Volume 105, Issue 2 p. 215-228
HIGHLIGHTED ARTICLE
Free Access

Frontline Science: Monocytes sequentially rewire metabolism and bioenergetics during an acute inflammatory response

Xuewei Zhu

Corresponding Author

Xuewei Zhu

Department of Internal Medicine/Molecular Medicine, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Department of Microbiology and Immunology, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Correspondence

Xuewei Zhu, Department of Internal Medicine, Section on Molecular Medicine, Wake Forest School of Medicine, Medical Center Blvd, Winston-Salem, NC 27157, USA.

Email: [email protected]

Charles E. McCall, Department of Internal Medicine, Section on Molecular Medicine, Wake Forest School of Medicine, Medical Center Blvd, Winston-Salem, NC 27157, USA.

Email: [email protected]

Search for more papers by this author
Allison Meyers

Allison Meyers

Department of Microbiology and Immunology, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Search for more papers by this author
David Long

David Long

Department of Internal Medicine/Molecular Medicine, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Search for more papers by this author
Brian Ingram

Brian Ingram

Metabolon, Inc., Morrisville, North Carolina, USA

Search for more papers by this author
Tiefu Liu

Tiefu Liu

Department of Internal Medicine/Molecular Medicine, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Search for more papers by this author
Barbara K. Yoza

Barbara K. Yoza

Department of Surgery/General Surgery and Trauma, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Search for more papers by this author
Vidula Vachharajani

Vidula Vachharajani

Department of Anesthesiology, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Search for more papers by this author
Charles E. McCall

Corresponding Author

Charles E. McCall

Department of Internal Medicine/Molecular Medicine, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Department of Microbiology and Immunology, Wake Forest School of Medicine, Winston-Salem, North Carolina, USA

Correspondence

Xuewei Zhu, Department of Internal Medicine, Section on Molecular Medicine, Wake Forest School of Medicine, Medical Center Blvd, Winston-Salem, NC 27157, USA.

Email: [email protected]

Charles E. McCall, Department of Internal Medicine, Section on Molecular Medicine, Wake Forest School of Medicine, Medical Center Blvd, Winston-Salem, NC 27157, USA.

Email: [email protected]

Search for more papers by this author
First published: 11 January 2019
Citations: 40

Abstract

Metabolism directs the severe acute inflammatory reaction of monocytes to guard homeostasis. This occurs by sequentially activating anabolic immune effector mechanisms, switching to immune deactivation mechanisms and then restoring immunometabolic homeostasis. Nuclear sirtuin 1 and mitochondrial pyruvate dehydrogenase kinase metabolically drive this dynamic and are druggable targets that promote immunometabolic resolution in septic mice and increase survival. We used unbiased metabolomics and a validated monocyte culture model of activation, deactivation, and partial resolution of acute inflammation to sequentially track metabolic rewiring. Increases in glycogenolysis, hexosamine, glycolysis, and pentose phosphate pathways were aligned with anabolic activation. Activation transitioned to combined lipid, protein, amino acid, and nucleotide catabolism during deactivation, and partially subsided during early resolution. Lipid metabolic rewiring signatures aligned with deactivation included elevated n-3 and n-6 polyunsaturated fatty acids and increased levels of fatty acid acylcarnitines. Increased methionine to homocysteine cycling increased levels of s-adenosylmethionine rate-limiting transmethylation mediator, and homocysteine and cysteine transsulfuration preceded increases in glutathione. Increased tryptophan catabolism led to elevated kynurenine and de novo biosynthesis of nicotinamide adenine dinucleotide from quinolinic acid. Increased branched-chain amino acid catabolism paralleled increases in succinyl-CoA. A rise in the Krebs cycle cis-aconitate-derived itaconate and succinate with decreased fumarate and acetyl-CoA levels occurred concomitant with deactivation and subsided during early resolution. The data suggest that rewiring of metabolic and mitochondrial bioenergetics by monocytes sequentially activates, deactivates, and resolves acute inflammation.

