IL-27-Ucp2-FoxO3 axis mediating the polarization of alternatively activated macrophages and ameliorating inflammatory pain

Abstract

Background

The signaling pathways of inflammatory pain are widely explored, but practical clinical approaches to ameliorate pain remain inadequate.

Methods

Quantitative PCR (qPCR) and ELISA methods were applied to measure the concentration of interleukin (IL)-27 in the inflammatory pain mouse model. Flow cytometry was conducted to identify the source of IL-27. Bone marrow-derived macrophages were stimulated by IL-27, IL-4, lipopolysaccharide, and/or interferon-gamma, followed by qPCR to assess pro-inflammatory and pro-resolving markers’ dynamic expression. Then, the molecule profiling of IL-27-primed macrophages was determined using transcriptomic and proteomic sequencing. The Agilent Seahorse XF analyzer calculated energy metabolism indicators. The adoptive cell transfer method was used to verify that forkhead box class O3 (FoxO3) mediates alternatively activated macrophage differentiation induced by IL-27-Ucp2, contributing to alleviating pain sensation in mice.

Results

IL-27 is highly expressed centrally and peripherally in rodent pain models. Selective downregulation of IL-27 intensifies pain sensitivity in mice. In macrophages, IL-27 promotes the secretion of anti-inflammatory molecules, such as Arginase-1. Further, transcriptome, energy metabolic examination, and proteome analyses identified that IL-27 restructures the metabolism in macrophages, which is mediated by uncoupling protein 2 (Ucp2) and subsequently activates transcription factor FoxO3. Conditional knockdown of FoxO3 (si-FoxO3) in macrophages refrains the production of anti-inflammatory genes in vitro; meanwhile, adoptive transfer of macrophages with si-FoxO3 but no wild-type macrophages (or IL-27 primed macrophages) prolongs mechanical hyperalgesia in mice.

Conclusions

These findings reveal that the IL-27-Ucp2-FoxO3 axis regulates macrophage plasticity distinct from the canonical IL-4-mediated pathway through metabolic rewiring and facilitates alleviating Inflammatory pain.

INTRODUCTION

Inflammation is integral to protecting organisms from pathogens [1,2]. After tissue injury, inflammation triggers the sequential recruitment of immune cells (such as neutrophils, macrophages, and mast cells) into the damaged tissue. These immune cells release pro-inflammatory cytokines (interleukin-1β [IL-1β], tumor necrosis factor-α [TNF-α], and bradykinin), which induce inflammatory pain by sensitizing and activating sensory neurons [3,4]. Extensive pioneering studies have described the initiation and escalation of inflammation, but the rationale underlying the resolution of inflammation is less clearly defined.

Macrophages, characterized by mighty plasticity, are crucial in tissue homeostasis, contain inflammation, and host defense [5,6]. Many experiments have strengthened the theory that macrophages actively participate in the pathological regulatory network of inflammatory pain [4,7]. The work from Niels Eijkelkamp’s team showed that complete Freund’s adjuvant (CFA)-induced mechanical hyperalgesia lasted for 12 days in mice with depleted macrophages and monocytes (microglia is unaffected) compared with normal wild type (WT) mice (CFA-induced inflammatory pain usually was resolved at around 5–6 days), which highlights the necessity and feasibility of manipulating macrophages to deliver pro-resolving properties in attenuated pain [8]. The tissue microenvironment profoundly shapes the characteristics of macrophages [9]. The peripheral nervous system imparts specialized traits to resident macrophages, differentiating them from microglia outside the nervous system [3]. Following nerve injury, monocyte-derived macrophages infiltrate the nervous tissue and polarize it into an M1-like (or classic) phenotype by nociceptors, accumulating and producing high amounts of inducible nitric oxide synthase, TNF-α, and IL-1β in the dorsal root ganglion (DRG). This bioprocess is essential for initiating and maintaining pain during the early stages of inflammation [10,11]. However, some anti-inflammatory cytokines have been identified to contain the inflammatory cascade. IL-4 has been known to promote the differentiation of an alternatively activated macrophage (AAM) phenotype, which exhibits a distinctive protein expression profile, such as Arginase-1 (Arg-1), resistin-like molecule-alpha (Retnla). AAMs fulfill their role in facilitating the resolution of inflammation and tissue healing during later phases of various types of infection [12,13]. Although the philosophy of AAM has garnered colossal attention, the exact nature of the signaling pathways that regulate the polarization of such pro-resolving and reparative macrophages is still elusive.

IL-27, an essential immunoregulator in the IL-12 family, is primarily produced by antigen-presenting cells (APCs), such as dendritic cells, monocytes, and macrophages [14]. IL-27 comprises IL-27 p28 (unique to IL-27) and Epstein-Barr virus-induced gene 3 protein (EBi3, shared with IL-35) subunits. IL-27 influences multiple immune cells via the IL-27 receptor, composed of Wsx-1 (unique to IL-27) and Gp130 (shared with IL-35), and mainly the downstream JAK-STAT signaling pathways [15]. The bio-function of IL-27 is documented to have pro-inflammatory and anti-inflammatory properties in a contextualized milieu [15,16]. IL-27 promotes the differentiation of naïve CD4+ T cells into Th1 cells by enhancing interferon-gamma (IFN-γ) production to kill viruses and certain bacteria [17,18]. Meanwhile, evidence also supports that up-expressed IL-27 dampens the inflammatory response in some autoimmune conditions, such as rheumatoid arthritis and multiple sclerosis, by curbing Th17 differentiation to inhibit IL-17 and indirectly increase IL-10 production [19]. In addition to its role in mediating immunity, IL-27 modulates the differentiation of hematopoietic stem cells toward myeloid progenitors, suggesting its involvement in regulating macrophage maturation and phenotype [20]. Several studies have described IL-27 as having a positive role in alleviating pain in mice [21]. However, the detailed mechanisms are not yet clearly formulated. Therefore, exploring how IL-27 regulates macrophage polarization and its biological consequences in inflammatory pain is of great interest.

In this investigation, the authors aimed to elucidate the molecular basis for the pro- and/or anti-inflammatory roles of IL-27-macrophage pairs (Fig. 1). They focused on identifying the underpinned signaling hubs that govern macrophage plasticity in inflammatory pain. The molecular profiling indicated that IL-27 skews macrophage polarization toward AAM by rewiring mitochondrial metabolism, leading to uncoupling protein 2 (Ucp2)-mediated uncoupling of the mitochondrial respiratory chain and enhancing the expression of forkhead box class O3 (FoxO3). The IL-27-Ucp2-FoxO3 axis induced AAM, contributing to inflammation resolution and pain relief.

Fig. 1.

A schematic overview presents the framework of this study. AAMs: alternatively activated macrophage, FoxO3: forkhead box class O3, IL: interleukin.

MATERIALS AND METHODS

1. Animals

All experiments were performed under international guidelines and approved by the Animal Experimental Ethical Inspection of The First Affiliated Hospital, Zhejiang University School of Medicine (license number: 20221036). The authors euthanized 162 mice for the in vivo experiments and 248 mice for the in vitro (cell-based) experiments in this study. They are deeply grateful for the contribution of the laboratory animals to biomedical research and respectfully acknowledge their role in this work.

Adult (8–10 weeks) male C57BL/6J mice (26 ± 2 g) were used and maintained in The First Affiliated Hospital of Zhejiang University animal facility. Mice were housed in groups under a 12 hour/12 hour light/dark cycle, with food and water available ad libitum. The cages contained tissue paper and shelter. To minimize bias, animals were randomly assigned to the different groups before the start of the experiment, and experimenters were blinded to the treatments. Adult male mice were used for the behavioral and biochemical studies. To establish the inflammatory pain models, mice received an intraplantar injection (i.pl.) of 5 μL λ-carrageenan (λ-carr) (1% w/v, Sigma-Aldrich) in the hind paws, as described previously [20].

2. Pain behavioral tests

Before the experiments, mice were habituated to the testing environment by being placed in the experimental setting at least 3 times for 1 hour, 1 week before the start of the experiments. Behavioral assays were conducted on the same day for each experiment. A minimum of three baseline measurements were taken on different days 1 week before the start of the experiment, with the final baseline measure occurring on the day of the experiment’s initiation. All experiments were conducted in the same room and with consistent experimental setups.

