Glia maturation factor-γ regulates murine macrophage iron metabolism and M2 polarization through mitochondrial ROS

Abstract

In macrophages, cellular iron metabolism status is tightly integrated with macrophage phenotype and associated with mitochondrial function. However, how molecular events regulate mitochondrial activity to integrate regulation of iron metabolism and macrophage phenotype remains unclear. Here, we explored the important role of the actin-regulatory protein glia maturation factor-γ (GMFG) in the regulation of cellular iron metabolism and macrophage phenotype. We found that GMFG was downregulated in murine macrophages by exposure to iron and hydrogen peroxide. GMFG knockdown altered the expression of iron metabolism proteins and increased iron levels in murine macrophages and concomitantly promoted their polarization toward an anti-inflammatory M2 phenotype. GMFG-knockdown macrophages exhibited moderately increased levels of mitochondrial reactive oxygen species (mtROS), which were accompanied by decreased expression of some mitochondrial respiration chain components, including the iron-sulfur cluster assembly scaffold protein ISCU as well as the antioxidant enzymes SOD1 and SOD2. Importantly, treatment of GMFG-knockdown macrophages with the antioxidant N-acetylcysteine reversed the altered expression of iron metabolism proteins and significantly inhibited the enhanced gene expression of M2 macrophage markers, suggesting that mtROS is mechanistically linked to cellular iron metabolism and macrophage phenotype. Finally, GMFG interacted with the mitochondrial membrane ATPase ATAD3A, suggesting that GMFG knockdown-induced mtROS production might be attributed to alteration of mitochondrial function in macrophages. Our findings suggest that GMFG is an important regulator in cellular iron metabolism and macrophage phenotype and could be a novel therapeutic target for modulating macrophage function in immune and metabolic disorders.

Graphical abstract

Figure 1.

GMFG regulates iron metabolism protein expression in murine macrophages. (A-B) RAW264.7 macrophages were treated with FAC (A; 0-150 µM) or DFO (B; 0-150 µM) in 10% FBS/DMEM for 24 hours. Immunoblot analysis of GMFG in cellular lysates (upper). α-tubulin was used as a loading control. Relative quantification of GMFG protein expression levels from densitometric scans after normalizing to the control α-tubulin (lower graphs). (C) Quantitative polymerase chain reaction (qPCR) analysis of mean relative GMFG messenger RNA (mRNA) expression in FAC- or DFO-treated RAW264.7 macrophages. The data were normalized to 18S mRNA expression. (D) Immunoblot analysis of iron metabolism proteins in cellular lysates of RAW264.7 macrophages or BMDMs transfected with control siRNA (Ctrl) or GMFG siRNA for 48 hours. α-tubulin was used as a loading control. Graphs (right) show relative quantification of immunoblot (left). Protein levels from densitometric scans are normalized to the control α-tubulin and presented as fold change relative to control siRNA-transfected cells. (E) qPCR analysis of mean relative mRNA expression levels of iron metabolism proteins in RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 48 hours. The data were normalized to 18S mRNA expression. (F) Immunoblot analysis of iron metabolism proteins in cellular lysates of RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 48 hours, then treated with vehicle, FAC (50-100 µM), or DFO (100 µM) for 24 hours. α-tubulin was used as a loading control. Graphs (right) show relative quantification of immunoblot (left). Protein levels from densitometric scans are normalized to the control α-tubulin and presented as fold change relative to control siRNA-transfected cells before treatments. (G) Immunoblot analysis of iron metabolism proteins in cellular lysates of RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 48 hours, followed by cotransfection of GFP vector or GMFG-GFP plasmid for another 24 hours. α-tubulin was used as a loading control. Data represent the mean ± standard deviation of at least 3 independent experiments. *P < .05 compared with control untreated cells or control siRNA-transfected cells. a.u., arbitrary units.

Figure 2.