Abbreviations

  • AcSDKP
  • Ac-Ser-Asp-Lys-Pro-OH
  • BCAAs
  • branched-chain amino acids
  • GC-MS
  • gas chromatography-mass spectroscopy
  • GPC
  • glycerolphosphorylcholine
  • GPE
  • glycerolphosphorylethanolamine
  • GSH
  • reduced glutathione
  • GSSG
  • oxidized glutathione
  • LC
  • liquid chromatography
  • PCA
  • principal component analysis
  • PDC
  • pyruvate dehydrogenase complex
  • PDK
  • pyruvate dehydrogenase kinase
  • PPP
  • pentose phosphate pathway
  • PUFA
  • polyunsaturated fatty acids
  • SAH
  • S-adenosylhomocysteine
  • SAM
  • S-adenosylmethionine
  • SDH
  • succinate dehydrogenase
  • SIRT
  • sirtuin
  • UPLC-MS/MS
  • ultrahigh performance liquid chromatography-tandem mass spectroscopy
  • 1 INTRODUCTION

    Acute inflammation supports evolution's universal survival principle to rapidly recognize, resist, or tolerate danger, and reestablish homeostasis.1, 2 Specialized immune cells evolved to support the survival principle,3 including proinflammatory monocytes and macrophages. When resistance and/or tolerance mechanisms fail, acute or chronic inflammatory diseases occur, such as sepsis, obesity, diabetes, atherosclerosis, and dementia.1 After decades of research, the dearth of specific molecular targets for treating inflammatory diseases remains a major gap in understanding the molecular details of the universal protection of homeostasis and survival principles encoded by inflammation.4

    To support evolution's survival principle, monocytes and tissue macrophages must be metabolically flexible.5 Metabolic flexibility in monocytes and macrophages sequentially mounts anabolic energy-consuming immune activation during resistance, switching it to a catabolic energy-conserving process during immune deactivation, and restoring homeostasis by reestablishing anabolic and catabolic energy balance during resolution.6 Innate and adaptive immune cells also become immune tolerant during sepsis.7 Organ cell-specific deactivation upon tolerance, while less recognized, also typifies the acute systemic inflammatory and potentially lethal response of sepsis.8, 9 Broad-based metabolic and bioenergy defects generate immunometabolic paralysis during sepsis and a catabolic low-energy state similar to hibernation and suspended animation.10, 11 The worldwide sepsis crisis highlights the need to identify molecular pivots, or rheostats, to target the immunometabolic survival mechanisms encoded by acute inflammation.12

    Our THP-1 monocyte cell culture model of acute inflammation first described key reprogramming axes, or pivots, of activation and deactivation of acute inflammation identified by endotoxin tolerance.13-15 Subsequently, these findings were reproduced and translated in primary monocytes and macrophages, human blood monocytes from septic patients and assessed ex vivo, and in a mouse model of sepsis.16-18 The first pivot we identified was transcription activation factor NFκB p65 and transcription repressor RelB, which sequentially exchange promoter occupancy on specific gene sets to convert from activation to deactivation during acute inflammation.13, 15, 19, 20 The indispensable pivot was extended to include the nuclear epigenetic metabolic and immune regulators controlled by nicotinamide adenine dinucleotide (NAD)+ redox-sensing nuclear sirtuins (SIRT) 1, 2, and 6, and the posttranslational regulators mitochondrial SIRT 3 and 4.13, 17, 21, 22 Pyruvate dehydrogenase complex (PDC), which catalyzes the rate-limiting step of glucose oxidation, is a master homeostat for regulating activation and deactivation,23 and increased expression of pyruvate dehydrogenase kinase (PDK)1 inhibits PDC in severe sepsis.21 The key pivots in monocytes, SIRT118 and PDK,24 are druggable targets for reversing deactivation to promote activation concomitant with increasing glucose oxidation, to rebalance mitochondrial anabolic and catabolic bioenergetics, and to improve survival in a mouse model of septic shock. The THP-1 monocyte cell model of acute inflammatory immunometabolic rewiring used in these molecular targeting studies has succeeded in informing and then translating molecular pathways to septic mice and blood monocytes from patients with sepsis.25

    Nonetheless, how the resistance, deactivation, and resolution metabolic bioenergetics are sequentially rewired in sepsis is unknown. Thus, we used unbiased metabolomics in our THP-1 human monocyte cell culture model of acute inflammation and sepsis to temporally follow metabolic rewiring. We found broad-based and temporally distinct metabolic rewiring of carbohydrates, lipids, proteins, nucleotides, and amino acids as activation and deactivation progressed from anabolism to catabolism to early resolution. The Krebs cycle opened at cis-aconitase and succinate dehydrogenase (SDH) concomitant with increases in itaconate and succinate during deactivation. We suggest that monocytes sequentially rewire anabolic and catabolic bioenergetics to drive activation, deactivation, and early resolution of an acute inflammatory response.