The von Frey test (Stoelting) assessed mechanical thresholds in both hind paws, employing the up-and-down technique to determine the 50% withdrawal threshold [22]. Each filament was applied to the plantar surface for 5 seconds. A higher-force filament was applied if no response was observed after the initial filament (0.4 g). If a response was elicited, a lower-force filament was subsequently used. A minimum of 30 seconds was allowed between applications of the filaments. After the first reversal in response direction, four measurements were obtained. Filaments were applied perpendicularly without horizontal movement. In case of an ambiguous response, the experimenter waited one minute and repeated the test.

3. Reagent administration

To determine the causal relationship between IL-27 and mechanical allodynia, the authors applied a loss/gain-of-function strategy via gene manipulation using a pAAV vector. The pAAV2/9-U6-shRNA (IL-27p28)-CMV-EGFP targeted to knock down IL-27 (pAAV-shIL-27) and pAAV2/9-U6-shRNA (scrambled)-CMV-EGFP (pAAV-scrambled shRNA) was synthesized by OBiO Technology Corp., Ltd. Then, an intrathecal injection (i.t.) was performed with a 30-gauge needle between the L5 and L6 levels to deliver pAAV-shIL-27 and/or pAAV-scrambled shRNA (10 μL) to the cerebrospinal fluid of mice. The forced expression of IL-27 in mice was achieved by injecting recombinant mouse IL-27 protein solution (228a.a HEK293, His, HY-P73200, MCE) at 100 ng/kg via the tail vein injection (i.v.). The elevation of IL-27 was confirmed in the serum of mice after 3 hours of i.v. injection.

4. The concentration of IL-27 readout

After establishing the Inflammatory pain model, the authors measured the level of IL-27 in the brain, spinal cord (L3–L5), DRG (L3–L5), serum, and spleen tissues, respectively. The mice’s blood was obtained by cardiac puncture. Freshly drawn whole blood (500 μL) was kept at room temperature (30 minute-1 hour) to facilitate clotting, followed by centrifugation at 1,000 × g for 10 minutes at 4°C. The serum was separated and either used immediately for ELISA or stored at -80°C for subsequent analysis. The concentration of IL-27p28 in the mouse serum was measured via a mouse IL-27p28 ELISA kit (ab171333, Abcam) according to the manufacturer’s instructions.

Animals were euthanized with 1% pentasorbital sodium (80 mg/kg, i.p.) and underwent intracardiac perfusion with chilled phosphate-buffered saline (PBS) buffer. Subsequently, the harvested spleen, brain, spinal cord (L3–L5), and bilateral DRG (L3–L5) were snap-frozen and stored in a –80°C freezer or immediately prepared for RNA (IL-27p28 and Ebi3) analysis.

5. BMDM culture

The bone marrow-derived macrophage (BMDM) was used in all experiments [8]. Bone marrow was extracted by flushing the Dulbecco’s Modified Eagle Medium (DMEM) medium through the mice’s femur and tibia with a 25-gauge needle. The marrow cells were then exposed to red blood cell lysis buffer (eBioscience) for 5 minutes at room temperature, centrifuged, resuspended, and passed through a 70 µm cell strainer (Falcon). The resulting cell suspension was plated in 10 cm Petri dishes containing DMEM medium with L-glutamine (Gibco), 10% fetal bovine serum (Gibco), 1% penicillin-streptomycin (Beyotime), and 10% L929 supernatant, and cultured for 6 days in a 37°C incubator with 5% CO₂. On day 7, cells were harvested, resuspended in complete DMEM, and seeded into multi-well plates at 5 × 10⁵ or 1 × 10⁶ cells/mL. For the in vitro experiment, this study stimulated macrophages with 100 ng/mL IL-27, 100 ng/mL lipopolysaccharide (LPS), 20 ng/mL IL-4, 100 mM Genipin, and 10 ng/mL IFN-γ for specific purposes.

6. Microglia and astrocytes culture

Neonatal C57BL/6J mice at postnatal day 1 were euthanized for the dissection of their cerebral cortices [23,24]. The meninges were meticulously removed, and the tissue was enzymatically dissociated using 0.25% trypsin with ethylenediaminetetraacetic acid (EDTA) solution B (Beyotime), followed by mechanical dispersion into a single-cell suspension. Cells obtained from each hemisphere, called “mixed glia,” were subsequently plated in 60 mm culture dishes and maintained at 37°C in a humidified incubator with 5% CO₂ and 95% air. The culture medium consisted of DMEM (Gibco) supplemented with 10% FBS (Gibco) and penicillin-streptomycin-amphotericin B solution (Beyotime), with medium changes performed every 3 days. After reaching confluence within 7–10 days, the mixed glial cultures were utilized for microglial isolation between days 12 and 14 post-culture initiation. A mild trypsinization approach was employed, wherein treatment with 0.06% trypsin-EDTA solution B (Beyotime) led to the detachment of an intact astrocyte-rich layer, yielding a highly purified microglial population. Microglia and astrocytes are cultured in vitro for the following experiments.

7. Real-time PCR

Total RNA from the harvested tissues was isolated per the manufacturer’s protocol using the QIAwave RNA Mini Kit (QIAGEN). Five hundred ng of obtained RNA was reverse-transcribed into cDNA (PrimeScriptTM RT reagent Kit). Quantitative real-time PCR reactions (TB Green^® Premix Ex TaqTM II) were performed using Applied Biosystems QuantStudio 5 (ThermoFisher). mRNA amounts are presented as relative expression normalized to GAPDH expression and the untreated group. The sequences of primers used are summarized in the Table 1. The efficiency of the primers was assessed using the standard curve and applied to the relative gene expression calculation based on the real-time PCR threshold values for different transcripts.

Table 1.

The sequences of primers, siRNA, and pAAV

Target	Forward	Reverse
siRNA
Si-Wsx-1	GACCGUCUGUGAAACUUCUTT	AGAAGUUUCACAGACGGUCTT
Si-Ucp2	GUGGUCAAGACGAGAUACATT	UGUAUCUCGUCUUGACCACTT
Si-Irg1	GACGUCCAGUACGUAAACATT	UGUUUACGUACUGGACGUCTT
Si-FoxO3	GGAGUUUGGUCAAUCAGAATT	UUCUGAUUGACCAAACUCCTT
Virus
pAAV2/9	TargetSeq	GATACCATCTTCCCAATGTTT
Vector (NC)	TargetSeq	CCTAAGGTTAAGTCGCCCTCG
M1 marker
IL-6	AACCGCTATGAAGTTCCTCTCTG	TGGTATCCTCTGTGAAGTCTCCT
IL-1β	CAGCACATCAACAAGAGCTTCAG	GAGGATGGGCTCTTCTTCAAAGA
TNF-α	GCCTCCCTCTCATCAGTTCTATG	ACCTGGGAGTAGACAAGGTACAA
M2 marker
Arg-1	TCTGCCAAAGACATCGTGTACAT	CGACATCAAAGCTCAGGTGAATC
Chi3l3	GAAGCTCTCCAGAAGCAATCCTG	TCCCTTCTATTGGCCTGTCCT
Retnla	CCTCCACTGTAACGAAGACTCTC	CTCCCAAGATCCACAGGCAAAG
Reference
GAPDH	CAGTGGCAAAGTGGAGATTGTTG	TCGCTCCTGGAAGATGGTGAT
β-actin	ACTGTCGAGTCGCGTCC	CTGACCCATTCCCACCATCA
IL-27p28	TCTTCCCAATGTTTCCCTGACTT	GTCCTCCTCCTCCTTTGAACATT
Ebi3	CCGGACATCTTCTCTCTCAAGTA	TTTCGAGTTCCTGAGGGTGAAAG
Wsx-1	CCTTCCAGACGCCATTCTTAGAT	TGTAAGTCACCTGCACACAAGG
IL12p35	GACCAAACCAGCACATTGAAGAC	TGATTCTGAAGTGCTGCGTTGAT
Ucp1	GATTCATCAACTCTCTGCCAGGA	CTGTCTGGACTTCATCAGCTCTT
Ucp2	CTTCTCCCAATGTTGCCCGTAAT	CCCAAGGCAGAGTTCATGTATCT
Ucp3	CCTACGACATCATCAAGGAGAAGTT	TCCAAAGGCAGAGACAAAGTGA
Ucp4	GAATGCCTATCGCCGAGGA	AGTAGGAACTTGCTCGTCCGG
Ucp5	TCCCAACTGCTCAGCGTG	GGTGCTTCTTGGTAATATCATAAACG
FoxO3	GGGGAACCTGTCCTATGCC	TCATTCTGAACGCGCATGAAG

8. Western blot analysis

After cells were pooled and lysed in RIPA buffer (Beyotime) containing a Protease inhibitor and Phosphatase inhibitor cocktail, the total protein samples (20–25 μg) were loaded and separated by 4%–10% SDS-PAGE and subsequently transferred onto a polyvinylidene difluoride membrane (0.22 μm PVDF, Millipore) for immunoblotting with the specific antibodies listed in the Table 2.