Knockdown of GMFG in macrophages increases intracellular iron content. (A) Representative images of immunofluorescence analysis of TfR1 in RAW264.7 macrophages transfected with control siRNA (Ctrl) or GMFG siRNA for 48 hours. TfR1 was visualized using anti-TfR1 antibody and Alexa Fluor 488–conjugated secondary antibody. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole. Scale bar, 50 μm. Quantification (right) of immunofluorescence intensity (left). Data represent the normalized mean corrected fluorescence intensity (MFI) ± standard deviation of 3 independent experiments. (B) Flow cytometry analysis of cell surface expression levels of TfR1 (CD71) in RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 48 hours. Representative histogram of unstained control siRNA-transfected cells (gray) or cells stained with the anti-CD71 antibody in control siRNA-transfected cells (red) and GMFG siRNA-transfected cells (blue) in control siRNA (left). Quantification (right) of CD71 cell surface expression flow cytometry results (left). Data represent the normalized MFI ± SD of 3 independent experiments. (C) Representative images of immunofluorescence analysis of Alexa Fluor 568–conjugated Tf internalization in RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 48 hours. Scale bar, 50 μm. (D) Analysis of total intracellular iron levels in RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 72 hours as measured by inductively coupled plasma mass spectrometry. Iron content was normalized to the total iron in control siRNA-transfected cells. (E) Analysis of labile iron pool in RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 72 hours as measured by calcein AM. Iron content was normalized to the total iron in control siRNA-transfected cells. Data represent the mean ± standard deviation of at least 3 independent experiments. *P < .05 compared with control siRNA-transfected cells.

Figure 3.

Knockdown of GMFG promotes M2 macrophage polarization. (A-F) BMDMs were transfected with control siRNA (Ctrl) or GMFG siRNA for 48 hours, then stimulated without (M0) or with M1 (LPS/IFN-γ) or M2 (IL-4/IL-13) macrophage inducers for 24 hours. Bar graphs represent quantitative polymerase chain reaction (qPCR) analysis of mean relative mRNA expression of M2 (A-E) or M1 (F) macrophage marker genes normalized to 18S mRNA expression. (G-I) qPCR analysis of mean relative mRNA expression of M2 macrophage marker genes in BMDMs transfected with control siRNA or GMFG siRNA for 48 hours, followed by cotransfection of GFP vector or GMFG-GFP plasmid for another 24 hours, then stimulated without (M0) or with M2 macrophage inducers for 24 hours, normalized to 18S mRNA expression. Data represent the mean ± standard deviation of at least 3 independent experiments. *P < .05 compared with control siRNA-transfected cells.

Figure 4.

Knockdown of GMFG in macrophages alters iron metabolism protein expression to mimic that observed in the M2 macrophage phenotype. (A) Immunoblot analysis of iron metabolism proteins in cellular lysates of RAW264.7 macrophages transfected with control siRNA (Ctrl) or GMFG siRNA for 48 hours, then stimulated without (M0) or with M1 (LPS/IFN-γ) or M2 (IL-4/IL-13) macrophage inducers for 24 hours. α-tubulin was used as a loading control. (B) Immunoblot analysis of IRP1, IRP2, and HIF-2α in cellular lysates of RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 48 hours, then stimulated without (M0) or with M1 or M2 macrophage inducers. α-tubulin was used as a loading control. (C) Representative RNA-binding activity of IRP1 analyzed by electrophoretic mobility shift assays. RAW264.7 macrophages were transfected with control siRNA or GMFG siRNA for 48 hours, then stimulated without (M0) or with M1 or M2 macrophage inducers for 24 hours. Cytoplasmic lysates were incubated with an excess of a phosphorus-32–labeled iron regulatory element probe. RNA-protein complexes were resolved on nondenaturing polyacrylamide gels and revealed by autoradiography.

Figure 5.

GMFG modulates mtROS production and the mitochondrial respiration chain in macrophages. (A) Immunoblot analysis of GMFG, TfR1, and IRP1 in cellular lysates of RAW264.7 macrophages treated with hydrogen peroxide (0-250 µM) in 10% FBS/DMEM for 24 hours. α-tubulin was used as a loading control. (B-I) RAW264.7 macrophages were transfected with control siRNA (Ctrl) or GMFG siRNA for 48 hours, then stimulated without (M0) or with M1 (LPS/IFN-γ) or M2 (IL-4/IL-13) macrophage inducers for 24 hours. (B-E) mtROS, total mitochondrial mass, and mitochondrial membrane potential (Δψ_m) were analyzed by labeling cells with MitoSOX (B), MitoTracker Green (C), MitoTracker Red (D), or TMRM (E), respectively. Stained cells were then subjected to flow cytometry. (F-G) Immunoblot analysis of mitochondrial respiratory chain complex subunit proteins (complex I [CI; NDUFV2], complex II [CII; SDHD], complex III [CIII; CORE2], and complex IV [CIV; COX5A]) (F) or ISCU and antioxidant proteins (G). α-tubulin was used as a loading control. (H-I) Oxygen consumption rate (OCR) were measured under basal conditions followed by the sequential addition of oligomycin (1 µM), carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone (0.5 µM), and rotenone (0.5 µM) plus antimycin A (0.5 µM) in control siRNA or GMFG siRNA-transfected Raw264.7 cells. Mean basal (H) and maximal (I) OCRs were measured using a Seahorse XF-24. OCRs were normalized by number of living cells in each condition. Basal OCR was measured over time for a single experiment. Data represent the mean ± standard deviation of at least 3 independent experiments. *P < .05 compared with control siRNA-transfected cells of the same phenotype.