    2 MATERIALS AND METHODS

    2.1 Cell model and sample preparation

    We used lipopolysaccharide (LPS) stimulation of THP-1 cells, a previously validated model of sepsis.26 THP-1 cells were maintained in RPMI 1640 media supplemented with 10% fetal bovine serum (Gemini Bioproducts, West Sacramento, CA USA), 2 mM glutamine, and penicillin/streptomycin at a density of between 0.3 × 106 and 1 × 106 cells per milliliter; we added fresh media every 2–3 days. On the day of experiments, cells were spun down and resuspended at a density of 0.7 × 106 cells per milliliter and allowed to recover in fresh media for a minimum of 2 h prior to stimulation. Cells were stimulated with a high dose of LPS (E. Coli 0111; B4 (Sigma-Aldrich, St. Louis, MO USA), 1 μg/mL), and metabolic profiles were assessed at time points of 8, 24, 48, 72, and 96 h after initial LPS exposure. Each sample contained 20 million viable cells. Samples were prepared using the automated MicroLab STAR® system from Hamilton Company. A recovery standard was added before the first step in the extraction process for purposes of quality control. To remove protein, dissociate small molecules bound to protein or trapped in the precipitated protein matrix, and to recover chemically diverse metabolites, proteins were precipitated with methanol under vigorous shaking for 2 min (Glen Mills GenoGrinder 2000) followed by centrifugation. The resulting extract was divided into five fractions: one for analysis by ultrahigh performance liquid chromatography-tandem mass spectroscopy (UPLC-MS/MS) with positive ion mode electrospray ionization, one for analysis by UPLC-MS/MS with negative ion mode electrospray ionization, one for analysis by UPLC-MS/MS polar platform (negative ionization), one for analysis by GC-MS, and one sample was reserved for backup. Samples were placed briefly on a TurboVap (Zymark, Hopkinton, MA USA) to remove the organic solvent. For liquid chromatography (LC), samples were stored overnight under nitrogen before preparation for analysis. For gas chromatography (GC), each sample was dried under vacuum overnight before preparation for analysis.

    2.1.1 UPLC-MS/MS

    The LC/MS portion of the platform was based on a Waters Acquity UPLC device and a Thermo Scientific Q-Exactive high-resolution/accurate mass spectrometer interfaced with a heated electrospray ionization (HESI-II) source and Orbitrap mass analyzer operated at 35,000 mass resolution. The sample extract was dried and then reconstituted in acidic or basic LC-compatible solvents, each of which contained eight or more injection standards at fixed concentrations to ensure injection and chromatographic consistency. One aliquot was analyzed using acidic positive ion optimized conditions and the other using basic negative ion optimized conditions in two independent injections on separate dedicated columns (Waters UPLC BEH C18-2.1 × 100 mm, 1.7 μm).

    Extracts reconstituted in acidic conditions were gradient eluted from a C18 column using water and methanol containing 0.1% formic acid. The basic extracts were similarly eluted from C18 using methanol and water, but with 6.5 mM ammonium bicarbonate added. The third aliquot was analyzed via negative ionization following elution from a hydrophilic interaction liquid chromatography column (Waters UPLC BEH Amide 2.1 ×  150 mm, 1.7 μm) using a gradient consisting of water and acetonitrile with 10 mM ammonium formate. The MS analysis alternated between MS and data-dependent MS2 scans using dynamic exclusion, and the scan range was from 80–1000 m/z. Raw data files were archived and extracted as described below.

