targeting V-ATPase in vivo delays resolution and suppresses SPM formation
To study the consequence of V-ATPase interference for SPM biosynthesis and resolution of inflammation in vivo, we made use of zymosan-induced peritonitis in mice (Fig. 7A), a well-established model of self-limited acute inflammation (19, 35). Application of ArchA (1 mg/kg, i.p., 1 h prior zymosan) caused marked elevation of neutrophil numbers in the peritoneum in the resolving phase 24 h postzymosan versus vehicle-treated mice, reflecting continuation of inflammation, whereas the number of neutrophils during the acute phase (4 h postzymosan) was hardly affected by ArchA (Fig. 7B). LM metabololipidomics of the exudates of these mice revealed strongly reduced levels of the SPM RvD2, RvD4, PD1, and LXA4 in ArchA-treated animals, and also RvD5 and MaR1 levels tended to be lower (Fig. 7C). Moreover, all 15-LOX-1–related SPM precursors (17-HDHA, 14-HDHA, 15-HETE, 15-HEPE, 12-HETE, 12-HEPE, 7-HDHA) were lower in exudates of mice that received ArchA, whereas the levels of 5-LOX-derived LTB4 and 5-HETE as well as of COX-derived PGs were not different between ArchA- and vehicle-treated groups (Fig. 7C). Similarly, the amounts of various circulating SPM (e.g., RvD1, RvD5) and their precursors in plasma were significantly lower 24 h post zymosan when mice received ArchA, whereas PG and LTB4 were not reduced but rather elevated (Fig. 7D). Collectively, targeting V-ATPase during inflammation in vivo locally and systemically suppresses SPM biosynthesis and delays resolution of inflammation.
One of the hallmarks of host defense during infectious inflammation is the engulfment of invaders by macrophages that transport the invader to the phagosome, an acidic digestive compartment, where the pathogen is degraded to regain homeostasis of the host. Efficient phagosome maturation requires active V-ATPase upon fusion with compartments of the endolysosomal pathway (11, 15). Also, V-ATPase plays a fundamental role in cytokine trafficking and secretion in inflammation (16, 36). However, the function of V-ATPase in the biosynthesis of either proinflammatory or proresolving LM and in resolution of inflammation has not been reported yet.
In this study, our results reveal a novel and critical role for V-ATPase in the biosynthesis of SPM in human M2-like MDM with potential consequences for resolution of inflammation. Alternative activation of macrophages by IL-4 results in proresolving M2 phenotypes that are featured by high capacities to generate SPM upon challenge with pathogenic bacteria (6) or during efferocytosis of apoptotic neutrophils (1, 5). SPM actively terminate inflammation and promote its resolution by inhibiting neutrophil infiltration, sequestering proinflammatory cytokines, promoting phagocytosis, and finally by clearance of apoptotic cells or cellular debris (7, 8). Moreover, SPM accelerate tissue repair and regeneration (37), typical features promoted by M2 macrophages. Intriguingly, M2-derived SPM, in particular the maresins, switch polarization of M1 toward the M2 phenotype (9). Our results from a pharmacological approach of targeting V-ATPase by two structurally different V-ATPase inhibitors (ArchA and bafilomycin) provide evidences for a critical role of V-ATPase in the induction of the expression of 15-LOX-1, the key enzyme in SPM biosynthesis, which is specifically upregulated during polarization toward M2 (6). Notably, the expression of 15-LOX-2 was not impaired by V-ATPase inhibition, corroborating the specificity of V-ATPase in regulating the 15-LOX-1 isoform. V-ATPase was shown to be required for IL-4–induced murine M2 polarization, but SPM biosynthesis or expression of 15-LOX-1 was not addressed (14). Our results with mice demonstrate a role of V-ATPase in SPM biosynthesis also in vivo and implicate its necessity for resolution of inflammation, reflected by unhindered neutrophil infiltration in ArchA-treated mice during peritonitis. Note that besides 15-LOX-1 and SPM, other LM pathways such as PGs and LTs were not affected by V-ATPase inhibition, neither in murine peritonitis nor in human M1. Furthermore, classical M1 and M2 polarization markers or cell viability were not influenced by V-ATPase inhibition and LM formation in the M1 phenotype; where PGs and LTs dominate (5, 6), ArchA had little to no impact. Instead, blocking V-ATPase in M1 increased COX-2 expression accompanied by elevated PG levels, in agreement with (16), a tendency observed also in our present in vivo experiments. Conclusively, V-ATPase is specifically required for induction of 15-LOX-1 and related SPM biosynthesis during M2 polarization and thus may be key for ensuring effective resolution of inflammation. However, 15-LOX-1 is a versatile enzyme, with roles not only in the termination but also in the onset of inflammation (38) and has been implicated in the pathogenesis of chronic inflammatory diseases, particularly related to the airways (39, 40). The induction of 15-LOX-1 expression by IL-4 in various cell types, including primary cultures of human monocytes, strongly depends on STAT6 (22, 23, 29). Therefore, interference with V-ATPase may prevent 15-LOX-1 induction by suppression of STAT6 signaling. However, targeting V-ATPase did neither influence STAT6 expression nor its phosphorylation, suggesting the existence of a STAT6-independent signaling route induced by IL-4 for 15-LOX-1 expression. Another prominent IL-4–induced pathway in mice that governs expression of classical M2 signature genes such as arginase-1, the mannose receptor (CD206), and IL-10, is the PI3K/Akt–mTORC1–LXR cascade in which Lamtor1 forms a complex with active V-ATPase (14). In our study using human MDM, ArchA inhibited phosphorylation of Akt and of the mTORC1 substrate p70S6K in M2, but the expression of Lamtor1, the target of mTORC1, was not affected. Moreover, inhibition of PI3K, of mTORC1, or of LXR did not influence 15-LOX-1 protein levels and SPM formation, suggesting that this pathway is negligible for the SPM-biosynthetic machinery in human M2.