Table 2.

The details of the reagents in this experiment

Reagent	Source	Identifier
Recombinant Mouse IL-27	Biolegend	Cat# 577404-25μg
Recombinant Mouse IL-4	Biolegend	Cat# 574304-25μg
IL-27 Protein, Mouse	MedChemExpress	HY-P73200
Mouse IL-27 p28 ELISA Kit	Abcam	ab171333
Mouse Chi3l3 ELISA Kit	LS Bio	LS-F36207
Mouse Retnla ELISA Kit	LS Bio	LS-F66446
LPS	MedChemExpress	HY-D1056
Genipin	MedChemExpress	HY-17389
IFN-γ Recombinant Protein	PeproTech	Cat# 315-05-20μg
Lipofectamine 3000 reagent	Thermofisher	L3000015
APC Anti-Mouse Ly6C	Elabscience	E-AB-F1121E
CD11b Monoclonal Antibody (M1/70), Alexa Fluor 488	eBioscience	53-0112-80
CD11c Monoclonal Antibody (N418), PerCP-Cyanine 5.5	eBioscience	45-0114-80
Ly-6G Monoclonal Antibody (1A8-Ly6g), PE-eFluor 610	eBioscience	61-9668-80
F4/80 Monoclonal Antibody (BM8), eFluor 450	eBioscience	48-4801-80
IL-27 p28 Monoclonal Antibody (MM27-7B1), PE	eBioscience	12-7285-80
Zombie NIR Fixable Viability Kit	BioLegend	423105
Agilent Seahorse XF Cell Mito Stress Test Kits	Agilent Technologies	103015-100
Agilent Seahorse XF Glycolysis Stress Test Kit	Agilent Technologies	103020-100
Arginase-1 (D4E3M) Rabbit mAb	CST	#93668
β-Actin	ProteinTech	66009-1-Ig
GAPDH	ProteinTech	60004-1-Ig
FoxO3 (D19A7) Rabbit mAb	CST, US	#12829
Mouse Wsx-1 Antibody	R&D System	MAB21091
CoraLite488-conjugated Goat Anti-Rabbit IgG (H + L) ROS Fluorometric Assay Kit (Green)	ProteinTech Elabscience	SA00013-2 E-BC-K138-F

9. Immunofluorescence labeling

Cells were grown on glass coverslips and fixed for 20 minutes with 4% PFA (pH 7.4). After washing with PBS, cells were permeabilized using 0.2% Triton X-100 (Beyotime) in PBS for 20 minutes at room temperature, followed by blocking with 0.05% Tween 20 (Beyotime), 1% BSA (Beyotime), and 10% goat serum (Solarbio) in PBS for 1 hour at room temperature. Primary antibodies were stained in a blocking buffer at 4°C overnight. Secondary staining was executed at room temperature for 4 hours. The antibodies used in the experiments are listed in the Table 2.

10. siRNA-mediated knockdown

In the in vitro study, the authors selectively knocked down target genes (Wsx-1, Ucp2, Irg-1, FoxO3) using small interfering RNA (siRNA) tools. As stated in the manufacturer’s manual, under specific conditions, the Lipofectamine 3000 reagent (Thermofisher) was applied to transfect macrophages with si-Wsx-1, si-Ucp2, si-Irg-1, and si-FoxO3, respectively.

11. Flow cytometry

Preparation of single-cell suspension from mouse spleen: spleen tissues were obtained from C57/BL mice euthanized by carbon dioxide asphyxiation. The intact spleens were placed in Hank’s Balanced Salt Solution (Beyotime), and peripheral adipose tissue and membranes were carefully removed using ophthalmic forceps. The spleens were minced into small fragments and collected in a centrifuge tube. Subsequently, 1 mL of tissue dissociation solution (abs9482, Absin) was added, and the samples were incubated at 37°C while being shaken at 300 rpm for 30 minutes. The dissociated cell mixture was then passed through a 200-mesh filter to obtain a single-cell suspension of the spleen.

After washing with PBS, the spleen single-cell suspension was resuspended in cell staining buffer (BL1136A, Biosharp) and stained with fluorophore-conjugated antibodies (CD11b, CD11c, Ly6G, Ly6C, F4/80) on ice for 40 minutes. Following PBS wash, the cells were resuspended in cell staining buffer and fixed with fixation buffer (abs9936A, Absin) at room temperature in the dark for 15 minutes. The cells were then washed with PBS, resuspended in cell staining buffer, and permeabilized using a cell permeabilization kit (abs9936B, Absin) at room temperature in the dark for 15 minutes. After PBS wash and resuspension in cell staining buffer, PE-IL-27p28 antibody was added, and the cells were incubated on ice in the dark for 40 minutes. After washing, cells were resuspended in a cell staining buffer and examined using Cytek Aurora. The outcomes were analyzed using FlowJo (v.10.10.0) software.

12. BulkSeq

Total RNA from BMDM was isolated using the QIAwave RNA Mini Kit under the manufacturer’s guidelines. RNA purity and concentration were assessed with the NanoDrop 2000 spectrophotometer (Thermo Scientific), while RNA integrity was evaluated via the Agilent 2100 Bioanalyzer (Agilent Technologies). Library preparation was performed with the VAHTS Universal V6 RNA-seq Library Prep Kit following the manufacturer’s instructions. Sequencing was outsourced to OE Biotech Co., Ltd. and conducted on an Illumina Novaseq 6000 platform to generate 150 bp paired-end reads. Raw sequencing data in FASTQ format underwent initial quality control using fastp, with low-quality reads removed to produce high-quality clean reads. Reference genome and gene annotation files were obtained from NCBI, UCSC, or Ensembl. Clean reads were aligned to the mouse reference genome (GRCm39/mm10) via HISAT2 (v.2.0.5). Gene expression levels were quantified as FPKM, and read counts were determined using HTSeq-count (v.0.6.1). Principal component analysis (PCA) was performed in R (v.3.2.0) to evaluate sample biological replication.

Differential expression analysis of genes (DEGs) was conducted with DESeq2 (v1.46.0). P values were adjusted for multiple comparisons using Benjamini and Hochberg’s method to control the false discovery rate, with a significance threshold of Q value (adjusted P value) < 0.05. Hierarchical clustering of DEGs was carried out in R (v.3.2.0) to illustrate gene expression patterns across groups and samples. Data visualization was performed using the R packages ggplot2 and pheatmap.

13. Seahorse assay

Mitochondrial metabolic activity was monitored in real-time by measuring the oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) using the XFe96 extracellular flux analyzer (Agilent Technologies). BMDMs were plated at a density of (5 x 10⁴ to 1 x 10⁵) cells per well and either remained untreated or were exposed to IL-27 (100 ng/mL, 6 hours), IL-4 (20 ng/mL, 6 hours), or Genipin (100 mM, 24 hours). Experiments were carried out following the manufacturer’s instructions. Briefly, the Cell Mito Stress Test involved the sequential addition of 1 mM oligomycin, 1.5 mM carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP), and 0.5 mM rotenone/antimycin A (R/A). Likewise, the Cell Glyco Stress Test was performed by sequentially injecting 10 mM glucose, 1 mM oligomycin, and 50 mM 2DG. Proton leak was quantified by subtracting the lowest OCR value obtained after R/A treatment from the minimum OCR recorded following oligomycin addition. During the Seahorse Glycolysis Stress Test, glycolytic capacity was determined as the maximal ECAR measured after oligomycin injection and corrected by subtracting non-glycolytic acidification (baseline ECAR before glucose addition). OCR and ECAR values were normalized to cell number per well. Briefly, equal numbers of cells were seeded across conditions, and values were expressed as rate per 10³ cells after background correction. Reactive oxygen species (ROS) were assessed per the kit manufacturer’s guidelines (E-BC-K138-F, Elabscience).