    2.1.2 GC-MS

    Samples for these analyses were dried under vacuum for a minimum of 18 h before being derivatized under dried nitrogen using bistrimethyl-silyltrifluoroacetamide. Derivatized samples were separated on a 5% diphenyl/95% dimethyl polysiloxane fused silica column (20 m × 0.18 mm ID; 0.18 μm film thickness) with helium as the carrier gas and a temperature ramp from 60 to 340°C in a 17.5 min period. Samples were analyzed on a Thermo-Finnigan Trace DSQ fast-scanning single-quadrupole mass spectrometer using electron impact ionization and operated at unit mass resolving power. The scan range was from 50–750 m/z. Raw data files were archived and extracted as described below.

    2.1.3 Data extraction, compound identification, and curation

    Raw data were extracted, peak-identified, and processed for quality control using Metabolon's hardware and software. Compounds were identified by comparison to library entries of purified standards or recurrent unknown entities. Biochemical identifications included three criteria: retention index within a narrow window of the proposed identification, accurate mass match to the library ±0.005 amu, and the MS/MS forward and reverse scores between the experimental data and authentic standards. The MS/MS scores compared the ions in the experimental and library spectra.

    Multiple curation procedures ensured a high-quality data set for robust statistical analysis and data interpretation. The quality control and curation processes were designed to ensure accurate and consistent identification of true chemical entities and to remove system artifacts, misassignments, and background noise. Metabolon uses proprietary visualization and interpretation software to confirm the consistency of peak identification among samples. Library matches for each compound were checked for each sample and corrected if necessary.

    2.1.4 Statistical analyses

    Peaks were quantified using area-under-the-curve. The biochemical data were normalized to total protein as determined by the Bradford assay. Following log transformation, Welch's t test was used to identify biochemicals that differed significantly between the experimental groups. P values ≤0.05 were considered statistically significant, and P values 0.05 ≥ 0.10 were reported as trends. Multiple comparisons were accounted for by estimating the false discovery rate using q values.27

    3 RESULTS

    3.1 Principal component analysis

    Principal component analysis (PCA) is a mathematical procedure that provides a high-level view of a data set's structure. PCA permits visualization of how individual samples in a data set differ from each other. Samples with similar biochemical profiles cluster together, whereas samples with different biochemical profiles segregate from one another. A total of 494 biochemicals were detected and four samples were analyzed for each time point. We chose time points of 0, 8, 24, 36, 48, 72, and 96 h to cover the continuum from basal state, to activation peak (0–8 h), to deactivation (24–48 h), and transition to early resolution (48–96 h).13 Untreated cells and LPS-treated cells distinctly separated across activation, deactivation, and resolution states, as suggested by PCA (Fig. 1). This significant reprogramming of the metabolome supports the concept of separate and sequential metabolic rewiring across the acute inflammatory reaction spectrum from activation to deactivation, and to early resolution, outside of the basal state.

    Details are in the caption following the image
    PCA of metabolites of THP-1 monocytes. PCA score plot generated from a total of 494 biochemicals from untreated and LPS-treated cells

    3.2 Carbohydrates

    Carbohydrates (and in some cases, amino acids) provide carbons to support anabolism during monocyte activation, to increase biomass for differentiation (monocytes including THP-1 cells in culture are slow replicators) and optimize immune resistance mechanisms.5, 6 Anabolism requires and consumes large quantities of ATP at a high rate. Figure 2 shows early and transient increases in glycogen metabolism, plotted as increases in glucose-1-phosphate and glycolysis, shown by increased fructose-1,6-bisphosphate at 8 h peak. Decreases occurred in UDP-glucuronate concomitant with transient increases in UDP-glucose. Both lactate and pyruvate decreased as activation transitions to deactivation. Glycolysis diversion into the pentose phosphate pathway (PPP) is supported by increased 6-phosphogluconate accumulation during resolution. Decreases in ribulose-5-phosphate occurred as deactivation developed, with a slight upturn during resolution. A transient increase in sedoheptulose-7-phosphate, a step in the PPP induced by the carbohydrate response kinase-like protein (an orphan reaction in the PPP),28, 29 supports the concept of a deactivation pivot.

