Metabolic Reprogramming of Mouse Bone Marrow Derived Macrophages Following Erythrophagocytosis

Abstract

Reticuloendothelial macrophages engulf ∼0.2 trillion senescent erythrocytes daily in a process called erythrophagocytosis (EP). This critical mechanism preserves systemic heme-iron homeostasis by regulating red blood cell (RBC) catabolism and iron recycling. Although extensive work has demonstrated the various effects on macrophage metabolic reprogramming by stimulation with proinflammatory cytokines, little is known about the impact of EP on the macrophage metabolome and proteome. Thus, we performed mass spectrometry-based metabolomics and proteomics analyses of mouse bone marrow-derived macrophages (BMDMs) before and after EP of IgG-coated RBCs. Further, metabolomics was performed on BMDMs incubated with free IgG to ensure that changes to macrophage metabolism were due to opsonized RBCs and not to free IgG binding. Uniformly labeled tracing experiments were conducted on BMDMs in the presence and absence of IgG-coated RBCs to assess the flux of glucose through the pentose phosphate pathway (PPP). In this study, we demonstrate that EP significantly alters amino acid and fatty acid metabolism, the Krebs cycle, OXPHOS, and arachidonate-linoleate metabolism. Increases in levels of amino acids, lipids and oxylipins, heme products, and RBC-derived proteins are noted in BMDMs following EP. Tracing experiments with U-¹³C₆ glucose indicated a slower flux through glycolysis and enhanced PPP activation. Notably, we show that it is fueled by glucose derived from the macrophages themselves or from the extracellular media prior to EP, but not from opsonized RBCs. The PPP-derived NADPH can then fuel the oxidative burst, leading to the generation of reactive oxygen species necessary to promote digestion of phagocytosed RBC proteins via radical attack. Results were confirmed by redox proteomics experiments, demonstrating the oxidation of Cys152 and Cys94 of glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and hemoglobin-β, respectively. Significant increases in early Krebs cycle and C₅-branched dibasic acid metabolites (α-ketoglutarate and 2-hydroxyglutarate, respectively) indicate that EP promotes the dysregulation of mitochondrial metabolism. Lastly, EP stimulated aminolevulinic acid (ALA) synthase and arginase activity as indicated by significant accumulations of ALA and ornithine after IgG-mediated RBC ingestion. Importantly, EP-mediated metabolic reprogramming of BMDMs does not occur following exposure to IgG alone. In conclusion, we show that EP reprograms macrophage metabolism and modifies macrophage polarization.

Keywords: OXPHOS; blood; lipid accumulation; macrophage metabolism; mitochondrial dysregulation; omics technologies; pentose phosphate pathway.

FIGURE 1

In vitro assessment of BMDM EP. Mouse RBCs were incubated with rabbit, anti-mouse RBC IgG. BMDMs were then incubated with PBS (control) or with IgG-coated RBCs for in vitro opsonization to reflect EP. (A) Metabolic correlates identified before (Pre) and after (Post) EP were plotted as a hierarchically-clustered heat map (B). The metabolome view map of relevant metabolic pathways showed significant changes in cellular metabolic pathways following EP (C). Univariate analysis of the BMDM metabolome using untargeted (D) and targeted (E) metabolomics methods to identify metabolites that change due to EP. The region highlighted in red (fold change (FC) ≥ 2.0; p-value < 0.05) indicates metabolites that are present in significantly higher amounts in BMDMs following EP (Post); whereas, the region highlighted in blue (fold change ≤ 0.5; p-value < 0.05) indicates metabolites that accumulate in BMDMs before EP (Pre). Amino acids identified clustered in similar regions of Pre were encircled (red).

FIGURE 2

EP reprograms glycolysis, the pentose phosphate pathway, and glutathione metabolism. Metabolites from glycolysis and the pentose phosphate pathway (PPP) (A). As part of heme metabolism, biliverdin (BILV) confirms successful EP of IgG-opsonized RBCs by BMDMs. Entry into glutathione (GSH) metabolism via methionine (MET) and glutamate (GLU) (B). For all plots, the y-axis represents relative intensity (a.u.). *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 (unpaired t-test, 2-tailed distribution). GLUC, glucose; G6P, glucose 6-phosphate; FBP, fructose bisphosphate; PG, phophoglycerate; PEP, phosphoenolpyruvate; PYR, pyruvate; LAC, lactate; GDL, 6-phosphate gluconolactone; 6PDG, 6-phospho-D-gluconate; SP, sedoheptulose phosphate; RP, ribose phosphate (isomers); GLN, glutamine; CYS, cysteine; CYSS, cystine; SER, serine; GS-CYS, S-gultathionyl-L-cysteine; HIS, histidine; DMGLY, dimethylglygine; SAM/H, S-adenosyl methionine/homocysteine.