14. Proteome

The collected samples were sonicated 3 times on ice using a high-intensity ultrasonic processor (Scientz) in a lysis buffer containing 8 M urea and 1% protease inhibitor cocktail (Sigma). Cellular debris was eliminated by centrifugation at 12,000 × g for 10 minutes at 4°C. The supernatant was then collected, and the protein concentration was measured using a BCA assay kit (Thermo) per the manufacturer’s guidelines. Subsequently, the protein solution was reduced with 5 mM dithiothreitol (Thermo) for 30 minutes at 56°C, alkylated with 11 mM iodoacetamide (Thermo) for 15 minutes at room temperature in the dark, and diluted with 100 mM NH₄HCO₃ to decrease the urea concentration to below 2 M. Trypsin (Promega) digestion was performed at a 1:50 trypsin-to-protein mass ratio overnight, followed by second digestion at a 1:100 ratio for 4 hours. The resulting tryptic peptides were dissolved in 0.1% formic acid (solvent A) and directly loaded onto a custom-packed reversed-phase analytical column (15 cm, 75 μm i.d.) containing 1.9 μm Reprosil-Pur C18 beads (Dr. Maisch). Chromatographic separation was achieved using a gradient from 3% to 8% solvent B (0.1% formic acid in 98% acetonitrile) over 3 minutes, increasing to 20% over 37 minutes, followed by a rise to 30% over 12 minutes, and then reaching 80% over 4 minutes with a final hold at 80% for 4 minutes at a constant flow rate of 300 nL/min on an UltiMate 3000 nanoLC system.

Peptides were ionized using a nano-spray ionization source and analyzed via tandem mass spectrometry (MS/MS) on an Orbitrap Exploris 480 (Thermo Fisher Scientific) coupled to the UPLC system. An electrospray voltage of 2.0 kV was applied. Full scans were acquired in the m/z range of 400 to 1,200, and intact peptides were detected in the Orbitrap at a resolution of 60,000. Peptides were selected for MS/MS analysis with an NCE setting of 27, and fragment ions were detected in the Orbitrap at a resolution of 15,000. A data-dependent acquisition method alternated between one MS scan and 20 MS/MS scans with a dynamic exclusion of 30 seconds. Automatic gain control was set to 5E4, and FAIMS compensation voltages were adjusted to –45 V and –65 V. The obtained MS/MS spectra were processed using Proteome Discoverer v2.5 (Thermo Fisher), and tandem mass spectra were searched against the SwissProt Mouse database (17,089 sequences) using the SEQUEST algorithm.

15. Transcription factor enrichment analysis

The analysis of transcriptional interactions was conducted using the TRANSFAC database [25] (version 2024/2) and the JASPAR platform [24] (version 2024/10), both of which systematically curated and annotated experimentally confirmed interactions derived from the literature. Transcription factor–gene associations identified in mouse models were deemed eligible for assessment. Transcription factors linked to a predefined target gene set were retrieved, and their frequency of occurrence (combined score) was computed and ranked accordingly.

16. Adoptive transfer of macrophages

As described, cells were injected intrathecally (i.t. 2 × 10⁵ cells/5 μL per mouse) under isoflurane anesthesia [22]. For experiments, 2 × 10⁶ macrophages were treated in various manners (wild-type macrophage, macrophage + si-FoxO3, IL-27-primed macrophage, IL-27-macrophage + si-FoxO3) in 400 mL complete medium for 30 minutes at 37°C, followed by three washes before i.t. injection of 2 × 10⁵ cells.

17. Statistical analysis

All data are shown as mean ± standard error of the mean and were analyzed using GraphPad Prism (version 9.5.0). Data were assessed for normality using the Shapiro–Wilk test. When the normality assumption was met, the statistical analyses involved unpaired two-tailed t-tests, one-way/two-way ANOVA, and post hoc testing (Tukey’s/Dunnett’s multiple comparisons test). A P value of less than 0.05 indicated statistical significance, with significance marked with P < 0.05; P < 0.01; P < 0.001.

RESULTS

1. The increased production of IL-27 in mice with inflammatory pain

To establish an inflammatory pain model, an i.pl. injection of 1% λ-carr was made into the left hind paw of the mice. Signs of pain hypersensitivity (allodynia and hyperalgesia) were examined in the mice using the von Frey test to confirm the building of an inflammatory pain model. The mechanical allodynia threshold in the mouse pain model was significantly lower than in normal and saline-treated groups (Fig. 2A), and λ-carr-induced hyperalgesia was relieved within 4 to 6 days.

Fig. 2.

Establishing an Inflammatory pain model and assessing the expression profiles of IL-27. (A) Course of mechanical hyperalgesia in the mouse pain model (n = 5), two-way ANOVA tested P values (vs. NC) with Dunnett’s multiple comparisons test. (B–G) The dynamic concentration of IL-27 (IL-27p28 and Ebi3) in the brain, spinal cord (L3–L5), spleen, ipsilateral and contralateral DRG (tested by qPCR), and the ELISA method to measure IL-27p28 in the serum. N = 4, GAPDH was used as an internal reference. (H, I) The mRNA level of Wsx-1 in the spinal cord (L3–L5) and ipsilateral DRG tissues. N = 4, GAPDH was used as an internal reference. (J, K) The mRNA level of IL-12p35 (IL-35) in the spinal cord (L3–L5) and ipsilateral DRG tissues. N = 4, GAPDH was used as an internal reference. The P value (vs. NC) was measured using one-way ANOVA with Dunnett’s post hoc test. *P < 0.05, **P < 0.01, ***P < 0.001. Data are shown as mean ± standard error of the mean. IL: interleukin, Ebi3: Epstein-Barr virus-induced gene 3 protein, Ips-DRG: ipsilateral-dorsal root ganglion, qPCR: quantitative PCR, λ-carr: λ-carrageenan.

Based on the clinically relevant inflammatory pain model, the concentration of IL-27 were measured in the brain, spinal cord (L3–L5), DRG (L3–L5), serum, and spleen tissues, respectively. mRNA readouts indicated that relative to the pre-treated group, IL-27p28 in the brain tissue did not significantly change over time (Fig. 2B). The production of IL-27p28 was increased within 12 hours in the spinal cord (L3–L5) with a peak level of around 1 day and then gradually declined over the next 5 days (Fig. 2C). The ELISA method showed a high expression of IL-27p28 in the serum (Fig. 2D). Even in spleen tissue, the authors also detected the elevated production of IL-27p28 (Fig. 2E). The expression of IL-27p28 was elevated within 12 hours and peaked in 1 day in the ipsilateral DRG (L3–L5, ips-DRG) but with no significant change in the contralateral DRG (L3–L5, contra-DRG) (Fig. 2F, G). The increased pattern of Ebi3 is similar to IL-27p28, but its induction persisted longer than IL-27p28 in the spinal cord and ips-DRG. The authors then found the upregulated Wsx-1, a specific receptor for IL-27, in the spinal cord and ips-DRG (Fig. 2H, I). Since IL-27 and IL-35 share the Ebi3 subunit [13], they next tested the gene expression of IL-12p35 at the mRNA level in the spinal cord and ips-DRG. The results suggest that the mRNA level of IL-12p35 did not significantly change (Fig. 2J, K). Therefore, the involvement of IL-35 cytokine was ruled out in this setting. In Inflammatory pain, IL-27p28 and its receptor (Wsx-1) are elevated in the spinal cord and ips-DRG.

2. The source of IL-27

The high production of IL-27 centrally and peripherally revealed that the IL-27 gene is probably secreted from multiple sources. Based on the pioneering study, IL-27 cytokines are chiefly synthesized and released by APCs, including dendritic cells, monocytes, macrophages, etc. [16]. Therefore, a flow cytometry method was conducted to identify the source of IL-27 protein in peripheral blood and spleen. As for the high expression of IL-27 in the spinal cord (L3–L5), the authors cultured the primary microglia and astrocytes in vitro and exposed them to LPS/IFN-γ, followed by an ELISA test to determine the concentration of IL-27p28 in the cultured medium. The results indicated that monocytes (CD11b⁺Ly6C⁺) and neutrophils (CD11b+Ly6G+) secrete plenty of IL-27 into the peripheral blood and tissues (Fig. 3); meanwhile, IL-27 is primarily released by microglia in the spinal cord during the early stage of inflammation (Fig. 4).