    Details are in the caption following the image
    Quantification of carbohydrate metabolites in glycolysis/gluconeogenesis, PPP, and hexosamine metabolic pathways. Intracellular metabolites of THP-1 monocytes represented by box and whisker plots (n = 4). Significant increases of P < 0.05 are depicted by dark red, and P < 0.10 as light red. Significant decreases of P < 0.05 as dark green and P < 0.10 are light green

    3.3 Lipids

    Lipid metabolites in plasma from septic animals and humans show increased acylcarnitine fatty acid derivatives and decreased glycerolphosphorylcholine (GPC) with glycerophosphorylethanolamine (GPE) membrane components, predicting higher mortality rates during sepsis.30, 31 Figure 3 shows significant increases in monocyte GPC, choline, and glycerol 3-phosphate at peak activation (8 h), which decreased during deactivation and remained low in early resolution. Ethanolamine derived from GPE similarly increased during deactivation and remained high during resolution (Fig. 3). PE levels also increased, but unlike GPC continued to rise through resolution (Supplemental Fig. 1S).

    Details are in the caption following the image
    Altered membrane lipid metabolism in LPS-treated THP-1 monocytes. Intracellular phospholipid and its metabolites represented by box and whisker plots (n = 4). Significant increases of P < 0.05 are depicted by dark red, and P < 0.10 as light red

    The GPE metabolic rewiring marks the endocannabinoid metabolic pathway, which alters bioenergetics.32 Acylcarnitine increased significantly across activation and deactivation, declining during resolution (Fig. 4). n-6 and n-3 polyunsaturated fatty acids (PUFA) increased during metabolic rewiring (Fig. 5), and sphingomyelin increased during activation (Supplemental Fig. 2S); ceramides and cardiolipin were not measured.

    Details are in the caption following the image
    Acylcarnitine metabolite profiles. Heatmap table (left) and box and whisker plots (right) showed the accumulation of acylcarnitines in LPS-treated THP-1 cells (n = 4). Significant increases of P < 0.05 are depicted by dark red, and P < 0.10 as light red
    Details are in the caption following the image
    Polyunsaturated fatty acids (PUFAs) of the n-3 and n-6 fatty acid elongation and desaturation pathway in THP-1 monocytes. Significant increases of P < 0.05 are depicted by dark red

    3.4 Proteins and amino acids

    ATP depletion, with increases in AMP and AMP-activated protein kinase activation and glucose deprivation, blocks mammalian target of rapamycin during the starvation response, which degrades proteins.33, 34 As a potential marker of protein degradation, broad decreases in dipeptides were accompanied by increases in the lysine degradation products cadaverine, 5-aminovaleric acid, and putrescine (Fig. 6). Peptides generated from proteins can be functional. Supplemental Fig. 3S shows increased accumulation of the polypeptide Ac-Ser-Asp-Lys-Pro-OH (AcSDKP), cleaved from the N-terminus of thymosin β4, which is a physiologic regulator of hematopoiesis and blocks lymphocyte proliferation.35

    Details are in the caption following the image
    Protein metabolites in LPS-treated THP-1 monocytes. (A) dipeptides and (B) lysine degradation products. Significant increases of P < 0.05 are depicted by dark red, and P < 0.10 as light red. Significant decreases of P < 0.05 as dark green and P < 0.10 are light green

    Figure 7 shows decreased levels of many amino acids during activation at 8 h, which persisted during deactivation. Tryptophan deficiency underlies immune deficiency phenotypes and may promote immune dysfunction during sepsis.36, 37 The significant increase in tryptophan-derived metabolites, including kynurenine and quinolinic acid, spanned activation and deactivation, whereas kynurenine decreased and quinolinate and NAD remained elevated during early resolution (Fig. 8). The data are consistent with increases in indoleamine 2,3-dioxygenase activity.

    Details are in the caption following the image
    Amino acids in LPS-treated THP-1 monocytes. Significant increases of P < 0.05 are depicted by dark red, and P < 0.10 as light red. Significant decreases of P < 0.05 as dark green and P < 0.10 are light green
    Details are in the caption following the image
    Tryptophan-derived metabolite profile in LPS-treated THP-1 monocytes. Intracellular metabolites represented by box and whisker plots (n = 4). Significant increases of P < 0.05 are depicted by dark red, and P < 0.10 as light red. Significant decreases of P < 0.05 as dark green

    Oxidized (GSSG) and reduced glutathione (GSH) increased by 8 h (Fig. 9) and decreased during resolution. GSH elevation follows increased transsulfuration of the amino acids methionine and cysteine.38, 39 The upstream methionine to cysteine cycle requires S-adenosylhomocysteine (SAH) and S-adenosylmethionine (SAM) derived from glycolysis metabolism to glycerol 3-phosphate, serine synthesis, and one-carbon metabolism. SAM, the rate-limiting step and sole methylation donor for epigenetic and other protein modifications,40 increased during deactivation and decreased during resolution.