Figure 6.

GMFG-knockdown modulation of iron metabolism protein expression and the M2 macrophage phenotype are associated with increased mtROS. (A) mtROS levels in RAW264.7 macrophages transfected with control siRNA (Ctrl) or GMFG siRNA for 48 hours, then treated with 8 mM of NAC for 30 minutes. Macrophages were then stained with MitoSOX and subjected to flow cytometry, and fluorescence-activated cell sorting MFI values were calculated. Intracellular ROS was expressed as the fold change of MFI normalized to the controls. Data represent the mean ± standard deviation of 3 independent experiments. (B) Immunoblot analysis of iron metabolism proteins in cellular lysates of RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 48 hours, then stimulated without (M0) or with M2 (IL-4/IL-13) macrophage inducers for 24 hours. Cells were subsequently treated with or without 8 mM of NAC for 30 minutes. α-tubulin was used as a loading control. (C-G) Quantitative polymerase chain reaction (qPCR) analysis of mean relative mRNA expression of M2 macrophage marker genes in RAW264.7 macrophages transfected with control siRNA or GMFG siRNA for 48 hours, then stimulated without (M0) or with M2 (IL-4/IL-13) macrophage inducers for 24 hours. Cells were subsequently treated with or without 8 mM of NAC for 30 minutes before RNA isolation. Expression levels were normalized to 18S mRNA expression. Data represent the mean ± standard deviation of at least 3 independent experiments. *P < .05 compared with no NAC treatment GMFG siRNA-transfected M2 cells.

Figure 7.

GMFG is associated with the mitochondrial membrane protein ATAD3A. (A) Cellular lysates of human THP-1 cells transfected with control siRNA (Ctrl) or GMFG siRNA for 48 hours were immunoprecipitated with control immunoglobulin G (IgG) or monoclonal anti-GMFG antibodies. The immunoprecipitated proteins were isolated using Dynabeads protein G, eluted, and subjected to liquid chromatography-tandem mass spectrometry analysis. Proteome data from Ctrl siRNA- or GMFG siRNA-transfected THP-1 cells were analyzed by MaxQuant, including protein name, intensity L (Ctrl siRNA- or GMFG siRNA-transfected cells immunoprecipitated with IgG antibody), intensity H (Ctrl siRNA-transfected cells immunoprecipitated with GMFG antibody), intensity M (GMFG siRNA-transfected cells immunoprecipitated with GMFG antibody), normalized H/L ratio, H/M ratio, and M/L ratio. (B-C) Immunoprecipitation analysis of GMFG association with the mitochondrial membrane protein ATAD3A in human THP-1 cells. Total cellular lysates were immunoprecipitated with anti-GMFG antibody (B) or anti-ATAD3A antibody (C), then the immunoprecipitants subjected to immunoblot analysis with anti-ATAD3A or anti-GMFG antibody. Samples of the total lysate after immunoprecipitated complexes/beads were isolated (lysates) are shown in the left 2 lanes; immunoprecipitates (IPs) are shown in the right 2 lanes. (D) Cellular lysates of human HEK-293T cells cotransfected with GFP-tagged GMFG plasmid and Myc-DDK–tagged ATAD3A plasmid for 48 hours were immunoprecipitated with control IgG, anti-GFP, or anti-Myc antibody; the IPs were then subjected to immunoblot analysis with anti-ATAD3A or anti-GMFG antibody. Samples of the total lysate after immunoprecipitated complexes/beads were isolated (lysates; bottom). Each sample corresponds to 5% of the cell lysate used in each immunoprecipitation. (E) Immunoblot analysis of ATAD3A proteins in cellular lysates of RAW264.7 macrophages transfected with control siRNA (Ctrl) or GMFG siRNA for 48 hours, then stimulated without (M0) or with M1 (LPS/IFN-γ) or M2 (IL-4/IL-13) macrophage inducers for 24 hours. α-tubulin was used as a loading control.