Fig. 3.

The identification of the source of IL-27. (A–D) The intracellular co-label staining by flow cytometry was applied to determine the source of IL-27. The four APCs were labeled by dendritic cell (CD11b, CD11c, B), neutrophil (CD11b, Ly6G, C), monocyte (CD11b, Ly6C, A), and macrophage (CD11b, F4/80, D), respectively. IL-27p28 antibody labeled the IL-27. The results indicated that neutrophil/monocyte-derived IL-27 was highly expressed in serum and spleen. IL: interleukin, APCs: antigen-presenting cells, SSC-A: side scatter area.

Fig. 4.

The source of IL-27 in the spinal cord is microglia. (A, B) The cultured primary microglia secret loads of IL-27 in response to LPS (100 ng/mL) insult for 12 hours, while cultured primary astrocytes released minimal IL-27 in the same condition (n = 3). (C, D) IFN-γ-stimulated (20 ng/mL) primary microglia were the main source of IL-27, compared with the astrocytes group (n = 3), which suggests IL-27 was primarily secreted from microglia in the spinal cord. IL: interleukin, LPS: lipopolysaccharide, IFN: interferon.

3. Inflammation induces the elevation of IL-27 protein

Then, the authors aimed to unveil the causal association between inflammation and the upregulation of IL-27 using a gain/loss-of-function strategy. The pAAV2/9-U6-shRNA (IL-27p28)-CMV-EGFP targeted to knock down IL-27 (pAAV-shIL-27) and pAAV2/9-U6-shRNA (scrambled)-CMV-EGFP (pAAV-scrambled shRNA) were synthesized and injected intrathecally to downregulate IL-27 gene expression in the spinal cord and DRG in the mice (Fig. 5A). After confirming the best effect of pAAV-shIL-27, three groups (non-control, λ-carr-treated, shIL-27-λ-carr) were designed to evaluate mechanical hyperalgesia in mice using the von Frey fibers (Fig. 5B, C). Data showed that the mice in the shIL-27-λ-carr group presented the most significant allodynia compared to the other two groups (Fig. 5E). The forced elevation of IL-27 concentration in the mice was obtained by i.v. injection of recombinant mouse IL-27 cytokine (rIL-27, MCE) (Fig. 5D). Four groups (non-control, λ-carr-treated, shIL-27-λ-carr, and λ-carr-rIL27) were set to measure mechanical pain threshold using the von Fray test. The results suggest that λ-carr-induced mechanical hypersensitivity is significantly alleviated, followed by injecting the rIL-27 agent (100 ng/kg) in mice (3 hours after λ-carr injection in the mice) (Fig. 5F). Therefore, it was inferred that IL-27 may be a protective cytokine, and upregulation of IL-27 contributes to mitigating inflammation in pain.

Fig. 5.

The knockdown of IL-27 intensified mechanical allodynia in mouse pain models. (A) Schematic diagram showing construction of pAAV2/9-U6-shRNA (IL-27p28)-CMV-EGFP vector. (B) The Sh-1 presents the most effective knockdown of IL-27 at the mRNA level in mouse DRG (L3–L5) tissue and was selected for use in the next operation (n = 3). One-way ANOVA with Tukey’s multiple comparisons test was applied. (C) The experiment procedures for i.t. injection of pAAV2/9-U6-shRNA (IL-27p28) to knock down IL-27 and behavior test in mice. (D) The timelines of mechanical hyperalgesia in WT, λ-carr, and sh-IL-27 mice after tail-vein injection of recombination mouse IL-27 in a hind paw (n = 5). P values (vs. λ-carr) were tested by two-way ANOVA with Tukey’s multiple comparisons test. (E, F) The timelines of mechanical hyperalgesia in mice with WT, sh-IL-27, IL-27 forced expression (IL-27 FE)-treated groups. Mice received the rIL-27 agent (100 ng/kg) at the indicated time (n = 5). P values (vs. λ-carr) were tested by two-way ANOVA with Tukey’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001. Data are shown as mean ± standard error of the mean. IL: interleukin, DRG: dorsal root ganglion, i.t.: intrathecal injection, WT: wild type, λ-carr: λ-carrageenan.

4. IL-27 polarizes macrophages into the AAM phenotype

In vitro, BMDM were exposed to IL-27 to study the signaling pathways underlying the IL-27-BMDM interaction (Fig. 6A). Initially, the authors tested the phenotype switch of BMDM by the M1 marker (IL-6, IL-1β, TNF-α) and M2 marker (Arg-1, Chi3l3, Retnla) after IL-27-stimulated macrophages [26]. IL-4 primed macrophages, a canonical AAM-inducer cytokine, were used as a positive control. The mRNA analysis revealed that M2 markers are highly expressed when macrophages respond to IL-27 and IL-4 insults (Fig. 6B–D); meanwhile, M1 markers are minimally produced and hardly changed under the IL-27-treated group compared to the LPS and/or IFN-γ insult groups (Fig. 6E–G). However, IL-27 and IL-4 induce AAM and upregulate M2 markers following different kinetics. IL-27-stimulated M2 markers increased and peaked at about 6–12 hours (early phase) and then declined over time, while IL-4-induced M2 markers were elevated gradually until 24 hours (later phase). The expression of Arg-1 at the protein level was examined, and the results suggest a high production of Arg-1 in macrophages after IL-27 and/or IL-4 priming, which follows the same kinetics pattern as observed with mRNA (Fig. 6H, I). Notably, the level of IL-27 also rises in the spinal cord; therefore, the authors examined whether IL-27 induces microglia toward an alternative activation phenotype. The Arg-1, M2 marker, did not change significantly after IL-27-primed microglia (Fig. 6J). Thus, IL-27 polarizes macrophages, not microglia, into an alternatively activated phenotype.

Fig. 6.

IL-27 induces the differentiation of AAM. (A) The picture shows the procedures of this part. (B–D) Quantitative real-time mRNA levels of M2 indicators, including Arg-1 (B), Chi3l3 (C), Retnla (D), when BMDMs were exposed to IL-27 (100 ng/mL for 24 hours) and IL-4 (20 ng/mL for 24 hours) (n = 3). GAPDH was regarded as a reference. The P value was compared with the IL-27-and IL-4-treated group, using two-way ANOVA followed by Tukey’s multiple comparisons test. (E–G) qRT-PCR levels of M1 indicators, including IL-1β, IL-6, TNF-α, when BMDMs were exposed to LPS (20 ng/mL for 24 hours), IL-27 (100 ng/mL for 24 hours), or LPS (20 ng/mL for 24 hours) + IFN-γ (20 ng/mL for 24 hours) (n = 3). (H, I) IL-27 (H) and IL-4 (I) promote the expression of Arg-1 at the protein level in BMDM with different patterns (n = 3). GAPDH was regarded as a reference. The P value was compared with the NC groups, using one-way ANOVA followed by Dunnett post-test. (J) IL-27 has no significant impact on the expression of Arg-1 when IL-27 stimulates microglia in vitro, which indicates that IL-27 doesn’t induce the phenotype switch of microglia (n = 3). ACTB was regarded as a reference. The P value was compared with the NC groups, using one-way ANOVA followed by Dunnett post-test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Data are shown as mean ± standard error of the mean. IL: interleukin, LPS: lipopolysaccharide, IFN: interferon, TNF: tumor necrosis factor, AAM: alternatively activated macrophage, BMDM: bone marrow-derived macrophage.

5. IL-27 and IL-4 induce distinct AAM phenotype

To identify the specific IL-27-mediate molecule profiling that promotes macrophage polarization toward a pro-resolving AAM phenotype, an mRNA expression analysis was conducted in IL-27 and IL-4-induced AAM by bulk RNA sequencing (the detailed readouts of RNA-seq are deposited in GEO platform with GSE285607) (Fig. 7A). A total of 977 differential expressed genes (DEGs, 541 upregulated genes, and 436 downregulated genes) were identified in IL-4 treated groups and 1,387 DEGs (with 670 higher expression genes and 717 downregulated genes) in IL-27-stimulated groups. This approach indicated that IL-27 and IL-4 induced a small portion of overlapping (only 340 shared DEGs in both IL-27 [roughly 24.5% of the whole DEGs] and IL-4-triggered [approximately 34.8% of the entire DEGs] DEGs in BMDMs), but mostly distinct gene expression in the macrophages (Fig. 7B, C, E). The comparative analysis found that the two cytokines triggered the upregulation of genes encoding M2 markers, including Arg-1 and Retnla (Supplementary Fig. 1). Further, the authors selectively knocked down Wsx-1, IL-27-specific receptor, using the si-Wsx-1 tool (Fig. 7D) and accordingly inhibited the IL-27-induced AAM differentiation but did not impair the M2 marker triggered by IL-4 (Fig. 7F–I). These results reveal that IL-27 and IL-4 organized different receptors and signaling networks to imprint an AAM phenotype.