    Details are in the caption following the image
    The metabolic profile of methionine metabolism and glutathione synthesis pathways. Quantification of metabolites in these pathways is shown in box and whisker plots (n = 4). Significant increases of P < 0.05 are depicted by dark red, and P < 0.10 as light red. Significant decreases of P < 0.05 as dark green

    3.5 Branched-chain amino acids

    The branched-chain amino acids (BCAAs) leucine, isoleucine, and valine increase in plasma of diabetic patients and may contribute to hyperglycemia.41-43 Catabolism of all three BCAAs increased, which requires reversible transamination by branched-chain aminotransferase followed by irreversible activation of branched-chain keto acid dehydrogenase.44 Figures 10 and 11 show sequential catabolism of leucine and isoleucine.

    Details are in the caption following the image
    Catabolism of branched-chain amino acids in THP-1 monocytes. Quantification of leucine and its metabolites is shown in box and whisker plots (n = 4). Significant increases of P < 0.05 are depicted by dark red and significant decreases of P < 0.05 as dark green
    Details are in the caption following the image
    Catabolism of branched-chain amino acids in THP-1 monocytes. Quantification of isoleucine and its metabolites is shown in box and whisker plots (n = 4). Significant increases of P < 0.05 are depicted by dark red and significant decreases of P < 0.05 as dark green

    3.6 Nucleotides

    Nucleic acid catabolites were enriched at early time points and diminished during resolution (Supplemental Fig. 4S). The purines adenosine, adenine, and AMP decreased throughout acute inflammation rewiring, whereas the metabolic products xanthosine 5'-monophosphate, xanthine, inosine, and xanthosine increased. Xanthine oxidation during uric acid cycling is a source of free radicals. Pyrimidine degradation products from UMP (including uracil, 5,6-dihydrouracil, and beta-alanine) increased in parallel with purine products. In contrast, pyrimidine thymidine monophosphate and thymidine decreased, possibly reflecting alterations in proximal one-carbon folic acid connections.

    3.7 Urea cycle

    Time-dependent changes in the metabolism of arginine (ornithine, proline, argininosuccinate, N-delta-acetylornithine, and polyamines) occurred during the acute inflammatory response spectrum (Supplemental Fig. 5S). The change may reflect alterations in nitric oxide synthase and/or arginase activity.

    3.8 Krebs cycle

    Krebs cycle support of anabolic and catabolic bioenergetics and its differential management of glucose, fatty acid, and amino acid carbons are critically important in directing activation, deactivation, and resolution.45-48 Figure 12 shows changes in the Krebs cycle and valine catabolism to succinyl-CoA, which precedes succinate generation. Significant decreases in succinylcarnitine, increases in succinate, and decreases in fumarate occurred as activation transitioned to catabolic deactivation; these changes subsided during resolution. The data support a block at SDH during acute inflammation, thereby limiting reducing agents (flavin adenine dinucleotide) needed to support electron transport and ATP synthesis. C5-dicarboxylic acid itaconate (methylene succinate) derived from aconitase (Fig. 12) supports a second Krebs cycle break at isocitrate.49, 50 Itaconate increased as activation transitioned into deactivation and decreased as deactivation entered resolution. Acetyl-CoA levels, which support acetylation of histones and proteins by acetate,40 decreased deactivation. Increased valine catabolism to succinylcarnitine and succinyl-CoA preceded increased succinate proximal to Krebs cycle breakage at SDH.