Our results propose that the MEK/ERK-1/2 pathway mediates IL-4–induced 15-LOX-1 expression during M2 polarization and that targeting of V-ATPase prevents 15-LOX-1 induction by blocking MEK/ERK-1/2 signaling. In fact, phosphorylation of ERK-1/2 by MEK was shown to mediate IL-13–induced 15-LOX-1 expression in human monocytes (22, 41). In our hands, the ERK-1/2 activation inhibitor U0126 reduced 15-LOX-1 expression and suppressed SPM biosynthesis, similar to V-ATPase inhibition. As mentioned above, ArchA inhibited Akt phosphorylation, and in fact, the PI3K/Akt cascade is known to activate ERK-1/2. Recently, a direct interaction of phosphorylated PI3K, Akt, and ERK with the subunit E of the V-ATPase V1 domain was reported in cancer cells (42). Thus, the MEK/ERK-1/2 cascade appears to be a focal point for cross-cascade regulation in M2 polarization involving V-ATPase. Along these lines, the transcription factor c-Myc downstream of the MEK/ERK-1/2 pathway was suppressed by ArchA in leukemic cells (33) as well as in our hands in M2, and the c-Myc inhibitor JQ-1 (32) abrogated 15-LOX-1 protein expression and SPM formation. Conclusively, our results suggest that induction of 15-LOX-1 in human M2-like MDM is accomplished by the MEK/ERK-1/2 pathway that requires V-ATPase activity.
In addition to SPM and 15-LOX–derived precursors, 5-LOX products were also reduced by either inhibition of V-ATPase, JAK-3, or ERK-1/2, without impairing 5-LOX protein levels. Cellular 5-LOX activity depends mainly on AA supply, FLAP, Ca2+, and phosphorylations by MAPKAPK-2 and ERK-1/2 (43). ArchA, CP-690,550, or U0126 reduced AA supply and ERK-1/2, and possibly other processes required for substantial 5-LOX product formation, without altering 5-LOX protein levels. Moreover, the ERK-1/2 pathway related to 5-LOX activation and LT biosynthesis is subject of a sex dimorphism due to transient modulation by testosterone in leukocytes (44), implying potential sex differences also for ERK-1/2–mediated 15-LOX-1 regulation. However, the lengthy monocyte/MDM culture conditions in sex hormone-containing FCS (10%) may mask potential sex-dependent modulation and was not explored in this study. Ongoing studies with male and female mice indicate the existence of sex-biased 15-LOX regulation in macrophages in accordance with observations by others (45).
Genetic or pharmacological interference with V-ATPase has been linked to defective wound healing events (46) in which V-ATPase was necessary for the ERK-1/2–dependent transcriptional activation of the repair response around the wound epidermis, and nonhealing chronic wounds display extracellular pH gradients that disrupt epidermal repair (47). This supports our hypothesis on a role of V-ATPase in resolution of inflammation, especially by ensuring effective SPM biosynthesis, because SPMs were shown to effectively increase tissue remodeling and regeneration together with superior wound healing events (48). In fact, we demonstrate in this study that targeting V-ATPase in zymosan-induced peritonitis in vivo, a self-limited model of acute inflammation (19, 35), delays resolution of inflammation reflected by impaired clearance of leukocytes from the inflamed cavity and by reduced SPM levels in both exudates and plasma. Again, other LM such as LTs and PGs were unaltered by V-ATPase inhibition in this model, supporting the selectivity of this pathway for SPM biosynthesis.
Taken together, we show that V-ATPase activity is essential for induction of 15-LOX-1 during polarization of human MDM toward the proresolving M2 phenotype and the consequent capacity to generate SPM. In particular, V-ATPase appears to be crucial for IL-4–induced MEK/ERK-1/2 signaling, which is directly connected to induction of 15-LOX-1 protein expression. We conclude that V-ATPase may be a key component in alternative macrophage activation by regulating 15-LOX-1 expression and related SPM biosynthesis and thus in the resolution of inflammation.
We thank Saskia Andreas and Bärbel Schmalwasser for expert technical assistance.
This work was supported by the Deutsche Forschungsgemeinschaft (FOR1406, SFB1127, ChemBioSys and SFB1278, Polytarget), and by the Free State of Thuringia and the European Social Fund (2016 FGR 0045). C.N.S. is supported by National Institutes of Health Grants P01-GM095467 and GM038765. Z.R. was partly financed by the China Scholarship Council, and J.G. received a postdoctoral stipend from the Carl Zeiss Foundation.
The online version of this article contains supplemental material.
Abbreviations used in this article:
late endosomal and lysosomal adaptor and MAPK and mTOR activator 1
liver X receptor
mTOR complex 1
specialized proresolving mediator
ultra-performance liquid chromatography–tandem mass spectrometry
The authors have no financial conflicts of interest.