FIGURE 3

Flux analysis of BMDMs incubated with heavy labeled substrates. BMDMs in PBS (red) or with IgG-opsonized RBCs (green) were uniformly labeled with ¹³C₆-glucose (A). Experimental results of ¹³C₆-glucose detection in BMDMs (B). Enrichment of the + 5 isotopologues of PPP metabolites, such as pentose phosphate isobaric isomers, were observed (top). Total percent labeled amounts of these isomers are shown in the bottom panel. These carbons are derived from glucose flux from the Embden-Meyerhof glycolytic pathway through the PPP. Continuous lines reflect the median of the three groups; whereas, the dashed lines represent interquartile ranges.

FIGURE 4

EP remodels purine and arginine metabolism, glutathione homeostasis, and the Krebs cycle in BMDMs. Overview and downstream metabolites of purine metabolism with entry into this pathway beginning with adenosine triphosphate (ATP) (A). Overview of arginine (ARG) metabolism, GSH homeostasis, and the Krebs cycle (B). For all plots, the y-axis represents relative intensity (a.u.). *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 (unpaired t-test, 2-tailed distribution). PO₃, phosphate; URA, urate; AMP, adenosine monophosphate; NAM, nicotinamide; HPX, hypoxanthine; INO, inosine; XAN, xanthine; ALA, aminolevulinate; ASP, aspartate; 2HG, 2-hydroxyglutarate; SUCC, succinate; ORN, ornithine; OH-ISOU, hydroxyisourate; SPMD, spermidine; SPM, spermine; ITA, itaconate; FUM, fumarate; CIT, citrate; MAL, malate; αKG, α-ketoglutarate.

FIGURE 5

EP reprograms arachidonate and linoleate metabolism in BMDMs. The top 50 metabolites in BMDMs that are part of eicosanoid and oxylipin synthesis and change significantly following EP (A). Values are plotted as a hierarchically-clustered heat map based on p-value. Overview of eicosanoid and oxylipin metabolism, stemming from arachidonic acid (ARA) and linoleic acid (LA) (B). For all plots, the y-axis represents relative intensity (a.u.). *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001 (unpaired t-test, 2-tailed distribution). PGG2, prostaglandin G2; PGH2, prostaglandin H2; PGE2, prostaglandin E2; PGF2, prostaglandin F2; HETE, hydroxyeicosatetraenoic acid; HPETE, hydroperoxyeicosatetraenoic acid; LTA4, leukotriene A4; LTB4, leukotriene B4; TX, thromboxane; TXB2, thromboxane B2; diHOME, dihydroxyoctadecanoic acids; triHOME, trihydroxyoctadecanoic acids; HODE, hydroxyoctadecadienoic acid; OxoODE, oxooctadecadienoic acid.

FIGURE 6

EP alters the BMDM proteome. (A). The top 50 proteins in BMDMs that are significantly changed by EP (A). Values are plotted as a hierarchically-clustered heat map based on p-value with proteins listed by gene name. Univariate analysis of the BMDM proteome using untargeted proteomics to identify proteins that change due to EP (B). The region highlighted in red (FC > 1.5; p < 0.05) indicates proteins present in significantly higher amounts in BMDMs after EP (Post); whereas, the region highlighted in blue (FC < 0.67; p < 0.05) indicates proteins found to be accumulated in BMDMs before EP (Pre). Representative tandem mass spectrometry spectra with deduced protein sequences for glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (C) and hemoglobin-β (D). Dehydroalanine modification of Cys152 and carbamidomethylation of Cys156 in the active site of GAPDH (C). Dehydroalanine modification of Cys94 adjacent to the heme binding site (H93, proximal) in the hemoglobin-β chain (D).