    Details are in the caption following the image
    Krebs cycle and valine metabolites in THP-1 monocytes treated with or without LPS. Significant increases of P < 0.05 are depicted by dark red, and P < 0.10 as light red. Significant decreases of P < 0.05 as dark green

    4 DISCUSSION

    We find sequential changes in the unbiased metabolome assessed during a severe acute inflammatory response in human monocytes, as it progressively departs basal homeostasis, transits from activation to deactivation, and then begins to resolve. The overall results support the broad-based metabolic and bioenergetic shifts previously reported in plasma30, 31 and primary monocytes of septic mice and humans, and when assessed at specific time points in the THP-1 human monocyte cell model of sepsis.17, 21, 25, 51-53 Overarching features of metabolic rewiring of acute inflammation during sepsis include decreases in ATP, NADH, and acetyl-CoA during catabolic deactivation, as sepsis evolves to tolerance and immune suppression.53, 54 The bioenergy demand and supply that separate resistance and tolerance of immune and organ-specific cells may underlie the poor outcomes and high mortality rates of individuals with severe acute inflammation and septic shock. However, the most important gap in understanding sepsis is knowing how inflammation succeeds in protecting homeostasis and resolves in sepsis survivors but fails in victims. Bioenergy management is one possibility.

    Our findings provide new information about the metabolic and bioenergetics rewiring axes driving activation, deactivation, and resolution. Stimulating mitochondrial PDC-dependent glucose oxidation increases bioenergy indices and survival in septic mice,24 highlighting the critical role of mitochondria in regulating energy demand and supply during acute inflammation, and supporting the Darwinian survival-of-the-fittest and resist and tolerate concepts.

    We show anabolic activation with increased glycogenolysis, glycolysis, and hexosamine metabolism before deactivation, followed by metabolic rewiring of lipids, proteins, amino acids, nucleotides, and Krebs cycle bioenergetics. The shift from a carbohydrate high-energy yielding and consuming state, to a starvation-like state of catabolism in monocytes, resembles the inertia phenotypes reported during organism-wide sepsis-induced deactivation.11 Our data suggest that catabolism includes combined lipolytic processes, such as increased glycerophosphorylcholine, glycerophosphorylethanomine, and fatty acid acylcarnitine derivatives, which may reflect increased fatty acid uptake and lipolysis. Increased proteolysis as activation progressed to deactivation was shown by decreased dipeptides and increased putrescine and cadaverine lysine degradation products. Nucleotide turnover, which included increased catabolism of purines and pyrimidines, was active throughout the inflammation spectrum. Some metabolites of lipids (e.g., acylcarnitines) and organic acids from the Krebs cycle (e.g., succinate) found in plasma signal poor sepsis outcomes in humans and animals.30, 55 The global rewiring of structural and functional molecular species observed in this study is compatible with increased lysosome activity, autophagy, and mitophagy activation, during which substrate is reused to conserve and repurpose energy and thus resolve acute inflammation.56, 57

    Pathway-specific lipid, amino acid, and Krebs cycle rewiring signatures could alter immunometabolic physiology and pathophysiology. First, increased PUFA levels could generate proinflammatory mediators from arachidonic acid and repress TLR4 signaling.58 In addition to altering lipid rafts, PUFAs repress TLR4 signaling and NLRP3 inflammasome complex assembly by directly activating receptor GPR120.59, 60 Moreover, increased activation of SREBP1 promotes the resolution phase of TLR4-induced gene activation by reprogramming macrophage lipid metabolism.61 Second, greater tryptophan catabolism increases the potentially immune repressive kynurenine-dependent xenobiotic pathways and sustained de novo production of NAD+ from quinolinic acid.36 These consequences of tryptophan catabolism might prolong tolerance by activating RelB62 or maintaining activation of NAD+-dependent SIRT family members.22 NAD+ support of the SIRT family homeostasis axis includes glycolysis repression by SIRT6,21 and fatty acid metabolism and autophagy activation by SIRT1.63 Third, increased catabolism of methionine and transsulfuration with cysteine support of antioxidant GSH synthesis could promote redox balance by increasing the antioxidative arm of redox controls.64, 65 The serine synthesis and one-carbon metabolism pathway to glycine and methionine catabolism lie proximal to transsulfuration and control transmethylation of proteins by increasing SAM.66 SAM is the rate-limiting step in transmethylation of reversible heterochromatin silencing by RelB and SIRT1.65-68 Methionine transsulfuration to homocysteine and cysteine cycling also controls cystathionine-dependent H2S decreases in ATP synthesis by inhibiting cytochrome C oxidase.69, 70 Fourth, breaks in the Krebs cycle at aconitase and SDH71 increased during deactivation and waned during resolution. During deactivation and tolerance, itaconate directly inhibits SDH, which would promote increases in succinate levels, disengage Complex II of the respiration chain, and limit mitochondrial ATP synthesis.72 Itaconate, a recent addition to Krebs cycle reprogramming, regulates expression of the master antioxidant transcription factor NRF2 by electrophilic binding to its nuclear repressor, KEAP.73 The poorly understood Krebs cycle and itaconate-driven bioenergy axis may play a pivotal role in controlling monocyte anabolism and catabolism by adjusting mitochondrial bioenergetics.74 The metabolic rewiring of itaconate also might control both immunity and organ regeneration during sepsis.