Fig. 7.

IL-27 distinctly induces the polarization of AAM from IL-4. (A) The picture of the study pipeline in this part. (B, C, E) Bulk RNA-seq comparing the BMDM treated by NC, IL-27 (100 ng/mL for 12 hours), and IL-4 (20 ng/mL for 24 hours). Data containing the heatmap presenting differential gene expression (B), PCA (C), and a Venn plot to show the shared and non-shared DEGs (E). (D) The knockdown effect of the three candidates’ siRNA was targeted at Wsx-1 at the protein level (GAPDH as a reference). (F) The decreased expression of Arg-1 in the si-Wsx-1 group when BMDM was insulted by IL-27, compared to the IL-4-treated and WT group (n = 3). β-actin as internal reference, P values (WT vs. si-Wsx-1 specifically under IL-27 stimulation) were tested by two-way ANOVA with Sidak’s multiple comparisons test. (G) The mRNA of Arg-1 was inhibited in the IL-27-si-Wsx-1 group, compared to the IL-4-treated and WT group (n = 4). GAPDH was regarded as a reference. (H, I) Compared to the WT and NC groups, the expression level of Chi3l3 induced by IL-27 decreased after si-Wsx-1 treatment (H), whereas the IL-4-treated group (I) remained unaffected (n = 4). The P values (si-Wsx-1 vs. WT in NC/IL-27/IL-4 treatment) were tested by two-way ANOVA with Tukey’s multiple comparisons test, n = 4, GAPDH as internal reference. ***P < 0.001, ****P < 0.0001. Data are shown as mean ± standard error of the mean. IL: interleukin, AAM: alternatively activated macrophage, BMDM: bone marrow-derived macrophage, PCA: principal component analysis, DEGs: differential expression analysis of genes, WT: wild type.

6. IL-27-Ucp2 signaling pathway mediates AAM

Gene ontology enrichment analysis was subsequently conducted based on the authors’ bulk RNA sequencing data to further elucidate the molecular pathways engaged in IL-4-independent differentiation of anti-inflammatory AAM in response to IL-27 (Supplementary Fig. 2). This analysis indicated that the primary activated genes under IL-27 stimulation were enriched in a cellular metabolic process in macrophages. Therefore, the authors seek to unfold the metabolic rewiring in IL-27-primed macrophages in detail measured by Seahorse energy metabolic analysis (Fig. 8A). The Extracellular flux analysis showed that IL-27 did not affect the ECAR in macrophages (Fig. 8B, C); meanwhile, the oxygen consumption results indicated that IL-4, but not IL-27, enhanced the OCR (maximal respiratory capacity) in macrophages followed a 6 hr treatment (Fig. 8F). However, IL-27-insulted macrophages showed a significant proton leak compared to those which were IL-4-treated, which indicated that uncoupling of the mitochondrial respiratory chain was involved in IL-27-induced metabolic reprogramming in macrophages (Fig. 8D–F). The production of Ucp families were then examined, including Ucp1–5, in IL-27 primed macrophages [27,28]. The mRNA analysis suggests that Ucp1 presents minimal expression, and Ucp2 is significantly increased compared to the expression of Ucp3, Ucp4, and Ucp5 in IL-27-stimulated macrophages (Fig. 8G). The elevated production of Ucp2 was also detected in transcriptional data. The upregulation of Ucp2 was blocked by the genetic knockdown of Wsx-1 (si-Wsx-1) after IL-27 stimulation for 3 hours (Fig. 8H). So, how does the role of Ucp2 mediate IL-27-triggered AAM? The pharmacological inhibition of Ucp2 by genipin (GNP, the specific inhibitor of Ucp2) and genetic knockdown of Ucp2 by si-Ucp2 on macrophages was conducted to investigate the potential relevance of Ucp2-regulated IL-27-polarized AAM. It was found that the OCR was decreased in GNP-treated macrophages, which implies the impaired mitochondrial function caused by dysfunction of Ucp2 (Fig. 8I, J). Notably, GNP-treated and si-Ucp2-manipulated macrophages exhibit impaired AAM polarization in response to IL-27, whereas IL-4-induced AAM differentiation is largely unaffected (Fig. 8K). These findings indicate that IL-27-activated AAM exhibits a metabolic reprogramming distinct from IL-4-treated macrophages, featuring Ucp2-mediated uncoupling of the mitochondrial respiratory chain.

Fig. 8.

IL-27-Ucp2 signaling pathway mediates AAM. (A) The schematic illustration of this part. GNP: Genipin explicitly inhibits the protein function of Ucp2. (B) ECAR of glycolysis stress test of BMDMs, either NC or treated with IL-27 (100 ng/mL for 6 hours). (C) The Glycolytic capacity was compared with the NC and IL-27-stimulated group (n = 3), and an unpaired, two-tailed t-test was used. (D) OCR of Mito Stress Test of BMDM either NC or treated with IL-27 (100 ng/mL for 6 hours), n = 3. (E) Proton leak measured from Mito Stress Test of BMDM either NC or treated with IL-27 (100 ng/mL for 6 hours) and IL-4 (20 ng/mL for 6 hours), n = 3, one-way ANOVA followed by Dunnett post-test. (F) OCR of BMDM in IL-27 (100 ng/mL for 6 hours) and IL-4 (20 ng/mL for 6 hours)-treated group. (G) The mRNA analysis of Ucp proteins. GAPDH was used as a reference, n = 4, and a unpaired, two-tailed t -test was used. (H) The mRNA level of Ucp2 was dampened in the si-Wsx-1 group when treated with IL-27, GAPDH was used as a reference, and n = 4, one-way ANOVA followed by Tukey’s multiple comparisons test. (I) OCR of BMDMs either NC or treated with IL-27 (100 ng/mL for 6 hours) and GNP (100 mM for 12 hours). (J) Proton leak measured from Mito Stress Test of BMDM either NC or treated with IL-27 (100 ng/mL for 6 hours) and GNP (100 mM for 12 hours), n = 3, one-way ANOVA followed by Tukey’s multiple comparisons test. (K) Compared with the WT group, si-Ucp2 hampered the IL-27-induced Arg-1 expression at the mRNA level. N = 4, The P values (si-Wsx-1 vs. WT in NC/IL-27/IL-4 treatment) were tested by two-way ANOVA with Tukey’s multiple comparisons test, n = 4, GAPDH as internal reference. *P < 0.05, **P < 0.01, ****P < 0.0001, ns: not significant. Data are shown as mean ± standard error of the mean. IL: interleukin, Ucp2: uncoupling protein 2, AAM: alternatively activated macrophage, ECAR: extracellular acidification rate, BMDM: bone marrow-derived macrophage, OCR: oxygen consumption rate, WT: wild type.