    Our study used cultures of undifferentiated human promonocyte THP-1 cells stimulated with high-dose LPS for up to 96 h, which we previously used to temporally track gene-specific chromatin modifications that drive immune resistance and tolerance phenotypes in the acute inflammatory response of sepsis.25 As examples, using the THP-1 cell model, we first identified the NFκB p65 and RelB transcription factor axis,19 the SIRT1 and six chromatin-directed glucose and fatty acid substrate fueling axis,21 and the PDK and PDC mitochondrial aerobic glycolysis and glucose carbon-based oxidative phosphorylation axis as essential contributors to sepsis immunometabolic and bioenergy reprogramming.17 These resistance and tolerance axes were translated to both normal and sepsis human blood monocytes and to normal mouse splenic macrophages and septic splenocytes from C57BL/6J adult mice with 50–60% mortality from our septic shock model of cecal ligation and puncture. Both nuclear SIRT1 and mitochondrial PDK targeting in septic mice reversed tolerance and improved survival.18, 24

    Despite this body of evidence, a major limitation of the current study is that its network rewiring findings are not yet translated into cultures of primary human monocytes or acute inflammatory responses in vivo in animal or human models. Additional limitations of the study are (1) no assessment of sequential metabolic rewiring before 8 h, thereby missing the initial launching of anabolic effector immune and excessive nitrogen and oxygen-derived free radical production; (2) no tracking of carbon flux of specific pathways by stable isotope radiolabeling; (3) no in-depth assessment of lipidomics (which precluded measuring resolvins) or a combination of metabolomics signatures with concomitant shifts in proteomics and epigenomics; and (4) lack of concomitantly tracked changes in surface markers (e.g., human leukocyte antigen-DR isotype, CD80/86, or programmed cell death protein 1) or physiologic responses such as phagocytosis in parallel with the sequential metabolic shifts.

    In summary, we provide new information about how monocytes sequentially rewire metabolism and bioenergetics as the acute inflammatory response progresses from anabolic activation, catabolic deactivation, to early resolution (see our concluding Supplemental Fig. 6S). Our model may inform metabolism-based molecular targeting sites to promote anabolic and catabolic homeostasis retrieval and ultimately improve survival rates among patients with sepsis.

    AUTHORSHIP

    X.Z., V.V., and C.E.M. conceived the study, interpreted data, and wrote the manuscript; B.I. helped design the study, manage data, and integrate informatics and biostatistics results; A.M. and B.Y. managed data and helped write the manuscript; D.L. conducted experiments and aligned data, generated graphics, and supported data interpretation; and T.L. helped design experiments and interpret the data. Dr. Zhu and Dr. McCall contribute equally to this work. Dr. Zhu is the primary corresponding author for this article.

      ACKNOWLEDGMENTS

      We acknowledge the editorial assistance of Karen Klein, MA in the Wake Forest Clinical and Translational Science Institute (UL1 TR001420; P.I.: McClain) and Peter Stacpoole, MD, Ph.D., (University of Florida). The research was supported by NIH grants R01 5R01AI065791, 1R01GM102497, 2R56 AI065791, and R35 GM126922 to C.E.M.; 1R01 GM099807-01A1 to V.V.; and R01 HL132035 to X.Z.

        DISCLOSURES

        The authors declare no conflicts of interest according to the academic institutions and NIH guidelines.

        DATA SHARING

        Metabolomics data including more in-depth method and statistical management are available on request, and the complete heat map of metabolites is available on request.