7. IL-27-mediated mitochondrial metabolic shift subsequently activates the transcription factor FoxO3

As these data suggested that Ucp2-mediated uncoupling of the respiratory chain was critical for mitochondrial functionality, macrophage plasticity, and AAM differentiation in response to IL-27, the authors subsequently sought to narrow down underlying mechanisms further and performed MS-based proteomic fingerprinting of IL-27-stimulated WT, GNP-treated, and si-Ucp2 macrophages (Fig. 9A). Based on the KEGG pathways from bulk RNA-seq and proteomics results (Supplementary Fig. 3), the authors primarily focused on the molecules that regulate metabolic procession in mitochondria, such as metabolic enzymes. This method identified IL-27-Ucp2-dependent shifts in metabolism-related products, including TCA (tricarboxylic acid cycle)-relevant metabolites. Among those changed molecules, a high aconitate decarboxylase 1 (Acod1) production was found in the IL-27-treated group; meanwhile, the Acod1 was inhibited in the GNP- and si-Ucp2-stimulated group (Fig. 9B). Acod1, encoded by the immune-responsive gene 1 (Irg1), is known to catalyze the cis-aconitate decarboxylase to produce itaconate [29]. Itaconate is reported to inhibit ROS production in mitochondria by limiting the activity of succinate dehydrogenase (SDH). It can transfer into the cytoplasm, exerting multiple anti-inflammatory actions that depend on activating transcription factors such as Nrf2 [29,30]. Next, intracellular ROS levels were assessed in BMDMs under IL-27 treatment and si-Ucp2 conditions. The results showed that IL-27 treatment significantly reduced the total ROS measured. Notably, Ucp2 knockdown markedly increased ROS in the presence of IL-27, indicating that Ucp2 is required to restrain oxidative stress (Supplementary Fig. 4). In this investigation, IL-27, but not IL-4, induced the expression of Irg1 in the early phase of 3 hours, which implied that IL-27 gives rise to the elevation of itaconate in macrophages (Fig. 9C). Data confirmed that Irg1 serves as an essential hub in IL-27-Ucp2-induced AAM polarization since AAM-concerned markers were inhibited from producing in si-Irg1 macrophages in response to IL-27; however, these effects were not replicated in IL-4-induced si-Irg1 in macrophages (Fig. 9D).

Fig. 9.

IL-27-Ucp2-mediated AAM and a subsequent activation of the transcription factor FoxO3. (A) A diagram of the experimental process of this section. (B) Volcano plot analysis of differentially expressed proteins (DEPs) comparing IL-27 vs. NC, IL-27 vs. si-Ucp2, and IL-27 vs. GNP-treated groups. The marked Acod1 was indicated. (C) IL-27 induced a higher expression of Irg1 in BMDM, n = 3, GAPDH was used as a reference, and an unpaired two-tailed t-test was applied. (D) Compared to the WT group, IL-27-induced Arg-1 expression was inhibited in the si-Irg1 treated group. GAPDH was used as a reference, n = 3; the P values (si-Irg1 vs. WT in NC/IL-27/IL-4 treatment) were tested by two-way ANOVA with Tukey’s multiple comparisons test. (E) Transcription factor (TF) prediction for both DEGs and DEPs induced by IL-27-stimulated BMDM. (F) The binding motif of FoxO3 predicted by the JASPAR platform. (G) The mRNA level of FoxO3 in BMDM with either NC, IL-27 (100 ng/mL), and IL-4 (20 ng/mL) treatment. N = 4, GAPDH was used as a reference, and an unpaired two-tailed t-test was applied. (H) Compared to the WT group, IL-27-induced FoxO3 expression was inhibited in si-Irg1/si-Ucp2/GNP-treated group. GAPDH was used as a reference, n = 4; two-way ANOVA tested the P values with Tukey’s multiple comparisons test. **P < 0.01, ***P < 0.001. Data are shown as mean ± standard error of the mean. IL: interleukin, Ucp2: uncoupling protein 2, AAM: alternatively activated macrophage, GNP: Genipin explicitly inhibits the protein function of Ucp2, BMDM: bone marrow-derived macrophage, WT: wild type, DEGs: differential expression analysis of genes.

To dive into the molecule profiling in IL-27-Ucp2-induced metabolic rewiring in macrophages, the potential transcriptional mediators were screened from the RNA-seq and proteome data in TRANSFAC and JASPAR (2024) platforms (Fig. 9E, F) [25,31]. This analysis suggested that IL-27-induced gene expression would specifically involve several transcription factors. This study will emphasize the function of FoxO3, a member of the FOX transcription factor family, in IL-27-mediated differentiation of AAM. Previous studies show that FoxO3 is a key transcription factor coordinating diverse metabolic processes, such as regulation of glucose, oxidative phosphorylation, and lipid metabolism. FoxO3 also acts as a critical regulator of oxidative stress, which protects cells from oxidative stress-induced damage by reducing ROS production [32–35]. These results highlight a potential association between IL-27-Ucp2-Irg1-induced mitochondrial reprogramming and the regulation of FoxO3. Next, it was confirmed that IL-27 selectively upregulated FoxO3 expression in BMDMs, whereas IL-4 did not enhance FoxO3 (Fig. 9G). Moreover, IL-27-induced expression of FoxO3 was dependent on both Ucp2-mediated mitochondrial uncoupling and Irg1-mediated process, as the addition of GNP, as well as selective inhibition of Ucp2 or Irg1, dampened FoxO3 expression in macrophages (Fig. 9H). These data suggest a central role of FoxO3 as a downstream signaling hub in macrophages that responds to an IL-27-induced mitochondrial reprogramming.

8. FoxO3 controls IL-27-induced AAM differentiation and mitigates inflammatory pain

The authors then aimed to identify the role of FoxO3 in IL-27-induced AAM polarization. They found that the FoxO3 protein was highly expressed in the nuclei of IL-27-primed BMDM (Fig. 10A, B). However, the production of FoxO3 in IL-27-stimulated BMDM decreased after the knockdown of Ucp2 (si-Ucp2) or Irg1 (si-Irg1) (Fig. 9H). siRNA, targeted to FoxO3, was synthesized to knock down the FoxO3 gene in BMDM (Fig. 10C, D). The production of the IL-27-induced AAM marker was dampened in the si-FoxO3 BMDM group, while the IL-4-induced AAM marker was intact, thus independent of the FoxO3 pathway (Fig. 10E–G). Further, the adoptive cell transfer method was applied to investigate whether the transcription factor FoxO3 contributes to pain alleviation in mice (Fig. 10H). The WT BMDMs were exposed to 4 different treatments: macrophages (M), macrophage + si-FoxO3 (MSF3), IL-27-primed macrophage (IL-27M), and IL-27-primed macrophage + si-FoxO3 (IL-27MSF3). Then, the authors transferred these 4 group BMDMs to a 1% λ-carr-induced mouse pain model (by i.t. injection) to observe mechanical hyperalgesia, respectively. Data indicated that the mice in the IL-27M group presented the most significant relief of mechanical allodynia compared to other groups (Fig. 10I, J). Analysis suggests that the IL-27-Ucp2-FoxO3 axis regulates AAM differentiation, an IL-4-independent pattern that resolves inflammation and mitigates pain.

Fig. 10.

FoxO3 controls IL-27-induced AAM differentiation and mitigates inflammatory pain in mice. (A) Immunofluorescence (IF) microscopy of FoxO3 in the nucleus of BMDMs upon stimulation with either NC, IL-27 (100 ng/mL for 12 hours), and IL-4 (20 ng/mL for 24 hours). n = 3, The scale bar indicates 1cm, and (B) the P values were tested by a one-way ANOVA with Tukey’s multiple comparisons test. (C) By comparing the knockdown efficiencies of three candidate siRNAs targeting FoxO3, Si#1 was identified as the most effective and will use it in subsequent experiments. GAPDH was used as a reference, n = 3, and the P values (vs. NC) were tested by a one-way ANOVA with Dunnett’s post hoc test. (D) Compared to the FoxO3 mRNA in WT and/or si-Fxox3-treated BMDM in response to IL-27 (100 ng/mL for 12 hours) and IL-4 (20 ng/mL for 24 hours) treatment. GAPDH was used as a reference, n = 4, and the P values (si-FoxO3 vs. WT in NC/IL-27/IL-4 treatment) were tested by two-way ANOVA with Tukey’s multiple comparisons test. (E–G) Three M2 markers (Arg-1, Chi3l3, Retnla) were assessed by qPCR (E) and ELISA methods (F, G) in WT and/or si-Fxox3-treated BMDM in response to IL-27 (100 ng/mL for 12 hours) and IL-4 (20 ng/mL for 24 hours) treatment. GAPDH was used as a reference, n = 3, and the P values (si-FoxO3 vs. WT in NC/IL-27/IL-4 treatment) were tested by two-way ANOVA with Tukey’s multiple comparisons test. (H) Flowchart of the experimental process of the adoptive transfer strategy. (I, J) Course of mechanical hyperalgesia in mice with five different intervention groups. (I) The BMDM transfer time was indicated, n = 5, and P values (M vs. MSF3) were tested by two-way ANOVA with Tukey’s multiple comparisons test. (J) The results of IL-27M-IL-27MSF3 groups. N = 5, and P values (IL-27M vs. IL-27MSF3) were tested by two-way ANOVA with Tukey’s multiple comparisons test. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant. Data are shown as mean ± standard error of the mean. FoxO3: forkhead box class O3, IL: interleukin, Ucp2: uncoupling protein 2, AAM: alternatively activated macrophage, BMDM: bone marrow-derived macrophage, WT: wild type, M: macrophages, MSF3: macrophage + si-FoxO3, IL-27M: IL-27-primed macrophage, IL-27MSF3: IL-27-primed macrophage+si-FoxO3.

DISCUSSION

Macrophages are pivotal in regulating inflammatory pain, acting as pro-inflammatory mediators that exacerbate pain and pro-resolve regulators that promote resolution. Upon tissue damage, macrophages rapidly infiltrate the injury site, adopting a pro-inflammatory phenotype and contributing to pain sensitization [36,37]. Conversely, the AAM produce pro-resolving mediators and perform efferocytosis, even donating their mitochondria to neurons to facilitate hyperalgesia resolution [8]. These properties allow macrophages to suppress nociceptive sensitization, promote tissue repair, and restore immune homeostasis.

As a key immune regulator in the IL-12 cytokine family, IL-27 has been reported to exert both pro-inflammatory and anti-inflammatory effects context-dependently. In the present study, the authors identify the neutrophil/monocyte-derived IL-27 as a highly produced pro-resolve mediator by polarizing macrophages toward AAM during the early phase of inflammation. Compared to IL-4, the well-known macrophage AAM-inducer, IL-27 triggers the signaling pathways in macrophages involved in Ucp2-mediated uncoupling of the respiratory chain in mitochondria and activates FoxO3-dependent differentiation of AAM. Therefore, these findings highlight that FoxO3 serves as an immune metabolic regulator that coordinates IL-27-induced AAM in response to inflammation. This investigation demonstrates that the neutrophils/monocytes-macrophages collaborate to dampen inflammatory cascades during the early phase of inflammatory pain. In this work, IL-27 exerts little impact on the production of pro-inflammatory markers, such as IL-1β. IL-4, and IL-27 enhancing anti-inflammatory indicators (Arg-1, Chi3l3, and Retnla) in macrophages with different expression patterns, which indicates they polarized macrophage AAM through distinct signaling networks. The identified role of IL-27 induced metabolic reprogramming in mitochondria, characterized by respiratory chain uncoupling, and subsequently facilitated FoxO3-driven AAM polarization. This finding implies that cellular metabolism, particularly mitochondrial dynamics and functionality, plays a fundamental role in shaping the phenotypic polarization of immune cells, including macrophages [38].

The experiment data suggest that IL-27 induced a significant proton leak in mitochondria, which indicates the Ucp may be engaged in the IL-27-regulated macrophage AAM polarization [38]. Previous work confirmed that IL-27 directly targets adipocytes, activating the p38 MAPK-PGC-1α signaling pathway, thereby stimulating the production of Ucp1, followed by enhancing thermogenic activity in brown and beige adipose tissues, increasing energy expenditure, and protecting against diet-induced obesity [39]. However, the data found that Ucp1 is minimally expressed in macrophages (or IL-27-stimulated macrophages). Otherwise, the Ucp2, primarily expressed in immune cells, was significantly upregulated in IL-27-primed macrophages. Ucp2 has been reported to be a protective factor against oxidative damage (limiting ROS generation) and inflammation, contributing to regulating immune responses, which suggests that IL-27-Ucp2-mediated uncoupling of the respiratory chain reduces ROS production and sustains mitochondrial respiratory activity. Further, the proteomic analysis identified that Acod1, encoded by Irg1, was increased in response to IL-27 but not IL-4; this piqued the authors’ interest, as Acod1 serves as the key enzyme catalyzing the decarboxylation of cis-aconitate to produce itaconate, implying that IL-27 stimulation facilitates itaconate production in macrophages. Itaconate is a pivotal anti-inflammatory immunometabolite derived from the TCA cycle, which acts as a metabolic checkpoint integrating mitochondrial metabolism with immune regulation [40,41]. Itaconate inhibits SDH, reducing succinate oxidation and functionally synergizing with Ucp2 to lower ROS production in mitochondria. In the cytoplasm, itaconate activates Nrf2, leading to the upregulation of glutathione (GSH) synthesis and superoxide dismutase (SOD2), which exert potent anti-inflammatory effects to restore cellular redox homeostasis. Interestingly, the transcription factor FoxO3 also enhances the expression of SOD2 to neutralize ROS during cellular stress response, suggesting the potential coordination of Nrf2 and FoxO3 [42].

This research identified FoxO3 as the metabolic transcription factor that regulates IL-27-induced AAM differentiation in macrophages. FoxO3 serves as a multifaceted integrator of metabolic and stress signals [43]. It modulates pathways such as glucose metabolism, oxidative phosphorylation, and autophagy in response to environmental cues, thereby preserving cellular homeostasis. The results demonstrate that IL-27, not IL-4, significantly enhances the FoxO3 protein in the nucleus of macrophages and thus instructs AAM polarization, further highlighting the distinct signaling pathways from IL-4. After the adoptive transfer of macrophages (pre-treated with or without knockdown of FoxO3) into the inflammatory pain mouse model, the observation shows that the IL-27-stimulated macrophages (upregulation of FoxO3) significantly alleviate mechanical hyperalgesia in mice.

IL-27 is produced predominantly by APCs and, in neuroinflammatory settings, also by microglia; it shapes adaptive and innate responses by promoting immunoregulation and constraining Th17-mediated inflammation [44]. Mechanistically, FoxO3 in macrophages has been linked to augmented IL-10 expression and resolution programs, providing a plausible conduit by which IL-27 cues integrate with macrophage-intrinsic metabolic reprogramming to influence T-cell polarization indirectly [45]. Beyond T cells, IL-27 exerts direct effects on neutrophils—attenuating pro-inflammatory effector functions and remodeling their maturation toward an iron-scavenging, less cytotoxic phenotype—which could synergize with Ucp2-driven control of macrophage ROS to dampen tissue injury [46]. In the nervous system, IL-27 produced by microglia and other glia has been implicated in antinociceptive pathways (in part via IL-10), suggesting compartment-specific circuits in which glial-derived IL-27 modulates macrophage–nociceptor crosstalk at peripheral lesions, the DRG, and the spinal cord [47]. Finally, emerging data indicate that Ucp2 levels reflect macrophage polarization and metabolic state, positioning Ucp2 as a metabolic “rheostat” through which IL-27/FoxO3 signals might be transduced to shape interactions with dendritic cells and B/T cell niches across tissues [48]. Collectively, these lines of evidence support a model in which IL-27-Ucp2-FoxO3 not only reprograms macrophage metabolism and alternative activation, but also orchestrates crosstalk with T cells, neutrophils, and glia across compartments, with implications for biomarker development and combination immunometabolic therapies in inflammatory pain. From a translational perspective, targeting the IL-27-Ucp2-FoxO3 axis in macrophages offers a coherent path to disease modification in inflammatory pain. However, some limitations remain, such as preclinical models incompletely mirroring human trajectories, and pharmacologic manipulation of mitochondrial and transcriptional pathways raising off-target concerns. Accordingly, the IL-27-Ucp2-FoxO3 signaling network warrants deeper mechanistic investigation and rigorously designed clinical trials.

Our study has several limitations shaped by methodological choices. First, the authors did not use genetically engineered mice to ablate IL-27 in the animal experiments, which limits causal inference about IL-27–specific effects. Second, they did not perform single-cell RNA sequencing or spatial transcriptomics, preventing a high-resolution view of cellular heterogeneity and interactions within the DRG neuro-immune microenvironment. Third, they did not include sex as a biological variable in the mouse studies, which may constrain the generalizability of the findings. Fourth, detecting thermal hyperalgesia will better confirm IL-27-Upp-2-FoxO3 signaling for alleviating inflammatory pain, and the authors intend to incorporate thermal testing in future work.

Inflammatory pain arises from the activation of the immune system in response to tissue injury or infection. Remodeling macrophages to promote an anti-inflammatory and pro-resolving phenotype is a promising strategy to suppress neuroinflammation, modulate pain pathways, and restore tissue homeostasis. The IL-27-Ucp2-FoxO3 axis mediates the macrophage’s AAM polarization, offering a novel avenue for ameliorating pain.

SUPPLEMENTARY MATERIALS

Supplementary materials can be found via https://doi.org/10.3344/kjp.25307.